Embed Size (px)
Citation preview
http://wrap.warwick.ac.uk
Original citation: Bell, Gavin R., Burrows, Christopher W., Hase, Thomas P. A., Ashwin, M. J., McMitchell, Sean R. C., Sanchez, Ana M. and Aldous, James D.. (2014) Epitaxial growth of cubic MnSb on GaAs AND InGaAs(111). SPIN, 4 (4). 1440025. Permanent WRAP url: http://wrap.warwick.ac.uk/74458 Copyright and reuse: The Warwick Research Archive Portal (WRAP) makes this work of researchers of the University of Warwick available open access under the following conditions. This article is made available under the Creative Commons Attribution 3.0 (CC BY 3.0) license and may be reused according to the conditions of the license. For more details see: http://creativecommons.org/licenses/by/3.0/ A note on versions: The version presented in WRAP is the published version, or, version of record, and may be cited as it appears here. For more information, please contact the WRAP Team at: [email protected]
EPITAXIAL GROWTH OF CUBIC MnSbON GaAs AND InGaAs(111)
GAVIN R. BELL*,‡, CHRISTOPHER W. BURROWS*,§, THOMAS P. A. HASE*,¶,MARK J. ASHWIN*,||, SEAN R. C. MCMITCHELL*,**,ANA M. SANCHEZ*,†† and JAMES D. ALDOUS†,‡‡
*Department of Physics, University of WarwickCoventry CV4 7AL, UK
†London Centre for NanotechnologyUniversity College London
17-19 Gordon Street London WC1H 0AH, UK‡[email protected]§[email protected]¶[email protected]||[email protected]
**[email protected]††[email protected]
Received 30 April 2014Accepted 1 September 2014Published 16 October 2014
The cubic polymorph of the binary transition metal pnictide (TMP) MnSb, c-MnSb, has beenpredicted to be a robust half-metallic ferromagnetic (HMF) material with minority spin gap& 1 eV. Here, MnSb epilayers are grown by molecular beam epitaxy (MBE) on GaAs andIn0:5Ga0:5As(111) substrates and analyzed using synchrotron radiation X-ray di®raction. We ¯ndpolymorphic growth of MnSb on both substrates, where c-MnSb co-exists with the ordinaryniccolite n-MnSb polymorph. The grain size of the c-MnSb is of the order of tens of nanometer onboth substrates and its appearance during MBE growth is independent of the very di®erentepitaxial strain from the GaAs (3.1%) and In0:5Ga0:5As (0.31%) substrates.
Keywords: MnSb; half-metallic ferromagnetism; polymorphism; epitaxy; X-ray di®raction.
1. Introduction
Half-metallic ferromagnetism1,2 is a highly desirableproperty for advanced spintronic devices which re-quire very high spin polarization at the Fermi level
and has been explored in several classes of materi-
als.3,4 The cubic polymorphs of many binary transi-
tion metal pnictides (TMPs), such as MnAs or CrSb,
have been predicted to be half-metallic ferromagnetic
‡Corresponding author.This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the CreativeCommons Attribution 3.0 (CC-BY) License. Further distribution of this work is permitted, provided the original work is properlycited.
SPINVol. 4, No. 4 (2014) 1440025 (8 pages)© The AuthorsDOI: 10.1142/S2010324714400256
1440025-1
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
(HMF) materials with Curie temperatures ðTCÞ wellabove room temperature which is essential if they areto be exploited in any functional device.5–7 ManyTMPs normally adopt a double hexagonal close-packedB81 niccolite structure (abbreviated n-), whichhas ABAC stacking along the c-axis (A=transitionmetal, B,C=pnictogen). The structure of the cubic(c-) TMP polymorphs are predicted to be signi¯cantlyless energetically favorable than the n-polymorphsunder normal growth conditions. However, becausethe TMPs have excellent engineering compatibilitywith many mainstream III–V and Group IV semi-conductor materials,a their cubic polymorphs inparticular have enormous potential as highly spin-polarized contacts for hybrid semiconductor spin-tronics. Combining TMPs with semiconductorsallows the greatest possible °exibility in spintronicdevice design: arbitrary all-epitaxial ferromagnet/an-tiferromagnet/semiconductor heterostructures couldbe grown (e.g., MnSb/CrSb/In1�xGaxAs).
