Single-Crystal Growth and Characterization of the Chalcopyrite

Letter Cite This: J. Phys. Chem. Lett. 2018, 9, 6833−6840

pubs.acs.org/JPCL

Single-Crystal Growth and Characterization of the Chalcopyrite Semiconductor CuInTe2 for Photoelectrochemical Solar Fuel Production Jessica J. Frick,† Andreas Topp,‡ Sebastian Klemenz,† Maxim Krivenkov,§ Andrei Varykhalov,§ Christian R. Ast,‡ Andrew B. Bocarsly,† and Leslie M. Schoop*,† †

Department of Chemistry, Princeton University, Princeton, New Jersey 08544, United States Max Planck Institute for Solid State Research, Heisenbergstraße 1, 70569 Stuttgart, Germany § Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Berlin, Germany J. Phys. Chem. Lett. 2018.9:6833-6840. Downloaded from pubs.acs.org by IOWA STATE UNIV on 01/12/19. For personal use only.

‡

S Supporting Information *

ABSTRACT: Transition-metal chalcogenides are a promising family of materials for applications as photocathodes in photoelectrochemical (PEC) H2 generation. A long-standing challenge for chalcopyrite semiconductors is characterizing their electronic structure, both experimentally and theoretically, because of their relatively high-energy band gaps and spin−orbit coupling (SOC), respectively. In this work, we present single crystals of CuInTe2, whose relatively small optically measured band gap of 0.9 ± 0.03 eV enables electronic structure characterization by angle-resolved photoelectron spectroscopy (ARPES) in conjunction with firstprinciples calculations incorporating SOC. ARPES measurements reveal bands that are steeply dispersed in energy with a band velocity of 2.5−5.4 × 105 m/s, almost 50% of the extremely conductive material graphene. Additionally, CuInTe2 single crystals are fabricated into electrodes to experimentally determine the valence band edge energy and confirm the thermodynamic suitability of CuInTe2 for water redox chemistry. The electronic structure characterization and band edge position presented in this work provide kinetic and thermodynamic factors that support CuInTe2 as a strong candidate for water reduction.

P

the energetics of the valence and/or the conduction band can be manipulated to shift the absolute energy and position of the band gap. This approach has been used to study water splitting with chalcopyrite variants such as Cu(In,Ga)(S,Se)210−12 and CuIn(SxSe1−x)2 and (AgxCu1−x)(GayIn1−y)S2 in our own work.13−15 Though thermodynamic considerations provide an initial screening for photoelectrode materials, kinetic factors also play a dominant role in achieving efficient photoelectrodes for water reduction or oxidation.16 Upon illumination of a photocathode composed of a p-type semiconductor, two types of charge transfer must be considered: that of the dominant holes (majority carriers) and the nondominant electrons (minority carriers). Charge transfer of majority carriers through the semiconductor bulk has been identified as a major kinetic limitation for photoelectrodes.14,17 The bulk structure of a material’s valence band, through which majority carriers are transported, can provide hints about the intrinsic recombination rates of a photocathode and therefore its

hotoelectrochemical conversion of solar to chemical energy is an attractive approach for sustainable production of storable and transportable hydrogen fuel.1−3 However, to efficiently harness light to reduce water involves thermodynamic criteria and kinetic challenges. Thermodynamic considerations of a semiconductor’s electronic structure, such as absolute band gap energy and band edge positions, often provide a starting point for evaluating a material class for photoelectrochemical performance.4 Using these criteria, ternary transition-metal chalcogenides with the chalcopyrite crystal structure have been identified as major targets for future sustainable light-harvesting materials.5 These ternary chalcogenides have the general composition MxM′yEn (M = metal; E = S, Se, Te). The two metal cations, Mx and M′y, adopt the oxidation state +1 and +3, respectively, resulting in a chargebalanced semiconductor. In general, this class of materials possesses the two thermodynamic requirements for water reduction: light absorption well-matched to the terrestrial solar spectrum and a conduction band energy more negative than the reduction potential of water.6−9 A wide range of studies have demonstrated the tunability of ternary chalcopyrites to meet both these thermodynamic requirements. For example, by varying the cation ratio Mx/M′y or the anion ratio En/E′m, © 2018 American Chemical Society

Received: October 8, 2018 Accepted: November 15, 2018 Published: November 15, 2018 6833

DOI: 10.1021/acs.jpclett.8b03100 J. Phys. Chem. Lett. 2018, 9, 6833−6840

Letter

The Journal of Physical Chemistry Letters

Figure 1. (a) PXRD pattern (black) for prereacted polycrystalline CuInTe2 recorded at 300 K. The dotted gray lines show the alignment of the high-symmetry cubic diffraction peaks with the tetragonal structure. The dotted purple lines show the alignment of the tetragonal diffraction peaks with our experimental CuInTe2 XRD pattern. The lattice parameters were determined from refined XRD patterns: a = 6.204(1) Å and c = 12.404(3) Å. (b) UV−vis−NIR reflectance data of CuInTe2 plotted as (αhν)2 (eV/cm2) vs hν (eV). The optical band gap energy, Eg = 0.9 ± 0.03 eV, was estimated from extrapolating the linear absorption edge to the x axis. The inset shows α (cm−1) vs hν (eV), giving α ≈ 104 cm−1. (c) Resistivity temperature profile of CuInTe2 single crystal from 300 to 10 K. Resistivity increases with decrease in temperature, signifying semiconducting behavior. The inset shows a representative Hall measurement for CuInTe2 single crystals. The sample resistivity was monitored as a function of magnetic field from −9 to +9 T at 300 K. The red solid line corresponds to the linear fit, extracted to calculate p-type carrier concentration.

