TK-NOTE/04-04
since: February 8, 2004 (September 1st, 2002)
last update: May 15, 2006
Two-dimensional Gauge Field Theory
and Mirror Symmetry
木村 哲士
School of Physics, Korea Institute for Advanced Study,
207-43, Cheongnyangni 2-dong, Dongdaemun-gu, Seoul 130-722, KOREA
Comments
Appendix A の SUSY algebra を訂正。(2005 6/17)
Appendix A の N ≤ (2, 2) supersymmetry を大幅追加。(2005 7/24)
Appendix A の N ≤ (2, 2) supersymmetry を修正。(2005 8/05)
Contents
I N = (2, 2) Gauged Linear Sigma Model 1
I.1 Lagrangian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
I.2 Renormalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
I.3 Classical vacua . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
I.3.1 Various phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
I.3.2 Classical limit and symmetry restoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
I.4 Low energy effective theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
I.5 Calabi-Yau/Landau-Ginzburg correspondence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
I.6 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
I.6.1 Projective space CPN−1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
I.6.2 O(−N) bundle on CPN−1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
I.6.3 CY hypersurface in CPN−1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
I.6.4 Degree ℓ hypersurface in CPN−1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
I.7 O(−N + ℓ) bundle on CPN−1[ℓ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
I.7.1 Field configuration and supersymmetric vacuum manifold . . . . . . . . . . . . . . . . . . . 24
I.7.2 Calabi-Yau phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
I.7.3 Orbifold phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
I.7.4 Singularity phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
I.8 Squashed GLSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
I.A General formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
I.B R-symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
ii CONTENTS
I.B.1 Charge assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
I.B.2 R-invariance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
I.B.3 R-symmetry and U(1) current in N = (2, 2) SCFT . . . . . . . . . . . . . . . . . . . . . . . 46
I.C Weighted projective space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
II N = (2, 2) Landau-Ginzburg Theory 49
II.1 Critical point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
II.2 Renormalization group flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
II.3 Central charges of CFT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
II.4 Chiral ring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
II.5 Three significant concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
II.5.1 Versal deformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
II.5.2 Modality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
II.5.3 Poincare polynomial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
II.6 Modality zero singularities: ADE classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
II.A Data for ADE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
II.B Highest charged state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
II.B.1 quintic hyperdurface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
III Two-dimensional N = (2, 2) Superconformal Theory 67
III.1 Superconformal algebras . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
III.2 NS chiral primary states and R ground states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
III.3 Spectral flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
III.4 Witten index, Poincare polynomial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
IV CY/LG Correspondence 77
IV.1 CY geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
IV.2 LG minimal model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
IV.3 Correspondence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
V Calabi-Yau Data from LG Superpotentials 83
CONTENTS iii
V.1 CY data from An type LG superpotentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
V.2 A few examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
V.2.1 CY 3-fold: (ki + 2) = (5, 5, 5, 5, 5) (quintic hypersurface in CP4) . . . . . . . . . . . . . . . . 90
V.2.2 CY 3-fold: (ki + 2) = (8, 8, 4, 4, 4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
V.2.3 K3 surface: (ki + 2) = (4, 4, 4, 4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
V.3 CY data from Dk type LG superpotentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
VI Mirror Symmetry: T-duality of GLSM 99
VI.1 T-duality theories on tori . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
VI.2 General formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
VI.2.1 Vortex and fermionic zero modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
VI.3 Example of mirror pair: complex projective space . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
VI.4 Period integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
VI.5 Mirror geometry descriptions for various theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
VI.6 Mirror dual of O(−N) bundle on CPN−1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
VI.6.1 mirror description in terms of homogeneous coordinates . . . . . . . . . . . . . . . . . . . . 118
VI.6.2 mirror description in terms of local coordinates . . . . . . . . . . . . . . . . . . . . . . . . . 120
VI.7 Mirror dual of CPN−1[N ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
VI.7.1 Mirror LG description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
VI.7.2 Mirror geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
VI.8 Mirror dual of CPN−1[ℓ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
VI.8.1 mirror description in terms of homogeneous coordinates . . . . . . . . . . . . . . . . . . . . 127
VI.8.2 mirror description in terms of local coordinates . . . . . . . . . . . . . . . . . . . . . . . . . 129
VI.9 Mirror dual of O(−N + ℓ) bundle on CPN−1[ℓ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
VI.9.1 Field configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
VI.9.2 Mirror Landau-Ginzburg descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
VI.9.3 Mirror geometry descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
VI.9.4 Return to the gauged linear sigma model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
VI.9.5 Summary and Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
iv CONTENTS
VI.A Linear dilaton CFT and Liouville theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
A Conventions 141
A.1 N = 1 supersymmetry in four dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
A.2 N = (2, 2) supersymmetry in two dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
A.2.1 Spinors in this notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
A.2.2 Weyl representation in two dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
A.2.3 Differential operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
A.2.4 Superfields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
A.2.5 Gauging of phase shifts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
A.3 N = (1, 1) supersymmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
A.3.1 A definition of N = (1, 1) supersymmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
A.3.2 (1, 1) superfields reduced from (2, 2) superfields . . . . . . . . . . . . . . . . . . . . . . . . . 161
A.3.3 Supersymmetry transformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
A.3.4 Superspace measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
A.4 N = (0, 2) supersymmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
A.4.1 Differential operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
A.4.2 (0, 2) superfields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
A.5 N = (0, 1) supersymmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
A.5.1 A definition of (0, 1) supersymmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
A.6 Propagator and Fourier Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
A.7 Left/right moving modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
B Calabi-Yau Data from Gepner Construction 185
B.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
C Field Theory Realization of SCFT 195
C.1 Landau-Ginzburg minimal model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
D Topological Sigma Model 199
D.1 Supersymmetric nonlinear sigma models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
CONTENTS v
D.2 Definition of topological twist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
D.3 Topological twist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
D.4 Topological gauged linear sigma models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
D.5 Example: Topological GLSM for quintics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
D.6 Example: Topological GLSM for resolved conifold . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
D.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
D.8 A-type branes, B-type branes, boundary CFT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Chapter I
N = (2, 2) Gauged Linear Sigma Model
2 N = (2, 2) Gauged Linear Sigma Model
Introduction
Gauged linear sigma model (GLSM) [113] は非常に有用な模型である。FI parameter の値によって様々な phase が
顔を出す。それらは FI parameter の違いに伴って現れる [113]。
GLSM は基本的に D = 4, N = 1 supersymmetric abelian gauge theory with chiral matter を D = 2, N = (2, 2)
に dimensional reduction する事で得られる。なお、D = 2 では spinor の irreducible representation がMajorana-
Weyl で与えられるため、chiral field の他に twisted chiral field が登場する。これら (twisted) chiral field は議論
の中心的役割を果たし、topological field theory を議論する時になどにはその性質が最大限利用される。T-dualized
theory に於いてもこれらは本質的な役割を果たす。
I.1 Lagrangian
まずは superfield levelでの gauge変換を定義しよう。但しここでは non-Abelianについては扱わず、Abelian (complex)
U(1) gauge symmetry のみを扱う。gauge coupling を e, chiral superfield Φi の U(1) gauge charge を Qi、gauge
parameter を Λ(x, θ, θ) (chiral superfield) として、次のように定義する:
Φ′i = e−2iQiΛ Φi , Φ′
i = e+2iQiΛ Φi , (I.1.1a)
V ′ = V + i(Λ− Λ) , Σ′ = Σ . (I.1.1b)
ここでは gauge coupling e が field strength にかかる normalization での記法を採用している。この表記では gauge
coupling が超くりこみ可能になるので都合がよい。すぐ後で記述する Lagrangian もこれに従う。component field
で real U(1) gauge 変換にするには Wess-Zumino gauge を採用すると良い。complex U(1) gauge 変換の前後で
Wess-Zumino gauge に固定する constraint を課すと、gauge parameter Λ の fermion と auxiliary field はゼロとな
り、scalar fieldは real partのみが non-zeroとなる。この下で chiral mutilplet Φi、gauge multiplet V の component
field の gauge 変換が与えられる。無限小変換を与えておこう:
δφi = −2iQiαφi , δψi = −2iQiαψi , δFi = −2iQiαFi , (I.1.2a)
δvm = 2∂mα , δσ = δλ = δD = 0 , (I.1.2b)
δ(Dmφi) = −2iQiα(Dmφi) . (I.1.2c)
α は gauge parameter である。また Dmφi は (I.1.7a) で定義される gauge covariant derivative である。
GLSM の一般的な Lagrangian をまず superfield で記述する:
S =1
2π
∫d2xLGLSM , (I.1.3a)
LGLSM =
∫d4θ
∑
a
(− 1
e2aΣaΣa
)+∑
i
Φi e2P
a Qai VaΦi
+(∫
d2θWGLSM(Φi) + (h.c.))
+( 1√
2
∫d2θ W (Σ) + (h.c.)
) (I.1.3b)
≡ Lg + LK + LW + LD,θ . (I.1.3c)
I.1 Lagrangian 3
index a は U(1) gauge field を labeling し、index i は chiral superfield を labeling している。Lg は gauge coupling
を overall に出した次の表記になるように normalize している:
Lg = − 1
e2
∫d4θΣΣ ∋ 1
2e2(F01)
2 = − 1
4e2FmnF
mn . (I.1.4)
Lg = − 14FmnF
mn の normalization に直すときには vm → evm とすれば良い。また、twisted superpotential W (Σ)
は twsited chiral superfield の関数である。起源は FI parameter r と Theta angle θ の寄与である。具体的には
W (Σ) = −Σ τ , τ = r − iθ , (I.1.5)
で定義される1。WGLSM(Φ) は chiral superfields Φ の holomorphic function で記述される superpotential である。
superfield で定義された Lagrangian LGLSM を component field で書き下そう:
Lg =∑
a
1
e2a
[12
(D2a + (F01,a)2)− σa (∂2
0 − ∂21)σa + iλ+,a (∂0 − ∂1)λ+,a + iλ−,a (∂0 + ∂1)λ−,a
], (I.1.6a)
LK =∑
i
[|Fi|2 − φi (∂2
0 − ∂21)φi + iψ−,i (∂0 + ∂1)ψ−,i + iψ+,i (∂0 − ∂1)ψ+,i
+∑
a
Qai
va,0 (−ψ−,iψ−,i − ψ+,iψ+,i) + va,1 (−ψ−,iψ−,i + ψ+,iψ+,i)
−√
2σa ψ−,iψ+,i −√
2σa ψ+,iψ−,i + 2i (va,0 φi ∂0φi + va,1 φi ∂
1φi)
− i√
2φi (λ−,aψ+,i − λ+,aψ−,i) + i√
2φi (ψ+,iλ−,a − ψ−,iλ+,a)
+ |φi|2∑
a
QaiDa −
∑
a,b
QaiQ
bi (va,m vm
b + 2σaσb)]
, (I.1.6b)
LW =∑
i
Fi∂WGLSM
∂φi+∂2WGLSM
∂φi∂φjψ−,iψ+,j + (h.c.)
, LD,θ =
∑
a
− raDa + θaF01,a
. (I.1.6c)
上の表式は完全に展開したものであるが、次の gauge covariant derivative を用いると幾分見易くなる:
Dmφi =(∂m + iQi vm
)φi , Dmψi =
(∂m + iQi vm
)ψi , etc. (I.1.7a)
gauge coupling e が Lagrangian の前に出て normalize されている事に注意する。つまり Lg = − 14FmnF
mn の
normalization の下では
Dmφi =(∂m + iQie vm
)φi , (I.1.7b)
となることに注意する。
Lagrangian を kinetic term と interaction terms に分解してLGLSM = Lkin + Lint と表そう。interaction terms
Lint は次のようになる:
Lint =∑
a
− raDa + θaF01,a +
1
2e2aD2
a
+∑
i
|Fi|2 +∑
i
Fi∂WGLSM
∂φi+∂2WGLSM
∂φi∂φjψ−,iψ+,j + (h.c.)
+∑
i
∑
a
Qai
[va,0 (−ψ−,iψ−,i − ψ+,iψ+,i) + va,1 (−ψ−,iψ−,i + ψ+,iψ+,i)
1ここで
Zdθ+dθ− θ+θ− = −1 を用いている。
4 N = (2, 2) Gauged Linear Sigma Model
−√
2σa ψ−,iψ+,i −√
2σa ψ+,iψ−,i
+ i (va,0 φi ∂0φi + va,1 φi ∂
1φi)− i (va,0 ∂0φiφi + va,1 ∂
1φiφi)
− i√
2φi (λ−,aψ+,i − λ+,aψ−,i) + i√
2φi (ψ+,iλ−,a − ψ−,iλ+,a) +Da|φi|2]
−∑
i
∑
a,b
QaiQ
bi |φi|2 (va,m vm
b + 2σaσb) . (I.1.8)
従って、auxiliary field Da, Fi の on-shell condition と、scalar field のみで記述される potential energy density
U(φ, σ) の一般形は次のように与えられる2:
1
e2aDa = ra −
∑
i
Qai |φi|2 , F i = −∂WGLSM
∂φi, (I.1.9a)
U(φ, σ) =∑
a
e2a2
(ra −
∑
i
Qai |φi|2
)2+∑
i
∣∣∣∂WGLSM
∂φi
∣∣∣2
+ 2∑
a,b
σaσb
(∑
i
QaiQ
bi |φi|2
). (I.1.9b)
SUSY vacuum が実現できているかどうかをみるためにはまず potential energy density を記述する必要がある。し
かしこれに fermion や gauge field を入れる必要はなく、scalar field のみの記述で十分 vacuum を考察できる。何故
なら、fermion や gauge field の真空期待値は、Lorentz symmetry を保つためには必ずゼロでないとならないので、
vacuum energy を調べる際には効いてこないからである。もちろん、low energy theory の potential energy density
を調べる際は、scalar field の揺らぎだけではなく、fermion や gauge field の揺らぎも必要となる。
Supersymmetric vacua の条件はもちろん
Da = 0 , U(φ, σ) = 0 , dU = 0 , (I.1.10)
で与えられる。
I.2 Renormalization
low energy theory として、
∑
i
Qai = 0 , (I.2.1)
という条件を課す事がある。これは FI parameter ra のくりこみが one-loop finiteであるという条件の帰結であり (see,
ref. [113] section 3; ref. [81] section 3)、CY condition と呼ぶ。これが実現される low energy theory は non-trivial
IR fixed point を持つと言う事が出来て、N = (2, 2) supersymmetric conformal theory が実現されていると考えられ
る。これを実現する NLSM は target space が CY manifold で記述できる。では、FI parameter が繰り込みを受け
ない条件を、どう記述するか。それは interaction term を component field で書き下せばよくわかる。cut-off theory
とみなしたときの、cut-off scale での Lagrangian LGLSM の D field の部分をみれば良い:
LGLSM ∋ 1
2e2D2 +D
(∑
i
Qi|φi,0|2 − r0)
(I.2.2)
2potential energy は U =R
d2xU である。
I.2 Renormalization 5
ただし簡単のため、U(1) gauge symmetry はひとつだけとした。また gauge coupling e は mass dimension +1 を持
つので super-renormalizable である。したがってこの coupling の繰り込みは考慮しなくて良い。r0 は cut-off scale
での bare FI parameter である。D のくりこみは考えない。φi,0 は bare fields である。これが、cut-off ΛUV よりは
下の scale µ での effective theory を与えるとき、 cut-off から
|φi,0|2 = |φi,R|2 + ∆φ , r0 = rR + ∆r (I.2.3)
だけ effective にずれる (µ ≤ k ≤ ΛUV scale の field が integrate out されたための効果)。ここで index R がついて
いるのは effective scale µ での fields や coupling である。ここで ∆φ を one-loop correction として与えよう3:
∆φ = 〈φi(x)φi(x)〉 =
∫ ΛUV
µ
d2k
(2π)2i
2π
k2= log
(ΛUV
µ
)(I.2.4)
導出は appendix A.6 を参照。(I.2.3), (I.2.4) を (I.2.2) に代入する:
LGLSM ∋ 1
2e2D2 +D
(∑
i
Qi|φi,0|2 − r0)
=1
2e2D2 +D
(∑
i
Qi(|φi,R|2 + ∆φ)− (rR + ∆r))
≡ 1
2e2D2 +D
(∑
i
Qi|φi,R|2 − rR)
(I.2.5)
∴ ∆r =∑
i
Qi∆φ =∑
i
Qi log(ΛUV
µ
). (I.2.6)
effective theory Lagrangian の形も microscopic (cut-off scale) Lagrangian と同じであるべし、という条件を用いて
いる。これより∆r が与えられた。具体的には FI parameter のくりこみは
r0 = rR +∑
i
Qi log(ΛUV
µ
)(I.2.7)
となる。cut-off theoryの bare FI parameterは ΛUV →∞で発散すべし、という条件がある。さらにくりこみ (I.2.7)
の形から、それが log 発散であるべし、ということから、さらに mass dimension を持つ dynamical scale parameter
Λ を導入することで
r0 =∑
i
Qi log(ΛUV
Λ
)
と表される。これより、constant partを除いて4、renormalized FI parameter r(µ)は、scale parameter µと dynamical
scale parameter Λ を用いて
r(µ) =∑
i
Qi log(µ
Λ
)(I.2.8)
となる。(classical theory では dimensionless parameter r が free parameter であったが、quantum correction を考
えると、もはやそれは free parameter ではなく、mass dimension を持つ Λ が parameter として登場するという、言
わば mass dimension の生成の様な現象を、dimensional transmutation と呼ぶ。他にも Gross-Neveu model [43] で
も登場する。) この dynamical scale Λ は RG invariant dynamical scale と呼ばれる。“ここでは”繰り込み点 µ を
Λ ≤ µ に設定する。つまり Λ を事実上 IR cut-off とみなすことが可能である。(後程 low energy effective theory を
3factor 2π は、action (I.1.3a) の規格化に由来する。4繰り込まれた FI parameter rR は constant part と corrected part r(µ) に分解されていると考えよう。
6 N = (2, 2) Gauged Linear Sigma Model
考えるが、このときは scale parameter µ を dynamical scale Λ より十分小さく採る。) 繰り込まれた complexified
FI parameter τ(µ) = r(µ)− iθ/2π は
τ(µ) =
∑iQi log(µ/Λ)− iθ if
∑iQi 6= 0
r − iθ if∑
iQi = 0(I.2.9)
となる [54]。また µ は繰り込み点、つまり観測する energy scale を表し、µ → Λ limit が low energy limit に、
µ→ ΛUV が high energy limit に対応する。これより、∑
iQi > 0 では
r(µ) →∞ if µ→ ΛUV
0 if µ→ Λ(I.2.10)
となる5。∑
iQi > 0 の場合は、GLSM は asymptotically free であることがすぐにわかる。FI parameter r(µ) は、
IR limit で出現する NLSM の coupling constant g(µ) と r(µ) ∝ 1/g(µ)2 の関係がある。そのため、FI parameter
の RG flow (I.2.8) から、coupling constant g(µ) の β function を即座に求めることができる:
µ∂
∂µr(µ) ∝ µ
∂
∂µ
1
g(µ)2= − 2
g(µ)2· µ ∂
∂µg(µ) ,
∴ µ∂
∂µg(µ) ≡ β(g) = −g(µ)3
4π
∑
i
Qi . (I.2.11)
別の見方をしよう。FI parameter は high energy limit で無限大になるというのは、その極限で target space の半
径が無限大になることに相当する。target space の半径が大きくなればなるほど、quantum fields の揺らぎの効果が
target space に及ぼす影響 (target space の変形に相当する) が小さくなる。つまり、free theory に見えることにな
り、asymptotically free theory であることがわかる。
high energy limit で target space の半径が無限大に広がり、low energy limit で半径が無限小に縮まる様
子は、dynamical compactification を彷彿とさせて非常に面白いと思われる。 (2004 3/02)
なお、ここでは GLSM の IR limit が NLSM であることを直接示していない。しかし [53] の chapter 15.4 を参照す
ると良い。具体的に、CPN−1 model について詳細な議論が展開されている。
∑iQi = 0 のときは FI parameter は one-loop くりこみがないことが理解できる。この
∑iQi = 0 を、CY
condition と呼ぶことは先程述べた。ちなみに、この CY condition が保たれない model では、RA-symmetry が
anomalous となる。(path integral measure が RA-symmetry 変換に対して不変ではない。)
supersymmetric vacua は FI parameter ra の値によって様々な構造を持つ。それぞれの同値でない vacua の上で
展開される low energy theory は、それぞれ別の物として登場する。しかし全くの別物ではなく、何らかの関係を持っ
ている。それぞれの low energy theory は、FI parameter の値によって異なるだけなので、ある系の “phase” が異
なるという解釈が出来る。よって FI parameter の張る空間を分類することが重要となる。この空間を phase space
と呼ぶことにしよう。
ここで phase space の議論で登場する用語のある程度の導入を行っておく6。
5TK-NOTE/02-04 [61] では、この RG flow の議論が不完全である。6詳細は Morrison-Plesser [81] section 3.1 にある。
I.3 Classical vacua 7
• Kahler cone KV
これは、phase space のうち、全ての FI parameter が正である region だと言って差し支えない。正確には、low
energy theory が Kahler manifold V 上の NLSM で記述でき、その Kahler form が FI parameter に比例した
形で記述できるような phase space の region である7。
• SUSY cone Kc
造語である。[81] で登場するが、具体的な呼び名が見つからないため、こう呼ばせてもらう。この cone の定義
は、SUSY condition (I.1.10) を満たす region を表す。これにより、一般的にKV ⊂ Kc である事がわかる。ま
た CY condition を課すことにより、この Kc は phase space 全体を張ることになる。
ここまでは一般の multi abelian gauge symmetry を持つ Lagrangian を構成してきたが、low energy theory など
を解析し、理解するためには、一般のままでは複雑で、構造がかえって理解できない場合がある。また一般的な議論
はそれほど重要ではなく、個々の具体的な模型を理解する方が本質が見えやすい。従って、これ以後、example を列
挙するまでは、大方の議論では single U(1) gauge theory のみを扱う事にする。multi gauge theory については、よ
り深い理解を得るために適宜議論する事にしよう。
I.3 Classical vacua
low energy effective theory を考察するためにはまず vacua を見つける必要がある。ここで取り上げたいのは su-
persymmetric theory であるとする。従って、SUSY condition (I.1.10) が満たされている必要がある。この際、FI
parameter ra の値によって、場の真空期待値の配位が異なる。以下では様々な配位について考察しよう。具体的な考察
は、後述の各模型で行う事にして、ここではごく簡単に一般論を述べるに留めておく。なお、ここでは CY condition
(I.2.1) が成立する場合のみを考える8。
具体的な模型を扱った方が、ここの議論は明解であろう。
I.3.1 Various phases
Calabi-Yau sigma model
FI parameter ra が全て正である場合を考えよう。いくつかの massless classical field が、Higgs mechanism によって
mass term を獲得する。具体的には chiral superfield のうちいくつかが真空期待値を持ち、これによっていくつかの
chiral superfield や vector superfield が massive になる。また U(1) gauge symmetry は完全に破れる。massive fields
を全て integrate out して得られる low energy effective theory は supersymmetric nonlinear sigma model として記
7Morrison-Plesser [81] の議論の展開は的確である。まず compact toric geometry V を GLSM として記述する。その low energy theory が
NLSM on V で記述できるとき、その region in phase space を KV と呼ぶ。次に additional fields を導入し、non-compact toric geometry
V + を vector bundle on V という形で定義する。NLSM が定義できる phase space では、Kahler cone は V で定義された Kahler cone と同
じであることがわかる。詳細は後ほど議論しよう。8CY condition (I.2.1) が成立しない場合は one-loop で発散を持つため、low energy theory は CFT にはならず、
Pi Qi > 0 の場合は漸
近的に自由な理論になる。また axial R-symmetry が anomalous になる。詳細は [113, 54] を参照。
8 N = (2, 2) Gauged Linear Sigma Model
述される。CY condition (I.2.1) の下では、CY sigma model と呼ばれるものになる9。この模型は superconformal
sigma model である (例えば Greene [38])。
なお、CY sigma model が実現されている ra の領域を Kahler cone と呼び、そこの理論を CY phase と呼ぶ。
Orbifold model
FI parameter が負になる場合でも、supersymmetric vacua が存在することがある。multi gauge theory ではわかり
づらいが、例えば single gauge theory のとき、r < 0 の領域では CY sigma model に変わって orbifold model が実現
されている。sigma model と同様にして、chiral superfield のいくつかが真空期待値を持ち、Higgs mechanism によっ
て chiral superfield や vector superfield が massive になる。また U(1) gauge symmetry は完全には破れず、discrete
symmetryが残る。massive fieldsを全て integrate outしたときに得られる effective theoryが、今度は orbifold model
となる。特に Landau-Ginzburg (LG) theory として登場する事が多い。基本的に LG theory は nontrivial infrared
fixed point を持ち、その極限は minimal model として記述できると理解されている。LG theory は sigma model と
異なり、superpotential が全ての dynamics などを決定している。minimal model とみなすと、各 superfield は chiral
primary field と見なす事ができ、central charge も容易に与えられる。LG model の一般論などは [78, 104, 72, 56, 86]
を参照。なお、この phase を orbifold phase と呼ぶ。
others
CY phase や orbifold phase は、Higgs branch に属していたが、FI parameter が r = 0 のときは Coulomb branch に
属する。chiral superfield は真空期待値を持たないが、vector superfield が期待値を持つ。この phase は CY phase と
も LG phase とも性質が異なる。なお、後述するCY/LG correspondence を議論する際には、この phase は重要とな
る。また、multi gauge theory の場合は、ある FI parameter は正で、他の FI parameter は負となる supersymmetric
vacua も存在する。この時には、一方では CY sigma model 的な見方が出来、他方では LG orbifold model 的な見方
ができる。これを hybrid model と呼ぼう。またこの phase を hybrid phase と呼ぶ。
I.3.2 Classical limit and symmetry restoration
上述の CY phase や orbifold phase は、FI parameter ra の正負によって vacuum manifold (もしくは VEVs の住む
空間)がガラリと変化する。この FI parameterは vacuum manifoldの大きさを規定する parameter、もしくは数とな
る。(CY condition (I.2.1) が成立するときは理論の parameter であり、成立しないときは (I.2.8) によって dynamical
scale parameter Λ の関数となる。) そして vacuum manifold 上の 1 点を SUSY vacuum と定め、fluctuation field
theory を考えると、vacuum manifold の 1 点を選んだために対称性が幾らか破れる。
ここで少し疑問が生ずる。Coleman による symmetry restoration [17] という、(1 + 1)-dimensional spacetime
theory では continuous symmetry breaking に対する no-go theorem が存在する。これは、(1 + 1)-dimensional
9N = 2 sigma model の target space は Kahler manifold で与えられるが [116] これが Ricci-flat であるとき、その sigma model は
one-loop finite であるということが示されている [3]。Ricci-flat な Kahler manifold は Calabi-Yau manifold であるので、これを target space
に持つ sigma model を Calabi-Yau sigma model と呼ぶ。
I.4 Low energy effective theories 9
spacetime では massless field が quantum theory の中ではwell-defined でないこと (propagater に IR divergence が
内蔵されていること) と、global symmetry breaking に伴って登場する massless Goldstone modes とが矛盾すること
から得られる定理である10。(なお、gauge symmetry breaking については、Higgs mechanism が存在すればmassless
Goldstone mode は他の field に吸収されるので、Coleman’s no-go theorem には抵触しない。)
よって、GLSM の SUSY vacua の廻りの fluctuation theory として NLSM が意味を成すのは、classical limit で
のみである。classical limit とは、vacuum manifold の大きさが無限大の極限である。fluctuation mode が vacuum
manifold 全体を走ることができないくらいの大きさという事である。これは、FI parameter が無限大 r → ∞ の極限を採る事に相当する。つまり、r →∞ 極限でのみ、CY phase では NLSM としての記述が良い。finite r では、も
はや NLSM としての記述は良いものではない。
I.4 Low energy effective theories
一般に GLSM は one-loop くりこみを受ける。FI parameter は理論の本来の parameter ではなくなり、dynamical
scale Λ が本来の parameter であることはすでに述べた。FI prameter の RG flow は (I.2.8) の様になる。従って
最初に r ≫ 0 である classical theory から始めた場合、asymptotic free ならば low energy に RG flow すると、FI
parameter がどんど小さくなり、やがて scale parameter µ が dynamical scale Λ となる時点で r(µ) = 0 となる。
さらに energy scale が下ると、FI parameter は負になる。古典的には、正の値を持つ FI parameter は NLSM の
target space の Kahler form に比例する数である。これが RG flow で負に流れると、もはや幾何学的な意味を持た
なくなるので、理論が破綻するように思われるが、幾何学的な意味は「古典的描像」に沿った概念であるので、負の
FI parameter も量子論では意味を持つ。詳細は個々の模型で議論しよう。
I.5 Calabi-Yau/Landau-Ginzburg correspondence
GLSM の中に登場する FI parameter を手で変化させると、ある値によって low energy theory が様々なものとなるこ
とを述べた。それらの間にはある対応関係が存在することが知られている。その対応関係はGLSMでは簡単に “phase
transition” として捉える事ができる。
CY/LG correspondence とは、ある CY sigma model と、ある LG orbifold model が、CFT の level で見て同じ
central charge を持ち、partition function が一致することを指す。massless field の数が一致しなければならない、と
いうことはない。
この対応が正当化されれば、LG orbifold model の ring structure Rcc, Rac と、CY の complex structure,
(complexified) Kahler class がそれぞれ対応することになる。
詳細は chapter IV を参照のこと。
10Coleman のこの文献 [17] は 1973 年に発表されている。
10 N = (2, 2) Gauged Linear Sigma Model
I.6 Example
I.6.1 Projective space CPN−1
まず、具体的な toric geometryを 2つ扱おう。ここで取り上げる模型は、後で扱う GLSM for non-comact CY manifold
を理解するのに重要であろうと思われる。特に KV や Kc の意味を理解するには必要であろう。具体的には、もっ
とも簡単な toric variety である projective space V = CPN−1 を low energy effective theory の field space に持つ
GLSM を構成しよう。GLSM は toric variety を field theory として記述するために開発されたと考えてもよいと思
われる [113]。GLSM for CPN−1 の configuration は以下の通り:
Φi =(S1, S2, · · · , SN
), Qi =
(1, 1, · · · , 1
), (I.6.1a)
L =
∫d4θ
− 1
e2ΣΣ +
N∑
i=1
SiSi e2V
+( 1√
2
∫d2θ W (Σ) + (h.c.)
), W (Σ) = −τ Σ . (I.6.1b)
auxiliary D field と potential energy density (contained only by scalar fields) は
1
e2D = r −
N∑
i=1
|si|2 , (I.6.2a)
U(s, σ) =1
2e2D2 + 2|σ|2
N∑
i=1
|si|2 , (I.6.2b)
となる。
この模型では Kahler cone KV と SUSY cone Kc は共に等しい:
KV = Kc =r > 0
. (I.6.3)
classical vacua
SUSY vacua を探そう。条件は (I.1.10) である。この条件を満たすのは KV 内部のみであるのはすぐにわかる。
D = 0 : r =
N∑
i=1
|si|2 , (I.6.4a)
U(si, σ) = 0 : 0 = 2|σ|2( N∑
i=1
|si|2). (I.6.4b)
これよりすぐに σ = 0 であることがわかる。この配位によって vacuum manifold は
M =
(s1, s2, · · · , sN ) ∈ CN∣∣∣ r =
N∑
i=1
|si|2 > 0/
U(1) = CPN−1 (I.6.5)
で与えられる。(U(1) gauge symmetry は、vacuum manifold のある一点での物理を考えた時に、field の期待値が
non-zero で与えられることで自発的に破れる。ここでは完全に破れる。)
classical field φ を、真空期待値 〈φ〉 と quantum fluctuation φ, φ に分離しよう。ここで φ, φ はそれぞれ vacuum
manifold M に並行な成分 (tangent modes)、並行でない成分 (non-tangent modes) を表す。
I.6 Example 11
vacuum manifold M が次で定義されているとする:
M =
(si, · · · , sN ) ∈ CN∣∣∣Fa(s1, · · · , sN ) = 0 for a = 1, · · · , k
.
このとき、classical fields si を vev 〈si〉, tangent modes si, non-tangent modes si に分解するとき、
vev, tangent modes は次をみたす:
Fa(〈s1〉, · · · , 〈sN 〉) = 0 ,
N∑
i=1
si ·∂Fa(sj)
∂si
∣∣∣M
= 0 .
すなわち、定義方程式 Fa(si) = 0 の一次の揺らぎに対して不変である方向が、tangent modes であ
る。もちろん 2 次以上の揺らぎに関しては non-zero である: s2∂2Fa|M 6= 0。よって N 個の fields
si は独立ではなく、N − k 個の独立自由度のみを持つ。non-tangent modes si はこれをみたさない。
現在考えている GLSM を具体的に検証しよう。vacuum manifoldM は (I.6.5) で与えられる。よって次のように
classical fields si の揺らぎを与えることになる:
si = 〈si〉+ si + si , F (si) = r −N∑
i=1
|si|2 , σ = 〈σ〉+ σ + σ , 〈σ〉 = 0 , (I.6.6a)
F (〈si〉) = 0 ,N∑
i=1
si∂F
∂si
∣∣∣M
= 0 , (I.6.6b)
∴
N∑
i=1
si〈si〉 = 0 , σ + σ ≡ σ . (I.6.6c)
最後の「σ+ σ ≡ σ」は、σ と σ を区別する条件が登場しないため、ひとつにまとめるということである。なお σ と記
述したのは、後にわかるように massive field であることを先に反映させているからである。この条件の下で potential
energy density の揺らぎを見よう:
U =e2
2
r −
N∑
i=1
∣∣〈si〉+ si + si
∣∣22
+ 2|σ|2( N∑
i=1
∣∣〈si〉+ si + si
∣∣2)
=e2
2
−
N∑
i=1
(|si|2 + |si|2 + 2Re[sisi + si〈si〉]
2
+ 2|σ|2N∑
i=1
(|〈si〉|2 + |si|2 + |si|2 + 2Re[sisi + si〈si〉]
). (I.6.7)
ここで VEV の条件 r =∑
i |〈si〉|2 > 0 を用いた。これより、si, σ は∑
i |〈si〉|2 に比例した mass m2 ∝ e2r を獲得
することがわかる。si は mass term を持たない事は定義より自明であったがここでもはっきりわかるであろう。そし
て N 個の si はすべて独立なのではなく、N − 1 個のみの独立自由度を持つ。ここで、mass dimension を持つ gauge
coupling constant e を無限大に飛ばす (IR limit)。これは massive fields を十分重くして理論から排除するという操
作に対応する。これにより effective theory は、classical theory の範囲においては r ≫ 0 ではM を target space と
する supersymmetric NLSM が実現される。
12 N = (2, 2) Gauged Linear Sigma Model
quantum vacua
さて one-loopくりこみで得られる RG flowによる effective theoryを考えよう。この模型は∑
iQi > 0である。(I.2.11)
より、これは asymptotic free theory であることがわかる。また (I.2.8) より、r(µ)≫ 0 という領域は Λ ≤ µ ≤ ΛUV
の energy scale による理論であることがわかる。だが、RG flow で low energy に流れると、つまり µ→ Λ となると
r(µ)→ 0 となる。さらに low energy になると、今度は r < 0 の領域まで流れ込む。古典的には r < 0 に解は存在し
ないのであるが、量子論的に存在することをこれから見ていこう。そのため、r = 0 付近、r < 0 の理論を考察する。
まず r(µ) = 0 付近を考えよう。このとき、potential energy density U のゼロ点は∀si = 0 , σ = arbitrary (I.6.8)
で実現されることがわかる。σ は gauge multiplet の一員であり、mass dimension を持っているので、low energy を
考えるときには σ の VEV 〈σ〉 は十分大きな値を持つとしよう。このまわりでの揺らぎを potential energy density Uで考察すると、chiral superfields Si は全て 2|〈σ〉|2 の質量を持つため、effective theory では integrate out されてし
まう。一方 gauge multiplet Σ は massless を維持するため、effective theory は Σ で与えられる理論となる。具体的
に Si を integrate out したときの effective Lagrangian は
Leff =
∫d4θ
(−Keff(Σ,Σ)
)+( 1√
2d2θ Weff(Σ) + h.c.
), (I.6.9a)
Weff = −Στ(µ)−N∑
i=1
QiΣ
log(QiΣ
µ
)− 1
= −Στ(µ)−NΣ
log(Σ
µ
)− 1, (I.6.9b)
となる。effective Kahler potential Keff(Σ,Σ) の関数形は一般にはわからない。しかし critical point、つまり SUSY
vacua を探す際には、ここでは殆ど寄与しないと見てよい。さて、supersymmetric vacua を探す。U ∋ e2eff
2 |∂σWeff |2
であるので、vacua は ∂Weff = 0 で与えられる。つまり
0 = ∂σWeff(σ) = −N log(σ/µ
)− τ(µ) (I.6.10)
であるので、critical point は
(σ/µ
)N= e−τ(µ) (I.6.11)
で与えられる。これをみたす解 σ = σcr は N 個存在する。つまり N 個の真空が存在することになる。なお、r(µ) = 0
は µ = Λ であるが、さらに low energy µ < Λ の時にも上の解は矛盾しない。これは r(µ) < 0 の領域に相当する。そ
のため quantum theoryとして r(µ) < 0まで含めて r(µ)の全領域で定義することができる。sigma model on CPN−1
の low energy limit に於ける quantum vacua は、vacuum manifoldM = CPN−1 ではなく、殆どが持ち上がってN
個の点になる。その様子を Figure I.1 で表しておこう:
ちなみに、r(µ) < 0 での真空の個数 N は、CPN−1 の cohomology ring の次元と一致する:
N = dimH∗(CPN−1) . (I.6.12)
I.6.2 O(−N) bundle on CPN−1
では、ある phase で NLSM on CY manifold を low energy theory を実現する最も簡単な GLSM を考察しよう。こ
こでは superpotential を導入しない。またほとんど全ての chiral superfield の U(1) charge が +1 で、CY condition
I.6 Example 13
µ < Λ µ = Λ µ > Λr(µ) < 0 r(µ) > 0
Figure I.1: Quantum Sigma Model on CPN−1.
(I.2.1) を満たすものとして、次のような field content と U(1) charge assignment を考える:
Φi :(S1, S2, · · · , SN , P
), Qi :
(+ 1,+1, · · · ,+1,−N
). (I.6.13)
従って bosonic field のみで記述される auxiliary field D と potential energy density U(φ, σ) は次のようになる:
1
e2D = r −
N∑
i=1
|si|2 +N |p|2 , (I.6.14a)
U(s, p, σ) =1
2e2D2 + 2|σ|2
( N∑
i=1
|si|2 +N2|p|2), (I.6.14b)
SUSY condition D = U(φ) = 0 の解を探そう11:
D = 0 : r =N∑
i=1
|si|2 −N |p|2 , (I.6.15a)
U(si, p, σ) = 0 : 0 = 2|σ|2( N∑
i=1
|si|2 +N2|p|2). (I.6.15b)
これは FI parameter r の値によって、D = 0 の解が異なることを示すので、詳細を見ていこう。
r ≫ 0 case: CY phase
特に r ≫ 0の場合を考える。この phaseの low energy effective theoryは Higgs branchとして記述されるのがわかる。
何故なら、chiral superfield (gauge theoryの言葉で言えば hypermultiplet)が low energy limitでは真空期待値を持つ
からである。この phaseでは∑
i |si|2 6= 0より、σ = 0となることがわかる。一方、pは D = 0 = r−∑i |si|2 +N |p|2
をみたす範囲であれば特に固定されない。すなわち、vacuum manifold MCY は p と si で構成され、U(1) gauge
11停留条件 dU = 0 も考慮に入れるのが正当であるが、これは後の解では全てきちんと満たされているので、省略する。
14 N = (2, 2) Gauged Linear Sigma Model
symmetry で割られる次の空間をなす12:
MCY =
(s1, s2, · · · , sN ; p) ∈ CN+1∣∣∣ r =
N∑
i=1
|si|2 −N |p|2 > 0/
U(1)
= O(−N) bundle over CPN−1
(I.6.16)
ここで si が CPN−1 の homogeneous coordinates で、p が fiber coordinate である。total space は noncompact CY
manifold である。
この解の下で massless effective theory を考察しよう。massless field がこの vacuum manifold と一致することを
見る。〈φ〉 と quantum fluctuation part φ+ φ に分離した φ = 〈φ〉+ φ+ φ を用いてゆらぎの条件を考えよう:
si = 〈si〉+ si + si , p = 〈p〉+ p+ p , σ = 〈σ〉+ σ + σ , (I.6.17a)
F (si, p, σ) = r −N∑
i=1
|si|2 +N |p|2 , (I.6.17b)
〈σ〉 = 0 , F (〈si〉, 〈p〉, 〈σ〉) = 0 , (I.6.17c)
N∑
i=1
si∂F
∂si
∣∣∣MCY
+ p∂F
∂p
∣∣∣MCY
= 0 =
N∑
i=1
si〈si〉 −Np〈p〉 , σ + σ ≡ σ . (I.6.17d)
この条件を用いて、potential energy density を記述しよう:
U =e2
2
r −
N∑
i=1
∣∣〈si〉+ si + si
∣∣2 +N∣∣〈p〉+ p+ p
∣∣22
+ 2|σ|2( N∑
i=1
∣∣〈si〉+ si + si
∣∣2 +N2∣∣〈p〉+ p+ p
∣∣2)
=e2
2
−
N∑
i=1
(|si|2 + |si|2 + 2Re[sisi + si〈si〉]
)+N
(|p|2 + |p|2 + 2Re[pp+ p〈p〉]
)2
+ 2|σ|2 N∑
i=1
(|〈si〉|2 + |si|2 + |si|2 + 2Re[sisi + si〈si〉]
)+N2
(|〈p〉|2 + |p|2 + |p|2 + 2Re[pp+ p〈p〉]
).
(I.6.18)
vector multiplet の scalar field σ は質量 m2 = 2e2∑
i |〈si〉|2 = 2e2r を獲得する。いわゆる Higgs mechanism で
ある。あらわには記述してないが、gauge field vi も同じ質量を Higgs mechanism で獲得し、その superpartner λ
(gaugino) も同じ質量を獲得する。また field si, p も同様に質量を獲得する事が理解できる。その質量も m2 = 2e2r
である。よってこれら massive field を全て integrate out する (e2 →∞ limit をとって、gauge field の kinetic term
を freeze することにも対応する。もちろんこの極限では質量は全て無限大に飛ぶ。) ことで massless effective theory
が得られる。なお、U(1) gauge symmetry は ∃〈si〉 6= 0 なので完全に破れる。
この理論は N = (2, 2) supersymmetric NLSM、特に CY manifold を target space として持つ conformal sigma
model と見なされる。この phase を CY phase と呼ぶ。
12moduli space の代わりに vacuum manifold とも呼ぶ [53]。
I.6 Example 15
r ≪ 0 case: orbifold phase
先の議論と同じようにして、特に r ≪ 0 の場合を考える。この phase も Higgs branch として記述される。やはり
low energy limit で hypermultiplet が真空期待値を持つからである。
この phase で実現される moduli は p 6= 0 である。これによりすぐさま σ = 0 が与えられる。si はすべて r < 0
をみたす範囲内で free である。そのため vacuum manifold Morbifold は
Morbifold =
(s1, s2, · · · , sN ) ∈ CN∣∣∣ r =
N∑
i=1
|si|2 −N |p|2 < 0/
U(1) (I.6.19)
となることがわかる。
effective theory を考えるときは vacuum manifold のある一点での物理を考えることになる。例えば、〈p〉 6= 0,
∀〈si〉 = 0 の点を選んだとすると、U(1) gauge symmetry は ZN まで自発的に破れる。一方 〈si〉 6= 0 となる点で理論
を展開すれば、ZN gauge symmetry は完全に破れる。
∀〈si〉 = 0 となる vacuum manifold の点、つまり原点で quantum fluctuated field を議論しよう。この場合の
effective potential U(φ, σ) は
U =e2
2
(−
N∑
i=1
|si|2 +N(〈p〉p+ 〈p〉p+ |p|2
))2
+ 2N2|〈p〉|2|σ|2 + 2|σ|2( N∑
i=1
|si|2 +N2(〈p〉p+ 〈p〉p
)),
(I.6.20)
となるので、σ, p は m2 = 2e2N2|〈p〉|2 = −2N2e2r の質量を持つ。これらの superpartner も同様の質量を獲得する。
すべて質量の獲得は Higgs mechanism で行われる。
特に r → −∞ の極限を考えよう。このときは fixed moduli 〈p〉 も無限大にとる。すると r =∑
i |〈si〉|2 −N |〈p〉|2
からの揺らぎには拘束がつかない。つまり Si はすべて flat space CN に住むことが出来るようになる。よってこの
orbifold phase での low energy effective theory は
CN/ZN orbifold theory (sigma model)
である。
r = 0 case: Coulomb branch and singularity
この場合は非常に特殊である。Theta angle θ の値 (zero/non-zero) で振舞いが全く変わってくる。ちなみにここでは
low energy theory は Coulomb branch として記述される。実はこの phase は非常に重要である。
supersymmetric condition (I.6.15) に r = 0 を代入して得られる解を考察しよう。
• Higgs branch
16 N = (2, 2) Gauged Linear Sigma Model
In the Higgs branch, 実はさらに 2 系統の解が存在する。一つは、Mr>0 やMr<0 で r → ±0 極限と滑かに繋
がる
∀〈si〉 → 0 , 〈p〉 → 0 (I.6.21a)
であり、もうひとつは
〈si〉 ≃ 〈p〉 = O(1) (I.6.21b)
である。CY phase に特化して考えよう。CY phase では、vacuum manifoldMr>0 の形状が CY geometry と
して well-defined であるが、r = 0 では Kahler 構造が潰れてしまう。[46] などのように、FI parameter は定数
に固定しておいて、他の parameter ([46] では積分定数 b) をゼロに飛ばすことで、base space がゼロに見える
極限、別の言葉では原点から無限に遠い極限での記述になり、依然 well-defined な geometry であった。しかし
FI parameter r そのものを zero にすると、base space の構造が壊れてしまう。もはや CY geometry とは言え
ないものになる。つまりもはや si, p は fiber bundle のような構造を持っていない。ただ CN+1 space で、こ
こに constraint∑
i |si|2 = N |p|2 があるだけである。(mod U(1) gauge symmetry。)
• Coulomb branch
ここで挙げる解は CY phase/LG phase には存在しないものである。端的にいえば、vector multiplet の scalar
field σ が nonzero の真空期待値を取る解である。
〈si〉 = 〈p〉 = 0 でも auxilary field 〈D〉 = 0 はみたされる。この解のとき potential energy density U = 0
condition は、〈σ〉 が任意の値で許される。つまり 〈σ〉 6= 0 が許されるのである。この configration では U(1)
gauge symmetry は破れない。これは非常に強い意味を持つ。CY phase での FI parameter r の値を徐々に手
で下げて行く。FI parameter r は CY manifold の Kahler class の moduli であり、ここでは CPN−1 の size
を表す。これを小さくするとは、CPN−1 を小さくすることに対応する。そして、r = 0 付近でこの CPN−1 が
1 点に潰れるのであるが、それに伴い、Coulomb branch が生成され、その branch では、CY phase で破れて
いた gauge symmetry が enhance される。
Coulomb branch についてもう少し詳しく見てみよう。Coulmob branch では 〈si〉 = 〈p〉 = 0 で、〈σ〉 が任意の値を取ることが可能であったので、〈σ〉 ≥ 1 としよう。この真空からの揺らぎを考えると、chiral multiplets Si, P は
|〈σ〉2| に比例した質量を持つ。そのため、massive fields を integrate out した low energy effective theory では、こ
れら chiral multiplets は全て登場しない。一方 vector multiplet Σ は質量を持たないため、effective theory でも登場
する。その際、scalar σ, gaugino λ は完全に free field として登場し、quantum effect に寄与しない。しかし gauge
field (特に F01 だけなので、electric field) は事情が多少異なる。gauge field の effective Lagrangian を記述すると
Leff =1
2e2(F01)
2 + θF01 , (I.6.22)
である。但し θ = θ mod 2πZ である。two dimensional gauge theory ではこの Theta term は非常に重要な役割を
果たす。この Theta angle θ のために、constant electric field F01 が vacuum energy に寄与する。(I.6.22) でその
constant electric field の値を求めると
F01 = −θe2 , (I.6.23)
I.6 Example 17
となる。よって vacuum energy density は
Evacuum = −L∣∣constant field
=e2
2θ2 , (I.6.24)
となる。
この Coulomb branch を r 6= 0 の場合まで持ち上げると、vacuum energy density として新たに e2
2 r が加わるの
で13、結果として
Evacuum =e2
2
r2 + θ2
, (I.6.25)
が得られることになる。ここで r → 0, θ → 0 (mod 2πZ) 極限を取ると、この Coulmob branch は Higgs branch
(〈σ〉 = 0) と縮退して、理論として異なる真空が同時に存在することになる。
また FI parameter が r = 0 前後での phase transition が smooth に行われる、という議論も存在する [113]。そ
れは、Theta angle θ が nonzero の場合に対応する。きちんと理解できていないので満足行く形でここに記載するこ
とは出来無いが、簡単に言うことはできる。string theory ではこの Theta angle の項は NS-NS 2-form field B の
worldsheet 上での積分に対応する。従って θ 6= 0 である場合は、例え r = 0 で CPN−1 が潰れても、そこに NS-NS
2-form fluxが存在するため、理論に特異点がなく smoothに繋がることができる。もし θ = 0であれば本当に singular
である。これについては Aspinwall [6] などが議論している。
なお、Higgs branch の singularity (I.6.21a) から逃げてきた effective theory は、∀ϕa = σ = 0 を介して Coulomb
branch に入り、σ 6= 0 の領域に解析的に繋がって、r = θ = 0 以外の FI parameter での effective theory に落ち着く
と期待できる。r ≪ 0 への接続では、また Coulomb branch と Higgs branch の接点 ϕa = 0 を通じて繋がり合う
だろう。また別の回避も考えられる。Coulomb branch を介さずに CY phase から r = 0 の singular phase の Higgs
branch (I.6.21b) を通り、そのまま orbifold phase に抜ける物も考えらえる:
phase CY r > 0 singular r = 0 orbifold r < 0
connection 1 (order r) Higgs ↔ Coulomb ↔ Higgs
connection 2 (arbitrary order) Higgs ↔ Higgs ↔ Higgs
これら 2 つの経路の存在の理由はは、この模型では不明のままである。
CY phase: r > 0
Coulomb branch at r = 0
Higgs branch at r = 0
orbifold phase: r < 0
〈σ〉 is free
〈φ〉 = O(r) 〈φ〉 = O(|r|)
〈φ〉 is free
singularity
r → +0
r → −0
13effective potential U に si = p = 0 を代入することで得られる。
18 N = (2, 2) Gauged Linear Sigma Model
I.6.3 CY hypersurface in CPN−1
O(−N) bundle on CPN−1 を与える configuration (I.6.13) に、次の superpotential WGLSM を付加する:
WGLSM = P ·GN (Si) . (I.6.26)
ここで GN (Si) は一般に homogeneous polynomial of degree N であり、次の性質を持つとする:
GN (Si) =∂GN
∂S1= · · · =
∂GN
∂SN= 0 implies S1 = S2 = · · · = SN = 0 . (I.6.27)
例えば、次にあるように Fermat type を用意すれば良い:
GN (Si) =
N∑
i=1
SNi . (I.6.28)
この setup の時には、potential energy density は次のようになる:
U(s, p, σ) =1
2e2D2 + 2|σ|2
( N∑
i=1
|si|2 +N2|p|2)
+ |GN (si)|2 + |p|2N∑
i=1
|∂iGN |2 , (I.6.29a)
1
e2D = r −
N∑
i=1
|si|2 +N |p|2 . (I.6.29b)
CY phase
r ≫ 0 phase において、supersymmetric vacuum U(φ) = 0 (I.6.29) をみたす真空期待値の配位は
N∑
i=1
|si|2 6= 0 , σ = 0 , GN = 0 , 〈p〉 · ∂iGN (sk) = 0 (I.6.30)
である。また、もし p 6= 0 とするならば、U(φ) = 0 の条件 (I.6.29) から読み取れるように GN = ∂1GN = ∂2GN =
· · · = ∂NGN = 0 となるが、これは (I.6.27) から ∀si = 0 であることを意味し、(I.6.30) にあるような∑
i |si|2 6= 0 に
反する。よって
p = 0 (I.6.31)
が得られる。これにより、vacuum manifold MCY は
MCY =
(s1, s2, · · · , sN ) ∈ CN∣∣∣ r =
N∑
i=1
|si|2 > 0, GN (si) = 0/
U(1)
= hypersurface CPN−1[N ] = CY (N − 2)-fold
(I.6.32)
であることがわかる。
ここで適当な moduli space の一点を選び、effective theory (r → ∞, e2 → ∞) を考えよう。ここで si は元々
CPN−1 の homogeneous coordinates であったので、これで構成される vacuum manifold 上の点はどこも同じ物理
を出す。前に議論した揺らぎの評価と同様の事をすると、揺らぎは
si = 〈si〉+ si + si , p = p . (I.6.33)
I.6 Example 19
で与えられる。記号の意味は前述の CPN−1 の場合と同じである14。この配位では massive field は si, σ とそれらの
superpartner である。さらに effective potential の最終項から p の質量項が登場し、p も si, σ と共に integrate out
される。よって、真空期待値を持たない field p が massive となって integrate out される。よって、massless field だ
けで記述される low energy theoryは、上で述べたように、vacuum manifoldと同じCPN−1 の hypersurface GN = 0
という compact manifold (やはり CY manifold。N = 5 の quintic hypersurface in CP4 が CY であることを用い
た議論は多い。) を target space とする NLSM となる。
orbifold phase
r ≪ 0の phaseは SUSY coneに入っても Kahler coneには入っていない。従って low energy effective theoryは sigma
model では記述されない。これは superpotential がない場合の議論と同様である。SUSY condition より、moduli は
p 6= 0 → σ = 0 , (I.6.34)
の配位を採る。これと (I.6.27), (I.6.29) より、
∀si = 0 (I.6.35)
が結論付けられる。よって vacuum manifold Morbifold はただの 1 点になる:
Morbifold =
(s1, s2, · · · , sN ) = (0, 0, · · · , 0)/
U(1) (I.6.36)
effective theory は、vacuum manifold の一点で展開される。〈p〉 6= 0 より U(1) gauge symmetry は ZN まで自
発的に破れる。そして p, σ とその superpartner は Higgs mechanism を通じてmassive field となる。
なお、homogeneous function GN が degree 2 以下ならば、potential energy の式の最終項から、全ての si も質
量項を持つ。したがって homogeneous function GN が degree N ≤ 2 であれば、全ての field が massive となり、
massless fieldで構成される low energy theoryが実現しない事になる15。なお、N ≥ 3の条件であれば、この最終項か
ら si の質量項が登場する事はない。この条件下では、low energy theory は si が flat kinetic term と superpotential
W ′ =√−rGN (si) を持ち (|〈p〉|2 = −r)、ZN gauge symmetry を持つ LG orbifold theory が実現している。そして
この superpotential は
W ′ = 〈p〉GN (Si) =√−r/N GN (Si) (I.6.37)
で与えられる。因子√−r/N は重要ではないので、適当に superfield Si に吸収させることができる。
この LG model の IR limit は CFT minimal model として解釈できる。この模型の central charge を求めてみよ
う。central charge は orbifolding されてない Lagrangian でも正しく与えられる (chiral ring は orbifolding しないと
正しく与えられない)。superpotential (I.6.37) を見よう。chiral superfield Si が全て chiral primary field であるとす
る。また LG model が scale invariant であることを要請して、Si の scale 変換則を考えよう。worldsheet coordinate
14最も複雑な O(−N + ℓ) bundle on CPN−1[ℓ] の議論では、きちんと定量的に評価するが、ここは面倒なので省略。
15もしくは massive field そのものを low energy theory で扱うことになる。
20 N = (2, 2) Gauged Linear Sigma Model
z が z → z′ = λ−1z (λ ∈ C∗) と変換される時、superspace coordinates θα や measure d2zd2θ は次のように変換さ
れる:
z′ = λ−1z , z′ = λ−1z , θα → λ−1/2θα (∵ ym = xm + iθσmθ) ,
dz → λ−1dz , dθα → λ1/2dθα ,
∴ d2z d2θ → λ−1d2z d2θ .
この変換の下で、superfield Si が
S′i(z
′, z′) =(∂z′z
)hi(∂z′z
)hiSi(z, z) = λ2hiSi(z, z) , (I.6.38)
と変換される。ここで hi, hi は それぞれ left/right の conformal weight である16。F-term が scale invariant であ
ることを要請する事で、hi の値が決定される。
superpotential W ′(Si) の変換則は
W ′(Si) → W ′(λ2hiSi) ≡ λW ′(Si) (I.6.39)
である。よって、
1 = 2hi ·N , → hi =1
2N, (I.6.40)
となった。つまり、superpotential W が left/right U(1) charge (1, 1) を持ち、それによって chiral superfield Si が
(qi, qi) = (1/N, 1/N) を持つ事になる。
Si が chiral superfield であり、homogeneous superpotential を持つ minimal model の場合、conformal weight hi
と U(1) charge は hi = 12qi で与えられる。またこの model にある highest charged state は hρ = c/6 であるので、
total central charge ctotal は次で与えられる (appendix III; ref. [105]):
ctotal = 6β , β =∑
i
(1
2− qi
). (I.6.41)
よって
ctotal = 3(N − 2) , (I.6.42)
である。N = 5 の時は ctotal = 9 であるが、これは r > 0 のCP4[5] が CY 3-fold であり、その上の conformal sigma
model の central charge が c = 9 であることと一致する。
CY/LG correspondence により、LG model の ring structure から CY の cohomology class が読みとれるので
あった。従って chiral ring を理解するのは重要である。chiral ring は、chiral primary field の OPE がなす環構造
Φ(1)(z)Φ(2)(w) ∼ C(123)(z − w)Φ(3)(w) である。C(123) はこの ring の structure constant である。この chiral ring
の元は
Rcc =C(Si)
[dW ′(Si)], (I.6.43)
で与える事が出来る。ここで C(Si) は、Si の多項式全体がなす関数体である。chiral superfield で構成されるので、
(c, c)-ring である。
なお、appendix V において、LG model から CY geometry の詳細を読み取る議論とその例を紹介しておく。16Si は chiral superfield であるが、left/right には分離していない。従って left/right それぞれの conformal 変換が施される。また superfield
の lowest component field は scalar であるので、spin s = hi − hi = 0 である。
I.6 Example 21
I.6.4 Degree ℓ hypersurface in CPN−1
次の configuration を考えてみよう:
S1 · · · SN P
U(1) 1 · · · 1 −ℓ
W = P ·Gℓ(Si) . (I.6.44)
ここで Gℓ(Si) は一般に homogeneous function of degree ℓ であり、次の性質を持つとする (3 ≤ ℓ ≤ N − 1 とする):
Gℓ(Si) = ∂iGℓ = 0 implies S1 = S2 = · · · = SN = 0 . (I.6.45)
この setup の時には、potential energy density は次のようになる:
U(s, p, σ) =e2
2
r −
N∑
i=1
|si|2 + ℓ|p|22
+ 2|σ|2( N∑
i=1
|si|2 + ℓ2|p|2)
+ |Gℓ(si)|2 + |p|2N∑
i=1
|∂iGℓ|2 . (I.6.46)
r ≫ 0 phase
r ≫ 0 phase において、supersymmetric vacuum U(φ) = 0 (I.6.46) をみたす真空期待値の配位は
N∑
i=1
|si|2 6= 0 , σ = 0 , Gℓ = 0 , 〈p〉 · ∂iGℓ(sj) = 0 (I.6.47)
である。また、もし p 6= 0 とするならば、U(φ) = 0 の条件 (I.6.46) から読み取れるように Gℓ = ∂1Gℓ = ∂2Gℓ =
· · · = ∂NGℓ = 0 となるが、これは (I.6.45) から ∀si = 0 であることを意味し、(I.6.47) にあるような∑
i |si|2 6= 0 に
反する。よって
p = 0 (I.6.48)
が得られる。これにより、vacuum manifold MCY は
MCY =
(s1, s2, · · · , sN ) ∈ CN∣∣∣ r =
N∑
i=1
|si|2 > 0, Gℓ(si) = 0/
U(1)
= hypersurface CPN−1[ℓ]
(I.6.49)
であることがわかる。
ここで適当な moduli space の一点を選び、effective theory (r → ∞, e2 → ∞) を考えよう。ここで si は元々
CPN−1 の homogeneous coordinates であったので、これで構成される vacuum manifold 上の点はどこも同じ物理
を出す。揺らぎを
si = 〈si〉+ si + si , p = p . (I.6.50)
で与える。記号の意味は前述の CPN−1 の場合と同じである。この配位では massive field は si, σ とそれらの
superpartner である。さらに effective potential の最終項から p の質量項が登場し、p も si, σ と共に integrate out
される。よって、真空期待値を持たない field p が massive となって integrate out される。よって、massless field だ
けで記述される effective theory は、上で述べたように、vacuum manifold と同じ CPN−1 の hypersurface Gℓ = 0
という compact manifold を target space とする NLSM となる。
22 N = (2, 2) Gauged Linear Sigma Model
r ≪ 0 phase
r ≪ 0の phaseは SUSY coneに入っても Kahler coneには入っていない。従って low energy effective theoryは sigma
model では記述されない。これは superpotential がない場合の議論と同様である。SUSY condition より、moduli は
p 6= 0 → σ = 0 , (I.6.51)
の配位を採る。これと (I.6.45), (I.6.46) より、
∀si = 0 (I.6.52)
が結論付けられる。よって vacuum manifold Morbifold はただの 1 点になる:
Morbifold =
(s1, s2, · · · , sN ) = (0, 0, · · · , 0)/
U(1) (I.6.53)
effective theory は、vacuum manifold の一点で展開される。〈p〉 6= 0 より U(1) gauge symmetry は Zℓ まで自
発的に破れる。そして p, σ とその superpartner は Higgs mechanism を通じてmassive field となる。この条件下で
は、effective theory は si が flat kinetic term と superpotential W ′ =√−rGℓ(si) を持ち (|〈p〉|2 = −r)、Zℓ gauge
symmetry を持つ LG orbifold theory が実現している。そしてこの superpotential は
W ′ = 〈p〉Gℓ(Si) =√−r/ℓGℓ(Si) (I.6.54)
で与えられる。因子√−r/ℓ は重要ではないので、適当に superfield Si に吸収させることができる。
さてここで挙げた effective theory は classical なものである。実際には
∑
a
Qa = N − ℓ > 0
であるために asymptotic free である。そのため FI parameter r はくりこみを受け、(I.2.8) のようになっている。そ
のため Λ≪ µ≪ ΛUV で r(µ) > 0 の値を持つ sigma model on CPN−1[ℓ] で始まる理論は RG flow で低エネルギー
に流れるに連れて r(µ)→ 0 となる。そして µ ∼ Λ で r(Λ) = 0 となるため sigma model description が破綻する。し
かし理論はここで破綻するのではない。sigma model on CPN−1 の low energy theory 同様、r(µ) = 0 では別の理論
に繋がる。それを議論する。
r = 0 での potential energy density U (I.6.46) から言えることは、σ が flat direction となることである。σ は
mass dimension を持つことと、low energy theory を考える際に mass dimension を持つものは無限大に飛ばすことか
ら、ここでも 〈σ〉 が十分大きい場所を考えると、そのまわりでは Si, P は十分大きな質量項を持つことから、massless
effective theory を考える時には integrate out される。massless limit では Σ だけであり、その effective Lagrangian
は次のようになる:
Leff =
∫d4θ
−Keff(Σ,Σ)
+( 1√
2
∫d2θ Weff(Σ) + h.c.
), (I.6.55a)
Weff(Σ) = −Στ(µ)−∑
a
QaΣ
log(QaΣ
µ
)− 1. (I.6.55b)
sigma model on CPN−1 の r = 0 における massless effective theory の議論でも上の effective Lagrangian は登場し
た。CPN−1 と CPN−1[ℓ] の違いは GLSM classical Lagrangian に chiral superpotential WGLSM = P1Gℓ(Si) が存
I.6 Example 23
continuous spectrum
discrete spectrum
0 σ
U(σ)
Ueff =e2eff2
∣∣τ(µ)− ℓ log(−ℓ)∣∣2
Figure I.2: Effective potential around r(µ) = 0.
在するかしないかである。この存在の有無は effective Kahler potential Keff(Σ,Σ) に影響を及ぼすが、twisted chiral
superpotential である Weff(Σ) には影響を及ぼさない。(c, c)-ring の構造は (a, c)-ring の構造に影響を与えないので
ある。
さて今の場合は
Weff(Σ) = −Σ(τ(µ)− ℓ log(−ℓ)
)(I.6.56)
である。これから得られる potential energy density は
Ueff(σ) =e2eff2
∣∣∂σWeff(σ)∣∣2 =
e2eff2
∣∣∣τ(µ)− ℓ log(−ℓ)∣∣∣2
(I.6.57)
となる。effective Kahler potential Keff からの寄与は全て eeff に押し込める (らしい)17。この値は Σ が非常に大き
な時には正しいが、小さな所では正しくないため、漸近的な形式に過ぎない。実際 supersymmetric であることを保
つためには potential energy density は何処でゼロ点を持っているはずである。そのため、potential energy density
の形は Figure I.2 のようになっていると予想される:
ここで
log(−ℓ) = πi+ log |ℓ|+ 2πin n ∈ Z (I.6.58)
であり、これから θ と Ueff の関係が Figure I.3 になることがわかる:
この potential 群の最低点 (赤線)を採り続ける θ を
θ ≡ θ − 2πℓn (I.6.59)
と定義する。先の CY sigma model on CPN−1[N ] で Coulomb phase を議論したときに登場させた θ はこれである。
なお、τ(µ) = ℓ log(−ℓ) という値を持つ時にはどんな scale parameter µ の範囲でも理論は連続スペクトルを持つ
ことになる。これは worldsheet が cylinder S1 ×R1 であることからその上の field theory は discrete spectrum を持
つべきだという描像と矛盾するため、t(µ) = ℓ log(−ℓ) では理論は singular となる。古典的には τ = 0 が singular で
17[53] chapter 15 参照。
24 N = (2, 2) Gauged Linear Sigma Model
U
0 θ
Figure I.3: Relation between θ and Ueff .
あるように見えるが、実は量子論的な効果を入れるとこのようにずれる。しかし量子論的にも τ(µ) 6= ℓ log(−ℓ) (mod
2πZ) であれば dicrete spectrum を持つことが可能であり、理論は健全なものとなる。RG flow ではその健全な部分
を流れる。その様子を Figure I.4 に表しておこう:
θ
singular point: τ(µ) = ℓ log(−ℓ)
sigma model on CPN−1[ℓ]
LG theory with WLG
µ≫ Λ
µ≪ Λ
Figure I.4: Parameter space τ(µ).
I.7 O(−N + ℓ) bundle on CPN−1[ℓ]
I.7.1 Field configuration and supersymmetric vacuum manifold
Now we are ready to analyze massless low energy effective theories in the GLSM for O(−N+ℓ) bundle on CPN−1[ℓ].
We consider a U(1) gauge theory with N + 2 chiral superfields Φa of charges Qa. We set the field configuration to
chiral superfield Φa S1 . . . SN P1 P2
U(1) charge Qa 1 . . . 1 −ℓ −N + ℓ(I.7.1)
In addition we introduce a superpotential WGLSM(Φ) = P1 ·Gℓ(S), where Gℓ(S) is a function of chiral superfields
Si. This is a holomorphic homogeneous polynomial of degree ℓ. Owing to the homogeneity, this polynomial has a
following property:
if Gℓ(s) = ∂1Gℓ(s) = . . . = ∂NGℓ(s) = 0 → then ∀si = 0 . (I.7.2)
I.7 O(−N + ℓ) bundle on CPN−1[ℓ] 25
By definition, the numbers N and ℓ are positive integers: ℓ,N ∈ Z>0. We assume that these two integers satisfy
1 ≤ ℓ ≤ N − 1 and 2 ≤ N . The sum of all charges Qa vanishes (I.2.1) in order to obtain non-trivial SCFTs on the
CY manifold.
Now we consider the potential energy density and look for supersymmetric vacua. Imposing the Wess-Zumino
gauge, we write down the bosonic part of the potential energy density U(ϕ):
U(ϕ) =e2
2D2 +
∣∣Gℓ(s)∣∣2 +
N∑
i=1
∣∣p1∂iGℓ(s)∣∣2 + Uσ(ϕ) , (I.7.3a)
D = r −N∑
i=1
|si|2 + ℓ|p1|2 + (N − ℓ)|p2|2 , (I.7.3b)
Uσ(ϕ) = 2|σ|2 N∑
i=1
|si|2 + ℓ2|p1|2 + (N − ℓ)|p2|2. (I.7.3c)
Imposing zero on them, we obtain the supersymmetric vacuum manifold M. Since the Lagrangian has N = (2, 2)
supersymmetry and the single U(1) gauge symmetry, the vacuum manifold becomes a Kahler quotient space:
M =
(ϕa) ∈ CN+3∣∣∣D = Gℓ = p1∂iGℓ = Uσ = 0
/U(1) , (I.7.4)
In attempt to study effective theories, we choose a point on M as a vacuum and give VEVs of scalar component
fields: ϕa ≡ 〈ϕa〉. Then we expand the fluctuation modes around the vacuum. In general, the structure of M is
different for r > 0, r = 0 and r < 0 and there appear various phases in the GLSM. The phase living in the r > 0
region is referred to the “CY phase,” and the phase in r < 0 is called to the “orbifold phase.” A singularity of the
model emerges in the phase at r = 0. Thus we sometimes call this the “singularity phase.” We will treat these
three cases separately. We comment that in each phase the vacuum manifold is reduced from the original M. We
often refer the reduced vacuum manifold toMr ⊂M. The VEVs of the respective phases can be set only inMr.
I.7.2 Calabi-Yau phase
In this subsection we analyze the CY phase r > 0. In this phase, D = 0 requires some si cannot be zero and
therefore σ must vanish. If we assume p1 6= 0, the equations Gℓ(s) = ∂iGℓ(s) = 0 with the condition (I.7.2) imply
that all si must vanish. However this is inconsistent with D = 0. Thus p1 must be zero. The variable p2 is free as
long as the condition D = 0 is satisfied. Owing to these, the vacuum manifoldM is reduced to MCY defined by
MCY =
(si; p2) ∈ CN+1∣∣∣D = Gℓ(s) = 0 , r > 0
/U(1) . (I.7.5)
Here we explain this manifold in detail. This is an (N − 1)-dimensional noncompact Kahler manifold. The
components si denote the homogeneous coordinates of the complex projective space CPN−1. The constraint
Gℓ(s) = 0 reduces CPN−1 to a degree ℓ hypersurface expressed to CPN−1[ℓ]. we find that p2 is a fiber coordinate
of the O(−N + ℓ) bundle on CPN−1[ℓ]. Furthermore the vanishing sum of U(1) charges indicates that the FI
parameter r is not renormalized. This is equivalent to c1(MCY) = 0. Thus we conclude that the reduced vacuum
manifold MCY is nothing but a noncompact CY manifold on which a non-trivial superconformal field theory is
realized.
26 N = (2, 2) Gauged Linear Sigma Model
Let us consider a low energy effective theory. We choose a vacuum and take a set of VEVs of the scalar
component fields. Because ∃〈si〉 6= 0, the U(1) gauge symmetry is spontaneously broken down completely. Next,
we expand all fields in terms of fluctuation modes such as ϕa = 〈ϕa〉+ ϕa + ϕa. We set 〈p1〉, 〈σ〉 and σ to be zero.
Substituting them into the potential energy density (I.7.3), we obtain
U =e2
2
2Re
[−
N∑
i=1
2si〈si〉+ (N − ℓ)p2〈p2〉]−
N∑
i=1
∣∣si + si
∣∣2 + ℓ|p1|2 + (N − ℓ)∣∣p2 + p2
∣∣22
+∣∣∣
N∑
i=1
si ∂iGℓ(〈s〉) +
ℓ∑
k=2
1
k!
∑
i1,··· ,ik
(s+ s)i1 · · · (s+ s)ik· ∂i1 · · · ∂ik
Gℓ(〈s〉)∣∣∣2
+ |p1|2N∑
i=1
∣∣∣∂iGℓ(〈s〉) +ℓ−1∑
k=2
1
k!
∑
j1,··· ,jk
(s+ s)j1 · · · (s+ s)jk· ∂i∂j1 · · · ∂jk
Gℓ(〈s〉)∣∣∣2
+ 2|σ|2 N∑
i=1
∣∣〈si〉+ si + si
∣∣2 + ℓ2|p1|2 + (N − ℓ)2∣∣〈p2〉+ p2 + p2
∣∣2.
Fluctuation modes si and p2 remain massless and move only tangent toMCY because they are subject to δD|VEV =
δGℓ|VEV = 0. The variation δ(p1∂iGℓ)|VEV = 0 indicates p1 = 0. The modes σ, p1, si and p2 have mass
m2 = O(e2r). The gauge field vm also acquires mass of order O(e2r) by the Higgs mechanism. The fermionic
superpartners behave in the same way as the scalar component fields because of preserving supersymmetry. In the
IR limit e→∞ and the large volume limit r →∞, the massive modes decouple from the system. Thus we obtain
N = (2, 2) supersymmetric NLSM onMCY (I.7.6)
as a massless effective theory. Notice that this description is only applicable in the large volume limit because
the NLSM is well-defined in the weak coupling limit from the viewpoint of perturbation theory. This effective
theory becomes singular if we take the limit r → +0 because the decoupled massive modes becomes massless. This
phenomenon also appears in the Seiberg-Witten theory [95, 96], the black hole condensation [100, 39], and so on.
Let us make a comment on the target space MCY. By definition, the number ℓ means the degree of the
vanishing polynomial Gℓ(s) = 0, which gives a hypersurface in the projective space CPN−1. We can see that if
ℓ = 1, Gℓ=1(s) = 0 gives a linear constraint with respect to the homogeneous coordinates si and the hypersurface
CPN−1[ℓ = 1] is reduced to (N − 2)-dimensional projective space CPN−2. This reduction does not occur if
2 ≤ ℓ ≤ N − 1. Here we summarize the shape of the target space MCY in Table I.1:
degree ℓ vacuum manifold MCY
ℓ = 1 O(−N + 1) bundle on CPN−2
2 ≤ ℓ ≤ N − 1 O(−N + ℓ) bundle on CPN−1[ℓ]
Table I.1: Classification of O(−N + ℓ) bundle on CPN−1[ℓ].
Although the ℓ = 1 case has been already analyzed in the original paper [113], the other cases 2 ≤ ℓ ≤ N − 1 are
the new ones which have not been analyzed.
