Improved Photovoltaic Characteristics and Grain Boundary Potentials

Jun 29, 2016 - This work introduces the incorporation of Na into the nontoxic precursor solution of CIGS to improve photovoltaic cell performance with...
1 downloads 3 Views 4MB Size
Letter www.acsami.org

Improved Photovoltaic Characteristics and Grain Boundary Potentials of CuIn0.7Ga0.3Se2 Thin Films Spin-Coated by Na-Dissolved Nontoxic Precursor Solution Ik Jin Choi,† Jin Woo Jang,† Bhaskar Chandra Mohanty,‡ Seung Min Lee,† and Yong Soo Cho*,† †

Department of Materials Science and Engineering, Yonsei University, Seoul 120-749, Korea School of Physics & Materials Science, Thapar University, Patiala 147004, India



S Supporting Information *

ABSTRACT: This work introduces the incorporation of Na into the nontoxic precursor solution of CIGS to improve photovoltaic cell performance with the optimized benefits of Na. The extensive incorporation range of 0.05 to 0.5 mol % Na is used for the simple spin-coating process of high quality absorber thin films. A cell efficiency of ∼8.21%, which corresponds to an improvement of ∼10.2% compared to the reference sample, is achieved for the 0.25 mol % Na sample with enhanced open-circuit voltage and fill factor. The improvement was further analyzed as related to InCu defects and grain boundary potentials.

KEYWORDS: thin film solar cells, chalcopyrite, solution deposition, grain boundary, Na-incorporation

C

solution-processing approach for fabricating only selenide CIGS solar cells is desirable, there are very rare reports on successful deposition of CIGS with a resultant high cell efficiency, which utilize the direct film coating of easily available nontoxic acetate, nitrate, or chloride precursor solutions without involvement of nanoparticles prepared separately. On the other hand, the effects of Na presence in the absorber are reportedly different depending on the specific condition of experiments and processing methods. Typically for synthesis routes involving a high temperature greater than 450 °C, Na is introduced into the CIGS layer by diffusion from soda-lime glass substrate through the Mo back contact.7 Other intentional incorporation methods of Na including postdeposition of Na on CIGS,8 NaF coevaporation on Mo9 and the use of Nadoped Mo,10 have also been reported. The Na incorporation is believed to improve the photovoltaic conversion efficiency by enhancing mainly open-circuit voltage Voc and fill factor FF while nearly not affecting the short circuit current density Jsc.3 However, the cell performance starts to degrade with adhesion problems after exceeding an optimal amount of incorporated Na.4 The high doping level also leads to small grain sizes with potentially changed nature of grain boundary.11 Nevertheless, the exact mechanism of the performance enhancement by Na is still being debated.

halcopyrite CuInSe2 (CIS), which benefits from its high optical absorption coefficient >1 × 105 cm−1 in the visible to near-IR spectral range, appropriate direct and tunable energy bandgap of 1−1.5 eV, high tolerance to native defects and longterm stability,1 has been the most popular candidate as the absorber layer in thin film solar cells. Recently, the record conversion efficiency of ∼22.3% was reported for the solar cells with Cu(In,Ga)Se2 (CIGS) absorber layer.2 At present, the successful high-efficiency CIGS solar cells are typically fabricated by vacuum deposition techniques, which, however, suffer from the limitations of high processing capital costs (i.e., initial investment and vacuum maintenance expense), low efficiency of resource material usage, and slow processing speed.3,4 Solution methods, which traditionally have the advantages of being simple and inexpensive and having large area deposition, high throughput, and better compatibility with flexible substrates, are believed to provide alternate routes of preparation of CIGS thin films and are, therefore, being widely studied. So far, the highest efficiency of ∼15.2% was reported for the solution-processed CIGS absorber layer prepared by dissolving chalcogenides (Cu2S, In2Se3, Ga2Se3, and Se) in hydrazine solution.5 However, hydrazine being highly toxic and explosive has limited acceptability in large-scale production. Consequently, there are increasing efforts to synthesize the absorber layers by nonhydrazine solution process. Most of the recent developments deal with the solution-processed sulfoselenide Cu(In,Ga)(S,Se)2 wherein reproducibility of the films remains a concern arising because of poor control over S/Se ratio during postdeposition selenization step.6 Although a © XXXX American Chemical Society

