INVITED FEATURE PAPER

Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

Kasra SardashtiMaterials Science Program University of CaliforniandashSan Diego La Jolla CA 92093 USA and Department ofChemistry and Biochemistry University of CaliforniandashSan Diego La Jolla CA 92093 USA

Dennis PaulPhysical Electronic Inc Chanhassen MN 55317 USA

Chuck HitzmanStanford Nano Shared Facilities Stanford CA 94305 USA

John HammondPhysical Electronic Inc Chanhassen MN 55317 USA

Richard HaightIBM T J Watson Research Center Yorktown Heights NY 10598 USA

Andrew C Kummela)

Department of Chemistry and Biochemistry University of CaliforniandashSan Diego La Jolla CA 92093 USA

(Received 31 July 2016 accepted 29 September 2016)

Kesterite Cu2ZnSn(SSe)4 (CZTSSe) absorbers are considered promising alternatives to commercialthin film technologies including CdTe and Cu(InGa)Se2 (CIGSe) owing to the earth abundance andnon-toxicity of their constituents However to be competitive with the existing technologies thephotovoltaic performance of CZTSSe solar cells needs to be improved beyond the current recordconversion efficiency of 126 In this study nanoscale elemental mapping using Auger nanoprobemicroscopy (NanoAuger) and nano secondary ion mass spectrometry (NanoSIMS) are used toprovide a clear picture of the compositional variations between the grains and grain boundaries inCu2ZnSn(SSe)4 kesterite thin films NanoAuger measurements revealed that the top surfaces of thegrains are coated with a Zn-rich (ZnSn)Ox layer While thick oxide layers were observed at thegrain boundaries their chemical compositions were found to be closer to SnOx NanoSIMSelemental maps confirmed the presence of excess oxygen deeper within the grain boundary groovesas a result of air annealing of the CZTSSe films

I INTRODUCTION

The market for thin film photovoltaics (PV) has beenexpanding over the past decade due to the technologicaladvances in developing low-cost and high-efficiencychalcogenide-based devices with CdTe and Cu(InGa)Se2absorbers However the total power generation capacityper year for these chalcogenide-based technologies isprojected to saturate at about 100 GWp

12 This is mainlydue to the limited accessible reserves of In and Te in theearthrsquos crust as well as the toxicity of Cd3 AlternativelyIn and Ga in the CIGSe films can be replaced withmore earth-abundant non-toxic elements such as Znand Sn forming Cu2ZnSn(SSe)4 (CZTSSe) whichhas kesterite crystal structure The current record

conversion efficiency of polycrystalline thin filmCZTSSe solar cell is 126 24 short of theefficiency threshold needed to make this materialcompetitive with the commercial thin film technolo-gies24 This deficit in performance has been mostlyascribed to the limited open-circuit voltage (Voc) due inlarge to the extent of carrier recombination at thedefects sites within the bulk of the absorbers or at theinterfaces that exist in the PV devices5ndash8

One of the critical interfaces in polycrystalline thinfilms is the one between adjacent grains called grainboundaries If not properly passivated grain boundariescan introduce a large density of carrier recombinationsites within the absorber region where the electronndashholepairs are generated Local surface electronic measure-ments of high-efficiency CIGSe thin films have shownmodest levels of downward band bending (100ndash300 mV)at the grain boundaries910 This downward band bending(equivalent to smaller work function of the grain bound-aries) has been attributed to a variety of factors includingCu depletion11 and Na or O accumulation12 at the grain

Contributing Editor Sam Zhanga)Address all correspondence to this authore-mail akummelucsdedu

This paper has been selected as an Invited Feature PaperDOI 101557jmr2016389

J Mater Res Vol 31 No 22 Nov 28 2016 Materials Research Society 2016 3473httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

boundaries In the case of CZTSSe thin films bothupward and downward band bending were observed atthe grain boundaries depending on the growth method13ndash15

Similar to CIGSe atomic scale substitutions with alkalimetals (such as Na and Li) or O in the crystal structureadjacent to the grain boundaries have been identified as themajor grain boundary passivation mechanisms1617