I.7 O(−N + ℓ) bundle on CPN−1[ℓ] 27
I.7.3 Orbifold phase
Here we consider the negative FI parameter region r < 0. In this region the total vacuum manifold (I.7.4) is
restricted to a subspace defined by
Morbifold =
(p1, p2; si) ∈ CN+2∣∣∣D = Gℓ = p1∂iGℓ = 0 , r < 0
/U(1) . (I.7.7)
Since D = 0 does not permit p1 and p2 to vanish simultaneously, σ must be zero. This subspace is quite different
from MCY in the CY phase. In addition, the shape of Morbifold is sensitive to the change of the degree ℓ because
of the existence of the constraints Gℓ = p1∂iGℓ = 0 and the property (I.7.2). Thus let us analyze Morbifold and
study massless effective theories on it in the case of 3 ≤ ℓ ≤ N − 1, ℓ = 2 and ℓ = 1, separately.
Effective theories of 3 ≤ ℓ ≤ N − 1
Here we analyze the vacuum manifold Morbifold and massless effective theories of 3 ≤ ℓ ≤ N − 1. Owing to the
constraints Gℓ = p1∂iGℓ = 0 and their property (I.7.2), the manifold Morbifold is decomposed into the following
two subspaces:
Morbifold
∣∣3≤ℓ≤N−1
= M1r<0 ∪M2
r<0 , (I.7.8a)
M1r<0 :=
(p1, p2) ∈ C2
∣∣∣D = 0 , r < 0/
U(1) , (I.7.8b)
M2r<0 :=
(p2; si) ∈ CN+1
∣∣∣D = Gℓ = 0 , r < 0/
U(1) . (I.7.8c)
In the former subspace the condition (I.7.2) is trivially satisfied whereas in the latter subspace it is satisfied non-
trivially. Both of the two subspace include a specific region p1 = ∀si = 0. The subspace M1r<0 is defined as a
one-dimensional weighted projective space WCP1ℓ,N−ℓ represented by two complex fields p1 and p2 of U(1) charges
−ℓ and −(N − ℓ), respectively. The precise definition of the weighted projective space is in appendix I.C.
Let us choose a supersymmetric vacuum and set VEVs of all scalar fields. Then we expand all the fields around
the VEVs. Expanding the potential energy density (I.7.3) in terms of VEVs and fluctuation modes, we obtain the
following form:
U =e2
2
2Re
[ℓ p1〈p1〉+ (N − ℓ)p2〈p2〉
]−∑
i
∣∣si + si
∣∣2 + ℓ∣∣p1 + p1
∣∣2 + (N − ℓ)∣∣p2 + p2
∣∣22
+∣∣Gℓ(s+ s)
∣∣2 +∣∣〈p1〉+ p1 + p1
∣∣2 ·∑
i
∣∣∂iGℓ(s+ s)∣∣2
+ 2|σ|2∑
i
∣∣si + si
∣∣2 + ℓ2∣∣〈p1〉+ p1 + p1
∣∣2 + (N − ℓ)2∣∣〈p2〉+ p2 + p2
∣∣2,
where 〈p1〉 and 〈p2〉 are VEVs of scalar components of P1 and P2, respectively. They live in the weighted projective
space (I.7.8b). Because the VEVs of si are all zero, the U(1) gauge symmetry is spontaneously broken to Zα,
where α is the great common number between ℓ and N − ℓ: α = GCMℓ,N − ℓ. This potential energy density
provides that all fluctuation modes si and si appear as linearly combined forms such as si + si, which do not acquire
any mass terms. The modes p1 and p2 remain massless and move tangent to the subspace (I.7.8b). The other
28 N = (2, 2) Gauged Linear Sigma Model
fluctuation modes acquire mass of order m2 = O(e2|r|). Thus, in the IR limit e → ∞, all the massive modes are
decoupled from the system. Thus we obtain the following massless effective theory:
N = (2, 2) supersymmetric NLSM on WCP1ℓ,N−ℓ
coupled to “LG” theory withWLG = (〈p1〉+ P1)Gℓ(S)
/Zα , (I.7.9)
where P1 and Si are massless chiral superfields. Note that the sigma model sector also contains the Zα orbifold
symmetry coming from the property of WCP1ℓ,N−ℓ. As is well known that the term 〈p1〉Gℓ(S) forms an ordinary LG
superpotential. Thus in the IR limit we can interpret that this term is marginal and flows to the N = (2, 2) minimal
model. The second term P1 ·Gℓ(S) is somewhat mysterious. Since this term has not any isolated singularities we
might not obtain well-defined unitary CFT. This difficulty causes the noncompactness of the manifoldMCY which
appears in the CY phase.
There are two specific points in the subspace WCP1ℓ,N−ℓ. One is the point p2 = 0 and the other is p1 = 0.
In the former point the gauge symmetry is enhanced to Zℓ. Furthermore the mode p1 disappears and p2 becomes
massless, which combines with a massless fluctuation modes p2 linearly. This combined mode is free from any
constraints. The other massless modes si + si in (I.7.9) remain massless and are also free from constraints. Thus
in the IR limit and the large volume limit, the massless effective theory becomes an N = (2, 2) supersymmetric
theory as
CFT on C1 ⊗ LG theory with WLG = 〈p1〉Gℓ(S)/
Zℓ . (I.7.10)
This effective theory consists of N + 1 massless chiral superfields such as P2 and Si, which live in the free and the
LG sectors, respectively. Since we take the IR limit, this effective theory becomes an SCFT. The LG sector flows
to a well-known LG minimal model [113]. Thus the sigma model sector is also a superconformal field theory. Here
we notice that we did not integrate out but just decomposed all massive modes in the above discussion because it is
generally impossible to calculate the integration of them. Thus the above effective theory is merely an approximate
one. If we will be able to integrate out all massive modes exactly, the obtaining effective theory will be different
from the above one. In later section we will discuss the exact form of the effective theory.
Next let us consider the latter point p1 = 0 in the space WCP1ℓ,N−ℓ. On this point the broken gauge symmetry
is partially restored to ZN−ℓ. The massless fluctuation mode p2 becomes zero whereas the massive mode p1 becomes
massless, which combines with p1 being free from any constraints. Thus P1 appears as a massless chiral superfield.
In the IR limit we obtain the supersymmetric massless effective theory such as
LG theory with WLG = P1 ·Gℓ(S) on CN+1/
ZN−ℓ , (I.7.11)
which consists of N + 1 massless chiral superfields such as P1 and Si. This theory is not a well-defined LG theory
because the superpotential WLG has no isolated singularities. We interpret the defect of isolated singularities as
a noncompactness of the manifold MCY in the CY phase via CY/LG correspondence (if this correspondence is
satisfied in the case of sigma models on noncompact CY manifolds.) This property prevents from calculating a
chiral ring of this model in the same way as unitary LG minimal models describing compact CY manifolds [72].
Here we study massless effective theories on the subspaceM2r<0 defined in (I.7.8c). As mentioned before, there
are non-trivial constraints in M2r<0. Thus, as we shall see, the effective theories are also under these constraints.
I.7 O(−N + ℓ) bundle on CPN−1[ℓ] 29
In the same way as discussed before, we choose one point in the subspace M2r<0 and make all the scalar fields
fluctuate around it. Then we write down the expanded potential energy density (I.7.3) in terms of VEVs and
fluctuation modes 〈ϕa〉, ϕa and ϕa:
U =e2
2
2Re
[−∑
i
si〈si〉+ (N − ℓ)p2〈p2〉]−∑
i
∣∣si + si
∣∣2 + ℓ∣∣p1
∣∣2 + (N − ℓ)∣∣p2 + p2
∣∣22
+∣∣∣∑
i
si ∂iGℓ(〈s〉) +
ℓ∑
k=2
1
k!
∑
i1,...,ik
(s+ s)i1 · · · (s+ s)ik· ∂i1 · · · ∂ik
Gℓ(〈s〉)∣∣∣2
+ |p1|2 ·∑
i
∣∣∣∂iGℓ(〈s〉) +ℓ−1∑
k=1
1
k!
∑
j1,...,jk
(s+ s)j1 · · · (s+ s)jk· ∂i∂j1 · · · ∂jk
Gℓ(〈s〉)∣∣∣2
+ 2|σ|2∑
i
∣∣〈si〉+ si + si
∣∣2 + ℓ2∣∣p1
∣∣2 + (N − ℓ)2∣∣〈p2〉+ p2 + p2
∣∣2.
This potential energy density indicates the following: The fluctuation modes si, p1 and p2 are massive; si and p2
move tangent to M2r<0. Thus, taking e→∞ and |r| → ∞, we obtain
N = (2, 2) supersymmetric NLSM onM2r<0 . (I.7.12)
In this theory there exist P2 and Si as massless chiral superfields, which move tangent to M2r<0. Notice that in
general points inM2r<0 the U(1) gauge symmetry is completely broken because of the existence of non-zero VEVs
〈si〉. However, taking ∀〈si〉 = 〈p1〉 = 0 and 〈p2〉 6= 0 in the subspace M2r<0, we find that the gauge symmetry is
partially restored to ZN−ℓ.
Even though the vacuum manifoldsM1r<0 andM2
r<0 are connected on p1 = ∀si = 0, the effective theories given
by (I.7.11) and (I.7.12) are quite different from each other. The reason is that while the subspaceM1r<0 is free from
constraints Gℓ = p1∂iGℓ = 0, in the subspace M2r<0 these constraints are still valid on the region p1 = ∀si = 0.
On account of the existence of these constraints, a phase transition occurs when the theory moves from one to
the other. Thus we conclude that a new phase appears on the subspace M2r<0, which has not been discovered in
well-known GLSMs such as the models for O(−N) bundle on CPN−1, for CPN−1[N ], for resolved conifold, and so
on. We refer this phase to the “3rd phase.” Here we refer the phase onM1r<0 to the orbifold phase, as usual.
Effective theories of ℓ = 2
Let us consider the orbifold phase of ℓ = 2. In the same way as the previous analysis, the constraints Gℓ = p1∂iGℓ =
0 and the property (I.7.2) decompose the manifoldMorbifold into two subspaces:
Morbifold
∣∣ℓ=2
= M1r<0 ∪M2
r<0 , (I.7.13a)
M1r<0 :=
(p1, p2) ∈ C2
∣∣∣D = 0 , r < 0/
U(1) ≡ WCP12,N−2 , (I.7.13b)
M2r<0 :=
(p2; si) ∈ CN+1
∣∣∣D = G2 = 0 , r < 0/
U(1) . (I.7.13c)
These two subspaces are glued in the region given by p1 = ∀si = 0. Although this situation is same as to the case
of 3 ≤ ℓ ≤ N − 1, the appearing massless effective theories are quite different.
30 N = (2, 2) Gauged Linear Sigma Model
Here let us analyze the effective theories on the subspace M1r<0 = WCP1
2,N−2. We choose a point in this
subspace as a supersymmetric vacuum and take VEVs of all scalar fields. Then we make all scalar fields fluctuate
around the VEVs. Fluctuation modes p1 and p2 are subject to the constraints such that they move only tangent
to WCP12,N−2. The fluctuation modes si have no degrees of freedom because of the variation of the constraint
p1∂iG2 = 0. (In the case of (I.7.8b), the equations Gℓ = 0 and p1∂iGℓ = 0 are trivially satisfied in WCP1ℓ,N−ℓ.
These variations are also trivial. However the case of ℓ = 2 is quite different. By definition, some ∂i∂jG2 must
have non-zero values. Thus even though the above equations are trivially satisfied in the subspace WCP12,N−2,
their variations give non-trivial constraints on the fluctuation modes.) Under these conditions we write down the
potential energy density (I.7.3) in terms of VEVs 〈ϕa〉 and fluctuation modes ϕa and ϕa:
U =e2
2
2Re
[2p1〈p1〉+ (N − 2)p2〈p2〉
]−∑
i
∣∣si
∣∣2 + 2∣∣p1 + p1
∣∣2 + (N − 2)∣∣p2 + p2
∣∣22
+∣∣G2(s)
∣∣2 +∣∣〈p1〉+ p1 + p1
∣∣2 ·∑
i
∣∣∂iG2(s)∣∣2
+ 2|σ|2∑
i
∣∣si
∣∣2 + 4∣∣〈p1〉+ p1 + p1
∣∣2 + (N − 2)2∣∣〈p2〉+ p2 + p2
∣∣2.
This function denotes the following: The fluctuation modes si, p1 and p2 acquire masses m2 = O(e2|r|); the modes
p1 and p2 remain massless and move tangent to WCP12,N−2. Thus taking e → ∞ and |r| → ∞, we obtain the
massless effective theory described by
N = (2, 2) supersymmetric NLSM on WCP12,N−2 . (I.7.14)
This sigma model has Zα orbifold symmetry coming from the property of WCP12,N−2, where α = GCM2, N − 2.
This effective theory does not include massless LG theory. The reason is that the degree two polynomial G2
generates mass terms such as |〈p1〉|2∑
i |∂iG2|2. (See, for example, [105].)
Now we consider the effective theory on two specific points in WCP12,N−2 like (I.7.10) and (I.7.11). Expanding
the theory on the one point (p1, p2) = (p1, 0), the gauge symmetry is partially restored to Z2. Thus we obtain the
effective theory on this specific point as
N = (2, 2) SCFT on C1/Z2 . (I.7.15)
Note that this theory can possess the LG theory with a quadratic superpotential WLG = 〈p1〉G2(S), which gives
massive modes of Si.
The effective theory drastically changes if we expand the theory on another point (p1, p2) = (0, p2) in WCP12,N−2.
On this point, the broken gauge symmetry is enhanced to ZN−2 and the fluctuation modes si become massless
with being free from any constraints. Both p1 and p1 are massless and linearly combined in the potential energy
density. The remaining field p2 becomes zero because there exists a non-trivial variation of the constraint D = 0.
Summarizing these results, we find that the following massless effective theory appears in the limit e, |r| → ∞:
N = (2, 2) “LG” theory with WLG = P1 ·G2(S) on CN+1
/ZN−2 . (I.7.16)
Although this superpotential also has no isolated singularities, this theory should describe a non-trivial SCFT. We
shall return here in later discussions.
I.7 O(−N + ℓ) bundle on CPN−1[ℓ] 31
We next study the massless effective theories on the subspace M2r<0 defined in (I.7.13c). The potential energy
density is obtained as
U =e2
2
2Re
[−∑
i
si〈si〉+ (N − 2)p2〈p2〉]−∑
i
∣∣si + si
∣∣2 + 2∣∣p1
∣∣2 + (N − 2)∣∣p2 + p2
∣∣22
+∣∣∣∑
i
si ∂iG2(〈s〉) +1
2!
∑
i,j
(s+ s)i(s+ s)j · ∂i∂jG2(〈s〉)∣∣∣2
+ |p1|2 ·∑
i
∣∣∣∂iG2(〈s〉) +∑
j
(s+ s)j · ∂i∂jG2(〈s〉)∣∣∣2
+ 2|σ|2∑
i
∣∣〈si〉+ si + si
∣∣2 + 4∣∣p1
∣∣2 + (N − 2)2∣∣〈p2〉+ p2 + p2
∣∣2
under the following constraints on fluctuation modes: The fluctuations si and p2 move tangent toM2r<0; the other
tangent mode p1 is zero; the fluctuations ϕa are all massive of m2 = O(e2|r|). Thus the effective theory expanded
around generic points inM2r<0 becomes
N = (2, 2) supersymmetric NLSM onM2r<0 (I.7.17)
in the IR and large volume limit: e, |r| → ∞. The U(1) gauge symmetry is completely broken if some 〈si〉 6= 0
exist. On the other hand, if we expand the theory on a specific point p1 = ∀si = 0, the gauge symmetry is partially
restored to ZN−2.
So far we have studied the effective theories on all regions of the vacuum manifoldMorbifold of ℓ = 2. From the
same reason discussed in the case of 3 ≤ ℓ ≤ N − 1, there exists a phase transition between the theories (I.7.16)
and (I.7.17) because of the non-trivial constraint coming from the variation of the equation G2 = 0. Thus we find
that the GLSM for the O(−N + 2) bundle on CPN−1[2] also includes two phases in the negative FI parameter
region. The phase on (I.7.14) is called the orbifold phase, and we refer the phase on (I.7.17) to the 3rd phase.
Here we illustrate the relation among the phases in the GLSM schematically in Figure I.5:
CY phase
orbifold phase
3rd phase
|si|2
|p1|2
|p2|2
|si|2
|p1|2
|p2|2
Figure I.5: Various phases in GLSM for O(−N + ℓ) bundle on CPN−1[ℓ] with 2 ≤ ℓ ≤ N − 1. The axes with
thin/thick lines represent the vacuum space coordinates in the positive/negative FI parameter regions, respectively.
32 N = (2, 2) Gauged Linear Sigma Model
In the large volume limit |r| → ∞, these three effective theories (I.7.6), (I.7.9) and (I.7.12) become well-defined.
In later discussions we shall consider how these effective theories deform in the small FI parameter limit |r| → 0.
There we must consider the singular phase [113].
Effective theory of ℓ = 1
Finally we investigate the ℓ = 1 case. Since the polynomial Gℓ=1(S) is of degree one, there exist some non-zero
values of ∂iG1(S). Thus, combining this condition with the other constraints which define Morbifold, we find that
p1 must be zero and obtain the following reduced vacuum manifold:
Morbifold
∣∣ℓ=1
=
(p2; si) ∈ CN+1∣∣∣D = G1 = 0 , r < 0
/U(1) =: M2
r<0 . (I.7.18)
Since this space is defined in the same way as (I.7.8c) and (I.7.13c), we also referred it toM2r<0.
After taking VEVs of scalar fields which live in (I.7.18), we make scalar fields fluctuate around the VEVs. These
fluctuation modes are subject to constraints: si and p2 move only tangent toM2r<0; p1 is zero. Substituting these
into (I.7.3), we obtain the expanded potential energy density
U =e2
2
2Re
[−∑
i
si〈si〉+ (N − 1)p2〈p2〉]−∑
i
∣∣si + si
∣∣2 +∣∣p1
∣∣2 + (N − 1)∣∣p2 + p2
∣∣22
+∣∣G1(s)
∣∣2 +∣∣p1
∣∣2 ·∑
i
∣∣∂iG1(〈s〉)∣∣2
+ 2|σ|2∑
i
∣∣〈si〉+ si + si
∣∣2 +∣∣p1
∣∣2 + (N − 1)∣∣〈p2〉+ p2 + p2
∣∣2.
This indicates that the modes si and p2 remain massless whereas the modes si, p1 and p2 become massive. Thus
the following massless effective theory appears in the limit e, |r| → ∞:
N = (2, 2) supersymmetric NLSM onM2r<0 , (I.7.19)
where the U(1) gauge symmetry is completely broken because ∃〈si〉 6= 0. While if we set the VEVs to ∀〈si〉 = 0, the
broken U(1) gauge symmetry is enhanced to ZN−1. In addition the modes si become massless and are combined
with the tangent modes si, which are still under constraint G1 = 0. The mode p2 becomes zero, which is derived
from the variation of D = 0. In this specific point we can see that the spaceM2r<0 is deformed to CN−1/ZN−1 and
the effective theory are
N = (2, 2) SCFT on CN−1/ZN−1 . (I.7.20)
The two effective theories (I.7.19) and (I.7.20) are smoothly connected without any phase transitions coming from
the variations of constraints. Thus we find that in the ℓ = 1 case there exists only one phase in the negative FI
parameter region. We refer this to the orbifold phase, as usual. Here we illustrate the schematic relation between
the CY phase and the orbifold phase in Figure I.6:
Note that the GLSM for O(−N + 1) bundle on CPN−1[1] is completely the same as the one for O(−N + 1)
bundle on CPN−2. This is because the hypersurface CPN−1[1] is nothing but CPN−2. Thus the vacuum structure
and phases are also equal to each other.
I.7 O(−N + ℓ) bundle on CPN−1[ℓ] 33
CY phase
orbifold phase
|si|2
|p1|2
|p2|2
|si|2
|p1|2
|p2|2
Figure I.6: Various phases in GLSM for O(−N + 1) bundle on CPN−1[1].
I.7.4 Singularity phase
In this subsection let us analyze the singularity phase. As mentioned before, the effective theory (I.7.6) becomes
singular if r → +0. The effective theories in the orbifold and the 3rd phases also become singular if r → −0. Thus
we will study the singularity phase r = 0 in order to avoid the singularities in effective theories. In this analysis
we will find that there appears a new branch. Then we will discuss how to avoid the singularity.
Here we study how the vacuum manifold (I.7.4) is reduced in the r = 0 phase. If we assume p1 6= 0, then we
obtain∑
i |si|2 6= 0 from D = 0. However the equations Gℓ(s) = 0 and p1∂iGℓ(s) = 0 insist that all si vanish. This
is a contradiction. Thus p1 must be zero. Under this condition, we obtain two solutions. One is obtained by D = 0
and σ = 0. In general this solution has non-zero φa, where φa are scalar component fields of chiral superfields. The
other solution is given by ∀φa = 0 and σ is free. We refer the former and latter solutions to the Higgs and Coulomb
branches, respectively. These branches are similar to the ones of N = 2 SQCD in four dimensions [96]. (The CY,
orbifold and 3rd phases are all in the “Higgs branch”.) They are connected if all scalar component fields vanish:
∀ϕa = 0. Now we analyze the effective theories on these two branches.
Higgs and Coulomb branches
Let us consider the Higgs and Coulomb branches in detail. In the Higgs branch, there exist two supersymmetric
vacuum solutions. One is
∀〈φa〉 = O(|r|) → 0 , 〈σ〉 = 0 . (I.7.21a)
This solution is smoothly connected with the supersymmetric vacuum solutions in the phases of non-vanishing FI
parameter. The other is
〈p1〉 = 0 , ∀〈si〉 , 〈p2〉 : arbitrary order , 〈σ〉 = 0 , (I.7.21b)
which is satisfied only on r = 0. Although the first solution (I.7.21a) appears in each GLSM, the second solution
(I.7.21b) does not satisfy the supersymmetric vacuum condition U(ϕ) = 0 in some GLSMs, for example, the GLSM
34 N = (2, 2) Gauged Linear Sigma Model
for CPN−1[N ].
In the Coulomb branch, we can set that ∀〈φa〉 = 0 and 〈σ〉 is free. This solution appears only when the FI
parameter vanishes. If we choose 〈σ〉 to be zero, i.e., all the VEVs of scalar fields vanish 〈ϕa〉 = 0, the Coulomb
branch connects with the Higgs branch.
Let us consider a massless effective theory in the Coulomb branch. Since the scalar field σ has mass dimension
one, we take the VEV 〈σ〉 to be very large. Owing to this, all chiral superfields Si, P1 and P2 acquire very large
masses via Uσ(ϕ) in (I.7.3). Taking 〈σ〉 → ∞ and integrating out all massive fields18, we obtain the following
effective Lagrangian:
Leff =
∫d4θ
−Keff(Σ,Σ)
+( 1√
2
∫d2θ Weff(Σ) + c.c.
),
Weff(Σ) = −Σ t−∑
a
QaΣ
log(QaΣ
µ
)− 1
= −Σ(t− ℓ log(−ℓ)− (N − ℓ) log(−N + ℓ)
).
Note that the twisted superpotential was deformed by the quantum effects coming from the integration of massive
fields. This effective Lagrangian presents the asymptotic form of the potential energy density
Ueff(σ) =e2eff2
∣∣∂σWeff(σ)∣∣2 =
e2eff2
∣∣∣t− ℓ log(−ℓ)− (N − ℓ) log(−N + ℓ)∣∣∣2
. (I.7.22)
In order that the effective theory remains supersymmetric, the potential energy density must be zero in a specific
value of σ. Notice that if the complexified FI parameter is given by
t = ℓ log(−ℓ) + (N − ℓ) log(−N + ℓ) , (I.7.23)
the potential energy density becomes always zero. If it happens, the effective theory does not have any mass gap
and becomes singular as a two-dimensional field theory. Thus (I.7.23) is the quantum singular point of the GLSM.
In the classical point of view, the value t = 0 looks like a singular point in the theory. Integrating out the massive
fields, we find that the singular point moves to (I.7.23). The massless effective theories in Coulomb and Higgs
branches are connected with each other avoiding this singular point.
CY/LG correspondence and topology change
As mentioned before, the massless effective theories are only valid if we take the FI parameter to be infinitely
large |r| → ∞. In this limit the effective theories are (partly) described by the NLSMs. However if we change
the FI parameter to be small, the NLSM representations are no longer well-defined and must be deformed. This
phenomenon has been already studied in [113] as following: If the FI parameter goes to zero r → 0, the effective
theory on the CY phase moves to the theory on the Coulomb branch in the singularity phase avoiding the singular
point. Furthermore the effective theory connects to the LG theory in the orbifold phase when r → −∞.
18Here we can integrate out Φa because the superpotential WGLSM does not contribute to the deformation of the effective twisted
superpotential fWeff(Σ). For the precise derivation, see chapter 15 in [53].
I.7 O(−N + ℓ) bundle on CPN−1[ℓ] 35
The above phenomenon suggests that, rather than the LG theory being equivalent to the sigma model on the CY
manifold, they are two different phases of the same system, i.e., the system of the single GLSM. Thus the CY/LG
correspondence [78, 103, 104, 41, 79, 56, 14] can be read from the phase transition. In fact, it has been proved that
the sigma model on CP4[5] and the Z5-orbifolded LG theory with WLG = G5(S), which are equivalent to each
other, appear as the distinct phases in the single GLSM. Furthermore the topology change is also understood in
the framework of the phase transition of the GLSM. The flop of the resolved conifold O(−1)⊕O(−1)→ CP1 is a
typical example [113].
Now let us apply the above discussions to the GLSM for O(−N + ℓ) bundle on CPN−1[ℓ]. For example we
consider the relations among the various phases of 3 ≤ ℓ ≤ N − 1, where we have found three phases: The CY
phase onMCY, the orbifold phase onM1r<0 and the 3rd phase onM2
r<0. Furthermore we have found four effective
theories. Let us discuss the relations among them:
• The effective theory on the CY phase (I.7.6) and the theory on the 3rd phase (I.7.12) are related to each
other via a topology change, because the defining equations of the target spaces are equal except for the sign
of the FI parameter. Furthermore these two effective theories are both sigma models, which do not include
the potential theory sectors such as a LG theory.
• (I.7.11) and (I.7.12) are connected at the point p1 = ∀si = 0 in Morbifold. Since the former is the sigma
model and the latter is the LG theory, there exists a phase transition between these two theories, which are
equivalent to each other by the CY/LG correspondence.
• Both the theories (I.7.10) and (I.7.11) are on the weighted projective space and are included in the theory
(I.7.9).
• The LG theory (I.7.10) is equivalent to the sigma model (I.7.6) by the CY/LG correspondence.
Notice that the sigma model (I.7.6) is not related to (I.7.11) directly, because the theory (I.7.11) has already
been connected to (I.7.12) while the CY/LG correspondence connects between two theories by one-to-one. These
connections are realized through the singularity phase. Even though we wrote down the connections only from the
qualitative point of view, we can acquire non-trivial relations among the effective theories as illustrated in Figure
I.7:
In the case of compact CY manifolds, we have already understood that the local rings in the LG theory
are identified with the chiral rings of the SCFT and these chiral rings are related to the harmonic forms on such
manifolds [72]. However we have no proof that this relation is also satisfied in the case of noncompact CY manifolds.
Thus we must investigate the spectra of the above effective theories as a future problem.
As discussed before, we have obtained various massless effective theories by decomposing all massive modes.
Thus they are just approximate descriptions which must be deformed if we can exactly integrate out massive modes.
In the next section we will study the T-dual theory of the GLSM [54]. This formulation is so powerful to obtain
the exact effective theories. Analyzing them exact theories we will re-investigate the massless effective theories in
the original GLSM.
36 N = (2, 2) Gauged Linear Sigma Model
PSfrag
CY phase
at r ≫ 0
Orbifold phase
at r ≪ 0
3rd phase
at r ≪ 0
singularity
Zℓ-orb.
LG
ZN−ℓ-orb.
“LG”
topology change
CY/LG corresp.CY/LG corresp.
Figure I.7: The relation among various phases around the singularity: a conjecture.
I.8 Squashed GLSM
Hori [49] から次の文章を抜粋しておこう。但し notation はこのノートにあわせておく。また前述した事も反復して
記載する場合がある。さらには途中で、chapter VI の記述も用いる。
Consider a U(1)k gauge theory with N chiral multiplets Φi of charge Qai (i = 1, · · · , N , a = 1, · · · , k). The
classical Lagrangian is given by
L =
∫d4θ−
k∑
a=1
1
e2a|Σa|2 +
N∑
i=1
Φi e2Qi·V Φi
+( 1√
2
∫d2θ
k∑
a=1
(− Σaτa
)+ c.c.
), (I.8.1)
where Qi · V =∑k
a=1Qai Va and Σa = 1√
2D+D−Va (fieldstrength). The theory is classically parametrized by
the dimensionful gauge coupling constants ea and dimensionless FI-Theta parameters τa = ra − iθa. It is super-
renormalizable with respect to ea but ra has to be renormalized as
ra(µ′) = ra(µ) + ba1 log(µ′/µ) , ba1 =
N∑
i=1
Qai
(causing a dimensional transmutation if ba1 6= 0).
(途中省略)
What we have done above is essentially T-duality with respect to the torus fibration determined by the U(1)N−k
global symmetry. This group action is not free and the circle fibre for arg Φi shrinks to zero size at the locus
|Φi| = 0. One would naively expect that the dual torus blows up at the same locus. However, what we have seen
shows that a superpotential term e−Yi is generated in the dual theory. This diverges toward the locus |Φi| = 0, and
breaks the rotational invariance of the dual theory, accounting for the loss of winding number in the sigma model
due to the degeneration of the circle. It would be interesting to see how universal this phenomenon is.
I.8 Squashed GLSM 37
A while ago, Fateev, Zamolodchikov and Zamolodchikov [24] conjectured a duality between the two-dimensional
black hole background and a potential theory called sine-Liouville theory.19 The former has a semi-infinite cigar
geometry which asymptotes to a flat cylinder with a linear dilaton. The momentum along the circle is conserved
but the winding number is not, because a winding string can strip off at the tip of the cigar. The latter is a theory
of a cylinder variable with (Liouville × cosine) potential. This also has an asymptotic region with a linear dilaton
where the potential is exponentially small. Winding number is conserved but the rotational symmetry is broken by
the cosind potential. Comparison of the asymptotic regions suggests that the two theories are related by T-duality.
The degeneration of the circle in the 2d black hole corresponds to the growing sine-Liouville potential. This is
strongly reminiscent of our mirror symmetry.
In [51] a proof is given for the supersymmetric version of the FZZ duality [34] (c.f. [82, 23]). It is between the 2d
black hole background for fermionic strings, SL(2,R)k+2 mod U(1) Kazama-Suzuki super-coset model, and N = 2
Liouville theory, the LG model of a periodic chiral superfield Y ≡ Y + 2πi with the superpotential W = e−Y and
the Kahler potential K = |Y |2/2k. The two theories are weakly coupled in the opposite regimes k ≫ 1 and k ≪ 1.
The essential part of the proof is to show that the SCFT of the 2d black hole arizes as the infra-red fixed point
of the following U(1) gauge theory. It is the theory with two chiral superfields Φ and P ≡ P +2πi which transforms
as Φ→ eiαΦ and P + iα under the U(1) gauge transformation, with the Lagrangian given by
L =
∫d4θΦ e2V Φ +
k
2
(P + P + 2V
)2 − 1
e2|Σ|2
. (I.8.2)
The classical vacuum manifold is a semi-infinite cigar, but the metric is not precisely that of the 2d black hole and
the dilaton is not linear in the asymptotic region. However, a one-loop renormalization group analysis shows that it
flows to the 2d black hole background, including the linear dilaton. This analysis is valid at large k. Next, following
[114, 98], one computes the central charge of the infra-red fixed point, by identifying the N = 2 superconformal
algebra in the ring of left-chiral operators. This is possible when one can identify the right-moving R-current. The
axial U(1) R-symmetry is anomalous in the present system, but one can use the field P to modify the current in
a gauge invariant way so that it is conserved. By reality of the currents in the asymptotic resion, one can also fix
the ambiguity due to the presence of the global symmetry. Thus, one can completely identify the N = 2 algebra.
This shows that the central charge of the IR fixed point is
c = 3(
1 +2
k
), (I.8.3)
which is the correct value for the super-coset model. The modified parts of the N = 2 currents are linear in
P and this exhibits the linear dilaton in the asymptotic region (with the correct slope). Finally, one excludes
the possibility that the theory flows not to the super-coset itself but to some other nearby fixed point with the
same central charge, symmetries, and asymptotic behaviour. This is done by showing that the super-coset has no
supersymmetric, parity invariant marginal operator that is small at the asymptotic region.
The rest is a straightforward generalization of the argument in the previous section. Dualization of the phase of
Φ and the imaginary part of P turns them into twisted chiral superfields Y and YP (both period 2πi). The exact
19This was recently used in the Matrix Model formulation of string theory in the black hole background [59]. More recent studies
are in [33, 26, 36].
38 N = (2, 2) Gauged Linear Sigma Model
superpotential of the dual system is
W = Σ(Y + YP ) + e−Y . (I.8.4)
No non-perturbative superpotential is generated for YP since there is no P -vortex. The Kahler potential is
K = − 1
e2|Σ|2 − 1
2k|YP |2 + · · · , (I.8.5)
where · · · are small in the asymptotic region. In the limit e→∞, it is appropriate to integrate out Σ and we have
the constraint Y + YP = 0. In fact, one can use the rigidity of the super-coset to show that the terms · · · in (I.8.5)
vanishes. Thus, we conclude that the dual theory flows to the N = 2 Liouville theory with the Kahler potential
K = −|Y |2/2k. This completes the proof of the claimed equivalence.
This story has a generalization which can be presented as a deformation of the standard linear sigma model
(I.8.1). We gauge the U(1)N−k flavor group, acting on Φi with charge Riq (q = 1, . . . , N − k), and introduce for
each U(1) a chiral superfield Pq transforming inhomogeneously. The Lagrangian of the system reads
L =
∫d4θ−
k∑
a=1
1
e2a|Σa|2 +
N∑
i=1
Φi e2Qi·V +2Ri·V ′
Φi
+( 1√
2
∫d2θ
k∑
a=1
(− Σaτa
)+ c.c.
)
+
∫d4θ−
N−k∑
q=1
1
e2q|Σ′
q|2 +
N−k∑
q=1
kq
2
(Pq + P q + 2V ′
q
)2,
(I.8.6)
where Ri · V ′ =∑N−k
q=1 RiqV′q . The vacuum manifold X ′ is again a toric manifold with the same Kahler class
as X, but with a different metric. Deep in the interior of the base of the torus fibration, the sizes of the fibers
are constants ∼√kq. Thus X ′ is a “squashed” version of the toric manifold. In the limit kq → ∞, the (P , Σ′)
pairs decouple, and we recover the sigma-model on the “round toric manifold” X. When all ba1 = 0, the theory is
expected to flow to a non-trivial superconformal field theory. The N = 2 currents can be uniquely identified in the
left-chiral ring if there are asymptotic regions, and they include linear terms in Pq proportional to bq =∑N
i=1Riq,
implying a linear dilaton if bq 6= 0. The central charge is found as
c = 3(N − k +
N−k∑
q=1
2b2qkq
). (I.8.7)
In the “round limit” kq →∞, we recover c/3→ N − k = dimX.
The dual theory is found as before. In the limit ea, eq →∞, we obtain a LG model on the (C∗)N−k of (VI.2.12)
with the following Kahler and superpotentials
K = −1
2
N∑
i,j=1
gijY iYj + · · · , W =
N∑
i=1
e−Yi , (I.8.8)
where gij =∑N−k
q=1 RiqRjq/kq. This is the mirror of the sigma-model on the squashed toric manifold X ′. Note that
the Kahler potential depends on the squashing parameters kq and tends to vanish in the “round limit” kq → ∞.
For X = CP1, X ′ is sausage-shaped, and we find that the mirror of the supersymmetric sausage model is the
N = 2 sine-Gordon model with a finite Kahler potential. This equivalence has been conjectured by Fendley and
Intriligator.
I.8 Squashed GLSM 39
Examples
簡単な 2 つの例を挙げておこう:
• (N, k, q) = (1, 0, 0)→ (1, 0, 1): C1 → cigar
C1
cigar
√k
この squashed C1 = cigar の計量を導出する。本質的な役割を果たすものは、gauge 化された U(1)′ flavor
symmetry と、supersymmetry condition (D′-flatness condition) である。但しここでは、CP1 の Fubini-Study
metric の導出に際して superfield のまま計算する方法は採用しない。なお、参考は Hori and Kapustin [52] で
ある。
もともとは C1 を張る空間であることから、それを与える chiral superfield (complex coordinates) Φ を用いて、
Kahler potential と line element は次で定義される:
K = |Φ|2 , ds2 = |dφ|2 . (I.8.9)
(但しここの line elementの定義は、real coordinates x, yで張られるflat spaceの line elementをds2 = dx2+dy2
で定義し、これを complex coordinates z = x+ iy, z = x− iy で与えたとき、line element が
ds2 =1
2dz ⊗ dz +
1
2dz ⊗ dz ≡ dz dz , (I.8.10)
となることを用いている。) ここで U(1)′ isometry (flavor symmetry) を gauge 化しよう。自由度勘定から、
gauge symmetry の他に余分な complex coordinate p を導入する:
K =|Φ|2 + k|P |2
/(gauge symmetry + SUSY) , (I.8.11a)
ds2 =|dφ|2 + k|dp|2
/(gauge symmetry + SUSY) . (I.8.11b)
ここで gauge symmetry、supersymmetry condition は次で与えられる:
gauge :
φ→ e−2iλφ
Im(p)→ Im(p)− 2λSUSY : |φ|2 + 2kRe(p) = 0 . (I.8.12)
ここで complex coordinates φ, p を次のように分解しておこう:
φ = reiθ , p = Re(p) + iIm(p) . (I.8.13)
この分解によって、line element (I.8.11b), gauge symmetry/SUSY condition (I.8.12) は次のようになる:
ds2 =(dr)2 + r2(dθ)2 + k(dRe(p))2 + k(dIm(p))2
/(gauge symmetry + SUSY) , (I.8.14a)
40 N = (2, 2) Gauged Linear Sigma Model
gauge :
θ → θ − 2λ
Im(p)→ Im(p)− 2λSUSY : r2 + 2kRe(p) = 0 . (I.8.14b)
具体的に (I.8.14b) を (I.8.14a) に反映させよう:
ds2 = dr2 + r2(dθ − 2dλ
)2+ k(− r
kdr)2
+ k(dIm(p)− 2dλ
)2
=(1 +
r2
k
)dr2 + r2dθ2 + k(dIm(p))2
+ (r2 + k)
(−2dλ) +1
r2 + k
(r2dθ + k dIm(p)
)2
− 1
r2 + k
(r2dθ + k dIm(p)
)2
=(1 +
r2
k
)dr2 + r2
(1 +
r2
k
)−1
dθ2 + r2(1 +
r2
k
)−1
(dIm(p))2 − 2r2k
r2 + kdθ(dIm(p))
+ (r2 + k)
(−2dλ) +1
r2 + k
(r2dθ + k dIm(p)
)2
.
(I.8.15)
gauge symmetry があると、gauge fix をしなければならない。ここでは
Im(p) ≡ 0 (I.8.16)
に固定する。すなわち dIm(p) ≡ 0 に固定する。最後に、e′ → ∞ 極限操作、つまり gauge field を integrate
out する。これは (−2dλ) + · · · 2 を除去することに相当する。(Gauss 積分できるようにすでに平方完成させ
てある。)よって得られる cigar metric は次のようになる:
ds2 =(1 +
r2
k
)dr2 + r2
(1 +
r2
k
)−1
dθ2 (I.8.17)
• (N, k, q) = (2, 1, 0)→ (2, 1, 1): CP1 → sausage
CP1
sausage
√k
この sausage の計量を導出する作業は省略する。詳細は Hori and Kapustin [52] を参照。原理的な導出方法は
cigar と同じである。
Derivation of linear dilation
Hori-Kapustin [51] の Lagrangian (I.8.2) は、classical metric として cigar type を導出し、RG flow で 2-dim. black
hole = N = 2 Kazama-Suzuki model or SL(2,R)k/U(1) gauged WZW model に繋がると見ている。具体的に RG
flowを走らせて結論を得たというより、高い対称性の要請でこれしかない、という議論の仕方をするため、IR limitで実
現される CFTを squashed GLSMから直接導いたわけではない。そのため、当然ながら squashed GLSM Lagrangian
(I.8.2) から直接 linear dilaton field を導出したわけでもない。
ちなみに、2-dim. black hole と思っている SL(2,R)k/U(1) gauged WZW model の large k での metric の導出
及び dilaton field の導出は、Witten [110] に詳細があって、実際に簡単に再現できる。
I.8 Squashed GLSM 41
NS5-branes and noncompact CY
Hori-Kapustin [52] では、NS5-brane 近傍の物理を worldsheet description (SCFT) で理解しようと努めるものであ
る。NS5-brane を CP1 や CP2 に巻き付けたり、2-dim. spacetime を ADE type の minimal model で埋めて全体
で 10-dim. にしたりしている。ここで、例えば次のような spacetime contents の setup が、[52] の冒頭で紹介され
ている:
spacetime 0 1 2 3 4 5 6 7 8 9
NS5-branes © © © © © ©(throat) © ©
minimal model © ©
C(1)m =[SL(2,R)m+2
U(1)× SU(2)m−2
U(1)
]/Zm (I.8.18)
これは NS5-branes の transverse direction が Kazama-Suzuki model と minimal model の direct product で記述さ
れている状況を指す。
spacetime 0 1 2 3 4 5 6 7 8 9
NS5-branes © © © © © ©CP1 © ©
CY 2-fold(?) © © © ©minimal model © ©
C(2)m =[4d backgroundm ×
SU(2)m−2
U(1)
]/Zm (I.8.19)
これは NS5-branes が CP1 に twisted compactification という形で巻き付いている状態を指す。
spacetime 0 1 2 3 4 5 6 7 8 9
NS5-branes © © © © © ©CP2 © © © ©
CY 3-fold © © © © © ©minimal model © ©
C(3)m =[6d backgroundm ×
SU(2)m−2
U(1)
]/Zm (I.8.20)
これは NS5-branes が CP2 に twisted compactification という形で巻き付いている状態を指す。なお CP2 は CY
3-fold である O(−3) bundle on CP2 の中の projective space である。
42 N = (2, 2) Gauged Linear Sigma Model
Appendix
I.A General formulation
表記の注意をしておく。ここで扱っている superfield においての chiral/anti-chiral の表記と、holormophic/anti-
holomorphic (left-moving/right-moving) の表記が、CFT で扱っている表記とは互いに逆である。appendix III 参
照。Wess-Bagger notation [107, 113] では、chiral/anti-chiral を Φ/Φ で表記し、holomorphic/anti-holomorphic を
ψ+/ψ− で表記している。CFT ではその逆なのである。この点を、個々の文献を読む際には混乱しないようにしてお
きたい。
super covariant derivative の定義は次の通りである:
Dα =∂
∂θα+ i(σm)αα θ
α ∂
∂xm, Dα = − ∂
∂θα− i(σm)αα θ
α ∂
∂xm. (I.A.1)
今後ずっと登場する superfield を定義し、component field で展開する。
• chiral superfield
chiral superfield の定義は DαΦ = 0 である。component で展開すると
Φ(x, θ) = φ(y) +√
2θαψα(y) + θαθαF (y) , (I.A.2)
となる。但し ym = xm + iθα(σm)ααθα である。
• vector superfield
vector superfield の定義は V † = V で与えられる。Wess-Zumino gauge での component field は
V = −θα(σm)ααθα vm + iθαθαθαλ
α − iθαθα θαλα +
1
2θαθα θαθ
αD . (I.A.3)
である。
これらは D = 4, N = 1 superfieldを単純に dimensional reductionする事で得られる (mass dimension はもちろん変
化する)。これら以外に、D = 2 theoryで便利な spinor indexを与えよう。(ψ1, ψ2) = (ψ−, ψ+), (ψ1, ψ2) = (ψ−, ψ+)
と表記する。rank 2 anti-symmetric tensor εαβ (ε12 = ε21 = 1) なども同様に ε−+ = ε+− = 1 となる。これより、
ψ− = ψ+, ψ+ = −ψ− となる。これらの hermitian conjugate は (ψ−, ψ+)† = (ψ−, ψ+) = (ψ+,−ψ−) と定義される。
(ψ1, ψ2) = (ψ−, ψ+) (ψ1, ψ2) = (ψ−, ψ+)
ε12 = ε21 = 1 → ε−+ = ε+− = 1
ψ− = ψ+ ψ+ = −ψ−
(ψ−, ψ+)† = (ψ−, ψ+) = (ψ+,−ψ−)
この記号を用いて、D = 2 theory に特有の superfield を考察しよう。次の superfield の定義を行う。
I.A General formulation 43
• chiral superfield
chiral superfield の定義は
D±Φ = 0 , (I.A.4)
である。但し super covariant derivative は
D± =∂
∂θ±− iθ±
(∂0 ± ∂1
), D± = − ∂
∂θ±+ iθ±
(∂0 ± ∂1
), (I.A.5)
と書き直されている (for example, see ref [54])。component field で展開すると
Φ(x, θ) = φ(x) +√
2θ+ψ+(x) +√
2θ−ψ−(x) + 2θ+θ−F (x) + · · · , (I.A.6)
· · · は φ, ψ の微分項である。
• vector superfield
vector superfield の定義は V † = V で与えられる。D = 4 spacetime coordinates xm を D = 2 に reduction の
作業を行う。そのとき x0, x3 を新たに x0, x1 と再定義する20。この reduction に伴い、vector field vm も変更
を受ける。D = 4 の x0, x3 成分はそのまま残り、x1, x2 成分は scalar field となる:
σ =1√2(v1 − iv2) , σ =
1√2(v1 + iv2) . (I.A.7)
この約束の下で、vector superfield V のWess-Zumino gauge での component field は
V = θ+θ+(v0 + v1
)+ θ−θ−
(v0 − v1
)−√
2θ−θ+σ −√
2θ+θ−σ
− 2i θ+θ−(θ+λ+ + θ−λ−
)+ 2i θ+θ−
(θ+λ+ + θ−λ−
)− 2θ+θ−θ+θ−D ,
(I.A.8)
となる。
• twisted chiral superfield
twisted chiral superfield の定義は
D+Y = D−Y = 0 , (I.A.9)
である。twisted chiral superfield Y を component field で展開すると
Y (x, θ, θ) = y(x) +√
2θ+χ+(x) +√
2θ−χ−(x) + 2θ+θ−G(x) + · · · , (I.A.10)
である。· · · は y, χ の微分項である。
vector superfield V を twisted chiral superfield の記述で表そう:
Σ =1√2D+D−V = σ − i
√2 θ+λ+ − i
√2 θ−λ− +
√2 θ+θ−(D − iF01)
− iθ−θ− (∂0 − ∂1)σ − iθ+θ+ (∂0 + ∂1)σ +√
2 θ−θ+θ− (∂0 − ∂1)λ+
20dimensioanl reduction で残る D = 2 spacetime を D = 4 spacetime の x0, x3 と選んだ理由は、Pauli matrices σ0 = −12, σ3 が対角
的で扱いやすいからである。
44 N = (2, 2) Gauged Linear Sigma Model
+√
2 θ+θ−θ+ (∂0 + ∂1)λ− − θ+θ−θ−θ+ (∂02 − ∂1
2)σ . (I.A.11)
もちろん ∂0, ∂1 は D = 2 での spacetime derivative である。以後、xm という記号は D = 2 spacetime
coordinates である。また F01 = ∂0v1 − ∂1v0 である。つまり U(1) gauge field の field strength である。これ
より Σ はしばしば superfield for field strength と呼ばれる。
superspace part の measure についての規約を列挙する。
d2θ = −1
4dθα dθβ εαβ = −1
2dθ+ dθ− , d2θ = −1
4dθα dθβ ε
αβ =1
2dθ+ dθ− , (I.A.12a)
d2θ = −1
2dθ+ dθ− , d2θ = −1
2dθ− dθ+ , (I.A.12b)
d4θ = d2θ d2θ = −d2θ d2θ = −1
4dθ+ dθ− dθ+ dθ− . (I.A.12c)
よって superspace coordinates の積分は次で与えられる:
∫d2θ θθ = 1 ,
∫d2θ θθ = 1 ,
∫d2θ θ+θ− =
1
2,
∫d2θ θ−θ+ =
1
2. (I.A.13)
chiral superfield Φ、twisted superfield Σ のそれぞれの component fields に対する supersymmetry 変換を記載し
ておこう [107, 113]:
δv0 = iα−λ− + iα+λ+ + iα+λ+ + iα−λ− , δv1 = −iα−λ− + iα+λ+ + iα+λ+ − iα−λ− , (I.A.14a)
δσ = −i√
2α+λ− − i√
2α−λ+ , δσ = −i√
2α+λ− − i√
2α−λ+ , (I.A.14b)
δD = −α+(∂0 − ∂1)λ+ − α−(∂0 + ∂1)λ− + α+(∂0 − ∂1)λ+ + α−(∂0 + ∂1)λ− , (I.A.14c)
δλ+ = iα+D +√
2(∂0 + ∂1)σα− − F01α+ , δλ− = iα−D +√
2(∂0 − ∂1)σα+ + F01α− ,
δλ+ = −iα+D +√
2(∂0 + ∂1)σα− − F01α+ , δλ− = −iα−D +√
2(∂0 − ∂1)σ α+ + F01α− ,(I.A.14d)
δφ =√
2(α+ψ− − α−ψ+) , δφ = −√
2(α+ψ− − α−ψ+) , (I.A.14e)
δψ+ = i√
2(D0 +D1)φα− +√
2α+F − 2Qφσ α+ , δψ− = −i√
2(D0 −D1)φα+ +√
2α−F + 2Qφσ α− ,
δψ+ = −i√
2(D0 +D1)φα− +√
2α+F − 2Qφσ α+ , δψ− = i√
2(D0 −D1)φα+ +√
2α−F + 2Qφσ α− ,
(I.A.14f)
δF = −i√
2α+(D0 −D1)ψ+ − i√
2α−(D0 +D1)ψ−
+ 2Q(α+σψ− + α−σψ+) + 2iQφ(α−λ+ − α+λ−) ,(I.A.14g)
δF = −i√
2α+(D0 −D1)ψ+ − i√
2α−(D0 +D1)ψ−
− 2Q(α+σψ− + α−σψ+) + 2iQφ(α−λ+ − α+λ−) .(I.A.14h)
但し α±, α± は fermionic parameter であり、Q は chiral superfield Φ の U(1) charge である。また Dm は U(1)
gauge covariant derivative である。さらに χ−λ+ = λ+χ− = −χ−λ+ に注意する。なお、この表記は topological
twist をした後の BRST transformation rule を構成する時にも役に立つ。
I.B R-symmetry 45
I.B R-symmetry
I.B.1 Charge assignment
Supersymmetric theory には superspace coordinates θ の phase shift の global な U(1) 対称性を持つ。この対称
性を R-symmetry と呼ぶが、 D = 4, N = 1 もしくは D = 2, N = (2, 2) supersymmetric theory の場合、この
R-symmetry は vector U(1) symmetry (RV -symmetry) と axial vector U(1) symmetry (RA-symmetry) が存在す
る。それぞれ、次のような phase shift の下での対称性である:
RV -symmetry : Φ(θ±, θ±) → eiqV αΦ(e−iαθ±, e+iαθ±) (I.B.1a)
RA-symmetry : Φ(θ±, θ±) → eiqAαΦ(e∓iαθ±, e±iαθ±) (I.B.1b)
但し Φ(θ±, θ±)は一般的な superfieldであるとする。chiral superfieldである必要はない。superfieldの積も superfield
であるので、そういった物も含めてこの記号で表す。(qV , qA) は RV /RA-symmetry における superfield Φ の charge
である。superfield の変換則に対して、component fields の変換則が与えられる。
chiral superfield:
chiral superfield の展開は
Φ(θ±) = φ+√
2θ+ψ+ +√
2θ−ψ− + 2θ+θ−F , (I.B.2)
であるので、R-symmetry 変換 (I.B.1) の下で
Φ(θ±) → eiqV αΦ(e−iαθ±) = eiqV α(φ+√
2e−iαθ+ψ+ +√
2e−iαθ−ψ− + 2e−2iαθ+θ−F), (I.B.3)
となる。したがってこの RV -symmetry を component fields の変換則に読み換えると、
φ → eiqV αφ , ψ± → ei(qV −1)αψ± , F → ei(qV −2)αF , (I.B.4)
と assign されることになる。同様に RA-symmetry に対する変換則は
Φ(θ±) → eiqAαΦ(ei∓αθ±) = eiqAα(φ√
2e−iαθ+ψ+ +√
2e+iαθ−ψ− + 2θ+θ−F), (I.B.5)
であるので、component fields の変換則に読み換えると
φ → eiqAαφ , ψ± → ei(qA∓1)αψ± , F → F , (I.B.6)
となる。まとめると、chiral superfield の component fields に対して R-charge (QV , QA) が次のように与えられる:
chiral superfield Φ φ ψ+ ψ− F
(QV , QA) (qV , qA) (qV , qA) (qV − 1, qA − 1) (qV − 1, qA + 1) (qV − 2, qA)
twisted chiral superfield:
twisted chiral superfield は次のように展開される:
Y (θ+, θ−) = y +√
2θ+χ+ +√
2θ−χ− + 2θ+θ−G . (I.B.7)
46 N = (2, 2) Gauged Linear Sigma Model
従って R-charge は次のようになる:
twisted chiral superfield Y y χ+ χ− G
(QV , QA) (qV , qA) (qV , qA) (qV − 1, qA − 1) (qV + 1, qA − 1) (qV , qA − 2)
I.B.2 R-invariance
D = 2, N = (2, 2) SUSY theory を記述する最も一般的な Lagrangian は、Kahler potential K(Φ,Φ, Y, Y )、super-
potential W (Φ) and W (Φ)、twisted superpotential W (Y ) and W (Y ) を用いて次で与えられる:
L =
∫d4θK(Φ,Φ, Y, Y ) +
(∫d2θW (Φ) + h.c.
)+(∫
d2θ W (Y ) + h.c.)
= DK +(FW + FW
)+(GfW +GfW
).
(I.B.8)
但しここで DK , FW , GfW はそれぞれ K, W , W の highest component である。RV - もしくは RA-transformation
に対して Lagrangian が classical invariance を保つ条件を考えよう。chiral superfield Φ の R-transformation (I.B.1)
と同様に、一般的な superfield F(θ±, θ±) は次の様に変換される:
RV -symmetry :
F(θ±, θ±) → eiqV αF(e−iαθ±, e+iαθ±)
θ+θ− → e−2iαθ+θ− , θ+θ− → θ+θ− , θ+θ−θ+θ− → θ+θ−θ+θ−
RA-symmetry :
F(θ±, θ±) → eiqAαF(e∓iαθ±, e±iαθ±)
θ+θ− → θ+θ− , θ+θ− → e−2iαθ+θ− , θ+θ−θ+θ− → θ+θ−θ+θ−
これにより、Kahler potential K は、RV -/RA-symmetry両方に対して neutralであれば、DK は変換に対して不変に
なる。一方、W (Φ) の RV -/RA-charges を (qWV , qW
A ) とし、W (Y ) の charges を (qfWV , q
fWA ) とすると、R-invariance
であるには
(qWV , qW
A ) = (2, 0) , (qfWV , q
fWA ) = (0, 2)
と assign すれば良いことがわかる。
I.B.3 R-symmetry and U(1) current in N = (2, 2) SCFT
次の問題は、N = (2, 2) SCFT で登場する left/right U(1) current JL や JR を vector/axial vector R-symmetry
を用いてどのように与えるか、もしくは left/right の charge を R-charge を用いてどの様に assign するかである。
vector/axial vector R-symmetry の current をそれぞれ JV , JA とする。left/right U(1) current を JL, JR と記述す
る。これらを次のようにして結び付ける21:
JV = JL + JR , JA = −JL + JR . (I.B.9)
この定義は Silverstein-Witten [99] に基づいているが、Hori-Vafa [54]の R-symmetry に合致している。これを chiral
superfield Φと twisted chiral superfield Y について当てはめる。chiral superfield ΦのR-chargeを (qVφ , q
Aφ )、twisted
chiral superfield Y の R-charge を (qVy , q
Ay ) と置くと、component fields それぞれの R-charge (QV , QA)、そして
U(1) left/right current の charge (QL, QR) は次のようになる:21JA = JL − JR でも問題はないが、Hori-Vafa [54] の RA-symmetry と向きを合わせるためにこの符号を採用する。
I.C Weighted projective space 47
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Φ qVφ qA
φ12 (qV
φ − qAφ ) 1
2 (qVφ + qA
φ )
φ qVφ qA
φ12 (qV
φ − qAφ ) 1
2 (qVφ + qA
φ )
ψ+ qVφ − 1 qA
φ − 1 12 (qV
φ − qAφ ) 1
2 (qVφ + qA
φ − 2)
ψ− qVφ − 1 qA
φ + 1 12 (qV
φ − qAφ − 2) 1
2 (qVφ + qA
φ )
F qVφ − 2 qA
φ12 (qV
φ − qAφ − 2) 1
2 (qVφ + qA
φ − 2)
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Y qVy qA
y12 (qV
y − qAy ) 1
2 (qVy + qA
y )
y qVy qA
y12 (qV
y − qAy ) 1
2 (qVy + qA
y )
χ+ qVy − 1 qA
y − 1 12 (qV
y − qAy ) 1
2 (qVy + qA
y − 2)
χ− qVy + 1 qA
y − 1 12 (qV
y − qAy + 2) 1
2 (qVy + qA
y )
G qVy qA
y − 2 12 (qV
y − qAy + 2) 1
2 (qVy + qA
y − 2)
anti-chiral superfield Φ、twisted anti-chiral superfield Y の R-charge は、これらすべてが逆符号になる。
I.C Weighted projective space
weighted projective space WCP1ℓ,N−ℓ は、通常の projective space CP1 と若干の違いがある。
CP1 の定義は、原点を除く 2-dim. complex plane W = C2 − 0 の座標 (z1, z2) の比 (z1 : z2) が張る空間であ
る。つまり W 上の 2 点 (z1, z2) と (λz1, λz2) (ここで λ ∈ C∗) は同一視されている:
(z1, z2) ≃ (λz1, λz2) .
そのため、CP1 の 1 点を選ぶときは、λ を固定して、代表元を選べばよい。別の言い方をすれば、「W 空間 の C∗
gauge symmetry を固定して、1 点 (z1, z2) を選ぶ」となる。特に z1 = 0 もしくは z2 = 0 の点では、CP1 は C1 に
reduce する。
これに対して、WCP1ℓ,N−ℓ の定義は、W の座標 (z1, z2) の比の取り方に “weight” をつけることである:
(z1, z2) ≃ (λℓz1, λN−ℓz2) .
この同一視でWCP1ℓ,N−ℓ が定義されるが、この空間は CP1 にはなかった余分な対称性がまだ内蔵されている。C∗
による同一視を採るとき、λ の scale は固定されるが、phaseについては離散的な対称性が固定されない。具体的には、
λ = exp(2πi α
), α = GCMℓ,N − ℓ
は上の同一視では固定されない自由度となる:
(λℓz1, λN−ℓz2) = (z1, z2) .
これはつまり、WCP1ℓ,N−ℓ は一般に Zα orbifold symmetry を内蔵していることになる。また特に WCP1
ℓ,N−ℓ は、
z2 = 0 の領域は C1/Zℓ に、z1 = 0 の領域は C1/ZN−ℓ にそれぞれ reduce する。
Chapter II
N = (2, 2) Landau-Ginzburg Theory
The basis of this chapter is the Warner’s lecture [105].
50 N = (2, 2) Landau-Ginzburg Theory
II.1 Critical point
Let us first write down the N = (2, 2) supersymmetric action in terms of the chiral superfields Φi:
S =
∫d2zL , L =
∫d4θK(Φ,Φ) +
(∫d2θW (Φ) + (h.c.)
), (II.1.1)
where K(Φ,Φ) is a Kahler potential whose simplest form is
K(Φ,Φ) =
N∑
i=1
ΦiΦi (II.1.2)
which represents the free field theory kinetic term. Since the Kahler potential term receive irrelevant perturbation
in the RG flow (we discuss later), we find that it is not important to write down the concrete expression of the
Kahler potential in IR theory.
For our present purpose, the only information we need about the component action is that the purely bosonic
potential is
V (φ) = |∇W (φ)|2 . (II.1.3)
It follows that the critical points of superpotential W (i.e. points at which ∇W vanishes1) correspond to zero
energy minima of potential of the theory V , or to vacuum states of the theory (i.e., the moduli spaceM). Note, in
particular, that if there is more than one vacuum then all the vacua are degenerate. We therefore wish to consider
superpotentials W for which there is at least one critical point, and one can always take one of these critical points
to be at Φi = 0, i.e.:
∂W
∂Φi
∣∣∣Φ=0
= 0 , (II.1.4)
because one can always make the chiral superfields Φ → Φ + α and the superpotential W → W + ∆W shift in
terms of constant values α and ∆W . For multi-critical behavior of the corresponding statistical mechanical system
one obviously wishes to consider those potentials where several degenerate minima have come together. Thus one
is particularly interested in those theories for which
ρ|Φ=0 = deti,j
( ∂2W
∂Φi∂Φj
)∣∣∣Φ=0
= 0 . (II.1.5)
Note that ρ = det(∂i∂jW ) is the highest charged state in LG theory (see appendix C.1). For simplicity it is also
conventional to shift the superpotential so that
W |Φ=0 = 0 . (II.1.6)
We will also assume that the critical point of W at Φ = 0 is isolated. (This point is called the isolated singularity.)
That is, in a small neighborhood of Φ = 0 the equation
∂W
∂Φi= 0 (II.1.7a)
1The symbol ∇ simply means the derivative.
II.2 Renormalization group flow 51
implies
Φi = 0 . (II.1.7b)
In particular, this means that the potential has no flat directions at Φi = 0. The condition (II.1.7) is, in fact, a
little stronger than requiring that there be no flat directions because it implies that a generic perturbation of W
will yield a potential that has, in the neighorhood of Φ = 0, a finite number (µ) of critical points at which ρ does
not vanish. Physically, such a potential corresponds to a system where µ phases meet at the multi-critical point.
The number µ is called the multiplicity of the singularity of W at Φ = 0. One can see that µ is well-defined
and finite by observing that µ is equal to the winding number of the map
Φi
|Φi| 7→∂iW
|∂iW |(II.1.8)
considered as a map of S2N−1 ⊆ CN to S2N−1 ⊆ CN .
The assumption that the critical point of W is isolated has been made for convenience. There is no physical
reason to impose this condition. As we will see later, the restriction to isolated critical points does have some
extremely important consequences for the structure of the conformal theory.
II.2 Renormalization group flow
The conformal theory emerges from the statistical mechanical theory at its multi-critical point only when viewed
at large distance scales. This means that we must consider the renormalization group (RG) flow of the theory and
look for infrared (IR) fixed points. (It should be noted that when we are talking about the RG flows, we will always
mean flows in the IR limit as in Figure II.1.)
UV IRRG Flow
Figure II.1: Renormalization Group (RG) flow from UV to IR.
At such a fixed point, if one exists, we will, by definition, have a conformally invariant theory. We will always
assume that for a given LG theory there is an IR fixed point, and therefore the action (II.1.1) may be viewed, in
this sense, as defining an N = (2, 2) SCFT (see appendix III). The only problem is that the action after the RG
flow may look nothing like the one with which we started.
Suppose a new LG action S′, that is a small perturbation of S by some (scalar) operator O, i.e.,
S′ = S + ǫ
∫d2zO(z, z) . (II.2.1)
The operator O is called irrelevant if this theory flows to the same fixed point as S, and the operator O is relevant
if it flows to a new fixed point. The operator O is called marginal if the theory flows to a new fixed point that
can be continuously connected to the old fixed point by shrinking ǫ to 0. If there are marginal operators then the
52 N = (2, 2) Landau-Ginzburg Theory
Irrelevant
Relevant
IR Fixed Point
Marginal
Figure II.2: RG flow around the IR fixed point.
fixed “point” is really a fixed “surface” defined by all the marginal operators of the theory. The image of the RG
flows is drawn in Figure II.2.
At a fixed point, where we have a conformal field theory, the question of whether an operator is relevant,
irrelevant or marginal is determined entirely by its conformal dimension [115]. Recall that the conformal (or
scaling) dimensions d of a field is the sum of the left-moving conformal weight h, and the right-moving conformal
weight h. Moreover, the spin s of a field is equal to the difference s = h − h of the conformal weight, and thus
scalar fields have h = h. The relevant, irrelevant and marginal operators are those scalars of conformal dimension
less than 2, greater than 2 and equal to 2, respectively. Note that the conformal dimensions of z and θ are −1 and
− 12 , and that the conformal dimensions of dz and dθ are −1 and + 1
2 .
conformal dimensions d = h+ h
spin s = h− h
relevant d < 2
irrelevant d > 2
marginal d = 2
z d = −1
θ d = − 12
dz d = −1
dθ d = + 12
Hence if the lowest component of a general superfield has conformal weights (h, h), then the highest component has
weight (h + 1, h + 1). When the conformal weight of the Kahler potential K(Φ,Φ) is (hK , hK) and its conformal
dimension is dK = hK +hK , the highest component of K(Φ,Φ) (or equivalently the term∫
d4θK(Φ,Φ), called the
D-term) has conformal weights (hK + 1, hK + 1) and its conformal dimension is dD = hK + hK + 2. However, in a
unitary CFT we have h, h ≥ 0 and thus the D-term will be marginal or irrelevant (hK + hK + 2 ≥ 2). In addition,
when the superpotential term∫
d2θW (Φ) (called F-term) is the relevant (or marginal) operator (dF ≤ 2), the
lowest component of W has conformal dimension dW less than (or equal to) 1:
K(Φ,Φ) : dK = hK + hK →∫
d4θK(Φ,Φ) : dD = hK + hK + 2
in unitary CFT (h, h ≥ 0) →∫
d4θK(Φ,Φ) is irrelevant or marginal (dD ≥ 2)
II.2 Renormalization group flow 53
W (Φ) : dW = hW + hW →∫
d2θW (Φ) : dF = (hW + 1) + hW
relevant or marginal
∫d2θW : dF ≤ 2 → hW + hW ≤ 1
Notice that the conformal weight of superfield is equal to one of its lowest component field.
The important point is that W (Φ) is the only relevant object for the RG flow. The kinetic term∫
d4θK is
irrelevant. So we choose some starting value for K(Φ,Φ) such as (II.1.2), and K will flow to whatever value it
need to have at the fixed point. The final value will not depend on the starting value2. Notice that we do not
impose that the conformal dimension of the Kahler potential term is equal to that of the superpotential term. If those
terms are equal conformal dimensions (more than two), they both are irrelevant operators and the IR theory becomes a
trivial CFT.
We will now make a major assumption. We will assume that the non-renormalization theorem hold. In other
words, we assume that the F-terms do not renormalize except for wave-function renormalization. These theorems
are known to hold in perturbation theory in the UV limit. However we really need to apply them in the IR limit,
and to the block-averaging renormalization of statistical mechanics. One would also like some guarantee that the
theorems are also true non-perturbatively. While little is known about the non-perturbative quantum field theory,
it seems likely that the non-renormalization theorems should still be valid in the IR limit and also for the effective
action obtained by the block-averaging procedure. Ultimately, perhaps the best justification for assuming that the
F-terms do not renormalize will be the results that we will derive based upon this assumption.
So we assume that in the effective action, the F-terms are not renormalized. This means, that up to wave-
function renormalizationm the superpotential W is invariant. Thus it is an invariant, and as we will see, a truly
topological characteristic of the flow3. Moreover, it contains the only relevant operators and hence it also dictates
the flow. Since the kinetic terms are irrelevant, the fixed points can thus only depend on W , and W is also the
superpotential at the fixed point as well as at the starting point of the flow.
To classify all fixed points of LG theories, we therefore only need classify all the possible superpotential W .
The superpotential is an analytic function. Clearly the multi-critical structure does not change under an analytic
(and, of course, invertible) change of variables, i.e., under field re-definitions. Moreover, the field re-definition in K
is simply an irrelevant perturbation of the original theory. Thus the LG theory will flow to the same fixed point
by making a field re-definition and we find that it is sufficient to classify W up to analytic changes of variables.
Next we make an important technical observation. If the potential of the LG theory is such that every field has
a nonzero mass, then the RG flow leads to a trivial CFT, i.e., one with central charge c = 0, or equivalently, one
whose Hilbert space is merely the complex numbers. The reason is simple: If there is a mass term, the two point
2This analysis is only really true in the neighborhood of a fixed point. It may be that the space of possible Kahler potentials K
actually decomposes into distinct regions, in each of which the RG flow goes to some distinct IR fixed point. Thus large changes in K
may be relevant. Therefore, K is only really irrelevant if there is an unique fixed point. In this chapter we shall, for simplicity, assume
that there is an unique fixed point. However we will not, in fact, use this assumption in deducing the properties of the SCFT that
appears at an IR fixed point. Consequently, these properties will hold at all the fixed points that can be obtained from a given W .3意味がわからない (2004 2/08)。
54 N = (2, 2) Landau-Ginzburg Theory
function 〈Φ(z)Φ(0)〉 behaves asymptotically as e−m|z| for large |z|. Thus, in the IR limit the correlation functions
are simply δ functions4 — the theory is trivial. In other words, the particles freeze out when the mass scale of
the theory becomes much less than the particle masses5. This means that given (m + n) fields (Φi,ΦA)
(i = 1, 2, · · · ,m and A = 1, 2, · · · , n) and a superpotential W (Φi,ΦA) of the form
W (Φi,ΦA) = W1(Φi) +W2(Φ
A) +∑
A
mA(ΦA)2 , (II.2.2)
then the CFT corresponding to W1 and W are identical6. Thus we can not only change variables, but also change
dimensions by adding quadratic pieces to W17.
Therefore we want to classify all superpotentials W with the multi-critical structures described earlier, up to
the addition of extra “quadratic dimensions” and changes of variables8 [5]. (The word “stable” refers to
the possibility of changing the dimension by addition of quadratic terms.)
The non-renormalization theorems tell us that W is invariant up to wave function renormalization, whereas
the F-term is unrenormalized. Under scaling z → λ−1z, we have θ → λ−1/2θ, dz → λ−1dz, dθ → λ+1/2dθ and thus∫
d2zd2θ → λ−1∫
d2xd2θ. Hence W (Φ) must scale according to
W (Φ) → λW (Φ) (II.2.3)
at the fixed point. However, wave function renormalization means that
Φi → λωiΦi (II.2.4)
for some numbers ωi. The numbers ωi are called the anomalous dimensions of Φi. Since Φi are scalars, the
conformal weights of Φi are given by hi = hi = ωi
2 .
z → λ−1z , dz → λ−1z , θ → λ−1/2θ , dθ → λ+1/2dθ
∴ d2zd2θ → λ−2+1d2zd2θ∫d2zd2θW (Φ) should be invariant under rescaling at fixed points
∴ W (Φ) → λ+1W (Φ)
further, we set rescaling factor of scalar Φi as Φi → λωiΦi, hi = hi =ωi
2
For a generic superpotential W (Φ) with some power expansion about Φi = 0, the only way to reconcile (II.2.3)
and (II.2.4) is if the coefficients of the power series are viewed as coupling constants that also rescale so that
when this scaling is combined with (II.2.4) one obtains (II.2.3). For example, given W (Φ) = g1Φ3 + g2Φ
5 with
Φ→ λωΦ, we must have g1 → λ1−3ωg1 and g2 → λ1−5ωg2. If all the conformal dimensions are strictly positive,
4δ function? asymptotically zero?5観測している energy scale では particle は重過ぎて dynamics を持てない。6長距離極限 (IR limit) では interaction が効かない (クラスター性) という仮定であるため、 W2(ΦA) は decouple する。7変数の数を変えるのは理解できるが、mass terms が付加されることで conformal dimensions が変更された fields は IR では freeze out す
るから、解析には寄与するのか? interaction を通じて、mass terms が付加されていない fields の conformal dimensions を変更するというこ
とか?8mass terms を付加するくらいで変化する superpotential をひとつの class としてまとめて考える。
II.2 Renormalization group flow 55
it follows that the lowest order terms in the power series of W (Φ) will dominate as we approach the critical point.
At the critical point the F-term must be scale invariant (because of CFT), and in the IR limit this means that all
the higher order terms must vanishes, and the scaling dimensions of the fields must be such that the lowest
order part W0 of the superpotential W satisfies
W0(λωiΦi) = λ+1W0(Φ
i) (II.2.5)
with no scaling of the remaining coupling constant9. Let us consider the above example. If we take λ→ 0 and let
higher order term be faded out (i.e., g2/g1 → 0) in the IR limit, then we find that ω = 1/3, g2 ≡ 0 and g1 does not
rescale.
Warner [105] では λ→∞ が IR limit であるかのように記述されているが、それは間違いである。coordinate
rescaling が z → λ−1z のように定義されているので、IR limit (long distance limit) は λ → 0 である。
z → λ−1z の rescale の下で、Kahler potential term (D-term)∫
d2zd4θK と superpotential term (F-term)∫
d2zd2θW はそれぞれ
∫d2z d4θK → λωK
∫d2z d4θK , ωK : conformal dim. of K , 0 ≤ ωK
∫d2zd2θW → λωW −1
∫d2z d2θW , ωW : conformal dim. of W , 0 ≤ ωW ≤ 1
と rescale されるが、先の議論にあったように前者は irrelevant or marginal、後者は relevant or marginal であ
ることが、λ→ 0 limitで dominateされるのか否かで理解できる。つまり [105]の記述 (IR limit⇔ “λ→∞”)
は間違いであり、IR limit ⇔ “λ→ 0” であると定義すれば、Figure II.2 の意味も納得できる。
In general, at the conformal fixed point, the superpotential W of the theory must satisfy (II.2.5) (without
rescaling of the coupling constants) for some choice of ωi. Such a superpotential is called homogeneous with
weight ωi. Note that these weights are precisely the anomalous dimensions of Φi at the conformal fixed point. ωi
are rational numbers and they are obviously completely determined by the form of W0(Φi), provided that W0
has an isolated critical point at Φi = 0.
The notion of the “lowest order part” W0 of W is, in general, ambiguious. These difficulties may give rise
to indeterminacies in the RG flow — there could be several fixed points or fixed surface to which the theory
might flow. It is conceivable that there might even be strange attractors. On a more down-to-earth level, the
presence od flat directions in W0 could reflect some kind of spontaneous ‘decompactification’ of the theory.
While the consideration of these possibilities might contain some extremely interesting physics10, we shall
take the simpler course and remove the problem by requiring that the starting superpotential W (Φi) be
semi-quasi-homogeneous. In other words, we require that W has the form
W = W0 +W1
where W0 is quasi-homogeneous, and has a single, isolated multi-critical point at Φi = 0, and W1 only contains
terms of scaling dimension strictly greater than one11. Finally one should observe that at the conformal fixed point,9CFT!