While most band structure calculations madeusing density functional theory (DFT) apply to zerotemperature it has been known for some years thatthe nonzero temperature behavior of HMFs is veryimportant in determining their real-world spin po-larization and hence device performance.8–10 Thepossibility of a critical temperature T � � TC for theonset of a reduction in the spin polarization washighlighted by a recent DFT calculation employingthe disordered local moments (DLM) approach tomodel nonzero temperature band structures.5 Inthat work, both NiMnSb (the canonical HMF semi-Heusler alloy,2 more consistently labeled MnNiSb11)and c-MnSb were studied. NiMnSb has a small mi-nority spin gap 0.5 eV giving a low T � in the regionof 100K. This low value of T � arises from defect-likestates in the minority spin gap, whose spectralweight increases with spin disorder, and easilyreaches the Fermi level with an overall magnetiza-tion reduction of only 5%. Conversely the large mi-nority spin gap � 1 eV of c-MnSb, together with amid-gap Fermi level, means that magnetization re-duction must reach around 20% before minority spinmagnetic disorder states reach the Fermi level,giving aT � probably in excess of 300K.5The prospectfor room-temperature HMF behavior in c-MnSbmakes it a highly promising spintronic material.
Several groups have investigated the growth ofcubic TMP polymorphs by molecular beam epitaxy(MBE) on substrates with square symmetry.6,12,13
The presence of cubic phases has been inferred fromstructural measurements such as X-ray di®raction(XRD) and transmission electron microscopy(TEM), while for some ultra-thin ¯lms (< 1 nm)normally antiferromagnetic materials such as CrAscan show slightly open hysteresis loops, suggestingthe presence of a ferromagnetic cubic polymorph.14
However, detailed DFT work suggested that theseultra-thin ¯lms were not actually cubic in structureand the observed magnetic hysteresis was likely dueto uncompensated spins in a highly strained dis-torted orthorhombic epilayer. Furthermore, for suchultra-thin ¯lms, interpretation of TEM and XRDdata is far from being unambiguous. More recently ithas been claimed that c-MnAs can be grown directlyon InP(001) alongside the ordinary hexagonalphase.15,16 We recently demonstrated the growth ofc-MnSb within n-MnSb ¯lms on GaAs(111), wherethe large grain sizes of the c-MnSb (� 10 nm) madestructural identi¯cation more straightforward.
In this work we extend our MBE growth of MnSbfrom InP,17 GaAs18 and Ge19 to relaxed In0:5Ga0:5As(111) virtual substrates. Furthermore, In1�xGaxAsstructures are attractive for semiconductor spin-tronic applications thanks to their high electronmobility, high Land�e g-factor and low Schottkybarriers.20 Furthermore, the nominal lattice mis-match to n-MnSb is only 0.31% for In0:5Ga0:5As,compared to 3.1% for GaAs. Hence, by comparingthe growth of MnSb grown directly on GaAs withthat on In1�xGaxAs, it is possible to determinewhether the substrate lattice strain plays a role inpolymorph formation. In fact, we will show that theformation of c-MnSb occurs in a very similar fashionon both GaAs and In0:5Ga0:5As substrates suggest-ing that the substrate strain is of minimal impor-tance, at least in this family of substrate materials.These results are consistent with our earlier sug-gestion that the c-MnSb forms epitaxially on top ofn-MnSb in a mixed c-MnSb/n-MnSb polymorphiclayer, rather than in contact with the substrate. Theoutlook for exploiting TMP polymorphs in spin-tronics is brie°y discussed.
aBy \engineering compatibilty" we mean straightforward epitaxial growth, usually favorable interface chemistry, and ease of deviceprocessing and fabrication.
G. R. Bell et al.
1440025-2
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
2. Experimental Details
All MnSb ¯lms were grown by MBE in a dedicatedhome-built chamber, using wafer pieces cut to typ-ically 8� 8mm. After sonication with organic sol-vents and rinsing with deionized water to removedust and cutting debris, substrate wafers werecleaned in vacuo by careful degassing, gentle Ar ionsputtering (500 eV, 0.5�A, 10min) and annealingwithout incident Group V °ux. For the (111)Aoriented substrates this produced the expected(2� 2) reconstructions on both GaAs andIn0:5Ga0:5As. Mn and Sb e®usion cells operatingclose to 857�C and 355�C, respectively produced anSb4:Mn beam equivalent pressure ratio of 6.6:1 asmeasured by a retractable ionization gauge. Thisappears to be the optimum °ux ratio for MnSbpolymorph formation on GaAs(111) and the °uxeswere balanced carefully before each growth. Samplerotation during growth is not possible within ourMBE system. The growth rate was 2.7 nmmin�1
and the substrate temperature was maintained atð415� 5Þ� as measured by a thermocouple on thesample manipulator. Films varying in thicknessfrom 1 nm to 300 nm were grown: here we focus onMnSb ¯lms in the thickness range 100 nm to 300 nmwhere cubic polymorphs can be found.
