Upload
andreea-prodescu
View
19
Download
0
Tags:
Embed Size (px)
Citation preview
5/19/2018 1-s2.0-S0109564114000207-main.pdf
1/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10
d ent al m at eri al s x x x ( 2 0 14 ) xxx.e1xxx.e10
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: www.int l .e lsevierheal th.com/ journals/dema
Metallurgical and interfacial characterization of
PFM CoCr dental alloys fabricatedvia casting,
milling or selective laser melting
Y.S. Al Jabbaria,b,, T. Koutsoukisa, X. Barmpagadakic, S. Zinelisd,a
a Dental Biomaterials Research and Development Chair, College of Dentistry, King Saud University, P.O. Box 60169,
Riyadh 11545, Saudi Arabiab
Department of Prosthetic Dental Sciences, College of Dentistry, King Saud University, P.O. Box 60169, Riyadh11545, Saudi Arabiac Private Practice, Githiou 81 Str, Pireas 18544, Greeced Department of Biomaterials, School of Dentistry, University of Athens, 2 Thivon Str, Goudi 11527, Athens, Greece
a r t i c l e i n f o
Article history:
Received 29 April 2013
Received in revised form
25 September 2013
Accepted 16 January 2014
Available online xxx
Keywords:
SLM
Fabrication techniques
CoCr dental alloy
Metalceramic interface
a b s t r a c t
Objectives. Bulk and interfacial characterization of porcelain fused to metal (PFM) CoCr
dental alloys fabricated via conventional casting, milling and selective laser melting.
Methods.Three groupsof metallic specimensmade of PFMCoCrdental alloys were prepared
using casting (CST), milling (MIL) and selective laser sintering (SLM). The porosity of the
groups was evaluated using X-ray scans. The microstructures of the specimens were evalu-
ated via SEM examination, EDX and XRD analysis. Vickers hardness testing was utilized to
measure the hardness of the specimens. Interfacial characterization was conducted on the
porcelain-covered specimens from each group to test the elemental distribution with and
without the application of INmetalbond. The elemental distribution of the probed elements
was assessed using EDX line profile analysis. Hardness results were statistically analyzed
using one-way ANOVA and HolmSidaks method (= 0.05).
Results. X-ray radiography revealed the presence of porosity only in the CST group. Different
microstructures were identified among the groups. Together with the phase matrix, a sec-
ond phase, believed to be the Co3Mo phase, was also observed by SEM and subsequent XRD
analysis. Cr7C3and Cr23C6carbides were also identified via XRD analysis in the CST and MIL
groups. The hardness values were 32012HV, 2975HV and 37110HV, and statistically
significant differences were evident among the groups.
Significance. The microstructure and hardness of PFM CoCr dental alloys are dependent
on the manufacturing technique employed. Given the differences in microstructural and
hardness properties among the tested groups, further differences in their clinical behavior
are anticipated.
2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
Correspondingauthor at: Dental Biomaterials Research and Development Research Chair, College of Dentistry, King Saud University, P.O.Box 60169, Riyadh 11545, Saudi Arabia. Tel.: +966 1 4698312; fax: +966 1 4698313.
E-mail addresses: [email protected], [email protected] (Y.S. Al Jabbari).
0109-5641/$ see front matter 2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.dental.2014.01.008
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://www.sciencedirect.com/science/journal/01095641http://www.intl.elsevierhealth.com/journals/demamailto:[email protected]:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008mailto:[email protected]:[email protected]://www.intl.elsevierhealth.com/journals/demahttp://www.sciencedirect.com/science/journal/01095641http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.0085/19/2018 1-s2.0-S0109564114000207-main.pdf
2/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10xxx.e2 dental mater i al s xxx ( 2 014 ) xxx.e1xxx.e10
1. Introduction
Technological developments have led to the implementa-
tion of novel manufacturing processes in everyday dental
practice. In recent decades, digitalized technologies have been
employed for the production of metallic structures, mainly in
prosthetic dentistry [15]. These technologies can be classifiedas based on subtractive manufacturing, such as the milling
of pre-manufactured materials assisted by computer-aided
design/computer-aided manufacturing (CAD/CAM) systems
[1,2,57], or on additive manufacturing, such as the recently
developed selective laser melting (SLM) technique [3,4,810].
Although CAD/CAM has long been directly associated with
the milling procedure in dental literature, it should be
mentioned that SLM is also classified as CAD/CAM technol-
ogy.