While experimental momentum-resolved electronic structure studies have been widely used to understand physical properties in semimetals (such as for example, the ultrahigh mobility in graphene33), they have been used less frequently in the study of photoactive materials. A reason might be that the most common technique for studying the electronic structure, angle-resolved photoemission spectroscopy (ARPES), is hard to perform on materials with a band gap that exceeds 1.5 eV in magnitude. For this reason, we chose to investigate the electronic structure of CuInTe2, which has a band gap of 0.9 ± 0.03 eV, which both meets the ARPES requirements and is a reasonable value with respect to the absorption of solar insolation. Utilizing a combination of traditional methods, including Xray diffraction and ultraviolet−visible (UV−vis) absorption spectroscopy, we have provided experimental verification of the known structural and optical properties of a bulk single crystal of CuInTe2 grown by the vertical Bridgman method. Bulk transport property measurements were used to estimate p-type carrier concentration and carrier mobility. Then, for the first time, ARPES is used to experimentally characterize CuInTe2’s valence electronic structure, revealing information about the slope and number of bands composing the valence band and indicating a high Fermi velocity. The measured data is compared with electronic structure calculations using DFT and is found to be in good agreement with our experimentally determined band gap value and valence band structure. Finally, we show evidence that the band edge position of CuInTe2 enables this variant of ternary chalcopyrite to serve as a photocathode for photoelectrochemical water splitting. The electronic structure characterization presented here provides critical insight into kinetic factors that govern efficiency of light-harvesting photocathodes similar to CuInTe2 that will influence future optimization of ternary chalcopyrites in photoelectrochemical H2 generation. Synthesis, Growth, and Characterization of CuInTe2. The PXRD pattern of prereacted polycrystalline CuInTe2 is shown in Figure 1a (black). The purple lines correspond to the tetragonal CuInTe2 chalcopyrite structure, which is derived from the cubic, zinc blende structure shown in gray. The

qualitative efficiency as a photocathode. A material that possesses a band structure composed of very steep bands in k-space is advantageous for photovoltaic applications, because the slope of a band is directly proportional to its band velocity. In addition, a linear band dispersion results in zero effective mass and thus high mobility.18−21 Carriers with high mobility have a higher chance to avoid charge recombination, especially those that are generated within a diffusion distance of the space charge region.14 However, few experimental electronic structure studies exist on such kinetic factors influencing the majority charge carrier transport through the bulk of a ternary chalcopyrite (or most other) light-harvesting semiconductor materials. Consider the popular photoanode material n-BiVO4 as an example. It is well-known that BiVO4 possesses poor majority charge carrier mobility as its major limiting factor for photoelectrochemical water splitting.17,22−26 Despite considerable attention drawn to this kinetic limitation, the literature has not yet come to a consensus on the nature of BiVO4’s fundamental band gap and valence band structure, an understanding of which is key to interpreting its majority carrier mobility limitations.16,27 Similar to BiVO4, computational studies predicting the band gap of ternary chalcopyrite derivatives, such as CuInTe2, are mostly inaccurate and imprecise, with the majority of calculations underestimating the reported optically measured band gap because of typical limitations of density functional theory (DFT).28−31 CuInTe2 has been known for decades to possess a direct band gap, determined through electroreflectance measurements,32 which gives rise to the material’s high optical absorption coefficient. However, the momentum resolved valence band structure of CuInTe2 has not been experimentally determined. However, it is exactly this property that will give important insight into the carrier mobility, and thus, it is important for assessing the ability of CuInTe2 to act as an efficient photoelectrode. Knowledge of the momentum-resolved electronic structure will therefore help establish rational approaches to improved photoelectrochemical quantum yield and energy conversion efficiency. 6834


Letter


previous studies have included spin−orbit coupling (SOC) effects, to the best of our knowledge, although optical measurements have suggested that SOC has an effect on the electronic structure of CuInTe2.38,39 In our calculations (calculation results seen in Figure S4), we compute a band gap of 0.77 eV if we employ the mBJ functional and include SOC and 0.91 eV if we omit SOC. Even if the latter value is in excellent agreement with the experimental band gap, we will show below that SOC has an effect on the experimentally determined electronic structure and can therefore not be omitted in the calculation. The obvious failure of DFT to provide a conclusive description of CuInTe2’s electronic structure highlights the importance of experimental investigations. ARPES Measurements. The unit cell type and size influence the resulting Brillouin zone (BZ) of a crystal lattice. For CuInTe2, a body-centered tetragonal BZ is anticipated, with a smaller size compared to a cubic BZ because of the larger unit cell. However, instead of observing the tetragonal BZ, our ARPES data reveals only the cubic BZ. Figure 2a shows the