In the champion CZTSSe solar cells with 126energy conversion efficiency a critical processing step isthe air anneal where the samples are heated up to300ndash400 degC in ambient air for several minutes Kelvinprobe force microscopy (KPFM) measurements on air-annealed CZTSSe films showed upward band bending of70ndash100 mV at the grain boundaries correlated with thepresence of 100ndash200 nm thick SnOx layer within thegrain boundary grooves18 This SnOx layer is believed togrow during the air anneal process where oxygen isprovided from the ambient to oxidize the excess Sn at thegrain boundaries In addition to the grain boundaries airanneal has been found to assist formation of a Cu-poortop surface that is critical in realizing optimal topjunctions between CdS buffers and CZTSSe grains19

Cd from CdS chemical bath deposition can replace Cuwithin tens of nanometers away from the junctionresulting in a sub-junction doping and improvement ofthe CdSabsorber junction quality2021 However verylimited work has been performed to compare the oxidecomposition on the top surfaces of the grains with grainboundaries formed during the air anneal process

In the present study nanoscale elemental analysis ofCZTSSe grain surfaces and grain boundaries is carriedout with Auger nanoprobe microscopy (NanoAuger) Tocompare the composition of the oxide formed during theair anneal NanoAuger spectroscopy and elemental map-ping were performed before and after oxide removalusing in situ Ar1 sputtering Elemental maps of theair-annealed surfaces showed formation of Zn-rich(SnZn)Ox on grain surfaces and SnOx at the grainboundaries Using nanoscale secondary ion spectrom-etry (NanoSIMS) elemental mapping on air-annealedCZTSSe films it was confirmed that excess oxygenexists deep within grain boundaries consistent with theSnOx layer covering a large fraction of the grain wallsthrough the films Comparison of NanoSIMS elementalmaps before and after air anneal led to the conclusion thatthe O accumulation is a result of air annealing process

II MATERIALS AND METHODS

CZTSSe films were grown by spin coating of ahydrazine-based precursor slurry onto soda lime glasssubstrates coated with 100 nm thick Mo layers Spincoating of slurry onto the substrates was performed ina nitrogen-filled glove box After the thin film depositionCZTSSe films underwent two steps of post deposition

annealing (i) hard bake (HB) annealing at 600 degC for 15min in a nitrogen-filled glove box (ii) air anneal (AA)annealing at 350 degC for 10 min in air with average relativehumidity of 40ndash50 Further details of the CZTSSegrowth procedure have been published elsewhere22ndash24

The bulk [Cu][Zn] and [Se][S] 1 [Se] measured byx-ray fluorescence (XRF) were 18 and 07 respectively

Auger nanoprobe measurements were performed usinga PHI-710 Auger nanoprobe microscope (NanoAugerPhysical Electronics Chanhassen Minnesota) TheNanoAuger tool provides 1 energy resolution forspectroscopy and 5 nm spatial resolution for elementalmapping A 20 kV 10 nA electron beam was used forboth Auger spectroscopy and elemental mapping Addi-tionally to remove the native oxide from the thin filmsrsquotop surfaces in situ Ar1 ion sputtering with 500 eVenergy was applied To minimize the sputtering ratedependence on the surface geometry the sputter gun wasperpendicular to the sample surface

NanoSIMS elemental mapping was achieved by usinga CAMECA NanoSIMS 50L tool (CAMECA Gennevil-liers Cedex France) with Cs1 as the primary ion beam(2ndash8 pA) with a nominal spot size of 100ndash200 nm Thebeam was rastered over areas as large as 5 5 lm on thesample surface Negative secondary ions for the elementsof interest were collected simultaneously on the Nano-SIMS multi-collection system

III RESULTS AND DISCUSSION

A Auger nanoprobe microscopy (NanoAuger)

Figure 1 displays the NanoAuger elemental maps inplanar mode for a CZTSSe surface after the air annealingat 350 degC Since no chemical treatment was performed onthe sample after air anneal the top surface is expected tobe covered by an oxide layer Distributions of all fiveelements (Cu Zn Sn Se and O) on the grainsrsquo surfaceswere relatively uniform In stark contrast the grainboundaries appeared to have higher concentrations ofSn and O and smaller concentrations of Cu Zn and Serelative to grains This is in consistent with the previousreport on the observation of a thin SnOx layer at the grainboundaries of air annealed CZTSSe thin films18