As a recently introduced technique, SLM has attracted
the worldwide interest of research groups for the manufac-
turing of dental metallic structures. In prosthetic dentistry,
most studies on SLM have focused on CoCr dental alloys[3,8,9,1120]; significantly fewer studies have been per-
formed on Ti alloys [4,21] or the costly precious alloys
[3,12]. These studies have primarily focused on the evalua-
tion of the marginal and/or internal fit of the restorations
[3,6,9,13,1517,20], whereas other studies have tested the
bond strength with dental porcelain [8,18,21], internal poros-
ity [4], effect of surface treatments on microroughness [14]
and electrochemical properties [19]. However, the compara-
tive analyses of specific properties among metallic structures
made using SLM and conventional techniques are limited
[13,21]; thus, the effect of the SLM technique on mechani-
cal, electrochemical, microstructural and other properties is
still unknown. Given the large differences in the manufac-turing process between casting, which uses the complete
melting and overheating of casting materials, the milling of
a prefabricated metal block and SLM of a fine metallic pow-
der, large differences in microstructural characteristics are
anticipated. These microstructural differences may also dif-
ferentiate the interfacial characterizationof metallic elements
at the metalporcelain interface. Although common for other
prosthetic dental alloys, interfacial analyses of CoCr alloys
cast or milled with porcelain are still absent from the dental
literature. Therefore, the aim of this study was to metal-
lurgically and interfacially characterize CoCr dental alloys
prepared by casting, milling and SLM techniques. The null
hypothesis was that there would be significant dissimilari-ties among the groups prepared by different manufacturing
techniques.
2. Materials and methods
2.1. Specimenpreparation
Three groups (CST, MIL, and SLM) were prepared using CoCr
dental alloys as indicated by the manufacturers. The spec-
imens of the CST group were fabricated by the traditional
casting technique using CoCr raw material; those of the MIL
groupwere milledoff a prefabricated block andthe specimens
of the SLM group were fabricated by the SLM technique using
CoCr mixed powder. The brand names, manufacturers and
elemental composition of the alloys tested are presented in
Table 1.
In the CST group, 12 wax patterns (IQ sticks, Yeti Den-
tal, Engen, Germany) were invested with phosphate-bonded
investment (GC Stellavest, GC Europe NV, Belgium) with
dimensions of 0.5mm3 mm25mm. The mold was pre-heated at 910 C and cast with VI-COMP alloy at 1450C using
a centrifugal casting machine (Ducatron S3, UginDentaire,
Seyssins, France).The mold was left to cool down to room tem-
perature and the specimens were then divested and cleaned
by sandblasting with alumina particles (100m).
A prefabricated block of a commercial CoCr dental
alloy (Okta-C) was milled to fabricate a dental restoration
using the Organical Multi Milling/Grinding CAD/CAM system
(R+K CAD/CAM Technologie, Berlin, Germany). A rectangular-
shaped wax pattern was digitized and the specimens were
cut to their final dimensions (0.5mm3 mm25mm) using
the Organical Multi Milling/grinding machine (R+K CAD/CAM
Technologie).
Thelaser-sinteredspecimens were prepared fromcommer-
cial CoCr powder (ST2725G) using a dental laser sintering
device (PM 100 Dental System, Phenix Systems, Clermont-
Ferrard, France) equipped with a 500W Yb-fiber laser, at a
temperatureof 1650C;the laser systemhadthe ability toweld
across a controlled (XY)-axis coordinate system with a Z-axis
tolerance of0.0254 mm. The CoCr powder was applied to a
stainless steel plate and was laser-sintered upwards in subse-
quentlayers after a 20-m-thick layer was completeduntil the
final product was generated. Following laser sintering, the sin-
tered parts were cooleddown to furnace temperature. In total,
12 specimens with dimensions of 0.5mm3 mm25mm
were fabricated using this technique.
2.2. X-ray testing
All specimens of all groups were then examined for internal
porosity using a dental X-ray unit (Orix 70, Ardet, Milan, Italy)
operating at 70kV and 5mA with an exposure time of 15s.
Digital images were collected from all specimens, and the X-
ray images were assessed by the naked eye.