10One of the examples is, maybe, the dual LG theory of the sigma model on line bundle on QN−2.11The scaling dimensions of the fields Φi and hence of ther terms in W1 are determined by W0.
56 N = (2, 2) Landau-Ginzburg Theory
W1 scales away leaving us with W0. Therefore we would have obtained precisely the same result if we had started
with W0 in the first place.
As a consequence of the foregoing considerations, we will, from now on, only consider a quasi-homogeneous
superpotential W satisfying (II.1.4), (II.1.5), (II.1.6), (II.1.7) and
W (λωiΦi) = λ+1W (Φi) . (II.2.6)
Observe that this fixes the conformal weights of the fields Φi to be
hi = hi =ωi
2> 0 (II.2.7)
and these are all strictly positive. This means that the D-term of (II.1.1) can only contain terms of conformal
dimension strictly greater than 2, which are irrelevant. There are thus no marginal operators in the D-terms of
the LG theory:
— well-defined LG theory —
L =
∫d4θK(Φ,Φ) +
(∫d2θW (Φ) + (h.c.)
)
∫d4θK is an irrelevant operator
W |Φ=0 =∂W
∂Φi
∣∣∣Φ=0
= 0 ⇔ existence of the critical point at Φi = 0
deti,j
( ∂2W
∂Φi∂Φj
)∣∣∣Φ=0
= 0
ρ ≡ deti,j
( ∂2W
∂Φi∂Φj
)6= 0 ⇔ existence of isolated singularity
W (λωiΦi) = λ+1W (Φi) where hi = hi =ωi
2> 0
This restriction to quasi-homogeneous superpotentials simplifies things considerably, but it turns out that
it is not as severe a restriction as one might expect. Many of the stable singularity types have quasi-
homogeneous (or semi-quasi-homogeneous) representatives. The whole point of the above discussion is
to point out that we are explicitly excluding a number of potentially interesting LG theories that do
not necessarily have well-defined or unique fixed points (or surfaces). Some of the excluded class are very
interesting from the point of view of singularity theory.
II.3 Central charges of CFT
Given a superpotential W satisfying (II.1.4), (II.1.5), (II.1.6), (II.1.7) and (II.2.6), we give a central charge c of
CFT at the fixed point of the RG flow (see also chapter III and appendix C):
c = 6β , β =
N∑
i=1
(1
2− ωi
). (II.3.1)
II.4 Chiral ring 57
The quantity β is called the singularity index of W . For a general superpotential W , the singularity index is
defined by the asymptotic expansion of the integral
∫
RN
dNx ei~
W (xi) ∼ ~( N2−β) (II.3.2)
as ~→ 0. By using this very suggestive definition one can make a path integral argument to establish that c = 6β.
There was a conjecture in singularity theory that the singularity index is upper semi-continuous [4]
in the sense that if one is given a function f whose singularity index is βf , then there is a small
neighborhood of the function f (in the space of all analytic functions) such that βf is the maximum
value of the singularity index on that neighborhood of f . That is, β is a locally non-increasing function
on the space of analytic functions. Given the identification (II.3.1) between β and c, we see that the
foregoing conjecture is merely a statement of the Zamolodchikov c-theorem [115]. The surprising
thing is that while the upper semi-continuity of β is valid for a broad class of singularities, there are
counterexamples to the conjecture [5]. It is not believed that this means that there are counterexamples
to the Zamolodchikov c-theorem. The counterexamples to the upper semi-continuity of β are not semi-
quasi-homogeneous, and so the lack of upper semi-continuity presumably reflects an ambiuity in the
conformal fixed point. It would be interesting to investigate this further.
II.4 Chiral ring
There is another extremely useful and natural structure that one introduces in singularity theory. It is the notion
of the local ring of W . Consider the ring of power series of functions expanded around Φi = 0. Call this ring P.
Let J denote the ideal of this ring generated by all the first partial derivative ∂W
∂Φi . The ideal J consists of all
linear combinations of power series that have one (or more) factors of one of the ∂W∂Φi . The local ring R of W is
defined by
R = P/J (II.4.1)
and it consists of all power series taken modulo the partial derivatives ∂W∂Φi . The dimension of R is finite and is, in
fact, equal to the multiplicity µ of the singularity:
dim R = µ .
As an example, let us consider
W (x, y) = x3 + y5 .
This is quasi-homogeneous with weights ωx = 1/3 and ωy = 1/5. The partial derivatives are
∂W
∂x= 3x2 ,
∂W
∂y= 5y4 .
58 N = (2, 2) Landau-Ginzburg Theory
Hence the ring R consists of all polynomials taken modulo x2 and y4, and is equal to all linear combination of
1, x, y, y2, y3, xy, xy2, xy3
,
which is an eight dimensional ring.
Physically the local ring R of a superpotential W corresponds to what is called the chiral ring of the SCFT.
In CFT all the chiral fields of the theory must be obtainable from polynomials in the Φi (i.e., the elements of P).
The partial derivatives ∂W/∂Φi are, by the equations of motion, proportional to some superderivatives acting on
some combination of Φi. Thus ∂W/∂Φi are polynomials in Φi that represent descendant fields12. Moreover the
set of ∂W/∂Φi and their products with all polynomials in Φi generate all polynomials of Φi that are descendant
fields. Hence R is precisely the ring of chiral primary fields in SCFT. We will see later that the operator product
algebra of chiral primary fields is in fact the polynomial algebra of R. That is, there is no powers of z by which one
needs to rescale, and the operator product vanishes if and only if the leading term is a descendant field. In other
words, the local ring of W and the ring of chiral primary fields are isomorphic as rings. As a result of this, one
can conclude that the conformal dimension of any monomial in R is simply the sum of the conformal dimensions
of the fields in the monomial. Thus, in the above example we know that hx = 12ωx = 1/6, hy = 1
2ωy = 1/10. The
conformal dimensions of the monomials of R are13
field dimension field dimension
1 030 y3 18
30
x 1030 xy 16
30
y 630 xy2 22
30
y2 1230 xy3 28
30
0BBBBBBBB@
W (x, y) = x3 + y5
W (λωiΦi) = λW (Φi) → ωx =1
3, ωy =
1
5
∂xW = 3x2 , ∂yW = 5y4
R = 1, x, y, y2, y3, xy, xy2, xy3
1CCCCCCCCA
Note that the dinominator 30 is the Coxeter number of E8 (for the dual Coxeter number g∨, see chapter 13.1 in
[19]), while the numerators are the degrees of the Casimirs of E8 minus two, and that the dimension of the ring 8,
which is the rank of E8. This is not accidental.
II.5 Three significant concepts
In order to apply the results of singularity theory, we need to introduce three other concepts: versal deformations,
modality and the Poincare polynomials.
II.5.1 Versal deformation
Given an analytic function f : U → C where U is some neighborhood of the origin in Cn, then an analytic function
F : U × V → C is said to be a versal deformation of f if14
12For present purposes, a combination of fields will be thought of as a descendant if it can be rewritten as some other combination
of fields that involves derivatives or superderivatives. This definition will coincide with the usual one at the conformal fixed point.13Remember that since the superfields are scalars, we have h = h.14分かったような分からんような。
II.5 Three significant concepts 59
1. V is some neighborhood of the origin in Cm with m <∞.
2. F (x, 0) = f(x), for x ∈ U .
3. There is a small neighborhood F of f in the space of analytic functions on neighborhoods of the origin in Cn,
such that for every g ∈ F there is an analytic change of variable y(x) such that for x in some neighborhood
of the origin in Cn we have
g(y(x)) = F (x, λ) (II.5.1)
for some value of λ near the origin in Cm.
In other words, in suitably small neighborhoods of the origin in Cn, every analytic function near f can be obtained
by analytic changes of variable of F (x, λ) for some value of λ.
If f has a singularity of multiplicity µ at the origin15, then it can be shown (the versality theorem [5]) that the
following is a versal deformation:
F : U × Cµ → C ; F (x, λ) = f(x) +
µ∑
ν=1
λνrν(x) , (II.5.2)
where rν(x) are bases for the local ring of f . In other words, the local ring provides a universal class of deformations
of f(x).
It is also useful to note that for general values of the parameters λν , the function F (x, λ) morsifies16 f(x). That
is, F (x, λ) completely resolves the singularity of f into µ distinct critical points17 at which detij(∂2F
∂xi∂xj ) = 0. It
turns out that the foregoing deformation of miniversal. That is, for a function of multiplicity µ, a versal deformation
must always have at least µ parameters, and thus the deformation (II.5.2) has the minimum number of possible
parameters.
II.5.2 Modality
The question naturally arises as to whether one requires a continuum or discrete set of values of the parameters λ so
as to get to everything in the neighborhood of f via a change of variable. A function f is said to have a singularity
of modality zero if every function in the neighborhood of f is coordinate transformation equivalent to a versal
deformation F (x, λ) of f(x) for a finite set of values of λ. This means that only a finite set of choices of λ are
needed to cover a small neighborhood F of f via the action of changes of variables. Note that no matter how small
we make F , a function of modality zero always has a finite set of choices for λ that enable one to cover F by the
orbits of the corresponding finite set of functions F (x, λ) under the diffeomorphism group18.
A function f has a critical point of modality m if F can be covered by the action of the diffeomorphism group
on F (x, λ) for finitely many m parameter families of deformation parameter λ, and m is the least such number of
parameters. These parameters are called modal parameters, or moduli, of the singularity.15multiplicity µ of W の値は、chiral ring R = C/[dW ] の次元 µ = dim R しか導出方法がないのであろうか。16“morsify” とはどういう意味だ?17Technically, a morsification of f(x) must also take different values at each of its critical points.18何を言っているのかわからない。(2004 3/04)
60 N = (2, 2) Landau-Ginzburg Theory
II.5.3 Poincare polynomial
Suppose that we have a quasi-homogeneous function W of n variables Φj with weights ωj . Let pj and N be (non-
negative) integers such that ωj = pj/N . For definiteness, let N and pj be reduced so that they have no overall
common factor. The Poincare polynomial of W is defined to be
P (t) =∞∑
k=1
nk tk , (II.5.3)
where nk is the number of monomials of scaling dimension k/N in a basis for the local ring R of W . It is relatively
easy to see that [5]
P (t) =
n∏
j=1
1− tN−pj
1− tpj
. (II.5.4)
The proof is simply to observe that the terms in the denominator generate the partition function of the free
polynomial algebra in Φj , while the numerator subtracts out the contribution of everything generated by ∂W∂Φj
(whose scaling dimension isN−pj
N ).
There are two immediate consequences of the foregoing result. First, from the definition of P (t) we see that
µ = P (1), but, applying l’Hopital’s rule in (II.5.4) we get
µ =
n∏
j=1
( 1
ωj− 1), (II.5.5)
where ωj = pj/N are the conformal weights of Φj . Now consider the limit as t→∞. From (II.5.4) it follows that
the highest power of t in P (t) must be tM where
M =
n∑
j=1
(N − 2pj) = N
n∑
j=1
(1− 2ωj) = 2Nβ , (II.5.6)
and moreover the coefficient of this term in P (t) is 1. Therefore there is an unique element ρ of maximal scaling
dimension in the local ring R of W , and the maximal value of the scaling dimension is 2β. Thus ρ must have
conformal weight:
h = h = β =n∑
j=1
(1
2− ωj
). (II.5.7)
One can also see that ρ must be the polynomial:
ρ(Φ) = detij
( ∂2W
∂Φi∂Φj
). (II.5.8)
Observe that P (t) has a (Poincare) duality:
P (t) = t2NβP (1/t) . (II.5.9)
This follows trivially from (II.5.4). Under this duality the state ρ interchanges with the unique state with h = h = 0,
that is, with the vacuum. From the viewpoint of N = (2, 2) SCFT, this duality is interpreted as the result of the
spectral flow.
II.6 Modality zero singularities: ADE classification 61
II.6 Modality zero singularities: ADE classification
Finally, it is interesting to note that given an arbitrary set of integers pj , and an integerN , then a quasi-homogeneous
function with weights ωj = pj/N can only exist if the function P (t) defined by (II.5.4) is a (finite) polynomial.
If (II.5.4) is a finite polynomial then such a quasi-homogeneous function still might not exist, but the interesting
observation here is that the set of weights of a quasi-homogeneous function are, in general, a highly restricted set
of rational numbers.
With this basic introduction to singularity theory, we are in a position to quote some of the theorems. All
singularities with modalitiy ≤ 2 have been classified up to stable equivalence [5]. All modality zero singularities
are stably equivalent to one of the following [5]:
type superpotential W central charge
Ak xk+1 c = 3− 6k+1 k ≥ 1
Dk xk−1 + xy2 c = 3− 62(k−1) k ≥ 3
E6 x3 + y4 c = 3− 612
E7 x3 + xy3 c = 3− 618
E8 x3 + y5 c = 3− 630
Table II.1: LG superpotentials with modality zero singularities.
where c is the central charge of the corresponding CFT, and has been computed using (II.3.1). Note that the
labelling of the singularities with the labels of the simply-laced, finite Lie algebra is no accident. First observe
that the central charge is given by c = 3− 6N where N is the Coxeter number of the algebra. One can check that
the local ring has dimension ℓ, where ℓ is the rank of the algebra, and there is a basis for the local ring in which the
conformal dimensions of the fields are di−2N , where di are the degrees of the Casimirs of the Lie algebra19. These
coincidences are not mysterious numerology. There is a close connection between the topology of the bifurcation
varieties of these functions and the braid group of the Weyl group of the corresponding Lie group. On a more down
to earth level, suppose W is some generic versal deformation of W , and one tries to find integral curves of ∇W ,
i.e., solutions of
dxi
dt=
∂W
∂xi(II.6.1)
and suppose one seeks solutions running between the critical points of W , i.e., the points which satisfy ∂W = 0.
Represent the existence of such curves by drawing a dot for each critcal point and a bond for each integral curve.
Then up to some mathematical niceties, the generic result is the Dynkin diagram for each Lie algebra20. These
integral curves have a beautiful physical interpretation. Observe that if (II.6.1) is satisfied then
d2xi
dt2=∑
j
(∂j(∂iW )
)dxj
dt=∑
j
(∂j∂iW )(∂jW ) =1
2∂i
[|∇W |2
].
19Furthermore we can identify the number “di − 1” with the “exponent” of the Lie algebra.20まだ具体的に AN series すら書けない (2004 4/14)。例が詳細に記述されている文献はないだろうか。
62 N = (2, 2) Landau-Ginzburg Theory
Now recall that V = |∇W |2 is the bosonic potential of the theory and that these integral curves represent rolling
motions on the inverted bosonic potential well. They they represent semi-classical tunelling solitions of the
theory. The Dynkin diagram thus tells us about the vacuum states and the tunelling solutions.
The list of modality zero singularities (Table II.1) corresponds precisely with the list of all the modular invariant
N = (2, 2) superconformal discrete series models [91, 28, 92, 29, 31]. Indeed, the discrete series models are known
to have c = 3− 6N where N is the Coxeter number of a simply-laced Lie algebra. In general the chiral fields in the
theory have
h = h =j
2N, j = 0, 1, 2, · · · , N − 2 .
Such theories corresponds to the AN−1 series discussed above. For N even, there is a Z2 orbifold of these AN−1
theories, which gives rise to the DN2−1 series. There are also three exceptional models in which many more fields
have been projected out, and which give rise to the E6, E7 and E8 theories.
There are several important things to note:
1. Modular invariance of the LG models is guaranteed; there is no chiral splitting and there is always left-right
symmetry and thus no global worldsheet anomalies can appear.
2. Given a quasihomogeneous superpotential W (Φi) with weight ωi, write the weights in the form ωi = pi/N
where pi and N are integers as above. It is clear that the LG theory has a ZN symmetry21
Φi → e2πipi/NΦi , Φi → e−2πipi/NΦi . (II.6.2)
For the discrete series models this symmetry has been discussed in [91, 92, 31].
3. Table II.1 strongly suggests isomorphisms
E6 ≡ A2 ⊗A3 , E8 ≡ A2 ⊗A4 , (II.6.3)
and indeed this is true up to ZN twist.
4. There are many, many more singularities beyond the modality zero singularities described in Table II.1. There
are vast tables up to modality 2, and of course it is elementary to write down higher modality potentials. For
example, the singularity of the function
W = x3 + y3k+1 + xy2k+1( k−2∑
j=0
ajyj)
has multiplicity µ = 6k, and modality k−1. (The coefficients ai are the modal parameters.) Some singularities
give rise to CFT that are obviously tensor products of simpler ones such as the discrete series, while other
singularities correspond to completely new CFT. For more recent results on this, see [72, 30].
21If one considers the action of this ZN on the Ramond sector, then this symmetry may well become a Z2N symmetry.
II.6 Modality zero singularities: ADE classification 63
There are several major virtues of the perspective that the application of singularity theory affords one. First it
yields new models. Secondly, the ring of chiral, primary fields captures some simple intrinsic structure of the theory
that tells one when two theories can or cannot be isomorphic. The example of E8 ≃ A2 ⊗ A4 is typical. Thus it
seems hopeful that one might use the structure of the chiral ring to classify the N = (2, 2) SCFT. The problem with
such a programme is that not all N = (2, 2) SCFT correspond to LG theories. The simplest counterexample is a
complex boson and a complex fermion compactified on a 2-torus. This is an N = (2, 2) theory, but has no LG from
since the critical the chiral fields are not left-right symmetric. There are many other counterexamples. However, it
turns out that in many of these counterexamples one can turn the theory into a LG theory by choosing the moduli
appropriately and dividing by a discrete symmetry. A generic property of many of the known counterexamples
is that the chiral ring is too small and the number of moduli os too large. Division by discrete symmetries
usually freezes out some of the moduli and introduces twist fields that fill out the chiral ring. So one might hope
that if one twists a theory sufficiently then it will have a LG form. It is not known whether this is true in general.
64 N = (2, 2) Landau-Ginzburg Theory
Appendix
II.A Data for ADE
Ak type singularity: (level = k − 1)
W = xk+1 , k ≥ 1
ωx = 1k+1
px = 1 , N = k + 1∂W∂x = (k + 1)xk
R = 1, x, x2, · · · , xk−1µ =
(1
ωx− 1)
= k = dimR
β = 12 − ωx = 1
2 − 1k+1
c = 6β = 3− 6k+1
field dimension
1 0k+1
x 1k+1
......
xk−1 k−1k+1
Dk type singularity: (if k = 2j + 2, level = 4j; if k = 2j + 1, level = 4j − 2)
W = xk−1 + xy2 , k ≥ 3
ωx = 1k−1 , ωy = k−2
2(k−1)
px = 2 , py = k − 2 , N = 2(k − 1)∂W∂x = (k − 1)xk−2 , ∂W
∂y = 2xy
R = 1, x, x2, · · · , xk−2, yµ =
∏i=x,y
(1ωi− 1)
= k = dim R
β =∑
i=x,y
(12 − ωi
)= 1
2 − 12(k−1)
c = 6β = 3− 62(k−1)
field dimension field dimension
1 02(k−1) xk−3 2(k−3)
2(k−1)
x 22(k−1) xk−2 2(k−2)
2(k−1)
...... y k−2
2(k−1)
E6 type singularity: (level = 10)
W = x3 + y4
ωx = 13 , ωy = 1
4
px = 4 , py = 3 , N = 12∂W∂x = 3x2 , ∂W
∂y = 4y3
R = 1, x, y, y2, xy, xy2µ =
∏i=x,y
(1ωi− 1)
= 6 = dimR
β =∑
i=x,y
(12 − ωi
)= 1
2 − 112
c = 6β = 3− 612
field dimension field dimension
1 012 y2 6
12
x 412 xy 7
12
y 312 xy2 10
12
II.B Highest charged state 65
E7 type singularity: (level = 16)
W = x3 + xy3
ωx = 13 , ωy = 2
9
px = 6 , py = 4 , N = 18∂W∂x = 3x2 + y3 , ∂W
∂y = 3xy2
R = 1, x, x2, y, y2, xy, x2yµ =
∏i=x,y
(1ωi− 1)
= 7 = dim R
β =∑
i=x,y
(12 − ωi
)= 1
2 − 118
c = 6β = 3− 618
field dimension field dimension
1 018 y2 8
18
x 618 xy 10
18
x2 1218 x2y 16
18
y 418
E8 type singularity: (level = 28)
W = x3 + y5
ωx = 13 , ωy = 1
5
px = 10 , py = 6 , N = 30∂W∂x = 3x2 , ∂W
∂y = 5y4
R = 1, x, y, y2, y3, xy, xy2, xy3µ =
∏i=x,y
(1ωi− 1)
= 8 = dimR
β =∑
i=x,y
(12 − ωi
)= 1
2 − 130
c = 6β = 3− 630
field dimension field dimension
1 030 y3 18
30
x 1030 xy 16
30
y 630 xy2 22
30
y2 1230 xy3 28
30
II.B Highest charged state
ここでは LG superpotential から読み取れる “highest charged state” ρ の具体的な形を見よう。定義は appendix C.1
にある。なお、この highest charged state がゼロにならないことと、LG superpotential に isolated singularity があ
ることとは同義である、と Lerche, Vafa and Warner は述べている (see [72], p.446)。これは compact manifolds につ
いてのみあてはまるのか、一般的に通用するのかはまだよく把握していない (2004 5/01)。ただ、LG superpotential
W に isolated singularity が存在することは、LG theory の IR limit が (direct products of) minimal models、つ
まり Gepner model で記述できることと同義であり、また LG superpotential から dual な CY geometry の Hodge
number が読み取れる事も意味する。
66 N = (2, 2) Landau-Ginzburg Theory
II.B.1 quintic hyperdurface
まずは最も良く知られた quintic hypersurface を考えよう。dual な LG superpotential WLG は
WLG =5∑
i=1
X5i = X5
1 +X52 +X5
3 +X54 +X5
5 (II.B.1)
で定義される22。これは先程までの議論に登場する A4 type singularity theory の 5 つの直積である。これについて
は chapter V でももう少し議論を展開するが、ここでは ρ を求めるに留めておく。
“highest charged state” ρ の定義は
ρ = deta,b
( ∂2W
∂Φa∂Φb
)
である。先程の superpotential (II.B.1) を代入すると
ρ = detij
(20Xiδij
)= 205 ·X1X2X3X4X5 (II.B.2)
となる。これはゼロではない。また原点ではゼロになる。これは isolated singularity を持つ模型であることを意味し
ている。
22より厳密には、WLG は Xi の 5 次の一般的な多項式で定義されるが、ここでは簡単に Fermat type にする。
Chapter III
Two-dimensional N = (2, 2) Superconformal Theory
This chapter deeply depends on the work by Lerche, Vafa and Warner [72].
68 Two-dimensional N = (2, 2) Superconformal Theory
III.1 Superconformal algebras
N = (2, 2) superconformal algebra を構成するため、energy-momentum tensor T (z) とその superpartner G±(z)、
そして N = (2, 2) supercharge の rotation から来る U(1) current J(z) の OPEを列挙しよう1:
T (z)T (w) ∼ c/2
(z − w)4+
2
(z − w)2T (w) +
1
z − w∂wT (w) , (III.1.1a)
T (z)G±(w) ∼ 3/2
(z − w)2G±(w) +
1
z − w∂wG±(w) , (III.1.1b)
T (z)J(w) ∼ 1
(z − w)2J(w) +
1
z − w∂wJ(w) , (III.1.1c)
J(z)J(w) ∼ c/3
(z − w)2, (III.1.1d)
J(z)G±(w) ∼ ± 1
z − wG±(w) , (III.1.1e)
G+(z)G−(w) ∼ 2c/3
(z − w)3+
2
(z − w)2J(w) +
1
z − w(2T (w) + ∂wJ(w)
). (III.1.1f)
なお chiral/anti-chiralの表記と、holormophic/anti-holomorphic (left-moving/right-moving)の表記が、Wess-Bagger
とは逆であることに注意。ここの表記である CFT [72] では、holomorphic/anti-holomorphic を T (z)/T (z) で表記
し、chiral/anti-chiral を G+/G− で表記している。Wess-Bagger notation [107, 113] はその逆である。
z-plane での Laurent expansion を与える:
T (z) =∑
n∈Z
Lmz−n−2 , Lm =
∮dz
2πizm+1T (z) , (III.1.2a)
J(z) =∑
n∈Z
Jmz−n−1 , Jm =
∮dz
2πizmJ(z) , (III.1.2b)
G±(z) =∑
r∈Z(+1/2)
G±r z
−r−3/2 , G±r =
∮dz
2πizr+1/2G±(z) . (III.1.2c)
OPE から、これらの mode 同士の交換関係が得られるが、それがN = (2, 2) superconformal algebra をなす:
[Lm, Ln] = (m− n)Lm+n +c
12m(m2 − 1)δm+n , (III.1.3a)
[Lm, G±r ] =
(n2− r)G±
n+r , (III.1.3b)
[Lm, Jn] = −nJm+n , (III.1.3c)
[Jm, Jn] =c
3mδm+n , (III.1.3d)
[Jn, G±r ] = ±G±
n+r , (III.1.3e)
G−r , G
+s = 2Lr+s − (r − s)Jr+s +
c
3
(r2 − 1
4
)δr+s . (III.1.3f)
bosonic generator Lm, Jm では m ∈ Z であるが、fermionic generator G±r では r ∈ Z もしくは r ∈ Z + 1
2 である。
前者に属するものを Ramond (R) sector と呼び、後者に属するものを Neveu-Schwarz (NS) sector と呼ぶ。
1holomorphic part と anti-holomorphic part とは同じ代数をなす。
III.2 NS chiral primary states and R ground states 69
III.2 NS chiral primary states and R ground states
まずは conformal vacuum について議論する。その後 bosonic primary states について記述しよう。NS sector と R
sector についても後半で議論する。なお、ここでも left-moving mode (holomorphic part) のみ議論するに留める。
まず、conformal theory を議論する際、vacuum を定義する必要がある。vacuum は conformal 変換で不変でなけ
ればならない。これを energy-momentum tensor の言葉でいうと、
limz→0
T (z)| 0 〉 = finite (III.2.1)
となる。つまり pole を持ってはならないという条件になる。energy-momentum tensor T (z) は conformal 変換の
generator であり、その (空間)積分が charge を与える。conformal invariance は
Q| 0 〉 =
∮dz
2πiT (z)| 0 〉 = 0 (III.2.2)
であるので、(III.2.1) が自然な定義となる。superconformal vacuum の場合は、同時に
limz→0
G±(z)| 0 〉 = finite (III.2.3)
でなければならない。これらを Virasoro generator で表すと、
Lm| 0 〉 = 0 for m ≥ −1 , G±r | 0 〉 = 0 for
r ≥ − 1
2 (NS sector)
r ≥ −1 (R sector), (III.2.4)
である。
次に primary states を定義しよう。operator φ(z, z) が次の変換をするとき、それを primary operator と呼ぶ:
φ′(z′, z′) =( ∂z∂z′
)h( ∂z∂z′
)h
φ(z, z) . (III.2.5)
h, h は conformal weight と呼ばれる。これらから dimension (∆ = h+ h), spin (s = h− h) が定義される。また、この operator を用いて、primary state |φ 〉 が、
|φ 〉 = limz,z→0
φ(z, z)| 0 〉 , (III.2.6)
で定義される。T (z) と φ(z, z) の OPE が (III.2.5) から
T (z)φ(w,w) ∼ h
(z − w)2φ(w,w) +
1
z − w∂wφ(w,w) , (III.2.7)
で与えられるので、primary states に関する条件が
L0|φ 〉 = h|φ 〉 , Lm|φ 〉 = 0 for m > 0 , (III.2.8)
となる。この primary states に L−m (m > 0) を作用させて得られるものを、descendant states と呼ぶ。
ここから superconformal modelの statesについて議論する。NS sectorから考えよう。この sectorではNS chiral
primary states が考えられるが、それは (III.2.8) の条件に
G+n−1/2|φ 〉 = G−
n+1/2|φ 〉 = 0 for n ≥ 0 , (III.2.9)
70 Two-dimensional N = (2, 2) Superconformal Theory
が加わるものである。但しこれは left chiral primary states の条件である。left anti-chiral primary states や right
(anti-)chiral primary states の定義は (III.2.9) の代わりに
G+n+1/2|φ 〉 = G−
n−1/2|φ 〉 = 0 for n ≥ 0 , left anti-chiral (III.2.10a)
G+n−1/2|φ 〉 = G−
n+1/2|φ 〉 = 0 for n ≥ 0 , right chiral (III.2.10b)
G+n+1/2|φ 〉 = G−
n−1/2|φ 〉 = 0 for n ≥ 0 , right anti-chiral (III.2.10c)
で与えられる。N = (2, 2) superconformal algebra (III.1.3) を用いて chiral primary states の性質を見ておこう。
chiral primary states として
L0|φ 〉 = h|φ 〉 , J0|φ 〉 = q|φ 〉 , (III.2.11a)
Lm|φ 〉 = Jm|φ 〉 = 0 for m > 0 , (III.2.11b)
を用意する。G−1/2, G
+−1/2 より
0 =G−
1/2, G+−1/2
|φ 〉 = (2L0 − J0)|φ 〉 (III.2.12)
であるので、
h =1
2q (III.2.13)
が得られる。anti-chiral primary の場合は h = − 12q である (q ≤ 0)。また、unitarity condition (positive norm
condition) より、一般の states は
h ≥ 1
2|q| (III.2.14)
である。さらに G−3/2, G
+−3/2 と unitarity condition より、
G−
3/2, G+−3/2
|φ 〉 =
(2h− 3q +
2
3c)|φ 〉 ≥
(− 4h+
2
3c)|φ 〉 , ∴ h ≤ c
6, (III.2.15)
が得られる。
chiral primary operator 同士の product は limz′→z φ(z′)χ(z) = (φχ)(z) で与えられる。この関係式を調べるため、
chiral (h ≥ 12q ≥ 0) と、U(1) charge が additive であることを用いよう。(φχ)(z) が primary operator であるとき、
hφχ ≥1
2qφχ =
1
2(qφ + qχ) , (III.2.16)
である。また、φ と χ の OPE は一般に
φ(z′)χ(z) = (z′ − z)hφχ−hφ−hχ(φχ)(z) , (III.2.17)
で与えられる。よってもし (φχ)(z) が chiral primary であるなら、すなわち等号が成立するなら、この積は有限で存
在する。もし chiral primary にならないなら、つまり descendant operator になっているなら、それは pole を持た
ずに消える。この性質より、chiral primary states は ring structure Rc を持つ事がわかる。chiral primary states の
数が有限なら、この ring の次元も有限である。
III.3 Spectral flow 71
では次に Ramond sectorについて考えよう。NS chiral primary statesのように、この sectorには Ramond ground
states | φ 〉 と呼ばれる次の states を定義する事が出来る:
G±n | φ 〉 = G±
n | φ 〉 = 0 for n ≥ 0 . (III.2.18)
但し | φ 〉 は (III.2.8) も満たしているとする。このとき、N = (2, 2) superconformal algebra (III.1.3) において
G−0 , G
+0 = 2L0 − c/12 であるので、これを R ground states に作用させると、
0 =G−
0 , G+0
| φ 〉 =
(2h− c
12
)| φ 〉 , ∴ h =
c
24, (III.2.19)
が得られる。これは U(1) charge に依存しない。さらに、R ground states に G±−m (m > 0) を作用させて得られる
descendant states の unitarity condition より、一般には
h ≥ c
24(III.2.20)
であることがわかる。
III.3 Spectral flow
N = (2, 2) SCFT で登場する non-trivial な state (NS and R sector) を、one parameter で連続的に接続する事が可
能である。NS sector と R sector を、U(1) charge を変換させる事でつなげるのである。具体的には U(1) charge q
を q − c3θ に変換させる作用を考える (θ ∈ R)。
Hilbert space H を、θ で次のように変換させる:
H0 → Hθ , (III.3.1a)
operator O や state |A 〉 に関しても、変換の operator Uθ を次のように作用させる:
O → Oθ = UθOU−1θ , |A 〉 → |Aθ 〉 = Uθ|A 〉 . (III.3.1b)
具体的に N = (2, 2) SCFT の generator に関して次のような変換を施す:
Lθn = UθLnU
−1θ = Ln + θJn +
c
6θ2δn,0 , (III.3.2a)
Jθn = UθJnU
−1θ = Jn +
c
3θδn,0 , (III.3.2b)
G±,θr = UθGr
±U
−1θ = G±
r±θ . (III.3.2c)
この変換された generator がなす superconformal algebra は、変換前のそれと同型である。実際に計算をすれば、変
換後の generator も N = (2, 2) SCFT algebra をなす事がわかる。
この変換は、operator の変換則が重要なのではなく、state の変換が重要である。spectral flow によって、ある
state が違った state にどう変換されるのか、conformal weight や U(1) charge はどうなったのか、を見るのが重要
なのである。
まずは left moving mode のみを考えよう。次の chiral primary state を用意する:
L0|φ 〉 = h|φ 〉 , J0|φ 〉 = q|φ 〉 , h =1
2q , h ≤ c
6, (III.3.3a)
72 Two-dimensional N = (2, 2) Superconformal Theory
Lm+1|φ 〉 = G+m−1/2|φ 〉 = G−
m+1/2|φ 〉 = 0 for m ≥ 0 . (III.3.3b)
spectral flow された operator が spectral flow された chiral primary state に作用した時の様子を見よう:
Lθ0|φθ 〉 = UθL0U
−1θ Uθ|φ 〉 = h|φθ 〉 , (III.3.4a)
Jθ0 |φθ 〉 = UθJ0U
−1θ Uθ|φ 〉 = q|φθ 〉 . (III.3.4b)
この関係を用いて、spectral flow された chiral primary state の conformal weight hθ や U(1) charge qθ を求める:
L0|φθ 〉 ≡ hθ|φθ 〉 , J0|φθ 〉 ≡ qθ|φθ 〉 , (III.3.5a)
Lθ0|φθ 〉 =
(L0 + θJ0 +
c
6θ2)|φθ 〉 =
(hθ + θqθ +
c
6θ2)|φθ 〉 , (III.3.5b)
Jθ0 |φθ 〉 =
(J0 +
c
3θ)|φθ 〉 =
(qθ +
c
3θ)|φθ 〉 , (III.3.5c)
∴ hθ = h− qθ +c
6θ2 , qθ = q − c
3θ . (III.3.5d)
これより、NS chiral primary state |φ 〉 (G+−1/2|φ 〉 = 0) は spectral flow θ = 1/2 で |φ 1
2〉 となる。この state は
G+0 |φ 1
2〉 = 0 を満たす。よって |φ 1
2〉 は Ramond ground state である。NS chiral primary state と R ground state
の conformal weight h、U(1) charge q を掲載しよう:
Neveu-Schwarz spectral flow Ramond
(h = q2 , q)
chiral primary state→ (θ = 1
2 )→ (c/24, q − c/6)
R ground state
さらに θ = 1 とすると、例えば NS chiral primary state は NS anti-chiral primary state と繋がる:
G+−1/2|φ 〉 = G−
1/2|φ 〉 = 0θ=1−−→ G+
1/2|φ1 〉 = G−−1/2|φ1 〉 = 0 , (III.3.6)
∴ |φ1 〉 is an anti-chiral primary state.
何故 anti-chiral なのかというと、annihilation condition がそうだからである。また、positive U(1) charge を chiral
としていると、θ = 1 spectral flow で得られた |φ1 〉 は negative charge を持つからである:
h1 = h− q +c
6= −h+
c
6≥ 0 , q1 = q − c
3= 2
(h− c
6
)≤ 0 . (III.3.7)
spectral flow を Figure III.1 に表現しよう:
先程までは holomorhic sector についてのみの議論であったが、勿論 anti-holomorhic sector についても同様であ
る。ただ、U(1) charge の符号が逆であることに注意すると、spectral flow の向きも逆に定義しておく方が望ましい
ことが分かる。NS anti-chiral primary state と R ground state の conformal weight h、U(1) charge q を掲載しよ
う (q < 0 としておく):
Neveu-Schwarz spectral flow Ramond
(h = − q2 ,−q)
anti-chiral primary state→ (θ = − 1
2 )→ (c/24,−q + c/6)
R ground state
III.4 Witten index, Poincare polynomial 73
hh h
q
θ = 12θ = 1
2
h = q2 h = − q
2c6
c6 c
24
c3 − c
6c6 − c
3(a) (b) (c)
Figure III.1: In this figure it is shown that starting fron the NS sector (a), the chiral primary states flow to the
ground states of R sector (b) with flow parameter θ = 1/2. Flow by an additional θ = 1/2 will take the ground
state of the R sector to the anti-chiral primary states of the NS sector (c). The effect of the spectral flow on charges
of states is simply to shift them.
さらに θ = −1 とすると、例えば NS anti-chiral primary state は NS chiral primary state と繋がる:
G+1/2|φ 〉 = G−
−1/2|φ 〉 = 0θ=−1−−−−→ G+
−1/2|φ−1 〉 = G−1/2|φ−1 〉 = 0 , (III.3.8)
∴ |φ−1 〉 is a chiral primary state.
N = (2, 2) superconformal model は left と right が同時に system に存在するため、同時に spectral flow を考
える必要がある。left、right が共に chiral primary である (c, c) primary states (N = (2, 2) supersymmetric theory
では chiral superfield がこの候補になる) の場合は、left/right を同じ方向に spectral flow θL = θR = 1/2 させる
(qL, qR > 0 とする):
θL = θR = 1/2 NS-NS (c, c) primary states R-R ground states
conformal weight (hL, hR) (c/24, c/24)
U(1) charge (qL, qR) = (2hL, 2hR) (qL − c/6, qR − c/6)
一方、left、rightがそれぞれ anti-chiral primary、chiral primaryとなる (a, c) primary states (twisted chiral superfield
がこれの候補になる。) については、left と right の spectral flow は逆であるとする (qL < 0, qR > 0 とする):
−θL = θR = 1/2 NS-NS (a, c) primary states R-R ground states
conformal weight (hL, hR) (c/24, c/24)
U(1) charge (−qL, qR) = (2hL, 2hR) (−qL + c/6, qR − c/6)
III.4 Witten index, Poincare polynomial
Witten index Tr(−1)F を定義しよう。これは Ramond ground states における (worldsheet) fermion zero mode の
個数を表す2[108]。定義は
Tr(−1)F = TrR
[(−1)J0−J0 qL0−c/24 qL0−c/24
](III.4.1)
2一般には field φ の zero mode は (−∂2 + m2) という differential operator が作用するとゼロになる mode φ (ker(−∂2 + m2)) で定義さ
れる。この mode は path integral (Gaussian integral) が発散するため、特別に議論される場合が多い。一方、Witten index の議論では、zero
74 Two-dimensional N = (2, 2) Superconformal Theory
である (q = exp(iπτ))。この indexを、left/right symmetric spectral flow (θL = θR ≡ θ)を用いて NS chiral primary
states でも計算する事が出来る。まずは一般的に次のように再定義しよう:
Tr(−1)F ≡ TrHθ
[(−1)Jθ
0−Jθ0 qLθ
0−c/24 qLθ0−c/24
]. (III.4.2)
θ = 0 では R-R ground state に関する演算にあたり、Witten index (III.4.1) を与える。θ = − 12 とおくと、R-R
ground state から NS-NS (c, c) primary state へと flow し、θ = 12 のときは NS-NS (a, a) primary state へと flow
する。どちらに flow してもよいが、(a, a) は (c, c) の complex conjugate と見なせるので、θ = − 12 の場合だけを考
えれば良い。θ = − 12 で得られる Witten index は
Tr(−1)F = TrH 1
2=NS
[(−1)J0−J0 qL0− 1
2J0 qL0− 1
2J0
](III.4.3)
となる。よって、(c, c) primary state (h = 12q) では、
Tr(−1)F =∑
R
exp(iπ(qL − qR)
)(III.4.4)
と、left/right U(1) charge の差で与えられる。ここで R は後で議論する chiral ring (正確には (c, c)-ring) を指す。
このWitten index (III.4.2)をさらに一般化しよう。ある任意の複素数 t, t ∈ C∗ を用意して、NS-NS (c, c) primary
states について次のようなもの (Poincare polynomial と呼ぶ) を定義する:
P(c,c)(t, t) = TrNS
[tJ0 tJ0
]∣∣∣R
. (III.4.5)
またこれの一般形 (spectral flow) を定義する:
P ≡ TrHθ
[tJ
θ0 tJ
θ0
]∣∣∣ground states
. (III.4.6)
(III.4.6) は θ = 0 のときに (III.4.5) を再現するように定義する。つまり NS-NS (c, c) primary states に関する和を
与える。θ = 12 ととると、NS-NS (c, c) primary states から R-R ground states へ spectral flow を起こすので、P を
介して関係がつく:
P = TrNS
[tJ0 tJ0
]∣∣∣R
= TrH 1
2
[tJ0+c/6 tJ0+c/6
]∣∣∣ground states
= (tt)c/6 TrR
[tJ0 tJ0
]∣∣∣G±
0 =G±
0 =0. (III.4.7)
但し central charge について c = c とした。これを、後の議論が明確になるように次のように並べ直す:
TrR
[tJ0 tJ0
]∣∣∣G±
0 =G±
0 =0= (tt)−c/6 Tr
NS
[tJ0 tJ0
]∣∣∣R
= (tt)−c/6 P(c,c)(t, t) . (III.4.8)
Ramond ground states は U(1) charge conjugation invariant である (flip などの convention ではなく、実際に
invariant である) ことを用いると、
TrR
[tJ0 tJ0
]∣∣∣G±
0 =G±
0 =0= Tr
R
[t−J0 t−J0
]∣∣∣G±
0 =G±
0 =0(III.4.9)
energy states の数を議論する。一般にはこれらは異なる。しかしここでは CFT を扱っている。CFT は dimensionful parameter を持たない。
つまり mass は m2 = 0 である。このため、differential operator は −∂2 だけとなる。spatial momentum が ~p = 0 の点で考察する限りは、
この 2 つの概念は同じとなる。
III.4 Witten index, Poincare polynomial 75
である。またこれより
TrR
[t−J0 t−J0
]∣∣∣G±
0 =G±
0 =0= (tt)c/6 Tr
NS
[t−J0 t−J0
]∣∣∣R
= (tt)c/6 P(c,c)(1/t, 1/t) . (III.4.10)
である。これらを用いると、P (t, t) に
P(c,c)(t, t) = (tt)c/3P(c,c)(1/t, 1/t) , (III.4.11)
の関係がつく事になる。この関係式は非常に重要である。この関係式から、NS-NS (anti-)chiral primary states 同
士の関係が得られる。例えば、highest charged state (hL, hR) = (c/6, c/6) 一つのみに着目しよう。これについて
P(c,c)(t, t) = (tt)c/3 が得られる。一方、vacuum (hL = hR = 0) に関しては P(c,c)(t, t) = 1 が得られる。これらと
(III.4.11) を比較すると、ちょうど繋がる事がわかる。よって、highest charged NS-NS chiral primary state と NS
vacuum は “Poincare duality” で繋がるという。なお、この関係の起源は R-R ground states の charge conjugation
invariance からくる:
(h, q) = (c/6, c/3) ← Poincare duality → (h, q) = (0, 0)
Poincare duality
between NS chiral primary states=
charge conjugation invariance
in R ground states
Poincare polynomial (III.4.5) の別の記述をしよう。NS-NS (c, c) primary states 同士の積は (c, c) primary states
になるか、もしくは zeroになることから、(c, c) primary statesの積は ringをなす。それをRcc と書くと、この element
は一般に left/right U(1) charge (p, q)を持つ (c, c) primary statesで与えられる。一方、Poincare polynomialは、NS-
NS (c, c) primary states を全て用いて与えられるので、この ring elements で定義されている。Poincare polynomial
の定義から、P(c,c)(t, t) を、次のように展開して表示する事が出来る3:
P(c,c)(t, t) =d∑
p,q=0
bd−p,q tp tq , d =
c
3. (III.4.12)
Poincare polynomial の構成より、この係数 bd−p,q は、left/right U(1) charge が (p, q) である (c, c) primary states
(ring elements) の個数を示す。但しここで、chiral primary states の性質から 0 ≤ p, q ≤ c/3 ≡ d であることを用いている。
ここで、少し視点を変えて、qL = −p, qR = q (p, q ≥ 0) の charge を持つ NS-NS (a, c) primary states が与える
polynomial P(a,c)(t, t) を定義しよう。(c, c) primary とは charge の符号が異なるだけなので、
P(a,c) = TrNS
[t−J0 tJ0
]∣∣∣Rac
, (III.4.13)
と記述できる。ここで Rac は、(a, c) primary states がなす ring である。
この P(a,c)(t, t) を (III.4.5) の様に表したい。そこでまず、P(c,c)(t, t) との関係を見よう。そのために spectral flow
をもう一度用いる。θ = 0 の時に P(a,c)(t, t) であるように定義しよう:
TrHθ
[t−Jθ
0 tJθ0
]∣∣∣ . (III.4.14)
3Lerche, Vafa and Warner [72] とは係数 bp,q の定義が異なる。このノートの定義にしておいた方が、chiral superfield で構成される
superpotential W の deformation が complex deformation に対応し、twisted chiral superfield で構成される twisted superpotential fW の
deformation が Kahler deformation に対応することが明解になる。
76 Two-dimensional N = (2, 2) Superconformal Theory
これは spectral flow θ の値に依存しないとする。これより、
P(a,c)(t, t) = TrHθ=0
[t−Jθ
0 tJθ0
]∣∣∣ = TrH
θ= 12
[t−Jθ
0 tJθ0
]∣∣∣ = TrR
[t−(J0+d/2) tJ0+d/2
]∣∣∣
= (t−1t)d/2 TrR
[t−J0 tJ0
]∣∣∣ = (t−1t)d/2 TrR
[tJ0 tJ0
]∣∣∣
= t−d TrNS
[tJ0 tJ0
]∣∣∣Rcc
= t−dP(c,c)(t, t) , (III.4.15)
となる。但し R ground states における charge conjugation invariance や、spectral flow θ = − 12 による引き戻しを
途中行っている。(III.4.12) を用いると、
P(a,c)(t, t) = t−dd∑
p,q=0
bd−p,q tp tq =
d∑
p,q=0
bd−p,q t−d+p tq , (III.4.16)
となる。ここで −d+ p ≡ −p′ と再定義しよう。これより 0 ≤ p′ ≤ d であるので、ベキを組み直して
P(a,c)(t, t) =
d∑
p′,q=0
bp′,q t−p′
tq (III.4.17)
と書き直す事が出来る。これより、係数 bp′,q は (−p′, q) charged (a, c) primary states の個数を表していることにな
る。(a, c)-ring Rac の elements のうち、(−p′, q) charged (a, c) primary states の個数を表す、と表現した方がより
正確であろう。
実は、SCFTの NS-NS (c, c) primary statesで構成される chiral ringと、ある geometryの Dolbeault cohomology
ringは同一視する事が出来る。このとき、SCFT sideを調べる事で、その対応する geometry (CY manifold)の topology
を調べる事が出来る。具体的には、(p, q) charged (c, c) primary states の数が、Hodge number bd−p,q そのものであ
る。また、(c, c)-ring から読みとられる Hodge number を持つ CY manifold M の他に、(a, c)-ring で構成された
polynomial の係数 bp,q を Hodge number bd−p,q と同一視する事も出来る。そのとき、対応する CY d-fold W は、先の (c, c)-ring から読みとられた CY d-fold M と mirror pair の関係にある。
Chapter IV
CY/LG Correspondence
78 CY/LG Correspondence
chapter II で扱った LG theory の IR limit で登場する minimal model は、幾何学的な情報を持つと見なす事が
出来る。特に、CY manifold 上の cohomology などと対応がつくと言われている [72, 41]。ここでは大雑把にその対
応を記述しよう。細かな定義などはここではしない。登場する tool は、appendix III を参照する事。
まず、N = (2, 2) SCFT の幾何学的側面を見るため、LG model と CY が持つ性質を列挙しておこう。但しここ
では chiral superfield からすぐに見える物のみを扱う。つまり (c, c)-ring、complex moduli である:
• complex moduli and its degree
• marginal deformation
• chiral primary field and its degree
• cohomology and its dimension
これらは互いに密接な繋がりがある。その関係を、LG side と CY side でそれぞれ簡単に述べ、最後につなげよう。
IV.1 CY geometry
ある CY metric gij を用意する。この metric は Ricci-flat である。この条件を保ちつつ perturbation (marginal
deformation) をかける。その時、deformation の仕方は大きく分けて 2 種類 δgij type と δgij がある。前者は Kahler
classの deformationに相当し、後者は complex structureの deformationに対応する。(Kahler manifoldでは gij = 0
であるが、defromation は δgij 6= 0 も許される。) 後者、つまり complex deformation についてのみ考察しよう。
deform された新しい metric g′ij
= gij + δgij + δgij は、最初の CY とは異なるが、deformation で繋がっている。こ
の deformation は、ある parameter (今の場合 complex moduli) で特徴づけられ、この deformation parameter で張
られる空間を complex moduli space と言う (see [13])。complex moduli space の次元は complex moduli の数で決
まるが、CY d-fold のとき、それは harmonic (d− 1, 1)-form の数 bd−1,1 で表される。
IV.2 LG minimal model
LG minimal model での議論をする。NS sector の field は chiral primary かつ、left/right U(1) charge が等しい
(qL = qR)、つまり (c, c)-ring element であるとする。
LG superpotential W ′ を用意する。これは chiral primary field の polynomial である。これに、marginal defor-
mation という形で perturbation をかけるが、付加させる term は、conformal weight が (h, h) = (1, 1) の chiral
primaryであるとする (chiral fieldの積は chiral field)。これにより、この付加させた termを構成する chiral primary
field の conformal weight が決定される。
chiral primary field の U(1) charge が (p, q) であるとする。もし整数でなければ、orbifolding をすることで整数
にする1。
1さらに言えば、chiral ring の元は、整数の U(1) charge になった chiral primary field である。分数のそれは、chiral ring の element に
はならない。orbifold model の場合は、orbifolding をする前に chiral ring の議論をしても意味がない。
IV.3 Correspondence 79
IV.3 Correspondence
Lerche-Vafa-Warner [72] から読みとれることをもう少し正確に記述しよう。CY/LG correspondence という場合、実
際には 2つの議論を同時に行っている。一つは “CY/SCFT”であり、もう一つは “SCFT/LG”である。前者は、SCFT
の Poincare polynomial が CY geometry の Poincare polynomial と同一視する事を指し、後者は LG superpotential
のなす ring と、SCFT の chiral ring を同一視することを指す。
1. CY/SCFT
Witten [108] の議論より、harmonic forms on CY が SCFT の Ramond ground states と同一視される。そし
て、spectral flow により、その R-R ground states が NS-NS chiral primary states に変換される。よって、最
終的に、harmonic (p, q)-forms on CY が NS-NS (d− p, q) charged (c, c) primary states と one-to-one 対応す
る (p, q ≥ 0)。もしくは、harmonic (p, q)-forms on CY と (−p, q) charged (a, c) primary states と one-to-one
対応する。
R-R states | φ 〉 は、
G+0 | φ 〉 = G+
0 | φ 〉 = 0 , and G+m| φ 〉 = G+
m| φ 〉 = 0 for m > 0 (IV.3.1)
が定義であり、spectral flow θL = θR = − 12 によって (c, c) primary states |φ 〉 にする:
G+0 | φ 〉 → UθL=−1/2G
+0 | φ 〉 = G+
−1/2|φ 〉 = 0 , (IV.3.2a)
G+0 | φ 〉 → UθR=−1/2G
+0 | φ 〉 = G+
−1/2|φ 〉 = 0 . (IV.3.2b)
この (c, c) primary state の U(1) charge を (qL, qR) = (p, q) であるとしよう。(そのように R-R ground state
| φ 〉 の U(1) charge を定義すれば良い。)
では SCFT の states と CY manifold 上の harmonic form の関係を Figure IV.1 に表現する2。
まとめると Tabel IV.1 のようになる:
CY d-fold M N = (2, 2) SCFT (d = c/3, c = c)
bd−p,q dimHd−p,q(M) # of (p, q) charged (c, c) primary states
P (t, t) Poincare polynomial “Poincare” polynomial ((c, c) primary)
bp,q = bd−p,d−q Poincare duality charge conjugation invariance on R ground states
b0,0 = 1 simply connected uniqueness of vacuum: (hL, hR) = (0, 0)
bd,d = 1 Poincare duality uniqueness of highest charge state: (hL, hR) = (c/6, c/6)
bd,0 = 1 holomorphic (d, 0)-form spectral flow from vacuum: (θL, θR) = (−1/2, 0)
Table IV.1: Correspondence and identification between CY geometry and SCFT.
2. SCFT/LG
2(c, c) primary state の charge を (p, q) (p, q ≥ 0) としておく。
80 CY/LG Correspondence
(c, c) ring element:
(p, q) charged
NS-NS chiral primary
p, q ≥ 0
Spectral flow
θL = 1, θR = 0
(a, c) ring element:
(p − d, q) charged
NS-NS chiral primary
p − d ≤ 0
One-to-one
correspondence
Spectral flow
−θL = θR = 12
harmonic (d − p, q)-form
on Calabi-Yau d-fold
d − p, q ≥ 0
d = c/3
One-to-one
correspondence
R-R gound state
(SUSY ground state)
(qL, qR) = (p − d2, q − d
2)
hL = hR = c/24
Figure IV.1: Relation among NS-NS primary states, RR ground states and harmonic (d−p, q)-form on CY d-fold.
LG model は non-trivial IR fixed point が存在し、その上では SCFT minimal model として記述される。LG
を記述する chiral superfield Φi と superpotential W (Φi) の left/right U(1) charge が、conformal invariance
から決定される。ここで、superpotential W から ring structure を構成する事が出来る:
RLGcc =
C(Φ)
[∂iW (Φ)]. (IV.3.3)
この local ring RLGcc が、SCFT の (c, c) ring RSCFT
cc と同一視される:
RLGcc ≃ R
SCFTcc . (IV.3.4)
ただ、canonical (つまり unorbifolded) LG には、chiral superfield (left-right symmetric charge qL = qR) の
みが模型を記述しているので、local ring RLGcc の element も当然 left-right symmetric ring element しかない。
さらに、(a, c)-ring element は存在しない (vacuum 以外の non-trivial element がない)。すなわち、ring から
読みとれる bd−p,q が、p 6= q の時には zero である。ここで bd−p,q は、(p, q) charged (c, c) ring element の数
を表す。SCFT/CY では、CY Hodge number と同一視される。つまり、LG から CY geometric data を得た
いなら、canonical LG では満たされない事になる。そこで orbifolding を行う。
orbifolding を行うと、SCFT の twisted sector から non-trivial (c, c) ring elements や non-trivial (a, c) ring
elements が登場する。また、untwisted sector では、orbifolding で invariant な elements 以外は project out
される。
3. CY/orbifolded LG
以上により、SCFT を介する事で、CY と “orbifolded” LG を同一視する事が出来る。道具は、Poincare poly-
nomial P (t, t), Hodge number bp,q である。
この correspondence を簡明に説明するのが、本文で扱っている GLSM と T-dualized theory である。
IV.3 Correspondence 81
GLSM/T-dualized theory は chiral superfield Φi と twisted chiral superfield Σ, Yi を用いて記述される。それぞ
れ left/right U(1) charge を定義できるとして、charge assignment などを
Φi : (qL, qR) = (qi, qi) (c, c) ring elements in LG
Yi : (qL, qR) = (−qi, qi) (a, c) ring elements in LG
とする。superpotential W , twisted superpotential W はそれぞれ chiral superfield, twisted chiral superfield で記述
される。よって、marginal deformation δW , δW はそれぞれ、(c, c)-ring deformation, (a, c)-ring deformation と考
える事が出来る。
ここで、marginal deformation δW , δW を別の視点で捉える。δW は、F-term deformation であり、target space
の complex structureの marginal deformationに相当する。一方、δW には (complexified) FI parameter τ が入ってお
り、これは (complexified) Kahler form on target spaceに比例する。よって、δW は δτ に相当し、さらに Kahler class
deformation に相当することになる。したがって、complex deformation, Kahler deformation はそれぞれ (c, c)-ring
deformation, (a, c)-ring deformation と呼ぶことが可能である。
Chapter V
Calabi-Yau Data from LG Superpotentials
84 Calabi-Yau Data from LG Superpotentials
Introduction
GLSM では、quintic hypersurface in CP4 と dual な LG superpotential は、例えば Fermat type
W (Si) =
5∑
i=1
S5i (V.0.1)
で与えられることが示される。これは、ADE classification で言えば、5 つの A4 type の direct product で与えられ
ていることに相当すると思われる:
LG dual of CP4[5] CY sigma model ∼ A4 ⊗A4 ⊗A4 ⊗A4 ⊗A4 .
事実、central charge は
c = 5×(3− 6
4 + 1
)= 9 (V.0.2)
であり、CFT on CY 3-fold の central charge と一致する。
実はもっと精密に議論が展開可能である。最も分かりやすい解説として Brunner, Hori, Hosomichi and Walcher
[9] section 2.1 がある。これにはAn type LG minimal model で構成される具体的な CY 3-fold の例を列挙してある。
ここでは原文を抜粋し、具体的な計算を実行する。
V.1 CY data from An type LG superpotentials
Consider a (2, 2) supersymmetric gauge theory in 1 + 1 dimensions with U(1) gauge group and r + 1 fields
X1, · · · ,Xr, P with tree level superpotential
W = P ·(Xk1+2
1 + · · ·+Xkr+2r
)(V.1.1)
and twisted superpotential
W = −Στ .
Σ = 1√2D+D−V is the superfieldstrength and τ = r − iθ/2π is the Fayet-Iliopoulos-Theta parameter. The gauge
transformations act on the fields as
P → e−iHλP , Xi → eiwiλXi ,
where
H := LCMki + 2 , wi :=H
ki + 2.
For large values of the FI parameter, the system reduces at low energies to the sigma model on the hypersurface
M = Xk1+21 + · · ·+Xkr+2
r = 0 in a weighted projective space of dimension r− 1. This gauge system, introduced
V.1 CY data from An type LG superpotentials 85
in [113], is called the linear sigma model for the manifold M . The condition that M is Calabi-Yau is reflected by
the vanishing of the sum of charges −1 +∑r
i=11
ki+2 = 0. Namely
c
3=
r∑
i=1
ki
ki + 2= r − 2 = dimC M . (V.1.2)
In this case, the beta function for the FI parameter vanishes and therefore τ is a free parameter of the system.
At large negative Re(t), the P field has a vacuum expectation value and breaks the U(1) gauge symmetry to
the subgroup in which eiHλ = 1. This unbroken subgroup Γ is generated by the one with λ = 2π/H which acts on
the fields as
γ : Xi → e2πi
ki+2Xi , (V.1.3)
and is a cyclic group of order H. The model at τ = −∞ is identified as the LG orbifold with superpotential
WG = Xk1+21 + · · ·+Xkr+2
r (V.1.4)
divided by the group Γ ≃ ZH acting on fields as (V.1.3). The LG model with superpotential W = Xk+2 flows in the
IR limit to a (2, 2) superconformal field theory with central charge c = 3kk+2 , called the (A-series) level k N = (2, 2)
minimal model, Mk (or Ak+1 minimal model). The IR limit of the above LG orbifold is thus the Γ-orbifold of the
product of the minimal models;
( r∏
i=1
Mki
)/Γ
This is the Gepner model. The generator (V.1.3) of the orbifold group Γ is identified as
γ = (g, · · · , g)︸ ︷︷ ︸r
(V.1.5)
in which g := e−2πiJ0(−1)bF where J0 is the U(1) current of the (right-moving) N = (2, 2) superconformal algebra
and (−1)bF is 1 on NSNS sector but −1 on RR sector. Note that the RR-ground states of lowest R-charge q = − c
6
survives the orbifold projection, since c6 = dim M
2 = r−22 and thus γ = eπi(r−2)(−1)r = 1. This state corresponds to
the holomorphic volume form of the Calabi-Yau manifold.
Type II string theory on M × RD is consistent only if 2 dimM +D = 10. If we denote the complex dimension
of the transverse space by d = (D − 2)/2, the criticality condition is
r + d = 6 . (V.1.6)
In this subsection we assume both the Calabi-Yau condition (V.1.2) and the criticality condition (V.1.6).
Remarks.
1. It is possible to have some ki = 0. The IR limit of W = X2 is empty, but can be regarded as the system with a unique
(ground) state in each of R/NS-sectors, with zero energy, zero charge. We will regard the ki = 0 factor as such a
quantum field theory. The orbifold group acts on this factor non-trivially: the generator γ acts as g = e−2πiJ0(−1)bF =
(−1)bF , namely, as identity on NSNS sector but as (−1) on the RR-sector. Thus, having this factor has a non-trivial
effect.
86 Calabi-Yau Data from LG Superpotentials
2. The behaviour of the system depends very much on whether there is an even ki. It is useful to note that when there
is at least one even ki there is actually an even number of i with largest factors of 2 in ki, under the Calabi-Yau
condition,Pr
i=1H
ki+2= H.
3. Let us present some examples that satisfy the Calabi-Yau and criticality conditions.
• (ki + 2) = (3, 3, 3); M = an elliptic curve, D = 7 + 1.
• (ki + 2) = (4, 4, 4, 4); M = a K3 surface. D = 5 + 1.
We will mainly consider the case with r = 5 and d = 1 since this corresponds to the string compactification to 3 + 1
dimensions. The examples of this type are
• (ki + 2) = (5, 5, 5, 5, 5); M = a quintic hypersurface in CP4.
• (ki + 2) = (8, 8, 4, 4, 4).
• (ki + 2) = (8, 8, 8, 8, 2).
• (ki + 2) = (12, 12, 6, 6, 2).
The first two will be our basic examples where we examine the general story in detail. A complete list can be found
in [25, 76, 69].
4. The non-chiral GSO projection of the minimal model Mk by (−1)F = e−πi(J0− eJ0) is the SU(2)k × U(1)2 mod U(1)
gauged WZW model, or simply SU(2)k × U(1)2/U(1)k+2 coset model. The latter model has primaries labeled by
(l, m, s) ∈ Mk; namely l ∈ Pk = 0, 1, ..., k, m ∈ Z2(k+2), s ∈ Z4, with l+m+s even, (l, m, s) ≡ (k−l, m+k+2, s+2).1
The product theory Mk1× · · · × Mkr should not be confused with the tensor product of the GSO projected models
of Mk1, ..., Mkr . In the latter the space of states would have mixture of NSNS and RR factors, while in the former
NS/R alignment is automatically imposed, as usual in ordinary supersymmetric quantum field theories.
5. The GSO projected model has global symmetries gn,s corresponding to simple currents (0, n, s) (n ∈ Z2(k+2), s ∈ Z4,
with n + s even) which act on the states in Hl′,m′,s′ ⊗ Hl′,−m′,−s′ as multiplication by a phase exp(πi(nm′
k+2− ss′
2)).
The symmetry g above induces one of them, g2,0.
6. “Gepner Model” usually refers to more general models based on orbifold of the product of minimal models. It doesn’t
have to come from linear sigma models of the above types. (In Appendix A of [9], we present more general models.)
In the main text of [9], we treat only the class of models introduced above, in particular the case D = 3+1 and r = 5.
In many cases, M has singularities that are inherited from the orbifold singularities of the ambient space, and
their resolution introduces extra Kahler parameters. This is accommodated in the linear sigma model by extending
the gauge group and adding charged fields. In general, the gauge group will be U(1)k =∏k
a=1 U(1)a gauge theory
with matter fields P,X1, · · · ,Xr+k−1 of certain charge QaP , Q
a1 , · · · , Qa
r+k−1 and certain (twisted) superpotential.
For example, for (ki + 2) = (8, 8, 4, 4, 4), the full system after the resolution has U(1)× U(1) gauge group and six
matter fields of the following charges [81]:
P X1 X2 X3 X4 X5 X6
U(1)1 −4 0 0 1 1 1 1
U(1)2 0 1 1 0 0 0 −2
(V.1.7)
1In [9], following the convention used by majority of people, the SU(2) spin j is labeled by L = 2j ∈ Pk = 0, 1, 2, · · · , k, rather
than j itself that is used in [77, 8].
V.1 CY data from An type LG superpotentials 87
The system has superpotential
W = P ·X4
6 (X81 +X8
2 ) +X43 +X4
4 +X45
,
and twisted superpotential
−W = τ1Σ1 + τ2Σ2 , (V.1.8)
where the τa = ra − iθa/2π and Σa = 1√2D+D−Va are the FI-Theta parameter and the fieldstrength of the U(1)a
gauge group. In the limit τ2 → −∞ with 2τ1 + τ2 fixed, X6 acquires a large absolute value and breaks the gauge
group except the one generated by (2i, i) ∈ u(1)1 ⊕ u(1)2. We are then left with the original system with one U(1)
gauge symmetry whose FI-Theta parameter is τ = 2τ1 + τ2. This corresponds to undoing the resolution.
Let us present the list of supersymmetric ground states of the system. The level k minimal model has (k + 1)
supersymmetric ground states | l 〉RR (l = 0, 1, · · · , k) which correspond to X l and have R-charges q = q = l+1k+2 − 1
2 .
Also, on a circle twisted by e−2πiνJ0 , there is a unique supersymmetric ground state | 0 〉ν which has R-charge
q = −q = lν+1k+2 − 1
2 where lν ∈ 0, 1, · · · , k is defined by
lν + 1 ≡ −ν (mod (k + 2)) . (V.1.9)
The RR ground states of the Gepner model are made of these states. Since the orbifold group is generated
by the tensor product of −e−2πiJ0 for the r = 5 factors, the condition is that the sum of R-charges is an
odd half-integer,∑
i qi ∈ 12 + Z. Untwisted sector states are thus the products ⊗5
i=1| li 〉RR with the condition∑
i( li+1
ki+2− 1
2) =
∑i
li
ki+2− 3
2∈ 1
2+ Z, or2
5∑
i=1
liki + 2
= 0, 1, 2, 3 . (V.1.10)
They correspond to harmonic forms of degree (3, 0), (2, 1), (1, 2) and (0, 3) respectively of the relevant Calabi-Yau
manifold3. There are also RR ground states from the twisted sectors labeled by ν = 1, 2, · · · ,H − 1. The orbifold
condition is the same as (V.1.10) where li is replaced by l(i)ν for such i where the twist is non-trivial, ν 6≡ 0 mod
(ki + 2). For the ν = 1 twist, we find l(i)1 = ki for all i and we find a unique ground state with q = −q = 3
2 . The
geometrical counterpart is the (0, 0)-form. For ν = (H − 1), we also find a unique ground state that corresponds to
the (3, 3)-form. The ground states from the twisted sectors are mostly related to (p, p)-forms. However, there can
be states corresponding to off-diagonal forms. For example, let us consider the case where H is even and twist by
ν = H2 . The twist in the i-th factor is non-trivial if and only if wi = H
ki+2 is odd. For such an i, l(i)H2
is ki
2 and the
ground state is qi = qi = 0. For other i, the twist is trivial and the ground states are ordinary ones | li 〉RR with
R-charges qi = qi = li+1ki+2 − 1
2 . They correspond to (2, 1) or (1, 2) forms. Let us show the number of ground states
in two examples.
2NS-NS (c, c) primary statesQ
i Xlii は ZH 変換に対して invariant であるべし、という要請である。
3 For non-linear sigma model on a Calabi-Yau manifold of dimension n, supersymmetric ground states with R-charge (q, eq) correspond
to harmonic (p, p) forms where (q, eq) = ( n2− p, p − n
2) or (p, p) = ( n
2− q, eq + n
2). これは appendix IV の議論と一致する。
88 Calabi-Yau Data from LG Superpotentials
(ki + 2) = (5, 5, 5, 5, 5)
Untwisted ground states correspond to monomials of Xi with degree 0, 5, 10, 15 (with relations X4i = 0) and there
are 1, 101, 101, 1 of them. Also there is a unique ground state ⊗i| 0 〉ν from each of ν = 1, 2, 3, 4 twisted sectors.
These numbers are organized into the “Hodge diamond”
b0,0
b1,0 b0,1
b2,0 b1,1 b0,2
b3,0 b2,1 b1,2 b0,3
b3,1 b2,2 b1,3
b3,2 b2,3
b3,3
=
1
0 0
0 1 0
1 101 101 1
0 1 0
0 0
1
(ki + 2) = (8, 8, 4, 4, 4)
X1,X2,X3,X4,X5 have weights wi = 1, 1, 2, 2, 2. Untwisted ground states correspond to monomials of Xi with
total weight 0, 8, 16, 24 (with relations X71 = X7
2 = X33 = X3
4 = X35 = 0). There are 1, 83, 83, 1 of them. There is
a unique ground state ⊗i| 0 〉ν from each of ν = 1, 2, 3, 5, 6, 7 twisted sectors. They corresponds to diagonal forms.
For the ν = 4 twisted sector, ground states are
⊗
i=1,2
| 0 〉ν ⊗⊗
i=3,4,5
| li 〉RR
where l3 + l4 + l5 = 1 (3 states) or 5 (3 states). The Hodge diamond is therefore
b0,0
b1,0 b0,1
b2,0 b1,1 b0,2
b3,0 b2,1 b1,2 b0,3
b3,1 b2,2 b1,3
b3,2 b2,3
b3,3
=
1
0 0
0 2 0
1 86 86 1
0 2 0
0 0
1
As usual [72], RR ground states are in one-to-one correspondence with chiral primaries by a spectral flow which
shifts the R-charge as q → q ± c6 , q → q ± c
6 . The spectral flow with the sign (++) maps the ground states to
NSNS states corresponding to chiral fields ((c, c)-fields), and the (−+)-spectral flow maps them to NSNS states
corresponding to twisted chiral fields ((a, c)-fields). They are marginal operators if q = q = 1. Marginal (c, c)
primaries correspond to (2, 1)-forms and marginal (a, c) primaries correspond to (1, 1)-forms.
V.2 A few examples 89
V.2 A few examples
先の sections で紹介したのは任意の次元の CY d-fold について適用できる。なお、この解説は結果のみを与えている
ので、CY 3-fold、K3 surface の例について具体的に計算しよう。
なお、ここでは全て Ak+1 type singularity (の direct products) を持つ LG superpotential から構成される物を
扱う。Ak+1 type については section II.A にあるように、superpotential W や central charge c は
W = Xk+2 , c = 3− 6
k + 2.
で与えられる。ここで X は chiral superfield であり、(c, c) field である。 direct product の場合はこれらの和を考
察する:
W =
r∑
i=1
Xki+2i , c =
r∑
i=1
ci =
r∑
i=1
(3− 6
ki + 2
).
各 Ak+1 sector において、monomial X l (l = 0, 1, · · · , k) (これは Ak+1 の chiral ring element であり、NS-NS
(c, c) primary states) に対応する untwisted R-R ground state (SUSY ground state) | l 〉RR は、appendix IV にある
ように spectral flow θL = θR = 12 を通じて与えられる。例えば、Ak+1 type superpotential W = Xk+2 における
NS-NS (c, c) primary state である X l は、left-/right U(1) charge は
(qL, qR)NS-NS =( l
k + 2,
l
k + 2
)(V.2.1a)
で与えられる。spectral flow θL = θR = 12 を行うと、対応する RR ground state | l 〉RR の U(1) charge は
(qL, qR)RR =( l
k + 2− c
6,
l
k + 2− c
6
)=( l + 1
k + 2− 1
2,l + 1
k + 2− 1
2
)(V.2.1b)
となる。同様に、⊗jAki+1 type LG minimal model において、NS-NS (c, c) primary state∏
iXlii の U(1) charge は
(qL, qR)NS-NS =( r∑
i=1
liki + 2
,
r∑
i=1
liki + 2
)(V.2.2a)
であり、spectral flow θL = θR = 12 を施して対応する R-R ground state ⊗i| li 〉RR を構成したとき、この U(1)
charge は
(qL, qR)RR =( r∑
i=1
( liki + 2
− ci6
),
r∑
i=1
( liki + 2
− ci6
))
=( r∑
i=1
( li + 1
ki + 2− 1
2
),
r∑
i=1
( li + 1
ki + 2− 1
2
)) (V.2.2b)
で与えられる。
left/right U(1) charge (qL, qR)RR を持つ R-R ground stateと、CY d-foldでのharmonic (p, p)-formは、appendix
IV の Figure IV.1から読み取れる様に(qL =
d
2− p , qR = p− d
2
)⇔
(p =
d
2− qL , p = qR +
d
2
)(V.2.3)
の関係がある。
以後の解析ではこの情報、特に (V.2.2) と (V.2.3) を用いる。
90 Calabi-Yau Data from LG Superpotentials
V.2.1 CY 3-fold: (ki + 2) = (5, 5, 5, 5, 5) (quintic hypersurface in CP4)
LG superpotential は
W =
5∑
i=1
X5i = X5
1 +X52 +X5
3 +X54 +X5
5 (V.2.4)
で与えられる。ki + 2 = 5 であるため、
H = LCMki + 2 = 5 (V.2.5)
となる。また、chiral ring C
[dW ] を考慮に入れるため、∂iW ≃ X4i ≡ 0 である。よって、1 つの A4 sector における
chiral ring element、つまり NS-NS (c, c) primary states X lii は
li = 0, 1, 2, · · · , ki = 0, 1, 2, 3 (V.2.6)
となる。
全 sector ⊗jAkj+1 を一度に考える。NS-NS (c, c) primary fields∏
iXlii に対応する untwisted RR ground states
⊗i| li 〉RR を考えよう。(V.2.2) より、⊗i| li 〉RR の U(1) charge (q, q) は
q = q =
5∑
i=1
( li + 1
ki + 2− 1
2
)∈ Z +
1
2(V.2.7)
であることがわかる。右辺は、NS-NS (c, c) primary states∏
iXlii が ZH = Z5 invariant であるべし、という要請か
らの帰結である。これをここで示しておこう。superpotential W が ZH invariant であることを要請するため、chiral
fields Xi は、ZH 変換で
Xi → ωXi , ω = e2πi/H
となる。従って NS-NS (c, c) primary states∏
iXlii は
∏
i
X lii → ωα
∏
i
X lii , α =
∑
i
li
であるが、この factor ωα が 1 となるべし、という条件が、NS-NS primary states の ZH invariance である:
ωα ≡ 1 →∑
i
li = Hn , n ∈ Z .