dotted purple lines are a guide for the eye, showing our CuInTe2 material matches with the tetragonal phase rather than the cubic phase, resulting in two distinct lattice parameters where c > 2a. Single-crystal X-ray diffraction was performed to further verify its pure tetragonal phase (Tables S1−S3). Refining the X-ray data resulted in a tetragonal unit cell with lattice parameters a and c equal to 6.204(1) and 12.404(3) Å, respectively. In addition to the single-crystal Xray diffraction, the composition of the material was confirmed with EDX (Figure S1), and XPS was used to probe the chemical composition and oxidation state of freshly cleaved CuInTe2 surfaces (Figure S2). The direct optical band gap (Eg) measurement is shown in Figure 1b along with an inset plotting the material’s absorption coefficient (α). CuInTe2’s Eg measured 0.9 ± 0.03 eV, with α on the order of 102 cm−1. Details on the calculations of Eg and α can be found in the Supporting Information. Transport Properties of CuInTe2 Single Crystals. The temperature-dependent bulk resistivity of CuInTe2 was probed on cooling from 300 K using a four-point probe configuration. A representative temperature profile of electrical resistivity from 300 to 10 K is plotted in Figure 1c. The sample displays an exponentially increasing resistivity with decreasing temperature, as expected for a semiconducting material. At room temperature, the resistivity (ρ) is 6 × 10−3 Ω·cm. The majority carrier concentration of the CuInTe2 crystals were estimated by Hall measurements. Representative Hall resistivity data as a function of magnetic field (−9 to +9 T) at 300 K is shown in the inset of Figure 1c. The positive trend of resistivity with magnetic field further confirms p-type carriers as the dominant carrier type in the CuInTe2 crystals. A reliable linear fit (R2 = 0.99) allowed a confident calculation of the hole concentration np = 1.2 × 1018 cm−3. With the resistivity and majority carrier concentration in hand, the mobility (μ) of the p-type carriers was estimated to be 870 cm2 V−1 s−1, a startling two orders of magnitude greater than mobilities found for polycrystalline samples of CuInS2 and CuInSe2 reported from our lab.13 Compared to other common light-harvesting semiconductors, this calculated mobility is eight times larger than that of a single crystalline CH3NH3PbI3 perovskite (105 cm2 V−1s−1 by Hall measurements34) and is 3 orders of magnitude greater than recently reported mobilities of n-type BiVO4 single crystals measured by Hall measurements, 0.2 cm2 V−1s−1.35 In the case of halide perovskite materials, the mobility has, at least theoretically, been shown to stem from high band velocities,36 which are common in perovskite-type materials.18,19 The link between carrier velocity and reduced recombination rates has also been drawn in the same publication.36 Details on the calculations of np and μ can be found in the Supporting Information, along with a plot of the resistivity data as log(ρ) vs 1/kBT (Figure S3) which shows the extraction of CuInTe2’s activation energy (Ea). Theoretical Electronic Band Structure. The band structure of CuInTe2 has been theoretically reported numerous times using several different methods of calculations (see Table S4). While these calculations all agree that CuInTe2 has a direct band gap at the Γ point, with highly dispersed conduction and valence bands, which is ideal for light-harvesting materials, the vast majority of these calculations greatly underestimate the band gap of CuInTe2. On the other hand, DFT calculations using the modified Becke−Johnson (mBJ) functional, which has been shown to result in more accurate band gaps, overestimates the band gap in the CuInTe2 case.37 To date, no

Figure 2. (a) Cu and In cations in the CuInTe2 unit cell with the corresponding primitive vectors a1, a2, and a3 for both the tetragonal (black) and cubic (blue) symmetry. Their corresponding Brillouin zones are presented in black and blue. (b) Two orthographic side views on the two BZs. The projections of the tetragonal bulk highsymmetry points onto the [112] surface are indicated. (c) Top view on the [112] surface. The black and blue dots indicate projections of the bulk Γ points of the cubic and tetragonal BZ, respectively.

different lattice vectors (blue for cubic; black for tetragonal) and the resulting combination of both BZs with their relative sizes. This BZ phenomenon has been previously reported by Pettenkofer and Hofmann40 for CuInSe2, the sister compound of CuInTe2. In addition, orthorhombic structures such as the perovskite CaTiO3 have also been reported to display pseudocubic unit cells in experimental band structure 6835


Letter


Figure 3. ARPES data of CuInTe2. (a) Constant energy slices at Ei = 0 eV (EF) and Ei = −0.87 eV, integrated over 15.5 meV. At the Fermi surface, there is only weak intensity visible at the Γ points corresponding to the cubic bulk BZ. The lower initial state energy was chosen to visualize the point positions better. The evolution in energy around Γ is shown in panel b. (c) Band structure dispersions along ky = 0 Å−1 and kx = 0 Å−1 with Shirley background subtraction and in negative second derivative representation. Three bands can be identified, which are schematically superposed in the ARPES data and marked with arrows in the second derivative plots. (d) DFT calculations of the respective k directions of panel c. The three visible ARPES bands are matched by theory except for a necessary stretching of the theoretical energy axis. A further small splitting of the theoretical bands by SOC along N−Γ−N is too small in energy to be resolved by ARPES.

measurements.41 Figure S5 contains detailed information on the unit cell of CuInTe2, explaining the pseudocubic unit cell displayed by this tetragonal compound. Because the ARPES data feature only the cubic BZ, it is important to understand how the high-symmetry points of the different BZ’s relate to each other in order to compare the measured data with the calculated electronic structure, as the latter is based on the tetragonal BZ. Figure 2b shows different orthographic projections on the two BZs from panel a. The natural cleavage plane of CuInTe2 is the (112) plane, which is the pseudohexagonal plane, and thus, the packing density is maximized. As can be seen from Figure 2, cuts along kx therefore correspond to the Γ−N, while cuts along ky correspond to the Γ−Z direction. Note that the exact highsymmetry point correspondence of the k|| directions depends on the chosen photon energy. However, a photon energy of 80 eV in Figure 3 agrees well with the aforementioned directional determination. Having the directions in momentum space resolved, we now proceed to investigate the ARPES spectra. Figure 3 shows ARPES data gathered at hν = 80 eV. Figure 3a displays an overview over the BZ in the form of constant energy cuts at Ei = 0 eV (at the Fermi level) and Ei = −0.8 eV. At Ei = 0 eV, the Fermi surface is composed of only very weak intensity spots at the Γ̅ and Γ̅ ′ point, indicating that the valence band maximum is at Γ̅ , consistent with DFT. Lower initial state energies show a clover-shaped feature surrounding Γ̅ and Γ̅ ′, indicating that the valence bands disperse away from the Γ̅ point. The distance between Γ̅ and Γ̅ ′ is 1.66 Å−1, matching the expected distances within the cubic BZ projected along the (112) direction. Figure 3b shows several constant energy cuts, which are stacked to give insight into the band dispersion. Cuts along ky = 0 Å−1 (Γ−Z) and kx = 0 Å−1 (Γ−N), shown in Figure 3c, reveal that the bands are steeply dispersed in energy. The Shirley background subtracted and second derivative ARPES data both show a pair of bands meeting close to the Fermi level at Γ̅ . A third band has its maximum at roughly −0.8 eV and is responsible for the inner structure of Figure 3b at lower initial state energies. All identified bands are schematically superimposed on the ARPES data and marked with arrows in the

second derivative plots. Because the calculated band structure was plotted for the tetragonal BZ, the DFT data must be stretched by a factor of 2 in kx and ky, which was also included in the high-symmetry point positions superimposed on the ARPES data of Figure 3a,c. In comparison to the experimental band structure, the calculated bands in Figure 3d, furthermore, must be stretched by a factor of 1.9 in energy to match the measured dispersion. Note that this trend is opposite to the commonly found reduced bandwidth in experimentally determined band structures, which is a result of electron− electron correlations.42 Thus, DFT underestimated the slope of the bands. Measuring the slope below the Fermi level, we find a band velocity of 2.5−5.4 × 105 m/s, depending on which band is considered. This is about 25−50% of the value found in graphene, ∼1 × 106 m/s.43 The conclusion that the band velocity is the reason for the high carrier mobility measured suggests itself, even though the velocity of course decreases at close vicinity to the Fermi level because of the parabolic band shape. However, because photoabsorption processes excite electrons well below the Fermi level, the initial band velocity of the created holes can still be very high. A closer look at the measured band dispersions reveals a few more interesting details about the electronic structure of CuInTe2. The three bands identified along the Z−Γ−Z direction are not reproduced in the DFT calculations without SOC, which predicts only two bands along this cut (Figure S4). As can be seen in Figure 3d, if SOC is included in the calculation, the upper of the two bands along Z−Γ−Z splits into two, while the lower band is shifted to lower binding energies. This matches the expected splitting of p bands in the chalcopyrite structure.44 The shift is relatively large, about 0.8 eV, and is clearly resolved in the measurement, meaning that SOC strongly affects the electronic structure of CuInTe2. Along N−Γ−N, in contrast, all three visible bands are already predicted in the calculation excluding SOC (Figure S4). Because CuInTe2 lacks inversion symmetry, SOC has the potential to also spin split bulk bands (of the p or d type) according to the Rashba/Dresselhaus effect.44,45 This SOCinduced splitting, visible in Figure 3d, is much smaller and thus not resolvable in experiment. Nevertheless, the lowest band 6836


Letter


Figure 4. Determination of CuInTe2 band edge thermodynamics. (a) Representative M−S plot of CuInTe2 at 1, 3, and 5 kHz measured under dark conditions. (b) Representative chopped-light LSV of CuInTe2. Dark current is subtracted to give the differential current density. The bottom right inset shows LSV scans under dark and illuminated conditions. (c) Schematic diagram of the electronic band positions of CuInTe2 at pH 9. The valence band edge was approximated from the flat band potentials measured from M−S and photocurrent onset techniques. The conduction band edge was approximated based on the material’s optically measured band gap. The black dashed line represents the water reduction potential at pH 9.

on the photocurrent density at positive applied potentials. The band edge energy of the photoanode was not reported. To ensure our p-type single-crystal material is thermodynamically suitable for water reduction, its flat band potential (Efb) was experimentally determined. In triplicate, CuInTe2 photoelectrodes were subjected to AC impedance measurements for Mott−Schottky (M−S) analysis and chopped-light linear sweep voltammetry (LSV). Figure 4a shows a representative M−S plot for CuInTe2. M−S analysis was performed at pH 9 in 0.6 M Na2SO4(aq) to ensure a high double layer capacitance, thus allowing the space charge capacitance to dominate. The negative slopes of each curve confirm the p-type behavior of the electrodes. The M−S equation (discussed in the Supporting Information) used to fit the data assumes that a semiconductor−electrolyte interface is ideally capacitive. However, these interfaces are almost never perfectly capacitive, leading to frequency-dependent slopes.49 This class of frequency dependence gives plots with varying slopes that converge to a common intercept, giving a valid Efb.50 At high frequencies, such as the 5 kHz plot in Figure 4a, a slight nonlinearity starts to creep in. This nonlinearity at higher frequencies is a result of different regions of capacitance that exist in the sample because of the presence of surface states.51 The Efb determined by M−S analysis was −0.62 ± 0.041 V vs Ag/AgCl at pH 9. Figure 4b shows a representative chopped LSV plot with the inset showing the dark and illuminated LSV scans. The dark current was subtracted from the chopped-light curve to form a j−V curve emphasizing the photocurrent. The Efb is estimated by identifying the potential at which the photocurrent starts to grow from its baseline,

maximum is again shifted to lower binding energies when considering SOC, matching the ARPES data. Therefore, we can unambiguously conclude that SOC influences the electronic structure of CuInTe2 and must be included in the discussion. It has previously been suggested that the resulting spin splitting is advantageous for light-harvesting materials.46 Thus, the observed SOC effect could also contribute to good performance of CuInTe2. A more in-depth discussion of the phenomena is given in the Supporting Information. Having resolved the valence band structure of CuInTe2, we can now compare the results to other light-harvesting materials such as CH3NH3PbI3 or BiVO4. While, with the exception of a low-resolution ARPES experiment on CH3NH3PbBr3,46 the electronic structures of these materials have not been experimentally determined (likely because of their large band gap sizes), band structure calculations have been performed instead.47 One could predict the slow charge transport kinetics of BiVO4, given the flat slopes seen in its band structure. Similarly, the high performance of CH3NH3PbI3 has been attributed to its calculated steep bands.18,19 However, as we have shown here, band velocity is another variable (such as band gap size) that DFT frequently fails to accurately determine. Hence, ARPES provides important insight into the potential of (small-gap) semiconductors as photoactive materials. Thermodynamic Characterization of CuInTe2 Photocathodes. To the best of our knowledge, CuInTe2 has not been studied for photoelectrocatalytic water reduction. Until this work, the only reported photoelectrochemical characterization of CuInTe2 was of n-type nanoplates by Zhang et al.48 reporting 6837



■

namely, the photocurrent onset potential (Vonset). The Efb determined by Vonset was −0.58 ± 0.073 V vs Ag/AgCl at pH 9. The Efb determined by each technique is well within the experimental error, enabling a confident approximation of the valence band edge with respect to Ag/AgCl at pH 9. Based on this data, Figure 4c shows a schematic diagram of the electronic band positions of CuInTe2. The conduction band energy was approximated by applying the optically measured band gap, shown in Figure 1b. The dark blue coloring represents the filled states of the valence band, while the light blue coloring represents the empty states of the conduction band. The diagram also includes a dotted line representing the potential versus Ag/AgCl for H2 evolution at pH 9. The valence and conduction band edges of CuInTe2 straddle this potential, confirming this material is indeed thermodynamically capable of photoelectrochemical water reduction. Indeed, hydrogen evolution using CuInTe2 photoelectrodes was confirmed by GC analysis. Comprehensive photoelectrochemical studies on these single-crystal photocathodes are currently underway. In conclusion, the merit of a semiconductor for lightharvesting applications depends heavily on its electronic structure. Thermodynamic factors such as band edge positions and band gap sizes are commonly investigated in this respect. Here, we showed that the momentum-resolved electronic structure is another important aspect of the electronic structure to consider. We showed with ARPES that the band velocity of the valence band in CuInTe2 can reach values of up to 50% of the value found in graphene, a material with extremely high mobility. Because of the parabolic (rather than linear) band dispersion, this value decreases in the vicinity to the Fermi level. Because photoabsorption processes excite electrons well below the Fermi level, this region of steep band dispersion can be accessed in the light-harvesting process. We further show evidence that CuInTe2 has energetically suitable band edges for photoelectrocatalytic water reduction, enforcing the argument that it possesses a favorable electronic band structure to efficiently harness solar energy for H2 fuel production.

Letter

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.8b03100. Details on synthetic procedure, crystal defect chemistry, crystal structure, spectroscopy techniques, transport property measurements, theoretical electronic band structure, and flat-band measurement techniques (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Maxim Krivenkov: 0000-0002-9889-9047 Andrew B. Bocarsly: 0000-0003-3718-0933 Leslie M. Schoop: 0000-0003-3459-4241 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank Prof. Robert Cava for instrumentation used in material synthesis and transport property characterization. The authors thank Nan Yao, John Schrieber, and Yao-Wen Yeh for invaluable assistance with X-ray photoelectron spectroscopy. We thank HZB for the allocation of synchrotron radiation beamtime. This manuscript is based upon work supported by the National Science Foundation Graduate Research Fellowship via Grant 1656466 awarded to J.J.F. A.T. was supported by the DFG proposal no. SCHO 1730/1-1. The authors gratefully acknowledge the U.S. Department of Energy for financial support. A.B.B. acknowledges Grant DE-SC0002133. L.M.S. acknowledges funding from the Princeton Catalysis Initiative (PCI).

■

■

REFERENCES

(1) Hisatomi, T.; Kubota, J.; Domen, K. Recent Advances in Semiconductors for Photocatalytic and Photoelectrochemical Water Splitting. Chem. Soc. Rev. 2014, 43, 7520−7535. (2) Götz, M.; Lefebvre, J.; Mörs, F.; McDaniel Koch, A.; Graf, F.; Bajohr, S.; Reimert, R.; Kolb, T. Renewable Power-to-Gas: A Technological and Economic Review. Renewable Energy 2016, 85, 1371−1390. (3) McKone, J. R.; Lewis, N. S.; Gray, H. B. Will Solar-Driven Water-Splitting Devices See the Light of Day? Chem. Mater. 2014, 26, 407−414. (4) Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Solar Water Splitting Cells. Chem. Rev. 2010, 110, 6446−6473. (5) Matthews, P. D.; McNaughter, P. D.; Lewis, D. J.; O’Brien, P. Shining a Light on Transition Metal Chalcogenides for Sustainable Photovoltaics. Chem. Sci. 2017, 8, 4177−4187. (6) Kang, D.; Kim, T. W.; Kubota, S. R.; Cardiel, A. C.; Cha, H. G.; Choi, K.-S. Electrochemical Synthesis of Photoelectrodes and Catalysts for Use in Solar Water Splitting. Chem. Rev. 2015, 115, 12839−12887. (7) Shin, D.; Ngaboyamahina, E.; Zhou, Y.; Glass, J. T.; Mitzi, D. B. Synthesis and Characterization of an Earth-Abundant Cu2BaSn(S,Se)4 Chalcogenide for Photoelectrochemical Cell Application. J. Phys. Chem. Lett. 2016, 7, 4554−4561. (8) Tachibana, Y.; Vayssieres, L.; Durrant, J. R. Artificial Photosynthesis for Solar Water-Splitting. Nat. Photonics 2012, 6, 511−518.

EXPERIMENTAL METHODS CuInTe2 single crystals were grown using the vertical Bridgman technique (for details, see the Supporting Information). The crystals were analyzed by powder X-ray diffraction on a Bruker D8 Advance Eco with Cu Kα radiation and a LynxEye-XE detector. For single-crystal X-ray diffraction, a Bruker APEX II CCD with Mo Kα radiation was used (for details, see the Supporting Information). The optical band gap was measured with the Agilent Technologies Cary 5000 UV− vis−NIR spectrometer. More details about the sample characterization can be found in the Supporting Information. Transport measurements were performed with a Quantum Design Physical Property Measurement System, and ARPES spectra were recorded with the 12 experiment on the UE112PGM2a beamline at the BESSY II synchrotron facility. Experimental details can be found in the Supporting Information. DFT calculations were performed within the framework of Wien2k52 using the modified Becke−Johnson functional.37 Details on CuInTe2 electrode assembly and band edge characterization techniques can be found in the Supporting Information. 6838


Letter


(29) Luo, Y.; Yang, J.; Jiang, Q.; Li, W.; Zhang, D.; Zhou, Z.; Cheng, Y.; Ren, Y.; He, X. Progressive Regulation of Electrical and Thermal Transport Properties to High-Performance CuInTe2 Thermoelectric Materials. Adv. Energy Mater. 2016, 6, 1600007. (30) Reshak, A. H.; Auluck, S. Electronic Structure, Linear, Nonlinear Optical Susceptibilities and Birefringence of CuInX2(X = S, Se, Te) Chalcopyrite-Structure Compounds. PMC Phys. B 2008, 1, 12. (31) Rashkeev, S. N.; Lambrecht, W. R. L. Second-Harmonic Generation of I-III-VI2 Chalcopyrite Semiconductors: Effects of Chemical Substitutions. Phys. Rev. B: Condens. Matter Mater. Phys. 2001, 63, 165212. (32) Thwaites, M. J.; Tomlinson, R. D.; Hampshire, M. J. The Observation of a Direct Energy Band Gap for CuInTe2 Single Crystals Using Electroreflectance Techniques. Solid State Commun. 1977, 23, 905−906. (33) Sprinkle, M.; Siegel, D.; Hu, Y.; Hicks, J.; Tejeda, A.; TalebIbrahimi, A.; Le Fèvre, P.; Bertran, F.; Vizzini, S.; Enriquez, H.; et al. First Direct Observation of a Nearly Ideal Graphene Band Structure. Phys. Rev. Lett. 2009, 103, 226803. (34) Dong, Q.; Fang, Y.; Shao, Y.; Mulligan, P.; Qiu, J.; Cao, L.; Huang, J. Electron-Hole Diffusion Lengths > 175 μm in SolutionGrown CH3NH3PbI3 Single Crystals. Science 2015, 347, 967−970. (35) Rettie, A. J. E.; Lee, H. C.; Marshall, L. G.; Lin, J.-F.; Capan, C.; Lindemuth, J.; McCloy, J. S.; Zhou, J.; Bard, A. J.; Mullins, C. B. Combined Charge Carrier Transport and Photoelectrochemical Characterization of BiVO4 Single Crystals: Intrinsic Behavior of a Complex Metal Oxide. J. Am. Chem. Soc. 2013, 135, 11389−11396. (36) Filippetti, A.; Mattoni, A.; Caddeo, C.; Saba, M. I.; Delugas, P. Low Electron-Polar Optical Phonon Scattering as a Fundamental Aspect of Carrier Mobility in Methylammonium Lead Halide CH3NH3PbI3 Perovskites. Phys. Chem. Chem. Phys. 2016, 18, 15352−15362. (37) Tran, F.; Blaha, P. Accurate Band Gaps of Semiconductors and Insulators with a Semilocal Exchange-Correlation Potential. Phys. Rev. Lett. 2009, 102, 226401. (38) Sridevi, D.; Reddy, K. V. Electrical Conductivity and Optical Absorption in Flash-Evaporated CuInTe2 Thin Films. Thin Solid Films 1986, 141, 157−164. (39) Neumann, H.; Perlt, B.; Hörig, W.; Kühn, G. The Optical Properties of CuInTe2 Epitaxial Layers. Thin Solid Films 1989, 182, 115−120. (40) Hofmann, A.; Pettenkofer, C. Electronic Band Structure of Epitaxial CuInSe2 Films. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 84, 115109. (41) Muff, S.; Fanciulli, M.; Weber, A. P.; Pilet, N.; Ristić, Z.; Wang, Z.; Plumb, N. C.; Radović, M.; Dil, J. H. Observation of a TwoDimensional Electron Gas at CaTiO3 Film Surfaces. Appl. Surf. Sci. 2018, 432, 41−45. (42) Nayak, J.; Filsinger, K.; Fecher, G. H.; Chadov, S.; Minár, J.; Rienks, E. D. L.; Büchner, B.; Parkin, S. P.; Fink, J.; Felser, C. Observation of a Remarkable Reduction of Correlation Effects in BaCr2As2 by ARPES. Proc. Natl. Acad. Sci. U. S. A. 2017, 114, 12425− 12429. (43) Gierz, I.; Riedl, C.; Starke, U.; Ast, C. R.; Kern, K. Atomic Hole Doping of Graphene. Nano Lett. 2008, 8, 4603−4607. (44) Thwaites, M. J.; Tomlinson, R. D.; Hampshire, M. J. The Observation of Optical Transitions from Valence Band States in CuInTe2. Phys. Status Solidi B 1979, 94, 211−214. (45) Dresselhaus, G. Spin-Orbit Coupling Effects in Zinc Blende Structures. Phys. Rev. 1955, 100, 580−586. (46) Niesner, D.; Wilhelm, M.; Levchuk, I.; Osvet, A.; Shrestha, S.; Batentschuk, M.; Brabec, C.; Fauster, T. Giant Rashba Splitting in CH3NH3PbBr3 Organic-Inorganic Perovskite. Phys. Rev. Lett. 2016, 117, 126401. (47) Zhao, Z.; Li, Z.; Zou, Z. Electronic Structure and Optical Properties of Monoclinic Clinobisvanite BiVO4. Phys. Chem. Chem. Phys. 2011, 13, 4746−4753.

(9) Maeda, K. Photocatalytic Water Splitting Using Semiconductor Particles: History and Recent Developments. J. Photochem. Photobiol., C 2011, 12, 237−268. (10) Iwashina, K.; Iwase, A.; Ng, Y. H.; Amal, R.; Kudo, A. ZSchematic Water Splitting into H2 and O2 Using Metal Sulfide as a Hydrogen-Evolving Photocatalyst and Reduced Graphene Oxide as a Solid-State Electron Mediator. J. Am. Chem. Soc. 2015, 137, 604−607. (11) Iwase, A.; Ng, Y. H.; Amal, R.; Kudo, A. Solar Hydrogen Evolution Using a CuGaS2 Photocathode Improved by Incorporating Reduced Graphene Oxide. J. Mater. Chem. A 2015, 3, 8566−8570. (12) Yokoyama, D.; Minegishi, T.; Maeda, K.; Katayama, M.; Kubota, J.; Yamada, A.; Konagai, M.; Domen, K. Photoelectrochemical Water Splitting Using a Cu(In,Ga)Se2 Thin Film. Electrochem. Commun. 2010, 12, 851−853. (13) Frick, J. J.; Kushwaha, S. K.; Cava, R. J.; Bocarsly, A. B. Characterization of Primary Carrier Transport Properties of the LightHarvesting Chalcopyrite Semiconductors CuIn(S1‑xSex)2. J. Phys. Chem. C 2017, 121, 17046−17052. (14) Frick, J. J.; Cava, R. J.; Bocarsly, A. B. Chalcopyrite CuIn(S1‑xSex)2 for Photoelectrocatalytic H2 Evolution: Unraveling the Energetics and Complex Kinetics of Photogenerated Charge Transfer in the Semiconductor Bulk. Chem. Mater. 2018, 30, 4422− 4431. (15) Ni, D.; Kuo, H.-Y.; Park, J. E.; Lee, T. S.; Sloman, S.-R. I.; Cava, R. J.; Bocarsly, A. B. Improved H2 Evolution in Quaternary SCIGS Chalcopyrite Semiconductors. J. Phys. Chem. C 2018, 122, 24512− 24519. (16) Sivula, K.; van de Krol, R. Semiconducting Materials for Photoelectrochemical Energy Conversion. Nat. Rev. Mater. 2016, 1, 15010. (17) Sivula, K. Metal Oxide Photoelectrodes for Solar Fuel Production, Surface Traps, and Catalysis. J. Phys. Chem. Lett. 2013, 4, 1624−1633. (18) Umari, P.; Mosconi, E.; De Angelis, F. Relativistic GW Calculations on CH3NH3PbI3 and CH3NH3SnI3 Perovskites for Solar Cell Applications. Sci. Rep. 2015, 4, 4467. (19) Connor, B. A.; Leppert, L.; Smith, M. D.; Neaton, J. B.; Karunadasa, H. I. Layered Halide Double Perovskites: Dimensional Reduction of Cs2AgBiBr6. J. Am. Chem. Soc. 2018, 140, 5235−5240. (20) Frost, J. M.; Butler, K. T.; Brivio, F.; Hendon, C. H.; van Schilfgaarde, M.; Walsh, A. Atomistic Origins of High-Performance in Hybrid Halide Perovskite Solar Cells. Nano Lett. 2014, 14, 2584− 2590. (21) Ashcroft, N. W.; Mermin, N. D. Solid State Physics; Saunders College: Philadelphia, PA, 1976. (22) Hisatomi, T.; Takanabe, K.; Domen, K. Photocatalytic WaterSplitting Reaction from Catalytic and Kinetic Perspectives. Catal. Lett. 2015, 145, 95−108. (23) Liang, Y.; Tsubota, T.; Mooij, L. P. A.; van de Krol, R. Highly Improved Quantum Efficiencies for Thin Film BiVO4 Photoanodes. J. Phys. Chem. C 2011, 115, 17594−17598. (24) Abdi, F. F.; van de Krol, R. Nature and Light Dependence of Bulk Recombination in Co-Pi-Catalyzed BiVO4 Photoanodes. J. Phys. Chem. C 2012, 116, 9398−9404. (25) Abdi, F. F.; Savenije, T. J.; May, M. M.; Dam, B.; van de Krol, R. The Origin of Slow Carrier Transport in BiVO4 Thin Film Photoanodes: A Time-Resolved Microwave Conductivity Study. J. Phys. Chem. Lett. 2013, 4, 2752−2757. (26) Kim, T. W.; Ping, Y.; Galli, G. A.; Choi, K.-S. Simultaneous Enhancements in Photon Absorption and Charge Transport of Bismuth Vanadate Photoanodes for Solar Water Splitting. Nat. Commun. 2015, 6, 8769. (27) Cooper, J. K.; Gul, S.; Toma, F. M.; Chen, L.; Liu, Y.-S.; Guo, J.; Ager, J. W.; Yano, J.; Sharp, I. D. Indirect Bandgap and Optical Properties of Monoclinic Bismuth Vanadate. J. Phys. Chem. C 2015, 119, 2969−2974. (28) Zhou, G.; Wang, D. High Thermoelectric Performance from Optimization of Hole-Doped CuInTe2. Phys. Chem. Chem. Phys. 2016, 18, 5925−5931. 6839


Letter

The Journal of Physical Chemistry Letters (48) Zhang, X.; Yang, L.; Guo, Z.; Su, G.; Gao, R.; Wang, W.; Dong, B.; Cao, L. Rapid Synthesis of CuInTe2 Ultrathin Nanoplates with Enhanced Photoelectrochemical Properties. Chem. Commun. 2017, 53, 5878−5881. (49) Lasia, A. Electrochemical Impedance Spectroscopy and Its Applications; Springer-Verlag: New York, 2014. (50) Govindaraju, G. V.; Wheeler, G. P.; Lee, D.; Choi, K.-S. Methods for Electrochemical Synthesis and Photoelectrochemical Characterization for Photoelectrodes. Chem. Mater. 2017, 29, 355− 370. (51) Bisquert, J. Nanostructured Energy Devices: Equilibrium Concepts and Kinetics; CRC Press, 2014. (52) K. S. Blaha, P.; Schwarz, K.; Madsen, G.; Kvasnicka, D.; Luitz, J. WIEN2k: An Augmented Plane Wave plus Local Orbitals Program for Calculating Crystal Properties. Technische Universität Wien: Wien, 2001; p 28.

6840


Single-Crystal Growth and Characterization of the Chalcopyrite

Recommend Documents