Both GaAs wafers (moderately n-doped, exactlyoriented) and 400 nm thick In0:5Ga0:5As virtualsubstrates were used. The latter were grown in aseparate Varian MBE system on 50mm (2 inch)GaAs(111)A wafers. In0:5Ga0:5As was depositedunder standard conditions on to a 100 nm GaAsbu®er layer (InGaAs growth rate 16.7 nmmin�1,substrate temperature 500�C, sample rotation) anda protective As cap was deposited after epilayercompletion. Due to the high lattice strain, the surfaceroughness and crystalline mosaic were signi¯cantlyhigher for the In0:5Ga0:5As virtual substrates than forthe GaAs wafers. The virtual substrates weretransferred through air, cut andmounted in the sameway as GaAs wafers and prepared for subsequentMnSb deposition using the same protocols except fora longer pre-anneal to desorb the As cap.
High-resolution XRD experiments were per-formed both in-house and at three synchrotron ra-diation facilitiesb all equipped with multi-circledi®ractometers. In this work, we show synchrotronXRD data only, all of which were obtained with
samples at room temperature under °owing dry ni-trogen using 10 keV photons monochromated usingSi(111) crystals. Angular data, which were recordedin a triple-axis geometry using suitable analyzercrystals, have been reduced to reciprocal latticeunits in which the Qz direction is de¯ned as normalto the substrate surface and hence parallel to a [111]direction. Scans were recorded as a function of bothQz and the orthogonal, in-plane, Qx, directions.Reciprocal space maps (RSMs) were obtained insymmetric (out-of-plane) and asymmetric (in-plane)di®raction geometries. The TEM experiments wereperformed using Jeol ARM-200F and 2100 micro-scopes operating at 200 keV.
3. Results and Discussion
We present TEM images from typical MnSb/In0:5Ga0:5As/GaAs(111)A heterostructures andthen compare XRD data from MnSb/In0:5Ga0:5Asand MnSb/GaAs. In Fig. 1, are shown two bright-¯eld TEM images from a 300 nm thick MnSb ¯lm onIn0:5Ga0:5As(111)A. The overview of the hetero-structure is shown in (a) and the arrows highlightthe GaAs-In0:5Ga0:5As interface (bottom) andIn0:5Ga0:5As-MnSb interface (top). The In0:5Ga0:5Aslayer is highly defective with many threading dis-locations, as expected from the large lattice mis-match (3.7%) with the substrate. However, itssurface is reasonably smooth and has a sharp in-terface with the MnSb overlayer. Since the MnSb isclosely matched to the relaxed In0:5Ga0:5As virtualsubstrate, the density of dislocations is much lowerin the MnSb ¯lm. However, non-niccolite structurescan be discerned in the MnSb ¯lms. In particular,the surface region of this ¯lm shows a granularmorphology with di®erent crystal structures. Thislayer is visible in Fig. 1(b) where the arrow high-lights the interface between a granular surface layerincorporating non-B81 structures and the underly-ing \pure" n-MnSb.
This surface layer is shown in more detail inFig. 2(a). The crystallite size within this layer istypically around 25 nm both laterally and in termsof overall layer thickness. A high-resolution TEMimage obtained on a single crystallite from withinthis surface layer is shown in Fig. 2(b). The atomiccolumns are readily resolved and their symmetry is
bBeamline X22C at NSLS (Brookhaven National Laboratory, USA), the XMaS facility at the ESRF (Grenoble, France) andbeamline I16 at Diamond Light Source, UK.