It should be noted that in the elemental maps a uniformbackground subtraction is applied smaller intensity in theelemental map does not necessarily correlate with the fulldepletion of the corresponding elements To determinethe extent of the compositional variation between thegrains and grain boundaries a line trace across the scanregion is measured as shown in Fig 2(a) Each elementalline trace in Fig 2(b) consists of 128 data points withtheir intensity corresponding to the peak-to-peak intensityof the elementrsquos derivative Auger spectrum (dNdEversus kinetic energy) For the grain boundary on theleft about 30 enhancement in Sn and 40

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 20163474httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

enhancement in O intensities were observed while only20 drop in Zn and Cu intensities were estimated Thiscan be in part due to the larger escape depths of Zn and CuAuger peaks which make them less surface sensitive2526

In addition the signal enhancement at the grain boundariesseemed to be gradual (not very sharp) consistent witha beam size larger than the width of the grain boundaryTherefore the rather small drop in the Zn and Cu signalcould be a result of the large beam spot size (larger thangrain boundary width) and the larger escape depth for theAuger peaks for Zn and Cu versus Sn and O

A mild Ar1 sputtering at 500 V ion energy (filamentcurrent 5 600 nA) was applied in situ for 30 s to removethe oxide layer from the CZTSSe top surface The samplestage was titled 30deg to align the ion gun with the samplersquossurface normal Secondary electron images and Augerelemental maps of the sample surface after 30 s of Ar1

sputtering are shown in Fig 3 It should be noted thatdifferent scan locations were chosen for Figs 1 and 2After sputtering the contrast in signal intensity betweenthe grains and grain boundaries for Sn and O has increasedrelative to the as-inserted surface Additionally lower

FIG 1 Secondary electron micrograph and NanoAuger elemental maps for Cu Zn Sn Se and O measured on as-inserted CZTSSe surface after 10 min airanneal at 350 degC The scan size is 2 2 lm2 Grain boundaries have higher concentrations of Sn and O compared to the top surfaces of the grains There isa void from which very small Auger signals were detected Therefore this void appeared as a dark region in all the five elemental maps

FIG 2 (a) Secondary electron micrograph and O elemental map for air-annealed CZTSSe surface with the position of the horizontal line traceshown by the dotted lines (b) Peak-to-peak intensities for Cu Zn Sn and O normalized to the relative sensitivity factors (RSFs) along the linetraces shown in (a) The line traces show that the variations in peak-to-peak intensities between the grains and grain boundaries are only 20ndash40Therefore darker or brighter regions in the elemental maps do not correspond to full depletion or accumulation of the elements

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 2016 3475httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

intensities for Cu Zn and Se were observed at the grainboundaries relative to the grains although the contrastbetween the grains and grain boundaries seems to besmaller than the as-inserted surface There were fewzones on the grain surfaces with O Zn and Sn signalenhancement (blue arrows) These could be due tononuniformities in the oxide thickness resulting in

incomplete removal of the oxides during the short Ar1

sputtering stepTo quantify the chemical composition of the grain and

grain boundaries after Ar sputtering single point Augerspectroscopy was carried out on top surface of a grain(area 1) and a grain boundary adjacent to it (point 2) asshown in Fig 4(a) The Auger survey spectra measured on

FIG 3 Secondary electron micrograph and NanoAuger elemental maps for Cu Zn Sn Se and O measured on air-annealed CZTSSe surface after30 s of Ar1 sputtering with 05 kV ion beam energy and 600 nA filament current The scan size for these measurements is 3 3 lm2 Similar to theas-inserted sample grain boundaries appear to have larger concentrations of Sn and O

FIG 4 (a) Secondary electron micrograph and Sn elemental map for air-annealed CZTSSe top surface after 30 s of Ar1 sputtering (b) Augersurvey spectra for area 1 (grain) and point 2 (grain boundary) Elemental compositions calculated from the Auger spectra are shown in the boxes onthe bottom right of the spectra

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 20163476httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

area 1 and point 2 are shown in derivative format in Fig 4(b) The elemental composition for each scan location wasestimated using the elemental peaks marked with sign(example Cu Zn etc) In comparison with the grainsurface in area 1 the grain boundary at point 2 has about2 larger concentration of Sn and O In addition the grainboundary oxide has [Sn][Zn] 5 187 Considering thelarge beam spot size (larger than grain boundary width)and longer escape depth for Zn Auger peak the majorityof the Zn signal in the single point measurements could bedetected from the CZTSSe grains adjacent to the grainboundary oxide Therefore it is expected that the grainboundary grooves near the top surface are terminated byan oxide with composition close to SnOx This oxide hasbeen found necessary for maintaining low grain boundaryrecombination rates and high Voc in CZTSSe solar cells18

To determine the composition of the oxide formed onthe grain surfaces variations in the Auger spectra of topsurfaces of grains were monitored before and after Ar1

sputtering [area 1 in Fig 5(a)] The elemental ratio foreach element was calculated by normalizing its Augerpeak-to-peak intensity (Ipndashp) to its relative sensitivityfactor (RSF) (In 5 IpndashpRSF) and then dividing by thesum of the normalized intensities of CZTSSe constituents[ie Ofrac12 frac14 InO= InCuthorn InZnthorn InSnthorn InSthorn I

nSethorneth THORN] Figure 5(b)

displays the elemental ratios for Cu Zn Sn S Se andO in area 1 before and after sputtering As a result ofsputtering the [Zn] and [Sn] ratios were reduced by about60 while [O] decreased by 12 from 06 to 005Concurrently [Cu] and [Se] were increased by 60 and76 respectively Therefore it can be concluded that thethin overlayer present on the as-inserted CZTSSe surface

FIG 5 (a) Secondary electron micrograph of air-annealed CZTSSe surface before Ar1 sputtering The area chosen for spectroscopy is marked bythe yellow box (b) Elemental ratios calculated for Cu Zn Sn S Se and O in area 1 before (red) and after (blue) Ar1 sputtering (c) Sn MNNAuger peak before and after surface oxide removal by Ar1 sputtering (d) Zn LMM Auger peak before and after surface oxide removal by Ar1

sputtering For both (c) and (d) Auger peaks are in absolute mode and the peak intensities are normalized to the maximum intensities It should benoted that the left shoulder for sputtered surface seems to be larger than the as-inserted surface This background increase could be due to theoverlap between Cu1 and Zn1 Auger peaks particularly considering the increase in [Cu] after sputtering

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 2016 3477httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

contains large amounts of Sn Zn and O Considering theshort amount of time needed to remove the oxide fromthe grainsrsquo top surfaces via Ar1 sputtering (30 s) theoverlayer thickness can be approximated to 15 nmDue to their large inelastic mean free path (IMFP) largefractions of the Cu (IMFP 16 nm) and Se (IMFP 24 nm) Auger signals are expected to be received fromthe CZTSSe underneath the oxide on air-annealedsurfaces

To confirm this Zn and Sn Auger peaks in absolutemode before and after sputtering were compared asshown in Figs 5(c) and 5(d) After Ar1 sputtering Snand Zn peaks shifted to higher kinetic energies by 1 and33 eV These shifts to higher kinetic energies for Zn andSn peaks after sputtering is consistent with these elementsbonding to O before sputtering and bonding to S or Seafter sputtering It should be noted that this difference inthe chemical shift is due to the three-electron nature ofthe Auger process resulting in variations in sensitivity ofAuger peaks to the changes in the elementsrsquo chemicalbonding state The difference in chemical shifts observedhere for Zn LMM and Sn MNN peaks are consistent withthe previous reports on Auger spectroscopy of Zn and Snsurface oxides2728 Considering the quantitative analysisgiven above grain surfaces in air annealed CZTSSe arehypothesized to be terminated by a thin layer of (ZnSn)Owith SnZn 1 This is consistent with previous large

area Auger spectroscopy measurements where the nativeoxide is identified as a Cu-free overlayer with large Znand Sn concentrations29

It should be noted that CZTSSe top surface oxide canbe removed by NH4OH in the chemical bath solutionused for CdS buffer deposition1819 Therefore thinsurface oxides are expected to have no effect on thequality of the CdSCZTSSe top junction in eventualphotovoltaic devices

B Nanoscale secondary ion mass spectrometry(NanoSIMS)

While NanoAuger measurements determined that grainboundaries are terminated by SnOx NanoAuger does notprovide direct information of the depth of this passivationlayer To clarify the depth distribution of the SnOx

passivation layer elemental mapping using NanoSIMSwas performed with the primary Cs1 was rastered overa 5 5 lm area over the surface of an air-annealedCZTSSe film NanoSIMS allow simultaneous sputteringwith elemental mapping with high chemical sensitivityand depth resolution but with lower spatial resolutionthan cyclic sputtering and NanoAuger (best lateralresolution of 50ndash60 nm) For NanoSIMS the amountof negatively charged secondary ions including O S andSe were measured simultaneously using three detectors inparallel Maps for secondary electrons as well as O S

FIG 6 Planar secondary electrons and elemental maps for O S and Se for three different frames measured by NanoSIMS (Cs1 source) on air-annealedCZTSSe surfaces at 3 different times during the profiling Frame 1 2 and 3 are subsequently showing deeper layers on the sample surface O is present atthe GBs of the air-annealed sample after more than 3 min of depth profiling (equivalent to 70ndash100 nm of CZTSSe removal)

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 20163478httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

and Se were recorded after 1 min long ion sputtering timeintervals

Secondary electron and elemental maps for O S andSe after 1 min (frame 1) 2 min (frame 2) and 3 min(frame 3) of Cs1 sputtering are shown in Fig 6 Similarto the NanoAuger maps grain boundaries on the startingsurfaces were found to be O-rich and depleted from S andSe As the films are eroded during the milling processalthough the O intensity at the grain boundaries de-creased the grain boundaries still appeared to be O-richEven after 3 min of ion milling excess O at the grainboundaries can be observed [Note The decrease in theintensity of O signal for longer sputtering times could bedue to the smaller width of SnOx layer at the grainboundaries deeper within the CZTSSe films As the widthof the O-rich region shrinks below the NanoSIMSelemental mapping resolution (100ndash150 nm) the signalis averaged over grain boundaries and adjacent grainssurfaces resulting in smaller total O intensity] This isconsistent with the (ZnSn)O grain boundary passivationlayer extending deeper within the grain boundary groovesat least for the thickness equivalent of 3 min long Cs1

milling (70ndash100 nm) In addition by moving from frame1 to 3 the S and Se intensities increase on the grainsurfaces consistent with the removal of the top surface(ZnSn)Ox

Figure 7 displays the NanoSIMS elemental maps forO S and Se elemental maps for hard baked [HB Fig 7(a)] and air-annealed [HB 1 AA Fig 7(b)] CZTSSefilms Maps were summed for 7 different imaging framesto enhance the overall signal intensity and contrast In

comparison with the hard-baked CZTSSe film [Fig 7(a)]grain boundaries in the air-annealed film appeared tohave larger O accumulation and smaller chalcogen (SSe)depletion This is consistent with the air anneal processassisting the grain boundary oxidation via oxidants in airor enhanced oxygen diffusion from the soda-lime glasssubstrate into CZTSSe back surface Previous Nano-Auger work demonstrated the presence of excess Sn atthe grain boundaries even for films that underwent hardbake only18 Therefore the air anneal process oxidizesthe Sn-rich grain boundaries forming a charge passivatingSnOx layer that induces up upward band bending in theregions adjacent to the grain boundaries1518

IV CONCLUSIONS

Qualitative and quantitative elemental analysis byNanoAuger as well as qualitative elemental mapping asa function of depth by NanoSIMS were used to determinethe chemical composition of the top surfaces of grainsand grain boundaries in air-annealed CZTSSe thin-filmabsorbers Grain boundaries appear to have larger con-centrations of Sn and O even before any oxide removalBy performing Auger spectroscopy on individual grainsurfaces it was determined that the oxide formed on thegrain surface during the air anneal process is composed of(ZnSn)O with ZnSn of about 11 Similar small areaspectroscopy measurements revealed that grain boundaryoxide composition is closer to pure SnOx NanoSIMSelemental maps showed that O-rich grain boundariesextend deeper within the grain boundary grooves for up

FIG 7 Planar secondary electrons and elemental maps for O S and Se measured by NanoSIMS (Cs1 source) on (a) hard-baked (HB) and (b) air-annealed (HB 1 AA) CZTSSe surfaces Data has been integrated over 7 measurement frames According to the O maps air annealing the sampleresulted in significantly higher O content at the GBs

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 2016 3479httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

to 3 min of ion milling This O-rich layer was determinedto be a direct result of the air annealing treatment sincemuch smaller O accumulation was observed at the grainboundaries of the CZTSSe films with hard bake only

ACKNOWLEDGMENTS

The information data or work presented herein wasfunded in part by the US Department of Energy EnergyEfficiency and Renewable Energy Program under AwardNumber DE-EE0006334 The information data or workpresented herein was funded in part by an agency of theUnited States Government Neither the United StatesGovernment nor any agency thereof nor any of theiremployees makes any warranty express or implied orassumes any legal liability or responsibility for theaccuracy completeness or usefulness of any informationapparatus product or process disclosed or representsthat its use would not infringe privately owned rightsReference herein to any specific commercial productprocess or service by trade name trademark manufac-turer or otherwise does not necessarily constitute or implyits endorsement recommendation or favoring by theUnited States Government or any agency thereof Theviews and opinions of authors expressed herein do notnecessarily state or reflect those of the United StatesGovernment or any agency thereof see the SupportingInformation section of this article for more informationPart of this work was performed at the Stanford Nano-Shared Facility (SNSF) The CAMECA nanoSIMS 50Lwas supported by the National Science Foundation underaward 0922648 Special thanks to Liang-Yi Chang forsupplying the CZTSSe films used in these investigationsFunding for the technique development was also providedby National Science Foundation Grant DMR 1207213

REFERENCES

1 C Candelise M Winskel and R Gross Implications for CdTeand CIGS technologies production costs of indium and telluriumscarcity Prog Photovoltaics 20(6) 816 (2012)

2 DB Mitzi O Gunawan TK Todorov and DAR BarkhouseProspects and performance limitations for CundashZnndashSnndashSndashSe photo-voltaic technology Philos Trans R Soc A 371 20110432 (2013)

3 XL Liu Y Feng HT Cui FY Liu XJ Hao G ConibeerDB Mitzi and M Green The current status and future prospectsof kesterite solar cells A brief review Prog Photovoltaics 24(6)879 (2016)

4 W Wang MT Winkler O Gunawan T Gokmen TK TodorovY Zhu and DB Mitzi Device characteristics of CZTSSe thin-film solar cells with 126 efficiency Adv Energy Mater 4(7)1301465 (2014)

5 E Chagarov K Sardashti AC Kummel YS Lee R Haight andTS Gershon Ag2ZnSn(SSe)(4) A highly promising absorber forthin film photovoltaics J Chem Phys 144(10) 104704 (2016)

6 NA Kattan IJ Griffiths D Cherns and DJ Fermin Observa-tion of antisite domain boundaries in Cu2ZnSnS4 by atomic-resolution transmission electron microscopy Nanoscale 8 14369(2016)

7 T Gokmen O Gunawan TK Todorov and DB Mitzi Bandtailing and efficiency limitation in kesterite solar cells Appl PhysLett 103(10) 103506 (2013)

8 JJ Scragg PJ Dale D Colombara and LM Peter Thermody-namic aspects of the synthesis of thin-film materials for solar cellsChemPhysChem 13(12) 3035 (2012)

9 CS Jiang MA Contreras I Repins HR Moutinho Y YanMJ Romero LM Mansfield R Noufi and MM Al-JassimHow grain boundaries in Cu(InGa)Se-2 thin films are chargedRevisit Appl Phys Lett 101(3) 033903 (2012)

10 CS Jiang IL Repins LM Mansfield MA ContrerasHR Moutinho K Ramanathan R Noufi and MM Al-JassimElectrical conduction channel along the grain boundaries of Cu(InGa)Se-2 thin films Appl Phys Lett 102(25) 253905 (2013)

11 K Sardashti R Haight R Anderson M ContrerasB Fruhberger and AC Kummel Grazing incidence cross-sectioning of thin-film solar cells via cryogenic focused ion beamA case study on CIGSe ACS Appl Mater Interfaces 8(24)14994 (2016)

12 D Abou-Ras SS Schmidt R Caballero T Unold HW SchockCT Koch B Schaffer M Schaffer PP Choi and O Cojocaru-Miredin Confined and chemically flexible grain boundaries inpolycrystalline compound semiconductors Adv Energy Mater2(8) 992 (2012)

13 ME Erkan V Chawla I Repins and MA Scarpulla Interplaybetween surface preparation and device performance in CZTSSesolar cells Effects of KCN and NH4OH etching Sol EnergyMater Sol Cells 136 78 (2015)