Received: April 13, 2016 Accepted: June 29, 2016

A

DOI: 10.1021/acsami.6b04407 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Letter

ACS Applied Materials & Interfaces

Figure 1. (a) Current density−voltage curves of samples containing different amounts of Na, (b) a representative cross-sectional SEM micrograph of photovoltaic cell, showing highly dense microstructure of 0.25% Na-CIGS, and (c) XPS depth profiles of elements through the 0.25% Na-absorber layer on Mo-coated SLS glass substrate.

Figure 2. (a) Variations in carrier concentration n, Hall mobility μ, and dark resistivity ρ as a function of Na content in the CIGS absorber films; and (b) photoluminescence spectra of the CIGS thin films with different amounts of Na, measured at the excitation wavelength of 325 nm.

amount of Na arising by its diffusion from the soda-lime silicate substrate. A typical solar cell structure of Al/ZnO:Al/i-ZnO/ CdS/CIGS/Mo/substrate with an active area ∼0.25 cm2 was fabricated. Detailed experimental procedure is described in the Supporting Information. Figure 1 shows the J−V curves of photovoltaic cells based on the CIGS absorber containing various levels of Na. Their characteristic values of short circuit current Jsc, open-circuit voltage Voc, fill factor FF, and conversion efficiency η are represented in the inserted table. The cell with the CIGS layer having 0% Na exhibited Jsc of 32.43 mA/cm2, Voc of 0.453 V, FF of 49.2% resulting in a conversion efficiency η of 7.45%. The devices with increasing Na content up to 0.25% revealed the increasing trend of Voc and FF, whereas Jsc showed little variation. The cell with a higher amount of Na than 0.25%, however, showed inferior photovoltaic characteristics. The cell containing 0.25% Na demonstrated the best efficiency of ∼8.21% with Jsc of ∼32.32 mA/cm2, Voc of 0.482 V, and FF of 54.8%. The increase in the efficiency of the device, which corresponds to an enhancement of ∼10.2% compared to the 0% Na sample, is primarily due to the combined effect of increase in Voc and FF.

In this study, a Na-incorporated absorber approach using nontoxic solutions is introduced with its potential beneficial effect on cell performance. An extended range of 0.05 to 0.5 mol % Na incorporated intentionally into precursor solutions was investigated to understand the versatile effects of Na on electrical properties, grain boundary potential and cell performance. As a promising achievement, an enhanced cell efficiency of ∼8.21% was achieved with enhanced Voc and FF, which corresponds to ∼10.2% enhancement compared to the reference sample. The efficiency value is so far the best one for pure solution-driven CIGS solar cells without involvement of toxic hydrazine and additional sulfurization processing. Our results not only provide a better understanding of the effects of Na in the CIGS thin films, but also demonstrate a scalable route of fabrication of CIGS thin films with controllable inclusion of Na from nontoxic clear precursor solutions. Na-incorporated Cu(In0.7Ga0.3)Se2 thin films were prepared by a nontoxic solution process containing different Na contents of 0.05 to 0.5 mol %. It should be mentioned here that the actual level of Na incorporation into the CIGS structure may be more than the intentional incorporation due to the additional B

DOI: 10.1021/acsami.6b04407 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Letter

ACS Applied Materials & Interfaces

Figure 3. (a−c) Topography images of the 0% Na, 0.25% Na, and 0.5% Na samples and (d−f) the corresponding potential images. The bottom plots represent quantitative fluctuations in height and z-potential along the designated A−B lines across grain boundary in images d−f.