Epitaxial Growth of Cubic MnSb on GaAs and InGaAs(111)
1440025-3
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
consistent with a cubic structure oriented with (111)planes parallel to the n-MnSb(0001) interface.A Fourier transform of the image is shown inthe inset of Fig. 2(b). This part of the MnSb ¯lmcomprises both c-MnSb grains and a continuation ofthe n-MnSb structure, i.e., it is not a simple epitaxiallayer of c-MnSb on top of n-MnSb. The appearanceof this granular structure is strikingly similar to c-MnSb on n-MnSb grown directly on GaAs(111).5
In Fig. 3, are shown selected synchrotron XRDdata for a typical MnSb/In0:5Ga0:5As/GaAs het-erostructure. Sharp line features on the symmetricdi®ractograms in (a) and (b) are due to multiplescattering from the substrate. The expected sub-strate and virtual substrate peaks are present alongwith n-MnSb with its c-axis out-of-plane. Panel (b)shows a smaller angular range around the lowestorder peaks. The sharpest and most intense peak is
due to the GaAs(111) substrate (the peak is within10�3 Å�1 of the position expected for GaAs at roomtemperature). The virtual substrate appears atlower Qz due to its larger lattice parameter, and thepeak is slightly broadened due to residual strain nearthe GaAs interface. The n-MnSb(0002) peak is in-tense and quite symmetric re°ecting the high crys-talline quality of the epilayer. Lattice parametersderived from ¯tting these peaks are a ¼ ð5:853�0:001ÞÅ and c ¼ ð5:768� 0:001ÞÅ for In0:5Ga0:5Asand n-MnSb, respectively. The value for the c-latticeparameter of n-MnSb is consistent with that foundin ¯lms grown on other substrates5,19 as well as thebulk material.21
Figure 3(c) shows a RSM about the GaAs(422)re°ection (asymmetric di®raction geometry). Sev-eral additional di®raction features appear due to thevirtual substrate and the two polymorphs of MnSb.The n-MnSbð1105Þ and ð1104Þ peaks are present inthe correct ratio derived when compared with in-tensities derived from the structure factors of theideal niccolite structure. The presence of both c-MnSb (422) and (133) re°ections is interesting. Fora single epitaxial orientation of c-MnSb on n-MnSb,
(a)
(b)
Fig. 2. Bright-¯eld (a) and high-resolution (b) TEM imagesfrom a typical MnSb/In0:5Ga0:5As/GaAs heterostructure.Image (a) highlights the grain size of around 25 nm in the non-niccolite structured surface layer, while (b) was obtained fromone of these grains and shows symmetry consistent with c-MnSb(111). A Fourier transform of image (b) is inset.
(a)
(b)
Fig. 1. Bright-¯eld TEM images of a typical MnSb/In0:5Ga0:5As/GaAs heterostructure. Black arrows highlight theprincipal interfaces: in (a) between GaAs and In0:5Ga0:5As(bottom) and between In0:5Ga0:5As and MnSb (top), and in (b)between the two polymorphs of MnSb.
G. R. Bell et al.
1440025-4
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
we would expect to see only one of these peaks in agiven RSM. However, two di®erent epitaxial orien-tations of c-MnSb are present corresponding togrowth on AB versus AC terminated n-MnSbregions. These are mutually misoriented in-plane by60�. The presence of two epitaxial orientationsenhances the grain contrast observed in TEM(Fig. 2). Having both in-plane (4 peaks) and out-of-plane (3 peaks) XRD data for the c-MnSb enables itsstrain state to be investigated. The lattice parameter
is a ¼ ð6:435� 0:007ÞÅ and we cannot detect anysigni¯cant distortion (¯tting three separate latticeparameters always converges to identical valueswithin the experimental error). This value is ap-proximately 1% smaller than previously reported forc-MnSb in ¯lms grown directly on GaAs.5 The epi-taxial strain for c-MnSb, at this lattice parameter, onn-MnSb is 10.2% and with small grain sizes of tens ofnanometer it is clear that substantial epitaxialstrains are to be expected. The lack of observeddistortion probably re°ects the complex structure ofthe polymorphic layer where the c-MnSb grainsembedded within a n-MnSb matrix are not subject tosimple biaxial stress.
The mosaic spread of the ¯lm and substrates wereobtained from rocking curves, Qx scans, at theprincipal XRD peak positions (not shown). The fullwidth at half maximum (FWHM) of the lowestdi®raction order rocking curves is as follows:In0:5Ga0:5As 0.543�, n-MnSb 0.537� and c-MnSb1.210�. Clearly, the mosaic spread of the n-MnSb¯lm is dominated by that of the virtual substrate,which in turn is due to the formation of the mis¯tdislocations at the GaAs/InGaAs interface shown inFig. 1(a). The c-MnSb gives rise to much broaderrocking curves due to mismatch-induced mosaic andthe smaller grain size, consistent with the TEM andXRD data.
We show in Fig. 4 typical XRD data for MnSb¯lms grown directly on GaAs(111). The standarddi®ractogram contains similar peaks to that ofFig. 3(a) with only the virtual substrate peaksmissing. In particular, strong substrate and n-MnSbpeaks appear and, despite the very di®erent sub-strate lattice parameter, c-MnSb is also present. Thelowest order family of peaks is shown in Fig. 3(b).Some additional features appear compared to thevirtual substrate samples. The peak labeled 1, lyingbetween the c-MnSb and GaAs(111) peaks, is dueto GaSb(111). This sometimes occurs at the GaAs/n-MnSb interface due to Ga droplets resulting fromthe surface preparation of GaAs. The presence ofsuch alternative III–V compounds at the substrate–¯lm interface which can also include InAs formingduring MnAs growth on InP, can make identi¯ca-tion of cubic TMP phases ambiguous without care.The peak labeled 2, between GaAs(111) andn-MnSb(0002), is due to n-MnSbð1101Þ. Smallcrystallites of this alternative epitaxial orientationcan appear in MnSb ¯lms on GaAs and have alsobeen observed for low-strain NiSb ¯lms on GaAs
Fig. 3. Synchrotron XRD data for a typical MnSb/In0:5Ga0:5As/GaAs heterostructure. Panel (a) shows a standardout-of-plane (symmetric) di®ractogram. All the main peaks arelabeled and correspond to the expected GaAs and In0:5Ga0:5As(111)-type re°ections and n-MnSb(0002) re°ections. A clear setof additional weaker peaks is assigned to c-MnSb. Panel (b)shows a more detailed view of the group of lowest order re°ec-tions. Panel (c) is an RSM in asymmetric geometry centered onthe GaAs(422) re°ection with di®raction features labeled.
Epitaxial Growth of Cubic MnSb on GaAs and InGaAs(111)
1440025-5
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
(111).22 However, the key point is the presence ofc-MnSb on both GaAs and In0:5Ga0:5As substrates.
A RSM obtained in symmetric di®raction condi-tions is shown in Fig. 4(c). This data reinforces theinterpretation of the principal features of Fig. 4(b).The n-MnSb(0002) and GaAs(111) features are in-tense and symmetric. The peak assigned to GaSb isalso symmetric and sharp, re°ecting good epitaxydirectly on the GaAs substrate; the lattice parame-ter of ð6:104� 0:003ÞÅ corresponds very well tobulk GaSb (0.13% expanded). The peak assigned ton-MnSbð1101Þ is much weaker and less distinctlysymmetrical as expected. The c-MnSb(111) featureis quite broad in Qx as expected for a granular phasewith higher mosaic spread; the corresponding rock-ing curve has FWHM of around 1.1�, similar to itscounterpart on In0:5Ga0:5As. The other interesting
feature of the c-MnSb peak is a broad shoulderextending to lower Qz, also very clear in the dif-fractogram. This can be explained by the presence ofvery small crystallites in varying high strain statesthroughout the mixed n-MnSb/c-MnSb polymor-phic layer.
4. Conclusions and Outlook
In this work, MnSb thin ¯lms have been grown byMBE on (111)-oriented GaAs and In0:5Ga0:5Assubstrates with lattice mismatches of 3.1% and0.31%, respectively. In both cases, high-qualityn-MnSb ¯lms can be grown using a suitable °uxratio [6.6:1 excess Sb] and a substrate temperatureof ð415� 5Þ�C. Under these conditions, all ¯lmsshowed the presence of the technologically impor-tant c-MnSb polymorph.
Structural characterization by TEM and XRD isconsistent with a model of small c-MnSb grainswithin a mixed polymorphic layer (c- plus n- withsometimes some wurtzite5). The c-MnSb is epitaxialon and within the n-MnSb matrix, and is under highepitaxial stress. However, this does not appear to besimple biaxial stress due to the mixed nature of thepolymorphic layers. There is evidence for a widerange of strain states within the c-MnSb (bothamong grains in a single ¯lm and from ¯lm to ¯lm)which most likely depends principally on crystallitesize. The in-plane epitaxy of the c-MnSb corre-sponds to (111) planes lying on the n-MnSb(0001)but with two possible in-plane orientations sepa-rated by 60�, most likely re°ecting matching to AB-layer or AC-layer terminated n-MnSb. FurtherTEM work is under way to understand in moredetail the strain states and epitaxial relationship ofthe c-MnSb in n-MnSb.