14 CS Jiang IL Repins C Beall HR Moutinho K Ramanathanand MM Al-Jassim Investigation of micro-electrical propertiesof Cu2ZnSnSe4 thin films using scanning probe microscopy SolEnergy Mater Sol Cells 132 342 (2015)

15 H Xin SM Vorpahl AD Collord IL Braly AR UhlBW Krueger DS Ginger and HW Hillhouse Lithium-dopinginverts the nanoscale electric field at the grain boundaries inCu2ZnSn(SSe)(4) and increases photovoltaic efficiency PhysChem Chem Phys 17(37) 23859 (2015)

16 T Gershon B Shin N Bojarczuk M Hopstaken DB Mitzi andS Guha The role of sodium as a surfactant and suppressor of non-radiative recombination at internal surfaces in Cu2ZnSnS4 AdvEnergy Mater 5(2) 1400849 (2015)

17 JH Kim SY Choi M Choi T Gershon YS Lee W WangB Shin and SY Chung Atomic-scale observation of oxygensubstitution and its correlation with hole-transport barriers inCu2ZnSnSe4 thin-film solar cells Adv Energy Mater 6(6)1501902 (2016)

18 K Sardashti R Haight T Gokmen W Wang LY ChangDB Mitzi and AC Kummel Impact of nanoscale elementaldistribution in high-performance kesterite solar cells Adv EnergyMater 5(10) 1402180 (2015)

19 R Haight XY Shao W Wang and DB Mitzi Electronic andelemental properties of the Cu2ZnSn(SSe)(4) surface and grainboundaries Appl Phys Lett 104(3) 033902 (2014)

20 K Hiepko J Bastek R Schlesiger G Schmitz R Wuerz andNA Stolwijk Diffusion and incorporation of Cd in solar-gradeCu(InGa)Se-2 layers Appl Phys Lett 99(23) 234101 (2011)

21 T Nakada and A Kunioka Direct evidence of Cd diffusion intoCu(InGa)Se-2 thin films during chemical-bath deposition processof CdS films Appl Phys Lett 74(17) 2444 (1999)

22 DAR Barkhouse O Gunawan T Gokmen TK Todorov andDB Mitzi Device characteristics of a 101 hydrazine-processedCu2ZnSn(SeS)4 solar cell Prog Photovoltaics 20(1) 6 (2012)

23 TK Todorov KB Reuter and DB Mitzi High-efficiency solarcell with earth-abundant liquid-processed absorber Adv Mater22(20) E156 (2010)

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 20163480httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at

24 TK Todorov J Tang S Bag O Gunawan T Gokmen Y Zhuand DB Mitzi Beyond 11 efficiency Characteristics of state-of-the-art Cu2ZnSn(SSe)(4) solar cells Adv Energy Mater 3(1)34 (2013)

25 A Jablonski and CJ Powell Information depth and the meanescape depth in Auger electron spectroscopy and X-ray photo-electron spectroscopy J Vac Sci Technol A 21(1) 274 (2003)

26 A Jablonski and CJ Powell Practical expressions for the meanescape depth the information depth and the effective attenuationlength in Auger-electron spectroscopy and x-ray photoelectronspectroscopy J Vac Sci Technol A 27(2) 253 (2009)

27 LC Lynn and RL Opila Chemical-shifts in the MNN Auger-spectra of Cd in Sn Sb and Te Surf Interface Anal 15(2) 180(1990)

28 I Milosev TK Mikic and M Gaberscek The effect of Cu-richsub-layer on the increased corrosion resistance of CundashxZn alloysin chloride containing borate buffer Electrochim Acta 52(2) 415(2006)

29 M Bar BA Schubert B Marsen S Krause S PookpanratanaT Unold L Weinhardt C Heske and HW Schock Nativeoxidation and Cu-poor surface structure of thin film Cu2ZnSnS4solar cell absorbers Appl Phys Lett 99(11) 112103 (2011)

K Sardashti et al Nano-scale compositional analysis of surfaces and interfaces in earth-abundant kesterite solar cells

J Mater Res Vol 31 No 22 Nov 28 2016 3481httpwwwcambridgeorgcoreterms httpdxdoiorg101557jmr2016389Downloaded from httpwwwcambridgeorgcore Access paid by the UCSD Libraries on 08 Dec 2016 at 183458 subject to the Cambridge Core terms of use available at