Figure 2a shows the variation of the electrical parameters such as carrier concentration n, Hall mobility μ, and dark resistivity ρ, measured at room temperature as a function of Na content. The carrier concentration increased from 2.6 × 1016 to 7.6 × 1016 cm−3 when the Na content increased from 0 to 0.25%, and thereafter it decreased. The mobility was not much changed as ∼3.40 cm2/(V s) up to 0.25% Na and then decreased to 1.93 cm2/(V s) for 0.5% Na. Correspondingly, the resistivity decreased slightly from 8.1 × 10−2 Ωcm for 0% Na to 4.4 × 10−2 Ωcm for 0.25% Na and increased to 9.0 × 10−2 Ωcm for 0.5% Na, consistent with the earlier reports.13 Numerous models have been proposed so far to account for the mechanism behind the increase in the hole carrier density induced by the incorporation of Na in CIGS film during the grain growth. Among those explanations, the annihilation of the donor-like InCu defect by Na14 and the formation of NaIn antisitic defects15 seem to be consistent with the results of this work. Pohl and Albe16 have shown that InCu antisites are shallow donors, presence of which decreases the net hole

A very dense microstructure with extensively intergrown grain structure was observed, as seen in Figure 1b for example, indicating that this proposed synthesis route is well capable of producing typical microstructure required for high efficiency CIGS devices. Figure S1 demonstrates more SEM images for the samples containing different contents of Na. Figure 1c shows the XPS line profiles of each constituent element through the absorber layer toward the Mo layer. As expected, uniform distributions of Se, Cu, and Na were observed, whereas the relative content of In was diminished toward the interior of the absorber film. Especially, Ga tends to distribute more inside of the film. All the observed distributions were the expected ones from high efficiency CIGS solar cells reported as having η > 10%.12 From the small contents of Na involved, there has been no distinguishable difference over the film, even in 1 mol % Na-incorporated sample as seen in Figure S2. In addition, XRD patterns of the Na-incorporated absorber films are included as Figure S3, which demonstrates no noticeable difference in phase evolution. C

DOI: 10.1021/acsami.6b04407 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Letter

ACS Applied Materials & Interfaces

more complex defect and/or defect complexes, which are responsible for diminished Jsc and FF. One of the reasons for the reduction of the shunt resistance and Jsc of the devices could be the unfavorable grain boundaries that induce recombination and impede the minority carrier transport through the grain boundaries. We have used the Kelvin probe force microscopy (KPFM) to investigate the electrostatic properties across the grains and grain boundaries of the samples. For photovoltaic materials such as chalcopyrites and kesterites, with nanoscale resolution, KPFM has been found to be highly useful to measure changes in electrical properties across grain boundaries and distinguish work functions coming from differently oriented crystal facets of absorber films.23 Figure 3 shows the simultaneously acquired topography and surface potential images of the CIGS thin films with varying contents of Na. In the surface potential images (Figure 3d−f), the brighter regions indicate higher electrochemical potential or smaller work function than those of the darker areas. The films, even the one with 0% Na, revealed that the grain boundaries are at higher surface potential than the surface of the grains away from the grain boundary. Quantitative fluctuations in surface potential and height across a grain boundary presented in the bottom panel of the figure indicate that the potential difference between the grain boundary and the grains was ∼0.49 V for the 0% Na. The potential difference increased to ∼0.77 V for the increase in Na content up to 0.25% and thereafter decreased. A positive surface potential at the grain boundaries is believed to induce downward band bending that helps in attracting minority carriers, electrons, into the grain boundaries and repelling the majority carriers, holes, away from the grain boundaries.24 Thus, the positively charged grain boundaries are beneficial in that they enhance minority carrier collection owing to the associated electric field in its vicinity and provide a current pathway for the electrons to reach the n-type layers and be collected. Yan et al.25 reported that grain boundaries in CIGS are electrically benign. From the difference in the potential profiles of surfaces of the CuInSe2 films grown on Mo-coated soda-lime glass and boron silicate glass, they suggested that Na from the soda-lime glass resulted in a higher potential variation compared to that in the boron silicate glass.25 Accordingly, it is plausible that the enhancement in the grain boundary potential in our samples is due to the increasing Na content up to 0.25%, which is directly reflected in improvement in cell performance. Although the defects at the grain boundaries are not well understood yet, interstitial Na atoms in grain boundaries in CuInSe2 are considered to be shallow donors and not act as recombination centers, thus positively affecting the devices.25 However, as discussed earlier, the presence of deep and trap level as a consequence of accumulation of Na beyond the optimum concentration can be detrimental to the performance of the devices. In conclusion, Na incorporation into the nontoxic CIGS precursor solution for the simple spin-coating method resulted in the enhancement of cell efficiency by ∼10.2%. Improved Voc and FF are believed to be responsible for the enhancement of cell efficiency. Photoluminescence spectra indicate that the incorporation of Na may induce the reduction of the donor-like InCu defects that incur the increase in net hole density and thus Voc. In addition, grain boundary potential is ascertained to change favorably with the incorporation of Na. It is concluded that the Na-dissolved solution processing is promising and may

density. However, being shallow donors, they would not act as recombination centers. Hence, the reduction of the density of the donor-like In Cu antisite defects facilitated by Na incorporation is expected to increase the net hole density and thus the Voc, whereas the Jsc would remain almost unaffected. The formation of electrically neutral NaCu defects may decrease the charge compensation. This hypothesis explains our observed results of increasing carrier density and the trends of Voc and Jsc up to 0.25% Na. The other factor such as the formation of acceptor-type NaIn antisitic defects could also boost the p-type conductivity. However, a density functional theory-based calculation revealed that NaCu compensating InCu defects forms more readily than NaIn, especially in the Cu-poor compositions typical of the high efficiency solar cells.17 On the other hand, for exceeding concentrations, after replacing InCu antisites, Na would start filling acceptor-like Cu vacancies,14 reducing the carrier concentration and hence, the Voc. However, this alone would not be sufficient to account for the loss of FF and Jsc for the 0.50% Na, inferior even to the 0% Na sample. Further study of the samples was carried out using the PL measurements and the typical PL spectra are shown in Figure 2b. The films with lower Na content exhibited a lone emission peak at ∼0.94 eV (P1) and its intensity decreased with increasing Na content. For the higher content of Na (0.50% Na film), the peak became broader suggesting evolution of another peak at higher energy around 1.00 eV (P2). To further explore this behavior, we prepared an additional sample with 1.0% Na, which showed a stronger P2 peak compared to the P1 peak. Many studies have revealed that the formation enthalpy of various point defects in CIGS thin films is very small and even negative.18,19 Correspondingly, large quantities of the native point defects are formed in the CIGS thin films. Owing to these abundant defects, the room temperature emission of the CIGS thin films is primarily determined by the localized carriers in the donor and the acceptor states and the emission peaks usually appear at energies below the bandgap energy of 1.1−1.2 eV.20,21 Thus, the observed peaks P1 and P2 are attributed to the donor−acceptor pair (DAP) luminescence. Considering the typical bandgap Eg of the films to be ∼1.2 eV, the sum of acceptor and donor ionization energies (ΔEa + ΔEd) corresponding to P1 and P2 becomes about 260 and 200 meV, respectively. In CIGS films, VCu and InCu defects occur followed by formation of (InCu-2 VCu) neutral defect complex and the uncompensated InCu acts as donors occurring at about 0.25 eV below the conduction band minimum19 and VCu as acceptor at about 0.03 eV above valence band maximum. Thus, P1 is believed to stem from the transitions associated with the InCu − VCu levels. We suggest that the added Na eliminates InCu by substituting for In on Cu site,21 and hence, with increasing Na amount results in a decrease in the InCu defect density, which may explain the decreasing tendency of the emission intensity of the peak P1 (InCu defect-related, 0.94 eV). This explanation appears consistent with that of Lee at al.22 who studied the carrier dynamics on CdS/CIGS samples using optical pump terahertz probe spectroscopy. They found that the CIGS films grown on borosilicate glass substrates had a PL peak at ∼0.92 eV and a shorter relaxation time compared to the PL peak of 0.97 eV with a relatively longer relaxation time observed for those grown on the soda-lime substrates. It was suggested that Na from the soda-lime substrates eliminated the InCu defects and relieved the defect level located in the band gap. However, excessive incorporation of Na possibly led to D

DOI: 10.1021/acsami.6b04407 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Letter

ACS Applied Materials & Interfaces

(12) Lindahl, J.; Zimmermann, U.; Szaniawski, P.; Torndahl, T.; Hultqvist, A.; Salome, P.; Platzer-Bjorkman, C.; Edoff, M. Inline Cu(In,Ga)Se2 Co-Evaporation for High-Efficiency Solar Cells and Modules. IEEE J. Photovolt. 2013, 3, 1100−1105. (13) Nakada, T.; Iga, D.; Ohbo, H.; Kunioka, A. Effects of Sodium on Cu(In,Ga)Se2-Based Thin Films and Solar Cells. Jpn. J. Appl. Phys. 1997, 36, 732−737. (14) Wei, S. H.; Zhang, S. B.; Zunger, A. Effects of Na on the Electrical and Structural Properties of CuInSe2. J. Appl. Phys. 1999, 85, 7214−7218. (15) Niles, D. W.; Ramanathan, K.; Hasoon, F.; Noufi, R.; Tielsch, B. J.; Fulghum, J. E. Na Impurity Chemistry in Photovoltaic CIGS Thin Films: Investigation with X-Ray Photoelectron Spectroscopy. J. Vac. Sci. Technol., A 1997, 15, 3044−3049. (16) Pohl, J.; Albe, K. Intrinsic Point Defects in CuInSe2 and CuGaSe2 as Seen via Screened-Exchange Hybrid Density Functional Theory. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 245203. (17) Oikkonen, L. E.; Ganchenkova, M. G.; Seitsonen, A. P.; Nieminen, R. M. Effect of Sodium Incorporation into CuInSe2 from First Principles. J. Appl. Phys. 2013, 114, 083503. (18) Zhang, S.; Wei, S.-H.; Zunger, A.; Katayama-Yoshida, H. Defect Physics of the CuInSe2 Chalcopyrite Semiconductor. Phys. Rev. B: Condens. Matter Mater. Phys. 1998, 57, 9642−9656. (19) Wei, S. H.; Zhang, S. B.; Zunger, A. Effects of Ga Addition to CuInSe2 on Its Electronic, Structural, and Defect Properties. Appl. Phys. Lett. 1998, 72, 3199−3201. (20) Bacewicz, R.; Zuk, P.; Trykozko, R. Photoluminescence Study of ZnO/CdS/Cu(In,Ga)Se2 Solar Cells. Opto-Electron. Rev. 2003, 11, 277−280. (21) Shin, Y. M.; Lee, C. S.; Shin, D. H.; Ko, Y. M.; Al-Ammar, E. A.; Kwon, H. S.; Ahn, B. T. Characterization of Cu(In,Ga)Se2 Solar Cells Grown on Na-Free Glass with an NaF Layer on a Mo Film. ECS J. Solid State Sci. Technol. 2013, 2, P248−P252. (22) Lee, W. J.; Cho, D. H.; Wi, J. H.; Han, W. S.; Chung, Y. D.; Park, J.; Bae, J. M.; Cho, M. H. Na-Dependent Ultrafast Carrier Dynamics of CdS/Cu(In,Ga)Se2 Measured by Optical PumpTerahertz Probe Spectroscopy. J. Phys. Chem. C 2015, 119, 20231− 20236. (23) Sadewasser, S.; Glatzel, T.; Rusu, M.; Jäger-Waldau, A.; LuxSteiner, M. C. High-Resolution Work Function Imaging of Single Grains of Semiconductor Surfaces. Appl. Phys. Lett. 2002, 80, 2979− 2981. (24) Jiang, C. S.; Noufi, R.; AbuShama, J. A.; Ramanathan, K.; Moutinho, H. R.; Pankow, J.; Al-Jassim, M. M. Local Built-In Potential on Grain Boundary of Cu(In,Ga)Se2 Thin Films. Appl. Phys. Lett. 2004, 84, 3477−3479. (25) Yan, Y.; Jiang, C. S.; Noufi, R.; Wei, S. H.; Moutinho, H. R.; AlJassim, M. M. Electrically Benign Behavior of Grain Boundaries in Polycrystalline CuInSe2 Films. Phys. Rev. Lett. 2007, 99, 235504.

have a room for further enhancement particularly when combined with postsulfurization process.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b04407. (i) Detailed experimental procedure, (ii) cross-sectional SEM images, (iii) XPS profiles, and (iv) XRD patterns of absorber films (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: 82-2-2123-5848. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was financially supported by grants (2011-0020285) from the National Research Foundation of Korea (NRF) funded by the Korean government.



REFERENCES

(1) Septina, W.; Kurihara, M.; Ikeda, S.; Nakajima, Y.; Hirano, T.; Kawasaki, Y.; Harada, T.; Matsumura, M. Cu(In,Ga)(S,Se)2 Thin Film Solar Cell with 10.7% Conversion Efficiency Obtained by Selenization of the Na-Doped Spray-Pyrolyzed Sulfide Precursor Film. ACS Appl. Mater. Interfaces 2015, 7, 6472−6479. (2) Gifford, J. Solar Frontier hits 22.3% on CIGS Cell. PV Mag. 2015, http://www.pv-magazine.com/news/details/beitrag/solar-frontierhits-223-on-cigs-cell_100022342/#axzz45roL5wnL. (3) Akhavan, V. A.; Goodfellow, B. W.; Panthani, M. G.; Reid, D. K.; Hellebusch, D. J.; Adachi, T.; Korgel, B. A. Spray-Deposited CuInSe2 Nanocrystal Photovoltaics. Energy Environ. Sci. 2010, 3, 1600−1606. (4) Garg, P.; Garg, A.; Rastogi, A. C.; Garg, J. C. Growth and Characterization of Electrodeposited CuInSe2 Thin Films from Seleno-Sulphate Solution. J. Phys. D: Appl. Phys. 1991, 24, 2026−2031. (5) Todorov, T. K.; Gunawan, O.; Gokmen, T.; Mitzi, D. B. SolutionProcessed Cu(In,Ga)(S,Se)2 Absorber Yielding a 15.2% Efficient Solar Cell. Prog. Photovoltaics 2013, 21, 82−87. (6) Choi, I. J.; Jang, J. W.; Lee, S. M.; Yeon, D. H.; Jo, Y. H.; Lee, M. H.; Yun, J. H.; Yoon, K. H.; Cho, Y. S. Electrical and Photovoltaic Characteristics of CuInSe2 Thin Films Processed by Nontoxic Cu−In Precursor Solutions. J. Phys. D: Appl. Phys. 2013, 46, 245102. (7) Hedstrom, J.; Ohlsen, H.; Bodegard, M.; Kylner, A.; Stolt, L.; Hariskos, D.; Ruckh, M.; Schock, H. W. ZnO/CdS/Cu(In,Ga)Se2 Thin Film Solar Cells with Improved Performance. Proceedings of the 23rd IEEE Photovoltaic Specialist Conference; Louisville, KY, May 10− 14, 1993 ; IEEE: Piscataway, NJ, 1993; pp 364−371. (8) Rudmann, D.; da Cunha, A. F.; Kaelin, M.; Kurdesau, F.; Zogg, H.; Tiwari, A. N.; Bilger, G. Efficiency Enhancement of Cu(In,Ga)Se2 Solar Cells due to Post-Deposition Na Incorporation. Appl. Phys. Lett. 2004, 84, 1129−1131. (9) Haug, F. J.; Rudmann, D.; Zogg, H.; Tiwari, A. N. Light Soaking Effects in Cu(In,Ga)Se2 Superstrate Solar Cells. Thin Solid Films 2003, 431−432, 431−435. (10) Lee, M.; Lee, S. M.; Jung, S.; Ahn, S.; Cho, J.-S.; Park, J.; Eh, Y.; Gwak, J.; Shin, K.; Yoon, K.; Cho, Y. S.; Yun, J. H. Highly Efficient Flexible CuIn0.7Ga0.3Se2 Solar Cells with a Thick Na/Mo Layer Deposited Directly on Stainless Steel. Appl. Surf. Sci. 2015, 346, 562− 566. (11) Rockett, A.; Britt, J. S.; Gillespie, T.; Marshall, C.; Al Jassim, M. M.; Hasoon, F.; Matson, R.; Basol, B. Na in Selenized Cu(In,Ga)Se2 on Na-Containing and Na-Free Glasses: Distribution, Grain Structure, and Device Performances. Thin Solid Films 2000, 372, 212−217. E

DOI: 10.1021/acsami.6b04407 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX