# A Versatile Solution Route to Efficient Cu2ZnSn(S,Se)4 Thin-Film

This material is available free of charge via the Internet at http://pubs.acs.org. .... Fabrication of $$\hbox {Cu}_{2}\hbox {ZnSn(S,Se)}_{4}$$ Cu 2 Z...

Subscriber access provided by UNIV LAVAL

Article

A Versatile Solution Route to Efficient Cu2ZnSn(S,Se )4 Thin Film Solar Cells Ruihong Zhang, Stephen M. Szczepaniak, Nathaniel J. Carter, Carol A. Handwerker, and Rakesh Agrawal Chem. Mater., Just Accepted Manuscript • Publication Date (Web): 02 Mar 2015 Downloaded from http://pubs.acs.org on March 4, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

A Versatile Solution Route to Efficient Cu2ZnSn(S,Se )4 Thin Film Solar Cells Ruihong Zhang†, Stephen M. Szczepaniak‡, Nathaniel J. Carter‡, Carol A. Handwerker†§, Rakesh Agrawal*‡ †

School of Materials Engineering, Purdue University, West Lafayette, IN 47907, USA

School of Chemical Engineering, Purdue University, West Lafayette, IN 47907, USA

§

Birck Nanotechnology Center, Purdue University, West Lafayette, IN 47907, USA

KEYWORDS Cu2ZnSnS4, Cu2ZnSnSe4, photovoltaics, molecular precursor, solar cells

ABSTRACT: A simple solution-based approach for the deposition of Cu2ZnSn(S,Se)4 using an amine-thiol mixture is presented. The versatility of this solvent mixture in dissolving different cation sources and chalcogens opens a door for a variety of metal chalcogenide molecular precursor designs. The process involves incorporating the metal sources and chalcogens into an amine-thiol solvent mixture at room temperature, spin coating a precursor film, and heat treating precursor films in both an inert gas and selenium atmosphere. With this solution approach, high-quality kesterite CZTS and CZTSSe thin films were formed after low-temperature annealing and selenization. Solar cells were fabricated based on these CZTSSe thin films, resulting in a total area power conversion efficiency of 7.86% for a cell area of 0.47 cm2 under standard AM 1.5 illumination.

INTRODUTION Kesterite copper zinc tin sulfide, selenide and sulfoselenide, Cu2ZnSnS4 (CZTS), Cu2ZnSnSe4 (CZTSe), Cu2ZnSn(S,Se)4 (CZTSSe) are attractive absorber materials that have been well-studied since they contain earth-abundant elements and have higher absorption coefficients (>104 cm-1) compared to silicon.1-9 Solutionbased deposition of the above materials offers a number of potential advantages over vacuum deposition in terms of lower cost of raw materials and equipment, higher throughput production, and better compatibility with flexible substrates. Solution-based deposition can be further divided into two groups: colloidal deposition2,6,8-13 and molecular precursor deposition.14-19 Compared to the colloidal deposition method, the molecular precursor method offers the potential of eliminating/reducing the undesired fine grain layer that results from residual carbon-containing capping ligands and the compositional non-uniformity of nanoparticles.12,20,21 Meanwhile, the molecular precursor route circumvents additional steps associated with the nanoparticle synthesis, offers a facile control of the thin-film composition, and promises the scalability for mass production. With a hydrazine-based slurry and a hydrazine-based pure solution, IBM has

achieved a record power conversion efficiency of 11.1% and 12.6%, respectively, for kesterite CZTSSe solar cells.16,18,22 Despite its efficacy, hydrazine is highly toxic and dangerously unstable, and thus various handling precautions are required during film preparation. Therefore, developing alternative CZTSSe molecular precursors that have lower toxicity, are more costeffective, and produce equal or higher efficiencies than hydrazine-based solutions is one of the motivations of our study. Currently, a few non-hydrazine approaches have been adopted to deposit CZTSSe thin films. They can be categorized into three groups: hydrazine-derivative solution approaches,23,24 metal salts/thiourea-based solution approaches,15,19,25 and metal oxide-based solution approaches.26,27As an example of the first approach, Hsu et al. used hydrazinocarboxylic acid to dissolve zinc powders in order to prepare pure solution containing copper, zinc, tin, and sulfur, and obtained a solar cell efficiency of 7.2% on a cell area of 0.12 cm2.23 Among all the studies using thiourea as the sulfur source, Xin et al. obtained the highest power conversion efficiency, i.e., 8.32%, on an active area of 0.43 cm2 .19 It is interesting to note that in terms of human toxicity, the lowest toxicity solutions reported for the direct solution approach are

ACS Paragon Plus Environment

1

Chemistry of Materials

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ethanol-water based solution with dissolved metal salts and thiourea; the highest efficiency obtained for such method is 5.29%.25 In a recent study using the metal oxide-based solution approach, copper (I) oxide, zinc oxide, tin (II) oxide were dissolved in a mixture of 2methoxyethanol, thioglycolic acid, and monoethanolamine, obtaining a solar cell efficiency of 6.83% on an active area of 0.368 cm2.26 Here, we report a new, versatile, and low-toxicity molecular solution method for depositing CZTSSe thin films using low-cost amine-thiol solvent mixtures. Different from the above reported studies, this precursor system offers three major advantages. First of all, the elemental metals, metal oxides, elemental chalcogens, and metal salts are soluble in these versatile solvent mixtures, providing many possible pure molecular precursor routes to CZTSSe thin film solar cells. Second, the cation-toanion ratio and the anion-to-anion ratio can be adjusted freely and independently during the precursor preparation. Third, we have found that many metal and chalcogen sources are soluble in a number of mixtures of primary, secondary, tertiary amines and/or diamines with monothiols and/or dithiols.28 Recently, Brutchey’s group also reported that tin (II) sulfide can be dissolved in diamine-dithiol mixture29 and elemental tellurium can be dissolved using diamine-ethanethiol mixture.30 The study presented here focuses on using mixtures of a primary amine and a monothiol to prepare molecular precursors. Our early experiments with this molecular precursor route have resulted in a power conversion efficiency of 7.86% (8.09%) for a total area of 0.47 cm2 (an active area of 0.456 cm2) under standard AM 1.5 illumination. These results indicate the promise of using this facile solution method to fabricate high-quality CZTSSe thin film solar cells. EXPERIMENTAL SECTION

Molecular precursor preparation: In the present work, “molecular precursor” or “precursor solution” refers to a pure solution in which the elements are homogeneously mixed at a molecular level. The CZTSSe molecular precursors were prepared by dissolving the cation sources and elemental chalcogens in a solvent mixture of a primary amine and a monothiol. In the experiment for device fabrication, solution A was prepared by dissolving CuCl, ZnCl2, and SnCl2 into a mixture of hexylamine and propanethiol (volume ratio 4:1). After stirring the solution at room temperature for 2 hrs, the metal chloride salts were fully dissolved. Solution B was prepared by dissolving the Se powder and S flakes into a combination of hexylamine and propanethiol (volume ratio 2:1). The total concentration of chalcogens (S+Se) was 2 M, and [S]:[Se] can be adjusted to modify the grain growth during thin film selenization and the band gap of the resulting thin film. For the devices

Page 2 of 9

reported here, the concentration ratio of [S]:[Se] was equal to 4:1. The CZTSSe precursor solution was prepared by mixing solution A with solution B in a volume ratio of 5:1. After complete mixing, the solutions are transparent orange, and the color varies slightly with concentration ratio of [S]:[Se] contained in solution B. All operations were performed at ambient temperature (25oC±2oC) in a N2 glovebox with water and oxygen less than 0.1 ppm. When exploring the other possible precursor routes, the solvent mixture of at least a primary amine and a monothiol was also used to dissolve alternative cationcontaining chemicals (see supporting information). Device fabrication: The CZTSSe molecular precursor was prepared following the aforementioned procedures. Then the CZTSSe precursor solution was spin coated on a one-by-one inch molybdenum-sputtered (~800 nm) soda lime glass (SLG) substrate. After spin coating a layer, the coated substrate was annealed inside a heating chamber with an argon atmosphere at around 250oC for 5 mins in order to evaporate the solvents. This coating step was repeated until the desired film thickness was obtained. Then the thin film was further annealed at 500oC in selenium atmosphere for 30 mins to achieve a higher crystallinity and promote grain growth. The selenization condition is described in detail in the supporting information. Solar cells were fabricated from the above-described CZTSSe films by creating the following additional layers: chemical bath deposition of ~50 nm cadmium sulfide (CdS), sputtering of ~80 nm intrinsic zinc oxide (ZnO) and ~220 nm indium-doped tin oxide (ITO). On the top of the device, Ni/Al metal contacts were deposited by electronbeam deposition. For some devices, a ~100 nm antireflective magnesium fluoride (MgF2) coating was applied on the top to enhance the light absorption. Characterization: The material phases, film morphology, and compositions were characterized for each step using X-ray diffraction (XRD), Raman spectroscopy, and scanning electron microscopy (SEM) equipped with energy X-ray dispersive spectroscopy (EDX) in FEI Quanta equipped with an Oxford Aztec Xstream-2 silicon drift detector. The J-V characteristics were measured with a four-point probe station using a Keithley 2400 series sourcemeter and a Newport Oriel simulator with AM 1.5 illumination. The solar simulator was calibrated to 100 mW/cm2 using a Si reference cell certified by NIST. External quantum efficiency was studied under 0 V and -1 V bias in order to gain more insight into the solar cell performance. RESULT AND DISCUSSION