ここの模型では H = LCMki + 2 = ki + 2 = 5 である。よって、
5∑
i=1
liki + 2
= n
となる。spectral flow を採った後の RR ground state で見ると、
q =
5∑
i=1
liki + 2
− c
6= n− 3
2
V.2 A few examples 91
となる。よって q は半整数であるべし、という事が示された。具体的には、q の取り得る値、つまり n はRR ground
state の U(1) charge と CFT の unitarity との関係
|q| ≤ c
6
から有限個に制限されて、以下のような解が得られる:
q = q = −3
2, −1
2,
1
2,
3
2. (V.2.8)
それぞれの q の値を持つ RR ground states ⊗i| li 〉RR がいくつあって、それが CY 3-fold の harmonic form として
何に対応するのかを考察する。そのため、(V.2.7) を変形して
5∑
i=1
li = q +15
2= 0 , 5 , 10 , 15 (V.2.9)
を用意しておく。ground states の数は、これをみたす (l1, · · · , l5) の場合の数を検討すれば良い。
• q = q = − 32 ,∑
i li = 0:
これをみたす (li) は唯一
(li) = (0, 0, 0, 0, 0)
だけである。よってこの場合の RR ground stateは 1つだけである。これに対応する harmonic formは、(V.2.3)
に d = 3 を代入すると harmonic (3, 0)-form であることがわかる。
• q = q = − 12 ,∑
i li = 5:
これは harmonic (2, 1)-form に対応する。場合の数は
(li) =
(1, 1, 1, 1, 1) 1 個
(2, 1, 1, 1, 0) 5× 4 = 20 個
(2, 2, 1, 0, 0) 5× 4× 3 = 60 個
(3, 2, 0, 0, 0) 5× 4 = 20 個
であり、合計 101 個ある。
• q = q = 12 ,∑
i li = 10:
これは harmonic (1, 2)-form に対応する。場合の数は
(li) =
(2, 2, 2, 2, 2) 1 個
(3, 2, 2, 2, 1) 5× 4 = 20 個
(3, 3, 2, 1, 1) 5× 4× 3 = 60 個
(3, 3, 3, 1, 0) 5× 4 = 20 個
であり、合計 101 個ある。
92 Calabi-Yau Data from LG Superpotentials
• q = q = 32 ,∑
i li = 15:
これは harmonic (0, 3)-form に対応する。場合の数は
(li) = (3, 3, 3, 3, 3)
のみの 1 個である。
次に twisted RR ground states | 0 〉ν を考察する。ν は twisted number であり、一般に
ν = 0, 1, 2, · · · ,H − 1 (V.2.10)
で与えられる。ここの場合では ν = 0, 1, 2, 3, 4 である。なお、ν = 0 は untwisted になるので、ここでは考察から除
外する。
各 twisted sector ν には liν +1 = −ν (mod ki +2) が対応し、各 ν につき 1 つ対応する twisted RR ground state
⊗i| li 〉RR の U(1) charge は
q = −q =∑
i=5
( liν + 1
ki + 2− 1
2
)(V.2.11)
で与えられることは既に述べた。よって ν の値ごとに考察していこう。
• ν = 1:
この時、liν + 1 = −1 (mod 5) より、liν ∼ 3 となる。よって、
q = −q = 5(3 + 1
5− 1
2
)=
3
2
となる。この ground state | 0 〉1 は、(V.2.3) から harmonic (0, 0)-form に対応することがわかる。
• ν = 2:
この時 liν ∼ 2 (mod 5) となり、
q = −q = 5(2 + 1
5− 1
2
)=
1
2
となる。この ground state | 0 〉2 は、(V.2.3) から harmonic (1, 1)-form に対応することがわかる。
• ν = 3:
この時 liν ∼ 1 (mod 5) となり、
q = −q = 5(1 + 1
5− 1
2
)= −1
2
となる。この ground state | 0 〉3 は、(V.2.3) から harmonic (2, 2)-form に対応することがわかる。
• ν = 4:
この時 liν ∼ 0 (mod 5) となり、
q = −q = 5(0 + 1
5− 1
2
)= −3
2
となる。この ground state | 0 〉4 は、(V.2.3) から harmonic (3, 3)-form に対応することがわかる。
V.2 A few examples 93
これ以上 RR ground states は存在しない。harmonic form の数などが上の議論で確定したので、Hodge diamond
は次の様に与えられることがわかる:
b0,0
b1,0 b0,1
b2,0 b1,1 b0,2
b3,0 b2,1 b1,2 b0,3
b3,1 b2,2 b1,3
b3,2 b2,3
b3,3
=
1
0 0
0 1 0
1 101 101 1
0 1 0
0 0
1
V.2.2 CY 3-fold: (ki + 2) = (8, 8, 4, 4, 4)
(ki + 2) = (8, 8, 4, 4, 4) で与えられる CY 3-fold の考察は多少複雑である。
まず、LG superpotential を記載しておこう:
W =
5∑
i=1
Xki+2i = X8
1 +X82 +X4
3 +X44 +X4
5 . (V.2.12)
これは A7 ⊗A7 ⊗A3 ⊗A3 ⊗A3 である。この LG minimal model の central charge を与えよう:
c =
5∑
i=1
ci =
5∑
i=1
(3− 6
ki + 2
)= 9 . (V.2.13)
また、他の情報を列挙する:
H = LCMki + 2 = 8 , ν = 0, 1, 2, 3, · · · , 7 , (V.2.14a)
l1, l2 = 0, 1, 2, · · · , 6 , l3, l4, l5 = 0, 1, 2 , (V.2.14b)
(wi) =( H
ki + 2
)= (1, 1, 2, 2, 2) . (V.2.14c)
まずは untwisted RR ground states ⊗i| li 〉RR を考察しよう。これは NS-NS (c, c) primary states∏
iXlii を
spectral flow θL = θR = 12 して得られる。よって U(1) charge は
q = q =
5∑
i=1
liki + 2
− c
6=
5∑
i=1
liki + 2
− 3
2= −3
2, −1
2,
1
2,
3
2, (V.2.15)
である。右辺の値は、∏
iXlii が ZH = Z8 invariant であるべし (Z8 LG orbifold) という要請からの帰結である。こ
こでもこれを示しておこう。superpotential W の形状から、各 chiral fields Xi の ZH 変換は次のようになる:
(X1,X2,X3,X4,X5
)→
(ωX1, ωX2, ω
2X3, ω2X4, ω
2X5
), ω = e2πi/H .
NS-NS primary states∏
iXlii が ZH invariant であることから
∏
i
X lii → ωα
∏
i
X lii ≡
∏
i
X lii , α =
∑
i
Hliki + 2
94 Calabi-Yau Data from LG Superpotentials
という要請が課される。つまり
∑
i
Hliki + 2
≡ Hn , n ∈ ZH
である。これを U(1) charge q for RR ground state で書き直すと、spectral flow より
q =∑
i
liki + 2
− c
6≡ n− 3
2
となる。よって上の「半整数条件」が示された。
RR ground state の U(1) charge と CFT の unitarity との関係
|q| ≤ c
6
が存在するため、整数 n は具体的に有限個に制限される。なおこれと li の関係式の並立から得られる解は
(l1 + l2) + 2(l3 + l4 + l5) = 0, 8, 16, 24 , (V.2.16)
となる。それぞれの解について考察しよう。
• q = − 32 , (l1 + l2) + 2(l3 + l4 + l5) = 0:
対応関係 (V.2.3) より、これは harmonic (3, 0)-form を与える。これをみたす (li) は
(li) = (0, 0, 0, 0, 0)
ただ 1 つである。
• q = − 12 , (l1 + l2) + 2(l3 + l4 + l5) = 8:
これは harmonic (2, 1)-form を与える。これをみたす (li) を考えよう。
l3 + l4 + l5 ≡ L , l1 + l2 ≡ 8− 2L
0 ≤ L ≤ 4 , 0 ≤ 8− 2L ≤ 8
と記述する (L の上限は 8− 2L の下限から決定される)。これらについて、(li) の場合の数を考える:
L =
0 5 通り
1 3× 7 = 21 通り
2 (3 + 3)× 5 = 30 通り
3 (6 + 1)× 3 = 21 通り
4 (3 + 3)× 1 = 6 通り
よって、合計 83 個ある。
• q = 12 , (l1 + l2) + 2(l3 + l4 + l5) = 16:
これは harmonic (1, 2)-form を与える。これをみたす (li) を考えよう。
l3 + l4 + l5 ≡ L , l1 + l2 ≡ 16− 2L
V.2 A few examples 95
2 ≤ L ≤ 6 , 4 ≤ 16− 2L ≤ 12
と記述する (16− 2L の下限は L の上限から決定される)。これらについて、(li) の場合の数を考える:
L =
2 (3 + 3)× 1 = 6 通り
3 (6 + 1)× 3 = 21 通り
4 (3 + 3)× 5 = 30 通り
5 3× 7 = 21 通り
6 1× 5 = 5 通り
よって、合計 83 個ある。
• q = 32 , (l1 + l2) + 2(l3 + l4 + l5) = 24:
これは harmonic (0, 3)-form を与える。これをみたす (li) は
(li) = (6, 6, 2, 2, 2)
ただ 1 つである。
次に twisted RR ground states | 0 〉ν (ν = 0, 1, · · · , 7) を考える。但し ν = 0 は untwisted なので除外する。
twisted ground states | 0 〉ν の U(1) charge は
q = −q =5∑
i=1
( liν + 1
ki + 2− 1
2
), liν + 1 = −ν (mod ki + 2) (V.2.17)
で与えられる。但しこれが適用出来るのは ν 6= H2 である ν = 1, 2, 3, 5, 6, 7 の場合である。ν = 4の場合は後で考える。
• ν = 1:
l1ν = l2ν = −2 ∼ 6 (mod 8)、かつ l3ν = l4ν = l5ν = −2 ∼ 2 (mod 4) なので、
q = −q =3
2
となる。(V.2.3) より、これは harmonic (0, 0)-form に対応する。
• ν = 2:
l1ν = l2ν = −3 ∼ 5 (mod 8)、かつ l3ν = l4ν = l5ν = −3 ∼ 1 (mod 4) なので、
q = −q =1
2
となる。(V.2.3) より、これは harmonic (1, 1)-form に対応する。
• ν = 3:
l1ν = l2ν = −4 ∼ 4 (mod 8)、かつ l3ν = l4ν = l5ν = −4 ∼ 0 (mod 4) なので、
q = −q = −1
2
となる。(V.2.3) より、これは harmonic (2, 2)-form に対応する。
96 Calabi-Yau Data from LG Superpotentials
• ν = 5:
l1ν = l2ν = −6 ∼ 2 (mod 8)、かつ l3ν = l4ν = l5ν = −6 ∼ 2 (mod 4) なので、
q = −q =1
2
となる。(V.2.3) より、これは harmonic (1, 1)-form に対応する。
• ν = 6:
l1ν = l2ν = −7 ∼ 1 (mod 8)、かつ l3ν = l4ν = l5ν = −7 ∼ 1 (mod 4) なので、
q = −q = −1
2
となる。(V.2.3) より、これは harmonic (2, 2)-form に対応する。
• ν = 7:
l1ν = l2ν = −8 ∼ 0 (mod 8)、かつ l3ν = l4ν = l5ν = −8 ∼ 0 (mod 4) なので、
q = −q = −3
2
となる。(V.2.3) より、これは harmonic (3, 3)-form に対応する。
• ν = 4:
X1, X2 は wi = 1 で、odd であるので、i = 1, 2 に関する twist は non-trivial となる。そのため li=1,24 = ki
2
となる。i = 1, 2 sector の U(1) charge は
qi = qi =li4 + 1
ki + 2− 1
2= 0
となる。
一方、i = 3, 4, 5 sector に関しては、wi = 2 と even であるので、この sector の twist は trivial (untwisted) と
なる。RR ground states | li 〉RR の U(1) charge は
q = q =li + 1
ki + 2− 1
2, li = 0, 1, 2
である。
これらをまとめると、ν = 4 での RR ground states は
⊗
i=1,2
| 0 〉ν=4 ⊗⊗
i=3,4,5
| li 〉RR
となり、U(1) charge は
q = q =
5∑
i=1
qi =
5∑
i=3
( li + 1
4− 1
2
)= ±1
2, li = 0, 1, 2 ,
と与えられる。
V.3 CY data from Dk type LG superpotentials 97
q = q = 12 を与える RR ground state は (V.2.3) harmonic (1, 2)-form に対応する。この数は
l3 + l4 + l5 = 5
をみたす場合の数だけ、つまり 3 個ある。一方、q = q = − 12 を与える RR ground state は (V.2.3) harmonic
(2, 1)-form に対応する。この数は
l3 + l4 + l5 = 1
をみたす場合の数だけ、つまり 3 個ある。
上の議論を全てまとめると、Hodge diamond が完成する:
b0,0
b1,0 b0,1
b2,0 b1,1 b0,2
b3,0 b2,1 b1,2 b0,3
b3,1 b2,2 b1,3
b3,2 b2,3
b3,3
=
1
0 0
0 2 0
1 86 86 1
0 2 0
0 0
1
, 86 = 83 + 3 .
V.2.3 K3 surface: (ki + 2) = (4, 4, 4, 4)
(ki + 2) = (4, 4, 4, 4) で与えられる K3 surface は、上の 2 つの考察を簡単に混ぜあわせた様な物で議論できる。もは
や計算は書かず、結論だけ記載しておこう:
b0,0
b1,0 b0,1
b2,0 b1,1 b0,2
b2,1 b1,2
b2,2
=
1
0 0
1 20 1
0 0
1
, 20 = 19 + 1
RR ground states ⊗j | lj 〉RR の議論は、例えば Hori and Vafa [54] section 7 でも登場する。
V.3 CY data from Dk type LG superpotentials
Dk type LG superpotentialを用いた CY 3-foldの例を考えよう。最も簡単だと思われる物として D5⊗D5⊗D5⊗D5、
LG superpotential としては
W =
4∑
i=1
X4
i +XiY2i
(V.3.1)
がある。これを選んだ理由は、全ての rank が同じで、全体の central charge が ctotal = 9 になるべしという条件を
みたすからである。同様の条件で D3 ⊗D3 ⊗D3 ⊗D3 ⊗D3 ⊗D3 も得られるが、これは field の数が多いので扱わ
ないことにする。
Chapter VI
Mirror Symmetry: T-duality of GLSM
100 Mirror Symmetry: T-duality of GLSM
ここでは Hori-Vafa [54] に従って新たな Lagrangian を構成する。基本的には、GLSM の chiral superfield を適
当に組み直して twisted chiral superfield にし、GLSM Lagrangian を twisted chiral field のみで記述することにな
る。具体的な構成は後で述べるが、T-dual を採ることに相当する。この T-dualized theory は GLSM の mirror に
なっていることを Hori-Vafa は示したので、これの review を行う。
VI.1 T-duality theories on tori
ここでは入門的に Hori-Vafa [54] section 2.2 の記述をそのまま掲載しよう。非常に基本的でためになると思われる。
In the case of supersymmetric sigma model on a flat torus, it has been known that mirror symmetry reduces
to the R → 1/R duality performed on a middle dimensional torus. Below, we review this in the simplest case of
sigma model into the algebraic torus C∗ = R× S1. We start with recalling the bosonic R→ 1/R duality.
R→ 1/R Duality
Let us consider the following action for a periodic scalar field ϕ of period 2π
Sϕ =R2
4π
∫
W
d2x√hhµν∂µϕ∂νϕ (VI.1.1)
where hµν is the metric on the worldsheet W which we choose to be of Euclidean signature. This is the action for
a sigma model into a circle S1 of radius R. This action can be obtained also from the following action for ϕ and a
one-form field Bµ
S′ =1
2π
1
2R2
∫
W
d2x√hhµνBµBν +
i
2π
∫
W
B ∧ dϕ . (VI.1.2)
Completing the square with respect to Bµ which is solved by
B = iR2 ∗ dϕ , (VI.1.3)
and integrating it out, we obtain the action (VI.1.1) for the sigma model.
If, changing the order of integration, we first integrate over the periodic scalar ϕ, we obtain a constraint dB = 0.
If the worldsheet W is a genus g surface, there is a 2g-dimensional space of closed one-forms modulo exact forms1.
One can choose a basis ωi (i = 1, · · · , 2g) such that each element has integral periods on one-cycles on W and that∫
Wωi ∧ ωj = J ij is a non-degenerate matrix of integral entry. Then, a general solution to dB = 0 is
B = dϑ0 +
2g∑
i=1
aiωi , (VI.1.4)
where ϑ0 is a real scalar field and ai’s are real numbers. Integration over ϕ actually yields constraints on aj ’s
as well. Recall that ϕ is a periodic variable of period 2π. This means that ϕ does not have to come back to
its original value when circling along a nontrivial one-cycles in W , but comes back to itself up to 2π shifts. For
1This can be easily extended to the case of worldsheets with boundaries.
VI.1 T-duality theories on tori 101
such a topologically nontrivial configuration, dϕ has an expansion like (VI.1.4) with non-zero coefficient ai for ωi
which is dual to the one-cycle. That the shift is only allowed to take integer multiples of 2π means that such ai is
constrained to be 2πni where ni is an integer. Thus, for a general configuration of ϕ we have
dϕ = dϕ0 +
2g∑
i=1
2πniωi , (VI.1.5)
where ϕ0 is a single valued function on W . Now, integration over ϕ means integration over the function ϕ0 and
summation over the integers ni’s. Integration over ϕ0 yields the constraint dB = 0 which is solved by (VI.1.4).
What about the summation over ni’s? To see this we substitute in∫B ∧ dϕ for B from (VI.1.4);
∫
W
B ∧ dϕ = 2π∑
i,j
aiJijnj . (VI.1.6)
Now, noting that J ij is a non-degenerate integral matrix and using the fact that∑
n eian = 2π∑
m δ(a − 2πm)
(the Poisson resummation formula), we see that summation over ni constrains ai’s to be an integer multiples of 2π;
ai = 2πmi , mi ∈ Z . (VI.1.7)
Inserting this into (VI.1.4), we see that B can be written as
B = dϑ , (VI.1.8)
where now ϑ is a periodic variable of period 2π. Now, inserting this to the original action we obtain
Sϑ =1
4π
1
R2
∫
W
dx2√hhµν∂µϑ∂νϑ (VI.1.9)
which is an action for a sigma model into S1 of radius 1/R.
Thus, we have shown that the sigma model into S1 of radius R is equivalent to the model with radius 1/R.
This is the R→ 1/R duality (which is called target space duality or T-duality in string theory).
Comparing (VI.1.3) and (VI.1.8), we obtain the relation
R dϕ =i
R∗ dϑ . (VI.1.10)
Since Rdϕ and iR∗dϕ are the conserved currents in the original system that count momentum and winding number
respectively, the relation (VI.1.10) means that momentum and winding number are exchanged under the R→ 1/R
duality. In particular, the vertex operator
exp(iϑ) (VI.1.11)
that creates a unit momentum in the dual theory must be equivalent to an operator that creates a unit winding
number in the original theory. This can be confirmed by the following path integral manipulation. Let us consider
the insertion of
exp(− i∫ q
p
B)
(VI.1.12)
102 Mirror Symmetry: T-duality of GLSM
in the system with the action (VI.1.2), where the integration is along a path τ emanating from p and ending on q.
Then, using (VI.1.8) we see that
exp(− i∫ q
p
B)
= e−iϑ(q) eiϑ(p) . (VI.1.13)
On the other hand, the insertion of e−iR
q
pB changes the B-linear term in (VI.1.2). We note that
∫ q
pB can be
expressed as∫
WB ∧ω, where ω is a one-form with delta function support along the path τ . This ω can be written
as ω = dθτ where θτ is a multi-valued function on W that jumps by one when crossing the path τ . Now, the
modification of the action (VI.1.2) can be written as
i
2π
∫
W
B ∧ dϕ −→ i
2π
∫
W
B ∧ dϕ+ i
∫ q
p
B =i
2π
∫
W
B ∧ d(ϕ+ 2πθτ ) . (VI.1.14)
Integrating out Bµ, we obtain the action (VI.1.1) with ϕ replaced by ϕ′ = ϕ + 2πθτ . Note that ϕ′ jumps by 2π
when crossing the path τ which starts and ends on p and q. In particular, it has winding number 1 and −1 around
p and q respectively. Thus, the insertion of eiϑ creates the unit winding number in the original system.
Mirror Symmetry as R→ 1/R Duality
We now proceed to a supersymmetric sigma model on the algebraic torus, or the cylinder C∗ = R× S1. We show
that R → 1/R duality performed on the S1 factor is indeed a mirror symmetry. We work now in Minkowski
signature.
We denote the complex coordinate of the cylinder R× S1 as
φ = + iϕ (VI.1.15)
where is the coordinates of R and ϕ is the periodic coordinate of S1 of period 2π. The Lagrangian of the system
is
L =
∫d4θ
R2
2|Φ|2 =
R2
2
(−ηµν∂µφ∂νφ+ iψ−(∂0 + ∂1)ψ− + iψ+(∂0 − ∂1)ψ+
), (VI.1.16)
where Φ is the chiral superfield whose lowest component is φ. The Kahler metric for φ is ds2 = R2|dφ|2 =
R2(d2 + dϕ2) so that S1 has radius R.2
We perform the duality transformation on ϕ. As we have seen, this yields another periodic variable ϑ of period
2π with the Kinetic term −(1/2R2)ηµν∂µϑ∂νϑ. Thus, the dual theory is also a sigma model into a cylinder, but
with a metric
ds2 = R2d2 +1
R2dϑ2 =
1
R2
(R4d2 + dϑ2
). (VI.1.17)
Thus, either R2 + iϑ or R2 − iϑ is the complex coordinates of the new cylinder. What is the superpartner
of this (anti-)holomorphic variable? We note the supersymmetry transformations δψ± = −i√
2(∂0 ± ∂1)φǫ± and
2 In this paper, we take the convention S = 12π
Rd2xL as the relation of the action and the Lagrangian. Thus, the weight factor
in Path-Integral is exp( i2π
Rd2xL ) (in Minkowski signature).
VI.1 T-duality theories on tori 103
δψ± = i√
2(∂0±∂1)φǫ±. From (VI.1.10), we see (after continuation back to Minkowski signature by x2 = ix0) that
R2(∂0 ± ∂1)ϕ = ∓(∂0 ± ∂1)ϑ and therefore R2(∂0 + ∂1)φ = (∂0 + ∂1)η and R2(∂0 − ∂1)φ = (∂0 − ∂1)η where
η = R2− iϑ . (VI.1.18)
Thus, the supersymmetry transformation is expressed as
R2δψ+ = −i√
2(∂0 + ∂1)ηǫ+ , R2δψ+ = i
√2(∂0 + ∂1)ηǫ
+ , (VI.1.19)
R2δψ− = −i√
2(∂0 − ∂1)ηǫ+ , R2δψ− = i
√2(∂0 + ∂1)ηǫ
− . (VI.1.20)
This is not a supersymmetry transformation for a chiral multiplet, but that for a twisted chiral multiplet. Indeed,
renaming the fermions as
R2ψ± = ±χ± , R2ψ± = ±χ± , (VI.1.21)
the Lagrangian for the dual theory becomes∫
d4θ(− 12R2 |Θ|2) for a twisted chiral superfield Θ = η +
√2(θ+χ+ +
θ−χ−) + · · · . Thus, we have seen that R→ 1/R duality on S1 transforms a theory of a chiral multiplet to another
theory of a twisted chiral multiplet. Thus, this is a mirror symmetry.
The above manipulation can be simplified by performing the dualization in superspace. We follow the procedure
developed in [93]. We start with the following Lagrangian for a real superfield B and a twisted chiral superfield Θ.
L′ =
∫d4θ
(R2
4B2 − 1
2(Θ + Θ)B
)(VI.1.22)
We first integrate over the twisted chiral field Θ, Θ. This yields the following constraint on B
D+D−B = D+D−B = 0 , (VI.1.23)
which is solved by
B = Φ + Φ , (VI.1.24)
where Φ is a chiral superfield. Now, inserting this into the original Lagrangian we obtain the Lagrangian (VI.1.16)
L =
∫d4θ
R2
4(Φ + Φ)2 =
∫d4θ
R2
2ΦΦ . (VI.1.25)
for the sigma model into the cylinder with radius R on S1. Now, reversing the order of integration, we consider
integrating out B first. Then, B is solved by
B =1
R2(Θ + Θ) . (VI.1.26)
Inserting this into L ′ we obtain
L =
∫d4θ
(− 1
2R2ΘΘ
), (VI.1.27)
which is again the Lagrangian for supersymmetric sigma model on the cylinder. This time, the radius of S1 is
1/R and the complex coordinate is described by the twisted chiral superfield Θ. From (VI.1.24) and (VI.1.26),
we obtain R2(Φ + Φ) = Θ + Θ which reproduces the relation between the component fields obtained above (e.g.
(VI.1.21)).
104 Mirror Symmetry: T-duality of GLSM
VI.2 General formulation
Hori-Vafa の T-dualized theory [54] を考察しよう。
T-dualized Lagrangian
ここでは U(1) gauge symmetryが一つ (よって vector superfield V は一種類)の記述を行うが、multi-gauge symmetric
theory も本質は同じなので、ここでは省略する。
ではまず、次のような Lagrangian を用意しよう:
L′ =
∫d4θ
− 1
e2ΣΣ +
∑
i
(e2QiV +Bi − 1
2(Yi + Yi)Bi
)+(− τ√
2
∫d2θΣ + (h.c.)
). (VI.2.1)
ここで Yi は auxiliary twisted chiral superfield (D+Yi = D−Yi = 0) であり、period 2π の周期関数 (Yi(x1 + 2π) ≡
Yi(x1) + 2πin) とする (n は winding number)。また Bi は auxiliary real superfield (Bi = Bi) である。もちろんこ
こでも CY condition (I.2.1) を課しておく。
この Lagrangian (VI.2.1) から、一方では GLSM を、他方では T-dualized Lagrangian を導出する。これらは、
導入されている auxiliary field Yi, Bi の integrate out の順番の違いで登場する。
ではまずL ′ の twisted chiral superfield Yi を integrate outしよう。ここで measureを super-covariant derivative
に書き換える作業を行う:
d4θ = −d2θ d2θ =1
2d2θ D+D− (VI.2.2)
これを用いると∫
d4θ∑
i
(Yi + Y i)Bi = −1
2
∫d2θ D+D−
∑
i
YiBi + (h.c.)
= −1
2
∫d2θ
∑
i
(D+D−Yi)Bi + Yi(D+D−Bi)
+ (h.c.)
= −1
2
∫d2θ
∑
i
Yi(D+D−Bi) + (h.c.) . (VI.2.3)
ここで D+Yi = D−Yi = 0 を用いた。これにより、Lagrangian L ′ から Yi を integrate out すると D+D−Bi = 0,
D+D−Bi = 0 が得られる。この解は
Bi = Ψi + Ψi , (VI.2.4)
である。但し Ψi は chiral superfield である。これを (VI.2.1) に代入すると GLSM そのものが得られる:
L = L′∣∣∣Bi=Ψi+Ψi
=
∫d4θ
− 1
e2ΣΣ +
∑
i
e2QiV +Ψi+Ψi
+(− τ√
2
∫d2θΣ + (h.c.)
)
=
∫d4θ
− 1
e2ΣΣ +
∑
i
Φi e2QiV Φi
+(− τ√
2
∫d2θΣ + (h.c.)
). (VI.2.5)
但し chiral superfield の性質を用いて Φi = eΨi と書き換えてある。
VI.2 General formulation 105
次に Lagrangian L ′ の real superfield Bi を integrate out したものを考察する。integrate out の解は
e2QiV +Bi =1
2(Yi + Y i) → Bi = −2QiV + log
(Yi + Yi
2
). (VI.2.6)
これを (VI.2.1) に代入する:
L = L′∣∣∣eq.(VI.2.6)
=
∫d4θ
− 1
e2ΣΣ +
∑
i
1
2(Yi + Y i)
(1 + 2QiV − log(Yi + Y i)
)+(− τ√
2
∫d2θΣ + (h.c.)
). (VI.2.7)
ここで Σ = 1√2D+D−V であることを用いると
∫d4θ QiV Yi = +
1
2Qi
∫d2θ D+D−V Yi = +
1
2Qi
∫d2θ (D+D−V )Yi +
1
2Qi
∫d2θ V (D+D−Yi) (VI.2.8)
= +1√2Qi
∫d2θΣYi , (VI.2.9)
であり、また∫
d4θYi = 0 や irrelevant term を省くと、最終的に次の Lagrangian を得る:
L =
∫d4θ
− 1
e2ΣΣ−
∑
i
(1
2(Yi + Y i) log(Yi + Y i)
)+ 1√
2
∫d2θ W + (h.c.)
, (VI.2.10a)
W = Σ(∑
i
QiYi − τ). (VI.2.10b)
この Lagrangian の scalar field が GLSM (VI.2.5) の scalar field と T-dual の関係にある [54]。また、auxiliary field
Bi を通じて、chiral superfield Φi と twisted chiral superfield Yi とは
Yi + Yi = 2Φi e2QiV Φi , (VI.2.11)
の関係がある。またこれより、on-shell D field の関係式 (I.1.9a) も twisted chiral superfield で書き直す事が出来る:
r =∑
i
Qi|Φi|2 →∑
i
QiYi = τ . (VI.2.12)
これは twisted superpotential W = Σ(∑
iQiYi− τ) に登場し、massive gauge field Σ を integrate out することで得
る事が出来る。つまり massive field を integrate out して得られる low energy theory には常に constraint (VI.2.12)
が課される。
しかし実はこれだけでは少し足りない。[54] にもあるように、axial U(1) R-symmetry が一般に anomalous であ
る (CY condition が成立する時はそうではない) ことから、theta angle θ が変換される効果 (instanton effect) が存
在し、twisted superpotential が dynamical に生成される。これを考慮に入れなければならない。ここではそれを簡
単に紹介しよう。生成される twisted superpotential を∆Wi と記述しよう (U(1) gauge symmetry の flavor 毎に定
義される。)。
aixal U(1) R-symmetry に伴う変換則は
Σ(θ±, θ±) → e2iαΣ(e∓iαθ±, e±iαθ±) , Yi(θ±, θ±) → Yi(e
∓iαθ±, e±iαθ±)− 2iα , (VI.2.13a)
τ → τ , Λ ≡ µ e−τ/Q → Λ , (VI.2.13b)
106 Mirror Symmetry: T-duality of GLSM
である。ここで Λ は dynamical scale parameter、µ は繰り込み点であり、complexified FI parameter τ とは
τ(µ) =
∑iQi log(µ/Λ)− iθ if
∑iQi 6= 0
r − iθ if∑
iQi = 0(VI.2.14)
という関係がある。また Q =∑
iQi である。このままでは twisted superpotential W = Σ(τ−∑iQiYi)は anomalous
term を出してしまう。そこでmodified axial U(1) R-symmetry transformation を次のように定義しよう:
Σ(θ±, θ±) → e2iαΣ(e∓iαθ±, e±iαθ±) , Yi(θ±, θ±) → Yi(e
∓iαθ±, e±iαθ±)− 2iα , (VI.2.15a)
τ → τ − 2iαQ . (VI.2.15b)
形式的に Y i ≡ Yi − t/Q を導入しておくと、
Σ → e2iαΣ , Λ → e2iαΛ , e−eY i → e−
eY i , (VI.2.16)
と表す事が出来る。
T-dualized theory で登場する field や parameter はこれで全てなので、これを用いて ∆W i を書き下そう。su-
persymmetric かつ R-charge は (qV , qA) = (0, 2) であるべきなどの一般的な条件から
∆Wi ≡ Σ∑
n,m
cn,m (Λ/Σ)n e−meY i
=∑
n,m
cn,m µn · Σn−m e−(n−m)t/Q e−meY i , n,m ∈ Z , (VI.2.17)
と書く事が可能である (cn,m は dimensionless constant)。ここで
Σ→ 0 limit で ∆Wi が pole を持たない : 1− n ≥ 0 , (VI.2.18)
r →∞ limit で ∆Wi ≪ W : n−m ≥ 0 , (VI.2.19)
Yi is not bounded in real direction : m ≥ 0 (VI.2.20)
という条件から3、解は
(n,m) = (0, 0) , (1, 0) , (1, 1) , (VI.2.21)
のみであることがわかる。しかし (0, 0) 解は∆W i = Σ となって W と同じ order で寄与するために排除され、(1, 0)
解は constant solution であるので適用されず、結局 (1, 1) 解のみが許される事になる。よって
∆Wi = c1,1 µ e−Yi (VI.2.22)
が得られる。ここで c1,1 6= 0 であることは保証される (see ref. [54]) がここでは議論しない。よって適当に c1,1 を
rescale して、くりこみを考慮する事で (ここでは CY condition にこだわらない一般的な表式である)、最終的な
T-dualized theory は
L =
∫d4θ
− 1
e2ΣΣ−
∑
i
(1
2(Yi + Y i) log(Yi + Y i)
)+ 1√
2
∫d2θ W + (h.c.)
, (VI.2.23a)
3[54] 参照。
VI.2 General formulation 107
W = Σ(∑
i
QiYi − τ(µ))
+ µ∑
i
e−Yi , Yi = Yi0 − log(ΛUV
µ
), (VI.2.23b)
となる4。この formulation では modified axial U(1) R-symmetry でなくても問題が起こる term は存在しなくなる。
実際に、(VI.2.23) から全ての twisted chiral superfield Yi を integrate out した Lagrangian は、(VI.2.5) から全て
の chiral superfield Φi を integrate out したものと一致する。
ここからの議論はこの (VI.2.23) の low energy theory の解析である。
low energy limit
[54] によると ΛUV/µ→∞ limit では Yi の scalar field yi の住む空間 (やはり Kahler manifold らしいが、きちんと
やっていない) が flat になる:
Yi0 = log(ΛUV
µ
)+ Yi , (VI.2.24a)
ds2 =∑
i
|dyi|22(2 log(ΛUV/µ) + yi + yi)
→∑
i
|dyi|24 log(ΛUV/µ)
. (VI.2.24b)
よって (VI.2.23) の kinetic term は gauge field Σ も含めて flat であるとしてよい。また、(VI.2.23) を component
field で全て書き下すとわかる (はずだ) が、gauge field Σ は質量 m2 ∼ e2r を持ち、(VI.2.12) に沿った方向 (flat
direction) の field は massless で (もしくは m2 ≃ µ2r, µ ∼ Λ) 登場する。よって low energy limit5 e ≫ µ では
m2 ≃ e2r の質量を持つ field は全て integrate out されて、effective Lagrangian は
Lac =
∫d4θ
1
4r
∑
a
|Ya|2 +( 1√
2
∫d2θ W + (h.c.)
), (VI.2.25a)
W = µ∑
i
e−Yi with constraint∑
i
QiYi = τ(µ) , (VI.2.25b)
として得られる。ここで index a は、twisted chiral superfield Yi のうちmassless field として low energy theory で
も生き残っているものだけを表す。
この Lagrangian (VI.2.23) は、GLSM に superpotential W が存在していたか否かを反映していない。この W
の有無が T-dualized theory にどう反映されるのかは、Hori-Vafa [54] の section 4, 6, 7 を参考にする必要がある。
U(1) gauge multiplet である twisted chiral superfield Σ を integrate out し、low energy limit を採ると、それは
twisted chiral field で書かれる LG theory (LGac theory) となる。それは次に記載する twisted superpotential W で
記述される:
W = µ∑
i
e−Yi with constraint∑
i
QiYi = τ . (VI.2.26)
ここで係数 µ は renormalization scale であるが、Hori-Vafa は往々にして省略 (もしくは field に吸収)する。
4Yi の表式、繰り込み点の導入など [54] で詳細を確認する事。ここでは繰り込みの議論の詳細は無視して、結果のみ用いる。5We call “sigma model limit” in [54].
108 Mirror Symmetry: T-duality of GLSM
VI.2.1 Vortex and fermionic zero modes
The twisted superpotential (VI.2.10b) obtained from dualization is an exact expression in perturbation theory with
resepect to 1/r; it is simply impossible to write down a perturbative correction that respects the R-symmetry
and/or anomaly, holomorphy in τ , and periodicity of Theta angle. The D-term is of course subject to perturbative
corrections.
However, the twisted superpotential is possibly corrected by non-perturbative effects. A typical non-perturbative
effect in quantum field theory is by the presence of instantons. The bosonic part of our theory is an abelian Higgs
model which can have an instanton configuration — vortex. It has been known that in an abelian Higgs model a
Theta dependent vacuum energy density is generated by the effect of the gas of vortices and anti-vortices [12]. As
in that case, and also as in Polyakov’s model of confinement where a bosonic potential for the dual field is generated
from the gas of monopoles and anti-monopoles, we expect that a superpotential for Yi’s can be generated by the
gas of vortices and anti-vortices.
Around the vortex for a charged scalar φi, the phase of φi has winding number one. As we have seen in R→ 1/R
duality, a winding configuration is dual to the insertion of the vertex operator eiϑi . The supersymmetric completion
of this operator is the twisted chiral superfield
e−Yi . (VI.2.27)
These exponentials have vector R-charge 0 and axial R-charge 2. Thus, we can add these to the twisted superpo-
tential without violating the U(1)V R-symmetry, and maintaining the correct anomaly of U(1)A.
In what follows we shall show that a correction of the form (VI.2.27) is indeed generated. In fact we will show
that the correction is simply the sum of them and the exact superpotential is given by
W = Σ
(N∑
i=1
QiYi − τ(µ)
)+ µ
N∑
i=1
e−Yi . (VI.2.28)
This is one of the main results of this paper. Note that the change of the renormalization scale µ can be absorbed
by the shift of Yi’s. The actual parameter of the theory is still the dynamical scale Λ for∑
iQi 6= 0 and the
FI-Theta parameter τ for∑
iQi = 0.
We continue our gauge theory to the Euclidean signature by Wick rotation x0 = −ix2. We choose the orientation
so that z = x1+ix2 is the complex coordinates. This leads to F01 → −iF12, D0+D1 → 2Dz and D0−D1 → −2Dz.
After solving for the auxiliary field as D = −e2(|φ|2 − r0) and F = 0, the Euclidean action is given by
SE =1
2π
∫d2x|Dµφ|2 + |σφ|2 +
1
2e2|∂µσ|2 +
1
2e2(F 2
12 +D2) + iθF12
− 2iψ−Dzψ− + 2iψ+Dzψ+ + ψ−σψ+ + ψ+σψ−
+1
e2(− iλ−∂zλ− + iλ+∂zλ+
)
+ i(φ†λ−ψ+ − φ†λ+ψ− − ψ+λ−φ+ ψ−λ+φ
).
(VI.2.29)
An instanton is a topologically non-trivial configuration that minimizes the bosonic part of this action.
VI.2 General formulation 109
An instanton can contribute to the (twisted) superpotential only when it carries two fermionic zero modes of
the right kind. Since a twisted F-term is obtained by the integration over two fermionic coordinates other than
θ− and θ+, a relevant configuration must be invariant under the supercharges Q− and Q+. The invariance of the
fermions under these supercharges requires
σ = 0 , (VI.2.30a)
Dzφ = 0 , (VI.2.30b)
F12 = e2(|φ|2 − r0) . (VI.2.30c)
The bosonic part of the action is 12π
∫d2x( 1
2e2 |∂µσ|2 + |σφ|2) plus
1
2π
∫d2x
(|Dµφ|2 +
1
2e2(F 2
12 +D2) + iθF12
)
=1
2π
∫d2x
(|2Dzφ|2 − F12|φ|2 +
1
2e2(F12 +D)2 − 1
e2DF12 + iθF12
)
=1
2π
∫d2x
(|2Dzφ|2 +
1
2e2(F12 +D)2
)− τ0
2π
∫d2xF12 ,
(VI.2.31)
where D = −e2(|φ|2 − r0) and τ0 = r0 − iθ. For a given topological number
k = − 1
2π
∫d2xF12 , (VI.2.32)
the real part of the action is bounded by kr0, and the minimum is indeed attained by a solution to the equations
(VI.2.30a)–(VI.2.30c). The value of the action for such an instanton is
SE = kτ0 . (VI.2.33)
Under the axial rotation by eiα, the path-integral measure in this topological sector changes by the phase e2ikα.
Since the twisted superpotential has axial R-charge 2, we see that the relevant configurations are those with k = 1.
A solution to (VI.2.30b) and (VI.2.30c) is the vortex. For each vortex with k = 1, φ has a single simple zero.
The moduli space of gauge equivalence classes of k = 1 vortices is complex one-dimensional and is parametrized
by the location of the zero of φ. To see this, we note the following well-known fact. The orbit of a solution to
(VI.2.30b) under the complexified gauge transformations contains vortex solutions in one gauge equivalence class,
and conversely, any gauge equivalence class of vortex solutions is contained in one such orbit. Here, a complexified
gauge transformation is a rotation (iAz, φ) → (iAz + h∂zh−1, hφ) by a function h with values in C∗. Thus, we
only have to find solutions to the equation Dzφ = 0 modulo the complexified gauge transformations. In other
words, we only have to find pairs of a holomorphic line bundle with a holomorphic section. Here, it is convenient to
compactify our Euclidean 2-plane to a Riemann sphere. Then, there is a unique holomorphic line bundle of k = 1.
Such a bundle has two-dimensional space of holomorphic sections, where each section has a single simple zero. The
residual complexified gauge symmetry is a multiplication by a constant and it does not change the location of the
zero of a section. Thus, the space of equivalence classes is complex one-dimensional and is parametrized by the
zero locus of φ.
Let us examine in more detail the behaviour of a vortex solution. We consider the vortex at z = 0. For
the finiteness of the action, |φ| must approach the vacuum value |φ|2 = r0 at infinity. By rescaling φ =√r0φ
110 Mirror Symmetry: T-duality of GLSM
where |φ| → 1 at infinity, it becomes clear that the equations depend only on one length parameter 1/e√r0. This
characterizes the size of the vortex. Thus, we expect that the gauge field is nearly flat on |z| ≫ 1/e√r0 and φ is
nearly covariantly constant there. Since (−1/2π)∫F12 = 1, we have Aµ = −∂µ arg(z) and φ =
√r0z/|z| at infinity.
The exact solution takes the form
iA = −1− f2
(dz
z− dz
z
), (VI.2.34a)
φ =√r0 exp
(−∫ ∞
|z|2
dwf(w)
2w
)z
|z| , (VI.2.34b)
where f is a function of w = |z|2 which satisfies the equation wf ′′ = e2r0
2 f + ff ′ and the boundary condition
f(0) = 1, f(+∞) = 0. The asymptotic behaviour of the function f(|z|2) at |z| ≫ 1/e√r0 is
f = const
√m|z| e−m|z| + · · · , (VI.2.35)
where const is a numerical constant and
m = e√
2r0 . (VI.2.36)
The vortex at z = z0 is obtained from the above solution simply by the replacement z → z − z0.
Fermionic Zero Modes
Now let us examine the fermionic zero modes in the instanton background. Since σ = 0 in this background, the
fermionic part of the action (VI.2.29) decomposes into two parts as
−i(ψ−, λ+)
(2Dz −φφ† − 1
e2 ∂z
)(ψ−
λ+
)+ i(ψ+, λ−)
(2Dz −φφ† − 1
e2 ∂z
)(ψ+
λ−
). (VI.2.37)
Using the index theorem of [106], one can see that the operators of the first and the second terms have index 1 and
−1 respectively in our background. Furthermore, there is no normalizable zero modes for (ψ−, λ+) nor (ψ+, λ−).
To see this we note that∫
d2z
(∣∣∣2Dzψ+ − φλ−∣∣∣2
+ 2e2∣∣∣φ†ψ+ −
1
e2∂zλ−
∣∣∣2)
=
∫d2z
(|2Dzψ+|2 + 2e2|φ|2|ψ+|2 +
2
e2|∂zλ−|2 + |φ|2|λ−|2
) (VI.2.38)
in the vortex background, where ψ+ and λ− are considered here as commuting spinors. The left hand side vanishes
for a zero mode of (ψ+, λ−), but then vanishing of the right hand side requires ψ+ = λ− = 0. The argument for
(ψ−, λ+) is the same.
Thus, each of (ψ−, λ+) and (ψ+, λ−) has exactly one zero mode. Actually, the expression for the zero mode in
terms of the vortex solution is available;
(ψ
(0)−
λ(0)
−
)=
(Dzφ
F12
),
(ψ
(0)+
λ(0)+
)=
(Dzφ
†
F12
). (VI.2.39)
VI.2 General formulation 111
It is easy to verify using the vortex equations (VI.2.30b) and (VI.2.30c) that these are indeed the zero modes.
These come from the supersymmetry transformation under the broken supercharges Q− and Q+. The asymptotic
behaviour of φ†ψ(0)− and ψ
(0)+ φ are
φ†ψ(0)− = ∂z|φ|2 = const × r0
z
|z|
√m
|z|e−m|z| + · · · (VI.2.40a)
ψ(0)+ φ = ∂z|φ|2 = const × r0
z
|z|
√m
|z|e−m|z| + · · · , (VI.2.40b)
for the vortex at z = 0. These agree with the asymptotic behaviour of the propagators for a Dirac fermion of mass
m = e√
2r0;
φ†ψ(0)− ∝ r0 S
F−−(z) , ψ
(0)+ φ ∝ −r0 SF
++(z) . (VI.2.41)
This is not an accident; After a suitable similarity transformation, the operators in (VI.2.37) near infinity (where
φ =√rz/|z| and Aµ = −∂µ arg(z)) become nothing but the Dirac operator for a free fermion of mass m = e
√2r.
Computation
In order to prove the generation of e−Y term in the twisted superpotential, we would like to compute a quantity
that is non-vanishing only when such a term is generated. To find out what is the appropriate object to look at,
let us return to our theory of Σ and Y . We recall that the Lagrangian is
L =
∫d4θ
(− 1
2e2ΣΣ− 1
4r0Y Y
)+
1√2
(∫d2θ
(Σ(Y − τ) + cµ e−Y
)+ h.c.
), (VI.2.42)
and we were asking whether c is zero or not. Here we have approximated the Kahler potential for Y by the one in
the continuum limit.
This theory involves a U(1) gauge field. However, there is no charged field and the only appearance of the gauge
field is in the kinetic term and in the Theta term (with the Theta angle being ϑ− θ where ϑ = −Im(y)). As is well
known [18], the effect of the gauge field is to generate a mass term for ϑ:
U =e2
2
(ϑ− θ
)2
, (VI.2.43)
where (α)2 = min(α + 2πn)2|n ∈ Z. Thus, we can treat Σ as an ordinary twisted chiral superfield which has a
complex auxiliary field. In particular, the theory without e−Y term is a free theory of two twisted chiral multiplets.
It is easy to diagonalize the ΣY mixing and it turns out that the combinations X(±) = ±Σ/(2e)+(Y − τ)/(2√2r0)
are superfields of mass
±m = ±e√
2r0 . (VI.2.44)
Now, it is easy to see that the fermionic components χ+ and χ− of Y has vanishing two point function in the free
theory;
〈χ+(x)χ−(y)〉 = 0 , if c = 0 . (VI.2.45)
112 Mirror Symmetry: T-duality of GLSM
However, if e−Y is generated, the twisted F-term would contain a term −e−yχ+χ− in the Lagrangian. This would
contribute to the two point function as
〈χ+(x)χ−(y)〉 = cµr20
∫d2z e−τ SF
++(x− z)SF−−(z − y) (VI.2.46)
where SFαβ(x− y) is the Dirac propagator for the fermions of mass m = e
√2r0.
Now, let us compute the two point function 〈χ+(x)χ−(y)〉 in the original gauge theory. There are some relations
between GLSM and T-dualized theories as
χ+ = 2ψ+φ , χ− = −2φ†ψ− . (VI.2.47)
Since the product of these carries an axial R-charge 2, only the vortex backgrounds with k = 1 can contribute to
the two point function. The contribution is expressed as an integration over the location z of the vortex
〈χ+(x)χ−(y)〉 = −4ΛUV
∫d2z e−τ0
(ψ
(0)+ φ(z)
)(x)(φ†(z)ψ
(0)−
)(y) , (VI.2.48)
where φ(z) is the vortex solution at z. The factor of ΛUV comes from the measure of bosonic zero modes (∝ Λ2UV)
and that for the fermionic zero modes (∝ Λ−1UV). As we have seen, the fermionic zero mode multiplied by φ is
proportional to the Dirac propagator (VI.2.41). Thus, we obtain
〈χ+(x)χ−(y)〉 = const × r20ΛUV
∫d2z e−τ0 SF
++(x− z)SF−−(z − y) , (VI.2.49)
which agrees with (VI.2.46) considering the relation ΛUVe−τ0 = µ e−τ . Thus, we have shown that our dual theory
correctly reproduces the gauge theory result if and only if c 6= 0.
VI.3 Example of mirror pair: complex projective space
ここで例を挙げる。まずは先程と同様に CPN−1 を考えよう。この時は N 個の chiral superfields Φi (Qi = 1) の dual
として N 個の twisted chiral superfields Yi を用意する。Yi の constraint は
N∑
i=1
Yi = τ (VI.3.1)
である。これは
Yi =τ
N−Θi (i = 1, 2, · · · , N − 1) , YN =
τ
N+
N−1∑
i=1
Θi (VI.3.2)
を導入する事で解かれる。ここで Θi は Yi と同じく twisted chiral field であり、Θi ∼ Θi + 2πi である。LG model
の twisted superpotential (VI.2.25b) は
W = Λ(eΘ1 + · · ·+ eΘN−1 +
N−1∏
i=1
e−Θi
), (VI.3.3)
と与えられる。ここで dynamical scale parameter Λ = µe−τ/N を導入した。この twisted superpotential は affine
AN−1 Toda field theory を与えている。
VI.4 Period integral 113
この LG superpotential (VI.3.3) の存在により、axial R-symmetry U(1)A は Z2N まで破れることを見ておこう
(vector R-symmetry は U(1)V を保つ)。まず、axial R-symmetry の変換を思い出す:
Yi → Yi − 2iα , τ → τ − 2iα . (VI.3.4)
ここで α は U(1)A の trsf parameter である。これより、Θi と Λ は次のように変換される:
Θi → Θi + 2iα , Λ → Λe2iα . (VI.3.5)
W の変換則は、各項が同じだけ変換されるべしという条件から、次のように与えられる:
Λ
N−1∑
i=1
eΘi → e4iα Λ
N−1∑
i=1
eΘi , Λ
N−1∏
i=1
e−Θi → e−2i(N−2)αΛ
N−1∏
i=1
e−Θi (VI.3.6a)
∴ e4iα = e−2i(N−2)α ⇒ α =2π
2Nn , n ∈ Z . (VI.3.6b)
これより、U(1)A → Z2N が示された。
また、LG model の twisted superpotential (VI.3.3) の critical points ∂W/∂Θi = 0 を考察しよう。この解は
eΘ1 = eΘ2 = · · · = eΘN−1 ≡ X かつ X = X−(N−1) で与えられる。つまり N 個の真空解 eΘi = X = e2πiℓ/N
(ℓ = 0, 1, · · · , N − 1) がそれぞれ
W = NΛe2πiℓ/N (VI.3.7)
を critical value として持つ。N 個の真空解は全て massive である。そしてそれぞれの真空を結ぶ解は N = 2
supersymmetry の central charge の絶対値を質量に持つ BPS soliton である。例えば ℓ = 0 真空と ℓ = k 真空を結
ぶ BPS soliton は
Z0k = NΛ(e2πik/N − 1) (VI.3.8)
の central charge で与えられる質量 m0k = |Z0k| を持つ。同様にm0k の質量を持つ BPS soliton は(Nk
)個存在する。
VI.4 Period integral
Hori and Vafa [54] が構成した GLSM の T-dualized Lagrangian は (VI.2.23) であった。しかしこれは GLSM side
に chiral superpotential WGLSM が存在しない場合である。WGLSM が存在する場合、一般に Kahler potential term
は複雑できちんと構成することは不可能であろうと考えられる。また古典的に求めることができても、wavefunction
renormalization などが働くため、RG flow の先にはどのような関数形になるのかは未知である。そのため WGLSM が
存在する GLSM の T-dualized theory は一般に次で記述される:
L =
∫d4θ−K(Σ,Σ, Yi, Y i)
+( 1√
2
∫d2θ W + h.c.
), (VI.4.1a)
W = Σ(∑
i
QiYi − τ(µ))
+ µ∑
i
e−Yi . (VI.4.1b)
つまり Hori-Vafa formulation で完全に理解できているのは twisted superpotential W だけである。この W を用い
て、どこまで非自明な事柄が導けるであろうか。
114 Mirror Symmetry: T-duality of GLSM
これを考えるために、period integral というものを導入する。GLSM side で構成される CY が noncompact V の
場合と、compact M の場合それぞれの period integral を考えよう。ここで V と M は、それぞれある toric variety
X の fiber bundle もしくは hypersurface として与えられる。それぞれ、GLSM for X に余分な chiral superfield Pα
を導入して構成されるが、両者の違いは、GLSM に chiral superpotential WGLSM =∑
α Pα ·Gα が存在するかしな
いか、である。
まずいきなり結果から記載しておくと、V , M 上の GLSMに対応するT-dualized theory sideでの period integral
は次のようになる6:
Π = 〈 γ | 1 〉V =
∫
γ
Ω exp(−WLG) , (VI.4.2a)
Π = 〈 γ | 1 〉M =
∫
γ
Ω δ · exp(−WLG) = k∂
∂τΠ =
∫
γ
Ω∂
∂YPexp(−WLG) . (VI.4.2b)
ここで Ω は T-dualized theory での volume form、そして δ = kΣ である。但し簡単のため、GLSM は 1 つの U(1)
gauge symmetry を持ち、P は −k の U(1) charge を持つとする。そして WGLSM = P ·Gk とする7。
何故この period integral (VI.4.2)が登場するのであろうか。それはCecotti and Vafaの tt∗-relation [15]、Morrison
and Ronen Plesser の topological A-twisted GLSM の議論 [81]、そして Hori-Vafa の T-dualized Lagrangian の形
から、これが必要になるからである。では、それぞれについて言及しよう。
• Cecotti-Vafa [15]
ある CY manifold 上の topological B-twisted sigma model を用意する。そして B-model の identity operator
1 (QB = Q+ +Q− singlet) と、A-boundary γA (QA = Q+ +Q− singlet) で構成される “partition function”
Π を定義する:
Π = 〈 γA | 1 〉 =
∫
γA
Ω . (VI.4.3)
さて、この Π は CY sigma model の上で定義されるが、[15] で議論されるところによると、CY/LG correspon-
dence から得られる LG (chiral) superpotential WLG を用いて
Π =
∫
γA
Ω exp(−WLG) ≡ Z
[ γA
A-boundary B-twist
← 1
]. (VI.4.4)
と記述できる。これは実は topological B-twist された LG theory の partition function である。topological
B-twisted theory では、Kahler potential, anti-chiral superpotential は QB-exact になるため、上の period
integral がそのまま partition function となる。そのため、改めて Z[∗] という記号を用いている。さらにわかりやすいように worldsheet で表しておいた。これは Kahler structure に依存しない。さらにこれの mirror dual
を取ると、これは
Π =
∫
eγB
Ω exp(−WLG) , (VI.4.5a)
6実は一足飛びに period integral = path integral というわけではない。conformal であること、topological であることなどを用いて、LG
quantum mechanics on CY manifold に reduction している。さらに登場する関数は全て bosonic scalar component だけで議論して、volume
とか LG potential とかの導出を行う。そして最後に superfield だと思え、という事をする。その詳細は Hori et al. [53] chapter 19 を参照。7複数の U(1) gauge symmetry を持つ、もしくは WGLSM が複数の polynomial Gα を持つ場合などに簡単に一般化できるので、ここでは
最も簡単な模型についてのみ言及する。
VI.4 Period integral 115
Z
[ γA
A-boundary B-twist
← 1
]= Z
[ γB
B-boundary A-twist
← 1
], (VI.4.5b)
となる。つまり mirror CY上の γB boundary statesと A-twisted fieldsとの inner product (B-sector in mirror
CY)が A-sector in orginal CYから理解できる8。ちなみにこれより、topological A-sector in CYから B-sector
in mirror CY についての period integral もここで理解できることになる。
• Morrison and Ronen Plesser [81]
Morrison and Ronen Plesser [81] は topological A-twisted GLSM についての議論である。具体的には、GLSM
superpotential WGLSM の有無が correlation function にどのように影響するのかを述べている。(A-twisted
GLSM では、WGLSM の変分は QA-exact であるため、一見 physical には影響しないと思われるが、存在する
か否かは影響する。)
まず GLSM for (compact) toric variety X を用意する。簡単のため U(1) gauge symmetry は 1 つだけとする。
これは chiral superfields Si は Qi > 0 の U(1) charge を持ち、WGLSM は存在しない。これに negative charge
(−k) を持つ chiral superfield P を用意すると、Si, P 全体で noncompact toric variety V が構成される。さ
らに WGLSM = PGk(Si) を導入すると、compact hypersurface M が得られる。例えば
X = CPN−1 , V = O(−k)→ CPN−1 , M = CPN−1[k]
である9。それらの関係を Figure VI.1 で表しておこう:
V
M
X
[example]
V = O(−k) bundle on CPN−1
X = CPN−1
M = hypersurface CPN−1[k]
Figure VI.1: Images of toric variety X, noncompact variety V and compact hypersurface M .
sigma model on V と sigma model on M の関係は、GLSM から見れば、RG flow として V から M への流れ
が存在する。理由は、WLG = Gk が RG flow として marginal operator であり、そのため CFT がこれから読
み取れるように設定すると、WGLSM = P ·Gk は relevant operator になるからである。
ここで topological A-twist されていることからくる重要な関係式を挙げる (ここでもその導出は述べない。詳
細は [81, 54] を見よ。):
| 1 〉M = | δ 〉V ≡ δ| 1 〉V , δ ≡ kΣ . (VI.4.6)
これより、B-boundary states と A-model ground state との inner product は
Π = 〈 γB | 1 〉M = 〈 γB | δ 〉V = 〈 γB |δ| 1 〉V , (VI.4.7a)8mirror dual であることをここでは用いているが、何故これが言えるのかは、ここではその証明はしない。どこかにあるのだろう。つまりこ
こでは mirror の証明をするのが目的ではなく、mirror dual ならば何を用いたら有効か、を明らかにする。9topological A-model は、Kahler variety であればよい。何も CY である必要はない。
116 Mirror Symmetry: T-duality of GLSM
Π = Z
[ γB
B-boundary A-twist
← 1
]
M
= Z
[ γB
B-boundary A-twist
← δ
]
V
, (VI.4.7b)
なる関係がある。(VI.4.7a) は A-twisted sigma model on CY M and V の period integral の記述であり、
(VI.4.7b) は A-twisted LG theory の partition function としての記述だと思えば分かりやすい。
• Hori and Vafa [54]
Hori and Vafa は上述の 2 つの関係式 (VI.4.5), (VI.4.7) を組み合わせ、topological A-twisted GLSM for
compact CY manifold M (つまりWGLSM ありの理論) の topological A-sector と mirror LG の B-sector 双方
を記述する period integral を定式化した:
Π = 〈 γB | 1 〉M =
∫
γB
Ω δtt∗=
∫
γB
Ω δ · exp(−WLG)
mirror=
∫
eγA
Ω δ · exp(−WLG)tt∗=
∫
eγA
Ω δ = 〈 γA | 1 〉fM ,
(VI.4.8a)
Z
[ γB
B-boundary A-twist
← 1
]
M
= Z
[ γB
B-boundary A-twist
← δ
]
V
= Z
[ γA
A-boundary B-twist
← δ
]
eV= Z
[ γA
A-boundary B-twist
← 1
]
fM.
(VI.4.8b)
ここで M は mirror CY である。
ここで注意が必要である。(VI.4.8) の根源にあるのは Morrison and Ronen Plesser の関係式 (VI.4.7) である
が、(VI.4.7) は GLSM の topological A-sector の議論しか存在しない。そのためこれを用いた (VI.4.8) は、GLSM
の topological A-sector と T-dualized theory の B-sector の関係が得られても、GLSM の B-sector と T-dualized
theory の A-sector に関しては何も説明できない。そのため、本当の意味での mirror symmetry の説明にはなってい
ないことがわかる。なお、chiral superpotential を必要とする CP4[5] が何故ここの議論だけで十分であるかを補足
しておく。CP4[5] を定義する時、その hypersurface は、単に “homogeneous” であれ、という条件しか持っていな
い。WGLSM = P ·G5(S) に課す要請としては一番緩いものであり、G5(S) の存在を述べているに過ぎない。そのた
め、topological A-twist された GLSM でも、period integral Π の計算に δ = 5Σ を挿入するだけで十分である。その
ため、chiral superpotential を持つとはいえ、CP4[5] の mirror dual が問題なく計算でき、その結果も正しいものと
なる。しかし、Gℓ(S) にそれ以上の条件を課すとなると、もはや δ = ℓΣ の挿入だけでは不十分となる。topological
B-twist してきちんと定義する必要が生じる。
VI.5 Mirror geometry descriptions for various theories 117
CY M
A-sector
(Kahler)
CY M
A-sector
(Kahler)
CY M
B-sector
(complex)
CY M
B-sector
(complex)
explained by Hori-Vafa
not explained
VI.5 Mirror geometry descriptions for various theories
ここではさらに一般的な geometry の mirror dual を考えよう。参考文献はHori and Vafa [54] section 7.3 及び Hori,
Iqbal and Vafa [50] section 8 である。[54] は mirror geometry の homogeneous coordinates による表示の導出であ
り、[50] は local coordinates による表示の導出である。どちらも本質的には同じであるのは当然であるが、ここでは
幾つかの geometry を両方のやり方で記述することにする。ここの議論の主役は “period integral” である。
period integral (VI.4.2) では、GLSM での superpotential WGLSM の存在から来る correlation function 及び
period integral の影響を δ = ℓΣ で表した後、これを complexified FI parameter τ を用いて書き換えたが、それと
は異なる表記を次のように採ろう。例えばWGLSM = PGℓ(Si) という superpotential が存在する場合を考える:
Π =
∫dΣ∏
i
dYi dYP (ℓΣ) exp− Σ
(∑
i
Yi − ℓYP − τ)−∑
i
e−Yi − e−YP
=
∫dΣ∏
i
dYi dYP∂
∂YP
[exp
− Σ
(∑
i
Yi − ℓYP − τ)]
exp(−∑
i
e−Yi − e−YP
)
= −∫
dΣ∏
i
dYi dYP e−YP exp− Σ
(∑
i
Yi − ℓYP − τ)−∑
i
e−Yi − e−YP
=
∫ ∏
i
dYi dYP e−YP δ(∑
i
Yi − ℓYP − τ)
exp(−∑
i
e−Yi − e−YP
).
(VI.5.1)
ここでは Π ≡ Πa,1 という略記を用いている。この (VI.5.1) を用いて理論を解析することで、mirror geometry を読
み取る作業を以下に列挙する。
以後、いくつかの例を挙げながら、period integral (VI.4.2) から得られる mirror LG 及び mirror CY geometry
を見ていこう。
118 Mirror Symmetry: T-duality of GLSM
VI.6 Mirror dual of O(−N) bundle on CPN−1
GLSM における field contents は次の通り:
S1 . . . SN P
U(1) charge 1 . . . 1 −N
chiral superpotential はない。そのため T-dualized theory における period integral は次のようにして与えられる:
Π =
∫dΣ
N∏
i=1
dYidYP exp− Σ
( N∑
i=1
Yi −NYP − τ)−
N∑
i=1
e−Yi − e−YP
=
∫ N∏
i=1
dYi dYP δ( N∑
i=1
Yi −NYP − τ)
exp(−
N∑
i=1
e−Yi − e−YP
). (VI.6.1)
canonical measure を与えるような場の変換を行うことで period integral から mirror geometry を読み取る。この方
法にはテクニカルに 2 通り、つまり homogeneous coordinates で求める方法と local coordinates で求める方法があ
る。それぞれをここで議論しよう。
VI.6.1 mirror description in terms of homogeneous coordinates
homogeneous coordinates の表示を得る方法を挙げる [54]。次の場の変換を行おう:
e−YP ≡ P , e−Yi ≡ PUi . (VI.6.2)
この変換によって P , Ui はそれぞれ C∗-valued fields として与えられるが、それらを用いた (VI.6.1) は次のように
なる:
Π =
∫ N∏
i=1
(dUi
Ui
)(dP
P
)δ(
log( N∏
i=1
Ui
)+ τ)
exp− P
( N∑
i=1
Ui + 1)
. (VI.6.3)
さらに P の measure を canonical にするために、次の Gaussian integral を用いよう:
1
P=
∫du dv exp
(P uv
). (VI.6.4)
但し u, v は C-valued fields である。これより
Π =
∫ N∏
i=1
(dUi
Ui
)dP du dv δ
(log( N∏
i=1
Ui
)+ τ)
exp− P
( N∑
i=1
Ui + 1− uv)
(VI.6.5)
となる。P はこれによって canonical measure を持ち、C-valued field とみなすことができるようになる。C-valued
になったため P も積分できる:
Π =
∫ N∏
i=1
(dUi
Ui
)du dv δ
(log( N∏
i=1
Ui
)+ τ)δ( N∑
i=1
Ui + 1− uv). (VI.6.6)
VI.6 Mirror dual of O(−N) bundle on CPN−1 119
さて、C∗-valued fields Ui も canonical measure を与えるように再度変換しよう。period integral の δ-functional を
考慮して次のように与える:
Ui ≡ e−τ/N ZNi
Z1 · · ·ZN, (VI.6.7)
この変換によって新しい fields Zi は次の対称性を持つ:
Zi 7→ λωiZi , where λ : C∗-value , ωNi = ω1 · · ·ωN = 1 (VI.6.8)
この変換でも Ui は不変であるので、これは C∗ × (ZN )N−2 gauge symmetry であるとみなすことができる。(ZN )N
ではない理由として、まず ωi は独立ではなく ω1 · · ·ωN = 1 という条件があるために ZN がひとつ不要となる。さ
らに Zi は実は一次独立ではない。そのためさらに ZN がひとつ不要となる。
measure の Jacobian を考えよう。ただ安直に変換を施すと
dUi
Ui= (N − 1)
dZi
Zi− dZj
Zj(1− δij)
であるために
N∏
i=1
(dUi
Ui
)= |J |
N∏
i=1
dZi , J = det( ∂Ui
∂Zj
)= 0
となり、well-defined になっていないように思われるが、これはそもそも Ui もしくは Zi がそれぞれ一次独立ではな
いために起こる現象である。また gauge symmetry を持っているので、正しくは
N∏
i=1
(dUi
Ui
)δ(
log( N∏
i=1
Ui
)+ τ)≃
N∏
i=1
(dZi
Zi
) 1
vol.(C∗)(VI.6.9)
となる。(但し余分に単なる数が登場するが、それは適当に measure の中に押し込んでおく。) これによって (VI.6.6)
はさらに書き換えられる:
Π =
∫1
vol.(C∗)
N∏
i=1
(dZi
Zi
)du dv δ
( e−τ/N
Z1 · · ·ZNG(Zi;u, v)
)
=
∫1
vol.(C∗)
N∏
i=1
dZi du dv δ(G(Zi;u, v)
), G(Zi;u, v) ≡
N∑
i=1
ZNi + eτ/NZ1 · · ·ZN (1− uv) . (VI.6.10)
つまり変換 (VI.6.7) によって C∗-valued fields Ui は C-valued fields Zi になる。
もう少し計算して volume form になるまで見ておこう。但し [54] を完全に模倣する。まず C∗ gauge symmetry
を、ZN ≡ 1 にすることで完全に固定しよう。すると
Π =
∫du dv
N−1∏
i=1
dZi δ(G(Z1, · · · , ZN−1, 1;u, v)
)
=
∫du dv
N−2∏
i=1
dZi
/∂G|ZN=1
∂ZN−1
∣∣∣G=0
=
∫du ∧ dv ∧ ΩN−2 (VI.6.11)
となる。
120 Mirror Symmetry: T-duality of GLSM
さて、表式 (VI.6.10)から algebraic equationを読みとることが可能である。O(−N) bundle on CPN−1 の mirror
geometry M は次のようになっている:
M =
(Z1, · · · , ZN ;u, v) ∈ CN+2∣∣∣ 0 =
N∑
i=1
ZNi + zZ1 · · ·ZN
/C∗ × (ZN )N−2
z = eτ/N − u′v′ , u′v′ = eτ/Nuv
(VI.6.12)
この algebraic equation は次の性質を持つ:
• Zi は CPN−1[N ]/(ZN )N−2 の homogeneous coordinates。これだけですでに compact CY (N−2)-foldを成す。
• その CPN−1[N ]/(ZN )N−2 は z = eτ/N − u′v′ の点毎に存在し、z は無限に伸びている。この “coordinates” の
存在のために、全体で noncompact geometry となる。
これにより、mirror CY geometry M は、CPN−1[N ]/(ZN )N−2 の 1-parameter family (noncompact direction) を
なす noncompact geometry であることがわかる。compact direction に DOF = N − 1, noncompact direction に
DOF = 1 である。
最後に、ZN = 1 とした local coordinates Xi ≡ Zi/ZN で上の algebraic equation を表しておこう:
M =
(Xi;u′v′) ∈ CN+1
∣∣∣ 0 = 1 +
N−1∑
i=1
XNi + zX1 · · ·XN−1
/(ZN )N−2 , z = eτ/N − u′v′ . (VI.6.13)
VI.6.2 mirror description in terms of local coordinates
続いて local description を議論する。出発点は (VI.6.1) である。
δ-function を YN について解く:
e−YN = e−τN∏
i=1
eYie−NYP . (VI.6.14)
これより period integral から δ-functional が外れる:
Π =
∫ N−1∏
i=1
dYi dYP exp−
N−1∑
i=1
e−Yi − e−τN∏
i=1
eYie−NYP − e−YP
. (VI.6.15)
ここで field re-definition を行う:
e−YP ≡ P , e−Yi ≡ P Vi . (VI.6.16)
ここで P , Vi は C∗-valued fields となっている。これによって period integral はさらに簡単になる:
Π =
∫ N−1∏
i=1
(dVi
Vi
)(dP
P
)exp
− P
(N−1∑
i=1
Vi + 1 +e−τ
V1 · · ·VN−1
)
≡∫ N−1∏
i=1
(dVi
Vi
)dP du dv exp
− P
(N−1∑
i=1
Vi + 1 +e−τ
V1 · · ·VN−1− uv
). (VI.6.17)
VI.6 Mirror dual of O(−N) bundle on CPN−1 121
ここで、C-valued fields u, v を導入することで P は canonical measure を持つことになる。これは P はもはや
C-valued field であることを意味する。従って P も積分できるようになり、δ-functional を生み出す:
Π =
∫ N−1∏
i=1
(dVi
Vi
)du dv δ
(N−1∑
i=1
Vi + 1 +e−τ
V1 · · ·VN−1− uv
)
≡∫ N−1∏
i=1
(dV ′i
V ′i
)du′ dv′ δ
( e−τ/N
V ′1 · · ·V ′
N−1
G(V ′i ;u′, v′)
)
=
∫ N−1∏
i=1
dV ′i du′ dv′ δ
(G(V ′
i ;u′, v′)), G(V ′
i ;u′, v′) ≡ 1 + V ′1 · · ·V ′
N−1
(N−1∑
i=1
V ′i + z
). (VI.6.18)
但し Vi = e−τ/NV ′i , uv = e−τ/Nu′v′ と rescale しておいた。さらに z ≡ eτ/N − u′v′ を導入している。なお、この表
示ではすでに V ′i も canonical measure を獲得している。もはや volume form の導出は省略するが、ここから mirror
geometry M が次の ように読み出せる:
M =
(V ′i ;u′, v′) ∈ CN+1
∣∣∣ 0 = 1 + V ′1 · · ·V ′
N−1
(N−1∑
i=1
V ′i + z
), z = eτ/N − u′v′
. (VI.6.19)
実はまだ orbifolding をしていない。それは次のように再定義して与えられる:
V ′i =
XNi
X1 · · ·XN−1,
N−1∏
i=1
dV ′i ≃
N−1∏
i=1
dXi , (VI.6.20a)
Xi 7→ ωiXi , ωNi = ω1 · · ·ωN = 1 (VI.6.20b)
これより orbifold symmetry は (ZN )N−2 であることがわかる。また Xi でも canonical measure となっている。こ
れより (VI.6.19) はさらに書き換えられる:
M =
(Xi;u′, v′) ∈ CN+1
∣∣∣ 0 = 1 +
N−1∑
i=1
XNi + zX1 · · ·XN−1 , z = eτ/N − u′v′
/(ZN )N−2
(VI.6.21)
これは (VI.6.13) と全く同じである。これから homogeneous coordinates に書き直すには
Xi ≡Zi
ZN(VI.6.22)
とすれば良いことがわかる。
algebraic equation としては homogeneous description (VI.6.12) が整っているが、導出方法としては local de-
scription の方がやりやすい気がする。
122 Mirror Symmetry: T-duality of GLSM
VI.7 Mirror dual of CPN−1[N ]
VI.7.1 Mirror LG description
CY hypersurface を与える GLSM field contents を思い出そう:
S1 . . . SN P
U(1) charge 1 . . . 1 −N
WGLSM = P ·GN (Si) , GN (Si) : homogeneous polynomial of degree N . (VI.7.1)
これを基に構成される T-dualized theory の period integral は、chiral superpotential が存在するために Π で与えら
れる:
Π =
∫dΣ
N∏
i=1
dYi dYP (NΣ) exp− Σ
( N∑
i=1
Yi −NYP − τ)−
N∑
i=1
e−Yi − e−YP
= N∂
∂τ
∫ N∏
i=1
dYi dYP δ( N∑
i=1
Yi −NYP − τ)
exp−
N∑
i=1
e−Yi − e−YP
. (VI.7.2)
ここで δ-functional を YP について解く:
−YP =1
N
(τ −
N∑
i=1
Yi
). (VI.7.3)
よって、これを代入した後に τ 微分を実行すると、(VI.7.2) の measure に fields が追加される:
Π = N∂
∂τ
∫ N∏
i=1
dYi exp−
N∑
i=1
e−Yi − eτN
N∏
i=1
e−1N
Yi
= eτN
∫ N∏
i=1
(e−
1N
YidYi
)exp
−
N∑
i=1
e−Yi − eτN
N∏
i=1
e−1N
Yi
. (VI.7.4)
canonical measure を与えるために次の field re-definition を行えばよいだろう:
Xi ≡ e−1N
Yi . (VI.7.5)
これより period integral が次のようになり、適切な LG twisted superpotential が読み取れる様になる:
Π = (factors)
∫ N∏
i=1
dXi exp(−
N∑
i=1
XNi − e
τN X1 · · ·XN
)
≡ (factors)
∫ N∏
i=1
dXi exp(− W
). (VI.7.6)
またこの変換では、Yi → Yi + 2πi shift から次の orbifold symmetry が登場する:
Xi → ωiXi , ωNi = ω1ω2 · · ·ωN = 1 . (VI.7.7)
よって得られる LG theory は
W = XN
1 + · · ·XNN + e
τN X1 · · ·XN
/(ZN )N−1 (VI.7.8)
である。
VI.7 Mirror dual of CPN−1[N ] 123
VI.7.2 Mirror geometry
すでに前にも記載したが、GLSM における field contents をもう一度記載する:
S1 . . . SN P
U(1) charge 1 . . . 1 −N
なお、GLSM には次の chiral superpotential がある:
WGLSM = P ·GN (Si) , GN (Si) : homogeneous polynomial of degree N . (VI.7.9)
これを基に構成される T-dualized theory の period integral は、chiral superpotential が存在するために Π で与えら
れる:
Π =
∫dΣ
N∏
i=1
dYi dYP (NΣ) exp− Σ
( N∑
i=1
Yi −NYP − τ)−
N∑
i=1
e−Yi − e−YP
=
∫dΣ
N∏
i=1
dYi dYP∂
∂YP
[exp
− Σ
( N∑
i=1
Yi −NYP − τ)]
exp(−
N∑
i=1
e−Yi − e−YP
)
=
∫ N∏
i=1
dYi dYP e−YP δ( N∑
i=1
Yi −NYP − τ)
exp(−
N∑
i=1
e−Yi − e−YP
). (VI.7.10)
mirror dual of O(−N) bundle on CPN−1 の時と同様に、homogeneous coordinates と local coordinates それぞれで
の構成を記載する。その際、今回はあまりコメントは記載しない。
mirror description in terms of homogeneous coordinates
(VI.7.10) の δ-functional に見合うように場の再定義を行う:
e−YP ≡ P , e−Yi ≡ PUi , P , Ui : C∗-valued fields . (VI.7.11)
∴ Π =
∫ N∏
i=1
(dUi
Ui
)dP δ
(log( N∏
i=1
Ui
)+ τ)
exp− P
( N∑
i=1
Ui + 1)
=
∫ N∏
i=1
(dUi
Ui
)δ(
log( N∏
i=1
Ui
)+ τ)δ( N∑
i=1
Ui + 1). (VI.7.12)
Ui から field re-definition で canonical measure を引き出す (C-valued field にする) ために次の変換を行う:
Ui ≡ e−τ/N ZNi
Z1 · · ·ZN,
dUi
Ui= (N − 1)
dZi
Zi− dZj
Zj(1− δij) , (VI.7.13a)
次の Zi の変換の下で Ui は不変である:
Zi 7→ λωiZi , ωNi = ω1 · · ·ωN = 1 , λ : C∗-value (VI.7.13b)
124 Mirror Symmetry: T-duality of GLSM
Ui, Zi はそれぞれ一次独立ではないことを反映して λ が登場している。つまり gauge symmetry が存在する。全体
の gauge symmetry は C∗ × (ZN )N−2 である。これによりmeasure は次のようになる:
dUi
Uiδ(
log( N∏
i=1
Ui
)+ τ)
=N∏
i=1
(dZi
Zi
) 1
vol.(C∗). (VI.7.14)
∴ Π =
∫ N∏
i=1
(dZi
Zi
) 1
vol.(C∗)δ(1 +
e−τ/N
Z1 · · ·ZN
N∑
i=1
ZNi
)
≡∫ N∏
i=1
(dZi
Zi
) 1
vol.(C∗)δ( e−τ/N
Z1 · · ·ZNG(Z1, · · · , ZN )
)
=
∫ N∏
i=1
dZi1
vol.(C∗)δ(G(Z1, · · · , ZN )
), G(Z1, · · · , ZN ) =
N∑
i=1
ZNi + eτ/NZ1 · · ·ZN . (VI.7.15)
canonical measure になっているので、Zi は C-valued fields とみなすことができる。
volume form 導出は省略する。この (VI.7.15) から、mirror geometry M の algebraic equation が得られる:
M =
(Z1, · · · , ZN ) ∈ CN∣∣∣ 0 =
N∑
i=1
ZNi + eτ/NZ1 · · ·ZN
/C∗ × (ZN )N−2
(VI.7.16)
local description にするには ZN = 1 もしくは Xi = Zi/ZN とすると良い:
M =
(X1, · · · ,XN−1) ∈ CN−1∣∣∣ 0 = 1 +
N−1∑
i=1
XNi + eτ/NX1 · · ·XN−1
/(ZN )N−2 . (VI.7.17)
これからわかることは、Zi, Xi はそれぞれ CPN−1[N ]/(ZN )N−2 の homogeneous/local coordinates を表すことであ
る。他に field は存在しないので、CPN−1[N ] の mirror dual は CPN−1[N ]/(ZN )N−2 であることが示される。これ
は Greene-Plesser 構成法による結果 [40] を再現している。また LG twisted superpotential による記述 (VI.7.8) と
比較すると良いであろう。
quintic hypersurface CP4[5] の mirror dual は次で与えられていた:
W/G =[ 5∑
i=1
X5i + eτ/5
5∏
i=1
Xi
]/(Z5)
4 . (VI.7.18)
twisted superpotential W の left/right U(1) は (−1, 1) であるべきなので、twisted chiral superfield Xi のそれは
(−1/5, 1/5) となるべきである。また、appendix III や ref. [72] の議論にあるように、LG model に内蔵されている
highest charged state から central charge を読みとると
c = 6β = 6
5∑
i=1
(1
2− qi
)= 9 (VI.7.19)
が得られる。これは GLSM side の LG model (LGcc と呼ぼう) の central charge と一致する。また、もともとの
GLSM side での CY/LG correspondence は
CP4[5] ←→ Z5 orbifolded LGcc (VI.7.20)
VI.7 Mirror dual of CPN−1[N ] 125
であった。一方 T-dualized theory から (Z5)4 orbifolded LGac が得られている。この (LGcc, LGac) は mirror の関
係であるのは Hori-Vafa [54] の知る所であるが、これを CY/LG を用いて CY geometry に読み変えると、
(Z5)4 orbifolded LGac ←→ CP4[5]/(Z5)
3 (VI.7.21)
となる。この CP4[5]/(Z5)3 は CP4[5] と mirror pair の関係であるのは、Greene-Plesser [40] によって明確に示され
ている。また (VI.7.16) でも構成している。つまりこの subsection では、最終的に
CY sigma model
on CP4[5]
LG orbifold theory
with WLG/Z5
GLSM
Hori-Vafa’s
CY sigma model
on CP4[5]/(Z5)3
LG orbifold theory
with WLG/(Z5)4
CY/LG
CY/LG
T-dualMirror
FI ≫ 0
algebraic
FI ≪ 0
geometric
WLG = G5(S)
WLG = X51 + X5
2 + X53 + X5
4 + X55 + et/5X1X2X3X4X5
b2,1(CP4[5]) = b1,1(CP4[5]/(Z5)3) = 101 , b1,1(CP4[5]) = b2,1(CP4[5]/(Z5)
3) = 1
の関係が得られる。一般の次元では
CY sigma model
on CPN−1[N ]
LG orbifold theory
with WLG/ZN
GLSM
Hori-Vafa’s
CY sigma model
on CPN−1[N ]/(ZN)N−2
LG orbifold theory
with WLG/(ZN)N−1
CY/LG
CY/LG
T-dualMirror
FI ≫ 0
algebraic
FI ≪ 0
geometric
が得られる。
この議論は Brunner and Hori [8] の section 8.1 に example として記載されている。([8] の本題は orientifold で
126 Mirror Symmetry: T-duality of GLSM
ある。つまり boundary GLSM もしくは boundary CFT, boundary LG model を議論している。その準備として、
まずは boundary のない議論を簡単に紹介しており、そこで CPN−1[N ] の CY mirror dual が登場する。)
mirror description in terms of local coordinates
period integral (VI.7.10) の δ-functional から YN について解くと
e−YN = e−τN−1∏
i=1
eYie−NYP (VI.7.22)
なので、これを (VI.7.10) に代入する:
Π =
∫ N−1∏
i=1
dYi dYP e−YP exp−
N−1∑
i=1
e−Yi − e−τN−1∏
i=1
eYie−NYP − e−YP
. (VI.7.23)
ここでさらに
e−YP ≡ P , e−Yi ≡ P Vi , P , Vi : C∗-valued fields (VI.7.24)
と field re-definition することで
Π =
∫ N−1∏
i=1
(dVi
Vi
)dP exp
− P
(N−1∑
i=1
Vi + 1 +e−τ
V1 · · ·VN−1
)
=
∫ N−1∏
i=1
(dV ′i
V ′i
)δ( e−τ/N
V ′1 · · ·V ′
N−1
G(V ′i ))
=
∫ N−1∏
i=1
dV ′i δ(G(V ′
i )), G(V ′
i ) ≡ 1 + V ′1 · · ·V ′
N−1
(N−1∑
i=1
V ′i + eτ/N
). (VI.7.25)
となる。但し V ′i = eτ/NVi である。V
′i はもはやこの時は C-valued fields となっている。これから mirror geometry
M の algebraic equation が得られる:
M =
(V ′1 , · · · , V ′
N−1) ∈ CN−1∣∣∣ 0 = 1 + V ′
1 · · ·V ′N−1
(N−1∑
i=1
V ′i + eτ/N
). (VI.7.26)
さらに orbifolding を生み出す re-definition を行う:
V ′i ≡
XNi
X1 · · ·XN−1, (VI.7.27)
この変換でも period integral は canonical measure を持つ。ここで Xi が受ける変換則
Xi 7→ ωiXi , ωNi = ω1 · · ·ωN−1 = 1 (VI.7.28)
は V ′i を不変に保つため、Xi で記述されるときは (ZN )N−2 orbifold gauge symmetry を持つ事になる。この Xi を
用いて M を表す:
M =
(X1, · · · ,XN−1) ∈ CN−1∣∣∣ 0 = 1 +
N−1∑
i=1
XNi + eτ/NX1 · · ·XN−1
/(ZN )N−1 (VI.7.29)
VI.8 Mirror dual of CPN−1[ℓ] 127
これは (VI.7.17) と同一である。
最後に homogeneous coordinates Zi で記述するには
Xi ≡Zi
ZN(VI.7.30)
とすればよい。上の local coordinates は ZN 6= 0 で記述していたと考えればよい。mirror geometry は (VI.7.16) で
与えられる。
VI.8 Mirror dual of CPN−1[ℓ]
ここでは CPN−1[ℓ] (但し 0 < ℓ < N) を扱う。これは一般に CY hypersurface ではない。
GLSM における field contents は次の通り:
S1 . . . SN P
U(1) charge 1 . . . 1 −ℓ
なお、GLSM には次の chiral superpotential がある:
WGLSM = P ·Gℓ(Si) , Gℓ(Si) : homogeneous polynomial of degree ℓ . (VI.8.1)
これを基に得られる T-dualized theory の period integral を定義する:
Π =
∫dΣ
N∏
i=1
dYi dYP (ℓΣ) exp− Σ
( N∑
i=1
Yi − ℓYP − τ)−
N∑
i=1
e−Yi − e−YP
=
∫ N∏
i=1
dYi dYP e−YP δ( N∑
i=1
Yi − ℓYP − τ)
exp(−
N∑
i=1
e−Yi − e−YP
). (VI.8.2)
VI.8.1 mirror description in terms of homogeneous coordinates
period integral (VI.8.2) の δ-functional から次の re-definition を行う:
e−YP ≡ P , e−Ya ≡ PUa , e−Yb ≡ Ub , a = 1, · · · , ℓ , b = ℓ+ 1, · · · , N , (VI.8.3a)
∴
N∏
i=1
dYi dYP e−YP =
N∏
i=1
(dUi
Ui
)dP ,
N∑
i=1
Yi − ℓYP − τ = − log( N∏
i=1
Ui
)− τ ,
N∑
i=1
e−Yi + e−YP = P( ℓ∑
a=1
Ua + 1)
+N∑
b=ℓ+1
Ub .
(VI.8.3b)
∴ Π =
∫ N∏
i=1
(dUi
Ui
)dP δ
(log( N∏
i=1
Ui
)+ τ)
exp− P
( ℓ∑
a=1
Ua + 1)−
N∑
b=ℓ+1
Ub
128 Mirror Symmetry: T-duality of GLSM
=
∫ N∏
i=1
(dUi
Ui
)δ(
log( N∏
i=1
Ui
)+ τ)δ( ℓ∑
a=1
Ua + 1)
exp( N∑
b=ℓ+1
Ub
). (VI.8.4)
場の再定義を行う:
Ua ≡ e−τ/ℓ Zℓa
Z1 · · ·ZN,
dUa
Ua= (ℓ− 1)
dZa
Za− dZi
Zi(1− δai) , (VI.8.5a)
Ub ≡ Zℓb ,
dUb
Ub= ℓ
dZb
Zb. (VI.8.5b)
Zi の次の変換の下で Ui は不変である:
Za 7→ λωaZa , Zb 7→ ωbZb
ωℓa = ωℓ
b = ω1 · · ·ωN = 1 , λ : C∗-value(VI.8.6)
つまり理論には C∗ × (Zℓ)N−2 gauge symmetry が存在することになる。場の変換で period integral の measure な
どは
N∏
i=1
(dUi
Ui
)δ(
log( N∏
i=1
Ui
)+ τ)
=N∏
i=1
(dZi
Zi
) 1
vol.(C∗), (VI.8.7a)
ℓ∑
a=1
Ua + 1 =e−τ/ℓ
Z1 · · ·ZNG(Z1, · · · , ZN ) , G(Z1, · · · , ZN ) =
ℓ∑
a=1
Zℓa + eτ/ℓZ1 · · ·ZN , (VI.8.7b)
N∑
b=ℓ+1
Ub =N∑
b=ℓ+1
Zℓb , (VI.8.7c)
になるので、Zi は C-valued fields とみなすことができる。そして (VI.8.4) は次のようになる:
Π =
∫ N∏
i=1
(dZi
Zi
) 1
vol.(C∗)δ( e−τ/ℓ
Z1 · · ·ZNG(Z1, · · · , ZN )
)exp
(−
N∑
b=ℓ+1
Zℓb
)
=
∫ N∏
i=1
dZi1
vol.(C∗)δ(G(Z1, · · · , ZN )
)exp
(−
N∑
b=ℓ+1
Zℓb
). (VI.8.8)
この (VI.8.8) から次の geometry M が読み取れる:
M =
(Z1, · · · , ZN ) ∈ CN∣∣∣ 0 =
ℓ∑
a=1
Zℓa + ψZ1 · · ·Zℓ , ψ = eτ/ℓZℓ+1 · · ·ZN
/C∗ × (Zℓ)
N−2
(VI.8.9a)
Z1, · · · , Zℓ は CPℓ−1[ℓ] を構成していることがわかる。なおこれが ψ で parametrize されて、全体で (N − 1)-dim.
space をなし、それが (Zℓ)N−2 orbifold symmetry を持っていることがわかる。この M を図解すると Figure VI.2
となる:
この M の上で twisted LG theory with twisted superpotential WfM が定義されている:
WfM =
N∑
b=ℓ+1
Zℓb (VI.8.9b)
VI.8 Mirror dual of CPN−1[ℓ] 129
CPℓ−1[ℓ](ψ)
ψ(Zb) = eτ/ℓZℓ+1 · · ·ZN
Figure VI.2: Mirror geometry M : the hypersurface CPN−1[ℓ].
local description にするには、例えば Z1 ≡ 1 の local patch で表示しよう:
Xa =Za
Z1, Xb = Zb . (VI.8.10)
すると total space M は
M =
(X2, · · · ,XN ) ∈ CN−1∣∣∣ 0 = 1 +
ℓ∑
a=2
Xℓa + ψX2 · · ·Xℓ , ψ = eτ/ℓXℓ+1 · · ·XN
/(Zℓ)
N−2 (VI.8.11)
で記述される。
VI.8.2 mirror description in terms of local coordinates
period integral (VI.8.2) の δ-functional から Y1 を解くと −Y1 =∑N
i=2 Yi − ℓYP − τ なので、period integral が次の
ようになる:
Π =
∫ N∏
i=2
dYi dYP e−YP exp(− e−τ/ℓ
N∏
i=2
eYie−ℓYP −N∑
i=2
e−Yi − e−YP
)(VI.8.12)
場の変換を行う:
e−YP ≡ P , e−Ya ≡ P Va , e−Yb ≡ Vb , a = 2, · · · , ℓ , b = ℓ+ 1, · · · , N , (VI.8.13a)
N∏
i=2
dYi dYP e−YP =N∏
i=2
(dVi
Vi
)dP , (VI.8.13b)
e−τ/ℓN∏
i=2
eYie−ℓYP +N∑
i=2
e−Yi + e−YP = P(1 +
ℓ∑
a=2
Va +e−τ
V2 · · ·VN
)+
N∑
b=ℓ+1
Vb
= P( e−τ
V2 · · ·VNG(Vi)
)+
N∑
b=ℓ+1
Vb
(VI.8.13c)
G(V2, · · · , VN ) ≡ 1 + eτV2 · · ·VN
( ℓ∑
a=2
Va + 1). (VI.8.13d)
130 Mirror Symmetry: T-duality of GLSM
これらの変換によって period integral が書き換えられる:
∴ Π =
∫ N∏
i=2
(dVi
Vi
)dP exp
− P
( e−τ
V2 · · ·VNG(Vi)
)−
N∑
b=ℓ+1
Vb
=
∫ N∏
i=2
(dVi
Vi
)δ( e−τ
V2 · · ·VNG(Vi)
)exp
(−
N∑
b=ℓ+1
Vb
)
=
∫ ℓ∏
a=2
dV ′a
N∏
b=ℓ+1
dVb δ(G(V ′
a, Vb))
exp(−
N∑
b=ℓ+1
Vb
), V ′
a = eτ/ℓVa . (VI.8.14)
ここでは canonical measure を持っている。
ここで次の re-definition を定義しよう:
V ′a ≡
Xℓa
X2 · · ·XN, Vb ≡ Xℓ
b . (VI.8.15a)
これは、Vi を不変に保つ次の対称性が存在する:
Xa 7→ ωaXa , Xb 7→ ωbXb , ωℓa = ωℓ
b = ω2 · · ·ωN = 1 (VI.8.15b)
これより、Xi で記述される理論には (Zℓ)N−2 orbifold symmetry が存在する。(VI.8.14) の G(Vi) = 0 から読み取れ
るように、その理論の fields Xi はまず次の space M に住んでいる:
M =
(X2, · · · ,XN ) ∈ CN−1∣∣∣ 0 = 1 +
ℓ∑
a=2
Xℓa + ψX2 · · ·Xℓ , ψ = eτ/ℓXℓ+1 · · ·XN
/(Zℓ)
N−2
(VI.8.16)
そして twisted LG theory は M 上に
WfM =
N∑
b=ℓ+1
Xℓb (VI.8.17)
という twisted superpotential を持つ。これは (VI.8.11) と全く同一である。これより homogeneous coordinates で
表すと (VI.8.9) で記述されることはすぐにわかる。
VI.9 Mirror dual of O(−N + ℓ) bundle on CPN−1[ℓ]
VI.9.1 Field configuration
Let us analyze the T-dual theory of the GLSM for the O(−N + ℓ) bundles on CPN−1[ℓ]. The field configuration
is assigned as follows:
chiral superfield Φa S1 . . . SN P1 P2
U(1) charge Qa 1 . . . 1 −ℓ −N + ℓ
twisted chiral superfield Ya Y1 . . . YN YP1YP2
(VI.9.1)
VI.9 Mirror dual of O(−N + ℓ) bundle on CPN−1[ℓ] 131
The twisted chiral superfields Ya are periodic variables Ya ≡ Ya + 2πi. They are defined from the chiral superfields
Φa via 2Φa e2QaV Φa = Ya+Ya. As we have already discussed, the twisted superpotential W and the period integral
Π, given by the followings, play key roles:
W = Σ( N∑
i=1
Yi − ℓYP1− (N − ℓ)YP2
− t)
+
N∑
i=1
e−Yi + e−YP1 + e−YP2 , (VI.9.2a)
Π =
∫dΣ
N∏
i=1
dYi dYP1dYP2
(ℓΣ) exp(− W
). (VI.9.2b)
Let us take the IR limit e → ∞ in order to consider the low energy effective theories. It is clear that the
dynamics of Σ is frozen and this superfield becomes just an auxiliary superfield. Thus we must replace the factor
ℓΣ in the period integral (VI.9.2b) to appropriate variables.
VI.9.2 Mirror Landau-Ginzburg descriptions
In this subsection we will derive LG theories with orbifold symmetries. In order to do this, we change the variable
ℓΣ in the period integral (VI.9.2b) to
ℓΣ → ℓ∂
∂t.
This replacing can be easily performed because of the existence of the term Σ(∑
aQaYa−t) in Π. Then we integrate
out the superfield Σ and obtain
Π = ℓ∂
∂t
∫ N∏
i=1
dYi dYP1dYP2
δ(∑
i
Yi − ℓYP1− (N − ℓ)YP2
− t)
exp(−∑
i
e−Yi − e−YP1 − e−YP2
). (VI.9.3)
Next let us solve the δ-function in this function. We note that there are two ways to solve it. One is to write the
variable YP1in terms of Yi and YP2
. The other is to solve YP2by Yi and YP1
. Both two solutions give consistent
LG theories with orbifold symmetries.
Solution one: Zℓ orbifolded LG theory
Let us solve the variable YP1via the δ-function in (VI.9.3):
−YP1=
1
ℓ
(t−
N∑
i=1
Yi + (N − ℓ)YP2
).
Performing the t-derivative in (VI.9.3) after the substitution of this solution, we obtain
Π = et/ℓ
∫ N∏
i=1
(e−
1ℓYidYi
) (e
N−ℓℓ
YP2 dYP2
)exp
(−∑
i
e−Yi − et/ℓ∏
i
e−1ℓYie
N−ℓℓ
YP2 − e−YP2
).
132 Mirror Symmetry: T-duality of GLSM
It is clear that the integral measure is not canonical. Transforming the variables into Xℓi := e−Yi and Xℓ
P2:=
e(N−ℓ)YP2 , we obtain the following period integral with a canonical measure up to an overall constant10:
Π =
∫ N∏
i=1
dXi dXP2exp
(−∑
i
Xℓi −X
− ℓN−ℓ
P2− et/ℓX1 · · ·XNXP2
). (VI.9.4)
Since Ya are periodic with respect to the shifts of their imaginary parts Ya ≡ Ya + 2πi, the new variables Xi and
XP2are symmetric under the following phase shifts:
Xi → ωiXi , XP2→ ωP2
XP2, ωℓ
i = ω− ℓ
N−ℓ
P2= ω1ω2 · · ·ωNωP2
= 1 . (VI.9.5)
We can read from (VI.9.4) and (VI.9.5) that the following orbifolded LG theory appears:
Wℓ =
N∑
i=1
Xℓi +X
− ℓN−ℓ
P2+ et/ℓX1 · · ·XNXP2
/(Zℓ)
N . (VI.9.6)
This theory is still ill-defined from the minimal model point of view. Even though the terms of positive powers
such as Xℓi are well-defined and they consist of N = 2 LG minimal model, there exists a term X
− ℓN−ℓ
P2, which does
not generate any critical points at finite XP2. However there is an interpretation to avoid this difficulty. Recall
a discussion on the linear dilaton CFT and the Liouville theory [35, 34]. (We prepare a brief review in appendix
VI.A.) Based on this argument, we can interpret the negative power term corresponds to Z−k0 in (VI.A.4), which
gives an N = 2 SCFT on the coset SL(2,R)k/U(1) at level k assigned by
k =ℓ
N − ℓ .
This assignment is correct because the conformal weights ra in the appendix VI.A are all ra = 1/ℓ, where n+1 = N .
Thus we obtain rΩ =∑
a ra − 1 = N/ℓ − 1 ≡ 1/k, which gives the above equation. This theory is given as an
N = 2 Kazama-Suzuki model on the coset SL(2,R)k/U(1) at level k [60], which is the gauged WZW model on
the two-dimensional Euclidean black hole [110]. Furthermore this theory is exactly equivalent to N = 2 Liouville
theory of background charge Q2 = 2/k via T-duality [34, 51]. We will continue to argue in later discussions.
Solution two: ZN−ℓ orbifolded LG theory
In the same analogy of the previous discussion11, we study the theory of the second solution
−YP2=
1
N − ℓ(t−
N∑
i=1
Yi + ℓYP1
),
which comes from the δ-function in the period integral (VI.9.3). Substituting this into (VI.9.3), we find
Π =
∫ N∏
i=1
(e−
1N−ℓ
YidYi
) (e
ℓN−ℓ
YP1 dYP1
)exp
(−∑
i
e−Yi − et
N−ℓ
∏
i
e−1
N−ℓYie
ℓN−ℓ
YP1 − e−YP1
).
10It is not serious to ignore an overall constant.11From now on we omit overall constant factors which appear in the period integral.
VI.9 Mirror dual of O(−N + ℓ) bundle on CPN−1[ℓ] 133
Performing the re-definitions XN−ℓi := e−Yi and XN−ℓ
P1:= eℓYP1 , we find that the period integral has a canonical
measure and the “ill-defined” LG theory with orbifold symmetry appears:
WN−ℓ =
N∑
i=1
XN−ℓi +X
−N−ℓℓ
P1+ e
tN−ℓX1 · · ·XNXP1
/(ZN−ℓ)
N . (VI.9.7)
Applying the discussions in appendix VI.A to the negative power term in the superpotential WN−ℓ, we find that the
theory is also described by the well-defined LG theory with an orbifold symmetry coupled to N = 2 Kazama-Suzuki
model on the coset SL(2,R)k/U(1) at level k, which is given by
N − ℓℓ
= k =2
Q2.
where Q is the charge of equivalent N = 2 Liouville theory.
VI.9.3 Mirror geometry descriptions
In the previous subsection we found two orbifolded LG theories as exact effective theories. They are obtained by
solving the twisted chiral superfields YP1and YP2
, respectively. Next we will read geometric informations from the
same period integral (VI.9.2b). Here we will also obtain two solutions which are related to the LG theories. The
derivation procedure is so complicated that we try to imitate the method discussed in section 7.3 of [54] and we
develop detailed calculations, explicitly. In order to obtain the geometric informations in the IR limit, we integrate
out the superfield Σ in the period integral (VI.9.2b) after the replacement of ℓΣ in (VI.9.2b) to other variables, as
we performed before.
Zℓ orbifolded geometry
Let us study how to obtain the geometry with Zℓ-type orbifold symmetry. Replacing ℓΣ in the period integral
(VI.9.2b) to
ℓΣ → ∂
∂YP1
,
we can perform the integration of Σ and obtain
Π =
∫ N∏
i=1
dYi
(e−YP1 dYP1
)dYP2
× δ(∑
i
Yi − ℓYP1− (N − ℓ)YP2
− t)
exp(−∑
i
e−Yi − e−YP1 − e−YP2
). (VI.9.8)
We perform the re-definitions of the variables Yi, YP1and YP2
:
e−YP1 =: P1 , e−Ya =: P1 Ua for a = 1, . . . , ℓ ,
e−YP2 =: P2 , e−Yb =: P2 Ub for b = ℓ+ 1, . . . , N .
134 Mirror Symmetry: T-duality of GLSM
Substituting these re-defined variables into (VI.9.8), we continue the calculation:
Π =
∫ N∏
i=1
(dUi
Ui
)dP1
(dP2
P2
)δ(
log(∏
i
Ui
)+ t)
exp− P1
( ℓ∑
a=1
Ua + 1)− P2
( N∑
b=ℓ+1
Ub + 1)
=
∫ ∏
i
(dUi
Ui
)dP2 du dv δ
(log(∏
i
Ui
)+ t)δ(∑
a
Ua + 1)
exp− P2
(∑
b
Ub + 1− uv)
=
∫ ∏
i
(dUi
Ui
)du dv δ
(log(∏
i
Ui
)+ t)δ(∑
a
Ua + 1)δ(∑
b
Ub + 1− uv), (VI.9.9)
where we introduced new variables u and v taking values in C and used a following equation
1
P2
=
∫du dv exp
(P2 uv
).
It is obvious that the resulting function (VI.9.9) still includes a non-canonical integral measure. Thus we perform
further re-definitions such as
Ua =: e−t/ℓ Zℓa
Z1 · · ·ZN, Ub =: Zℓ
b .
Note that the period integral (VI.9.9) is invariant under the following transformations acting on the new variables
Zi:
Za 7→ λωa Za , Zb 7→ ωb Zb , ωℓa = ωℓ
b = ω1 · · ·ωN = 1 ,
where λ is an arbitrary number taking in C∗. The ωi come from the shift symmetry of the original variables
Yi ≡ Yi + 2πi. Combining these transformations we find that the period integral has C∗ × (Zℓ)N−2 symmetries.
Substituting Zi into (VI.9.9), we obtain
Π =
∫1
vol.(C∗)
N∏
i=1
dZi du dv δ( ℓ∑
a=1
Zℓa + et/ℓZ1 · · ·ZN
)δ( N∑
b=ℓ+1
Zℓb + 1− uv
),
which indicates that the resulting mirror geometry is described by
Mℓ =
(Zi;u, v) ∈ CN+2∣∣∣F(Zi) = 0
/C∗ , G(Zb;u, v) = 0
/(Zℓ)
N−2 , (VI.9.10a)
F(Zi) :=
ℓ∑
a=1
Zℓa + ψZ1 · · ·Zℓ , G(Zb;u, v) :=
N∑
b=ℓ+1
Zℓb + 1− uv , (VI.9.10b)
ψ := et/ℓZℓ+1 · · ·ZN . (VI.9.10c)
This is an (N − 1)-dimensional complex manifold. It is guaranteed that Mℓ is a CY manifold because of the
following reason: We have already seen that the FI parameter t in (VI.9.2b) does not renormalized owing to the
CY condition∑
aQa = 0, which is also valid in the T-dual theory. In addition, we took the IR limit e → ∞ and
obtained the above non-trivial result. This means that the sigma model on the above geometry is a superconformal
sigma model.
Let us study the manifold Mℓ defined in (VI.9.10) more in detail. The equation F(Zi) = 0 denotes that the
complex variables Za consist of the degree ℓ hypersurface in the projective space: CPℓ−1[ℓ]. This subspace itself is
VI.9 Mirror dual of O(−N + ℓ) bundle on CPN−1[ℓ] 135
a compact CY manifold, which is parametrized by a parameter ψ which is subject to the equation G(Zb;u, v) = 0.
Moreover we can also interpret that the total space is a noncompact CY manifold whose compact directions are
described by Zi, while the variables u and v run in the noncompact directions under the equations (VI.9.10b).
Here let us comment on a relation between the manifold Mℓ and the LG twisted superpotential (VI.9.6). As
we have described in (VI.9.10), Mℓ has (Zℓ)N−2 orbifold symmetry, while the LG theory (VI.9.6) also holds this
type of orbifold symmetry, i.e., the (Zℓ)N orbifold symmetry. When we combine the two equations in (VI.9.10b)
as follows:
F (Zi, u, v) ≡ F(Zi) + G(Zb, u, v) =
N∑
i=1
Zℓi + et/ℓZ1 · · ·ZN + (1− uv) = 0 .
This function F = 0 is quite similar to the LG twisted superpotential Wℓ including negative power term (VI.9.6).
Recall that a LG theory written by a superpotential W is identical with a CY space defined by W = 0 in a
(weighted) projective space. (See, for examples, [78, 41].) If we can apply this argument to the above result, the
LG theory (VI.9.6) is identical with the sigma model on (VI.9.10) and there also exists the CY/LG correspondence
in the T-dual theory.
ZN−ℓ orbifolded geometry
We have constructed the two LG theories: (Zℓ)N orbifolded LG theory and (ZN−ℓ)
N orbifolded LG theory. The
former is related to the CY geometry Mℓ. It is natural to consider there also exists a dual geometry related to
the latter LG theory. In the previous calculation, we replaced the ℓΣ in the period integral (VI.9.2b) to ∂∂YP1
and
we obtained the (Zℓ)N−2 orbifolded geometry. Here we replace ℓΣ to the differential with respect to YP2
, which is
dual of the chiral superfield P2 of charge −(N − ℓ):
ℓΣ → ℓ
N − ℓ∂
∂YP2
.
Substituting this into (VI.9.2b), we obtain the following expression:
Π =
∫ N∏
i=1
dYi dYP1
(e−YP2 dYP2
)
× δ(∑
i
Yi − ℓYP1− (N − ℓ)YP2
− t)
exp(−∑
i
e−Yi − e−YP1 − e−YP2
). (VI.9.11)
Let us perform the following re-definitions of the variables Yi, YP1and YP2
:
e−YP1 =: P1 , e−Ya =: P1 Ua for a = 1, . . . , ℓ ,
e−YP2 =: P2 , e−Yb =: P2 Ub for b = ℓ+ 1, . . . , N .
Substituting the re-defined variables into (VI.9.11) and introducing auxiliary variables u and v in order to integrate
out P1 completely, we obtain
Π =
∫ N∏
i=1
(dUi
Ui
)du dv δ
(log( N∏
i=1
Ui
)+ t)δ( N∑
b=ℓ+1
Ub + 1)δ( ℓ∑
a=1
Ua + 1− uv). (VI.9.12)
136 Mirror Symmetry: T-duality of GLSM
The integral measure still remains non-canonical. We next introduce further re-definitions of Ui:
Ua =: ZN−ℓa , Ub =: e−t/(N−ℓ) ZN−ℓ
b
Z1 · · ·ZN.
We can see that the map from Zi to Ui is one-to-one modulo the C∗ × (ZN−ℓ)N−2 action given by
Za 7→ ωaZa , Zb 7→ λωbZb , ωN−ℓa = ωN−ℓ
b = ω1 · · ·ωN = 1 ,
where λ takes value in C∗. On account of the above re-definitions and symmetries we find that the period integral
is re-written as
Π =
∫1
vol.(C∗)
N∏
i=1
dZi du dv δ( ℓ∑
a=1
ZN−ℓa + 1− uv
)δ( N∑
b=ℓ+1
ZN−ℓb + et/(N−ℓ)Z1 · · ·ZN
),
from which we can read the geometric information described by
MN−ℓ =
(Zi;u, v) ∈ CN+2∣∣∣F(Za;u, v) = 0 ,
G(Zi) = 0
/C∗/
(ZN−ℓ)N−2 , (VI.9.13a)
F(Za;u, v) :=ℓ∑
a=1
ZN−ℓa + 1− uv , G(Zi) :=
N∑
b=ℓ+1
ZN−ℓb + ψZℓ+1 · · ·ZN , (VI.9.13b)
ψ := et/(N−ℓ)Z1 · · ·Zℓ . (VI.9.13c)
This is also a noncompact CY manifold including a compact CY hypersurface CPN−ℓ−1[N − ℓ], which is defined
by G(Zi) = 0 and parametrized by ψ with being subject to F(Za;u, v) = 0. Since the variables are the twisted
chiral superfields, we obtained the N = 2 supersymmetric NLSM on MN−ℓ as a low energy effective theory of the
T-dual theory. We can see that the sigma model on this manifold is identical with the LG theory described by
(VI.9.7).
VI.9.4 Return to the gauged linear sigma model
As discussed before, it has been proved that the N = 2 SCFT on coset SL(2,R)k/U(1) at level k is exactly T-dual
with the N = 2 Liouville theory of background charge Q under the relation Q2 = 2/k. Let us apply this argument
to the GLSM and its T-dual. Notice that the massless effective theories in the T-dual theory are exact, whereas
the ones in the original GLSM are approximately realized.
Now let us recall that if a CFT C has an abelian discrete symmetry group Γ, the orbifold CFT C′ = C/Γ has
a symmetry group Γ′ which is isomorphic to Γ and a new orbifold CFT C′/Γ′ is identical to the original CFT C.Including this argument into the effective theories of the GLSM and its T-dual theory of 2 ≤ ℓ ≤ N − 1, we find
that the theories (I.7.10) and (I.7.11) are equivalent to (VI.9.6) and (VI.9.7), respectively. Furthermore we can
interpret that the theories (I.7.10) and (I.7.11) are described by N = 2 Liouville theories coupled to the well-defined
LG minimal models as exact effective theories. As a result we obtain the non-trivial relations among the various
effective theories in the GLSM. Here we refer one typical result. The CY sigma model on (I.7.5) corresponds to
(I.7.10), which is deformed to the LG theory coupled to the Liouville theory as an exact quantum theory. This
is equivalent to (VI.9.6) via T-duality. On account of the CY/LG correspondence, (VI.9.6) and sigma model on
VI.9 Mirror dual of O(−N + ℓ) bundle on CPN−1[ℓ] 137
(VI.9.10) are identical with each other. Finally the original CY manifold (I.7.5) and (VI.9.10) are mirror dual with
each other. Notice that the CY manifold MCY is also deformed because the Liouville theory indicates that the
dilaton field propagates on the target space [84, 85]. Of course we find that there are the same relations among
effective theories (I.7.11), (I.7.12), (VI.9.7) and (VI.9.13).
Let us consider the case ℓ = 1. As discussed before, the GLSM has only two massless effective theories (I.7.6)
and (I.7.20). In addition, the subspace CPℓ−1[ℓ] in (VI.9.10) is ill-defined if ℓ = 1 and then the LG description
(VI.9.6) is also ill-defined. Thus the T-dual theory has only two descriptions (VI.9.7) and (VI.9.13) in the IR limit.
This situation is consistent with the result in [54], where the GLSM for O(−N) bundle on CPN−1 and its T-dual
was discussed.
VI.9.5 Summary and Discussions
We have studied the GLSM for noncompact CY manifolds realized as a line bundle on a hypersurface in a projective
space. This gauge theory has three non-trivial phases and includes two types of four massless effective theories in
the IR limit. Two theories are NLSMs on two distinct manifolds, whereas the other two are LG theories coupled
to complex one-dimensional SCFTs. Following the conventional arguments, we have interpreted that these four
theories are related to each other under phase transitions such as CY/LG correspondences and a topology change.
Performing the T-duality, we have also obtained two types of four exact massless effective theories; the two theories
are the sigma models on newly appeared mirror CY manifolds, while the other two are the LG theories including
the terms of negative power −k, which may be regarded as indicating N = 2 SCFTs on coset SL(2,R)k/U(1) at
level k. Since the SCFT on this coset is exactly equivalent to the N = 2 Liouville theory via T-duality, we have
argued that the LG effective theories derived from the original GLSM are exactly realized by the Liouville theories
coupled to the well-defined LG minimal models. The relations among the theories are illustrated in Figure VI.3:
CY sigma model
on noncompact MCY
N = 2 Liouville × LG
Zℓ orbifold theory
GLSM
Hori-Vafa
CY sigma model
on noncompact Mℓ
SL(2, R)k/U(1) × LG
(Zℓ)N orbifold theory
CY/LG
CY/LG
T-dualMirror T-dual
FI ≫ 0
ℓΣ → ℓ∂
∂t
FI ≪ 0
ℓΣ →∂
∂YP1
Figure VI.3: Relations among IR effective theories of GLSM and its T-dual.
Utilizing the above relations, we will obtain the topological charges of a CY manifold from the exact effective
theories in the T-dual theory, even though we cannot directly calculate them in the original sigma model. Further-
138 Mirror Symmetry: T-duality of GLSM
more we will understand noncompact CY manifolds in detail from the mathematical point of view. In addition,
we can interpret the holographic duality in type II string theory on noncompact (singular) CY manifolds [34] as
the phase transition and T-duality of the two-dimensional worldsheet theory and will be able to understand this
duality more closely in the framework of the worldsheet sigma model description [23, 80, 77, 52].
As mentioned in the introduction, we have constructed the noncompact CY manifolds as line bundles on HSSs
[46]. The base spaces HSSs can be seen as the submanifolds in the projective spaces obtained by polynomials with
additional symmetries [45]: the quadric surface SO(N)/[SO(N − 2)×U(1)] is given by a polynomial of degree two
with SO(N) symmetry, and E6/[SO(10) × U(1)] has a set of differential equations including E6 isometry group.
These symmetries give the information of the complex structures of not only the base spaces HSSs but also the
noncompact CY manifolds. However, the T-dual theory [54] is only valid when we consider the GLSM without a
superpotential or with a superpotential given simply by a homogeneous polynomial such as WGLSM = P · Gℓ(S).
Even though the polynomial Gℓ(S) has an additional symmetry, the period integral (VI.4.2) cannot recognize the
existence of this additional symmetry. Thus the T-dual theory does not map all structures of the CY M to the
mirror geometry completely. For example, we can argue the sigma model on the resolved conifold and its mirror
dual in the framework of GLSM and its T-dual, however we have not even understood any correct descriptions
for the deformed conifold represented by the GLSM. Therefore, if we wish to obtain the correct T-dual theories of
the sigma models on such noncompact CY manifolds, we must improve the formulation so that it may recognize
the complex structure of the manifold. It is quite significant to solve this problem in order to understand mirror
symmetry for more general noncompact CY manifolds.
VI.A Linear dilaton CFT and Liouville theory 139
Appendix
VI.A Linear dilaton CFT and Liouville theory
In this appendix we demonstrate the linear dilaton CFT and the Liouville theory discussed in [35, 34]. Let us
consider the superstring propagating on the following the ten-dimensional spacetime:
Rd−1,1 ×X2n , 2n = 10− d ,
where X2n is a 2n-dimensional singular CY manifold. Sending the zero string coupling limit gs → 0 at fixed string
length ls gives rise to a d-dimensional theory without gravity describing the dynamics of modes living near the
singularity on X2n. This theory is holographic dual to string theory on a following background which approaches
at weak coupling region:
Rd−1,1 × Rφ ×M = Rd−1,1 × Rφ × S1 ×M/U(1) ,
where M is a compact and non-singular manifold. The real line Rφ is parameterized by φ.
We can define CFT on each subspace. On the flat space Rd−1,1 we can define N = 1 SCFT whose central
charge is
cd =3
2d . (VI.A.1)
We describe the theory on Rφ in terms of a linear dilaton given by Φ = −Q2 φ. The linear dilaton CFT has a central
charge
cφ = 1 + 3Q2 . (VI.A.2)
From the consistency of superstring propagation, the worldsheet theory on M should be an N = 1 SCFT with
central charge cM = 3(n − 1/2 − Q2). Moreover, if the manifold has a U(1) symmetry, the theory on the coset
manifold M/U(1) must be an extended N = 2 SCFT with central charge
cM/U(1) = 3(n− 1−Q2) . (VI.A.3)
Let us specialize the N = 2 SCFT onM/U(1) to the N = 2 LG minimal model whose superpotential is defined in
terms of n+ 1 chiral superfields Za:
WLG = F (Za) , F (λraZa) = λF (Za) , a = 1, 2, . . . , n+ 1 ,
where ra are the conformal weights of the chiral superfields Za, respectively. Note that we have already understood
properties of this minimal model [72, 105]. The worldsheet central charge should correspond to (VI.A.3) such as
cLG = 3
n+1∑
a=1
(1− 2ra
)≡ cM/U(1) .
If we introduce a new variable rΩ with respect to the conformal weights ra such as rΩ ≡∑
a ra− 1, we can express
the background charge Q to Q2 = 2rΩ.
140 Mirror Symmetry: T-duality of GLSM
Here let us combine the above discussion with the conjectures proposed by Muhki and Vafa [82], Ghoshal and
Vafa [32], and Ooguri and Vafa [88], where they insisted that an N = 2 SCFT on the noncompact space Rφ ×Mcan be given formally by the “LG” superpotential
W = −µZ−k0 + F (Za) , (VI.A.4)
where Z0 is an additional chiral superfield and
k =1
rΩ=
2
Q2. (VI.A.5)
This formulation is useful to describe the sigma model on deformed conifold [32]. The first term in the superpotential
appears to be ill-defined from the LG minimal model point of view. The corresponding potential does not have a
minimum at the finite value of Z0. The topological LG model with such a superpotential has already been studied
by Ghoshal and Mukhi [37], and Hanany, Oz and Ronen Plesser [44] in order to investigate two-dimensional string
theory. Moreover, in general, k is not an integer, which makes (VI.A.4) non-single valued. Thus, it was proposed
that this first term can be interpreted as an N = 2 SCFT on the coset SL(2,R)/U(1) at level k. From the geometric
point of view, this coset space corresponds to a semi-infinite cigar, and in the IR limit this geometry deforms to
the two-dimensional Euclidean black hole [110]. This SCFT on the coset had been believed to be isomorphic to the
Liouville theory in the sense of SCFT. They are related by strong-weak coupling duality on the worldsheet: The
theory (VI.A.4) can be valid as a Liouville theory in the large Q limit, while this can be seen as a coset SCFT in the
large k limit (k = 2/Q2). Finally, it has been proved that the N = 2 SCFT on the coset SL(2,R)k/U(1) is exactly
equivalent (or T-dual) to the N = 2 Liouville theory to each other in any values of k > 0 [51]. This equivalence was
also proved by Tong in the framework of two-dimensional domain wall physics in three-dimensional theory [102].
To summarize, we find that the string theory on a singular CY manifold X2n can be holographic dual to string
theory as a product theory of the N = 2 SCFT on the coset SL(2,R)k/U(1) and the N = 2 LG minimal model on
M/U(1). The coset SCFT sector is also equivalent to the N = 2 Liouville theory on Rφ × S1.
Appendix A
Conventions
142 Conventions
A.1 N = 1 supersymmetry in four dimensions
In this note we introduce some notation and convention about N = 1 supersymmetry in four dimensions defined
by Wess and Bagger [107]. It is quite useful to reduce these notations when we consider two-dimensional N = 2
supersymmetry.
First of all, we define the Minkowski signature metric ηµν in four dimensions:
ηµν = ηµν = diag(−1, 1, 1, 1) (A.1.1)
d’Alembertian in four-dimensions is defined by using this metric as
ηµν∂µ∂ν = ∂µ∂µ = −∂2
0 +∇2 ≡ (A.1.2)
Levi-Civita tensor in four-dimensional spacetime is defined by
ε0123 = −1 , ε0123 = 1 (A.1.3)
Let us consider the Weyl spinor convention in four dimensions. Levi-Civita tensor acting on the spinor indices is
defined by
εαβ =
(0 −1
1 0
), εαβ =
(0 1
−1 0
)(A.1.4)
Weyl spinor conventions are as follows:
ψα = εαβψβ , ψα = εαβψβ (A.1.5a)
ψχ = ψαχα = −ψαχα = χαψα = χψ (A.1.5b)
ψχ = ψαχα = −ψαχα = χαψ
α = χψ (A.1.5c)(χψ)†
=(χαψα
)†= ψαχ
α = ψχ = χψ (A.1.5d)
Utilizing the above Weyl spinor, we can construct the Dirac spinor:
ψD =
(χα
ψα
), ψD =
(χα , ψα
)γ0 =
(− ψα , −χα
)(Dirac spinor) . (A.1.6)
Dirac gamma matrices in four dimensions are
γµ =
(0 σµ
σµ 0
), γ5 = iγ0γ1γ2γ3 =
(12 0
0 −12
). (A.1.7)
Dirac equation is described as
(iγµ∂µ +m14
)ψD = 0 . (A.1.8)
Note that the matrices σµ are the Pauli matrices defined by
σ0 =
(−1 0
0 −1
), σ1 =
(0 1
1 0
), σ2 =
(0 −ii 0
), σ3 =
(1 0
0 −1
)(A.1.9)
A.1 N = 1 supersymmetry in four dimensions 143
These Pauli matrices have various relations:
(σµ)αα = εαβεαβ(σµ)ββ =⇒ εαβ(σµ)ββ = εβρ(σµ)αρ (A.1.10)
σ0 = σ0 , σi = −σi (A.1.11)(σµ
αβ
)†= σµ
βα , [(σµ)αβ ]† = (σµ)βα (A.1.12)
Identities
δβαδ
σρ = −1
2ηµν(σµ)αρ(σ
ν)σβ , δβαδ
σρ =
1
2
[δσαδ
βρ − (σµν)α
σ(σµν)ρβ]
(A.1.13)
(σµσν + σνσµ
)α
β = −2ηµνδβα ,
(σµσν + σνσµ
)αβ = −2ηµνδα
β(A.1.14a)
(σµ)αβ(σν)βσ = −ηµνδσα + 2(σµν)α
σ , (σµ)αβ(σν)βσ = −ηµνδασ + 2(σµν)α
σ (A.1.14b)
(σµν
)α
β =1
4
[(σµ)αβ(σν)ββ − (σν)αβ(σµ)ββ
](A.1.15a)
(σµν
)αβ =
1
4
[(σµ)αα(σν)αβ − (σν)αα(σµ)αβ
](A.1.15b)
(σµν
)α
α = Tr(σµν
)= 0 ,
(σµν
)α
βεβγ =(σµν
)γ
βεβα (A.1.15c)
Tr(σµνσρσ
)= −1
2
(ηµρηνσ − ηµσηνρ
)− i
2εµνρσ (A.1.16a)
Tr(σµσνσρσσ
)=(σµσν
)α
β(σρσσ
)β
α =[− ηµνδβ
α + 2(σµν)αβ][− ηρσδα
β + 2(σρσ)βα]
= 2ηµνηρσ − 2ηµν Tr(σρσ)− 2ηρσ Tr
(σµν
)+ 4Tr
(σµνσρσ
)
= 2ηµνηρσ − ηµρηνσ + ηµσηνρ − iεµνρσ
(A.1.16b)
Spinor algebra
θαθβ = −1
2εαβθθ , θαθβ =
1
2εαβθθ , θαθβ =
1
2εαβθθ , θαθβ = −1
2εαβθθ (A.1.17)
(θσµθ
)(θσνθ
)= −1
2θθθθηµν (A.1.18a)
(θφ)(θψ)
= −1
2
(φψ)(θθ),
(θφ)(θψ)
= −1
2
(φψ)(θθ)
(A.1.18b)
εαβ ∂
∂θβ= − ∂
∂θα, εαβ ∂
∂θα
∂
∂θβθθ = 4 , εαβ
∂
∂θα
∂
∂θβ
θθ = 4 (A.1.19)
144 Conventions
χσµψ = −ψσµχ ,(χσµψ
)†= ψσµχ (A.1.20a)
χσµσνψ = ψσνσµχ ,(χσµσνψ
)†= ψσνσµχ (A.1.20b)
(ψφ)χβ = −1
2ηµν
(φσµχ
)(ψσν
)β
(A.1.20c)
Differential operators
Dα =∂
∂θα+ i(σµ)ααθ
α ∂
∂xµ, Dα = − ∂
∂θα− iθα(σµ)αα
∂
∂xµ(x, θ, θ) (A.1.21a)
Dα =∂
∂θα+ 2i(σµ)ααθ
α ∂
∂yµ, Dα = − ∂
∂θα(y, θ, θ) (A.1.21b)
Dα =∂
∂θα, Dα = − ∂
∂θα− 2iθα(σµ)αα
∂
∂y†µ(y†, θ, θ) (A.1.21c)
Supercharges in differential representation
Qα =∂
∂θα− i(σµθ
)α∂µ , Qα = − ∂
∂θα+ i(θσµ
)α∂µ (x, θ, θ) (A.1.22a)
Qα =∂
∂θα, Qα = − ∂
∂θα+ 2i
(θσµ
)α∂µ (y, θ, θ) (A.1.22b)
Qα =∂
∂θα− 2i
(σµθ
)α∂µ , Qα = − ∂
∂θα(y†, θ, θ) (A.1.22c)
Anticommutation relations
Qα, Qα = 2i(σµ)αα∂µ = 2(σµ)ααPµ , Qα, Qβ = Qα, Qβ = 0 (A.1.23a)
Dα,Dα = −2i(σµ)αα∂µ = −2(σµ)ααPµ , Dα,Dβ = Dα,Dβ = 0 (A.1.23b)
Dα, Qβ = Dα, Qβ = Dα, Qβ = Dα, Qβ = 0 (A.1.23c)
A.2 N = (2, 2) supersymmetry in two dimensions
A.2.1 Spinors in this notes
The Clifford algebra in two dimensions (m = 0, 1) is reduced from the one in four dimensions (µ = 0, 1, 2, 3):
γm, γn = −2ηmn , ηmn = diag.(−,+) , (A.2.1)
where γm are Dirac’s gamma matrices in two dimensions.
Now we introduce another useful convention which appeared in [113]:
(θ1, θ2) = (θ−, θ+) , (A.2.2a)
A.2 N = (2, 2) supersymmetry in two dimensions 145
ε12 = ε21 = 1 → ε−+ = ε+− = 1 , (A.2.2b)
θ− = θ+ , θ+ = −θ− , (A.2.2c)
(θ−, θ+)∗ = (θ−, θ+) = (θ+,−θ−) . (A.2.2d)
Notice that we defined the complex conjugate of spinors as (A.2.2d), whose assignment comes from the fact that the
chirality is not flipped in two-dimensional Minkowski spacetime under the complex conjugate. The reason is that
the chirality operator γ3 in two-dimensional Minkowski spacetime, which will be defined in the next subsection, is
invariant under the complex conjugate. (Notice that the Lorentz symmetry group in two-dimensional Euclidean
space is U(1, 1). On the other hand, the chirality of the Weyl spinor in four-dimensional Minkowski spacetime
should be flipped. The Lorentz group in four-dimensional Euclidean space is SL(2,C). The positive (negative)
chiral spinors transform as the 2 (2) representations under SL(2,C).) The relations (A.2.2d) also implies that the
left-moving modes do not flip to the right-moving modes in two-dimensional Minkowski spacetime.
On the other hand, when we define supercovariant derivative (defined later), we must define the complex
conjugate of spinors in Euclidean spacetime (x0 = −ix2) as following relations (e.g., [55]) in order to preserve
consistency in Euclidean spacetime:
(θ−, θ+)∗ = (θ+, θ−) = (−θ−, θ+) . (A.2.3)
This definition comes from the fact that the chirality of the Weyl spinor in two-dimensional Euclidean space should
be flipped under the complex conjugate. We should also notice that the chirality in four-dimensional Eucildean
space is not changed. (Notice that the Lorentz symmetry group in two-dimensional Euclidean space is U(1), while
the Lorentz group in four-dimensional Euclidean space is SU(2) × SU(2). The former (latter) SU(2) denotes the
positive (negative) chiral spinors.) However, in order to avoid confusion, we only discuss supersymmetric theories
in two-dimensional Minkowski spacetime and use (A.2.2).
A.2.2 Weyl representation in two dimensions
It is useful to study the Weyl representation in two dimensions:
γ0 = σ2 , γ1 = iσ1 , γ3 = γ0γ1 = σ3 , C2 = −γ0 = −σ2 , (A.2.4)
where C2 is a charge conjugation matrix. In this convention, a Dirac spinor in two dimensions is nothing but a
Weyl spinor in four dimensions, and Weyl spinors Θ± in two dimensions are defined as
ΘL,R ≡1± γ3
2ΨDirac =
1± γ3
2
(χ
ξ
), ΘL = χ , ΘR = ξ . (A.2.5)
In two-dimensional Minkowski spacetime, the Lorentz group should be represented as SO(1, 1) ≃ U(1,C). We
can define two complex Weyl spinors (a left chiral χ and a right-chiral ξ) transforming in 1 and 1 representations
under the U(1,C) rotation. Note that the complex conjugate χ ≡ χ∗ acts as the spinors in 1 representation, i.e.,
the complex conjugation flips the chirality of Weyl spinors as in the same way as four-dimensional complex Weyl
spinors.
146 Conventions
However, we should notice that the spinor components χ and ξ do not correspond to the spinor components θ1
and θ2 in the previous subsection; the spinors θα are expressed as linear combinations of the Weyl spinors χ and
ξ. In this note we only follow the spinors described in the Witten’s article [113] and in the previous subsection.
A.2.3 Differential operators
The covariant superderivatives Dα, Dα and supercharges Qα, Qα are rewritten as follows:
D± =∂
∂θ±− iθ±
(∂0 ± ∂1
), D± = − ∂
∂θ±+ iθ±
(∂0 ± ∂1
), (A.2.6a)
Q± =∂
∂θ±+ iθ±
(∂0 ± ∂1
), Q± = − ∂
∂θ±− iθ±
(∂0 ± ∂1
). (A.2.6b)
The (anti-)commutators are also rewritten (−i∂0 ≡ H and i∂i ≡ Pi):
Q−, Q− = 2(σm)11Pm = 2(H + P3) , Q+, Q+ = 2(σm)22Pm = 2(H − P3) (A.2.7a)
Q−, Q+ = 2(σm)12Pm = 2(P1 − iP2) , Q+, Q− = 2(σm)21Pm = 2(P1 + iP2) . (A.2.7b)
Now we change the 4-dimensional translational generators such as P3 ≡ P and P1− iP2 ≡ 0. Then we describe the
most general N = (2, 2) supersymmetry (anti-)commutation relations1:
Q2+ = Q2
− = Q2+ = Q2
− = 0 , (A.2.8a)
Q±, Q± = −2i(∂0 ± ∂1) = 2(H ∓ P ) , (A.2.8b)
Q+, Q− = 0 , Q+, Q− = 0 , Q−, Q+ = 0 , Q+, Q− = 0 , (A.2.8c)
D±,D± = 2i(∂0 ± ∂1) , (A.2.8d)
Dα,Dβ = Dα,Dβ = D±,D∓ = 0 , (A.2.8e)
Dα, Qβ = Dα, Qβ = Dα, Qβ = Dα, Qβ = 0 , (A.2.8f)
[M,Q±] = ∓Q± , [M,Q±] = ∓Q± , (A.2.8g)
[FV , Q±] = −Q± , [FV , Q±] = Q± , (A.2.8h)
[FA, Q±] = ∓Q± , [FA, Q±] = ±Q± . (A.2.8i)
where M is a two-dimensional Lorentz generator; FV , FA are vector R-rotation, axial R-rotation generators,
respectively. These R-rotations are defined by
eiαFV : F(x, θ±, θ±) 7→ eiαqV F(x, e−iαθ±, eiαθ±) , (A.2.9a)
eiβFA : F(x, θ±, θ±) 7→ eiβqAF(x, e∓iβθ±, e±iβθ±) , (A.2.9b)
where F(x, θ±, θ±) is an arbitrary superfield and qV , qA are R-charges of F .
Notice that [53, 1] incorporates the following definition of the worldsheet light-cone coordinates:
x± ≡ x0 ± x1 , ∂± ≡∂
∂x±=
1
2
( ∂
∂x0± ∂
∂x1
), (A.2.10)
1When we regard the Qα as operators, we can relax the relations (A.2.8) by introducing central charges Z and eZ such as Q+, Q− =
Z and Q−, Q+ = eZ. See chapter 12 in [53]. However we mainly discuss theories without these central charges.
A.2 N = (2, 2) supersymmetry in two dimensions 147
which is different from the one in [74], defined as
∂++=
=∂
∂x0± ∂
∂x1= ∂0 ± ∂1 . (A.2.11)
We must carefully use the above symbols (or ∂0 ± ∂1) in order not to confuse as well as possible.
Here we define the measure of fermionic coordinates in the superspace:
d2θ = −1
4dθα dθβ εαβ = −1
2dθ+ dθ− , d2θ = −1
4dθα dθβ ε
αβ =1
2dθ+ dθ− , (A.2.12a)
d2θ = −1
2dθ+ dθ− , d2θ = −1
2dθ− dθ+ , (A.2.12b)
d4θ = d2θ d2θ = −d2θ d2θ = −1
4dθ+ dθ− dθ+ dθ− . (A.2.12c)
We define integrals of superspace coordinates as follows:
∫d2θ θθ = 1 ,
∫d2θ θθ = 1 ,
∫d2θ θ+θ− =
1
2,
∫d2θ θ−θ+ =
1
2. (A.2.13)
A.2.4 Superfields
Here we briefly mention various superfields inN = (2, 2) supersymmetric theories. A general superfield F(x, θ±, θ±)
can be expanded by fermionic coordinates θ± and θ±:
F = φ+ i√
2θ+ψ+ + i√
2θ−ψ− + i√
2θ+χ+ + i√
2θ−χ−
+ 2iθ+θ−F + 2iθ+θ−M + 2iθ+θ−G+ 2iθ+θ−N + θ−θ−(v0 − v1) + θ+θ+(v0 + v1)
− 2iθ+θ−θ+ζ+ − 2iθ+θ−θ−ζ− + 2iθ+θ−θ+λ+ + 2iθ+θ−θ−λ− − 2θ+θ−θ+θ−D . (A.2.14)
where the terms θ−θ−(v0− v1) + θ+θ+(v0 + v1) came from (θ−, θ+)(γmvm)
(θ−
θ+
)= θσµθvµ in four-dimensional
theory. Note that coefficients of component fields are determined in terms of the closure of supersymmetry and so
on. Here we (partly) adapted the descriptions of [107]2.
Here we prepare some constraints applied to superfields in order to obtain well-known representations:
0 = D+F= −i
√2χ+ + θ+
(v0 + v1) + i(∂0 + ∂1)φ
− 2iθ−M − 2iθ−N
+ i√
2θ+θ−√
2ζ+ + i(∂0 + ∂1)ψ−
+ i√
2θ+θ−√
2λ+ + i(∂0 + ∂1)χ−
+ 2iθ−θ−λ−
−√
2θ+θ+(∂0 + ∂1)χ+ − 2θ+θ+θ−(∂0 + ∂1)M − 2θ+θ+θ−(∂0 + ∂1)N
+ θ+θ−θ−2D + i(∂0 + ∂1)(v0 − v1)
− 2θ+θ−θ+θ−(∂0 + ∂1)λ− , (A.2.15a)
0 = D−F= −i
√2χ− + θ−
(v0 − v1) + i(∂0 − ∂1)φ
+ 2iθ+M + 2iθ+G
2Of course we can always change coefficients of component fields as the ones which we demand. The definitions of them in [53] are
slightly different.
148 Conventions
+ i√
2θ+θ−√
2ζ− − i(∂0 − ∂1)ψ+
+ i√
2θ+θ−√
2λ− − i(∂0 − ∂1)χ+
− 2iθ+θ+λ+
−√
2θ−θ−(∂0 − ∂1)χ− + 2θ−θ−θ+(∂0 − ∂1)M + 2θ−θ−θ+(∂0 − ∂1)G
+ θ+θ+θ−2D + i(∂0 − ∂1)(v0 + v1)
+ 2θ+θ−θ+θ−(∂0 − ∂1)λ+ , (A.2.15b)
0 = D+F= i√
2ψ+ + θ+(v0 + v1)− i(∂0 + ∂1)φ
+ 2iθ−F + 2iθ−G
+ i√
2θ+θ−√
2ζ+ − i(∂0 + ∂1)ψ−
+ i√
2θ+θ−√
2λ+ − i(∂0 + ∂1)χ−− 2iθ−θ−ζ−
−√
2θ+θ+(∂0 + ∂1)ψ+ − 2θ+θ+θ−(∂0 + ∂1)F − 2θ+θ+θ−(∂0 + ∂1)G
+ θ+θ−θ−2D − i(∂0 + ∂1)(v0 − v1)
− 2θ+θ−θ+θ−(∂0 + ∂1)ζ− , (A.2.15c)
0 = D−F= i√
2ψ− + θ−(v0 − v1)− i(∂0 − ∂1)φ
− 2iθ+F − 2iθ+N
+ i√
2θ+θ−√
2ζ− + i(∂0 − ∂1)ψ+
+ i√
2θ+θ−√
2λ− + i(∂0 − ∂1)χ+
+ 2iθ+θ+ζ+
−√
2θ−θ−(∂0 − ∂1)ψ− + 2θ−θ−θ+(∂0 − ∂1)F + 2θ−θ−θ+(∂0 − ∂1)N
+ θ+θ+θ−2D − i(∂0 − ∂1)(v0 + v1)
+ 2θ+θ−θ+θ−(∂0 − ∂1)ζ+ . (A.2.15d)
We define the supersymmetry transformation rule for a superfield F as follows:
δεF ≡[ai(−ε−Q+ + ε+Q−) + bi(−ε−Q+ + ε+Q−)
]F
≡ δεφ+ i√
2θ+δεψ+ + i√
2θ−δεψ− + i√
2θ+δεχ+ + i√
2θ−δεχ−
+ 2iθ+θ−δεF + 2iθ+θ−δεM + 2iθ+θ−δεG+ 2iθ+θ−δεN + θ−θ−δε(v0 − v1) + θ+θ+δε(v0 + v1)
− 2iθ+θ−θ+δεζ+ − 2iθ+θ−θ−δεζ− + 2iθ+θ−θ+δελ+ + 2iθ+θ−θ−δελ− − 2θ+θ−θ+θ−δεD , (A.2.16)
where two coefficients a and b are arbitrary and we can always choose suitable values. We usually set them to
a = b = 1.
Chiral superfield
A chiral superfield Φ is a superfield that satisfies the constraint
D±Φ = 0 . (A.2.17)
We expand this superfield by using component fields
Φ(x, θ±, θ±) = φ(x) + i√
2θ+ψ+(x) + i√
2θ−ψ−(x) + 2iθ+θ−F (x)
− iθ+θ+(∂0 + ∂1)φ− iθ−θ−(∂0 − ∂1)φ+ θ+θ−θ+θ−(∂20 − ∂2
1)φ
+√
2θ+θ+θ−(∂0 + ∂1)ψ− +√
2θ−θ−θ+(∂0 − ∂1)ψ+ , (A.2.18)
where F (x) is a complex auxiliary field and the terms in second and third lines of the right-hand side involve
only the derivatives of φ and ψ±. φ is a complex scalar field and ψ± are Weyl fermions (complex fermions) with
positive and negative chirality. (Here we remark that the name “chiral superfield” does not means two-dimensional
A.2 N = (2, 2) supersymmetry in two dimensions 149
chiral field. This comes from a four-dimensional “chiral” field.) We also write down the expansion of an anti chiral
superfield3:
Φ = φ+ i√
2θ+ψ+ + i√
2θ−ψ− + 2iθ+θ−F
+ iθ+θ+(∂0 + ∂1)φ+ iθ−θ−(∂0 − ∂1)φ+ θ+θ−θ+θ−(∂20 − ∂2
1)φ
−√
2θ+θ+θ−(∂0 + ∂1)ψ− −√
2θ−θ−θ+(∂0 − ∂1)ψ+ . (A.2.19)
We consider supersymmetry transformations of chiral superfields. First we prepare the following equations:
Q+Φ = i√
2ψ+ + 2iθ−F
+√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ+ + 2θ+θ+θ−(∂0 + ∂1)F
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+ , (A.2.20a)
Q−Φ = i√
2ψ− − 2iθ+F
+√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ− − 2θ−θ−θ+(∂0 − ∂1)F
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+ , (A.2.20b)
Q+Φ = −2iθ+(∂0 + ∂1)φ+ 2√
2θ+θ−(∂0 + ∂1)ψ− − 2θ+θ−θ−(∂20 − ∂2
1)φ , (A.2.20c)
Q−Φ = −2iθ−(∂0 − ∂1)φ− 2√
2θ+θ−(∂0 − ∂1)ψ+ − 2θ−θ+θ+(∂20 − ∂2
1)φ , (A.2.20d)
D+Φ = i√
2ψ+ + 2iθ−F − 2iθ+(∂0 + ∂1)φ+ 2√
2θ+θ−(∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ+ − 2θ+θ+θ−(∂0 + ∂1)F
− 2θ−θ−θ+(∂20 − ∂2
1)φ− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+ , (A.2.20e)
D−Φ = i√
2ψ− − 2iθ+F − 2iθ−(∂0 − ∂1)φ− 2√
2θ+θ−(∂0 − ∂1)ψ+
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ− + 2θ−θ−θ+(∂0 − ∂1)F
− 2θ+θ+θ−(∂20 − ∂2
1)φ− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ− , (A.2.20f)
D±Φ = 0 . (A.2.20g)
Q+Φ = 2iθ+(∂0 + ∂1)φ− 2√
2θ+θ−(∂0 + ∂1)ψ− − 2θ+θ−θ−(∂20 − ∂2
1)φ
= Q+Φ , (A.2.21a)
Q−Φ = 2iθ−(∂0 − ∂1)φ+ 2√
2θ+θ−(∂0 − ∂1)ψ+ − 2θ+θ+θ−(∂20 − ∂2
1)φ
= Q−Φ , (A.2.21b)
Q+Φ = −i√
2ψ+ − 2iθ−F +√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ+ + 2θ+θ+θ−(∂0 − ∂1)F
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+
= Q+Φ , (A.2.21c)
Q−Φ = −i√
2ψ− + 2iθ+F +√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ− − 2θ−θ−θ+(∂0 − ∂1)F
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ−3The conjugate rule for fermions is (η+χ−)† ≡ χ−η+.
150 Conventions
= Q−Φ , (A.2.21d)
D±Φ = 0 , (A.2.21e)
D+Φ = −i√
2ψ+ − 2iθ−F + 2iθ+(∂0 + ∂1)φ− 2√
2θ+θ−(∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ+ − 2θ+θ+θ−(∂0 + ∂1)F
− 2θ+θ−θ−(∂20 − ∂2
1)φ+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+
= D+Φ , (A.2.21f)
D−Φ = −i√
2ψ− + 2iθ+F + 2iθ−(∂0 − ∂1)φ+ 2√
2θ+θ−(∂0 − ∂1)ψ+
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ− + 2θ−θ−θ+(∂0 − ∂1)F
− 2θ−θ+θ+(∂20 − ∂2
1)φ+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ−
= D−Φ . (A.2.21g)
We find that the variation δεΦ is also a chiral superfield under the supersymmetry transformation (A.2.16). Thus
we can define the supersymmetry transformation of component fields as4
δεΦ ≡[aiεαQα + biεβQβ
]Φ ≡ δεφ+
√2θ+δεψ+ +
√2θ−δεψ− + 2θ+θ−δεF + · · · , (A.2.22)
and the resulting transformation rules of component fields are described as
δεφ =√
2ai(−ε−ψ+ + ε+ψ−) , (A.2.23a)
δεψ+ =√
2aiε+F +√
2bε−(∂0 + ∂1)φ , δεψ− =√
2aiε−F −√
2bε+(∂0 − ∂1)φ , (A.2.23b)
δεF = −√
2bε−(∂0 + ∂1)ψ− + ε+(∂0 − ∂1)ψ+
. (A.2.23c)
Twisted chiral superfield
A twisted chiral superfield Y is also an irreducible superfield defined by
D+Y = D−Y = 0 , D+Y = D−Y = 0 , (A.2.24)
and can be expanded as
Y (x, θ±, θ±) = y(x) + i√
2θ+χ+(x) + i√
2θ−χ−(x) + 2iθ+θ−G(x)
− iθ+θ+(∂0 + ∂1)y + iθ−θ−(∂0 − ∂1)y − θ+θ−θ+θ−(∂20 − ∂2
1)y
−√
2θ−θ−θ+(∂0 − ∂1)χ+ +√
2θ+θ+θ−(∂0 + ∂1)χ− , (A.2.25a)
Y (x, θ±, θ±) = y + i√
2θ+χ+ + i√
2θ−χ− + 2iθ+θ−G
+ iθ+θ+(∂0 + ∂1)y − iθ−θ−(∂0 − ∂1)y − θ+θ−θ+θ−(∂20 − ∂2
1)y
+√
2θ−θ−θ+(∂0 − ∂1)χ+ −√
2θ+θ+θ−(∂0 + ∂1)χ− , (A.2.25b)
where G(x) is a complex auxiliary field and the terms in second and third lines of the right-hand side involve only
the derivatives of y, χ− and χ+; y(x), χ− and χ+ are a complex scalar, a Weyl fermion with negative chirality, and
4Here we assume that coefficients a and b are arbitrary constant numbers.
A.2 N = (2, 2) supersymmetry in two dimensions 151
a Weyl fermion with positive chirality, respectively. We prepare the following operations:
Q+Y = i√
2χ+ + 2iθ−G
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ+ + 2θ+θ+θ−(∂0 + ∂1)G
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+ , (A.2.26a)
Q−Y = 2iθ−(∂0 − ∂1)y + 2√
2θ+θ−(∂0 − ∂1)χ+ + 2θ+θ+θ−(∂20 − ∂2
1)y , (A.2.26b)
Q+Y = −2iθ+(∂0 + ∂1)y + 2√
2θ+θ−(∂0 + ∂1)χ− + 2θ+θ−θ−(∂20 − ∂2
1)y , (A.2.26c)
Q−Y = −i√
2χ− + 2iθ+G
−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ− − 2θ−θ−θ+(∂0 − ∂1)G
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ− , (A.2.26d)
D+Y = i√
2χ+ + 2iθ−G− 2iθ+(∂0 + ∂1)y + 2√
2θ+θ−(∂0 + ∂1)χ−
−√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
χ+ − 2θ+θ+θ−(∂0 + ∂1)G
+ 2θ−θ−θ+(∂20 − ∂2
1)y + i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+ , (A.2.26e)
D−Y = 0 , D+Y = 0 , (A.2.26f)
D−Y = −i√
2χ− + 2iθ+G+ 2iθ−(∂0 − ∂1)y + 2√
2θ+θ−(∂0 − ∂1)χ+
−√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
χ− + 2θ−θ−θ+(∂0 − ∂1)G
+ 2θ+θ+θ−(∂20 − ∂2
1)y − i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ− , (A.2.26g)
Q+Y = 2iθ+(∂0 + ∂1)y − 2√
2θ+θ−(∂0 + ∂1)χ− + 2θ+θ−θ−(∂20 − ∂2
1)y
= Q+Y , (A.2.27a)
Q−Y = i√
2χ− − 2iθ+G−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ− − 2θ−θ−θ+(∂0 − ∂1)G
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ−
= Q−Y , (A.2.27b)
Q+Y = −i√
2χ+ − 2iθ−G+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ+ + 2θ+θ+θ−(∂0 + ∂1)G
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+
= Q+Y , (A.2.27c)
Q−Y = −2iθ−(∂0 − ∂1)y − 2√
2θ+θ−(∂0 − ∂1)χ+ + 2θ+θ+θ−(∂20 − ∂2
1)y
= Q−Y , (A.2.27d)
D+Y = 0 , D−Y = 0 , (A.2.27e)
D−Y = i√
2χ− − 2iθ+G− 2iθ−(∂0 − ∂1)y − 2√
2θ+θ−(∂0 − ∂1)χ+
−√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
χ− + 2θ−θ−θ+(∂0 − ∂1)G
+ 2θ+θ+θ−(∂20 − ∂2
1)y + i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ−
= D−Y , (A.2.27f)
D+Y = −i√
2χ+ − 2iθ−G+ 2iθ+(∂0 + ∂1)y − 2√
2θ+θ−(∂0 + ∂1)χ−
152 Conventions
−√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
χ+ − 2θ+θ+θ−(∂0 + ∂1)G
+ 2θ−θ−θ+(∂20 − ∂2
1)y − i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+
= D+Y . (A.2.27g)
Vector superfield
A vector superfield V is defined as a real superfield which contains two-dimensional vector fields:
V ≡ V † = φ+ i√
2(θ+ψ+ + θ+ψ+) + i√
2(θ−ψ− + θ−ψ−)
+ 2i(θ+θ−F + θ+θ−F ) + 2i(θ+θ−G+ θ+θ−G) + θ−θ−(v0 − v1) + θ+θ+(v0 + v1)
− 2i θ+θ−(θ+λ+ + θ−λ−
)+ 2i θ+θ−
(θ+λ+ + θ−λ−
)− 2θ+θ−θ+θ−D
= θ+θ+(v0 + v1
)+ θ−θ−
(v0 − v1
)−√
2θ−θ+σ −√
2θ+θ−σ
− 2i θ+θ−(θ+λ+ + θ−λ−
)+ 2i θ+θ−
(θ+λ+ + θ−λ−
)− 2θ+θ−θ+θ−D , (A.2.28)
where we fixed V on the Wess-Zumino gauge because we can always add a pair of chiral and anti-chiral superfield
Φ+Φ to V in order to delete some component fields; vm are gauge fields and D is a real auxiliary field; σ and σ are
complex conjugate with each other. Notice that the Wess-Zumino gauge is not invariant under the supersymmetry
transformation, i.e., after performing supersymmetry transformations, we should impose a new Wess-Zumino gauge
fixing condition on the new superfield V ′ = V + δεV . Here we describe the supersymmetry operations without
Wess-Zumino gauge-fixing condition5:
Q+V = i√
2ψ+ + 2iθ−F + 2iθ−G+ θ+(v0 + v1) + i(∂0 + ∂1)φ
+ i√
2θ+θ−√
2λ+ + i(∂0 + ∂1)ψ−
+ i√
2θ+θ−√
2λ+ + i(∂0 + ∂1)ψ−
+√
2θ+θ+(∂0 + ∂1)ψ+ − 2iθ−θ−λ−
+ θ+θ−θ−2D + i(∂0 + ∂1)(v0 − v1)
+ 2θ+θ+(∂0 + ∂1)(θ
−F + θ−G)
+ 2θ+θ−θ+θ−(∂0 + ∂1)λ− , (A.2.29a)
Q−V = i√
2ψ− − 2iθ+F − 2iθ+G+ θ−(v0 − v1) + i(∂0 − ∂1)φ
+ i√
2θ+θ−√
2λ− − i(∂0 − ∂1)ψ+
+ i√
2θ+θ−√
2λ− − i(∂0 − ∂1)ψ+
+√
2θ−θ−(∂0 − ∂1)ψ− + 2iθ+θ+λ+
+ θ+θ+θ−2D + i(∂0 − ∂1)(v0 + v1)
− 2θ−θ−(∂0 − ∂1)(θ
+F + θ+G)
− 2θ+θ−θ+θ−(∂0 − ∂1)λ+ , (A.2.29b)
Q+V = −i√
2ψ+ − 2iθ−F − 2iθ−G+ θ+(v0 + v1)− i(∂0 + ∂1)φ
+ i√
2θ+θ−√
2λ+ − i(∂0 + ∂1)ψ−
+ i√
2θ+θ−√
2λ+ − i(∂0 + ∂1)ψ−
+√
2θ+θ+(∂0 + ∂1)ψ+ + 2iθ−θ−λ−
+ θ+θ−θ−2D − i(∂0 + ∂1)(v0 − v1)
+ 2θ+θ+(∂0 + ∂1)(θ
−F + θ−G)
5When we consider the supersymmetry transformation we want to obtain the transformation rules only in the superspace formalism.
In order to check this, we can calculate with the component fields but we should not fix to the Wess-Zumino gauge.
A.2 N = (2, 2) supersymmetry in two dimensions 153
+ 2θ+θ−θ+θ−(∂0 + ∂1)λ− , (A.2.29c)
Q−V = −i√
2ψ− + 2iθ+F + 2iθ+G+ θ−(v0 − v1)− i(∂0 − ∂1)φ
+ i√
2θ+θ−√
2λ− + i(∂0 − ∂1)ψ+
+ i√
2θ+θ−√
2λ− + i(∂0 − ∂1)ψ+
+√
2θ−θ−(∂0 − ∂1)ψ− − 2iθ+θ+λ+
+ θ+θ+θ−2D − i(∂0 − ∂1)(v0 + v1)
− 2θ−θ−(∂0 − ∂1)(θ
+F + θ+G)
− 2θ+θ−θ+θ−(∂0 − ∂1)λ+ , (A.2.29d)
D+V = i√
2ψ+ + 2iθ−F + 2iθ−G+ θ+(v0 + v1)− i(∂0 + ∂1)φ
+ i√
2θ+θ−√
2λ+ − i(∂0 + ∂1)ψ−
+ i√
2θ+θ−√
2λ+ − i(∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1)ψ+ − 2iθ−θ−λ−
+ θ+θ−θ−2D − i(∂0 + ∂1)(v0 − v1)
− 2θ+θ+(∂0 + ∂1)(θ
−F + θ−G)
− 2θ+θ−θ+θ−(∂0 + ∂1)λ− , (A.2.29e)
D−V = i√
2ψ− − 2iθ+F − 2iθ+G+ θ−(v0 − v1)− i(∂0 − ∂1)φ
+ i√
2θ+θ−√
2λ− + i(∂0 − ∂1)ψ+
+ i√
2θ+θ−√
2λ− + i(∂0 − ∂1)ψ+
−√
2θ−θ−(∂0 − ∂1)ψ− + 2iθ+θ+λ+
+ θ+θ+θ−2D − i(∂0 − ∂1)(v0 + v1)
+ 2θ−θ−(∂0 − ∂1)(θ
+F + θ+G)
+ 2θ+θ−θ+θ−(∂0 − ∂1)λ+ , (A.2.29f)
D+V = −i√
2ψ+ − 2iθ−F − 2iθ−G+ θ+(v0 + v1) + i(∂0 + ∂1)φ
+ i√
2θ+θ−√
2λ+ + i(∂0 + ∂1)ψ−
+ i√
2θ+θ−√
2λ+ + i(∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1)ψ+ + 2iθ−θ−λ−
+ θ+θ−θ−2D + i(∂0 + ∂1)(v0 − v1)
− 2θ+θ+(∂0 + ∂1)(θ
−F + θ−G)
− 2θ+θ−θ+θ−(∂0 + ∂1)λ− , (A.2.29g)
D−V = −i√
2ψ− + 2iθ+F + 2iθ+G+ θ−(v0 − v1) + i(∂0 − ∂1)φ
+ i√
2θ+θ−√
2λ− − i(∂0 − ∂1)ψ+
+ i√
2θ+θ−√
2λ− − i(∂0 − ∂1)ψ+
−√
2θ−θ−(∂0 − ∂1)ψ− − 2iθ+θ+λ+
+ θ+θ+θ−2D + i(∂0 − ∂1)(v0 + v1)
+ 2θ−θ−(∂0 − ∂1)(θ
+F + θ+G)
+ 2θ+θ−θ+θ−(∂0 − ∂1)λ+ . (A.2.29h)
We can define the field strength of this gauge field:
Σ =1
2√
2
D+,D− =
1√2D+D−V
= σ − i√
2 θ+λ+ − i√
2 θ−λ− +√
2 θ+θ−(D − iF01)
− iθ−θ− (∂0 − ∂1)σ − iθ+θ+ (∂0 + ∂1)σ +√
2 θ−θ+θ− (∂0 − ∂1)λ+
+√
2 θ+θ−θ+ (∂0 + ∂1)λ− − θ+θ−θ−θ+ (∂02 − ∂1
2)σ , (A.2.30)
154 Conventions
where F01 = ∂0v1 − ∂1v0 is a field strength of the gauge field; the gauge covariant superderivatives D± are
D± = e−V D±eV , D± = eV D±e−V . (A.2.31)
Furthermore we also write down the anti-chiral field strength:
Σ = (Σ)† =1√2D−D+V
= σ − i√
2 θ+λ+ − i√
2 θ−λ− +√
2 θ−θ+(D + iF01)
+ iθ−θ− (∂0 − ∂1)σ + iθ+θ+ (∂0 + ∂1)σ +√
2 θ−θ+θ− (∂0 − ∂1)λ+
+√
2 θ+θ−θ+ (∂0 + ∂1)λ− − θ+θ−θ−θ+ (∂02 − ∂1
2)σ . (A.2.32)
Spinor superfield
In four-dimensional theory Wess and Bagger [107] defined a spinor superfield Wα from vector superfield V such as
Wα = − 14DβD
βDαV . We can also define a spinor superfield in the same way as the above one:
W± = −1
2D+D−D±V , W± = −1
2D+D−D±V . (A.2.33)
These spinor superfields are invariant under chiral and gauge transformations. Explicit expressions of them are
W+ = − 1√2D+Σ
= −iλ+ − θ−(D + iF01)− i√
2θ+(∂0 + ∂1)σ + 2θ−θ+(∂0 + ∂1)λ−
+− iθ−θ−(∂0 − ∂1)− iθ+θ+(∂0 + ∂1)
(−iλ+) + iθ+θ+θ−(∂0 − ∂1)(D + iF01)
− iθ−θ−(∂0 − ∂1)− i√
2θ+(∂0 + ∂1)σ
+− iθ−θ−(∂0 − ∂1)
− iθ+θ+(∂0 + ∂1)
(−iλ+) , (A.2.34a)
W− = − 1√2D−Σ
= −iλ− − θ+(D − iF01)− i√
2θ−(∂0 − ∂1)σ + 2θ+θ−(∂0 − ∂1)λ+
+− iθ−θ−(∂0 − ∂1)− iθ+θ+(∂0 + ∂1)
(−iλ−) + iθ−θ−θ+(∂0 − ∂1)(D − iF01)
− iθ+θ+(∂0 + ∂1)−i√
2θ−(∂0 − ∂1)σ+− iθ−θ−(∂0 − ∂1)
− iθ+θ+(∂0 + ∂1)
(−iλ−) , (A.2.34b)
W+ = − 1√2D+Σ
= iλ+ − θ−(D − iF01) + i√
2θ+(∂0 + ∂1)σ − 2θ−θ+(∂0 + ∂1)λ−
+iθ+θ+(∂0 + ∂1) + iθ−θ−(∂0 − ∂1)
(iλ+)− iθ+θ+θ−(∂0 + ∂1)(D − iF01)
+ iθ−θ−(∂0 − ∂1)i√
2θ+(∂0 + ∂1)σ
+iθ−θ−(∂0 − ∂1)
iθ+θ+(∂0 + ∂1)
(iλ+) , (A.2.34c)
W− = − 1√2D−Σ
= iλ− − θ+(D + iF01) + i√
2θ−(∂0 − ∂1)σ − 2θ+θ−(∂0 − ∂1)λ+
+iθ−θ−(∂0 − ∂1) + iθ+θ+(∂0 + ∂1)
(iλ−)− iθ−θ−θ+(∂0 − ∂1)(D + iF01)
+ iθ+θ+(∂0 + ∂1)+i√
2θ−(∂0 − ∂1)σ+iθ−θ−(∂0 − ∂1)
iθ+θ+(∂0 + ∂1)
(iλ−) . (A.2.34d)
A.2 N = (2, 2) supersymmetry in two dimensions 155
Semi-chiral superfield
A semi-chiral superfield X appears in [75] when they consider nonlinear sigma models on generalized complex
geometries. These superfields were proposed by Ref.[10] and have been discussed in many papers, for example,
Ref.[58, 97, 42, 7] and references therein. The definitions of a left chiral superfield X and a left anti-chiral superfield
X are
D+X = 0 , D+X = 0 . (A.2.35)
Their expansions are as follows:
X = φ+ i√
2θ+ψ+ + i√
2(θ−ψ− + θ−χ−) + 2iθ+(θ−F + θ−G) + θ−θ−(A0 −A1) + 2θ+θ−θ−ζ−
− iθ+θ+(∂0 + ∂1)φ+√
2θ+θ+(∂0 + ∂1)(θ−ψ− + θ−χ−) + iθ+θ−θ+θ−(∂0 + ∂1)(A0 −A1) , (A.2.36a)
X = φ+ i√
2θ+ψ+ + i√
2(θ−ψ− + θ−χ−) + 2iθ+(θ−F + θ−G) + θ−θ−(A0 −A1)− 2θ+θ−θ−ζ−
+ iθ+θ+(∂0 + ∂1)φ−√
2θ+θ+(∂0 + ∂1)(θ−ψ− + θ−χ−)− iθ+θ−θ+θ−(∂0 + ∂1)(A0 −A1) , (A.2.36b)
where F , G and ζ− are auxiliary fields without any derivative terms; φ, F and G are complex fields, Am are real
vector fields; ψ±, χ− and ζ− are Weyl fermions. For later convenience, we prepare the following equations, which
we will use when we consider (1, 1) supersymmetry as in [75]:
Q+X = i√
2ψ+ + 2iθ−F + 2iθ−G+ 2θ−θ−ζ−
+√
2θ+θ+(∂0 + ∂1)ψ+ + 2θ+θ+θ−(∂0 + ∂1)F + 2θ+θ+θ−(∂0 + ∂1)G
+ 2iθ+θ−θ+θ−(∂0 + ∂1)ζ− , (A.2.37a)
Q−X = i√
2ψ− − 2iθ+F + θ−(A0 −A1) + i(∂0 − ∂1)φ
−√
2θ+θ−√
2ζ− − (∂0 − ∂1)ψ+
+√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ− − 2θ+θ−θ−(∂0 − ∂1)F
− iθ−θ+θ+(∂0 + ∂1)(A0 −A1) + i(∂0 − ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ− , (A.2.37b)
Q+X = −2iθ+(∂0 + ∂1)φ+ 2√
2θ+θ−(∂0 + ∂1)ψ− + 2√
2θ+θ−(∂0 + ∂1)χ−
− 2iθ+θ−θ−(∂0 + ∂1)(A0 −A1) , (A.2.37c)
Q−X = −i√
2χ− + 2iθ+G+ θ−(A0 −A1)− i(∂0 − ∂1)φ
−√
2θ+θ−√
2ζ− + (∂0 − ∂1)ψ+
−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ− − 2θ+θ−θ−(∂0 − ∂1)G
− iθ−θ+θ+(∂0 + ∂1)(A0 −A1)− i(∂0 − ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ− , (A.2.37d)
D+X = i√
2ψ+ + 2iθ−F + 2iθ−G− 2iθ+(∂0 + ∂1)φ
−√
2θ+θ+(∂0 + ∂1)ψ+ + 2θ−θ−ζ− + 2√
2θ+θ−(∂0 + ∂1)ψ− + 2√
2θ+θ−(∂0 + ∂1)χ−
− 2iθ+θ−θ−(∂0 + ∂1)(A0 −A1)− 2θ+θ+θ−(∂0 + ∂1)F − 2θ+θ+θ−(∂0 + ∂1)G
− 2iθ+θ−θ+θ−(∂0 + ∂1)ζ− , (A.2.37e)
D−X = i√
2ψ− − 2iθ+F + θ−(A0 −A1)− i(∂0 − ∂1)φ
−√
2θ+θ−√
2ζ− + (∂0 − ∂1)ψ+
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ− + 2θ+θ−θ−(∂0 − ∂1)F
− iθ−θ+θ+(∂0 + ∂1)(A0 −A1)− i(∂0 − ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ− , (A.2.37f)
156 Conventions
D+X = 0 , (A.2.37g)
D−X = −i√
2χ− + 2iθ+G+ θ−(A0 −A1) + i(∂0 − ∂1)φ
−√
2θ+θ−√
2ζ− − (∂0 − ∂1)ψ+
−√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
χ− + 2θ+θ−θ−(∂0 − ∂1)G
− iθ−θ+θ+(∂0 + ∂1)(A0 −A1) + i(∂0 − ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ− , (A.2.37h)
Q+X = 2iθ+(∂0 + ∂1)φ− 2√
2θ+θ−(∂0 + ∂1)ψ− − 2√
2θ+θ−(∂0 + ∂1)χ−
+ 2iθ+θ−θ−(∂0 + ∂1)(A0 −A1)
= Q+X , (A.2.38a)
Q−X = i√
2χ− − 2iθ+G+ θ−(A0 −A1) + i(∂0 − ∂1)φ
+√
2θ+θ−√
2ζ− + (∂0 − ∂1)ψ+
−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ− − 2θ+θ−θ−(∂0 − ∂1)G
+ iθ−θ+θ+(∂0 + ∂1)(A0 −A1) + i(∂0 − ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ−
= Q−X , (A.2.38b)
Q+X = −i√
2ψ+ − 2iθ−F − 2iθ−G+ 2θ−θ−ζ− +√
2θ+θ+(∂0 + ∂1)ψ+
+ 2θ+θ+θ−(∂0 + ∂1)F + 2θ+θ+θ−(∂0 + ∂1)G− 2iθ+θ−θ+θ−(∂0 + ∂1)ζ−
= Q+X , (A.2.38c)
Q−X = −i√
2ψ− + 2iθ+F + θ−(A0 −A1)− i(∂0 − ∂1)φ
+√
2θ+θ−√
2ζ− − (∂0 − ∂1)ψ+
+√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ− − 2θ+θ−θ−(∂0 − ∂1)F
+ iθ−θ+θ+(∂0 + ∂1)(A0 −A1)− i(∂0 − ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ−
= Q−X , (A.2.38d)
D+X = 0 , (A.2.38e)
D−X = i√
2χ− − 2iθ+G+ θ−(A0 −A1)− i(∂0 − ∂1)φ
+√
2θ+θ−√
2ζ− + (∂0 − ∂1)ψ+
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ− + 2θ+θ−θ−(∂0 − ∂1)F
+ iθ−θ+θ+(∂0 + ∂1)(A0 −A1)− i(∂0 − ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ−
= D−X , (A.2.38f)
D+X = −i√
2ψ+ − 2iθ−F − 2iθ−G+ 2iθ+(∂0 + ∂1)φ
−√
2θ+θ+(∂0 + ∂1)ψ+ + 2θ−θ−ζ− − 2√
2θ+θ−(∂0 + ∂1)ψ− − 2√
2θ+θ−(∂0 + ∂1)χ−
+ 2iθ+θ−θ−(∂0 + ∂1)(A0 −A1)− 2θ+θ+θ−(∂0 + ∂1)F − 2θ+θ+θ−(∂0 + ∂1)G
+ 2iθ+θ−θ+θ−(∂0 + ∂1)ζ−
= D+X , (A.2.38g)
D−X = −i√
2ψ− + 2iθ+F + θ−(A0 −A1) + i(∂0 − ∂1)φ
+√
2θ+θ−√
2ζ− + (∂0 − ∂1)ψ+
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ− + 2θ+θ−θ−(∂0 − ∂1)F
+ iθ−θ+θ+(∂0 + ∂1)(A0 −A1) + i(∂0 − ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ−
= D−X . (A.2.38h)
A.2 N = (2, 2) supersymmetry in two dimensions 157
The definitions of a right chiral superfield Y and a right anti-chiral superfield Y are
D−Y = 0 , D−Y = 0 . (A.2.39)
Notice that Lindstrom [75] represents this right “anti-”chiral superfield as Y. Unfortunately, however, this sym-
bolization is quite confusing and we often misunderstand this field as a right “chiral” superfield. Thus in this note
we represent this superfield as Y, while we do not express component fields with overline such as φ. We must also
take care when we read [75].
The expansions of these superfields are as follows:
Y = φ+ i√
2(θ+ψ+ + θ+χ+) + i√
2θ−ψ− + 2iθ+θ−F + 2iθ+θ−N − iθ−θ−(∂0 − ∂1)φ+ θ+θ+(B0 +B1)
− 2θ−θ+θ+ζ+ +√
2θ−θ−(∂0 − ∂1)(θ+ψ+ + θ+χ+) + iθ+θ−θ+θ−(∂0 − ∂1)(B0 +B1) , (A.2.40)
Y = φ+ i√
2(θ+ψ+ + θ+χ+) + i√
2θ−ψ− + 2iθ+θ−F + 2iθ+θ−N + iθ−θ−(∂0 − ∂1)φ+ θ+θ+(B0 +B1)
− 2θ−θ+θ+ζ+ +√
2θ−θ−(∂0 − ∂1)(θ+ψ+ + θ+χ+)− iθ+θ−θ+θ−(∂0 − ∂1)(B0 +B1) , (A.2.41)
where M , G and ζ+ are auxiliary fields without any derivative terms; φ, M and G are complex fields, Bm are real
vector fields; ψ±, χ+ and ζ+ are Weyl fermions. For later convenience, we prepare the following equations, which
we will use when we consider (1, 1) supersymmetry as in [75]:
Q+Y = i√
2ψ+ + 2iθ−F + θ+(B0 +B1) + i(∂0 + ∂1)φ
−√
2θ+θ−√
2ζ+ + (∂0 + ∂1)ψ−
+√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ+ + 2θ+θ+θ−(∂0 + ∂1)F
− iθ+θ−θ−(∂0 − ∂1)(B0 +B1) + i(∂0 + ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+ , (A.2.42a)
Q−Y = i√
2ψ− − 2iθ+F − 2iθ+N − 2θ+θ+ζ+ +√
2θ−θ−(∂0 − ∂1)ψ−
− 2θ+θ−θ−(∂0 − ∂1)F − 2θ+θ−θ−(∂0 − ∂1)N − 2iθ+θ−θ+θ−(∂0 − ∂1)ζ+ , (A.2.42b)
Q+Y = −i√
2χ+ − 2iθ−N + θ+(B0 +B1)− i(∂0 + ∂1)φ
−√
2θ+θ−√
2ζ+ − (∂0 + ∂1)ψ−
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ+ + 2θ+θ+θ−(∂0 + ∂1)N
− iθ+θ−θ−(∂0 − ∂1)(B0 +B1)− i(∂0 + ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+ , (A.2.42c)
Q−Y = −2iθ−(∂0 − ∂1)φ− 2√
2θ+θ−(∂0 − ∂1)ψ+ − 2√
2θ+θ−(∂0 − ∂1)χ+
− 2iθ−θ+θ+(∂0 − ∂1)(B0 +B1) , (A.2.42d)
D+Y = i√
2ψ+ + 2iθ−F + θ+(B0 +B1)− i(∂0 + ∂1)φ
−√
2θ+θ−√
2ζ+ − (∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ+ − 2θ+θ+θ−(∂0 + ∂1)F
− iθ+θ−θ−(∂0 − ∂1)(B0 +B1)− i(∂0 + ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+ , (A.2.42e)
D−Y = i√
2ψ− − 2iθ+F − 2iθ+N − 2iθ−(∂0 − ∂1)φ−√
2θ−θ−(∂0 − ∂1)ψ−
− 2θ+θ+ζ+ − 2√
2θ+θ−(∂0 − ∂1)ψ+ − 2√
2θ+θ−(∂0 − ∂1)χ+
− 2iθ−θ+θ+(∂0 − ∂1)(B0 +B1) + 2θ−θ−θ+(∂0 − ∂1)F + 2θ−θ−θ+(∂0 − ∂1)N
+ 2iθ+θ−θ+θ−(∂0 − ∂1)ζ+ , (A.2.42f)
D+Y = −i√
2χ+ − 2iθ−N + θ+(B0 +B1) + i(∂0 + ∂1)φ
−√
2θ+θ−√
2ζ+ + (∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
χ+ − 2θ+θ+θ−(∂0 + ∂1)N
158 Conventions
− iθ+θ−θ−(∂0 − ∂1)(B0 +B1) + i(∂0 + ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+ , (A.2.42g)
D−Y = 0 , (A.2.42h)
Q+Y = i√
2χ+ + 2iθ−N + θ+(B0 +B1) + i(∂0 + ∂1)φ
+√
2θ+θ−√
2ζ+ − (∂0 + ∂1)ψ−
+√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
χ+ + 2θ+θ+θ−(∂0 + ∂1)N
+ iθ+θ−θ−(∂0 − ∂1)(B0 +B1) + i(∂0 + ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+
= Q+Y , (A.2.43a)
Q−Y = 2iθ−(∂0 − ∂1)φ+ 2√
2θ+θ−(∂0 − ∂1)ψ+ + 2√
2θ+θ−(∂0 − ∂1)χ+ + 2iθ−θ+θ+(∂0 − ∂1)(B0 +B1)
= Q−Y , (A.2.43b)
Q+Y = −i√
2ψ+ − 2iθ−F + θ+(B0 +B1)− i(∂0 + ∂1)φ
+√
2θ+θ−√
2ζ+ + (∂0 + ∂1)ψ−
+√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
ψ+ + 2θ+θ+θ−(∂0 + ∂1)F
+ iθ+θ−θ−(∂0 − ∂1)(B0 +B1)− i(∂0 + ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+
= Q+Y , (A.2.43c)
Q−Y = −i√
2ψ− + 2iθ+F + 2iθ+N − 2θ+θ+ζ+ +√
2θ−θ−(∂0 − ∂1)ψ−
− 2θ+θ−θ−(∂0 − ∂1)F − 2θ+θ−θ−(∂0 − ∂1)N + 2iθ+θ−θ+θ−(∂0 − ∂1)ζ+
= Q−Y , (A.2.43d)
D+Y = i√
2χ+ + 2iθ−N + θ+(B0 +B1)− i(∂0 + ∂1)φ
+√
2θ+θ−√
2ζ+ + (∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1) + θ−θ−(∂0 − ∂1)
χ+ − 2θ+θ+θ−(∂0 + ∂1)N
+ iθ+θ−θ−(∂0 − ∂1)(B0 +B1)− i(∂0 + ∂1)φ
− i√
2θ+θ−θ+θ−(∂20 − ∂2
1)χ+
= D+Y , (A.2.43e)
D−Y = 0 , (A.2.43f)
D+Y = −i√
2ψ+ − 2iθ−F + θ+(B0 +B1) + i(∂0 + ∂1)φ
+√
2θ+θ−√
2ζ+ − (∂0 + ∂1)ψ−
−√
2θ+θ+(∂0 + ∂1)− θ−θ−(∂0 − ∂1)
ψ+ − 2θ+θ+θ−(∂0 + ∂1)F
+ iθ+θ−θ−(∂0 − ∂1)(B0 +B1) + i(∂0 + ∂1)φ
+ i√
2θ+θ−θ+θ−(∂20 − ∂2
1)ψ+
= D+Y , (A.2.43g)
D−Y = −i√
2ψ− + 2iθ+F + 2iθ+N + 2iθ−(∂0 − ∂1)φ−√
2θ−θ−(∂0 − ∂1)ψ−
− 2θ+θ+ζ+ + 2√
2θ+θ−(∂0 − ∂1)ψ+ + 2√
2θ+θ−(∂0 − ∂1)χ+
+ 2iθ−θ+θ+(∂0 − ∂1)(B0 +B1) + 2θ−θ−θ+(∂0 − ∂1)F + 2θ−θ−θ+(∂0 − ∂1)N
− 2iθ+θ−θ+θ−(∂0 − ∂1)ζ+
= D−Y . (A.2.43h)
A.2 N = (2, 2) supersymmetry in two dimensions 159
Linear superfields
Here we introduce other useful superfields, so-called linear superfields [27, 42], even though we do not try to expand
them in terms of component fields. Note that, as we shall explain, there are four linear superfields.
A real linear superfield G and a real twisted linear superfield H are defined as
D+D−G = D+D−G = 0 , D+D−H = D+D−H = 0 . (A.2.44)
Because of their definitions, we can describe them in terms of the combinations of chiral and twisted chiral superfields
as follows:
G = Y + Y , H = Φ + Φ . (A.2.45)
In the same way, we also define a complex linear superfield Σ and a complex twisted linear superfields Σ:
D+D−Σ = 0 , D+D−Σ = 0 . (A.2.46)
Notice that these are not twisted chiral superfields of a vector multiplet. These are also expressed in terms of
semi-chiral superfields X and Y:
Σ = X + Y , Σ = X + Y . (A.2.47)
A.2.5 Gauging of phase shifts
Here we introduce a U(1) gauge symmetry which is one of the most important symmetries of two-dimensional field
theories [113].
First we consider local phase shift of a chiral superfield Φ:
Φ → Φ′ = e−2iQΛΦ , (A.2.48)
where Q is a real constant which means a charge of this U(1) phase shift. Imposing Φ′ on a new chiral superfield,
we find that the gauge parameter Λ should be a chiral superfield
D±Λ = 0 . (A.2.49)
If this parameter is just a bosonic one, we immediately find that the condition is not satisfied such as D±Φ′ =
2Q(∂0 ± ∂1)Λ · Φ′ 6= 0 and Φ′ is no longer a chiral superfield. In the same reason, we cannot introduce the same
gauge transformation to a twisted chiral superfield Y : D−(e−2iQΛY ) 6= 0. Thus any twisted chiral superfields are
neutral under the above gauge transformation.
A vector superfield V is transformed as follows:
V → V ′ = V + i(Λ− Λ) , (A.2.50)
where Λ is an anti-chiral superfield satisfying D±Λ = 0. According to this, we find that Σ and W± are invariant
under this gauge transformation.
160 Conventions
We also find that a semi-chiral superfield X can possess this phase shift because the transformed semi-chiral
superfield satisfies the semi-chiral condition:
X → X′ = e−2iQΛX , D+X′ = −2iQD+Λ · e−2iQΛX + e−2iQΛD+X = 0 . (A.2.51)
Here we note that if we extend the above gauge transformation parameter Λ to a semi-chiral superfield, any
chiral superfield cannot transform under this extended parameter because D−Λ = 0 is not satisfied. Furthermore
it is difficult to understand how a vector superfield transforms under the extended gauge transformation.
A.3 N = (1, 1) supersymmetry
A.3.1 A definition of N = (1, 1) supersymmetry
Differential operators
Here we consider N = (1, 1) supersymmetry which has two real supercharges, one with positive chirality and the
other with negative chirality6. The relevant sub-superspace is the one where θ+ and θ− are real up to phases.
Namely, the subspace such that
θ±1 ≡ −ie−iν±θ± = ie+iν±θ± ,
or θ± = ieiν±θ±1 , θ± = −ie−iν±θ±1 ,where θ±1 is real , (A.3.1)
for arbitrary (but fixed) phases eiν± , where “θ±1 real” means (θ±1 )† = θ±1 . This also means the following equations
e−iν±θ± + eiν±θ± = 0 . (A.3.2)
θ±1 are the fermionic coordinates of this subspace, which we call (1, 1) superspace. Under this constraint we
introduce the following differential operators7
Q1± ≡
1√2
eiν±Q± + e−iν±Q±
∣∣∣eq.(A.3.2)
= − i√2
∂
∂θ±1+√
2θ±1(∂0 ± ∂1
), (A.3.3a)
D1± ≡
1√2
eiν±D± + e−iν±D±
∣∣∣eq.(A.3.2)
= − i√2
∂
∂θ±1−√
2θ±1(∂0 ± ∂1
). (A.3.3b)
Notice that after we impose the constraint (A.3.2) on the spinor coordinate derivatives ∂∂θ± and ∂
∂θ±, a half degrees
of freedom are frozen. These operators obey the anti-commutation relations
Q1±, Q
1± = −2i
(∂0 ± ∂1
)= 2(H ∓ P ) , Q1
+, Q1− = 0 , (A.3.4a)
D1±,D
1± = +2i
(∂0 ± ∂1
), D1
+,D1− = 0 , (A.3.4b)
Q1α,D
1β = 0 . (A.3.4c)
6Here we introduce the convention described in the Hori’s book [53].7The definitions of Q1
± and D1± in (A.3.3) are equivalent to the ones in [53] up to the scaling factors.
A.3 N = (1, 1) supersymmetry 161
We also define the following “differential operators”:
Q1± ≡
i√2
eiν±Q± − e−iν±Q±
, under the constraint (A.3.2): Q1
±
∣∣∣ = 0 , (A.3.5a)
D1± ≡
i√2
eiν±D± − e−iν±D±
, under the constraint (A.3.2): D1
±
∣∣∣ = 0 . (A.3.5b)
Under the constraint (A.3.2) these operators are trivially zero. However, they are another two supercharges
and two covariant derivatives in the original N = (2, 2) supersymmetry. We can easily find this “second” (1, 1)
supersymmetry operators satisfy the following anti-commutation relations in the (2, 2) supersymmetry level:
Q1±, Q
1± = −2i
(∂0 ± ∂1
)= 2(H ∓ P ) , Q1
+, Q1− = 0 , (A.3.6a)
D1±, D
1± = +2i
(∂0 ± ∂1
), D1
+, D1− = 0 , (A.3.6b)
Q1α, D
1β = 0 . (A.3.6c)
Furthermore we can check that the first (1, 1) supersymmetry and the second (1, 1) supersymmetry commute with
each other:
Q1α, Q
1β = D1
α, D1β = Q1
α, D1β = D1
α, Q1β = 0 . (A.3.7)
This result is consistent with the original N = (2, 2) supersymmetry because Q1± and Q1
± are, roughly speaking,
the real and imaginary part of the complex supercharges Q±. This phenomena is required when we consider the
nonlinear sigma models on generalized complex geometries such that the manifest (first) and non-manifest (second)
supersymmetries must commute with each other [75].
Let us consider more detail. By using Q1± and Q1
±, we can rewrite the N = (2, 2) supersymmetry transformation
rule8:
δ(2,2)F = (iε±Q± + ε±Q±)F= (iε±1 Q
1± + iε±1 Q
1±)F ≡ δ(1,1)F + δ(1,1)F , (A.3.8)
where F is an arbitrary N = (2, 2) superfield. In [10, 75], the authors regard the first term in (A.3.8) as a
standard and manifest N = (1, 1) supersymmetry transformation, while the second term is regarded as an another
(sometimes non-manifest)N = (1, 1) supersymmetry transformation. This second term is often modified to an extra
transformation in order to incorporate some extended information of, for instance, generalized complex structures
[74, 75].
A.3.2 (1, 1) superfields reduced from (2, 2) superfields
We can easily obtain N = (1, 1) superfields from (2, 2) superfields via the reductions (A.3.1). We obtain a superfield
from a (2, 2) chiral superfield Φ(2,2) as follows9:
Φ(1,1) = Φ(2,2)∣∣∣eq.(A.3.1)
= φ+ i√
2θ+1 (ieiν+ψ+) + i√
2θ−1 (ieiν−ψ−) + 2iθ+1 θ−1 (−eiν++iν−F )
8where we set a = b = 1.9We sometimes describe an N = (p, q) superfield as Φ(p,q).
162 Conventions
≡ φ+ i√
2θ+1 ψ+ + i√
2θ−1 ψ− + 2iθ+1 θ−1 F , (A.3.9a)
Φ(1,1) = Φ(2,2)∣∣∣eq.(A.3.1)
= φ+ i√
2θ+1 ψ+ + i√
2θ−1 ψ− + 2iθ+1 θ−1 F = Φ(1,1) , (A.3.9b)
where we performed phase shifts on ψ± = ieiν±ψ± and F = −eiν++iν−F .
We also obtain the same superfield from a (2, 2) twisted chiral superfield:
Y (1,1) = Y (2,2)∣∣∣eq.(A.3.1)
= y + i√
2θ+1 (ieiν+χ+) + i√
2θ−1 (−ie−iν−χ−) + 2iθ+1 θ−1 (eiν+−iν−G)
≡ y + i√
2θ+1 χ+ + i√
2θ−1 χ− + 2iθ+1 θ−1 G , (A.3.10a)
Y (1,1) = Y (2,2)∣∣∣eq.(A.3.1)
= y + i√
2θ+1 χ+ + i√
2θ−1 χ− + 2iθ+1 θ−1 G = Y (1,1) . (A.3.10b)
We can identify y; χ+; χ−; G in (A.3.10) with φ; ψ+; ψ−; F in (A.3.9), respectively. Thus we find there are no
differences between a chiral superfield and a twisted chiral superfield in (1, 1) supersymmetric theories.
Gauge multiplets are described by (2, 2) superfields V (2,2) and Σ(2,2):
V (1,1) = V (2,2)∣∣∣eq.(A.3.1)
= −√
2θ−1 θ+1 (e−iν++iν−σ)−
√2θ+1 θ
−1 (eiν+−iν−σ)
≡√
2θ+1 θ−1
(σ − σ
), (A.3.11a)
Σ(1,1) = e−iν++iν−Σ(2,2)∣∣∣eq.(A.3.1)
= (e−iν++iν−σ)− i√
2θ+1 (ieiν−λ+)− i√
2θ−1 (−ie−iν+λ−) +√
2θ+1 θ−1 (D − iF01)
≡ σ − i√
2θ+1 λ+ − i√
2θ−1 λ− +√
2θ+1 θ−1 (D − iF01) . (A.3.11b)
V (1,1) is also a real superfield because of (θ−1 θ+1 )† = θ+1 θ
−1 and of (σ)† = σ. Since D and F01 are real fields we
should incorporate an additional phase e−iν++iν− in front of Σ(2,2).
Spinor superfields are also reduced from (2, 2) theories:
W(1,1)+ = −ie−iν−W
(2,2)+
∣∣∣eq.(A.3.1)
= −iλ+ − θ−1 (D + iF01)− i√
2θ+1 (∂0 + ∂1)σ + 2θ−1 θ+1 (∂0 + ∂1)λ− , (A.3.12a)
W(1,1)− = ie−iν+W
(2,2)−
∣∣∣eq.(A.3.1)
= −iλ− − θ+1 (D − iF01)− i√
2θ−1 (∂0 − ∂1)σ + 2θ+1 θ−1 (∂0 − ∂1)λ+ . (A.3.12b)
We finally reduce semi-chiral superfields10:
X(1,1) = X(2,2)∣∣∣eq.(A.3.1)
= φ+ i√
2θ+1 ψ+ + i√
2θ−1 (ψ− + χ−) + 2iθ+1 θ−1 (F + G) , (A.3.13a)
X(1,1) = X(2,2)∣∣∣eq.(A.3.1)
= φ+ i√
2θ+1 ψ+ + i√
2θ−1 (ψ− + χ−) + 2iθ+1 θ−1 (F + G) = X(1,1) , (A.3.13b)
Y(1,1) = Y(2,2)∣∣∣eq.(A.3.1)
= φ+ i√
2θ+1 (ψ+ + χ+) + i√
2θ−1 ψ− + 2iθ+1 θ−1 (F + N) , (A.3.13c)
10The symbols of (1, 1) superfields from (2, 2) semi-chiral superfields are different from the ones in [75].
A.3 N = (1, 1) supersymmetry 163
Y(1,1) = Y(2,2)∣∣∣eq.(A.3.1)
= φ+ i√
2θ+1 (ψ+ + χ+) + i√
2θ−1 ψ− + 2iθ+1 θ−1 (F + N) = Y(1,1) . (A.3.13d)
A.3.3 Supersymmetry transformations
It is quite important to investigate how (1, 1) superfields transform under the first supersymmetry Q1± and under
the second supersymmetry Q1±. If we introduce a non-manifest, extended second supersymmetry, the target space
of the theory becomes a more general geometry such as a generalized complex geometry [74, 75].
As discussed before, the (1, 1) reduced supersymmetry transformations of (2, 2) supersymmetry transformations
are described by
δ(2,2)F∣∣∣eq.(A.3.1)
=δ(1,1)F + δ(1,1)F
∣∣∣eq.(A.3.1)
, (A.3.14)
where F in the left hand side is an arbitrary superfield in (2, 2) supersymmetric theory. If we impose the constraint
(A.3.1) and if we do not assume an extended supersymmetry to the second supersymmetry transformations, the
above equation are simply reduced to
δ(2,2)F∣∣∣eq.(A.3.1)
= δ(1,1)F (1,1) . (A.3.15)
If we wish to incorporate a (non-manifest) extended supersymmetry as discussed in [74, 75], we must consider
the reduced representations of the second supersymmetry δ(1,1)F (1,1), i.e., we must understand the transformation
rules of all the (1, 1) reduced superfields such as
Q1±F (1,1) , Q1
αQ1βF (1,1) , Q1
±F (1,1) , Q1αQ
1βF (1,1) , Q1
αQ1βF (1,1) . (A.3.16)
In order that the supersymmetry algebra is an off-shell closure, we should consider the operations of two su-
percharges. Thus in this subsection we discuss how the above operations are realized within the N = (1, 1)
supersymmetry level (under the constraint (A.3.1)).
Before discussing a supersymmetric theory itself, we investigate the two kinds of supersymmetry transformations
by using various differential operators Q1±, Q1
±, D1± and D1
±. Here we only consider them on a chiral superfield
Φ(2,2) and a semi-chiral superfield X(2,2), their reduced superfields Φ(1,1), X(1,1), and new (1, 1) superfields derived
from (2, 2) superfields.
Chiral superfields
First we write down (1, 1) reduced representations of (2, 2) supersymmetry operations. Second we consider the
supersymmetry operations in the (1, 1) supersymmetry level and investigate the relations among them.
Q1+Φ(2,2)
∣∣∣eq.(A.3.1)
= ψ+ +√
2θ−1 F +√
2θ+1 (∂0 + ∂1)φ+ 2iθ+1 θ−1 (∂0 + ∂1)ψ− ≡ Γ
(1,1)+ , (A.3.17a)
Q1−Φ(2,2)
∣∣∣eq.(A.3.1)
= ψ− −√
2θ+1 F +√
2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ−1 (∂0 − ∂1)ψ+ ≡ Γ
(1,1)− , (A.3.17b)
Q1+Φ(2,2)
∣∣∣eq.(A.3.1)
= iψ+ + i√
2θ−1 F − i√
2θ+1 (∂0 + ∂1)φ+ 2θ+1 θ−1 (∂0 + ∂1)ψ− ≡ Γ
(1,1)+ , (A.3.17c)
164 Conventions
Q1−Φ(2,2)
∣∣∣eq.(A.3.1)
= iψ− − i√
2θ+1 F − i√
2θ−1 (∂0 − ∂1)φ− 2θ+1 θ−1 (∂0 − ∂1)ψ+ ≡ Γ
(1,1)− , (A.3.17d)
D1+Φ(2,2)
∣∣∣eq.(A.3.1)
= ψ+ +√
2θ−1 F −√
2θ+1 (∂0 + ∂1)φ− 2iθ+1 θ−1 (∂0 + ∂1)ψ− = −iΓ(1,1)
+ , (A.3.17e)
D1−Φ(2,2)
∣∣∣eq.(A.3.1)
= ψ− −√
2θ+1 F −√
2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ−1 (∂0 − ∂1)ψ+ = −iΓ(1,1)
− , (A.3.17f)
D1+Φ(2,2)
∣∣∣eq.(A.3.1)
= iψ+ + i√
2θ−1 F − i√
2θ+1 (∂0 + ∂1)φ+ 2θ+1 θ−1 (∂0 + ∂1)ψ− = Γ
(1,1)+ , (A.3.17g)
D1−Φ(2,2)
∣∣∣eq.(A.3.1)
= iψ− − i√
2θ+1 F − i√
2θ−1 (∂0 − ∂1)φ− 2θ+1 θ−1 (∂0 − ∂1)ψ+ = Γ
(1,1)− , (A.3.17h)
Note that, surprisingly, the reductions of Q1±Φ(2,2) correspond not only to the reductions of D1
±Φ(2,2) but also to the
reductions of iD1±Φ(2,2), respectively. This means that we can discuss the second supersymmetry transformations
within the first supersymmetry language (differential operators iD1± and superfields Γ±(1,1), Γ
(1,1)± ). By using
this result, Gates et al [27] and Lindstrom et al [75] introduced extended N = (1, 1) supersymmetry charges in
standard N = (1, 1) supersymmetric theories. Here we mention that the above results are obtained within the
(1, 1) supersymmetry level:
Q1+ · Φ(1,1) = ψ+ +
√2θ−1 F +
√2θ+1 (∂0 + ∂1)φ+ 2iθ+1 θ
−1 (∂0 + ∂1)ψ− = Γ
(1,1)+ , (A.3.18a)
Q1− · Φ(1,1) = ψ− −
√2θ+1 F +
√2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ
−1 (∂0 − ∂1)ψ+ = Γ
(1,1)− , (A.3.18b)
Q1+ · Φ(1,1) = trivially zero , (A.3.18c)
Q1− · Φ(1,1) = trivially zero , (A.3.18d)
D1+ · Φ(1,1) = ψ+ +
√2θ−1 F −
√2θ+1 (∂0 + ∂1)φ− 2iθ+1 θ
−1 (∂0 + ∂1)ψ− = −iΓ(1,1)
+ , (A.3.18e)
D1− · Φ(1,1) = ψ− −
√2θ+1 F −
√2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ
−1 (∂0 − ∂1)ψ+ = −iΓ(1,1)
− , (A.3.18f)
D1+ · Φ(1,1) = trivially zero , (A.3.18g)
D1− · Φ(1,1) = trivially zero . (A.3.18h)
Even though Q± ·Φ(1,1) is trivially zero, we can obtain non-trivial relations of the operations of the second super-
charges. Now we write down (1, 1) supersymmetry operations of the above superfields Γ(1,1)± :
Q1+ · Γ(1,1)
+ = −i(∂0 + ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 ψ− + 2iθ+1 θ
−1 F
= −i(∂0 + ∂1)Φ(1,1) =
1
2Q1
+, Q1+Φ(2,2)
∣∣∣ , (A.3.19a)
Q1− · Γ(1,1)
+ = −iF −√
2θ+1 (∂0 + ∂1)ψ− +√
2θ−1 (∂0 − ∂1)ψ+ − 2θ+1 θ−1 (∂2
0 − ∂21)φ
= −Q1+ · Γ(1,1)
− = Q1−Q
1+Φ(2,2)
∣∣∣ , (A.3.19b)
D1+ · Γ(1,1)
+ = −i(∂0 + ∂1)φ− i
√2θ+1 ψ+ + i
√2θ−1 ψ− − 2iθ+1 θ
−1 F
= iQ1+ · Γ(1,1)
+ = −iQ1+Q
1+Φ(2,2)
∣∣∣
= −iQ1+ · Γ(1,1)
+ , (A.3.19c)
D1− · Γ(1,1)
+ = −iF −√
2θ+1 (∂0 + ∂1)ψ− −√
2θ−1 (∂0 − ∂1)ψ+ + 2θ+1 θ−1 (∂2
0 − ∂21)φ
= iQ1+ · Γ(1,1)
− = −iQ1−Q
1+Φ(2,2)
∣∣∣
= −iQ− · Γ(1,1)+ , (A.3.19d)
A.3 N = (1, 1) supersymmetry 165
Q1+ · Γ(1,1)
− = iF +√
2θ+1 (∂0 + ∂1)ψ+ −√
2θ−1 (∂0 − ∂1)ψ− + 2θ+1 θ−1 (∂2
0 − ∂21)φ
= −Q1− · Γ(1,1)
+ = Q1+Q
1−Φ(2,2)
∣∣∣ , (A.3.19e)
Q1− · Γ(1,1)
− = −i(∂0 − ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 ψ− + 2iθ+1 θ
−1 F
= −i(∂0 − ∂1)Φ(1,1) =
1
2Q1
−, Q1−Φ(2,2)
∣∣∣ , (A.3.19f)
D1+ · Γ(1,1)
− = iF −√
2θ+1 (∂0 + ∂1)ψ− −√
2θ−1 (∂0 − ∂1)ψ+ − 2θ+1 θ−1 (∂2
0 − ∂21)φ
= iQ1− · Γ(1,1)
+ = −iQ1+Q
1−Φ(2,2)
∣∣∣
= −iQ1+ · Γ(1,1)
− , (A.3.19g)
D1− · Γ(1,1)
− = −i(∂0 − ∂1)φ+ i
√2θ+1 ψ+ − i
√2θ−1 ψ− − 2iθ+1 θ
−1 F
= iQ1− · Γ(1,1)
− = −iQ1−Q
1−Φ(2,2)
∣∣∣
= −iQ1−Γ
(1,1)− . (A.3.19h)
We also write down (1, 1) supersymmetry operations of the above superfields Γ(1,1)± :
Q1+ · Γ(1,1)
+ = −(∂0 − ∂1)φ− i
√2θ+1 ψ+ + i
√2θ−1 ψ− − 2iθ+1 θ
−1 F
= −iD1+ · Γ(1,1)
+ = Q1+Q
1+Φ(2,2)
∣∣∣
= −Q1+ · Γ(1,1)
+ , (A.3.20a)
Q1− · Γ(1,1)
+ = F + i√
2θ+1 (∂0 + ∂1)ψ− + i√
2θ−1 (∂0 − ∂1)ψ+ + 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −iD1+ · Γ(1,1)
− = Q1−Q
1+Φ(2,2)
∣∣∣
= −Q1+ · Γ(1,1)
− , (A.3.20b)
D1+ · Γ(1,1)
+ = −(∂0 + ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 ψ− + 2iθ+1 θ
−1 F
= −(∂0 + ∂1)Φ(1,1) = − i
2Q1
+, Q1+Φ(2,2)
∣∣∣ , (A.3.20c)
D1− · Γ(1,1)
+ = F + i√
2θ+1 (∂0 + ∂1)ψ− − i√
2θ−1 (∂0 − ∂1)ψ+ − 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −D1+ · Γ(1,1)
− = −iQ1−Q
1+Φ(2,2)
∣∣∣
= −iQ1− · Γ(1,1)
+ , (A.3.20d)
Q1+ · Γ(1,1)
− = −F + i√
2θ+1 (∂0 + ∂1)ψ− + i√
2θ−1 (∂0 − ∂1)ψ+ − 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −iD1− · Γ(1,1)
+ = Q1+Q
1−Φ(2,2)
∣∣∣
= −Q1− · Γ(1,1)
+ , (A.3.20e)
Q1− · Γ(1,1)
− = −(∂0 − ∂1)φ+ i
√2θ+1 ψ+ − i
√2θ−1 ψ− − 2iθ+1 θ
−1 F
= −iD1− · Γ(1,1)
− = Q1−Q
1−Φ(2,2)
∣∣∣
= −Q1− · Γ(1,1)
− , (A.3.20f)
D1+ · Γ(1,1)
− = −F − i√
2θ+1 (∂0 + ∂1)ψ− + i√
2θ−1 (∂0 − ∂1)ψ+ + 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −D1− · Γ(1,1)
+ = −iQ1+Q
1−Φ(2,2)
∣∣∣
166 Conventions
= −iQ1+ · Γ(1,1)
− , (A.3.20g)
D1− · Γ(1,1)
− = −(∂0 − ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 ψ− + 2iθ+1 θ
−1 F
= −(∂0 − ∂1)Φ(1,1) = − i
2Q1
−, Q1−Φ(2,2)
∣∣∣ . (A.3.20h)
Notice that the above obtained spinor superfields Γ(1,1)± and Γ
(1,1)± are not independent of the chiral (scalar) superfield
Φ(1,1) because all the component fields of spinor superfields are represented by the component fields of the scalar
superfield.
Here we summarize the non-trivial supersymmetry operations of the chiral superfield in the N = (1, 1) super-
symmetry language:
Q1±Φ(1,1) = Γ
(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Φ
(1,1) , Q1±Γ
(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.21a)
Q1±Φ(1,1) = iD1
±Φ(1,1) = Γ(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Φ
(1,1) , (A.3.21b)
Q1±Γ
(1,1)∓ = iD1
±Γ(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.21c)
Q1αΓ
(1,1)β = iD1
αΓ(1,1)β = −Q1
βΓ(1,1)α . (A.3.21d)
The above relations are also valid but extended if we consider the supersymmetry transformations of a (1, 1)
reduced semi-chiral superfield X(2,2). The reason is, as we will discuss later, the the operation Q1−X(2,2)| is not
equivalent to iD1−X(2,2)| in the semi-chiral superfield case, even though both the differential operations Q1
±Φ(2,2)|are equivalent to iD1
±Φ(2,2)| under the constraint (A.3.1) in the chiral superfield case. There exists a new spinor
superfield Ψ(1,1)− in the (1, 1) reductions of semi-chiral superfield. This superfield is represented by new component
fields, which are not contained in the (1, 1) reduced superfield X(1,1).
There is a significant comment. In the above calculcations, we found non-trivial relations Q1±Φ(2,2)
∣∣eq.(A.3.1)
=
D1±Φ(2,2)
∣∣eq.(A.3.1)
. It seems very strange because the differential operators themselves are completely different from
each other and the square of them are (Q1±)2 = −(D1
±)2 = −i(∂0 ± ∂1). However if we carefully and explicitly
calculate the operations by using component fields, we find the following result
Q1±Q
1±Φ(2,2)
∣∣∣ = −i(∂0 ± ∂1)Φ(1,1) = −D1
±D1±Φ(2,2)
∣∣∣ . (A.3.22)
This is consistent with the supersymmetry algebra. Fortunately we do not have to consider the operations
D1±D
1±Φ(2,2)
∣∣ in the second supersymmetry transformations. The reason is that we use the differential operators
of second supercharges Q1± if we perform the supersymmetry transformation. Thus we can avoid some confusion
of some kind of naive guess.
Here we also pick up the supersymmetry operations on the anti-chiral superfield Φ:
Q1+Φ(2,2)
∣∣∣ = ψ+ +√
2θ+1 (∂0 + ∂1)φ+√
2θ−1 F + 2iθ+1 θ−1 (∂0 + ∂1)ψ− = Γ
(1,1)+ , (A.3.23a)
Q1−Φ(2,2)
∣∣∣ = ψ− −√
2θ+1 F +√
2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ−1 (∂0 − ∂1)ψ+ = Γ
(1,1)− , (A.3.23b)
Q1+Φ(2,2)
∣∣∣ = −iψ+ + i√
2θ+1 (∂0 + ∂1)φ− i√
2θ−1 F − 2θ+1 θ−1 (∂0 + ∂1)ψ− = Γ
(1,1)+ , (A.3.23c)
Q1−Φ(2,2)
∣∣∣ = −iψ− + i√
2θ+1 F + i√
2θ−1 (∂0 − ∂1)φ+ 2θ+1 θ−1 (∂0 − ∂1)ψ+ = Γ
(1,1)− , (A.3.23d)
A.3 N = (1, 1) supersymmetry 167
D1+Φ(2,2)
∣∣∣ = ψ+ −√
2θ+1 (∂0 + ∂1)φ+√
2θ−1 F − 2iθ+1 θ−1 (∂0 + ∂1)ψ− = iΓ
(1,1)+ , (A.3.23e)
D1−Φ(2,2)
∣∣∣ = ψ− −√
2θ+1 F −√
2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ−1 (∂0 − ∂1)ψ+ = iΓ
(1,1)− , (A.3.23f)
D1+Φ(2,2)
∣∣∣ = −iψ+ + i√
2θ+1 (∂0 + ∂1)φ− i√
2θ−1 F − 2θ+1 θ−1 (∂0 + ∂1)ψ− = Γ
(1,1)+ , (A.3.23g)
D1−Φ(2,2)
∣∣∣ = −iψ− + i√
2θ+1 F + i√
2θ−1 (∂0 − ∂1)φ+ 2θ+1 θ−1 (∂0 − ∂1)ψ+ = Γ
(1,1)− , (A.3.23h)
Q1+ · Φ(1,1) = ψ+ +
√2θ+1 (∂0 + ∂1)φ+
√2θ−1 F + 2iθ+1 θ
−1 (∂0 + ∂1)ψ− = Q1
+ · Φ(1,1) , (A.3.24a)
Q1− · Φ(1,1) = ψ− −
√2θ+1 F +
√2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ
−1 (∂0 − ∂1)ψ+ = Q1
− · Φ(1,1) , (A.3.24b)
Q1+ · Φ(1,1) = trivially zero , (A.3.24c)
Q1− · Φ(1,1) = trivially zero , (A.3.24d)
D1+ · Φ(1,1) = ψ+ −
√2θ+1 (∂0 + ∂1)φ+
√2θ−1 F − 2iθ+1 θ
−1 (∂0 + ∂1)ψ− = D1
+ · Φ(1,1) , (A.3.24e)
D1− · Φ(1,1) = ψ− −
√2θ+1 F −
√2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ
−1 (∂0 − ∂1)ψ+ = D1
− · Φ(1,1) , (A.3.24f)
D1+ · Φ(1,1) = trivially zero , (A.3.24g)
D1− · Φ(1,1) = trivially zero , (A.3.24h)
Thus we find the following supersymmetry transformation rules in the N = (1, 1) level:
Q1±Φ(1,1) = Γ
(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Φ
(1,1) , Q1±Γ
(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.25a)
Q1±Φ(1,1) = −iD1
±Φ(1,1) = Γ(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Φ
(1,1) , (A.3.25b)
Q1±Γ
(1,1)∓ = −iD1
±Γ(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.25c)
Q1αΓ
(1,1)β = −iD1
αΓ(1,1)β = −Q1
βΓ(1,1)α . (A.3.25d)
Note that we can easily derive the above relations by using hermitian conjugate procedure such as
Q1±Φ(1,1) = Q1
±Φ(1,1) = Γ(1,1)± , (A.3.26a)
Q1±Γ
(1,1)± = −Q1
±Γ(1,1)± = −(+i(∂0 + ∂1)Φ
(1,1)) = −i(∂0 + ∂1)Φ(1,1) . (A.3.26b)
Twisted chiral superfield
Here we discuss the reduced (1, 1) twisted chiral superfield and its (2, 2) supersymmetry transformations. First we
write down the (1, 1) reductions of (2, 2) supersymmetry operations:
Q1+Y
(2,2)∣∣∣eq.(A.3.1)
= χ+ +√
2θ+1 (∂0 + ∂1)y +√
2θ−1 G+ 2iθ+1 θ−1 (∂0 + ∂1)χ− ≡ Γ
(1,1)+ , (A.3.27a)
Q1−Y
(2,2)∣∣∣eq.(A.3.1)
= χ− −√
2θ+1 G+√
2θ−1 (∂0 − ∂1)y − 2iθ+1 θ−1 (∂0 − ∂1)χ+ ≡ Γ
(1,1)− , (A.3.27b)
Q1+Y
(2,2)∣∣∣eq.(A.3.1)
= iχ+ − i√
2θ+1 (∂0 + ∂1)y + i√
2θ−1 G+ 2θ+1 θ−1 (∂0 + ∂1)χ− ≡ Γ
(1,1)+ , (A.3.27c)
Q1−Y
(2,2)∣∣∣eq.(A.3.1)
= −iχ− + i√
2θ+1 G+ i√
2θ−1 (∂0 − ∂1)y + 2θ+1 θ−1 (∂0 − ∂1)χ+ ≡ Γ
(1,1)− , (A.3.27d)
168 Conventions
D1+Y
(2,2)∣∣∣eq.(A.3.1)
= χ+ −√
2θ+1 (∂0 + ∂1)y +√
2θ−1 G− 2iθ+1 θ−1 (∂0 + ∂1)χ− = −iΓ(1,1)
+ , (A.3.27e)
D1−Y
(2,2)∣∣∣eq.(A.3.1)
= χ− −√
2θ+1 G−√
2θ−1 (∂0 − ∂1)y + 2iθ+1 θ−1 (∂0 − ∂1)χ+ = +iΓ
(1,1)− , (A.3.27f)
D1+Y
(2,2)∣∣∣eq.(A.3.1)
= iχ+ − i√
2θ+1 (∂0 + ∂1)y + i√
2θ−1 G+ 2θ+1 θ−1 (∂0 + ∂1)χ− = Γ
(1,1)+ , (A.3.27g)
D1−Y
(2,2)∣∣∣eq.(A.3.1)
= −iχ− + i√
2θ+1 G+ i√
2θ−1 (∂0 − ∂1)y + 2θ+1 θ−1 (∂0 − ∂1)χ+ = Γ
(1,1)− . (A.3.27h)
Next we calculate the (1, 1) supersymmetry operations purely in the (1, 1) supersymmetry level:
Q1+ · Y (1,1) = χ+ +
√2θ+1 (∂0 + ∂1)y +
√2θ−1 G+ 2iθ+1 θ
−1 (∂0 + ∂1)χ− = Γ
(1,1)+ , (A.3.28a)
Q1− · Y (1,1) = χ− −
√2θ+1 G+
√2θ−1 (∂0 − ∂1)y − 2iθ+1 θ
−1 (∂0 − ∂1)χ+ = Γ
(1,1)− , (A.3.28b)
Q1+ · Y (1,1) = trivially zero , (A.3.28c)
Q1− · Y (1,1) = trivially zero , (A.3.28d)
D1+ · Y (1,1) = χ+ −
√2θ+1 (∂0 + ∂1)y +
√2θ−1 G− 2iθ+1 θ
−1 (∂0 + ∂1)χ− = −iΓ(1,1)
+ , (A.3.28e)
D1− · Y (1,1) = χ− −
√2θ+1 G−
√2θ−1 (∂0 − ∂1)y + 2iθ+1 θ
−1 (∂0 − ∂1)χ+ = +iΓ
(1,1)− , (A.3.28f)
D1+ · Y (1,1) = trivially zero , (A.3.28g)
D1− · Y (1,1) = trivially zero . (A.3.28h)
From the above results, we simply obtain the following (1, 1) reduced supersymmetry operation rules:
Q1±Y
(1,1) = Γ(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Y
(1,1) , Q1±Γ
(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.29a)
Q1±Y
(1,1) = ±iD1±Y
(1,1) = Γ(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Y
(1,1) , (A.3.29b)
Q1±Γ
(1,1)∓ = ±iD1
±Γ(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.29c)
Q1±Γ
(1,1)β = ±iD1
±Γ(1,1)β = −Q1
βΓ(1,1)± . (A.3.29d)
Note that there are no newly obtained, independent spinor superfields after performing the second supersymmtry
transformations on the twisted chiral superfield. All the component fields of all the spinor superfields Γ(1,1)± ; Γ
(1,1)±
are described by the component fields of the (1, 1) scalar superfield Y (1,1). This phenomenon is also occured in the
chiral superfield case, while we will obtain new spinor superfields from the reductions of a vector superfield and a
semi-chiral superfield, which we call Ψ(1,1)α .
Here we also pick up the supersymmetry operations on the twisted anti-chiral superfield Y :
Q1±Y
(1,1) = Γ(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Y
(1,1) , Q1±Γ
(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.30a)
Q1±Y
(1,1) = ∓iD1±Y
(1,1) = Γ(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)y
(1,1) , (A.3.30b)
Q1±Γ
(1,1)∓ = ∓iD1
±Γ(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.30c)
Q1±Γ
(1,1)β = ∓iD1
±Γ(1,1)β = −Q1
βΓ(1,1)± . (A.3.30d)
Vector superfield
We should discuss the supersymmetry transformation rules of the (1, 1) reduced vector superfield V (1,1) in order
to extend the supersymmetric nonlinear sigma models on generalized complex geometries [75] to supersymmetric
A.3 N = (1, 1) supersymmetry 169
gauge theories. We have already understood semi-chiral superfields X(2,2) as well as chiral superfields Φ(2,2) possess
non-trivial U(1) charges and they can be coupled to vector superfield V (2,2) in N = (2, 2) supersymmetry level. In
order to obtain generic transformation rules in the superfield formalism, we should not impose the Wess-Zumino
gauge condition on the vector superfield, i.e., we must calculate and check the supersymmetry transformation rules
by using component fields without the Wess-Zumino gauge.
First we write down the reduced (1, 1) vector superfield without imposing the Wess-Zumino gauge on the original
(2, 2) vector superfield:
V (1,1) ≡ V (2,2)∣∣∣eq.(A.3.1)
= φ+ i√
2θ+1 (ψ+ + ψ+) + i√
2θ−1 (ψ− + ψ−) + 2iθ+1 θ−1
(F + F ) + (G+ G)
.
(A.3.31)
We calculate the supersymmetry operations as discussed in the chiral superfield and the semi-chiral superfield:
Q1+V
(2,2)∣∣∣ = (ψ+ + ψ+) +
√2θ+1 (∂0 + ∂1)φ+
√2θ−1
(F + F ) + (G+ G)
+ 2iθ+1 θ
−1 (∂0 + ∂1)(ψ− + ψ−)
≡ Γ(1,1)+ , (A.3.32a)
Q1−V
(2,2)∣∣∣ = (ψ− + ψ−)−
√2θ+1
(F + F ) + (G+ G)
+√
2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ−1 (∂0 − ∂1)(ψ+ + ψ+)
≡ Γ(1,1)− , (A.3.32b)
Q1+V
(2,2)∣∣∣ = i(ψ+ − ψ+) +
√2θ+1 (v0 + v1) + i
√2θ−1
(F − F ) + (G− G)
+ 2√
2iθ+1 θ−1 (λ+ + λ+)
≡ Ξ(1,1)+ , (A.3.32c)
Q1−V
(2,2)∣∣∣ = i(ψ− − ψ−)− i
√2θ+1
(F − F )− (G− G)
+√
2θ−1 (v0 − v1)− 2√
2iθ+1 θ−1 (λ− + λ−)
≡ Ξ(1,1)− , (A.3.32d)
D1+V
(2,2)∣∣∣ = (ψ+ + ψ+)−
√2θ+1 (∂0 + ∂1)φ+
√2θ−1
(F + F ) + (G+ G)
− 2iθ+1 θ
−1 (∂0 + ∂1)(ψ− + ψ−)
≡ −iΓ(1,1)+ , (A.3.32e)
D1−V
(2,2)∣∣∣ = (ψ− + ψ−)−
√2θ+1
(F + F ) + (G+ G)
−√
2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ−1 (∂0 − ∂1)(ψ+ + ψ+)
≡ −iΓ(1,1)− , (A.3.32f)
D1+V
(2,2)∣∣∣ = i(ψ+ − ψ+) + i
√2θ−1
(F − F ) + (G− G)
+√
2θ+1 (v0 + v1) + 2√
2iθ+1 θ−1 (λ+ + λ+)
= Ξ(1,1)+ , (A.3.32g)
D1−V
(2,2)∣∣∣ = i(ψ− − ψ−)− i
√2θ+1
(F − F )− (G− G)
+√
2θ−1 (v0 − v1)− 2√
2iθ+1 θ−1 (λ− + λ−)
= Ξ(1,1)− . (A.3.32h)
We also calculate them only within (1, 1) supersymmetry level:
Q1+ · V (1,1) = (ψ+ + ψ+) +
√2θ+1 (∂0 + ∂1)φ+
√2θ−1
(F + F ) + (G+ G)
+ 2iθ+1 θ
−1 (∂0 + ∂1)(ψ− + ψ−)
= Γ(1,1)+ , (A.3.33a)
Q1− · V (1,1) = (ψ− + ψ−)−
√2θ+1
(F + F ) + (G+ G)
+√
2θ−1 (∂0 + ∂1)φ− 2iθ+1 θ−1 (∂0 − ∂1)(ψ+ + ψ+)
= Γ(1,1)− , (A.3.33b)
170 Conventions
Q1± · V (1,1) = trivially zero , (A.3.33c)
D1+ · V (1,1) = (ψ+ + ψ+)−
√2θ+1 (∂0 + ∂1)φ+
√2θ−1
(F + F ) + (G+ G)
− 2iθ+1 θ
−1 (∂0 + ∂1)(ψ− + ψ−)
= −iΓ(1,1)+ , (A.3.33d)
D1− · V (1,1) = (ψ− + ψ−)−
√2θ+1
(F + F ) + (G+ G)
−√
2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ−1 (∂0 − ∂1)(ψ+ + ψ+)
= −iΓ(1,1)− , (A.3.33e)
D1± · V (1,1) = trivially zero . (A.3.33f)
Thus we found that, except for the trivial operators in (1, 1) supersymmetry level, the operations after the reductions
are equal to the reductions after the operations as we have already discussed in the previous examples. The non-
trivial results from the second supercharges operations generate new spinor superfields Ξ(1,1)± , which are completely
independent of V (1,1), while Γ±(1,1) and Γ(1,1)± do not contain independent component fields.
Once we obtain the above non-trivial relations, we can easily apply them to the second operations. Here we
only summarize the supersymmetry operations on the vector superfield:
Q1±V
(1,1) = Γ(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)V
(1,1) , Q1±Γ
(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.34a)
iD1±V
(1,1) = Γ(1,1)± , iD1
±Γ(1,1)± = −i(∂0 ± ∂1)V
(1,1) , iD1±Γ
(1,1)∓ = −iD1
∓Γ(1,1)± , (A.3.34b)
Q1±V
(2,2)∣∣∣ = Ξ
(1,1)± , Q1
±Ξ(1,1)± = −i(∂0 ± ∂1)V
(1,1) , Q1±Ξ
(1,1)∓ = −Q1
∓Ξ(1,1)± , (A.3.34c)
Q1αΓ
(1,1)β = −iD1
βΓ(1,1)α , Q1
αΓ(1,1)β = −Q1
βΞ(1,1)α , Q1
αΓ(1,1)β = −iD1
βΞ(1,1)α . (A.3.34d)
Here we also mention that the result of two D1± operations are different from the result of two Q1
± operations even
though Q1±V
(2,2)∣∣ = D1
±V(2,2)
∣∣ are satisfied. The result of the action of two differential operators are
Q1±Q
1±V
(2,2)∣∣∣ = −i(∂0 ± ∂1)V
(1,1) = −D1±D
1±V
(2,2)∣∣∣ . (A.3.35)
We can show this explicitly by simple calculations with component field expansions. Notice that we need not
use D1± explicitly if we calculate N = (2, 2) and N = (1, 1) supersymmetry transformations. We only use the
supercharges Q1± and Q1
±.
Semi-chiral superfields
Here let us discuss the supersymmetry operations on a (2, 2) semi-chiral superfield X(2,2) [75]:
Q1+X(2,2)
∣∣∣eq.(A.3.1)
= ψ+ +√
2θ+1 (∂0 + ∂1)φ+√
2θ−1 (F + G) + 2iθ+1 θ−1 (∂0 + ∂1)(ψ− + χ−)
≡ Γ(1,1)+ , (A.3.36a)
Q1−X(2,2)
∣∣∣eq.(A.3.1)
= (ψ− + χ−)−√
2θ+1 (F + G) +√
2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ−1 (∂0 − ∂1)ψ+
≡ Γ(1,1)− , (A.3.36b)
Q1+X(2,2)
∣∣∣eq.(A.3.1)
= iψ+ − i√
2θ+1 (∂0 + ∂1)φ+ i√
2θ−1 (F + G) + 2θ+1 θ−1 (∂0 + ∂1)(ψ− + χ−)
≡ Γ(1,1)+ , (A.3.36c)
A.3 N = (1, 1) supersymmetry 171
Q1−X(2,2)
∣∣∣eq.(A.3.1)
= i(ψ− − χ−)− i√
2θ+1 (F − G) +√
2θ−1 (A0 −A1) + 2√
2θ+1 θ−1 ζ−
≡ Ψ(1,1)− , (A.3.36d)
D1+X(2,2)
∣∣∣eq.(A.3.1)
= ψ+ −√
2θ+1 (∂0 + ∂1)φ+√
2θ−1 (F + G)− 2iθ+1 θ−1 (∂0 + ∂1)(ψ− + χ−)
= −iΓ(1,1)+ , (A.3.36e)
D1−X(2,2)
∣∣∣eq.(A.3.1)
= (ψ− + χ−)−√
2θ+1 (F + G)−√
2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ−1 (∂0 − ∂1)ψ+
≡ −iΓ(1,1)− , (A.3.36f)
D1+X(2,2)
∣∣∣eq.(A.3.1)
= iψ+ − i√
2θ+1 (∂0 + ∂1)φ+ i√
2θ−1 (F + G) + 2θ+1 θ−1 (∂0 + ∂1)(ψ− + χ−)
= Γ(1,1)+ , (A.3.36g)
D1−X(2,2)
∣∣∣eq.(A.3.1)
= i(ψ− − χ−)− i√
2θ+1 (F − G) +√
2θ−1 (A0 −A1) + 2√
2θ+1 θ−1 ζ−
= Ψ(1,1)− , (A.3.36h)
where we re-defined the spinor field ζ− = −ie−iν+ζ−. Notice that Γ(1,1)− is only defined by the operator iD1
−, while
Γ(1,1)± ; Γ
(1,1)+ are obtained not only from the operators Q1
±; Q1+ but also from iD1
±, D1+, respectively. We
define a new spinor superfield Ψ(1,1)− via the supercharge Q1
−.
We also write down the supersymmtry operations in the N = (1, 1) supersymmetry level:
Q1+ · X(1,1) = ψ+ +
√2θ+1 (∂0 + ∂1)φ+
√2θ−1 (F + G) + 2iθ+1 θ
−1 (∂0 + ∂1)(ψ− + χ−) = Γ
(1,1)+ , (A.3.37a)
Q1− · X(1,1) = (ψ− + χ−)−
√2θ+1 (F + G) +
√2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ
−1 (∂0 − ∂1)ψ+ = Γ
(1,1)− , (A.3.37b)
Q1+ · X(1,1) = trivially zero , (A.3.37c)
Q1− · X(1,1) = trivially zero , (A.3.37d)
D1+ · X(1,1) = ψ+ −
√2θ+1 (∂0 + ∂1)φ+
√2θ−1 (F + G)− 2iθ+1 θ
−1 (∂0 + ∂1)(ψ− + χ−) = −iΓ(1,1)
+ , (A.3.37e)
D1− · X(1,1) = (ψ− + χ−)−
√2(F + G)−
√2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ
−1 (∂0 − ∂1)ψ+ = −iΓ(1,1)
− , (A.3.37f)
D1+ · X(1,1) = trivially zero , (A.3.37g)
D1− · X(1,1) = trivially zero . (A.3.37h)
Notice that spinor superfields Γ(1,1)± ; Γ
(1,1)± are not independent of the (1, 1) reduced semi-chiral superfield (called
a scalar superfield) X(1,1) because all the component fields in the spinor superfields also appear in the scalar
superfield. However, surprisingly, since the spinor superfield Ψ(1,1)− has independent component fields of the scalar
superfield, we understand this newly obtained spinor superfield Ψ(1,1)− is really independent of X(1,1). We also find
that this should be a candidate of the spinor superfields which represent cotangent bundle coordinates of generalized
complex geometries [75].
Let us continue to investigate the supersymmetry operations:
Q1+ · Γ(1,1)
+ = −i(∂0 + ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 (ψ− + χ−) + 2iθ+1 θ
−1 (F + G)
= −i(∂0 + ∂1)X(1,1) =
1
2Q1
+, Q1+X(2,2)
∣∣∣ , (A.3.38a)
172 Conventions
Q1− · Γ(1,1)
+ = −i(F + G)−√
2θ+1 (∂0 + ∂1)(ψ− + χ−) +√
2θ−1 (∂0 − ∂1)ψ+ − 2θ+1 θ−1 (∂2
0 − ∂21)φ
= −Q1+ · Γ(1,1)
− = Q1−Q
1+X(2,2)
∣∣∣ , (A.3.38b)
D1+ · Γ(1,1)
+ = −i(∂0 + ∂1)φ− i
√2θ+1 ψ+ + i
√2θ−1 (ψ− + χ−)− 2iθ+1 θ
−1 (F + G)
= iQ1+ · Γ(1,1)
+ = iQ1+Q
1+X(2,2)
∣∣∣
= −iQ1+ · Γ(1,1)
+ , (A.3.38c)
D1− · Γ(1,1)
+ = −i(F + G)−√
2θ+1 (∂0 + ∂1)(ψ− + χ−)−√
2θ−1 (∂0 − ∂1)ψ+ + 2θ+1 θ−1 (∂2
0 − ∂21)φ
= iQ1+ · Γ(1,1)
− = −Q1+D
1−X(2,2)
∣∣∣ , (A.3.38d)
Q1+ · Γ(1,1)
− = i(F + G) +√
2θ+1 (∂0 + ∂1)(ψ− + χ−)−√
2θ−1 (∂0 − ∂1)ψ+ + 2θ+1 θ−1 (∂2
0 − ∂21)φ
= −Q1− · Γ(1,1)
+ = Q1+Q
1−X(2,2)
∣∣∣ , (A.3.39a)
Q1− · Γ(1,1)
− = −i(∂0 − ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 (ψ− + χ−) + 2iθ+1 θ
−1 (F + G)
= −i(∂0 − ∂1)X(1,1) =
1
2Q1
−, Q1−X(2,2)
∣∣∣ , (A.3.39b)
D1+ · Γ(1,1)
− = i(F + G)−√
2θ+1 (∂0 + ∂1)(ψ− + χ−)−√
2θ−1 (∂0 − ∂1)ψ+ − 2θ+1 θ−1 (∂2
0 − ∂21)φ
= iQ1− · Γ(1,1)
+ = iQ1−Q
1+X(2,2)
∣∣∣
= −iQ1+ · Γ(1,1)
− , (A.3.39c)
D1− · Γ(1,1)
− = −i(∂0 − ∂1)φ+ i
√2θ+1 ψ+ − i
√2θ−1 (ψ− + χ−)− 2iθ+1 θ
−1 (F + G)
= iQ1− · Γ(1,1)
− = −Q1−D
1−X(2,2)
∣∣∣ , (A.3.39d)
Q1+ · Γ(1,1)
+ = −(∂0 + ∂1)φ− i
√2θ+1 ψ+ + i
√2θ−1 (ψ− + χ−)− 2iθ+1 θ
−1 (F + G)
= −iD1+ · Γ(1,1)
+ = Q1+Q
1+X(2,2)
∣∣∣
= −Q1+ · Γ(1,1)
+ , (A.3.40a)
Q1− · Γ(1,1)
+ = (F + G) + i√
2θ+1 (∂0 + ∂1)(ψ− + χ−) + i√
2θ−1 (∂0 − ∂1)ψ+ + 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −iD1+ · Γ(1,1)
− = Q1−Q
1+X(2,2)
∣∣∣
= −Q1+ · Γ(1,1)
− , (A.3.40b)
D1+ · Γ(1,1)
+ = −(∂0 + ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 (ψ− + χ−) + 2iθ+1 θ
−1 (F + G)
= −(∂0 + ∂1)X(1,1) = − i
2Q1
+, Q1+X(2,2)
∣∣∣ , (A.3.40c)
D1− · Γ(1,1)
+ = (F + G) + i√
2θ+1 (∂0 + ∂1)(ψ− + χ−)− i√
2θ−1 (∂0 − ∂1)ψ+ − 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −D1+ · Γ(1,1)
− = D1−Q
1+X(2,2)
∣∣∣
= iQ1+ · Γ(1,1)
− , (A.3.40d)
A.3 N = (1, 1) supersymmetry 173
Q1+ · Γ(1,1)
− = −(F + G) + i√
2θ+1 (∂0 + ∂1)(ψ− + χ−) + i√
2θ−1 (∂0 − ∂1)ψ+ − 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −iD1− · Γ(1,1)
+ = iQ1+D
1−X(2,2)
∣∣∣ , (A.3.41a)
Q1− · Γ(1,1)
− = −(∂0 − ∂1)φ+ i
√2θ+1 ψ+ − i
√2θ−1 (ψ− + χ−)− 2iθ+1 θ
−1 (F + G)
= −iD1− · Γ(1,1)
− = iQ1−D
1−X(2,2)
∣∣∣ , (A.3.41b)
D1+ · Γ(1,1)
− = −(F + G)− i√
2θ+1 (∂0 + ∂1)(ψ− + χ−) + i√
2θ−1 (∂0 − ∂1)ψ+ + 2iθ+1 θ−1 (∂2
0 − ∂21)φ
= −D1− · Γ(1,1)
+ = Q1+D
1−X(2,2)
∣∣∣
= −iQ1+ · Γ(1,1)
− , (A.3.41c)
D1− · Γ(1,1)
− = −(∂0 − ∂1)φ+ i
√2θ+1 ψ+ + i
√2θ−1 (ψ− + χ−) + 2iθ+1 θ
−1 (F + G)
= −(∂0 − ∂1)X(1,1) =
i
2D1
−,D1−X(2,2)
∣∣∣ . (A.3.41d)
The results of the above relations are same as the ones of the (2, 2) chiral superfield Φ(2,2). The following relations
are newly obtained because they are derivatives of the new spinor superfield Ψ(1,1)− , which are independent of the
scalar superfield X(1,1):
Q1+ ·Ψ(1,1)
− = −(F − G) + i√
2θ+1 (∂0 + ∂1)(ψ− − χ−) + 2iθ−1 ζ− + 2θ+1 θ−1 (∂0 + ∂1)(A0 −A1)
= Q1+Q
1−X(2,2)
∣∣∣ = −Q1− · Γ(1,1)
+ , (A.3.42a)
Q1− ·Ψ(1,1)
− = −i(A0 −A1) + 2iθ+1 ζ− + i√
2θ−1 (∂0 − ∂1)(ψ− − χ−) + 2iθ+1 θ−1 (∂0 − ∂1)(F − G)
= Q1−Q
1−X(2,2)
∣∣∣ = −Q1− · Γ(1,1)
− , (A.3.42b)
D1+ ·Ψ(1,1)
− = −(F − G)− i√
2θ+1 (∂0 + ∂1)(ψ− − χ−)− 2iθ−1 ζ− − 2θ+1 θ−1 (∂0 + ∂1)(A0 −A1)
= −iQ1+Q
1−X(2,2)
∣∣∣ = iQ1− · Γ(1,1)
+
= −iQ1+ ·Ψ(1,1)
− , (A.3.42c)
D1− ·Ψ(1,1)
− = −i(A0 −A1) + 2iθ+1 ζ− − i√
2θ−1 (∂0 − ∂1)(ψ− − χ−)− 2iθ+1 θ−1 (∂0 − ∂1)(F − G)
= D1−Q
1−X(2,2)
∣∣∣ = iQ1− · Γ(1,1)
− . (A.3.42d)
Furthermore we find that the following equation from purely (2, 2) supersymmetry algebra:
Q1− ·Ψ(1,1)
− =1
2Q1
−, Q1−X(2,2)
∣∣∣ = −i(∂0 − ∂1)X(1,1) . (A.3.43)
As we mentioned before, since the two operations of Q1−X(2,2)| and iD1
−X(2,2)| are not equivalent to each other,
there appears a new spinor superfield Ψ(1,1)− in the (1, 1) reductions of semi-chiral superfield. This superfield is
represented by new component fields, which are not contained in the (1, 1) reduced superfield X(1,1).
Here we summarize the above non-trivial relations in the (1, 1) supersymmetry language:
Q1±X(1,1) = Γ
(1,1)± , Q1
±Γ(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.44a)
Q1+X(1,1) = iD1
+X(1,1) = Γ(1,1)+ , Q1
−X(1,1) = Ψ(1,1)− , iD1
−X(1,1) = Γ(1,1)− , (A.3.44b)
174 Conventions
Q1+Ψ
(1,1)− = iD1
+Ψ(1,1)− = −Q1
−Γ(1,1)+ , (A.3.44c)
Q1+Γ
(1,1)+ = Q1
+Γ(1,1)+ = −i(∂0 + ∂1)X
(1,1) , (A.3.44d)
Q1−Γ
(1,1)− = Q1
−Ψ(1,1)− = iD1
−Γ(1,1)− = −i(∂0 − ∂1)X
(1,1) , (A.3.44e)
Q1±Γ
(1,1)+ = −Q1
+Γ(1,1)± , Q1
±Γ(1,1)− = −iD1
−Γ(1,1)± , Q1
±Ψ(1,1)− = −Q1
−Γ(1,1)± . (A.3.44f)
Here we also mention that the result of two D1± operations are different from the result of two Q1
± operations even
though Q1±X(2,2)
∣∣ = D1±X(2,2)
∣∣ are satisfied. The result of the action of two differential operators are
Q1±Q
1±X(2,2)
∣∣∣ = −i(∂0 ± ∂1)X(1,1) = −D1
±D1±X(2,2)
∣∣∣ . (A.3.45)
We can show this explicitly by simple calculations with component field expansions. Notice that we need not
use D1± explicitly if we calculate N = (2, 2) and N = (1, 1) supersymmetry transformations. We only use the
supercharges Q1± and Q1
±.
Here we also pick up the supersymmetry operations on the left anti-chiral superfield X:
Q1±X(1,1) = Γ
(1,1)± , Q1
±Γ(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.46a)
Q1+X(1,1) = −iD1
+X(1,1) , Q1−X(1,1) = Ψ
(1,1)− , −iD1
−X(1,1) = Γ(1,1)− , (A.3.46b)
Q1+Ψ
(1,1)− = −iD1
+Ψ(1,1)− = −Q1
−Γ(1,1)+ , (A.3.46c)
Q1+Γ
(1,1)+ = Q1
+Γ(1,1)+ = −i(∂0 + ∂1)X
(1,1) , (A.3.46d)
Q1−Γ
(1,1)− = Q1
−Ψ(1,1)− = −iD1
−Γ(1,1)− = −i(∂0 − ∂1)X
(1,1) , (A.3.46e)
Q1±Γ
(1,1)+ = −Q1
+Γ(1,1)± , Q1
±Γ− = iD1−Γ
(1,1)± , Q1
±Ψ(1,1)− = −Q1
−Γ(1,1)± . (A.3.46f)
Next we consider the supersymmetry operation rules for the right chiral superfield Y. The rules are quite similar
to the ones for left chiral superfield. Here we write down the rules explicitly:
Q1+Y(2,2)
∣∣∣ = (ψ+ + χ+) +√
2θ+1 (∂0 + ∂1)φ+√
2θ−1 (F + N) + 2iθ+1 θ−1 (∂0 + ∂1)ψ− ≡ Γ
(1,1)+ , (A.3.47a)
Q1−Y(2,2)
∣∣∣ = ψ− −√
2θ+1 (F + N) +√
2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ−1 (∂0 − ∂1)(ψ+ + χ+) ≡ Γ
(1,1)− , (A.3.47b)
Q1+Y(2,2)
∣∣∣ = i(ψ+ − χ+) +√
2θ+1 (B0 +B1) + i√
2θ−1 (F − N)− 2√
2θ+1 θ−1 ζ+ ≡ Υ
(1,1)+ , (A.3.47c)
Q1−Y(2,2)
∣∣∣ = iψ− − i√
2θ+1 (F + N)− i√
2θ−1 (∂0 − ∂1)φ− 2θ+1 θ−1 (∂0 − ∂1)(ψ+ + χ+) ≡ Γ
(1,1)− , (A.3.47d)
D1+Y(2,2)
∣∣∣ = (ψ+ + χ+)−√
2θ+1 (∂0 + ∂1)φ+√
2θ−1 (F + N)− 2iθ+1 θ−1 (∂0 + ∂1)ψ− = −iΓ(1,1)
+ , (A.3.47e)
D1−Y(2,2)
∣∣∣ = ψ− −√
2θ+1 (F + N)−√
2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ−1 (∂0 − ∂1)(ψ+ + χ+) = −iΓ(1,1)
− , (A.3.47f)
D1+Y(2,2)
∣∣∣ = i(ψ+ − χ+) +√
2θ+1 (B0 +B1) + i√
2θ−1 (F − N)− 2√
2θ+1 θ−1 ζ+ = Υ
(1,1)+ , (A.3.47g)
D1−Y(2,2)
∣∣∣ = iψ− − i√
2θ+1 (F + N)− i√
2θ−1 (∂0 − ∂1)φ− 2θ+1 θ−1 (∂0 − ∂1)(ψ+ + χ+) = Γ
(1,1)− , (A.3.47h)
where we re-defined ζ+ = ieiν−ζ+ and N = e−iν++iν−N .
Q1+ · Y(1,1) = (ψ+ + χ+) +
√2θ+1 (∂0 + ∂1)φ+
√2θ−1 (F + N) + 2iθ+1 θ
−1 (∂0 + ∂1)ψ− , (A.3.48a)
Q1− · Y(1,1) = ψ− −
√2θ+1 (F + N) +
√2θ−1 (∂0 − ∂1)φ− 2iθ+1 θ
−1 (∂0 − ∂1)(ψ+ + χ+) , (A.3.48b)
A.3 N = (1, 1) supersymmetry 175
Q1+ · Y(1,1) = trivially zero , (A.3.48c)
Q1− · Y(1,1) = trivially zero , (A.3.48d)
D1+ · Y(1,1) = (ψ+ + χ+)−
√2θ+1 (∂0 + ∂1)φ+
√2θ−1 (F + N)− 2iθ+1 θ
−1 (∂0 + ∂1)ψ− , (A.3.48e)
D1− · Y(1,1) = ψ− −
√2θ+1 (F + N)−
√2θ−1 (∂0 − ∂1)φ+ 2iθ+1 θ
−1 (∂0 − ∂1)(ψ+ + χ+) , (A.3.48f)
D1+ · Y(1,1) = trivially zero , (A.3.48g)
D1− · Y(1,1) = trivially zero , (A.3.48h)
Thus we can easily find the following supersymmetry transformation rules:
Q1±Y(1,1) = Γ
(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Y
(1,1) , Q1±Γ
(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.49a)
Q1+Y(1,1) = Υ
(1,1)+ , Q1
−Y(1,1) = iD1−Y(1,1) = Γ
(1,1)− , iD1
+Y(1,1) = Γ(1,1)+ , (A.3.49b)
Q1+Υ
(1,1)+ = −i(∂0 + ∂1)Y
(1,1) , Q1−Υ
(1,1)+ = iD1
−Υ(1,1)+ = −Q1
+Γ(1,1)− , (A.3.49c)
iD1+Γ
(1,1)+ = −i(∂0 + ∂1)Y
(1,1) , (A.3.49d)
Q1±Γ
(1,1)− = −Q1
−Γ(1,1)± , Q1
±Γ(1,1)+ = −iD1
+Γ(1,1)± , Q1
±Υ(1,1)+ = −Q1
+Γ(1,1)± . (A.3.49e)
Finally we write down the supersymmetry operation rules for the right anti-chiral superfield Y:
Q1±Y(1,1) = Γ
(1,1)± , Q1
±Γ(1,1)± = −i(∂0 ± ∂1)Y
(1,1) , Q1±Γ
(1,1)∓ = −Q1
∓Γ(1,1)± , (A.3.50a)
Q1+Y(1,1) = Υ
(1,1)+ , Q1
−Y(1,1) = −iD1−Y(1,1) = Γ
(1,1)− , −iD1
+Y(1,1) = Γ(1,1)+ , (A.3.50b)
Q1+Υ
(1,1)+ = −i(∂0 + ∂1)Y
(1,1) , Q1−Υ
(1,1)+ = −iD1
−Υ(1,1)+ = −Q1
+Γ(1,1)− , (A.3.50c)
−iD1+Γ
(1,1)+ = −i(∂0 + ∂1)Y
(1,1) , (A.3.50d)
Q1±Γ
(1,1)− = −Q1
−Γ(1,1)± , Q1
±Γ(1,1)+ = iD1
+Γ(1,1)± , Q1
±Υ(1,1)+ = −Q1
+Γ(1,1)± . (A.3.50e)
A.3.4 Superspace measure
In this subsection we discuss the modification of the measures of fermionic coordinates defined in (A.2.12). We
prepare an N = (2, 2) supersymmetric integral as follows:∫
d2xd4θ f(Fi) = −1
4
∫d2xdθ+dθ−dθ+dθ− f(Fi) , (A.3.51)
where f(Fi) is a function of arbitrary (2, 2) superfields Fi. Due to the definitions (A.2.13), we can regard the
fermionic measures as derivatives of their coordinates. Furthermore, we can rewrite them as the supercovariant
derivatives within the above integral (A.3.51) such as
dθ± =∂
∂θ±= D± , dθ± =
∂
∂θ±= −D± . (A.3.52)
Thus if we reduce an N = (2, 2) theory to an N = (1, 1) theory, we should transform the above measure to the
suitable (1, 1) measure via the definitions (A.3.3) and (A.3.5):∫
d4θ f(Fi) = −1
4
∫D+D−D+D− f(Fi) =
1
4
∫D1
+D1−D
1+D
1− f(Fi)
176 Conventions
= −1
8
∫dθ+1 dθ−1 Q
1+Q
1− f(Fi) . (A.3.53)
Notice that we changed the operators because we have already found that the second supersymmetry covariant
derivatives D1± are equivalent to the second supercharges Q1
± in N = (2, 2) supersymmetry level. Furthermore we
again changed the derivatives of θ±1 to their measures inside the integral (A.3.51):
D1+D
1− = −1
2
∂
∂θ+1
∂
∂θ−1= −1
2dθ+1 dθ−1 . (A.3.54)
Thus we derived the N = (1, 1) supersymmetric “D-term” Q1+Q
1−f(Fi)
∣∣eq.(A.3.1)
from the N = (2, 2) supersym-
metric D-term up to an overall factor.
Next we discuss the reduction of F-term in (2, 2) supersymmetric theories. A generic holomorphic F-term is
given by∫
d2xd2θW (Φi) = −1
2
∫d2xdθ+dθ−W (Φi) , (A.3.55)
where we obeyed the definitions (A.2.13). Notice that W (Φi) is a function of arbitrary chiral superfields Φi defined
by D±Φi = 0. We cannot add any semi-chiral superfields into this function. Here we modify the integral measure
of fermionic coordinates:
−1
2
∫dθ+dθ−W (Φi) = −1
2
∫D+D−W (Φi) = −1
2e−iν+−iν−
∫(eiν+D+)(eiν−D−)W (Φi)
= −1
2e−iν+−iν−
∫ (eiν+D+ + e−iν+D+
)(eiν−D− + e−iν−D−
)W (Φi)
= −e−iν+−iν−
∫D1
+D1−W (Φi)
=1
2e−iν+−iν−
∫dθ+1 dθ−1 W (Φi) , (A.3.56)
where we used the relations D±W (Φi) = 0. We also obtain the following:
∫d2θW (Φ) = −1
2e+iν++iν−
∫dθ+1 dθ−1 W (Φ) =
(∫d2θW (Φ)
)†. (A.3.57)
We can also reduce any (2, 2) twisted F-terms, which are represented by twisted chiral superfields Yi. As
discussed before, twisted chiral superfields are defined by D+Yi = D−Yi = 0. Here we modify an arbitrary twisted
F-term which is given by∫
d2xd2θ W (Yi) = −1
2
∫d2xdθ+dθ− W (Yi) , (A.3.58)
where W (Yi) is a function of arbitrary twisted chiral superfields. As in the same way as the F-term, the integral
measure of fermionic coordinates are modified as follows:
−1
2
∫dθ+dθ− W (Yi) =
1
2
∫D+D− W (Yi) =
1
2e−iν++iν−
∫(eiν+D+)(e−iν−D−) W (Yi)
=1
2e−iν++iν−
∫ (eiν+D+ + e−iν+D+
)(eiν−D− + e−iν−D+
)W (Yi)
= e−iν++iν−
∫D1
+D1− W (Yi)
A.4 N = (0, 2) supersymmetry 177
= −1
2e−iν++iν−
∫dθ+1 dθ−1 W (Yi) . (A.3.59)
Here we summarize the above discussion. A generic (2, 2) supersymmetric Lagrangian containing a D-term, a
F-term and a twisted F-term is reduced to the following function:
L =
∫d4θ f(Fi) +
(∫d2θW (Φa) + (c.c.)
)+(∫
d2θ W (Yb) + (c.c.))
=
∫dθ+1 dθ−1
− 1
8Q1
+Q1− f(Fi) +
1
2
(e−iν+−iν−W (Φa) + (c.c.)
)− 1
2
(e−iν++iν−W (Yb) + (c.c.)
)∣∣∣eq.(A.3.1)
.
(A.3.60)
We find that there are no distinctions among the “D-term”, “F-term” and “twisted F-term” in the N = (1, 1)
supersymmetry level.
A.4 N = (0, 2) supersymmetry
A.4.1 Differential operators
We next consider N = (0, 2) supersymmetry, which has two supercharges of positive (or “right”) chirality. The
relevant subspace of the N = (2, 2) superspace is the (0, 2) superspace defined by
θ− = θ− = 0 . (A.4.1)
In this subspace the following differential operators are defined:
Q+ =∂
∂θ++ iθ+
(∂0 + ∂1
), Q+ = − ∂
∂θ+− iθ+
(∂0 + ∂1
), (A.4.2)
D+ =∂
∂θ+− iθ+
(∂0 + ∂1
), D+ = − ∂
∂θ++ iθ+
(∂0 + ∂1
). (A.4.3)
R-rotation of the superfield is defined by
eiαF+ : F(x, θ+, θ+) 7→ eiαq+F(x, e−iαθ+, eiαθ+) , (A.4.4)
where F(x, θ+, θ+) is an arbitrary (0, 2) superfield. The supersymmetry algebra is described as follows (see N =
(2, 2) algebra (A.2.8), etc.):
Q+, Q+ = −2i(∂0 + ∂1
)= 2(H − P ) , Q2
+ = Q2+ = 0 , (A.4.5a)
D+,D+ = 2i(∂0 + ∂1
), D+,D+ = D+,D+ = 0 , (A.4.5b)
[M,Q+] = −Q+ , [M,Q+] = −Q+ , (A.4.5c)
[F+, Q+] = −Q+ , [F+, Q+] = +Q+ . (A.4.5d)
178 Conventions
A.4.2 (0, 2) superfields
Gauge multiplet
To construct gauge theories11, we need to extend our superspace derivatives, D+ and D+, to gauge covariant
superderivatives. The gauge covariant superderivatives D+,D+ acting on charge Q fields, and Dm (m = 0, 1)
satisfy the algebra
D2+ = D2
+ = 0 , D+,D+ = 2i(D0 +D1) . (A.4.6)
The first two equations imply that D+ = e−QΨD+eQΨ and D+ = eQΨD+e−QΨ where Ψ is a superfield taking
values in the Lie algebra of the gauge group. We will restrict to abelian theories in our discussion. In Wess-Zumino
gauge, the component expansion of Ψ gives
Ψ = θ+θ+(v0 + v1) , (A.4.7)
while
D0 +D1 = ∂0 + ∂1 + i(v0 + v1) , D0 −D1 = ∂0 − ∂1 + iV , (A.4.8a)
D+ =∂
∂θ+− iθ+(D0 +D1) , D+ = − ∂
∂θ++ iθ+(D0 +D1) . (A.4.8b)
The vector superfield V is given by
V = v0 − v1 − 2iθ+λ− − 2iθ+λ− + 2θ+θ+D . (A.4.9)
We see that the v0− v1 component of the gauge-field has two real gaugino partners, while v0 + v1 does not. Under
a gauge transformation with gauge parameter Λ satisfying a chiral constraint D+Λ = 0, the two gauge-fields V and
Ψ transform as follows
δΛV = ∂−(Λ + Λ) , δΛΨ = i(Λ− Λ) .
Finally, the natural field strength is an uncharged fermionic chiral superfield
Υ = [D+,D0 −D1] = D+(∂−Ψ + iV ) = −2λ−(z)− iθ+(D − iF01) , (A.4.10)
Chiral multiplets
An uncharged chiral superfield is one which satisfies D+Φ0 = 0. Chiral superfields are therefore naturally expanded
by components fields as follows,
Φ0 = φ(x) + i√
2θ+ψ+(x)− iθ+θ+(∂0 + ∂1)φ(x) . (A.4.11)
There are no auxiliary fields12.
11Here we introduce (0, 2) superfields from [1].12Coefficients in the expanded representaion is slightly different from the ones in [1].
A.4 N = (0, 2) supersymmetry 179
Fermi multiplets
In addition to bosonic chiral multiplets, there are also fermionic multiplets which, for uncharged fields, satisfy the
condition
D+Γ0 =√
2E0 (A.4.12)
where E0 satisfies
D+E0 = 0 .
A component expansion gives the terms
Γ0 = iχ− − i√
2θ+G− iθ+θ+(∂0 + ∂1)χ− −√
2θ+E0 . (A.4.13)
Note that fermi multiplets have negative chirality, i.e., χ−. To satisfy the covariant chirality condition, we again
define Γ0 = e−QΨΓ, and E0 = e−QΨE so that
D+Γ =√
2E . (A.4.14)
The choice of E plays an important role in our discussion for reasons that we will describe later. We follow [113]
and assume that E is a holomorphic function of chiral superfields Φi.
Additional chiral multiplet
We can also define a superfield including scalar component of vector multiplet in (2, 2) supersymmetry. Σ is an
uncharged chiral superfield with component expansion13
Σ = σ − i√
2θ+λ+ − iθ+θ+(∂0 + ∂1)σ . (A.4.15)
Reduction from (2, 2) superfields
We always obtain (0, 2) superfields F (0,2) defined above from (2, 2) superfields F (2,2) as follows14.
An uncharged (2, 2) chiral multiplet gives a (0, 2) chiral and Fermi multiplet (E0 = 0 case),
Φ(0,2) = Φ(2,2)∣∣∣θ−=θ−=0
, Γ(0,2) =1√2D−Φ(2,2)
∣∣∣θ−=θ−=0
. (A.4.16)
Similarly, a twisted chiral multiplet (which is always uncharged) also gives a chiral and Fermi multiplet,
Y (0,2) = Y (2,2)∣∣∣θ−=θ−=0
, F (0,2) = − 1√2D−Y
(2,2)∣∣∣θ−=θ−=0
. (A.4.17)
There is also a (2, 2) vector superfield, V , whose field strength is a twisted chiral multiplet (often denoted Σ). On
reduction to (0, 2), we obtain a chiral multiplet, Σ(0,2), and a vector multiplet, V (0,2), as follows:
Σ(0,2) = Σ(2,2)∣∣∣θ−=θ−=0
, V (0,2) =
∫dθ− dθ− V (2,2) , (A.4.18a)
13The definition of λ+ is slightly different from the one in [1].14The derivation rules are slightly different from the ones in [1], though the results are equivalent.
180 Conventions
Ψ(0,2) = V (2,2)∣∣∣θ−=θ−=0
, Υ(0,2) = −i√
2
∫dθ− Σ(2,2) , (A.4.18b)
where we used∫
dθ− dθ− θ−θ− = −1. Lastly, we note that a (2, 2) chiral multiplet with U(1) charge Q reduces to
a charged (0, 2) chiral multiplet, Φ(0,2), and a charged Fermi multiplet, Γ(0,2), with a particular non-vanishing E
so that
D+Γ(0,2) =√
2E
where E is given by [113]
E =√
2QΣ(0,2) Φ(0,2) . (A.4.19)
A.5 N = (0, 1) supersymmetry
The (0, 1) subspace in N = (2, 2) superspace is also obtained. First we define (1, 1) subspace, then we project the
subspace written only by the fermionic with positive chirality. As we discussed in the previous section, we know two
useful procedure to realize (1, 1) supersymmetry. Thus here we also discuss two types of (0, 1) supersymmetries.
A.5.1 A definition of (0, 1) supersymmetry
Differential operators
Here we project the superspace defined by (A.3.1) into the subspace of positive chirality. The obtaining (0, 1)
supersymmetry has only one real supercharge with positive chirality. The differential operators are defined by
Q(0,1)+ ≡ − i√
2
∂
∂θ+1+√
2θ+1(∂0 + ∂1
), D
(0,1)+ ≡ − i√
2
∂
∂θ+1−√
2θ+1(∂0 + ∂1
). (A.5.1)
The supersymmetry anti-commutation relations are
Q(0,1)+ , Q
(0,1)+ = −2i
(∂0 + ∂1
)= 2(H − P ) , D(0,1)
+ ,D(0,1)+ = +2i
(∂0 + ∂1
), (A.5.2a)
Q(0,1)+ ,D
(0,1)+ = 0 . (A.5.2b)
These are projection of (A.3.4) onto the subalgebra with positive chirality.
We also introduce another pair of differential operators as in N = (1, 1) supersymmetry:
Q(0,1)+ ≡ − i√
2
∂
∂θ+1+√
2θ+1(∂0 + ∂1
), D
(0,1)+ ≡ − i√
2
∂
∂θ+1−√
2θ+1(∂0 + ∂1
). (A.5.3)
They also satisfy the following anti-commutation relations:
Q(0,1)+ , Q
(0,1)+ = −2i
(∂0 + ∂1
)= 2(H − P ) , D(0,1)
+ , D(0,1)+ = +2i
(∂0 + ∂1
), (A.5.4a)
Q(0,1)+ , D
(0,1)+ = 0 , (A.5.4b)
Q(0,1)+ , Q
(0,1)+ = D(0,1)
+ , D(0,1)+ = Q(0,1)
+ , D(0,1)+ = Q(0,1)
+ , D(0,1)+ = 0 . (A.5.4c)
A.6 Propagator and Fourier Transformation 181
A.6 Propagator and Fourier Transformation
参考文献は TK-NOTE/04-01 [62] である。
Minkowski metric ηmn = (−,+) での massive free scalar field Lagrangian
S =1
α
∫d2xL , (A.6.1a)
L = −ηmn∂mφ∗∂nφ−m2φ∗φ = −φ∗(−∂2 +m2)φ , ∂2 = −∂2
0 + ∂21 (A.6.1b)
を用意する (α は適当な normalization factor)。この propagator は [65] での propagator の定義を真似て15
∆F (x− y) = 〈φ∗(x)φ(y)〉 =
∫d2k
(2π)2i
αe−ik(x−y)
k2 +m2 − iε (A.6.2)
である。これより、同一点での propagator、つまり one-loop は (−iε を無視して)
〈φ∗(x)φ(x)〉 =
∫d2k
(2π)2i
α
k2 +m2(A.6.3)
となる。worldsheet を Euclid 化しよう:
k0 = ik2 , k1 = k1 . (A.6.4)
momentum は元々が coordinates の微分なので、下が標準位置である。これより
k2 = −(k0)2 + (k1)
2 = (k1)2 + (k2)
2 ≡ t , d2k = dk0dk1 = idk1dk2 =i
2dtdθ (A.6.5)
となる。これより
〈φ∗(x)φ(x)〉 =1
2 · 2π
∫ Λ2UV
µ2
dtα
t+m2=
α
2πlog(ΛUV
µ
)(A.6.6)
が得られる。
これを基に、field φ(x) の Fourier 変換は次で与えられる:
φ(x) =
∫d2k√(2π)2
φ(k) e−ikx , φ(k) =
∫d2x√(2π)2
φ(x) e+ikx . (A.6.7)
(注意) Euclid 化に際して注意がある。coordinates については x0 = −ix2 である [113, 90]。coordinate space
と momentum space とは互いに逆に回転させている。よってここでは
kx = k0x0 + k1x
1 = (ik2)(−ix2) + k1x1 = k2x
2 + k1x1 , (A.6.8)
で、Euclid 化に対して不変である。
15この phase e−ikx の符号は任意ではない。Feynman propagator を定義したとき、−iε が正しく pole を拾うためには、この計量 ηµν =
diag.(−+) では e+ikx ではみたされない。
182 Conventions
A.7 Left/right moving modes
N = (2, 2) SCFT algebra を、CY sigma model の field を用いて表現する。ここでは結果だけ記述しよう。なお、こ
こで、数箇所で注意してあるが、field の chirality index と left/right index の表記が、SCFT generator のそれと反
転している。ここでも残念ながら統一はしない。従って、これから明記する SCFT generator の field representation
において、左辺 (SCFT side) の +/− index は chiral/anti-chiral を表すのに対して、右辺 (field side) の +/− はleft/right moving field を表す。逆に、左辺 (SCFT side) の bar の有無は left/right generator を表すのに対して、右
辺 (field side) の bar の有無は chiral/anti-chiral を表す。表記に対して混乱してはならない。
このノートでの worldsheet coordinates x0, x1 を Euclidean metric (x0 = −ix2) で書き直してみよう:
z ≡ x1 + ix2 = x1 − x0 ≡ −x− , z ≡ x1 − ix2 = x1 + x0 ≡ x+ , (A.7.1a)
2∂− =(∂0 − ∂1
)= −
(∂1 − i∂2
)= −2∂ , (A.7.1b)
2∂+ =(∂0 + ∂1
)=(∂1 + i∂2
)= 2∂ . (A.7.1c)
これは cylinder coordinates である。これにより、Lagrangian (I.1.6) の on-shell では
∂φ = ∂φ(z) , ∂φ = ∂φ(z) , ∂φ = ∂φ(z) , ∂φ = ∂φ(z) , (A.7.2a)
ψ+ = ψ+(x+) = ψ+(z) , ψ− = ψ−(x−) = ψ−(z) , (A.7.2b)
ψ+ = ψ+(x+) = ψ+(z) , ψ− = ψ−(x−) = ψ−(z) , (A.7.2c)
と分解できる。
注意であるが、様々な文献ごとに、left/right の意味が逆転する。ある文献では、x+ を left といい、他の文献で
は x− をそういう。また複素座標で書いた際にも同様の逆転が登場する。元来 left/right という「単語」そのものは
ここでは意味をなさない。意味を成すのは「x− or z dependent」なのか「x+ or z dependent」なのか、別の言葉
では “holomorphic か anti-holomorphic か” である16。どちらを left/right というかは著者の好みである。しかし、
R-symmetry の分類で登場する “vector” や “axial-vector” には意味がある。それぞれ field transformation の向きが
明確に定義されているからである。そのため、vector/axial-vector current を left/right current の和や差で定義する
が、第一義的な意味を成すのは vector/axial-vector であり、left/right という「言葉」は二次的である。
一見混乱するのが metric の index gij であろう。これは、本来は
gij =∂
∂φi
∂
∂φjK
であるので (K は Kahler potential)、bar の有無は、やはり chiral/anti-chiral (field space の holomorphic/anti-
holomorphic) であって、left/right (worldsheet の holomorphic/anti-holomorphic) ではない。なお、例えば N = 2
superconformal operator を field operator で記述すると次のように記述される:
J =∑
i,j
gij ψ+,i ψ+,j(x+) , J =
∑
i,j
gij ψ−,i ψ−,j(x−) , (A.7.3a)
G+ =∑
i,j
gij ψ+,i ∂φj(x+) , G− =
∑
i,j
gij ψ−,i ∂φj(x−) , (A.7.3b)
16但しここでも z を holomorphic、z を anti-holomorphic であると「定義」していれば、の話である。
A.7 Left/right moving modes 183
G+ =∑
i,j
gij ψ+,j ∂φi(x+) , G− =
∑
i,j
gij ψ−,j ∂φi(x−) , (A.7.3c)
ここで G+, G+, G−, G− はそれぞれ left-chiral, left-antichiral, right-chiral, right-antichiral である。但し文献によっ
ては plus/minus 記号、bar の有無の表記は入れ替わる事に注意。
Appendix B
Calabi-Yau Data from Gepner Construction
186 Calabi-Yau Data from Gepner Construction
B.1 Data
Lynker and Schimmrgk [76] の付録を掲載する:
1 5 41 1804 1 5 42 922 1 5 43 628 1 5 44 481
AAAA 0 251 251 AAAA 240 257 137 AAAA 336 263 95 AAAA 0 143 143
AAAD 0 251 251 AAAD −240 137 257 AAAD 336 263 95 AADA 0 143 143
AADA −240 137 257
AADA 240 257 137
1 5 46 334 1 5 47 292 1 5 49 236 1 5 52 187
AAAA 480 287 47 AAAA 480 287 47 AAAA 0 107 107 AAAA 240 173 53
AAAD 96 149 101 AAAD 480 287 47 AAAD 0 107 107 AADA 240 173 53
AADA 96 149 101
AADD 480 287 47
1 5 54 166 1 5 58 138 1 5 61 124 1 5 68 103
AAAA 624 335 23 AAAA 144 131 59 AAAA 720 377 17 AAAA 384 221 29
AAAD 240 173 53 AAAD −144 59 131 AAAD 720 377 17 AADA 384 221 29
AADA 240 173 53 AADA −144 59 131
AADD 624 335 23 AADD 144 131 59
1 5 76 89 1 5 82 82
AAAA 0 83 83 AAAA 960 491 11
AADA 0 83 83 AAAD 480 263 23
AADD 960 491 11
1 6 23 598 1 6 24 310 1 6 25 214 1 6 26 166
AAAA 96 167 119 AAAA 216 174 66 AAAA 264 180 48 AAAA 312 190 34
AAAD −96 119 167 AAAD 72 124 88 AAAD 120 128 68 AAAD 168 134 50
ADAA −96 119 167 AADA 0 97 97 AADA 120 128 68 AADA 120 106 46
ADAD 96 167 119 AADD 0 69 69 ADAD 264 180 48 AADD 72 74 38
ADAA 72 124 88 ADAA 168 134 50
ADAD 216 174 66 ADAD 312 190 34
ADDA 0 69 69 ADDA 72 74 38
ADDD 0 97 97 ADDD 120 106 46
1 6 28 118 1 6 30 94 1 6 31 86 1 6 34 70
AAAA 360 204 24 AAAA 408 222 18 AAAA 48 81 57 AAAA 456 242 14
AAAD 216 144 36 AAAD 168 118 34 AAAD −48 57 81 AAAD 312 176 20
AADA 168 118 34 AADA 168 118 34 ADAA −48 57 81 AADA 216 130 22
AADD 120 84 24 AADD 264 158 26 ADAD 48 81 57 AADD 168 98 14
ADAA 216 144 36 ADAA 264 158 26 ADAA 312 176 20
ADAD 360 204 24 ADAD 120 84 24 ADAD 456 242 14
ADDA 120 84 24 ADDA 120 84 24 ADDA 168 98 14
ADDD 168 118 34 ADDD 264 158 26 ADDD 216 130 22
1 6 38 58 1 6 40 54 1 6 46 46
AAAA 240 143 23 AAAA 144 105 33 AAAA 624 321 9
AAAD 48 67 43 AAAD 48 73 49 AAAD 336 181 13
AADA 144 105 33 AADA 0 55 55 AADD 624 321 9
AADD 48 53 29 AADD 0 39 39 ADAA 432 229 13
ADAA 144 105 33 ADAA 48 73 49 ADAD 240 129 9
ADAD 48 53 29 ADAD 144 105 33 ADDD 432 229 13
ADDA 240 143 23 ADDA 0 39 39
ADDD 48 67 43 ADDD 0 55 55
B.1 Data 187
1 7 17 340 1 7 18 178 1 7 19 124 1 7 20 97
AAAA 144 143 71 AAAA 216 150 42 AAAA 264 160 28 AAAA 96 93 45
AAAD 144 143 71 AAAD 24 86 74 AAAD 264 160 28 AADA 96 93 45
AADA 24 86 74
AADD 216 150 42
1 7 22 70 1 7 25 52 1 7 28 43 1 7 34 34
AAAA 336 183 15 AAAA 408 214 10 AAAA 216 126 18 AAAA 528 271 7
AAAD 144 103 31 AAAD 408 214 10 AADA 216 126 18 AAAD 288 157 13
AADA 144 103 31 AADD 528 271 7
AADD 336 183 15
1 8 14 238 1 8 16 88 1 8 18 58
AAAA 168 134 50 AAAA 276 155 17 AAAA 336 178 10
AAAD 24 80 68 AAAD 96 78 30 AAAD 144 98 26
AADA 24 80 68 AADA 144 98 26 AADA 144 98 26
AADD 168 134 50 AADD 96 65 17 AADD 336 178 10
ADAA 0 89 89 ADAA 120 99 39 ADAA 192 117 21
ADAD 0 53 53 ADAD 96 65 17 ADAD 96 65 17
ADDA 0 53 53 ADDA 96 65 17 ADDA 96 65 17
ADDD 0 89 89 ADDD 96 65 17 ADDD 192 117 21
AAEA 36 62 44
AAED 24 41 29
ADEA 24 41 29
ADED 24 41 29
1 8 22 38 1 8 28 28
AAAA 120 82 22 AAAA 492 251 5
AAAD −24 36 48 AAAD 264 140 8
AADA −24 36 48 AADD 192 101 5
AADD 120 82 22 ADAA 312 167 11
ADAA 0 49 49 ADAD 192 101 5
ADAD 0 29 29 ADDD 192 101 5
ADDA 0 29 29 AAEE 0 21 21
ADDD 0 49 49 ADEE 0 21 21
1 9 12 229 1 9 13 108 1 9 20 31
AAAA 0 79 79 AAAA 0 59 59 AAAA 240 129 9
AADA 0 79 79 AAAD 0 59 59 AADA 240 129 9
188 Calabi-Yau Data from Gepner Construction
1 10 11 154 1 10 12 82 1 10 13 58 1 10 14 46
AAAA 240 143 23 AAAA 264 147 15 AAAA 288 155 11 AAAA 312 164 8
AAAD 48 83 59 AAAD 120 91 31 AAAD 144 97 25 AAAD 168 100 16
ADAA 48 83 59 AADA 96 85 37 ADAA 144 97 25 AADA 168 100 16
ADAD 240 143 23 AADD 96 63 15 ADAD 288 155 11 AADD 312 164 8
ADAA 120 91 31 ADAA 168 102 18
ADAD 264 147 15 ADAD 120 68 8
ADDA 96 63 15 ADDA 120 68 8
ADDD 96 85 37 ADDD 168 102 18
1 10 16 34 1 10 18 28 1 10 19 26 1 10 22 22
AAAA 360 185 5 AAAA 192 106 10 AAAA 120 76 16 AAAA 480 243 3
AAAD 216 121 13 AAAD 72 58 22 AAAD 24 46 34 AAAD 288 151 7
AADA 192 109 13 AADA 96 70 22 ADAA 24 46 34 AADD 480 243 3
AADD 144 77 5 AADD 72 46 10 ADAD 120 76 16 ADAA 288 153 9
ADAA 216 121 13 ADAA 96 70 22 ADAD 192 99 3
ADAD 360 185 5 ADAD 72 46 10 ADDD 288 153 9
ADDA 144 77 5 ADDA 192 106 10
ADDD 192 109 13 ADDD 72 58 22
AAEA 120 77 17
AAED 72 49 13
ADEA 72 49 13
ADED 120 77 17
1 11 11 76
AAAA 240 143 23
AAAD 240 143 23
1 12 12 40 1 12 13 33 1 12 19 19
AAAA 348 180 6 AAAA 0 43 43 AAAA 288 151 7
AAAD 120 85 25 ADAA 0 43 43 ADAA 288 151 7
AADA 192 109 13
AADD 96 63 15
ADDA 96 63 15
ADDD 96 63 15
1 13 13 28 1 13 18 18
AAAA 288 208 4 AAAA 192 107 11
AAAD 288 208 4 AAAD 96 67 19
AADD 192 107 11
1 14 14 22 1 14 14 22
AAAA 240 127 7 AADD 48 41 17
AAAD 48 55 31 ADDA 240 127 7
AADA 144 85 13 ADDD 48 55 31
1 16 16 16 1 16 16 16 1 16 16 16
AAAA 540 272 2 AAAE 204 111 9 AAEE 108 56 2
AAAD 312 161 5 AADE 144 76 4 ADEE 108 56 2
AADD 216 110 2 ADDE 144 76 4 AEEE 54 35 8
ADDD 216 110 2
B.1 Data 189
2 3 19 418 2 3 20 218 2 3 22 118 2 3 23 98
AAAA 0 119 119 AAAA 120 125 65 AAAA 216 141 33 AAAA 216 141 33
AAAD 0 119 119 AAAD 120 125 65 AAAD 72 77 41 AAAD 216 141 33
AADA 0 71 71 AADA 72 77 41
AADD 0 71 71 AADD 216 141 33
2 3 26 68 2 3 28 58 2 3 34 43 2 3 48 38
AAAA 96 87 39 AAAA 312 173 17 AAAA 0 55 55 AAAA 432 227 11
AAAD 0 47 47 AAAD 312 173 17 AADA 0 55 55 AAAD 240 131 11
AADA 96 87 39 AADA 168 101 17 AADD 432 227 11
AADD 0 47 47 AADD 168 101 17
AAEA 48 47 23
AAED 48 47 23
2 4 11 154 2 4 12 82 2 4 13 58 2 4 14 46
AAAA 72 89 53 AAAA 132 96 30 AAAA 168 104 20 AAAA 192 112 16
AAAD 72 89 53 AAAD 132 96 30 AAAD 168 104 20 AAAD 96 67 19
ADAA 0 71 71 AADA 48 58 34 ADAA 120 86 26 AADA 96 67 19
ADAD 0 71 71 AADD 48 58 34 ADAD 120 86 26 AADD 192 112 16
ADAA 72 76 40 ADAA 120 86 26
ADAD 72 76 40 ADAD 72 56 20
ADDA 0 45 45 ADDA 72 56 20
ADDD 0 45 45 ADDD 120 86 26
2 4 16 34 2 4 18 28 2 4 19 26 2 4 22 22
AAAA 228 126 12 AAAA 108 74 20 AAAA 48 52 28 AAAA 312 164 8
AAAD 228 126 12 AAAD 24 37 25 AAAD 48 52 28 AAAD 192 103 7
AADA 120 73 13 AADA 108 74 20 ADAA 0 41 41 AADD 312 164 8
AADD 120 73 13 AADD 24 37 25 ADAD 0 41 41 ADAA 240 131 11
ADAA 168 100 16 ADAA 48 55 31 ADAD 144 81 9
ADAD 168 100 16 ADAD 0 31 31 ADDD 240 131 11
ADDA 72 56 20 ADDA 48 55 31
ADDD 72 56 20 ADDD 0 31 31
AAEA 84 53 11 AAAE 48 39 15
AAED 84 53 11 AADE 48 39 15
ADEA 72 46 10 ADAE 0 21 21
ADED 72 46 10 ADDE 0 21 21
2 5 8 138 2 5 10 40 2 5 12 26
AAAA 72 80 44 AAAA 72 59 23 AAAA 216 116 8
AAAD 72 80 44 AAAD 0 35 35 AAAD 216 116 8
AADA 0 53 53 AADA 72 59 23 AADA 144 80 8
AADD 0 53 53 AADD 0 35 35 AADD 144 80 8
AAEA 0 35 35
AAED 0 21 21
190 Calabi-Yau Data from Gepner Construction
2 6 7 70 2 6 8 38 2 6 10 22 2 6 14 14
AAAA 144 91 19 AAAA 168 96 12 AAAA 216 114 6 AAAA 288 148 4
AAAD 48 59 35 AAAD 96 67 19 AAAD 144 83 11 AAAD 192 101 5
ADAA 48 59 35 AADA 72 59 23 AADA 120 70 10 AADD 288 148 4
ADAD 144 91 19 AADA 72 49 13 AADD 96 55 7 ADAA 192 104 8
ADAA 96 67 19 ADAA 144 83 11 ADAD 144 75 3
ADAD 168 96 12 ADAD 216 114 6 ADDD 192 104 8
ADDA 72 49 13 ADDA 96 55 7
ADDD 72 59 23 ADDD 120 70 10
AAEA 120 68 8
AAED 72 46 10
ADEA 72 46 10
ADED 120 68 8
2 7 7 34 2 7 10 16
AAAA 144 91 19 AAAA 120 70 10
AAAD 144 91 19 AAAD 72 50 14
AADA 120 70 10
AADD 72 50 14
AAAE 72 44 8
AAEA 48 43 19
AAED 0 21 21
AAEE 0 21 21
2 8 8 18 2 8 10 13
AAAA 228 120 6 AAAA 48 42 18
AAAD 228 120 6 AADA 48 42 18
AADA 144 79 7 ADAA 0 29 29
AADD 144 79 7 ADAA 0 29 29
ADDA 72 49 13 AAEA 0 27 27
ADDD 72 49 13 ADEA 0 21 21
2 10 10 10 2 10 10 10 2 10 10 10
AAAA 324 165 3 AAAE 168 89 5 AAEE 72 46 10
AAAD 204 105 3 AADE 120 61 1 ADEE 72 40 4
AADD 132 69 3 ADDE 72 41 5 AEEE 0 21 21
ADDD 108 57 3
3 3 9 108 3 3 10 58 3 3 12 33
AAAA 80 79 39 AAAA 120 85 25 AAAA 64 59 27
AAAD 80 79 39 AAAD 24 53 41 AADA 64 59 27
AADA 24 53 41
AADD 120 85 25
3 3 13 28 3 3 18 18
AAAA 216 117 9 AAAA 272 143 7
AAAD 216 117 9 AAAD 160 91 11
AAAE 144 75 3 AADD 272 143 7
B.1 Data 191
3 4 6 118 3 4 7 43 3 4 8 28
AAAA 72 69 33 AAAA 96 67 19 AAAA 168 91 7
AAAD 24 49 37 ADAA 96 67 19 AAAD 72 49 13
AADA 24 49 37 AADA 120 69 9
AADD 72 69 33 AADD 96 55 7
ADAA 0 55 55 ADAA 96 67 19
ADAD 0 39 39 ADAD 96 55 7
ADDA 0 39 39 ADDA 96 55 7
ADDD 0 55 55 ADDD 96 55 7
AAAE 48 39 15
AADE 0 21 21
ADAE 0 21 21
ADDE 0 21 21
3 4 10 18 3 4 13 13
AAAA 72 49 13 AAAA 192 103 7
AAAD −24 21 33 ADAA 192 103 7
AADA −24 21 33
AADD 72 49 13
ADAA 0 31 31
ADAD 0 23 23
ADDA 0 23 23
ADDD 0 31 31
3 5 5 68
AAAA 96 71 23
ADAA 96 71 23
3 6 6 18
AAAA 112 63 7
AAAD −16 23 31
AADA 80 51 11
AADD 16 27 19
ADDA 12 63 7
ADDD −16 23 31
3 8 8 8
AAAA 288 145 1
AAAD 192 99 3
AADD 168 85 1
ADDD 168 85 1
192 Calabi-Yau Data from Gepner Construction
4 4 5 40 4 4 6 22 4 4 7 16
AAAA 156 86 8 AAAA 168 90 6 AAAA 192 99 3
AAAD 24 41 29 AAAD 96 61 13 AAAD 96 59 11
ADAA 96 65 17 AADA 96 61 13 ADAA 144 79 7
ADAD 48 47 23 AADD 168 90 6 ADAD 120 68 8
DDAA 48 47 23 ADAA 120 70 10 DDAA 120 68 8
DDAD 48 47 23 ADAD 72 48 12 DDAD 120 68 8
ADDA 72 48 12 AAAE 96 53 5
ADDD 120 70 10 ADAE 120 62 2
DDAA 72 52 16 DDAE 120 62 2
DDAD 24 36 24
DDDA 24 36 24
DDDD 72 52 16
4 4 8 13 4 4 10 10 4 4 10 10
AAAA 108 61 7 AAAA 252 128 2 AAAE 144 74 2
AADA 24 35 23 AAAD 180 94 4 AADE 144 74 2
ADAA 72 48 12 AADD 252 128 2 ADAE 120 62 2
ADDA 48 41 17 ADAA 192 101 5 ADDE 120 62 2
DDAA 48 41 17 ADAD 144 75 3 DDAE 120 62 2
DDDA 48 41 17 ADDD 192 101 5 DDDE 120 62 2
DDAA 144 79 7 AAEE 96 51 3
DDAD 96 57 9 ADEE 48 35 11
DDDD 144 79 7 DDEE 0 21 21
4 5 5 19
AAAA 96 65 17
DAAA 96 65 17
4 6 6 10 4 6 6 10 4 6 6 10 4 6 6 10
AAAA 120 66 6 ADDD 24 32 20 DDDA 48 43 19 DADE 0 21 21
AAAD 24 32 20 DAAA 48 43 19 DDDD 48 35 11 ADDE 96 55 7
AADA 96 55 7 DAAD 48 35 11 AAAE 96 55 7 DDDE 0 21 21
AADD 48 35 11 DADA 48 37 13 AADE 96 51 3
ADDA 120 66 6 DADD 48 33 9 DAAE 0 21 21
4 7 7 7
AAAA 216 112 4
DAAA 216 112 4
5 5 5 12
AAAA 240 122 2
AAAD 240 122 2
6 6 6 6
AAAA 296 149 1
AAAD 208 106 2
AADD 152 77 1
ADDD 112 58 2
DDDD 104 53 1
B.1 Data 193
1 4 4 4 4
AAAAA 204 103 1
AAAAD 168 84 0
AAADD 144 73 1
AADDD 168 84 0
ADDDD 168 84 0
2 2 2 3 18 2 2 2 4 10 2 2 2 6 6
AAAAA 120 65 5 AAAAA 132 69 3 AAAAA 168 86 2
AAAAD 24 29 17 AAAAD 60 39 9 AAAAD 144 73 1
AAADA 72 46 10 AAADD 168 86 2
AAADD 72 40 4
AAAAE 120 61 1
AAADE 0 21 21
2 2 3 3 8 2 2 4 4 4 3 3 3 3 3
AAAAA 48 39 15 AAAAA 108 60 6 AAAAA 200 101 1
AAAAD 0 21 21 AAAAD 96 51 3
AAADD 48 35 11
AADDD 0 21 21
2 2 2 2 2 2
AAAAAA 180 90 0
表の意味をここで記載しておこう:
k1 k2 k3 k4
G1 G2 G3 G4 −χ b2,1 b1,1
例えば上記の場合、4 つの ADE series で Gepner model が構成される。それぞれが G1, G2, G3, G4 の series であ
る。また ki は各 Gi の level を表す。そして χ, b2,1, b1,1 はそれぞれ CY 3-fold の Euler 数、complex moduli の数、
そして Kahler moduli の数である。
ADE series の level k と、ADE minimal model における partition function の関係を挙げておこう [76]:
type level partition function
Ak+1 k ≥ 1 Z =k
X
l=0
χlχl
D2(j+1) k = 4j (j ≥ 1) Z =
j−1X
l=0
|χ2l + χk−2l|2 + 2χk/2χk/2
D2j+1 k = 4j − 2 (j ≥ 2) Z =
k/2X
l=0
|χ2l|2 +
2j−2X
l=0
χ2l+1χk−2l−1
E6 k = 10 Z = |χ0 + χ6|2 + |χ3 + χ7|
2 + |χ4 + χ10|2
E7 k = 16 Z = |χ0 + χ16|2 + |χ4 + χ12|
2 + |χ6 + χ10|2 + |χ8|
2 + (χ2 + χ14)χ8 + χ8(χ2 + χ14)
E8 k = 28 Z = |χ0 + χ10 + χ28|2 + |χ6 + χ12 + χ16 + χ22|
2
Appendix C
Field Theory Realization of SCFT
196 Field Theory Realization of SCFT
C.1 Landau-Ginzburg minimal model
ここの記述は Lerche, Vafa and Warner [72] に大きく依存する。Warner による講義録 [105] も存在するが、それは
chapter II に紹介する。
なお、注意すべき所は、chiral superfield と chiral primary field の区別である。前者は LG model を記述するも
ので、left と right が内包されている。anti-chiral なものは含まれない。一方後者は SCFT を記述する。chairal と
anti-chiral を内包するが、right-moving mode は含まれていない。LG model は non-trivial IR fixed point が存在す
るといわれている。その fixed point 上では、chiral superfield が chiral primary field (の合成) で記述され、
left/right U(1) が等しいものになっている。
まずは Landau-Ginzburg model を記述しよう。LG model の RG flow を考えると、基本的に Kahler poten-
tial (D-term) は irrelevant になるので議論から外しても構わない。また、superpotential term (F-term) には (non-
)perturbative non-renormalization theoremが働くとして、microscopic theoryも、IR fixed point上での modelも同
じ関数であるとする。つまり superpotentialは LG model universality classを定義しているとする。また superpoten-
tial は、原点 (全ての fieldの zero point) で quasi-homogeneous singularity を持つとしよう。その時、superpotential
(の scale 変換則)は次のように定義される:
W (λwiΦi) = λedW (Φi) . (C.1.1)
但し Φi は chiral superfield である。また wi, d は共に整数であるとする1。さらに、標準的な LG model の superpo-
tential は isolated singularity を持つと仮定する。これは ∂iW (Φ) = 0 を満たす点がただ一点 (例えばΦi = 0) に限ら
れるというものである。この仮定がない場合は、対応する CY は non-compact になっているということができる2。
IR fixed point では N = (2, 2) superconformal theory が実現されているとするので、その言葉を用いよう。chiral
superfield Φi は chiral primary field の合成で記述されているとする。従って、chiral primary field の性質を持つ
F-term∫
d2θW は left/right U(1) charge が zero (neutral) である一方、superspace coordinate の measure d2θ は
(−1,−1) の charge を持つ3。よって superpotential W は (1, 1) charge を持つ事になる。今、superpitential W は
chiral superfield Φi の a 次のベキであるとしよう:
W (Φi) ≃ Φai . (C.1.2)
chiral field Φi が (qi, qi) charge を持つ。(LG universality により、Φi は left-right symmetric charge を持つ。これ
は chiral superfield の性質からの帰結である。) よって、
1 = a · qi (C.1.3)
という関係式が成立する。一方で (C.1.1) より、a ·wi = d という関係が成立しているので、Φi の U(1) charge qi は
qi =1
a=
wi
d(C.1.4)
1Lerche, Vafa and Warner [72] では ed とは書かずに d と書いてある。これは先の Poincare polynomial で登場する d と同じなのか否か。
ちなみに、marginal deformation をするなら ed = 1 である。2本文ではそのようなものを扱う。3仮定。
C.1 Landau-Ginzburg minimal model 197
で表される事になる。
LG model の IR fixed point 上では、superpotential が構成する ring R と、N = (2, 2) SCFT の NS-NS (c, c)
primary states のなす (c, c)-ring Rcc とが同型である (R ≃ Rcc) と見なす事が出来る。この、superpotential から構
成される ring R は、chiral superfield とその積で得られるあらゆる単項式 (monomial) から、∂iW (Φ) の zero point
を除いたもので構成される。すなわち、
R =C(Φ)
[∂iW (Φ)](C.1.5)
この ring を構成する element の数 (次元) を µ = dimR と表し、“multiplicity of W” と呼ぶ。
では、LG model の chiral superfield Φi を、“left/right symmetric (c, c) primary states” (qL = qR ≡ qi ≤ 1) と
して考えよう4。local ring (C.1.5) から Poincare polynomial を構成する事が可能である。((p, q) charged primary
states の個数を bd−p,q としてベキで表すので、ring element がわかっていたら構成可能。) Poincare polynomial は
P (t, t) =∏
i
1− (tt)1−qi
1− (tt)qi=∏
i
1− (tt)1−wi/ ed
1− (tt)wi/ ed, (C.1.6)
で与えられる。
Lerche, Vafa and Warner [72] では、(C.1.6) の表示を見易くするために tt を ted に置き換えている。また、
chapter II の (II.5.4) で記述される Poincare polynomial に照し合わせるには tt ≡ tN , pi ≡ wi
ed N , そして
marginal deformation condition d = 1 を用いると良い。但しここではこれらの置き換えはしないまま議論
を進める。
分母だけ見れば、C(Φ) で構成される全ての monomial の partition function そのものであるが、ring の定義で
∂iW (Φ) ∼ 0 を除いている。∂iW (Φ) の U(1) charge は (1− qi, 1− qi) となるので、この除いた部分を分子で表している。
polynomial (C.1.6) からもちろん Witten index を読みとる事が可能である (t = t = −1)。単純にすると解がない
ので、l’Hopital’s rule を用いよう:
Tr(−1)F = P (−1,−1) =∏
i
−(1− qi)(tt)−qi
−qi(tt)qi−1
∣∣∣t=t=−1
=∏
i
( 1
qi− 1). (C.1.7)
local ring element の中で最も charge の大きなものは、polynomial では最もベキの高いものに相当する。その
charge を求めよう。t, t が十分大きな極限をとると、polynomial (C.1.6) は、その “highest charged” states (これを
ρ と名付ける) が dominant になる性質を用いる:
P (t, t) =∏
i
(tt)1−qi
(tt)qi
(tt)−1+qi − 1
(tt)−qi − 1∼ (tt)
Pi(1−2qi) . (C.1.8)
よって “highest charge” は qρ =∑
i(1− 2qi) であり、その state の conformal weight hρ が
hρ =1
2qρ =
∑
i
(1
2− qi
)=∑
i
(1
2− wi
d
)≡ β , (C.1.9)
4正しい仮定なのかどうかまだ良くわからない (2002 10/31)。
198 Field Theory Realization of SCFT
となる。ここで β は chapter II でも紹介した “singularity index” である。ここで、(C.1.8) で登場する “highest
charged” state ρ は本当に highest charge を持つのであろうか。つまり、h = c/6 を満たしているのであろうか。
Poincare polynomial の定義より、最も大きな charge は p = q = c/3 であり、(C.1.8) はその項を引きずり出す作業そ
のものである。しかしこれは、p = q = c/3 なる primary state が ring element として含まれることが仮定となってい
る。この chiral primary state は、その定義より、highest charged state となり、h = c/6 を満たしている。しかしも
し、charge が fractional で、chiral primary の積をどううまくとってもこのような state を構成する事が出来なかっ
た場合、この highest charged state は ring を構成しない。つまり bd,d = 0 となってしまう。よって、h = c/6 を満た
す element が存在しない場合があり得ると考えられる。しかし、実は心配しなくて良い。NS vacuum (hL = hR = 0)
は必ず存在し、Poincare duality (III.4.11) により、highest charged states が必ず (1 つだけ) 存在する事が保証され
る。そしてこの state は、ring element 1 との積でまた chiral primary state になる。従って highest charged state
は、必ず ring element として含まれるので、(C.1.8) で得られる “highest charge” は、本当に highest charge であ
る。よって、highest charged state ρ の持つ関係式 hρ = c/6 より、この系の central charge c は
c = 6β = 6∑
i
(1
2− qi
), (C.1.10)
で与えられる事になる。
なお、この highest charged state ρ は、(II.5.8) にあるように、LG superpotential W (Φ) を用いて
ρ = deti,j
[∂2W (Φ)
∂Φi∂Φj
], (C.1.11)
で与えられる。なお、LG superpotential W が isolated singularity を持つということと、この highest charged state
ρ が non-vanishing であることは同義である [72]。
Appendix D
Topological Sigma Model
200 Topological Sigma Model
Introduction
CY sigma model や LG orbifold model から全ての amplitude を求めるのは容易ではない (CFT としては求められる
が、その field theory realizationは難しいという意味である)。そこである特定の状況に於ける amplitudeを計算するの
に有用な模型を導入する。例えば topological sigma model である。supercharge を手で twist することで BRST-like
な charge を構成するのであるが、twist には 2 種類存在し (A-twist/B-twist)、それぞれの作用によって A-model,
B-model と呼ばれる topological sigma model が構成される [111]。これらはそれぞれ CY manifold の (complexfied)
Kahler class、complex structure にのみ依存するものである。A-model は worldsheet instanton の寄与が含まれるた
め amplitude の計算は厄介であるが、B-model には worldsheet instanton の寄与がないため、zero modes で構成さ
れる amplitude の計算は、純粋に CY geometry のみを用いる事で理解できる。但しここでは amplitude の計算まで
深入りせず、単に supersymmetry から BRST への変換則を与えるに留める。amplitude などの議論は Witten [111]
を参照。
ここでは前半で topological nonlinear sigma model を扱い、後半で topological gauged linear sigma model の議
論を展開する。Witten [111] では supersymmetric NLSM を、on-shell の下で twist して、A-model, B-model を構成
したが (A-model 自体はもっと以前に、やはり Witten によって構成された [109])、ここでは GLSM の段階で twist
を施してみる。まず [111] の簡単な復習を行った後、GLSM の twist を行おう。
D.1 Supersymmetric nonlinear sigma models
まずは Witten [111] に従い、D = 2, N = 2 の Euclidean supersymmetric nonlinear sigma model Lagrangian を記
述しよう。但し superfield ではなく、component field のみを用いる1:
L = 2r
1
2gIJ∂φ
I∂φJ + iψj−Dzψ
j−gij + iψj
+Dzψi+gij +Rijkl ψ
i+ψ
j+ψ
k−ψ
l−
. (D.1.1)
但し ∂ = ∂z, ∂ = ∂z である。また r は gauged linear sigma model での FI parameter に相当する scale parameter
である。index I は target space index で、Riemann geometry の時には I, J, · · · を用いている。Kahler (complex)
geometry の時には i, j, · · · , i, j, · · · を用いる。Dz, Dz は covariant derivative であり、
Dzψi+ = ∂ψi
+ + (∂φj)Γijkψ
k+ , (D.1.2)
などで定義される。Γijk は affine connection、また Rijkl は Riemann tensor である。
supersymmetry parameter α を用いて、supersymmetry 変換を施そう:
δφi = iα−ψi+ + iα+ψ
i− , δφi = iα−ψ
i+ + iα+ψ
i− , (D.1.3a)
δψi+ = −α−∂φ
i − iα+ψj−Γi
jmψm+ , δψi
+ = −α−∂φi − iα+ψ
j−Γi
jmψm+ , (D.1.3b)
δψi− = −α+∂φ
i − iα−ψj+Γi
jmψm− , δψi
− = −α+∂φi − iα−ψ
j+Γi
jmψm+ . (D.1.3c)
1この subsection だけ他の section とは notation が若干異なり Witten [111] の記述を採用しているが、本文のそれに合わせるためには重大
な変更を必要としないのでこのままにする。
D.2 Definition of topological twist 201
D.2 Definition of topological twist
元々、D = 2, N = (2, 2) theory の topological twist の “定義” とは、Euclidean energy-momentum tensor Tmn に、
U(1) current Jm を加えることで、Lorentz group を U(1)E から U(1)E ×U(1)R にとることである2。ここで U(1)R
とは、SCFT algebra に入っている left/right U(1) を適当に組み合わせたものである。A-twist の時は JR = JL + JR
で与え、B-twist の時は JR = −JL + JR で与える。そして twisted energy-momentum tensor を
T twistedmn = Tmn +
1
4
(ǫm
k∂kJRn + ǫn
k∂kJRm
), (D.2.1)
とする3。この帰結として、Lorentz generator T が
T → T ′ = T +1
2JR (D.2.2)
のように変更を受ける4。これにより、spin 1/2の fermionが変更を受けて (U(1)E の spin表現だったものが U(1)E×U(1)R のある表現になって)、spin zero fermion になったり spin one fermion になったりする。
D.3 Topological twist
ここでは簡単に A-twist と B-twist の方法を述べよう。そのためには + twist と − twist という概念を導入する必要
がある。まず、fermion に on-shell condition を課すことで sigma model の mode を left/right に分解する。ここの
場合 ψ+, ψ+ などが left moving mode (ψ+ = ψ+(z) etc.) であり、ψ−, ψ− などが right moving mode (ψ− = ψ−(z)
etc.) である5。on-shell fermion ψ や superspace coordinates θ、global supersymmetry parameter α は、それぞれ
数学的には以下のものとして導入される。
left-moving mode:
ψi+ : section of K1/2 ⊗ Φ∗(T (1,0)X) ψi
+ : section of K1/2 ⊗ Φ∗(T (0,1)X)
α+ = −α− : section of K−1/2 α+ = α− : section of K−1/2
θ+ = −θ− : section of K−1/2 θ+ = −θ− : section of K−1/2 , (D.3.1a)
right-moving mode:
ψi− : section of K1/2 ⊗ Φ∗(T (1,0)X) ψi
− : section of K1/2 ⊗ Φ∗(T (0,1)X)
α− = α+ : section of K−1/2 α− = α+ : section of K−1/2
θ− = θ+ : section of K−1/2 θ− = θ+ : section of K−1/2 . (D.3.1b)
ここで T (1,0)X, T (0,1)X は それぞれ、GLSM の target space X の holomorphic, anti-holomorphic tangent bundle
である。つまりこれらはそれぞれ chiral/anti-chiral を表す。一方で K∗, K∗ は canonical/anti-canonical line bundle2U(1)E の index E は worldsheet が Euclidean という意味である。3notation は Hori-Vafa [54] である。したがって自分の notation でもあると思ってよい。4energy-momentum tensor Tmn は translation generator であり、Lorentz generator ではないのだが、topological twist は (D.2.1) で
行うとある。ここがまだ理解できない部分である。52-dimensional Minkowski metric で与えられている worldsheet coordinates xm を Euclidean metric に変換して complex coordinates
z, z にする方法は appendix C を参照。
202 Topological Sigma Model
on Σ (worldsheet) を表す。つまりこれらは left/right-moving mode を表す。ここで注意すべきことは、superspace
coordinates θ や supersymmetry parameter α は、fermion ψ と θψ = θ+ψ+ + θ−ψ− や αψ = α+ψ+ + α−ψ− の様
に組んで Lorentz scalar を構成するため、canonical bundle K のベキが逆で定義されることである。そのため、θ, α
の left/right-moving mode の表示が fermion とは逆になっている (θ, α は global parameter であるので、厳密な意
味では left/right-moving という呼び方は間違いであるが、fermion の Lorentz 変換の方向と逆である意味を込めてこ
う呼ぶ方が都合がよい。)。chiral/anti-chiral の表示は fermion と同じでよい。
これら ψ, θ, α はすべて fermionic spinor on Σ (spinor bundle K±1/2, K±1/2) であるが、twist をすることで、
あるものは scalar on Σ に、あるものは 1-form on Σ (つまり (anti-)canonical bundle K±1, K±1) になる。その方法
が 2 種類ある。以下に挙げる + twist と − twist がそれである:
+ twist:
ψi+ : K1/2 ⊗ Φ∗(T (1,0)X) −→ Φ∗(T (1,0)X)
ψi+ : K1/2 ⊗ Φ∗(T (0,1)X) −→ K1 ⊗ Φ∗(T (0,1)X)
ψi− : K1/2 ⊗ Φ∗(T (1,0)X) −→ Φ∗(T (1,0)X)
ψi− : K1/2 ⊗ Φ∗(T (0,1)X) −→ K1 ⊗ Φ∗(T (0,1)X)
(D.3.2a)
superspace coodinates θ や supersymmetry parameter α も同様に twist される。但し Lorentz invariance を保つた
め、twist の方向は field ψ とは逆である:
α− : K−1/2 −→ 1
α− : K−1/2 −→ K−1
α+ : K−1/2 −→ 1
α+ : K−1/2 −→ K−1
θ− : K−1/2 −→ 1
θ− : K−1/2 −→ K−1
θ+ : K−1/2 −→ 1
θ+ : K−1/2 −→ K−1
(D.3.2b)
− twist:
ψi+ : K1/2 ⊗ Φ∗(T (1,0)X) −→ K1 ⊗ Φ∗(T (1,0)X)
ψi+ : K1/2 ⊗ Φ∗(T (0,1)X) −→ Φ∗(T (0,1)X)
ψi− : K1/2 ⊗ Φ∗(T (1,0)X) −→ K1 ⊗ Φ∗(T (1,0)X)
ψi− : K1/2 ⊗ Φ∗(T (0,1)X) −→ Φ∗(T (0,1)X)
(D.3.3a)
− twist での supersymmetry parameter の twist も書いておこう:
α− : K−1/2 −→ K−1
α− : K−1/2 −→ 1
α+ : K−1/2 −→ K−1
α+ : K−1/2 −→ 1
θ− : K−1/2 −→ K−1
θ− : K−1/2 −→ 1
θ+ : K−1/2 −→ K−1
θ+ : K−1/2 −→ 1
(D.3.3b)
これらを用いて A-twist, B-twist を定義する。
topological A-twist
A-twistは 2通りの方法がある。left/right-moving modeをそれぞれ + twist, − twistする (+,−) twistと、その逆の
(−,+) twist がある。何方にせよ、left-moving mode と right-moving mode の twist を互いに逆にするのが A-twist
である。そこでそれぞれについて議論しておこう。
D.3 Topological twist 203
(+,−) twist
(+,−)-twist とは、left moving mode ψi+, ψi
+ については + twist を行い、right moving mode ψi−, ψi
− について
は− twist を行うことである。それに伴い、superspace coodinates θ や supersymmetry parameter α、 supercharge
Q も twist される。supercharge は
Q− ∼∂
∂θ−+ · · · ∼ ∂
∂θ++ · · · , Q+ ∼
∂
∂θ++ · · · ∼ ∂
∂θ−+ · · · , (D.3.4)
Q− ∼∂
∂θ−+ · · · ∼ ∂
∂θ++ · · · , Q+ ∼
∂
∂θ++ · · · ∼ ∂
∂θ−+ · · · , (D.3.5)
で記述されるので、Q+, Q− が spin zero charge となる。この線形和 Q+ +Q− ≡ QB (or = Qac) をとって、新たな
charge とする。この QB は nilpotency の性質を持つので、BRST charge とみなされる。まとめると、
ψi+ , ψi
− , θ− = −θ+ , θ+ = θ− , α− = −α+ , α+ = α− , QB = Q+ +Q− , (D.3.6)
が topological (+,−) twist によって spin zero となる。
(−,+) twist
(−,+)-twist とは、left-moving mode ψi+, ψi
+ については − twist を行い、right moving mode ψi−, ψi
− について
は+ twist を行うことである。それに伴い、superspace coodinates θ や supersymmetry parameter α、 supercharge
Q も twist される。supercharge は Q−, Q+ が spin zero charge となる。この線形和 Q− + Q+ ≡ QB (or = Qac)
をとって、新たな charge とする。この QB は nilpotency の性質を持つので、BRST charge とみなされる。まとめ
ると、
ψi− , ψi
+ , θ+ = θ− , θ− = −θ+ , α+ = α− , α− = −α+ , QB = Q− +Q+ , (D.3.7)
が topological (−,+) twist によって spin zero となる。
topological B-twist
B-twistにも 2通りの方法がある。left/right-moving modeをそれぞれ − twistする (−,−) twistと、その逆の (+,+)
twist がある。何方にせよ、left-moving mode と right-moving mode の twist を同じにするのが B-twist である。そ
こでそれぞれについて議論しておこう。
(−,−) twist
(−,−)-twist とは、left moving mode ψi+, ψi
+ と right moving mode ψi−, ψi
− を共に − twist することである。こ
れに伴い、superspace coodinates θ、supersymmetry parameter α、supercharge Q も twist される。その結果 spin
zero になるものは
ψi+ , ψi
− , θ− = −θ+ , θ+ = θ− , α− , α+ , QB = Q+ +Q− ,
である。
(+,+) twist
204 Topological Sigma Model
(−,−)-twist とは、left moving mode ψi+, ψi
+ と right moving mode ψi−, ψi
− を共に − twist することである。こ
れに伴い、superspace coodinates θ、supersymmetry parameter α、supercharge Q も twist される。その結果 spin
zero になるものは
ψi+ , ψi
− , θ− = −θ+ , θ+ = θ− , α− , α+ , QB = Q+ +Q− ,
である。これにより、(I.A.14)を参考にして、BRST transformationは以下のようにして与えられる。ここでも α± = 0,
α± ≡ α と置くと議論が簡単になる。
D.4 Topological gauged linear sigma models
Gauged linear sigma model の場合も nonlinear sigma model の topological twist と殆ど同じ議論ができるが、gauge
symmetry が存在するので chiral multiplet の twist の方向が場合によって変化することがある。つまりある chiral
multiplet の left-moving fermion が spin zero fermion に twist されると同時に、他の chiral multiplet の left-moving
fermion は spin one の worldsheet one-form に twist される事がある。その場合でも、最初模型に導入されていた
supersymmetry parameter α は 1 種類であるから、supersymmetry parameter α の twist の方向はどの sector で見
ても同じである必要がある。
gauged linear sigma modelの場合は (twisted) superpotentialを用いた記述をする。この (twisted) superpotential
が存在すると appendix I.Bの議論から left/right U(1) chargeを定義することが可能である。つまり modified spinを
R-charge を用いて簡単に与えることができる。但し、Lorentz spin に R-charge を単に付加しただけでは、(半) 整数
spin を与えることができない模型も存在する。そのため、U(1) gauge charge を新たに付加することがある。付加出
来る場合とは、U(1) gauge current J が BRST exact に書ける場合に限られる。そして最終的には spin 1/2 fermion
を spin zero fermion や spin one fermion に twist することができる。しかしこうなると fermion ψ± の topological
twist の方向を見て、model の twist の種類を分類するのは意味を成さなくなる。分類をするならば、supersymmetry
parameter α± や superspace coordinates θ±、そして supersymmetry generator Q± の topological twist の方向で
分類すべきであるということになる。その分類はこれまでの議論に合わせて (D.3.2), (D.3.3) と同じであるとする。
念のためここでもう一度列挙しておこう。但し fermion ψ± についてはもはや記述しても意味を成さない:
+ twist:
α− : K−1/2 −→ K0
α− : K−1/2 −→ K−1
α+ : K−1/2 −→ K0
α+ : K−1/2 −→ K−1
θ− : K−1/2 −→ K0
θ− : K−1/2 −→ K−1
θ+ : K−1/2 −→ K0
θ+ : K−1/2 −→ K−1
(D.4.1a)
− twist:
α− : K−1/2 −→ K−1
α− : K−1/2 −→ K0
α+ : K−1/2 −→ K−1
α+ : K−1/2 −→ K0
θ− : K−1/2 −→ K−1
θ− : K−1/2 −→ K0
θ+ : K−1/2 −→ K−1
θ+ : K−1/2 −→ K0
(D.4.1b)
D.4 Topological gauged linear sigma models 205
topological A-twist
それでは left/right moving mode を適当に topological twist した物を議論しよう。
(+,−) twist:
topological twist の rule として (D.4.1a), (D.4.1b) を用いる。(+,−) twist の時には supersymmetry parameter
のうち α− と α+ が spin zero に、α+ と α− が spin one に twist される。component field に関しては、模型によっ
て gauge charge が付加された U(1) charge を用いて twist するため、それが spin zero になっているかなどは一般
的な議論が展開できない6。一つの模型の中で、ある chiral multiplet については left-moving chiral fermion ψ+,i が
spin zero に twist される一方、他の chiral multiplet について同じ left-moving chiral fermion ψ+,j が spin one に
twist されている場合が存在する場合がある。後で具体例として挙げる GLSM for CP4[5] などがこの場合に相当す
る。このため、一見 supersymmetry parameter の twist の方向が矛盾を起こすと思うかもしれないが、それについて
は問題がない。単に BRST 変換則を全ての field について与えることは機械的にできる。注意すべきことは、「(+,−)
twist だというだけで、全ての left-moving chiral fermion ψ+ が spin zero fermion になっているとは限らない。twist
の分類は supercharge Q± の twist に起因するとみなすべきである。」ということである。
この (+,−) twist の場合は、supersymmetry parameter α−, α+ が spin zero になる場合である。別の言い方を
すれば、supercharge Q+, Q− が spin zero となり BRST charge を構成する場合である。
(I.A.14) から、twisted chiral multiplet (gauge multiplet) の BRST 変換が以下の様に与えられる。但し BRST
parameter を α− = α+ ≡ α のように簡略化する:
δB v0 = iα(λ+ + λ−) , δB v1 = iα(λ+ − λ−) ,
δB σ = −i√
2α(λ− + λ+) , δB σ = 0 ,
δBD = −α[− 2∂λ+ − 2∂λ−
],
δB λ+ = 2√
2α∂σ , δB λ− = αi(D − F12) ,
δB λ+ = −αi(D − F12
), δB λ− = −2
√2α∂σ . (D.4.2a)
また chiral multiplet についても BRST 変換を書き下すことができる:
δB φ = −√
2αψ+ , δB φ = −√
2αψ− ,
δB ψ+ = −2Qφσ α , δB ψ− =√
2α[2iDzφ+ F
],
δB ψ+ =√
2α[− 2iDzφ+ F
], δB ψ− = 2Qφσ α ,
δB F = α[2i√
2Dzψ+ + 2Q(σψ− − iφλ−)],
δB F = α[− 2i√
2Dzψ− − 2Q(σψ+ − iφλ+)]. (D.4.2b)
但しここではWitten [111] にあるような fermion の記号の書き換えは行わない。[111] は NLSM の twist であり、基
本的に target space は curved space (Kahler manifold) であるので、index i の上下には意味があるが、GLSM では
target space は flat (low energy limit ではじめて curved space となる) なので、index の上下の区別は存在しない。
6少なくとも 2003 2/16 の時点では一般的な議論は知らない。
206 Topological Sigma Model
ちなみに、BRST charge との関係は次の様になっている:
δBW = −iαQB,W . (D.4.3)
なお、これらは on-shell 上で BRST nilpotent δ2B = 0 であり、cohomology をなす (BRST cohomology)。
topological (+,−) twist を行うと、先に書いたように、superchrage のうち Q+, Q− が spin zero operator に変
化する。ここで、GLSM Lagrangian を思い起こそう。一般的には
LGLSM =
∫d4θK +
(∫d2θW + (h.c.)