2.3. SEM-EDX characterization
For microstructural characterization, three specimens of each
group were examined using a SEM (JSM 6610 LV, Jeol Ltd.,
Tokyo, Japan) equippedwith an X-ray EDS microanalysis (EDX)
unit (Oxford Instruments, Abingdon, UK). All examined sur-
faces were ground using SiC paper (2202000 grit) under
continuous water cooling andpolished in a grinding polishing
machine (Ecomet III, Bueler, Lake Bluff, IL, USA) using a dia-
mond paste (DP Paste, Struers, Copenhagen, Denmark). The
specimens were then cleaned in an ultrasonic water bath for
5 min. Specimens fabricated by the casting, milling or SLM
technique (0.5 mm3 mm25mm) were examined on the
3 mm25mm surface. The examined surfaces were imaged
using a backscattered electron detector (BSE) under an accel-
erating voltage of 30 kV and a beam current of 48A at a
working distance of 10mm. The elemental composition was
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.0085/19/2018 1-s2.0-S0109564114000207-main.pdf
3/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10
dental mater i al s xxx ( 20 14 ) xxx.e1xxx.e10 xxx.e3
Table 1 Brand names, elemental compositions and coefficients of thermal expansion (CTE) of the tested alloys, asprovided by themanufacturers.
Brand names Manufacturer Co Cr Mo Si Mn Fe
Okta-C Sae Dental Products
Inc. Bremerhaven,
Germany
61.6 30.0 6.5 0.8 0.8 N/A
ST2725G SINT-TECH, Riom,France. Bal (max 62.5) 29 5.5
5/19/2018 1-s2.0-S0109564114000207-main.pdf
4/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10xxx.e4 dental mater i al s xxx ( 2 014 ) xxx.e1xxx.e10
Table 2 Firing schedules of the veneering procedure for GC Initial MC porcelain (according to themanufacturersinstructions).
Product name Pre-heating temp. (C)
Drying time(min)
Heating rate(C/min)
Vacuum Final temp. (C) Holding time(min)
INmetalBond 550 6 80 Yes 980 1
Opaque 550 6 80 Yes 940 1
Dentin 580 6 55 Yes 900 1Glaze 480 2 45 No 850 1
Fig. 1 RepresentativeX-ray images of the (A) CST, (B) MIL and (C) SLM groups. Porositywas identified only in the CST group.
enriched mainly in Mo and less in Cr, which was also the
case for the Am phase. In the MIL group, the Bm was heavily
enriched in Mo but depleted in Cr relative to the matrix con-
tent. All of the dispersed phase primarily had low Co content
(Table 3).
3.3. XRD analysis
Following the XRD analysis, the diagrams recorded from all
groups were indexed as presented in Fig. 3. In addition to the
face-centered cubic (fcc) phase of Co and Cr, the hexagonal
close-packed (hcp) Co3Mo phase was identified in all speci-
mens. Additionally, Cr7C3and Cr23C6carbides were identified
in the CST and MIL groups.
3.4. Hardness
In total, 12 measurements of Vickers hardness were acquired
for the specimens of each dental alloy. The mean values
and the standard deviations were calculated as 32012 HV,2975HV and 37110HV for the CST, MIL and SLM groups,
respectively. Statistically significant differences were found
among all the groups tested (p< 0.05).
3.5. Interfacial characterization
Representative BEI from the interface of all materials tested
are presented in Fig. 4, and they show a well-formed and
defined interface between the CoCr alloy and the opaque or
INmetalbond. Image contrast revealed that the opaque is a
Fig. 2 Representative BE images of the microstructure observed in (A) cast, (B) milled and (C) SLM specimens. Note the
dispersion of a second phase in the cast and milled specimens (white contrast indicated with white arrows) and the
absence of such a phase in the SLM specimen. Precipitates of a third phase were also detected in the milled specimen (BM
in Fig. 2B). Porosity can also be observed in the images.
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.0085/19/2018 1-s2.0-S0109564114000207-main.pdf
5/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10
dental mater i al s xxx ( 20 14 ) xxx.e1xxx.e10 xxx.e5
Fig. 3 Indexed XRD diagrams of all groups included in the study. SLM showed only the presence of(Co,Cr) and Co3Mo
phases, in contrast to the CST and MIL groups, in which two Cr carbides were also identified.
Fig. 4 Representative BE images of the (A) metalINmetalbondopaque and (B) metalopaque interfaces. EDX area analyses
were performed as indicated on the images.
Table 3 Quantitative results of EDX analysis (wt%) of alloys (area analysis) and different phases (spot analysis) based onthe mean atomic contrast of BEI (Fig. 2). For the sake of clarity, the accuracy intervals of the EDX analysis are given onlyfor the alloy composition.