ACS Paragon Plus Environment

2

Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Molecular precursors with desired elements mixed on a molecular scale have been actively studied in the deposition of semiconducting materials, especially metal chalcogenides, for thin film transistors (TFTs), 31-33 solar cells,16-19,24 and thermoelectrics34 applications. The molecular-level homogeneity enables the precise control of stoichiometry and the enhanced spatial uniformity in as-deposited films, and in turn, an improved device performance. In the present work, amine-thiol based solvent mixtures were employed to dissolve metal chloride salts and elemental chalcogens in order to form CZTSSe molecular precursors. The metal chlorides, CuCl, ZnCl2, and SnCl2, were simultaneously dissolved in a hexylamine-propanethiol solvent mixture (vol. ratio 4:1), ultimately yielding a transparent yellow solution. Figure 1a shows the example solution A containing the concentration ratio of [Cu]:[Zn]:[Sn]=1.53:1.05:1 and [Sn]=0.1 M. Note that the highest concentration of Sn dissolved in this solvent mixture was 0.3 M and other cations were in the same ratio with respect to Sn as in the example solution A. The chalcogen precursor, solution B, was prepared by dissolving elemental S and Se into hexylamine and propanethiol (vol. ratio 2:1). The example solution B in Figure 1b contains 2 M of [S]+[Se] with [S]:[Se]=4:1, showing a dark red color. Higher concentration, up to 4 M of [S]+[Se] (any concentration ratio between S and Se) can be prepared by the same method. The CZTSSe molecular precursor was then obtained by mixing the cation precursor (solution A) and chalcogen precursor (solution B) at a desired volume ratio (e.g. vol. ratio 5:1). After mixing solution A and solution B, a transparent orange precursor solution (CZTSSe-1) was obtained, as shown in Figure 1c. The color of the precursor becomes lighter when the amount of dissolved S relative to Se increases. For example, CZTSSe-2 precursor was prepared by mixing solution A and another chalcogen precursor with [S]+[Se]=2 M, [S]:[Se]=6:1. In addition to metal chlorides and elemental chalcogenides, this solution mixture can dissolve many other cation precursors, including elemental Zn, ZnO, Zn(OAc)2, Cu2O, Cu(OAc)2, CuCl2, Cu(acac)2, and Sn(acac)2Cl2. The pictures of the solutions are shown in Figure S1. It should be noted that the amine and thiol used in the amine-thiol mixture can be different types of primary amines and monothiols. For example, the primary amine can be butylamine and hexylamine, while the monothiol can be ethanethiol and propanethiol. The images of the solutions prepared using different aminethiol solvent mixtures are shown in Figure S2. It is believed that there is a universal dissolution mechanism related to this versatile solvent mixture. The proposed dissolution mechanism of elemental selenium has been described in our previous work by Walker et al.35 and indepth mechanistic studies of the CZTSSe molecular precursors is in progress. For fulfilling the purpose of

fabricating carbon-free thin films with no fine-grained layer, the use of relative low boiling point amine/s and thiol/s is necessary, where “low boiling point” generally means <250oC.