The key point of this paper is that the presence ofthe c-MnSb polymorphic layers does not depend onwhether the n-MnSb is sitting on In0:5Ga0:5As(111)or GaAs(111) substrates. Thus, we can rule out thein°uence of epitaxial mismatch at the substrate–¯lminterface as the precursor for polymorphic growth, atleast in this pair of epitaxial systems. We have neverobserved c-MnSb grown directly on any III–V or Gesubstrate, even for (100) substrates (not discussedhere). Hence, we do not believe that c-MnSb isreadily stabilized directly on semiconductor sub-strates. At present, the precise mechanism for poly-morph nucleation within the n-MnSb ¯lm remainsunknown. It is plausible that the polymorphic
Fig. 4. Synchrotron XRD data for a typical MnSb/GaAs thin¯lm. Panel (a) shows a standard out-of-plane (symmetric) dif-fractogram. All the main peaks are labeled and correspond tothe expected GaAs and In0:5Ga0:5As(111)-type re°ections andn-MnSb(0002) re°ections. A clear set of additional weakerpeaks is assigned to c-MnSb. Panel (b) shows a more detailedview of the group of lowest order re°ections. Panel (c) is anRSM in symmetric geometry showing the lowest order re°ec-tions; other features of this RSM are discussed in the text.
G. R. Bell et al.
1440025-6
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
transition is related to the growth kinetics on par-ticular surface reconstructions of the n-MnSb(0001)23; layer growth kinetics and nanostructureself-assembly are long known to be a®ected by sur-face reconstructions in ordinary III–V semiconductorMBE.24 With a better understanding of the TMPpolymorph nucleation mechanism, improved preci-sion in the MBE process should allow c-MnSb to beformed more reliably and with tighter control overcrystallite sizes. Vicinal substrates may be usefulinsuppressing the twin epitaxial orientation of thec-MnSb.
The half-metallic cubic TMP polymorphs sharekey features with the most promising spintronicHeusler alloys, principally high Curie temperatureand large minority spin gap. Considering also theirgood epitaxial compatibility with mainstreamsemiconductors, these qualities bode well for appli-cations in room-temperature spintronics. Becausethe c-MnSb crystal structure only contains two face-centered cubic sublattices rather than the four of afull Heusler alloy, there is probably less scope foratomic-scale disorder to disrupt half-metallicity.However, the growth of cubic TMPs is less advancedthan that of Heusler alloys and it is not clear whatkey defect properties may be important in largecrystallites. Further DFT calculations of defectivecubic TMPs would be valuable. Presently, ourexperiments on cubic TMPs focus on the crystallitesizes and strain states, which are also issues sharedwith polycrystalline Heusler alloy ¯lms. Experi-mental demonstration of true half-metallicity in c-MnSb remains challenging due to the co-existence ofn-MnSb but surface-speci¯c techniques such as spin-resolved photoemission may initially be most useful,since surface layers dominated by the cubic poly-morph can already be grown. Polymorphic layersthick enough to dominate transport in a point con-tact Andreev re°ection (PCAR) experiment wouldallow more direct comparison to Heusler alloys forwhich a signi¯cant body of PCAR data exists. Theability to control the strain states of highly spin-polarized cubic TMP layers formed on or withinordinary niccolite-structured layers o®ers excitingnew possibilities in the ¯eld of hybrid semiconductorspintronic devices.
Acknowledgments
This work was supported by EPSRC, UK, undergrant numbers EP/I00114X/1 and EP/K032852/1,
by the USA's Department of Energy (DE-AC02-98CH10886) and by Diamond Light Source. XMaSis a mid-range facility funded by EPSRC. We aregrateful to the beamline sta® of I16, XMaS andX22C, and to R. Johnston and S. York for experttechnical support in Warwick.
References
1. P. Dowben, J. Phys. Condens. Matter 19, 310301(2007).
2. R. A. de Groot, F. M. Mueller, P. G. van Engen andK. H. J. Buschow, Phys. Rev. Lett. 50, 2024 (1983).
3. M. I. Katsnelson, V. Y. Irhkin, L. Chioncel, A. I.Lichtenstein and R. A. de Groot, Rev. Mod. Phys.80, 315 (2008).
4. B. Hillebrands and C. Felser, J. Phys. D: Appl. Phys.42, 080301 (2009).
5. J. D. Aldous, C. W. Burrows, A. M. S�anchez,R. Beanland, I. Maskery, M. K. Bradley, M. dosSantos Dias, J. B. Staunton and G. R. Bell, Phys.Rev. B 85, 060403(R) (2012).
6. H. Akinaga, T. Manago and M. Shirai, Jpn. J. Appl.Phys. 39, L1118 (2000).
7. J. E. Pask, L. H. Yang, C. Y. Fong, W. E. Pickettand S. Dag, Phys. Rev. B 67, 224420 (2003).
8. M. Ležaić, P. Mavropoulos, J. Enkovaara, G. Bihl-mayer and S. Blügel, J.Mater. Sci. 97, 026404 (2006).
9. R. Skomski, J. Appl. Phys. 103, 07D714 (2008).10. P. A. Dowben and R. Skomski, J. Appl. Phys. 95,
7543 (2004).11. T. Graf, S. S. P. Parkin and C. Felser, Prog. Solid
State Chem. 39, 1 (2011).12. J. H. Zhao, F. Matsukura, E. A. K. Takamura, D.
Chiba andH.Ohno,Appl. Phys. Lett. 79, 2776 (2001).13. V. H. Etgens, P. C. deCamargo, M. Eddrief,
R. Mattana, J. M. George and Y. Garreau, Phys.Rev. Lett. 92, 167205 (2004).
14. S. J. Hashemifar, P. Kratzer and M. Sche®ler, Phys.Rev. B 82, 214417 (2010).
15. H. Oomae, S. Irizawa, Y. Jinbo, H. Toyota,T. Kambayashi and N. Uchitomi, J. Cryst. Growth378, 410 (2013).
16. H. Oomae, J. T. Asubar, S. Nakamura, Y. Jinbo andN. Uchitomi, J. Cryst. Growth 338, 129 (2012).
17. S. A. Hat¯eld and G. R. Bell, J. Cryst. Growth 296,165 (2006).
18. J. D. Aldous, C.W. Burrows, I. Maskery,M. S. Brewer,T. P. A. Hase, J. A. Du®y, M. R. Lees, C. S�anchez-Hanke, T. Decoster, W. Theis, A. Quesada, A. K.Scmhid and G. R. Bell, J. Phys-Condens. Matter 24,146002 (2012).
19. C. W. Burrows, A. Dobbie, M. Myronov, T. P. A.Hase, S. B.Wilkins,M.Walker, J. J.Mudd, I.Maskery,
Epitaxial Growth of Cubic MnSb on GaAs and InGaAs(111)
1440025-7
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
M. R. Lees, C. F. McConville, D. R. Leadley and G. R.Bell, Cryst. Growth Des. 13, 4923 (2013).
20. P. J. Simmonds, S. N. Holmes, H. E. Beere, I. Farrer,F. S¯gakis, D. A. Ritchie and M. Pepper, J. Vac. Sci.Technol. B 27, 2066 (2009).
21. B. T. M. Willis and H. P. Rooksby, Proc. Phys. Soc.B 67, 290 (1954).
22. J. D. Aldous, C. W. Burrows, I. Maskery, M. Brewer,D. Pickup, M. Walker, J. J. Mudd, T. P. A. Hase,
J. A. Du®y, S. Wilkins, C. S�anchez-Hanke and G. R.Bell, J. Crystal Growth 357, 1 (2012).
23. S. A. Hat¯eld and G. R. Bell, Surf. Sci. 601, 5368(2007).
24. B. A. Joyce, D. D. Vvedensky, G. R. Bell, J. G. Belk,M. Itoh and T. S. Jones, Matter. Sci. Eng. B-Solid67, 7 (1999).
G. R. Bell et al.
1440025-8
SPIN
201
4.04
. Dow
nloa
ded
from
ww
w.w
orld
scie
ntif
ic.c
omby
UN
IVE
RSI
TY
OF
WA
RW
ICK
on
11/3
0/15
. For
per
sona
l use
onl
y.