)+(∫
d2θ W + (h.c.)), (D.4.4a)
d2θ = −1
2dθ+dθ− , d2θ =
1
2dθ+dθ− , (D.4.4b)
d2θ = −1
2dθ+dθ− , d2θ = −1
2dθ−dθ+ , (D.4.4c)
で与えられている。(+,−) twist を施すと、この 4 つの “superpotential term” すべてが理論に non-trivial に寄与す
るのではなく、殆どが BRST exact の形で書くことができる。
fermionic measure は dθ± = ∂/∂θ± のように微分演算と同等であるので、supercharge に読み換えることができ
る。よって (+,−) twist しているときには、
d2θ ∼ Q+Q− , d2θ ∼ Q+Q− , d2θ ∼ Q+Q− , d2θ ∼ Q−Q+ , (D.4.5)
と書き換えられる。これにより、Kahler potential term はもちろんのこと、W , W , W term は全て BRST exact な
QB, · · · の形になって irrelevant term となる。唯一 relevant なのは
∫d2θ W (D.4.6)
だけである。GLSM では、この twisted superpotential は complexified FI parameter τ = r − iθ と twisted chiral
multiplet Σ を用いて W = −τΣ と書き表される。よって、(+,−) twisted GLSM は、complexified FI parameter τ
に anti-holomorphic にのみ依存し、superpotential には依存しない。
(−,+) twist:
(+,−) twist の時と同様に、supersymmetry parameter α もしくは supercharge Q± の twist の方向でこの twist
を定義する。fieldについての twistは modelによってまちまちである事に注意しておく。しかし全ての fieldの BRST
変換則は (I.A.14) から機械的に得ることができる。ここの場合は、α+ と α− が spin zero になる場合である。つま
り supercharge Q− と Q+ が spin zero になり BRST charge を構成する場合である。
まず、twisted chiral multiplet (gauge multiplet) の BRST 変換が以下の様に与えられる。但し BRST parameter
を α+ = α− ≡ α のように簡略化する:
δB v0 = iα(λ− + λ+) , δB v1 = −iα(λ− − λ+) ,
δB σ = 0 , δB σ = −i√
2α(λ+ + λ−) ,
δBD = α[− 2∂λ+ − 2∂λ−
],
δB λ+ = αi(D + F12) , δB λ− = −2√
2α∂σ ,
D.4 Topological gauged linear sigma models 207
δB λ+ = 2√
2α∂σ , δB λ− = −αi(D + F12) . (D.4.7a)
また chiral multiplet についても BRST 変換を書き下しておこう:
δB φ =√
2αψ− , δB φ =√
2αψ+ ,
δB ψ+ =√
2α[2iDzφ+ F
], δB ψ− = 2Qαφσ ,
δB ψ+ = −2Qαφσ , δB ψ− =√
2α[− 2iDzφ+ F
],
δB F = α[− 2√
2iDzψ− + 2Q(σψ+ + iφλ+)],
δB F = α[2√
2iDzψ+ − 2Q(σψ− + iφλ−)]. (D.4.7b)
(−,+) twistを行うと、superchrageのうちQ−, Q+ が spin zero operatorに変化する。これにより、今度は GLSM
Lagrangian の中で
∫d2θ W (D.4.8)
だけでが relevant となる。GLSM では、この twisted superpotential は complexified FI parameter τ = r − iθ とtwisted chiral multiplet Σ を用いて W = −τΣ と書き表される。よって、(−,+) twisted GLSM は、complexified
FI parameter τ に holomorphic にのみ依存し、superpotential には依存しない。
A-twisted GLSM は、(+,−) twist にせよ、(−,+) twist にせよ、target space の Kahler class にのみ依存し、
worldsheet と target space の complex structure に依存しない。ちなみに、Witten [111] で記述される topological
A-model は NLSM の twist であるが、そこで登場する parameter t とは、FI parameter r のことである。これによ
ると、GLSM も FI parameter r には依存しないことになる。従って Theta angle にのみ依存することになるが、こ
れは A-model が worldsheet instantons の寄与を受けることにちょうど対応している。
topological B-twist
(−,−) twist:
topological A-twist での議論と同様、(D.4.1a), (D.4.1b) から twist の分類を行っている。field の twist はmodel
によって異なる。ここの場合は supersymmetry parameter α± が spin zero になる場合である。別の言い方をすれば、
supercharge Q± が spin zero generator となり、BRST charge となる場合である。これにより、(I.A.14) を参考にし
て、BRST 変換を以下のように与える。ここでも α± = 0, α± ≡ α と置くと議論が簡単になる:
δB v0 = iα(λ+ + λ−) , δB v1 = iα(λ+ − λ−) ,
δB σ = −i√
2αλ− , δB σ = −i√
2αλ+ ,
δBD = −2α[− ∂λ+ + ∂λ−
],
δB λ+ = 0 , δB λ− = 0 ,
δB λ+ = α[− iD + 2
√2∂σ + iF12
], δB λ− = α
[− iD + 2
√2∂σ − iF12
]. (D.4.9a)
208 Topological Sigma Model
chiral multiplet についても BRST 変換を与えておこう:
δB φ = 0 , δB φ = −√
2α(ψ− − ψ+) ,
δB ψ+ = 2α[√
2iDzφ−Qφσ], δB ψ− = 2α
[√2iDzφ+Qφσ
],
δB ψ+ =√
2αF , δB ψ− =√
2αF ,
δB F = α[− 2√
2i−Dzψ+ +Dzψ−
+ 2Q(σψ− + σψ+) + iφ(λ+ − λ−)
], δB F = 0 . (D.4.9b)
但しここでもWitten [111] にあるような fermion の記号の書き換えは行わない。[111] は NLSM の twist であり、基
本的に target space は curved space (Kahler manifold) であるので、index i の上下には意味があるが、GLSM では
target space は flat (low energy limit ではじめて curved space となる) なので、index の上下の区別は存在しない。
特に B-model では index を下げた θi = gij(ψj+ − ψj
−) が登場するが、GLSM では gij はまだ意味を持っていないか
らである。なお、この (−,−) twist においても nilpotency δ2B = 0 を満たすので、BRST は cohomology を定義する
ことができる。
(−,−) twist を施すと、GLSM Lagrangian の中で W term 以外は全て BRST exact であることがわかる。
(+,+) twist:
(D.4.1a), (D.4.1b) から twist の分類を行っている。field の twist は model によって異なる。ここの場合は su-
persymmetry parameter α± が spin zero になる場合である。別の言い方をすれば、supercharge Q± が spin zero
generator となり、BRST charge となる場合である。これにより、(I.A.14) を参考にして、BRST 変換を以下のよう
に与える。ここでも α± = 0, α± ≡ α と置くと議論が簡単になる:
δB v0 = iα(λ+ + λ−) , δB v1 = iα(λ+ − λ−) ,
δB σ = −i√
2αλ+ , δB σ = −i√
2αλ− ,
δBD = 2α[− ∂λ+ + ∂λ−
],
δB λ+ = α[iD + iF12 + 2
√2∂σ
], δB λ− = α
[iD − iF12 − 2
√2∂σ
],
δB λ+ = 0 , δB λ− = 0 . (D.4.10a)
chiral multiplet についても BRST 変換を与えておこう:
δB φ =√
2α(ψ− − ψ+) , δB φ = 0 ,
δB ψ+ =√
2αF , δB ψ− =√
2αF ,
δB ψ+ = −2α[√
2iDzφ+Qφσ], δB ψ− = 2α
[−√
2iDzφ+Qφσ],
δB F = 0 , δB F = −α[2√
2i−Dzψ+ +Dzψ−
+ 2Q(σψ− + σψ+)− iφ(λ+ − λ−)
]. (D.4.10b)
(+,+) twist を施すと、GLSM Lagrangian の中で W term 以外は全て BRST exact であることがわかる。
これにより、topological B-modelは、(−,−) twistにせよ (+,+) twistにせよ、target spaceの complex structure
にのみ依存することになる。
D.5 Example: Topological GLSM for quintics 209
D.5 Example: Topological GLSM for quintics
Silverstein-Witten [99] の記法にあるように、GLSM level での topological twist では何故 U(1) gauge charge の符号
が逆だと twist も逆 ((−,+) twist が (+,−) twist に) なるのか、その考察を具体的な模型で行う。具体例を 2 つ挙げ
る。まずは GLSM for quintics (CP4[5]) であり、ふたつめに GLSM for resolved conifold (O(−1)⊕O(−1)→ CP1)
を挙げる。
まずは topological twist そのものの定義を振り返る。そして Silverstein-Witten [99] ではどのように modify し
ているのかを考察しよう。
chiral superfield Si のR-chargeを (qVs , q
As )、chiral superfield P の R-chargeを (qV
p , qAp )、twisted chiral superfield
Σ の R-charge を (qVσ , q
Aσ ) と置く。このとき、先程と同様にして、component fields それぞれの R-charge (QV , QA)、
そして U(1) left/right current の charge (QL, QR) は次のようになる:
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Si qVs qA
s12 (qV
s − qAs ) 1
2 (qVs + qA
s )
si qVs qA
s12 (qV
s − qAs ) 1
2 (qVs + qA
s )
ψ+,i qVs − 1 qA
s − 1 12 (qV
s − qAs ) 1
2 (qVs + qA
s − 2)
ψ−,i qVs − 1 qA
s + 1 12 (qV
s − qAs − 2) 1
2 (qVs + qA
s )
Fs qVs − 2 qA
s12 (qV
s − qAs − 2) 1
2 (qVs + qA
s − 2)
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
P qVp qA
p12 (qV
p − qAp ) 1
2 (qVp + qA
p )
p qVp qA
p12 (qV
p − qAp ) 1
2 (qVp + qA
p )
ψ+,0 qVp − 1 qA
p − 1 12 (qV
p − qAp ) 1
2 (qVp + qA
p − 2)
ψ−,0 qVp − 1 qA
p + 1 12 (qV
p − qAp − 2) 1
2 (qVp + qA
p )
Fp qVp − 2 qA
p12 (qV
p − qAp − 2) 1
2 (qVp + qA
p − 2)
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Σ qVσ qA
σ12 (qV
σ − qAσ ) 1
2 (qVσ + qA
σ )
σ qVσ qA
σ12 (qV
σ − qAσ ) 1
2 (qVσ + qA
σ )
λ+ qVσ − 1 qA
σ − 1 12 (qV
σ − qAσ ) 1
2 (qVσ + qA
σ − 2)
λ− qVσ + 1 qA
σ − 1 12 (qV
σ − qAσ + 2) 1
2 (qVσ + qA
σ )
D ± F12 qVσ qA
σ − 2 12 (qV
σ − qAσ + 2) 1
2 (qVσ + qA
σ − 2)
これから R-charge を決定する作業に入る。Kahler potential は R-invariant であるのは自明である。non-trivial
な部分は superpotential W 及び twisted superpotential W である。それぞれ
W = P
5∑
i=1
S5i , W = −τΣ , (D.5.1)
210 Topological Sigma Model
である。model が R-invariant であるためには、superpotential W のR-charge が (2, 0)、twisted superpotential W
の R-charge が (0, 2) でなければならない [54]。これより、
2 = qVp + 5qV
s , 0 = qAp + 5qA
s , (D.5.2a)
0 = qVσ , 2 = qA
σ , (D.5.2b)
という束縛条件が得られる。GLSMだけでは chiral superfieldのR-charge は残念ながら決定できない。決定させるた
めには low energy theory で得られる LG theory の考慮も必要である。FI parameter r が負の領域では、LG theory
は superpotential が W ′ =∑
i S5i で与えられる。これも当然 R-invariant である必要があるので、ここで新たに
2 = 5qVs , 0 = 5qA
s , (D.5.2c)
という条件が得られる。これらによって、全ての R-charge が決定できる:
qVs =
2
5, qA
s = 0 , qVp = 0 , qA
p = 0 , qVσ = 0 , qA
σ = 2 . (D.5.3)
この値を代入してもう一度一覧表を作成しよう:
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Si25 0 1
515
si25 0 1
515
ψ+,i − 35 −1 1
5 − 45
ψ−,i − 35 1 − 4
515
Fs − 85 0 − 4
5 − 45
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
P 0 0 0 0
p 0 0 0 0
ψ+,0 −1 −1 0 −1
ψ−,0 −1 1 −1 0
Fp −2 0 −1 −1
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Σ 0 2 −1 1
σ 0 2 −1 1
λ+ −1 1 −1 0
λ− 1 1 0 1
D ± F12 0 0 0 0
この結果は Silverstein-Witten [99] をきちんと再現している。
D.5 Example: Topological GLSM for quintics 211
さてここでようやく準備が整ったのだが、Silverstein-Witten [99]は、この topological twistの定義 (D.2.1), (I.B.9)
を “for convenience as indicated in the footnote following Eq. (3.13)” といって変更している7。具体的には U(1)R
current の付加に伴う spin (ここでは A-twist) が次のように与えられる8:
1
2JR = −1
2(JL + JR) +
1
5Qg = −1
2JV +
1
5Qg . (D.5.4)
ここで Qg は U(1) gauge charge である。この一見不思議に見える再定義は、実はとても自然なものであるといえる。
currentの定義式 (I.B.9)は、conformal な NLSM level での定義であった。即ち field content は chiral multiplet のみ
で、twisted chiral multiplet (gauge multiplet) は low energy limit では積分されている。即ち U(1) gauge symmetry
は自発的に破れ、Higgs mechanism によってなくなっている。しかし今考察しているのは GLSM である。gauge
symmetry が存在している。従って current の定義は NLSM のそれとは異なっているはずで、gauge symmetry の産
物である U(1) gauge current が current JR に入っているべきである。従って charge の再定義が (D.5.4) であるの
はもっともであろうと思われる。1/5 があるのは、model の maximal gauge charge が P の Qg = −5 であるので、
その値で割っているということであろう9。これより、GLSM for CP4[5] での topological twist は
T twistedµν = Tµν +
1
4
(ǫµ
λ∂λJRν + ǫν
λ∂λJRµ
), (D.5.5a)
1
2JR =
− 1
2 (JL + JR) + 15Jg = − 1
2JV + 15Jg A-twist
− 12 (−JL + JR) + 1
5Jg = − 12JA + 1
5Jg B-twist, (D.5.5b)
で定義されると思われる (Jg は U(1) gauge current)。また B-twist についても gauge current が BRST exact に
なっているかどうかは未確認である [2003 2/18]。
ちなみに、Silverstein-Witten [99] での twist は、Witten [111] に合った言い方をすれば、(+,−) twist である。
Silverstein-Witten method
最初に twist をする前の段階での chiral superfield (Si, P ), twisted chiral superfield Σ に属する全ての component
fieldの spin T U(1)E Lorentz group)、RV -chargeと RA-charge (QV , QA)、それから派生する left/right U(1) charge
(QL, QR)、U(1) gauge charge Qg を列挙する:
superfield components T QV QA QL QR Qg
Si si 0 25 0 1
515 1
ψ+,i − 12 − 3
5 −1 15 − 4
5 1
ψ−,i12 − 3
5 1 − 45
15 1
Fs 0 − 85 0 − 4
5 − 45 1
7理由は、gauge current J は Q+, · · · の形で書ける、つまり BRST exact だから、という事らしい。しかしまだよく理解していない [2003
2/18]。8spin と書いているが、current の charge と思ってもいいだろう。9実はもっと深遠である。subsection D.6 にある GLSM for resolved conifold の議論を参照。
212 Topological Sigma Model
superfield components T QV QA QL QR Qg
P p 0 0 0 0 0 −5
ψ+,0 − 12 −1 −1 0 −1 −5
ψ−,012 −1 1 −1 0 −5
Fp 0 −2 0 −1 −1 −5
superfield components T QV QA QL QR Qg
Σ σ 0 0 2 −1 1 0
λ+ − 12 −1 1 −1 0 0
λ−12 1 1 0 1 0
D ± F12 (0, 1) 0 0 0 0 0
ここで、Lorentz group U(1)E の spin T の assign は
(ψ+, ψ−
)=(− 1
2,
1
2
)chiral superfield , (D.5.6a)
(ψ+, ψ−
)=(− 1
2,
1
2
)anti-chiral superfield , (D.5.6b)
である10。また U(1) R-symmetry と left/right U(1) current の関係を次のようにしている:
JV = JL + JR , JA = −JL + JR . (D.5.7)
topological twist を行う。Lorentz generator T を
T → T ′ = T +1
2JR , (D.5.8)
で再定義する。ここで JR は twist の方法に依存する:
1
2JR =
− 12 (QL +QR) + 1
5Qg (+,−) twist
+ 12 (QL +QR)− 1
5Qg (−,+) twist
− 12 (−QL +QR) (+,+) twist
+ 12 (−QL +QR) (−,−) twist
(D.5.9)
A-twist については U(1) gauge current を付加させると spin が上手く整数となるが、その定義では B-twist は分数
になってしまう。U(1) gauge current は A-twist の BRST charge Q+ を用いて J ∼ Q+, · · · と記述できるらしいが [99]、B-twist ではこうならない (BRST exact ではない)のであろうか。そうであれば、B-twist ではこの gauge
current の付加は適切ではないというのは納得できる。どちらにせよ、一度 gauge current Jg を Noether current と
して構成する必要がある。
10この逆の定義でも構わないが Witten [109] など一連の文献に合せてある。left/right で spin が逆向きであることはすぐ納得できよう。
chiral/anti-chiral で同じ向きであるのは、spin は worldsheet spin であり、target の holomorphicity とは関係ないからである。
D.5 Example: Topological GLSM for quintics 213
では各 twist において component fields がどのように spin の変更を受けるのか具体的に見よう:
si ψ+,i ψ−,i Fi p ψ+,0 ψ−,0 Fp σ λ+ λ− D ± F12
T 0 − 12
12 0 0 − 1
212 0 0 − 1
212 (0, 1)
(+,−) T ′ 0 0 1 1 −1 −1 0 0 0 0 0 (0, 1)
(−,+) T ′ 0 −1 0 −1 1 0 1 0 0 −1 1 (0, 1)
(+,+) T ′ 0 0 0 0 0 0 0 0 −1 −1 0 (0, 1)
(−,−) T ′ 0 −1 1 0 0 −1 1 0 1 0 1 (0, 1)
anti-chiral superfield, twisted anti-chiral superfield
anti-chiral superfield (Si, P )、twisted anti-chiral superfield Σ の component fields についても与えておこう。様々
な U(1) charge の符号は逆になるが、spin は逆になっていないので、ここできちんと考察しておいた方が良い。もう
一度 spin について書いておくと、Lorentz group U(1)E の spin T の assign は11
(ψ+, ψ−
)=(− 1
2,
1
2
)chiral superfield , (D.5.10a)
(ψ+, ψ−
)=(− 1
2,
1
2
)anti-chiral superfield , (D.5.10b)
である。この逆の定義でも構わないが Witten [109] など一連の文献に合せてある。left/right で spin が逆向きであ
ることはすぐ納得できよう。chiral/anti-chiral で同じ向きであるのは、spin は worldsheet spin であり、target の
holomorphicity とは関係ないからである。
anti-chiral superfield についても R-symmetry を次で定義しよう:
RV -symmetry : Φ(θ±, θ±) → eiqV αΦ(e−iαθ±, e+iαθ±) , (D.5.11a)
RA-symmetry : Φ(θ±, θ±) → eiqAαΦ(e∓iαθ±, e±iαθ±) . (D.5.11b)
anti-chiral superfield、twisted anti-chiral superfield の展開を記述しよう:
Φ(θ±) = φ+√
2θ+ψ+ +√
2θ−ψ− + 2θ+θ−F , (D.5.12a)
Y (θ+, θ−) = y +√
2θ+χ+ +√
2θ−χ− + 2θ+θ−G . (D.5.12b)
一般的な anti-chiral superfield Φ と twisted anti-chiral superfield Y の R-charge を以下で定義する:
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Φ qVφ qA
φ12 (qV
φ − qAφ ) 1
2 (qVφ + qA
φ )
φ qVφ qA
φ12 (qV
φ − qAφ ) 1
2 (qVφ + qA
φ )
ψ+ qVφ + 1 qA
φ + 1 12 (qV
φ − qAφ ) 1
2 (qVφ + qA
φ + 2)
ψ− qVφ + 1 qA
φ − 1 12 (qV
φ − qAφ + 2) 1
2 (qVφ + qA
φ )
F qVφ + 2 qA
φ12 (qV
φ − qAφ + 2) 1
2 (qVφ + qA
φ + 2)
11generator と eigenvalue を同じ記号で表すが混乱はないであろう。
214 Topological Sigma Model
fields QV QA QL = 12 (QV −QA) QR = 1
2 (QV +QA)
Y qVy qA
y12 (qV
y − qAy ) 1
2 (qVy + qA
y )
y qVy qA
y12 (qV
y − qAy ) 1
2 (qVy + qA
y )
χ+ qVy + 1 qA
y + 1 12 (qV
y − qAy ) 1
2 (qVy + qA
y + 2)
χ− qVy − 1 qA
y + 1 12 (qV
y − qAy − 2) 1
2 (qVy + qA
y )
G qVy qA
y + 2 12 (qV
y − qAy − 2) 1
2 (qVy + qA
y + 2)
quintics CP4[5] で議論する。(twisted) anti-superpotential は
W = P
5∑
i=1
S5i , W = −τΣ , (D.5.13)
であるので、
qVs = −2
5, qA
s = 0 , qVp = 0 , qA
p = 0 , qVσ = 0 , qA
σ = −2 , (D.5.14)
となる。よって、各 charge は次のように与えられる:
superfield components T QV QA QL QR Qg
Si si 0 − 25 0 − 1
5 − 15 −1
ψ+,i − 12
35 1 − 1
545 −1
ψ−,i12
35 −1 4
5 − 15 −1
F s 0 85 0 4
545 1
superfield components T QV QA QL QR Qg
P p 0 0 0 0 0 5
ψ+,0 − 12 1 1 0 1 5
ψ−,012 1 −1 1 0 5
F p 0 2 0 1 1 5
superfield components T QV QA QL QR Qg
Σ σ 0 0 −2 1 −1 0
λ+ − 12 1 −1 1 0 0
λ−12 −1 −1 0 −1 0
D ± F12 (0, 1) 0 0 0 0 0
topological twist を行う。Lorentz generator T を
T → T ′ = T +1
2JR ,
で再定義する。ここで JR は twist の方法に依存する:
1
2JR =
− 12 (QL +QR) + 1
5Qg (+,−) twist
+ 12 (QL +QR)− 1
5Qg (−,+) twist
− 12 (−QL +QR) (+,+) twist
+ 12 (−QL +QR) (−,−) twist
D.5 Example: Topological GLSM for quintics 215
では各 twist において component fields がどのように spin の変更を受けるのか具体的に見よう:
si ψ+,i ψ−,i F i p ψ+,0 ψ−,0 F p σ λ+ λ− D ± F12
T 0 − 12
12 0 0 − 1
212 0 0 − 1
212 (0, 1)
(+,−) T ′ 0 −1 0 −1 1 0 1 0 0 −1 1 (0, 1)
(−,+) T ′ 0 0 1 1 −1 −1 0 0 0 0 0 (0, 1)
(+,+) T ′ 0 0 1 0 0 −1 1 0 1 0 1 (0, 1)
(−,−) T ′ 0 −1 0 0 0 0 0 0 −1 −1 0 (0, 1)
chiral superfield, twisted chiral superfield の spin assignment とは若干異なるのが理解できるであろう。
Problems
先にも書いたが、gauge current J が BRST exact な形、もしくは supercharge の反交換で記述できるなら、spin
を (半)整数にするように適当に付加させることができる。ならば、どの場合がそれが出来て、どの場合ができない
のか、model 毎に考察する必要があるのかないのか、議論を展開する必要がある。これが一般的に理解できれば、
Silverstein-Witten [99] の path integral measure の deformation の一般的考察が可能になり、最終的に Ooguri-Vafa
[89] の C domain の計算と理解に非常に大きく貢献することになる。
もうひとつ問題がある。Ooguri-Vafa [89] で使われている geometry は resolved conifold である。それを super-
potential 無しで記述している。従って、chiral superfield の R-charge が決まらない。だからこれまでの議論で登場す
るような、left/right U(1) current を R-symmetry current から導出する手順が使えないと思われる。twisted chiral
superfield については使えるが、あまり有用ではない。この点に関しては Klebanov-Witten [68] の定義を採用するこ
とにしよう。
(D.2.1)では energy-momentum tensorに U(1) R-currentを付加させて modified tensorを定義した。しかし元々、
spin とは Lorentz group の表現であり、回転群の表現である。しかるに energy-momentum tensor は translation
つまり推進の演算子であり、回転ではない。ここの点がまだ頭の中で繋がらない。
Witten [109] でも、Lorentz generator T に left/right U(1) generator を付加させている (D = 2 Euclidean
worldsheet では Lorentz group は U(1) なので generator は一つである)。そして
T ′ = T +1
2(JL + JR) (D.5.15)
と変形させて定義している ([109] とは符号が異なるがここではこれまでの議論に合わせている)。
supersymmetry transformation
先の subsectionまでで GLSM for CP4[5]の全ての fieldsについて topological twistが完了したので、supersymmetry
変換の twist で得られる BRST 変換を考察しよう。まずは twist 以前の supersymmetry 変換則を列挙する:
gauge multiplet (twisted chiral):
δv0 = iα−λ− + iα+λ+ + iα+λ+ + iα−λ− δv1 = −iα−λ− + iα+λ+ + iα+λ+ − iα−λ−
216 Topological Sigma Model
δσ = −i√
2α+λ− − i√
2α−λ+ δσ = −i√
2α+λ− − i√
2α−λ+
δλ+ = iα+D + 2√
2∂σα− + iF12α+ δλ− = iα−D − 2√
2∂σα+ − iF12α−
δλ+ = −iα+D + 2√
2∂σα− + iF12α+ δλ− = −iα−D − 2√
2∂σ α+ − iF12α−
δD = 2α+∂λ+ − 2α−∂λ− − 2α+∂λ+ + 2α−∂λ− (D.5.16a)
chiral multiplet Si (Qg = 1):
δsi =√
2(α+ψ−,i − α−ψ+,i)
δsi = −√
2(α+ψ−,i − α−ψ+,i)
δψ+,i = 2√
2iDzsiα− +√
2α+Fi − 2siσ α+
δψ−,i = 2√
2iDzφα+ +√
2α−Fi + 2siσ α−
δψ+,i = −2√
2iDzsiα− +√
2α+F i + 2siσ α+
δψ−,i = −2√
2iDzsiα+ +√
2α−F i − 2siσ α−
δFi = 2√
2iα+Dzψ+,i − 2√
2iα−Dzψ−,i + 2(α+σψ−,i + α−σψ+,i) + 2isi(α−λ+ − α+λ−)
δF i = 2√
2iα+Dzψ+,i − 2√
2iα−Dzψ−,i + 2(α+σψ−,i + α−σψ+,i)− 2isi(α−λ+ − α+λ−) (D.5.16b)
chiral multiplet P (Qg = −5):
δp =√
2(α+ψ−,0 − α−ψ+,0)
δp = −√
2(α+ψ−,0 − α−ψ+,0)
δψ+,0 = 2√
2iDzpα− +√
2α+Fp + 10pσ α+
δψ−,0 = 2√
2iDzpα+ +√
2α−Fp − 10pσ α−
δψ+,0 = −2√
2iDzpα− +√
2α+F p − 10pσ α+
δψ−,0 = −2√
2iDzpα+ +√
2α−F p + 10pσ α−
δFp = 2√
2iα+Dzψ+,0 − 2√
2iα−Dzψ−,0 − 10(α+σψ−,0 + α−σψ+,0)− 10ip(α−λ+ − α+λ−)
δF p = 2√
2iα+Dzψ+,0 − 2√
2iα−Dzψ−,0 − 10(α+σψ−,0 + α−σψ+,0) + 10ip(α−λ+ − α+λ−) (D.5.16c)
topological twist として A-twist 特に (−,+) twist を採用する。ここで注意が必要である。それは、twist する際に
Si と P では逆向きに twist されることである:
si ψ+,i ψ−,i Fi p ψ+,0 ψ−,0 Fp σ λ+ λ− D ± F12
T 0 − 12
12 0 0 − 1
212 0 0 − 1
212 (0, 1)
(−,+) T ′ 0 −1 0 −1 1 0 1 0 0 −1 1 (0, 1)
si ψ+,i ψ−,i F i p ψ+,0 ψ−,0 F p σ λ+ λ− D ± F12
T 0 − 12
12 0 0 − 1
212 0 0 − 1
212 (0, 1)
(−,+) T ′ 0 0 1 1 −1 −1 0 0 0 0 0 (0, 1)
D.5 Example: Topological GLSM for quintics 217
ここで、Si sector と P sector で twist が異なっているが、問題はない。supersymmetry parameter α の twist は一
意であるので、それを与える。Si sector で合わせておこう。Si sector では (ψ−,i, ψ+,i) が spin zero になるので、そ
れに合わせて
α− = α+ , α+ = −α− (D.5.17)
が spin zero になる。これらを α+ = α− ≡ α としておこう。そして spin one になる (α−, α+) を α− = α+ ≡ 0 に
しておこう。この定義における BRST 変換則をここに記載しておこう:
gauge multiplet (twisted chiral):
δB v0 = iα(λ− + λ+) , δB v1 = −iα(λ− − λ+) ,
δB σ = 0 , δB σ = −i√
2α(λ+ + λ−) ,
δBD = α[∂λ+ − ∂λ−
],
δB λ+ = αi(D + F12) , δB λ− =√
2α∂σ ,
δB λ+ =√
2α∂σ , δB λ− = −αi(D + F12) . (D.5.18a)
chiral multiplet Si (Qg = 1)
δB si =√
2αψ−,i , δB si =√
2αψ+,i ,
δB ψ+,i =√
2α[iDzsi + Fi
], δB ψ−,i = 2α siσ ,
δB ψ+,i = −2α siσ , δB ψ−,i =√
2α[iDzsi + F i
],
δB Fi = α[− i√
2Dzψ−,i + 2(σψ+,i + isiλ+)],
δB F i = α[− i√
2Dzψ+,i − 2(σψ−,i + isiλ−)]. (D.5.18b)
chiral multiplet P (Qg = −5)
δB p =√
2αψ−,0 , δB p =√
2αψ+,0 ,
δB ψ+,0 =√
2α[2iDzp+ Fp
], δB ψ−,0 = −10α pσ ,
δB ψ+,0 = 10αpσ , δB ψ−,0 =√
2α[− 2iDzp+ F p
],
δB Fp = α[− 2√
2iDzψ−,0 − 10(σψ+,0 + ipλ+)],
δB F p = α[2√
2iDzψ+,0 + 10(σψ−,0 + ipλ−)]. (D.5.18c)
BRST exact modes
次に BRST exact modes を探し出す。BRST exact modes は path integral に寄与しないが、exact でない mode は
寄与する。それが effective theory に影響を与えるので、しっかりとした理解が必要となる。
(D.5.18) より、gauge multiplet Σ、chiral multiplets (Si, P ) の情報は次のようになる:
218 Topological Sigma Model
Σ,Σ modified spin BRST
σ 0 ∂σ, ∂σ が exact
σ 0 -
λ− 1 exact
λ+ −1 exact
λ+ 0 λ+ + λ− が exact
λ− 0 λ+ + λ− が exact
D 0 D ± F12 が exact
vm 1 D ± F12 が exact
これより、σ の constant mode σconst は BRST exact mode ではないことがわかる。
chiral multiplets の考察は gauge multiplet の結果を用いる:
Si, Si modified spin BRST
si 0 siσ が exact
si 0 siσ が exact
ψ+,i −1 -
ψ−,i 0 -
ψ+,i 0 exact
ψ−,i 1 exact
Fi −1 -
F i 1 -
P, P modified spin BRST
p 1 pσ が exact
p −1 pσ が exact
ψ+,0 0 -
ψ−,0 1 exact
ψ+,0 −1 exact
ψ−,0 1 -
Fp 0 -
F p 0 -
注意であるが、例えば siσ が BRST exact でも si が BRST exact とは言えない。同様にして λ+ が BRST exact で
も siλ+ が exact とは言えない。さらに連動して σψ+,i は BRST exact とは言えない。
path integral
chiral multiplets (Si, P )を path integralして gauge multiplet Σだけの effective actionを得よう。その際、component
field σ の path integral measure が変更を受ける (effective action に取り込んでもいいが)。
まず、元々 supersymmetric theory だったので、chiral multiplets (Si, P ) の boson と fermion の one-loop は
“non-constant modes” は互いにキャンセルをする。実際 Gauss-Fresnel 積分をすればそれがわかる12。topological
theory なので one-loop correction で exact である [111]。non-trivial に寄与するのは constant modes である。つま
り scalar fields の constant mode である。この議論は Witten index に通じるものがあると思われる [108]。
元々 supersymmetry変換則 (D.5.16)に入っている fermionは、spin 1/2の spinorなので、constant modeは入っ
ていない。その supersymmetry 変換則 (D.5.16) を twist して BRST 変換則 (D.5.18) に書き直したので、(D.5.18)
は spin zeroの fermionについては constant mode抜きの変換則で、constant modeは BRST exact ではない。では、
12積分される chiral multiplet は求めたい effective theory では非常に重いと考えて、interaction term は効かないとして無視している。そ
のため 2 次の項までしか考慮しない。
D.6 Example: Topological GLSM for resolved conifold 219
この spin zero fermion の “constant mode” の自由度は何処から来たのであろうか。ちなみにこの spin zero fermion
の “constant mode” の数は、Witten [111] の “the number of zero modes of the scalar χi” = an (n は instanton
number) に相当すると考えられる。そしてこれはさらに “the number of zero modes of one-form ψiz” = bn が無い場
合の、“virtual dimensions of the moduli space Mn” = wn となると思われる。
この解釈の下で、実際に path integral を実行しよう。GLSM for CP4[5] は次の様になっている:
L = −5∑
i=1
si
∂∂ + 2|σ|2
si − p
∂∂ + 50|σ|2
p
+
5∑
i=1
(ψ−,i , ψ+,i
)( 2i∂ −√
2σ
−√
2σ −2i∂
)(ψ−,i
ψ+,i
)+(ψ−,0 , ψ+,0
)( 2i∂ 5√
2σ
5√
2σ −2i∂
)(ψ−,0
ψ+,0
)
+ · · · . (D.5.19)
すでに auxiliary field は on-shell にしてあるとする。ここで non-constant modes は
det( 1
4∂∂ + 2|σ|2)5
·
det( 1
4∂∂ + 50|σ|2)×
det(4∂∂ + 2|σ|2
)5 ·
det(4∂∂ + 50|σ|2
)(D.5.20)
の寄与があり、これが path integral measure に寄与するが、良く見ると打ち消しあっていることがわかる。一方、
constant modes については、spin zero scalar field は constant mode があるが、spin 1/2 fermion には存在しない。
そのため、一般の supersymmetric theory でも constant mode の寄与が重要となる。今 topological A-twist を行っ
て、spin zero fermions (ψ−,i, ψ+,i), (ψ+,0, ψ−,0) が登場した。これらは BRST exact ではない constant modes を
持つことができる13。また、scalar field であった (p, p) は topological twist により spin zero ではなくなっているた
め、constant mode を失う14。したがって、constant modes の path integral measure への寄与は、si から 1/2|σ|2
が、(ψ−,i, ψ+,i) から −√
2σ が、そして (ψ+,0, ψ−,0) から 5√
2σ が寄与する。まとめると
( 1
2σσ
)5
·(−√
2σ)5 ·
(5√
2σ)∼ 1
σ4, (D.5.21)
が σ の path integral measure に寄与することになる (factor は path integral の規格化に含ませるため無視する)。
D.6 Example: Topological GLSM for resolved conifold
resolved conifold ではどうだろうか。
• model が内蔵する left/right U(1) charge を各 field 毎に決定する
• left/right U(1) charge を用いて modified spin を決定する
• modified spin が (半)整数にならない場合、うまい調節を行うと gauge current が BRST exact になることを
確認する
13今のところ出所は理解していない。14行先は不明。
220 Topological Sigma Model
Lagrangian
まずは Lagrangian を書いておこう。left/right U(1) charge の決定などには全ての情報をまず与えておくと後に都合
がよい。Lagrangian は GLSM for O(−1)⊕O(−1)→ CP1 である [113]:
L =
∫d4θ
− 1
e2ΣΣ +
2∑
i=1
Aie2V Ai +
2∑
k=1
Bke−2V Bk
+(− 1√
2
∫d2θ τΣ + (h.c.)
). (D.6.1)
この Lagrangian には superpotential term が存在しない。したがって chiral multiplets の R-charge が決定できず、
これを用いて topological twist をする手法は困難である。しかし幸い R-symmetry と left/right U(1) symmetry に
は関係があったので、そちらを考察しよう。そのために Lagrangian を component fields で書き直しておこう:
L =
2∑
i=1
ai
4∂∂ +D − 2|σ|2 − vmv
mai + 2i
2∑
i=1
vm ai∂mai
+2∑
k=1
bk4∂∂ −D − 2|σ|2 − vmv
mbk − 2i
2∑
k=1
vm bk∂mbk
+2∑
i=1
(ψ−,i , ψ+,i
)( 2i∂ − 2vz −√
2σ
−√
2σ −2i∂ + 2vz
)(ψ−,i
ψ+,i
)
+
2∑
k=1
(ψ−,k , ψ+,k
)( 2i∂ + 2vz
√2σ
√2σ −2i∂ − 2vz
)(ψ−,k
ψ+,k
)
− i√
2
2∑
i=1
ai
(λ−ψ+,i − λ+ψ−,i
)− ai
(ψ+,iλ− − ψ−,iλ+
)
+ i√
2
2∑
k=1
bk(λ−ψ+,k − λ+ψ−,k
)− bk
(ψ+,kλ− − ψ−,kλ+
)
− 1
2e2F 2
12 − iθF12 +4
e2σ∂∂σ +
(λ− , λ+
)( 2i∂ 0
0 −2i∂
)(λ−
λ+
)
+1
2e2D2 − rD +
2∑
i=1
|Fi|2 +
2∑
k=1
|Fk|2 . (D.6.2)
covariant derivative Dmφi = (∂m + iQivm)φi などを用いて記述してもよかったが、敢て全て展開しておいた。また、
2vz ≡ −(v0 − v1), 2vz ≡ (v0 + v1) である。
topological twist
先の section の議論を応用して、left/right U(1) charge から R-charge を読み取る表を載せておこう:
multiplet component QV QA QL QR
Ai ai qLi + qR
i −qLi + qR
i qLi qR
i
ψ+,i qLi + qR
i − 1 −qLi + qR
i − 1 qLi qR
i − 1
ψ−,i qLi + qR
i − 1 −qLi + qR
i + 1 qLi − 1 qR
i
Fi qLi + qR
i − 2 −qLi + qR
i qLi − 1 qR
i − 1
D.6 Example: Topological GLSM for resolved conifold 221
multiplet component QV QA QL QR
Bk bk qLk + qR
k −qLk + qR
k qLk qR
k
ψ+,k qLk + qR
k − 1 −qLk + qR
k − 1 qLk qR
k − 1
ψ−,k qLk + qR
k − 1 −qLk + qR
k + 1 qLk − 1 qR
k
Fk qLk + qR
k − 2 −qLk + qR
k qLk − 1 qR
k − 1
multiplet component QV QA QL QR
Σ σ 0 2 −1 1
λ+ −1 1 −1 0
λ− 1 1 0 1
D ± F12 0 0 0 0
但し QV = QL +QR, QA = −QL +QR を用いた。(twisted) anti-chiral superfield の U(1) charge は全て逆符号で
与えられるのでここではもはや列挙しない。また twisted superpotential W = −τΣ が与えられているために Σ の
R-charge はすぐに決定できる。
chiral multiplets (Ai, Bk) の U(1) charge がこれで決定できるかと思ったが、決定できない。superfield
level で不定であることから明らかであったかと思われるが、念のため component field でチェックしてもやはり
決まらない。 classical theory だけでは決定できない。しかし CY condition を課しておくと chiral anomaly free で
もあるので charge を決定することはできない。つまり GLSM だけでは決まらない。
決まらないのであれば、計算が簡単になる方で表示しよう。つまり R-charge で書いておこう:
multiplet component QV QA QL QR
Ai ai qVi qA
i12 (qV
i − qAi ) 1
2 (qVi + qA
i )
ψ+,i qVi − 1 qA
i − 1 12 (qV
i − qai ) 1
2 (qVi + qV
i − 2)
ψ−,i qVi − 1 qA
i + 1 12 (qV
i − qai − 2) 1
2 (qVi + qV
i )
Fi qVi − 2 qA
i12 (qV
i − qai − 2) 1
2 (qVi + qV
i − 2)
multiplet component QV QA QL QR
Bk bk qVk qA
k12 (qV
k − qAk ) 1
2 (qVk + qA
k )
ψ+,k qVk − 1 qA
k − 1 12 (qV
k − qak) 1
2 (qVk + qV
k − 2)
ψ−,k qVk − 1 qA
k + 1 12 (qV
k − qak − 2) 1
2 (qVk + qV
k )
Fk qVk − 2 qA
k12 (qV
k − qak − 2) 1
2 (qVk + qV
k − 2)
ここで制限を課す。この Lagrangian は GLSM for resolved conifold であり、FI parameter が正でも負でも、low
energy effective theory は共に CY sigma model on resolved conifold であり、flop で互いに移り合う。topology と
して殆ど同じで、異なるのは triangulation の向きだけである。よって Ai と Bk は共に同じ性質を持つべきである
(U(1) gauge charge だけは flop のために逆符号である)。また ai, bk はそれぞれ CP1 の homogeneous coordinates
とみなすため、これも同じ性質を持つ。従って全ての scalar field の R-charge は等しいとする:
qVi = qV
k ≡ q , qAi = qA
k ≡ q′ . (D.6.3)
222 Topological Sigma Model
これでもまだ不定性が大きい。ここで Klebanov-Witten [68] の記述を採用する。Klebanov-Witten では
q =1
2, (D.6.4)
としている。その理由は chiral anomaly cancellationによるらしい。Klebanov-Witten [68]での anomaly cancellation
は、U(N) gauge group の Casimir と R-charge を用いて行っている。この U(N) は N 枚の D3-branes 上の gauge
groupである。Ooguri-Vafa [89]でも open string sideではほとんどこれと同じ状況なので、この値を採用しておこう。
topological twist をする際、gauge current が BRST exact になることを確認した方が良いのだが、おそらく一
般的に証明ができると見越して、ここでは gauge current を加えた形で spin の変更を行う。変形は U(1)E Lorentz
generator T を
T ′ = T +1
2JV +
1
αJg , (D.6.5)
で与える。α は適当な整数である。GLSM for CP4[5] の場合は、gauge charge の maximal absolute value α = | − 5|がとられている (何故かははっきりしないが、これを採用すると modified spin が整数を取ることができる)。ちなみ
にこの twist は supercharge Q−, Q+ が spin zero charge (BRST charge) となる topological A-twist ((−,+) twist)
である。
上の再定義により、chiral multiplets (Ai, Bk) の component field の modified spin が次のように与えられる:
T QV Qg T ′ = T + 12QV + 1
αQg
Ai ai 0 12 1 1
4 + 1α
ψ+,i − 12 − 1
2 1 − 34 + 1
α
ψ−,i12 − 1
2 1 14 + 1
α
Ai ai 0 − 12 −1 − 1
4 − 1α
ψ+,i − 12
12 −1 − 1
4 − 1α
ψ−,i12
12 −1 3
4 − 1α
T QV Qg T ′ = T + 12QV + 1
αQg
Bk bk 0 12 −1 1
4 − 1α
ψ+,k − 12 − 1
2 −1 − 34 − 1
α
ψ−,k12 − 1
2 −1 14 − 1
α
Bk bk 0 − 12 1 − 1
4 + 1α
ψ+,k − 12
12 1 − 1
4 + 1α
ψ−,k12
12 1 3
4 + 1α
U(1) gauge current は BRST exact に記述できる (はず) なので、この定数 α は任意であると思われる。しかし
modified spin も 1 以下の (半) 整数であるべきなので、その条件を考察しよう。この条件は、整数 n,m ∈ Z を用意
して、
−1 ≤ 1
4+
1
α≡ 1
2n ≤ 1 and − 1 ≤ −3
4+
1
α≤ 1 , (D.6.6a)
−1 ≤ 1
4− 1
α≡ 1
2m ≤ 1 and − 1 ≤ −3
4− 1
α≤ 1 , (D.6.6b)
D.6 Example: Topological GLSM for resolved conifold 223
と記述できる。これより
n = 0, 1, 2 m = 0, 1, 2 , (D.6.7)
のみが許されることがわかる。この制限の下で α は (n,m) = (1, 0), (0, 1) でのみ以下の解を持つことが許される15:
α = 4 for (n,m) = (1, 0) , α = −4 for (n,m) = (0, 1) . (D.6.8)
このどちらを採用するかは自由である。α = −4 での modified spin は、α = 4 での modified spin で Ai と Bk を入
れ替えただけの、ちょうど flop の関係にある。ここでは ai が spin zero scalar を保つ α = −4 を採用しよう。
twisted chiral multiplet Σ については容易に次の結果が得られる:
T QV Qg T ′ = T + 12QV + 1
αQg
Σ σ 0 0 0 0
λ+ − 12 −1 0 −1
λ−12 1 0 1
D − iF01 (0, 1) 0 0 (0, 1)
Σ σ 0 0 0 0
λ+ − 12 1 0 0
λ−12 −1 0 0
ここでもう一度、twist 前の spin と twist 後の modified spin の一覧を挙げる:
ai ψ+,i ψ−,i bk ψ+,k ψ−,k σ λ+ λ− D ± F12
T 0 − 12
12 0 − 1
212 0 − 1
212 (0, 1)
(−,+) T ′ 0 −1 0 12 − 1
212 0 −1 1 (0, 1)
ai ψ+,i ψ−,i bk ψ+,k ψ−,k σ λ+ λ− D ± F12
T 0 − 12
12 0 − 1
212 0 − 1
212 (0, 1)
(−,+) T ′ 0 0 1 − 12 − 1
212 0 0 0 (0, 1)
BRST exact modes
supersymmetry 変換をここで挙げることはもうしない。一般的な記述はWitten [113] を参照すること。また、gauge
multiplet Σ の BRST 変換などすべて (D.5.18) と同様となるので、細かな議論はここでは省略する。必要になったと
きに改めて議論を行う。
path integral measure
chiral multiplets (Ai, Bk) を path integral する。その際、non-constant modes については元々の supersymmetry
により boson/fermion でキャンセルが起こる。しかし、constant modes については一般にキャンセルが起こらない。15 1
4+ 1
α= 1
2n かつ 1
4− 1
α= 1
2m より、 1
2= 1
2(n + m) が得られる。よって n + m = 1 のみが許される。
224 Topological Sigma Model
そこで topological (−,+) twist により spin zero になった field の constant modes を考察する。spin zero になって
いる field は boson では (ai, ai) のみであり、fermion は (ψ−,i, ψ+,i) のみである。これらの path integral の結果、
gauge multiplet の path integral measure にそれぞれ 1/|σ|2, σ の寄与がかかるので、
DAiDAiDBkDBkDΣDΣ → DΣDΣ ·( 1
σσ
)2
(σ)2 = DΣDΣ ·( 1
σ2
)(D.6.9)
となる。
D.7 Problems
先の議論の展開では追究しきれていない部分、topological theory における一般的な問題点を列挙する。
• superpotential のない model から、どうやって R-charge を決定するか。classical theory では決定できない
なら、Klebanov-Witten [68] に述べられているように quantum theory でも破れていない、つまり “anomaly
cancellation” を満たすように与える手法はどのようなものか。単に worldsheet (super)conformal theory だと
思っていると、CY condition さえあれば R-symmetry は anomaly を持たない。string theory だと思って、
spacetime にも gauge theory が存在するとして (例えば N D-branes wrapped on cycles in CY の状況など)、
gauge anomaly cancellation を考慮しないと決まらないのか。
• topological A-twist をする際に U(1) gauge current を left/right U(1) current に余分に付加できる理由をどう
確立するか。gauge current が BRST charge となる supercharge を用いて J ∼ Q, · · · と記述できることを証明せよ。
• twist によって spin zero になった fermion の constant mode の起源は何か。また twist で spin zero でなく
なった scalar boson の constant mode はどこへ行ったのか。moduli space の virtual dimensions との関係は
どうなっているのか。
• topological twisted theory では BRST exact な mode は理論に効かないので path integral はしなくても良い。
しかし一方で、topological twist されたことによって新たに登場する spin zero field の constant mode の寄与
が重要な役割を果す。元々 supersymmetric theory においても constant mode の寄与は重要であったことと同
じである。この constant mode の由来を明らかにすべきである。
D.8 A-type branes, B-type branes, boundary CFT
Ooguri, Oz and Yin [87] 参照。
Bibliography
[1] A. Adams, A. Basu and S. Sethi, “(0, 2) duality,” Adv. Theor. Math. Phys. 7 (2004) 865 [arXiv: hep-
th/0309226].
[2] M. Aganagic and C. Vafa, “Mirror Symmetry, D-branes and Counting Holomorphic Discs,” hep-th/0012041.
[3] L. Alvarez-Gaume and P. Ginsparg, “Finiteness of Ricci Flat Supersymmetric Non-linear Sigma-models,”
Commun. Math. Phys. 102 (1985) 311.
[4] V.I. Arnold, Russian Mathematical Survey, 28 (1973) 19.
V.I. Arnold, Russian Mathematical Survey, 29 (1974) 11.
V.I. Arnold, Russian Mathematical Survey, 30 (1975) 1.
[5] V.I. Arnold, “Singularity Theory,” London Mathematical Lecture Notes Series: 53, Cambridge University Press
(1981).
V.I. Arnold, S.M. Gusein-Zade and A.N. Varchenko, “Singualrities of Differentiable Maps,” volume 1 and 2,
Birkhauser (1985).
[6] P.S. Aspinwall, “Enhanced Gauge Symmetries and K3 Surfaces,” Phys. Lett. B357 (1995) 329, hep-th/9507012.
[7] J. Bogaerts, A. Sevrin, S. van der Loo and S. Van Gils, “Properties of semi-chiral superfields,” Nucl. Phys. B
562 (1999) 277, [arXiv:hep-th/9905141].
[8] I. Brunner and K. Hori, “Orientifolds and Mirror Symmetry,” hep-th/0303135.
[9] I. Brunner, K. Hori, K. Hosomichi and J. Walcher, “Orientifolds of Gepner Models,” hep-th/0401137.
[10] T. Buscher, U. Lindstrom and M. Rocek, “New supersymmetric sigma models with Wess-Zumino terms,”
Phys. Lett. B 202 (1988) 94.
[11] E. Calabi, “Metriques Kahleriennes et Fibres Holomorphes (in French),” Ann. Scient. Ec. Norm. Sup. 12 (1979)
269.
[12] C.G. Callan, R. Dashen and D.J. Gross, “A Mechanism for Quark Confinement”, Phys. Lett. 66B (1977) 375.
[13] P. Candelas and X.C. de la Ossa, “Moduli Space of Calabi-Yau Manifolds,” Nucl. Phys. B355 (1991) 455.
ii BIBLIOGRAPHY
[14] S. Cecotti, “N = 2 Landau-Ginzburg vs. Calabi-Yau σ-models: Non-perturbative aspects,” Int. J. Mod. Phys.
A 6 (1991) 1749.
[15] S. Cecotti and C. Vafa, “Topological and antitopological Fusion,” Nucl. Phys. B367 (1991) 359.
[16] T.M. Chiang, A. Klemm, S.-T. Yau and E. Zaslow, “Local Mirror Symmetry: Calculations and Interpreta-
tions,” Adv. Theor. Math. Phys. 3 (1999) 495, hep-th/9903053.
[17] S.R. Coleman, “There are No Goldstone Bosons in Two-dimensions,” Commun. Math. Phys. 31 (1973) 259.
[18] S.R. Coleman, “More about the Massive Schwinger Model”, Annals Phys. 101 (1976) 239.
[19] P. Di Francesco, P. Mathieu and D. Senechal, “Conformal Field Theory,” Springer.
[20] J. Distler, “Resurrecting (2, 0) compactifications,” Phys. Lett. B 188 (1987) 431.
[21] J. Distler and B. R. Greene, “Aspects of (2, 0) string compactifications,” Nucl. Phys. B 304 (1988) 1.
[22] L.J. Dixon, D. Friedan, E.J. Martinec and S. Shenker, “The Conformal Field theory of Orbifolds,” Nucl. Phys.
B282 (1987) 13.
[23] T. Eguchi and Y. Sugawara, “Modular invariance in superstring on Calabi-Yau n-fold with A-D-E singularity,”
Nucl. Phys. B577 (2000) 3, hep-th/0002100.
T. Eguchi and Y. Sugawara, “D-branes in singular Calabi-Yau n-fold and N = 2 Liouville theory,” Nucl. Phys.
B598 (2001) 467, hep-th/0011148.
T. Eguchi and Y. Sugawara, “SL(2,R)/U(1) Supercoset and Elliptic Genera of Non-compact Calabi-Yau
Manifolds,” JHEP 0405 (2004) 014, hep-th/0403193.
[24] V.A. Fateev, A.B. Zamolodchikov, Al.B. Zamolodchikov, unpublished.
[25] J. Fuchs, A. Klemm, C. Scheich and M.G. Schmidt, “Spectra and Symmetries of Gepner Models Compared
to Calabi-Yau Compactifications,” Annals Phys. 204 (1990) 1.
[26] T. Fukuda, K. Hosomichi, “Three-point Functions in Sine-Liouville Theory,” hep-th/0105217.
[27] S. J. Gates, C. M. Hull and M. Rocek, “Twisted multiplets and new supersymmetric nonlinear sigma models,”
Nucl. Phys. B 248 (1984) 157.
[28] D. Gepner, “Exactly Solvable String Compactifications on Manifolds of SU(N) Holonomy,” Phys. Lett. B199
(1987) 380.
[29] D. Gepner, “Space-time Supersymmetry in Compactified String Theory and Superconformal Models,” Nucl.
Phys. B296 (1988) 757.
[30] D. Gepner, “Scalar Field Theory and String Compactification,” Nucl. Phys. B322 (1989) 65.
[31] D. Gepner and Z. Qiu, “Modular Invariant Partition Functions for Parafermionic Field Theories,” Nucl. Phys.
B285 (1987) 423.
BIBLIOGRAPHY iii
[32] D. Ghoshal and C. Vafa, “c = 1 string as the topological theory of the conifold,” Nucl. Phys. B453 (1995) 121,
hep-th/9506122.
[33] G. Giribet, C. Nunez, “Correlators in AdS3 String Theory,” hep-th/0105200.
[34] A. Giveon, D. Kutasov, “Little String Theory in a Double Scaling Limit,” JHEP 9910 (1999) 034,
hep-th/9909110.
[35] A. Giveon, D. Kutasov and O. Pelc, “Holography for Non-critical Superstrings,” JHEP 0010 (1999) 035,
hep-th/9907178.
[36] A. Giveon, D. Kutasov, “Notes on AdS3,” Nucl. Phys. B621 (2002) 303, hep-th/0106004.
[37] D. Ghoshal and S. Mukhi, “Topological Landau-Ginzburg Model of Two-dimensional String Theory,” Nucl.
Phys. B 425 (1994) 173, hep-th/9312189.
[38] B.R. Greene, “String Theory on Calabi-Yau Manifolds,” hep-th/9702155.
[39] B.R. Greene, D.R. Morrison and A. Strominger, “Black Hole Condensation and the Unification of String
Vacua,” Nucl. Phys. B 451 (1995) 109, hep-th/9504145.
[40] B.R. Greene and M. Ronen Plesser, “Duality in Calabi-Yau Moduli Space,” Nucl. Phys. B338 (1990) 15.
[41] B.R. Greene, C. Vafa and N.P. Warner, “Calabi-Yau Manifolds and Renormalization Group Flows,” Nucl.
Phys. B324 (1989) 371.
[42] M. T. Grisaru, M. Massar, A. Sevrin and J. Troost, “Some aspects of N = (2, 2), D = 2 supersymmetry,”
Fortsch. Phys. 47 (1999) 301 [arXiv:hep-th/9801080].
[43] D.J. Gross and A. Neveu, “Dynamical Symmetry Breaking in Asymptotically Free Field Theories,” Phys. Rev.
D10 (1974) 3235.
[44] A. Hanany, Y. Oz and M. Ronen Plesser, “Topological Landau-Ginzburg Formulation and Integrable Structure
of 2d String Theory,” Nucl. Phys. B 425 (1994) 150, hep-th/9401030.
[45] K. Higashijima and M. Nitta, “Supersymmetric Nonlinear Sigma Models as Gauge Theories,” Prog. Theor.
Phys. 103 (2000) 635, hep-th/9911139.
[46] K. Higashijima, T. Kimura and M. Nitta, “Gauge Theoretical Construction of Non-compact Calabi-Yau Man-
ifolds,” Ann. Phys. 296 (2002) 347, hep-th/0110216.
[47] K. Higashijima, T. Kimura and M. Nitta, “Calabi-Yau Manifolds of Cohomogeneity One as Complex Line
Bundles,” Nucl. Phys. B645 (2002) 438, hep-th/0202064.
[48] N. J. Hitchin, A. Karlhede, U. Lindstrom and M. Rocek, “Hyperkahler metrics and supersymmetry,” Commun.
Math. Phys. 108 (1987) 535.
[49] K. Hori, “Mirror Symmetry and Some Applications,” hep-th/0106043.
iv BIBLIOGRAPHY
[50] K. Hori, A. Iqbal and C. Vafa, “D-branes and Mirror Symmetry,” hep-th/0005247.
[51] K. Hori and A. Kapustin, “Duality of the Fermionic 2d Black Hole and N = 2 Liouville Theory as Mirror
Symmetry,” JHEP 0108 (2001) 045, hep-th/0104202.
[52] K. Hori and A. Kapustin, “Worldsheet Descriptions of Wrapped NS Five-branes,” JHEP 0211 (2002) 038,
hep-th/0203147.
[53] K. Hori, S. Katz, A. Klemm, R. Pandharipande, R. Thomas, C. Vafa, R. Vakil and E. Zaslow, “Mirror
Symmetry,” Clay Mathematics Monographs vol. 1 (2003), American Mathmatical Society.
[54] K. Hori and C. Vafa, “Mirror Symmetry,” hep-th/0002222.
[55] K. Hosomichi, “N = 2 Liouville theory with boundary,” arXiv:hep-th/0408172.
[56] K. Intriligator and C. Vafa, “Landau-Ginzburg Orbifolds,” Nucl. Phys. B339 (1990) 95.
[57] Y. Ishimoto and S. Yamaguchi, “Minimal String Theory is Logarithmic,” hep-th/0406262.
[58] I. T. Ivanov, B. b. Kim and M. Rocek, “Complex structures, duality and WZW models in extended superspace,”
Phys. Lett. B 343 (1995) 133, [arXiv:hep-th/9406063].
[59] V. Kazakov, I.K. Kostov, D. Kutasov, “A Matrix Model for the Two Dimensional Black Hole,”
hep-th/0101011.
[60] Y. Kazama and H. Suzuki, “New N = 2 Superconformal Field Theories and Superstring Compactification,”
Nucl. Phys. B 321 (1989) 232.
[61] T. Kimura, “Towards Mirror Symmetry on Non-compact Calabi-Yau Manifolds: version 4,” TK-NOTE/0204.
[62] T. Kimura, “Feynman Propagators in Flat Spacetime with Almost Plus Signature,” TK-NOTE/04-01.
[63] T. Kimura, “Towards Mirror Symmetry on Non-compact Calabi-Yau Manifolds,” contribution to the pro-
ceedings of the 12th International Conference on Supersymmetry and Unification of Fundamental Interactions
(SUSY 2004), hep-th/0409003.
[64] T. Kimura, “Gauged Linear Sigma Models for Noncompact Calabi-Yau Varieties,” hep-th/0411258.
[65] 九後 汰一郎, “ゲージ場の量子論 I,” 培風館.
[66] 九後 汰一郎, “ゲージ場の量子論 II,” 培風館.
[67] I.R. Klebanov, J. Maldacena and N. Seiberg, “Unitary and Complex Matrix Models as 1− d Type 0 Strings,”
hep-th/0309168.
[68] I.R. Klebanov and E. Witten, “Superconformal Field Theory on Threebranes at a Calabi-Yau Singularity,”
Nucl. Phys. B536 (1998) 199, hep-th/9807080.
[69] M. Kreuzer and H. Skarke, “Calabi-Yau Data,” http://hep.itp.tuwien.ac.at/˜kreuzer/CY/
BIBLIOGRAPHY v
[70] D. Kutasov, “Some Properties of (Non) Critical Strings,” hep-th/9110041.
[71] D. Kutasov, K. Okuyama, J.W. Park and N. Seiberg, “Annulus Amplitudes and ZZ Branes in Minimal String
Theory,” hep-th/0406030.
[72] W. Lerche, C. Vafa and N.P. Warner, “Chiral Rings in N = 2 Superconformal Theories,” Nucl. Phys. B324
(1989) 427.
[73] U. Lindstrom, “Generalized N = (2, 2) supersymmetric nonlinear sigma models,” Phys. Lett. B 587 (2004)
216, [arXiv:hep-th/0401100].
[74] U. Lindstrom, R. Minasian, A. Tomasiello and M. Zabzine, “Generalized complex manifolds and supersym-
metry,” Commun. Math. Phys. 257 (2005) 235, [arXiv: hep-th/0405085].
[75] U. Lindstrom, M. Rocek, R. von Unge and M. Zabzine, “Generalized Kahler geometry and manifest N = (2, 2)
supersymmetric nonlinear sigma models,” [arXiv: hep-th/0411186].
[76] M. Lynker and R. Schimmrigk, “ADE Quantum Calabi-Yau Manifolds,” Nucl. Phys. B339 (1990) 121.
[77] J.M. Maldacena, G.W. Moore and N. Seiberg, “Geometrical Interpretation of D-branes in Gauged WZW
Models,” JHEP 0107 (2001) 046, hep-th/0105038.
[78] E.J. Martinec, “Algebraic Geometry and Effective Lagrangians,” Phys.Lett. B217 (1989) 431.
[79] E.J. Martinec, “Criticality, Catastrophes and Compactifications,” in the V.G. Knizhnik memorial volume, eds.
L. Brink, et al., Physics and Mathematics of Strings, Print-89-0373 (EFI,CHICAGO).
[80] S. Mizoguchi, “Modular Invariant Critical Superstrings on Four-dimensional Minkowski Space × Two-
dimensional Black Hole,” JHEP 0004 (2000) 014, hep-th/0003053.
[81] D.R. Morrison and M. Ronen Plesser, “Summing the Instantons: Quantum Cohomology and Mirror Symmetry
in Toric Varieties,” Nucl. Phys. B440 (1995) 279, hep-th/9412236.
[82] S. Mukhi, C. Vafa, “Two-dimensional Black Hole as a Topological Coset Model of c = 1 String Theory,” Nucl.
Phys. B407 (1993) 667, hep-th/9301083.
[83] Y. Nakayama, “Liouville Field Theory – a Decade after the Revolution,” hep-th/0402009.
[84] N. Nakayama and K. Sugiyama, “Construction of Supergravity Backgrounds with a Dilaton Field,”
hep-th/0411143.
[85] M. Nitta, “Conformal Sigma Models with Anomalous Dimensions and Ricci Solitons,” hep-th/0411149.
[86] 太田 裕二, “Mirror 対称性の世界,” 素粒子論研究 90 巻 5 号 (1995).
[87] H. Ooguri, Y. Oz and Z. Yin, “D-branes on Calabi-Yau Spaces and Their Mirrors,” Nucl. Phys. B477 (1996)
407, hep-th/9606112.
vi BIBLIOGRAPHY
[88] H. Ooguri and C. Vafa, “Two-dimensional Black Hole and Singularities of CY Manifolds,” Nucl. Phys. B463
(1996) 55-72, hep-th/9511164.
[89] H. Ooguri and C. Vafa, “Worldsheet Derivation of a Large N Duality,” Nucl. Phys. B641 (2002) 3,
hep-th/0205297.
[90] J. Polchinski, “String Theory,” Cambridge University Press.
[91] Z. Qiu, “Nonlocal Current Algebra and N = 2 Superconformal Field Theory in Two-dimensions,” Phys. Lett.
B188 (1987) 207.
[92] Z. Qiu, “Modular Invariant Partition Functions for N = 2 Superconformal Field Theories,” Phys. Lett. B198
(1987) 497.
[93] M. Rocek and E. Verlinde, “Duality, Quotients, and Currents”, Nucl. Phys. B373 (1992) 630, hep-th/9110053.
[94] N. Seiberg and D. Shih, “Branes, Rings and Matrix Models in Minimal (Super)string Theory,” JHEP 0402
(2004) 021, hep-th/0312170.
[95] N. Seiberg and E. Witten, “Electric-magnetic Duality, Monopole Condensation, and Confinement in
N = 2 Supersymmetric Yang-Mills Theory,” Nucl. Phys. B426 (1994) 19 [Erratum-ibid. B430 (1994) 485],
hep-th/9407087.
[96] N. Seiberg and E. Witten, “Monopoles, Duality and Chiral Symmetry Breaking in N = 2 Supersymmetric
QCD,” Nucl. Phys. B431 (1994) 484, hep-th/9408099.
[97] A. Sevrin and J. Troost, “Off-shell formulation of N = 2 non-linear sigma-models,” Nucl. Phys. B 492 (1997)
623 [arXiv:hep-th/9610102].
[98] E. Silverstein, E. Witten, “Global U(1) R-symmetry and Conformal Invariance of (0, 2) Models,” Phys. Lett.
B328 (1994) 307, hep-th/9403054.
[99] E. Silverstein and E. Witten, “Criteria for Conformal Invariance of (0, 2) Models,” Nucl. Phys. B444 (1995)
161, hep-th/9503212.
[100] A. Strominger, “Massless Black Holes and Conifolds in String Theory,” Nucl. Phys. B 451 (1995) 96,
hep-th/9504090.
[101] K. Sugiyama and S. Yamaguchi, “D-branes on a Noncompact Singular Calabi-Yau Manifold,” JHEP 0102
(2001) 015, hep-th/0011091.
[102] D. Tong, “Mirror Mirror on the Wall: on Two-dimensional Black Holes and Liouville Theory,” JHEP 0304
(2003) 031, hep-th/0303151.
[103] C.Vafa, “String Vacua and Orbifoldized LG Models,” Mod. Phys. Lett. A4 (1989) 1169.
[104] C. Vafa and N.P. Warner, “Catastrophes and the Classification of Conformal Theories,” Phys.Lett. B218
(1989) 51.
BIBLIOGRAPHY vii
[105] N.P. Warner, “Lectures on N = 2 Superconformal Theories and Singularity Theory,” in Superstrings ’89,
eds. M. Green et al., (World Scientific, Singapore, 1990).
[106] E.J. Weinberg, “Index Calculations for the Fermion - Vortex System”, Phys. Rev. D24 (1981) 2669.
[107] J. Wess and J. Bagger, “Supersymmetry and Supergravity: 2nd edition, revised and expanded,” Princeton
University Press (1992).
[108] E. Witten, “Constraints on Supersymmetry Breaking,” Nucl. Phys. B202 (1982) 253.
[109] E. Witten, “Topological Sigma Models,” Commun. Math. Phys. 118 (1988) 411.
[110] E. Witten, “String Theory and Black Holes,” Phys. Rev. D 44 (1991) 314.
[111] E. Witten, “Mirror Manifolds and Topological Field Theory,” hep-th/9112056.
[112] E. Witten, “The N Matrix Model and Gauged WZW Models,” Nucl. Phys. B371 (1992) 191.
[113] E. Witten, “Phases of N = 2 Theories in Two Dimensions,” Nucl. Phys. B403 (1993) 159, hep-th/9301042.
[114] E. Witten, “On the Landau-Ginzburg Description of N = 2 Minimal Models,” Int. J. Math. Phys. A9 (1994)
4783, hep-th/9304026.
[115] A.B. Zamolodchikov, “Irreversibility of the Flux of the Renormalization Group in a 2D Field Theory,” JETP
Lett. 43 (1986) 731.
A.B. Zamolodchikov, “Renormalization Group and Perturbation Theory near Fixed Points in Two-Dimensional
Field Theory,” Sov. J. Nucl. Phys. 46 (1987) 1090.
[116] B. Zumino, “Supersymmetry and Kahler Manifolds,” Phys. Lett. B87 (1979) 203.
Index
ADE classification 61
β function 6
Calabi-Yau
— condition 4
— hypersurface CPN−1[N ] 18, 122
— phase 8, 14
CY/LG correspondence 20
noncompact — 14
CY/LG correspondence 9, 79
central charge 20, 56
chiral ring 19, 57
cigar 39
classical limit 9
cohomology ring 12
Coleman’s no-go theorem 8
complex structure 9
conformal
— dimension 52
— fixed point 55
— vacuum 69
— weight 20, 52
Coulomb branch 15
Coxeter number 58
critical point 12
cut-off 5
descendant field 58
dimensional transmutation 5
discrete spectrum 23
dynamical scale parameter 5
Fermat type 18
fermionic zero mode 110
FI parameter 3, 106
renormalized — 6
fiber coordinate 14
fluctuation modes 10
gauged WZW model 40
Gepner model 86
GLSM
Lagrangian 2
squashed — 36
T-dualized theory 100
Higgs branch 13
Higgs mechanism 7, 14
highest charged state 65, 197
Hodge diamond 88
homogeneous 55
homogeneous coordinate 14
homogeneous polynomial 18
hybrid phase 8
IR limit 11
isolated singularity 55, 196
Kahler class 9, 16
Kahler cone 7
Kahler potential 12
Landau-Ginzburg theory 8, 19
minimal model 8
modality 59
multiplicity of the singularity 51
NLSM 11
ii INDEX
non-renrmalization theorem 53
NS sector 69
NS-NS 2-form field 17
NS5-brane 41
O(−N) bundle on CPN−1 12, 118
O(−N + ℓ) bundle on CPN−1[ℓ] 24, 130
orbifold phase 8, 15
period integral 113, 117
Poincare polynomial 60, 74, 197
Poisson resummation formula 101
potential energy density 4
primary state 69
N = (2, 2) NS chiral — 69
projective space 10, 112
weighted — 47
quantum vacua 12
R-symmetry 45
modified axial R-transformation 106
Ramond ground state 71
Ramond sector 71
Renormalization group flow 51
irrelevant — 51
marginal — 51
relevant — 51
resolved conifold 219
ring structure 9
sausage 40
Silverstein-Witten method 211
singularity index 57
spectral flow 60, 71
super covariant derivative 42
superconformal algebra 68
superfield 42
chiral — 42
twisted chiral — 43
vector — 42
superpartner 15
superpotential 3, 18, 20
SUSY cone 7
SUSY vacua 4
Theta angle 15
topological twist 201
+ twist 202, 204
− twist 202, 204
A-twist 202, 205
B-twist 203, 207
toric variety 10
tt∗-relation 114
twisted superpotential 3
U(1) charge 20
vacuum manifold 10
versal deformation 58
vortex 108
Witten index 73, 197
Zamolodchikov’s c-theorem 57