Figure 1. Images of molecular precursor solutions prepared using hexylamine-propanethiol solvent mixtures. a) Solution A: cation precursor containing CuCl, ZnCl2, and SnCl2 with a concentration ratio of [Cu]:[Zn]:[Sn] =1.53:1.05:1 and [Sn]=0.1 M in hexylamine and propanethiol (vol. ratio 4:1). b) Solution B: chalcogen precursor with a concentration ratio of [S]:[Se]=4:1 [S]+[Se]=2 M in hexylamine and propanethiol (vol. ratio 2:1). c) CZTSSe-1: mixture of Solutions A and B (vol. ratio 5:1). d) CZTSSe-2: mixture of Solution A with another chalcogen precursor with [S]:[Se]=6:1, shown to illustrate color changes with changing [S]:[Se].

After eight spin coating steps, a CZTSSe precursor film of 800 nm to 1 µm thick was fabricated. The film showed good optical reflectivity after the solvent evaporation process, indicating a low film porosity (Figure 2). The top-view SEM image (Figure 3a) confirms that this film is smooth and continuous despite a few shallow cracks. EDX linescans were employed in order to measure the composition fluctuation at the microscale within the CZTSSe precursor thin film. Figure 4 shows the elemental distribution of Cu, Zn, Sn, S, and Se within the precursor film along a >10 µm distance on the top surface and indicates that a more uniform elemental distribution is obtained using our pure solution method in comparison to the hydrazine slurry-processed CZTSSe precursor film.16 Figure 3b is a typical cross-sectional SEM image of an 8layer precursor film showing a uniform film structure throughout the thickness.

Figure 2. Images of the precursor film and the selenized film on Mo-sputtered soda-lime glass substrates.

ACS Paragon Plus Environment

3

Chemistry of Materials

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. a) Top-view SEM image of the precursor film fabricated by spin-coating and subsequent annealing, and b) Cross-sectional SEM image of the precursor film.

Page 4 of 9

the kesterite CZTSSe phase was obtained after the solvent evaporation process. Based on the SEM-EDX analysis (see Table I), the film contains copper, zinc, tin, sulfur, selenium, and trace chloride. The concentration of chloride ions dropped to a very low level in the as-annealed precursor film. It has been demonstrated that a higher temperature annealing (e.g. ≥300oC) can decrease the concentration of chloride in the precursor film to below the detection limit of EDX. The chloride ions may leave the film in various forms, such as chlorine gas, Se2Cl2 (boiling point at 127oC) and/or SeCl4 (sublimes at 191.4oC). The disassociation mechanism of the chloride ions deserves further study. After the film deposition and solvent evaporation/annealing steps, the films that formed contained a combination of CZTSSe nanoparticles and possible amorphous phases.

Figure 4. EDX linescan along precursor film surface (>10 µm scan) showing compositional uniformity.

Grazing incident XRD was employed to measure the nucleated phase after low-temperature annealing at 250oC. As shown in Figure 5, all the XRD peaks are consistent with the kesterite CZTS standard with a very slight shift to lower angles, indicating the substitution of some Se atoms at S positions. In order to get more insight into the phases present in the deposited thin films, Raman spectra were obtained using a HORIBA HR800 system with an excitation laser wavelength of 632.8 nm. In Figure 6, the peaks at 334 cm-1, and 364 cm-1 correspond to kesterite CZTS, while peaks at 223 cm-1, 238 cm-1, and 297 cm-1 correspond to sulfoselenide which contains both S and Se at the anion sites in the crystal lattice. The peak position and the details of peak fitting are shown in Figure S3. This bimodal behavior in the Raman spectra has been previously reported for samples with intermediate values of [S]/([S]+[Se]), with similar peak broadening and position shifts as we report here.5,24,36 In summary, four objectives have been achieved using this newly developed CZTSSe molecular precursor: (1) four to five cations/anions were dissolved at significant concentrations to obtain the above molecular precursor; (2) the precursor solution remained stable allowing the deposition of compositionally homogeneous precursor thin films; (3) the choice of the specific amine-thiol mixture and subsequent drying process enabled the formation of continuous and smooth precursor films; (4)

Figure 5. XRD patterns of the precursor film and the selenized ഥ, film. The standards of kesterite CZTSe (space group: ࡵ૝ ഥ , simulated JCPDS 01-070-8930) and CZTS (space group: ࡵ૝ pattern based on JCPDS 26-057512) are marked at the bottom of the plot.

ACS Paragon Plus Environment

4

Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Figure 6. Raman spectra of the precursor film and the selenized film. The Raman standards for CZTS and CZTSe are labeled at the bottom of the plot.36-38 The position of the fitting peaks are labeled on the spectra. The peak fittings of the Raman spectra are plotted in Figure S3 and the details of the peak fitting are provided in supporting information.

The selenization was performed at 500oC for 30 min under the saturated selenium atmosphere in a tube furnace with a constant Ar flow at 10 sccm. After selenization, the film converted into a high-crystallinity CZTSSe thin film as shown in Figure 7. The top-view SEM image in Figure 7a shows a typical microstructure which consists of a high-density film with a rough, highly facetted surface, and an average in-plane grain size of approximately 600 nm. The fracture cross-section of the device in Figure 7b shows that the fine grain layer/unsintered layer near the molybdenum substrate is very thin (< 150 nm) compared to other solution method studies.19,23,26

composition of the selenized film is Cu1.73±0.01Zn1.07±0.01Sn1(S,Se)4.43±0.02, and the concentration ratio of [S]/([S]+[Se])=0.08. Table I. Composition of precursor film and selenized film. Atomic ratio

Solution

Precursor film

Selenized film

[Cu]/[Sn]

1.53

1.62±0.02

1.73±0.01

[Zn]/[Sn]

1.05

1.06±0.02

1.07±0.01

4

4.27±0.01

4.43±0.02

5.63

0.3±0.01

-

[S]+[Se]/[Sn] [Cl]/[Sn]

Figure 7. a) Top-view SEM image of large-grain selenized film, and b) fractured cross-section of the CZTSSe device prepared following procedures described in Experimental Section.

The XRD pattern in Figure 5 also indicates the formation of a high-crystallinity kesterite CZTSSe film, with a small amount of S substituting for Se. In the corresponding Raman spectrum (Figure 6), peaks at 174 cm-1, 197 cm-1 (A1 mode), 234 cm-1, and 244 cm-1 correspond to kesterite CZTSe while the peak at 330 cm-1 corresponds to kesterite CZTS. Although the peak at 252 cm-1 might indicate the presence of ZnSe in the selenized film, ZnSe phase should not contribute to the spectra under a 633 nm laser excitation.37,38 Since XRD and Raman spectroscopy with 633 nm laser excitation cannot totally rule out the existence of secondary phase (e.g. ZnS/ZnSe), other characterization techniques are required for more accurate phase identification (e.g. Raman spectroscopy with 325 nm laser excitation). Focused ion beam milling (FIB) was used to prepare a cross-section of a CZTSSe solar cell and then SEM-EDX mapping was performed on this cross-section. The compositional uniformity of Cu, Zn, Sn, S, and Se in the absorber thin film is confirmed by the element mapping (Figure 8). The SEM-EDX linescan on a fractured cross-section (Figure S4) also confirms this. As shown in Table 1, the

Figure 8. EDX mapping of the CZTSSe device. A platinum coating was employed to protect the film during the FIB-cross section preparation.

Since the concentration of copper ions relative to other cations is critical to the defect density (e.g., [2CuSn +SnZn], [VCu+ZnCu]) in CZTSSe bulk materials,39 solution recipes with different [Cu]/[Sn] ratios varying from 1.45 to 1.74 were used to fabricate the CZTSSe thin film solar cells (Table S1). Based on the GIXRD and Raman spectra (see Figure S5), the kesterite CZTSSe phase was obtained for all the solution recipes without the appearance of secondary phases after the solvent evaporation and the subsequent selenization. The photovoltaic performance of the solar cells prepared using the aforementioned precursor route was measured under standard AM1.5 illumination. The characteristics of the champion cells for four precursor recipes are summarized in Table II and the efficiency statistics are plotted in Figure S6. In general, solar cells fabricated using solution recipe 2 exhibit the best solar