Element CST MIL SLM
Alloy Mc Ac Alloy Mm Am Bm Alloy
Co 59.9 [56.962.9] 62.7 46.9 59.1 [56.162.1] 60.2 47.9 43.2 62.7 [59.665.8]
Cr 32.2 [30.633.8] 31.7 35.8 33.1 [31.434.8] 33.3 37.9 26.9 29.2 [27.730.7]
Mo 6.2 [5.66.8] 4.3 15.3 6.0 [5.46.6] 4.9 12.3 28.2 6.3 [5.76.9]
Si 1.0 [0.81.2] 0.9 1.3 0.9 [0.41.3] 0.6 1.0 0.6 0.9 [0.41.3]
Mn 0.5 [0.20.7] 0.4 0.7 0.4 [0.20.6] 0.5 0.5 0.8 0.8 [0.41.2]
Fe 0.2 [0.10.3] 0.2 0.1 0.5 [0.20.7] 0.5 0.4 0.3 0.1 [0.10.2]
Table 4 EDX area analysis (wt%) of the INmetalbond and the opaque.
Material Element
O Si Ti Zr Fe Na Al K Ca Mg Zn
INmetalbond 22.9 19.4 29.2 12.3 1.7 3.7 3.4 6.6 0.8
Opaque 24.7 21.9 1.5 30.0 0.2 3.1 5.7 9.3 2.3 0.3 1.0
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.0085/19/2018 1-s2.0-S0109564114000207-main.pdf
6/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10xxx.e6 dental mater i al s xxx ( 2 014 ) xxx.e1xxx.e10
Fig. 5 EDX line profile analysis performed across the interface between the alloy and the opaque (A, C and E) or
INmetalbond (B, D and F). The white horizontal line represents the scanning route. The plateau observed in the Cr profile is
indicated by an arrow.
multiphase material, whereas the INmetalbond exhibited a
matrix with a dispersed phase that had a higher mean atomic
number. The elemental composition of both materials as
determined by EDX analysis is presented in Table 4. INmetal-
bond showed increased Ti content relative to the opaque,
mainly at the expense of Zr content. The results of the EDX
line profile analysis, recorded across either the metalopaque
interface or the metal-INmetalbond interface, are presented
in Fig. 5. All probed elements demonstrated a steady decrease
(Co) or increase (Si, K and Ti) from the alloy toward the porce-
lain or INmetalbond. An exception to this general trend was
observed for the Cr profile line in the groups with opaque,
which demonstrated an obvious plateau at the interface, as
indicated by the arrows in Fig. 5A, C and E.
4. Discussion
This study focused on the microstructural and interfacial
characterization of CoCr PFM alloys fabricated using casting,
milling and selective laser sintering. All of these manu-
facturing processes are currently used for the production
of dental prosthetic restorations. Based on the data pre-
sented above, the null hypothesis must be rejected, as
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.0085/19/2018 1-s2.0-S0109564114000207-main.pdf
7/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10
dental mater i al s xxx ( 20 14 ) xxx.e1xxx.e10 xxx.e7
significant dissimilarities were found among the groups
tested, apart from the interfacial distribution of metallic
elements at the metalceramic joint with or without INmetal-
bond.
Although great effort was invested in finding a single alloy
that could be used for the three manufacturing processes,
suchan alloy was notavailable at thetime of this study. There-
fore, the best matching nominal compositions as presentedin Table 1 were selected from a large pool of commercially
available CoCr dental alloys. EDX results indicated similar
values for all elements among different groups within the rel-
ative internal error of the quantitative standardless analysis
(Table 2), although a slightly higher Co/Cr ratio was identified
for the SLM group relative to the rest of the alloys. Although
Fe was identified in all groups, it is only known to exist in
the nominal composition of the ST2724G alloy. Interestingly,
the presence of Fe in the nominal composition of the ST2724G
CoCr alloy hasbeen both reported[9] and not reported[14,15].
An explanation for this inconsistency may be that the manu-
facturers slightly modified the formulations without changing
the brand name. Alternatively, the presence of these traces
may be attributed to the contamination of the raw materials
and/or contamination during the manufacturing process of
the alloys.
BEI demonstrated microstructural differences among the
tested groups. The CST group showed the typical cast struc-
ture of a CoCr dental alloy [2225], as the specimens were
composed of a matrix (Mc) and a heavier dispersed phase
(Ac) that occupy the interdendritic spaces. Although the den-
dritic structure was not clearly outlined, this magnification
was kept constant among the different groups for compar-
ison purposes. This phase was expected to be rich in the
heavier available elements, primarily Mo (Table 1), because of
its lighter contrast relative to the matrix in the BEI mode, and
this phenomenon was confirmed by EDX analysis (Table 2).
Because of its different composition and crystal structure, the
formation of this bulky phase is considered to be undesirable
because it increases the brittleness and deteriorates the cor-
rosion resistance of the alloy by removing Mo from the solid
solution [26]. A second phase (Am) with a higher mean atomic
number relative to the matrix (Mm) was also observed in the
microstructure of the MIL group (Fig. 2B), similarly to the cast
material, but with a more bulky morphology. Similarly to the
CST group, Am demonstrated increased Mo and Cr content,
andan even higherMo content was detectedin the third phase
Bm(Table 2), which agrees with previous observations in sim-
ilar alloys [26]. The formation of this phase, which preferably
nucleates and coarsens at the interface of the second phase
and the matrix, promotes the further depletion of Mo in the
matrix, which implies the deterioration of the corrosion resis-
tance [27]. Mo is also added to CoCr alloys to achieve a finer
grain structure, thereby enhancing the mechanical proper-
ties of the material [26]. This favorable effect is diminished
when Mo is segregated in Mo-rich compounds rather than
being dispersed within the matrix. In contrast to the CST and
MIL groups, no mean atomic contrast was revealed for the
SLM material a finding which is in agreement with a recently
published study [25]. This difference implies a completely
different solidification and/or thermomechanical history of
the tested groups [22,27,28].
XRD analysis results (Fig. 3) indicated that the microstruc-
ture of all groups consisted of the face-centered cubic (fcc)
phase, which mainly comprised Co and Cr which is in agree-
ment with a previous study [29]. As an allotropic element,
Co has a fcc crystal structure above temperatures of approx-
imately 417C and a hexagonal crystal packing (hcp) below
this temperature, but in CoCr alloys, the fcc structure is
maintained at room temperature because of the low fcchcp
transformation rate. The fcc phase should be attributed to
the matrix of the microstructure for the CST and MIL groups
(Fig. 2A andB), whereas theCo3Mo phase, whichhas also been
identified in previous studies of cast alloys [27,30] as a Mo-
rich compound (Table 2), should be attributed to the dispersed
phases Ac and Am of the CST and MIL groups, respectively.
Although this phase was identified by XRD analysis in the
SLM group, no mean atomic number contrast was found in
BEI, possibly because of therapid solidification of fused metal-
lic particles that led to a very fine phase size that was below
the resolution of backscattered electron imaging. Additionally,
the previously found carbides [22] Cr23C6and Cr7C3were also
identified by XRD analysis only in the CST and MIL groups
(Fig. 3). Even when the manufacturer produces a raw mate-
rial with a low C content, carbide formation may occur during
the manufacturing process and thus affect the nominal prop-
erties [24,26,28,3033] of the produced alloys. In addition to
the detected carbides, it is possible that other carbides could
have formed in the microstructure of the examined materi-
als [24,30], but the amount of these carbides is most likely
below the detection limit of the technique. The application of
advanced electron microscopy techniques, such as transmis-
sion electron microscopy (TEM), could provide further crucial
information, especially on the microstructural characteriza-
tion and the formation mechanism of the developed phases.
Despite the detrimental effect of internal porosity, which
is a common complication of the casting procedure for CoCr
alloys [34,35], the significantlyhigher hardness observedin the
CST group (32012HV) relative to the MIL group (2975HV)
may be attributed to the finer distribution of the dispersed
phase (Fig. 2). However, the increased hardness observed in
the SLM group (37110HV) could be attributed to the sinter-
ing technique,which not onlydiminishes undesirable porosity
but also provides a much more fine-grained structure [1]. This
is in agreement with a recently published study indicating
that SLM technique provides CoCr alloys with enhanced ulti-
mate tensile strengthand elongationcompared to casting[29].
Additionally, the presence of residual stresses during sinter-ing is another possible explanation for the increased hardness
in the SLM [36,37].
The surface of the three tested groups was intentionally
polished using a 1-m diamond paste prior to the veneering
procedure. Despite the manufacturers recommendations for
surface roughening before porcelain application, a flat surface
is required to obtaina very small area forEDX line profile anal-
ysis (Fig. 5). In addition, the retained alumina fragments on
metallic surfaces after sandblasting[38] mask the real distri-
bution of Al and O at the interface. Interfacial analysis with
BE imaging revealed that both the opaque and INmetalbond
adhered well to the metallic substrates for all groups tested
(Fig. 4), thereby providing a continuous interface with thesubstrate. In addition to the typical layering procedure with
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.0085/19/2018 1-s2.0-S0109564114000207-main.pdf
8/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10xxx.e8 dental mater i al s xxx ( 2 014 ) xxx.e1xxx.e10
Fig. 6 Standard free energy of different reactions with O for Co, Cr, Si and Ti, commonly known as Ellinghams Diagrams.
A lower position reflects higher chemical affinity of the element with O. Both axes have the same range for comparison
purposes.
opaque directly applied on the metallic surface, various prod-
ucts have been introduced to themarket that claim to increase
either the bond strength or the esthetics of PFM restorations
[39]. Specifically, INmetalbond is advertised as a material that
aims to block the escaping metal oxides and neutralize differ-
ences in the thermal expansion coefficient between porcelain
and metal. The application of this bonder is optional and
is recommended by the company for all precious and non-
precious PFM dental alloys.
The line profile analysis did not reveal any concentra-
tion gradient at the interface, but interestingly, the Cr profile
demonstrated a plateau at the interface for all groups coveredwith opaque, but not with INmetalbond. An explanation for
this behavior may include the chemical affinity of the involved
elements to O. This information is provided by typical Elling-
ham diagrams presented in Fig. 6 [40], which show the free
energy of a given oxide as a function of temperature. A lower
position reflects higher chemical affinity of an element to O
and thus a higher tendency to react with O. Given that Cr has
a higherchemical affinity to O than Co andMo (the latteris not
shown in Fig. 6), it reacts first, thereby increasing the contribu-
tion of Cr compounds at the interface. It should be mentioned
that even if the other metal reacts first because of kinetic rea-
sons, Cr ions can still reduce the oxides of the other metals
based on themetallothermic reaction, in whicha metal with a
higher chemicalaffinityto O (MH) reduces theoxides of metals
with a lower affinity to O (ML) based on the following general
formula:
(MH) + (ML)xOy (MH)Oy+x(ML)
Surprisingly, this behavior vanished when the INmetal-
bond was applied for all groups tested, which may be
associated with a higher chemical affinity of Ti with oxygen,
as Ti is the predominant element of INmetalbond. Of course,
this mechanism requires that Ti be capable of reacting with
O or participating in metallothermic reactions. In contrast,
although Si has a similar chemical affinity to O as Ti, it is
widely known that Si is anchored in the glassy matrix as an
oxide [41], and thus its capability for further reaction is elim-
inated. Although this thermodynamic approach agrees with
the data presented above, it is worthwhile to note that these
calculations represent the reaction of pure elements with O
and must be corrected when they are used for alloying ele-
ments in different alloys. However, such data are not available
and their calculation requires extensive research. In conclu-
sion, it is clear that the application of INmetalbond affects the
interfacial distribution of involved elements, and therefore its
effect should be further examined.
These findings regarding microstructural properties may
have clinical implications. Mechanical properties (i.e., thefatigue resistance of dental clasps), electrochemical proper-
ties and other properties may be altered by microstructural
changes, and further research is thus required in this field,
especially for the recently introduced SLM technology. How-
ever, no differences were found for the elemental distribution
of the probed elements at the metalporcelain interface,
which agrees with previous data that showed no difference
in bond strength between cast and laser-sintered structures
[8,18].
Given that the microstructure and hardness were signif-
icantly different in the SLM group, further differences in
the clinical behavior of prosthetic restorations manufactured
using this technique are anticipated. SLM technology has
recently been introduced in the dental field, and a vast spec-
trum of factors should be tested and/or optimized to increase
its efficacy in the production of metallic dental restorations.
5. Conclusions
Within the limitations of this study, the following conclusions
can be derived:
CoCr dental alloys fabricated via casting, milling or SLM
techniques show significant differences in microstructure
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.0085/19/2018 1-s2.0-S0109564114000207-main.pdf
9/10
Please cite this article in press as: Al Jabbari YS, et al. Metallurgical and interfacial characterization of PFM CoCr dental alloys fabricated viacasting, milling or selective laser melting. Dent Mater (2014), http://dx.doi.org/10.1016/j.dental.2014.01.008
ARTICLE IN PRESSDENTAL-2311; No.of Pages 10
dental mater i al s xxx ( 20 14 ) xxx.e1xxx.e10 xxx.e9
and hardness but not at the elemental distribution level of
metalceramic interface.
The application of INmetalbond has a profound effect on
the elements at the interface between the metal and the
opaque porcelain.
Acknowledgments
This study has been funded by a research grant (RGP-VPP-
206) from the Research Group Program, Deanship of Scientific
Research, King Saud University, Riyadh, Saudi Arabia.
The authors would like to thank Phenix Systems and Mr
Briakos Dental Lab for the manufacturing of metallic speci-
mens.
r e f e r ence s
[1] van Noort R. The future of dental devices is digital. DentMater 2012;28:312.
[2] Willer J, Rossbach A, Weber HP. Computer-assisted milling ofdental restorations using a new CAD/CAM data acquisitionsystem. J Prosthet Dent 1998;80:34653.
[3] Quante K, Ludwig K, Kern M. Marginal and internal fit ofmetalceramic crowns fabricated with a new laser meltingtechnology. Dent Mater 2008;24:13115.
[4] Traini T, Mangano C, Sammons RL, Mangano F, Macchi A,Piattelli A. Direct laser metal sintering as a new approach tofabrication of an isoelastic functionally graded material formanufacture of porous titanium dental implants. DentMater 2008;24:152533.
[5] Miyazaki T, Hotta Y. CAD/CAM systems available for thefabrication of crown and bridge restorations. Aust Dent J
2011;56(Suppl. 1):97106.[6] Witkowski S, Komine F, Gerds T. Marginal accuracy oftitanium copings fabricated by casting and CAD/CAMtechniques. J Prosthet Dent 2006;96:4752.
[7] Alt V, Hannig M, Wostmann B, Balkenhol M. Fracturestrength of temporary fixed partial dentures: CAD/CAMversus directly fabricated restorations. Dent Mater2011;27:33947.
[8] Akova T, Ucar Y, Tukay A, Balkaya MC, Brantley WA.Comparison of the bond strength of laser-sintered and castbase metal dental alloys to porcelain. Dent Mater2008;24:14004.
[9] Ucar Y, Akova T, Akyil MS, Brantley WA. Internal fitevaluation of crowns prepared using a new dental crownfabrication technique: laser-sintered CoCr crowns. J
Prosthet Dent 2009;102:2539.[10] Tian XY, Gunster J, Melcher J, Li DC, Heinrich JG. Processparameters analysis of direct laser sintering and posttreatment of porcelain components using Taguchis method.
J Eur Ceram Soc 2009;29:190315.[11] Williams RJ, Bibb R, Eggbeer D, Collis J. Use of CAD/CAM
technology to fabricate a removable partial dentureframework. J Prosthet Dent 2006;96:969.
[12] Tara MA, Eschbach S, Bohlsen F, Kern M. Clinical outcome ofmetalceramic crowns fabricated with laser-sinteringtechnology. Int J Prosthodont 2011;24:468.
[13] Ortorp A, Jonsson D, Mouhsen A, Vult von Steyern P. The fitofcobaltchromium three-unit fixed dental prosthesesfabricated with four different techniques: a comparativein vitro study. Dent Mater 2011;27:35663.
[14] Castillo-Oyague R, Osorio R, Osorio E, Sanchez-Aguilera F,Toledano M. The effect of surface treatments on the
microroughness of laser-sintered and vacuum-cast basemetal alloys for dental prosthetic frameworks. Microsc ResTech 2012;75:120612.
[15] Castillo-de-Oyague R, Sanchez-Turrion A, Lopez-Lozano JF,Albaladejo A, Torres-Lagares D, Montero J, et al. Verticalmisfit of laser-sintered and vacuum-cast implant-supportedcrown copings luted with definitive and temporary lutingagents. Med Oral Patol Oral Cir Bucal 2012;17:e6107.
[16] Oyague RC, Sanchez-Turrion A, Lopez-Lozano JF, Montero J,Albaladejo A, Suarez-Garcia MJ. Evaluation of fit ofcement-retained implant-supported 3-unit structuresfabricated with direct metal laser sintering and vacuumcasting techniques. Odontology 2012;100:24953.
[17] Oyague RC, Sanchez-Turrion A, Lopez-Lozano JF,Suarez-Garcia MJ. Vertical discrepancy and microleakage oflaser-sintered and vacuum-cast implant-supportedstructures luted with different cement types. J Dent2012;40:12330.
[18] Xiang N, Xin XZ, Chen J, Wei B. Metalceramic bond strengthofCoCr alloy fabricated by selective laser melting. J Dent2012;40:4537.
[19] Reclaru L, Ardelean L, Rusu L, Sinescu C. CoCr materialselection in prosthetic restoration: laser sinteringtechnology. Solid State Phenom 2012;188:4125.
[20] Castillo-Oyague R, Lynch CD, Turrion AS, Lopez-Lozano JF,Torres-Lagares D, Suarez-Garcia MJ. Misfit and microleakageofimplant-supported crown copings obtained by lasersintering and casting techniques, luted with glass-ionomer,resin cements and acrylic/urethane-based agents. J Dent2013;41:906.
[21] Iseri U, Ozkurt Z, Kazazoglu E. Shear bond strengths ofveneering porcelain to cast, machined and laser-sinteredtitanium. Dent Mater J 2011;30:27480.
[22] Hero H, Syverud M, Gjonnes J, Horst JA. Ductility andstructure of some cobalt-base dental casting alloys.Biomaterials 1984;5:2018.
[23] Qiu J, Yu WQ, Zhang FQ, Smales RJ, Zhang YL, Lu CH.
Corrosion behaviour and surface analysis of a CoCr andtwo NiCr dental alloys before and after simulated porcelainfiring. Eur J Oral Sci 2011;119:93101.
[24] Ramirez LE, Castro M, Mendez M, Lacaze J, Herrera M,Lesoult G. Precipitation path of secondary phases duringsolidification of the Co25.5%Cr5.5%Mo0.26%C alloy. ScrMater 2002;47:8116.
[25] Xin XZ, Chen J, Xiang N, Wei B. Surface properties andcorrosion behavior of CoCr alloy fabricated with selectivelaser melting technique. Cell Biochem Biophys 2013,http://dx.doi.org/10.1007/s12013-013-9593-9.
[26] Karpuschewski B, Pieper HJ, Krause M, Doring J. CoCr is notthe same: CoCr-blanks for dental machining. In: Schuh G,Neugebauer R, Uhlmann E, editors. Future trends inproduction engineering. Berlin, Heidelberg: Springer-Verlag;
2013. p. 26174.[27] Matkovic T, Matkovic P, Malina J. Effects of Ni and Mo on themicrostructure and some other properties of CoCr dentalalloys. J Alloys Compd 2004;366:2937.
[28] Yoda K, Suyalatu, Takaichi A, Nomura N, Tsutsumi Y, Doi H,et al. Effects of chromium and nitrogen content on themicrostructures and mechanical properties of as-castCoCrMo alloys for dental applications. Acta Biomater2012;8:285662.
[29] Takaichi A, Suyalatu, Nakamoto T, Joko N, Nomura N,Tsutsumi Y, et al. Microstructures and mechanicalproperties of Co29Cr6Mo alloy fabricated by selective lasermelting process for dental applications. J Mech BehavBiomed Mater 2013;21:6776.
[30] Podrez-Radziszewska M, Haimann K, Dudzikski W,
Morawska-Soltysik M. Characteristic of intermetallic phasesin cast dental CoCrMo alloy. Arch Foundry Eng 2010;10:516.
http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.dental.2014.01.008http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0005http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0005http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0005http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0010http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0010http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0010http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0010http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0015http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0015http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0015http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0015http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0015http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0020http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0020http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0020http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0020http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0020http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0020http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0025http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0025http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0025http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0025http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0030http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0030http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0030http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0030http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0035http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0035http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0035http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0035http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0035http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0040http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0040http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0040http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0040http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0040http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0045http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0045http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0045http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0045http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0045http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0045http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0050http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0050http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0050http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0050http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0050http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0050http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0055http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0055http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0055http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0055http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0060http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0060http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0060http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0060http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0060http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0065http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0065http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0065http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0065http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0065http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0065http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0065http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0070http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0070http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0070http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0070http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0070http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0070http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0075http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0075http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0075http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0075http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0075http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0075http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0075http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1007/s12013-013-9593-9http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0150http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0145http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0140http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0135http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0130http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1007/s12013-013-9593-9http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0120http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0115http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0110http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0105http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0100http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0095http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0090http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0085http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(14)00020-7/sbref0080http://refhub.elsevier.com/S0109-5641(1