ACS Paragon Plus Environment

5

Chemistry of Materials

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

cell performance. Figure 9 shows the current densityvoltage (J-V) data measured under standard AM1.5 illumination based on a total cell area of 0.47 cm2. The solar cell has a total area power conversion efficiency of 7.86 % (8.09% for an active area of 0.456 cm2) with Voc=382 mV, Jsc=34.4 mA/cm2, and fill factor=60.1%. For further insight into device performance, external quantum efficiency (EQE) response was measured at 0 V and -1 V (Figure 10). For both measurements, the EQE is >95% in the visible range and gradually decays in the longer wavelengths. The collection efficiency at the longer wavelength range is more likely limited by the low carrier lifetime, which is confirmed by the measurement of EQE (-1 V)/EQE (0 V) ratio. The linear increase of EQE (-1 V)/EQE (0 V) ratio at longer wavelength indicates that when the depletion width is increased, the collection efficiency is increased. The low lifetime could be a result of a high defect density in the absorber layer or a high recombination loss at the interface.5,40-42 The bandgap of the CZTSSe film is estimated based on the plot of [ln(1EQE)]2 versus the photon energy as well as the inflection of the EQE curve (i.e. the peak of the dEQE/dE curve, where E=hc/λ) as shown in Figure 10. The two estimation methods result in the same bandgap value of 1.08 eV. This low bandgap value might be due to a higher concentration of Se in CZTSSe phase after selenization, which is shown in the composition measurement (Table I). Table II. Device characteristics of CZTSSe cells. Solution recipe [Cu]:[Sn]

Voc (mV)

Jsc (mA/cm2)

FF (%)

PCE (%)

R1

1.45:1

381

34.1

58.4

7.24*

R2

1.53:1

382

34.4

60.1

7.86*

R3

1.64:1

367

31.4

55.5

6.34

R4

1.74:1

270

30.0

30.7

2.47

For a total area of 0.47 cm2. * With an antireflective coating.

Page 6 of 9

Figure 9. J–V curves and performance parameters for the best performing cell in the dark and under AM1.5 illumination. The dotted line shows the dark current density, while the solid line represents the light current density. The series resistance and shunt resistance under light are listed. The inset is the final device.

As described above, different types of primary amine can be used in the solvent mixture to prepare CZTSSe molecular precursors. Kesterite CZTSSe phase was also obtained after the final selenization when butylamine and propanethiol were used as the solvents to dissolve CuCl, ZnCl2, SnCl2, S, and Se. The XRD of the resulting CZTSSe films are shown in Figure S7. However, it is interesting that the precursor films formed at different annealing temperatures under N2 atmosphere contain wurtzite CZTS phase rather than kesterite CZTS phase. This indicates that the choice of amine or the amine-tothiol ratio might lead to a metastable phase during the low-temperature annealing. As expected, after selenization at 500oC, the wurtzite film transformed into large grains of kesterite phase.13,43 Devices were found to have lower power conversion efficiencies compared to the case of using hexylamine (Figure S7).

ACS Paragon Plus Environment

6

Page 7 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Raman spectra of CZTSSe thin films prepared using different [Cu]/[Sn] ratios, the statistics of CZTSSe solar cells prepared using different [Cu]/[Sn] ratios, the characterization of CZTSSe solar cells prepared using butylamine-propanethiol based precursor solutions, selenization conditions, and dynamic light scattering measurement. This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION

Corresponding Author *Rakesh Agrawal. E-mail: [email protected] Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS Figure 10. EQE measurement of 7.86% device under 0 V and 1 V bias. The inset shows the band gap estimated using [ln(1EQE)]2 versus the photon energy as well as the inflection of the EQE curve. The estimated band gap is 1.08 eV.

CONCLUSION

We present a novel, versatile, and lower-toxicity molecular precursor route to CZTSSe thin films for solar cell applications. The versatility of amine-thiol based solutions has been demonstrated in three aspects: 1) pure solutions can be obtained by dissolving different types of cation and anion sources in the amine-thiol solvent mixture. 2) amine/s and thiol/s employed in a solvent mixture can be any type of primary amine and monothiol, offering the flexibility to change the dissolution process, film fabrication process, and film morphology. 3) cationto-anion and anion-to-anion ratios can be adjusted freely and independently. As an example of using this versatile solvent mixture, a CZTSSe molecular precursor is obtained by dissolving CuCl, ZnCl2, SnCl2, S, and Se in hexylamine and propanethiol. After the deposition and the annealing processes, kesterite CZTS phase formed with no secondary phases detected. The kesterite CZTS was converted into a highly crystalline CZTSSe phase during selenization. The homogeneity of the precursor solution ensures the purity of the resulting phase, a thin fine grain layer, and in turn a good solar cell performance. With this molecular precursor route, a power conversion efficiency of 7.86% (8.09%) has been achieved for a total area of 0.47 cm2 (an active area of 0.456 cm2) under AM 1.5 illumination. We believe that further optimization in the precursor composition, annealing temperature, and selenization conditions will result in higher PCEs. ASSOCIATED CONTENT The supporting information includes the images of aminethiol based solutions and precursors, the peak fittings for Raman spectra corresponding to Figure 6, XRD and

The authors acknowledge Wei-Chang Yang and Charles J. Hages for assistance in SEM sample preparation and data interpretation. The authors also thank Dr. Bryce Walker and Caleb K. Miskin for very helpful discussion, and Kevin Brew and Brian Graeser for their assistance and expertise in preparing the Mo-coated soda lime glass. This research was funded from the NSF Solar Economy IGERT program (0903670-DGE) and the DOE SunShot program (DE-EE0005328). REFERENCES (1) Ahn, S.; Jung, S.; Gwak, J.; Cho, A.; Shin, K.; Yoon, K.; Park, D.; Cheong, H.; Yun, J. H. Appl Phys Lett 2010, 97, 2, 021905. (2) Guo, Q.; Ford, G. M.; Yang, W.-C.; Walker, B. C.; Stach, E. A.; Hillhouse, H. W.; Agrawal, R. J Am Chem Soc 2010, 132, 17384. (3) Ito, K.; Nakazawa, T. Japanese Journal of Applied Physics 1998, 27, 2094. (4) Katagiri, H.; Jimbo, K.; Maw, W. S.; Oishi, K.; Yamazaki, M.; Araki, H.; Takeuchi, A. Thin Solid Films 2009, 517, 2455. (5) Mitzi, D. B.; Gunawan, O.; Todorov, T. K.; Wang, K.; Guha, S. Solar Energy Materials and Solar Cells 2011, 95, 1421. (6) Riha, S. C. P., B. A.; Prieto, A.L. J Am Chem Soc 2009, 131, 12054. (7) Zhou, H.; Hsu, W.-C.; Duan, H.-S.; Bob, B.; Yang, W.; Song, T.B.; Hsu, C.-J.; Yang, Y. Energ Environ Sci 2013, 6, 2822. (8) Guo, Q.; Hillhouse, H. W.; Agrawal, R. Journal of the American Chemical Society 2009, 131, 11672. (9) Steinhagen, C.; Panthani, M. G.; Akhavan, V.; Goodfellow, B.; Koo, B.; Korgel, B. A. Journal of the American Chemical Society 2009, 131, 12554. (10) Cao, Y.; Denny, M. S., Jr.; Caspar, J. V.; Farneth, W. E.; Guo, Q.; Ionkin, A. S.; Johnson, L. K.; Lu, M.; Malajovich, I.; Radu, D.; Rosenfeld, H. D.; Choudhury, K. R.; Wu, W. J Am Chem Soc 2012, 134, 15644. (11) Guo, Q.; Kim, S. J.; Kar, M.; Shafarman, W. N.; Birkmire, R. W.; Stach, E. A.; Agrawal, R.; Hillhouse, H. W. Nano Lett 2008, 8, 2982. (12) Miskin, C. K.; Yang, W.-C.; Hages, C. J.; Carter, N. J.; Joglekar, C. S.; Stach, E. A.; Agrawal, R. Progress in Photovoltaics: Research and Applications 2014, DOI: 10.1002/pip.2472. (13) Yang, W. C.; Miskin, C. K.; Hages, C. J.; Hanley, E. C.; Handwerker, C.; Stach, E. A.; Agrawal, R. Chem Mater 2014, 26, 3530.

ACS Paragon Plus Environment

7

Chemistry of Materials

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(14) Bag, S.; Gunawan, O.; Gokmen, T.; Zhu, Y.; Mitzi, D. B. Chem Mater 2012, 24, 4588. (15) Schnabel, T.; Löw, M.; Ahlswede, E. Solar Energy Materials and Solar Cells 2013, 117, 324. (16) Todorov, T.; Sugimoto, H.; Gunawan, O.; Gokmen, T.; Mitzi, D. B. Ieee J Photovolt 2014, 4, 483. (17) Todorov, T. K.; Gunawan, O.; Gokmen, T.; Mitzi, D. B. Prog Photovoltaics 2013, 21, 82. (18) Todorov, T. K.; Tang, J.; Bag, S.; Gunawan, O.; Gokmen, T.; Zhu, Y.; Mitzi, D. B. Adv Energy Mater 2013, 3, 34. (19) Xin, H.; Katahara, J. K.; Braly, I. L.; Hillhouse, H. W. Adv Energy Mater 2014, 1301823. (20) Carter, N. J.; Yang, W. C.; Miskin, C. K.; Hages, C. J.; Stach, E. A.; Agrawal, R. Solar Energy Materials and Solar Cells 2014, 123, 189. (21) Yang, W.-C.; Miskin, C. K.; Carter, N. J.; Agrawal, R.; Stach, E. A. Chem Mater 2014, DOI: 10.1021/cm502930d. (22) Wang, W.; Winkler, M. T.; Gunawan, O.; Gokmen, T.; Todorov, T. K.; Zhu, Y.; Mitzi, D. B. Adv Energy Mater 2014, 4, 1301465. (23) Hsu, C.-J.; Duan, H.-S.; Yang, W.; Zhou, H.; Yang, Y. Adv Energy Mater 2014, 4, 1301287. (24) Yang, W.; Duan, H.-S.; Bob, B.; Zhou, H.; Lei, B.; Chung, C.H.; Li, S.-H.; Hou, W. W.; Yang, Y. Adv Mater 2012, 24, 6323. (25) Park, S.-N.; Sung, S.-J.; Son, D.-H.; Kim, D.-H.; Gansukh, M.; Cheong, H.; Kang, J.-K. Rsc Adv 2014, 4, 9118. (26) Tian, Q.; Wang, G.; Zhao, W.; Chen, Y.; Yang, Y.; Huang, L.; Pan, D. Chem Mater 2014, 26, 3098. (27) Wang, G.; Zhao, W.; Cui, Y.; Tian, Q.; Gao, S.; Huang, L.; Pan, D. ACS Appl Mater Interfaces 2013, 5, 10042. (28) Agrawal, R.; Zhang, R. H.; Walker, B. C.; Handwerker, C. A. Provisionl Patent Application, United States, Aug. 12, 2014, No. 62036233.

Page 8 of 9

(29) Antunez, P. D.; Torelli, D. A.; Yang, F.; Rabuffetti, F. A.; Lewis, N. S.; Brutchey, R. L. Chem Mater 2014, 26, 5444. (30) Webber, D. H.; Buckley, J. J.; Antunez, P. D.; Brutchey, R. L. Chemical Science 2014, 5, 2498. (31) Mitzi, D. B.; Kosbar, L. L.; Murray, C. E.; Copel, M.; Afzali, A. Nature 2004, 428, 299. (32) Milliron, D. J.; Mitzi, D. B.; Copel, M.; Murray, C. E. Chem Mater 2006, 18, 587. (33) Mitzi, D. B. J Mater Chem 2004, 14, 2355. (34) Cho, S.; Kim, Y.; Ketterson, J. B. Applied Physics A 2004, 79, 7, 1729. (35) Walker, B. C.; Agrawal, R. Chemical Communications 2014, 50, 8331. (36) Grossberg, M.; Krustok, J.; Raudoja, J.; Timmo, K.; Altosaar, M.; Raadik, T. Thin Solid Films 2011, 519, 7403. (37) Fella, C. M.; Uhl, A. R.; Hammond, C.; Hermans, I.; Romanyuk, Y. E.; Tiwari, A. N. J Alloy Compd 2013, 567, 102. (38) Šćepanović, M.; Grujić-Brojčin, M.; Nesheva, D.; Levi, Z.; Bineva, I.; Popović, Z. V. Acta Physica Polonica A 2009, 116, 708. (39) Chen, S.; Walsh, A.; Gong, X.-G.; Wei, S.-H. Adv Mater 2013, 25, 1522. (40) Gokmen, T.; Gunawan, O.; Todorov, T. K.; Mitzi, D. B. Appl Phys Lett 2013, 103. (41) Hages, C. J.; Carter, N. J.; Agrawal, R.; Unold, T. J Appl Phys 2014, 115, 234504. (42) Hegedus, S. S.; Shafarman, W. N. Progress in Photovoltaics: Research and Applications 2004, 12, 155. (43) Jiang, C.; Liu, W.; Talapin, D. V. Chem Mater 2014, 26, 4038.

ACS Paragon Plus Environment

8

Page 9 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials