Transport Evidence for Sulfur Vacancies as the Origin of

Importantly, the certificate of analysis for the standard FeS2 precursor ..... the behavior of the mobility with doping also provides evidence of comp...
0 downloads 0 Views 2MB Size
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Transport Evidence for Sulfur Vacancies as the Origin of Unintentional n‑Type Doping in Pyrite FeS2 Bryan Voigt,† William Moore,† Michael Manno,† Jeff Walter,†,‡ Jeff D. Jeremiason,§ Eray S. Aydil,*,†,∥ and Chris Leighton*,†

ACS Appl. Mater. Interfaces Downloaded from pubs.acs.org by ROBERT GORDON UNIV on 04/22/19. For personal use only.



Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota 55455, United States ‡ Department of Physics and Astronomy, Carleton College, Northfield, Minnesota 55057, United States § Department of Chemistry, Gustavus Adolphus College, Saint Peter, Minnesota 56082, United States ∥ Department of Chemical and Biomolecular Engineering, New York University Tandon School of Engineering, Brooklyn, New York 11201, United States S Supporting Information *

ABSTRACT: Pyrite FeS2 has long been considered a potential earth-abundant low-cost photovoltaic material for thin-film solar cells but has been plagued by low power conversion efficiencies and open-circuit voltages. Recent efforts have identified a lack of understanding and control of doping, as well as uncontrolled surface conduction, as key roadblocks to the development of pyrite photovoltaics. In particular, while n-type bulk behavior in unintentionally doped single crystals and thin films is speculated to arise from sulfur vacancies (VS), proof remains elusive. Here, we provide strong evidence, from extensive electronic transport measurements on high-quality crystals, that VS are deep donors in bulk pyrite. Otherwise identical crystals grown via chemical vapor transport under varied S vapor pressures are thoroughly characterized structurally and chemically, and shown to exhibit systematically different electronic transport. Decreased S vapor pressure during growth leads to reduced bulk resistivity, increased bulk Hall electron density, reduced transport activation energy, onset of positive temperature coefficient of resistivity, and approach to an insulator−metal transition, all as would be expected from increased VS donor density. Impurity analyses show that these trends are uncorrelated with metal impurity concentration and that extracted donor densities significantly exceed total impurity concentrations, directly evidencing a native defect. Well-controlled, wide-range n-doping of pyrite is thus achieved via the control of VS concentration, with substantial implications for photovoltaic and other applications. The location of the VS state within the gap, the influence of specific impurities, unusual aspects to the insulator−metal transition, and the influence of doping on surface conduction are also discussed. KEYWORDS: solar cells, photovoltaic absorbers, crystal growth, semiconductors, doping, defects, electronic transport, sulfur vacancies, insulator−metal transition theoretical maximum.1 These outcomes limited the interest in pyrite PV in the 1990s. Around 2009, interest in FeS2 grew, however, motivated by sustainability concerns with commercial thin-film solar cell absorbers such as CdTe and Cu(In,Ga)Se2.8 Importantly, several efforts in this second wave of interest in pyrite have focused on fundamental reasons for poor PV performance.9−15 Three main issues have been identified: the possibility of deleterious secondary phases,11,12 surface conduction phenomena, 9,15 and a lack of understanding and control of doping.13,14,16 Most recently, however, concerns over secondary phases such as pyrrhotite Fe1−xS and the FeS2 polymorph marcasite have abated due to syntheses resulting in rigorously phase-pure single crystals,9,10,14,15,17 thin films,13,14,17−20 and

1. INTRODUCTION Pyrite structure FeS2 is a potentially ideal photovoltaic (PV) material for sustainable, low-cost, large-scale solar-to-electric energy conversion as it is composed of earth-abundant, nontoxic, inexpensive elements, has a band gap near the Shockley−Queisser maximum1 (0.95 eV2), and absorbs visible light so well that a 100 nm thick film absorbs >90% of sunlight.2 These properties, combined with a room-temperature electron mobility3 of up to 360 cm2 V−1 s−1 and minority carrier diffusion lengths (100−1000 nm) exceeding the thickness for complete light absorption, 2 led to the identification of FeS2 in the 1980s as a promising material for thin-film solar cells. Attempts to employ FeS2 as the absorber in Schottky,2,4 photoelectrochemical,2,5,6 and p−i−ntype7 solar cells all failed, however, due to low open-circuit voltages (VOC < 0.2 V) and power conversion efficiencies ( 430 °C.62 Extrapolation to T ≤ 430 °C predicts the formation of 102−104 nm thick Fe1−xS surface layers in 24 h, 3 orders of magnitude larger than the S diffusion length (Supporting Information, Figure S1). We have verified this, observing ∼104 nm thick Fe1−xS skins on the surface of FeS2 crystals after vacuum annealing at 380 °C. While VS may accelerate S diffusion, the 6-order-of-magnitude increase in D required to compete with decomposition (Supporting Information Section A) is unlikely. Postgrowth annealing in reducing atmosphere is therefore not a promising route to control VS in bulk FeS2 crystals. In this work, we thus pursue the second approach, i.e., controlling S vapor pressure during growth (and cooling), to incorporate variable concentrations of VS. In this approach, slow S diffusion becomes an advantage, allowing for kinetic trapping of VS during growth. 2.2. Materials and Methods. Unintentionally doped pyrite FeS2 single crystals were grown via CVT.63 Cylindrical quartz ampoules (50 cm3) were first cleaned by sonication in deionized water, acetone, and methanol, and dried at 400 °C. Ampoules were then loaded with 100 mg of FeBr2 transport agent (Sigma-Aldrich, 99.999% purity) and 2.2 g of pyrite FeS2 precursor powder, evacuated to ∼10−6 Torr and flame-sealed. Both commercial (Alfa Aesar, 99.9% purity) and labsynthesized FeS2 precursor powders were used, hereafter referred to as “standard” and “high-purity”, respectively. High-purity FeS2 was synthesized by reacting S (Alfa Aesar, 99.9995% purity) and Fe (Alfa Aesar, 99.998% purity) in evacuated ampoules (6:1 S/Fe molar ratio) for 6 days at 500°C. The resulting powder was ground and repeatedly sulfidized until phase-pure pyrite FeS2 was obtained, verified by powder X-ray diffraction (PXRD). Key for this study, the S vapor pressure during CVT crystal growth, was controlled via the S/Fe molar ratio loaded in the ampoules, hereafter referred to as the “S:Fe loading”. We emphasize that this is the molar ratio of precursors in the CVT growth, not the S/Fe ratio of resultant crystals. To increase S:Fe loading above 2.0, excess S powder (Alfa Aesar, 99.9995% purity) was added. To decrease S:Fe loading below 2.0, high-purity FeS2 precursor was partially decomposed in N2 at 600 °C to form controlled fractions of pyrrhotite Fe7S8. The molar % of each phase was estimated from PXRD (using pyrite64 and pyrrhotite65 reference patterns). Sealed ampoules were then placed in a two-zone furnace with precursors at one end (the source zone). During growth, the source and growth zones were maintained at 670 and 590 °C, respectively. An inverted temperature configuration was used for the first 3 days, to “clean” the growth zone,66 followed by 17 days of growth. The average growth rate was determined by normalizing the average masses of the 10 largest crystals from each ampoule by the growth duration. Based on the above conditions, S vapor pressures during growth can be estimated to be 0.003 and 3.5 atm at low (2.3 results in ND ≈ 1019−1021 cm−3, exceeding the Co concentration by at least 2 orders of magnitude. This rules out explanations for the observed trends with S:Fe loading in terms of varying Co concentrations. The N D of 10 19 −10 21 cm −3 even exceeds the total MBI concentration (by at least 1 order of magnitude), ruling out any observed impurity as the primary donor in these crystals. In addition, as noted in Section 3.1, the fact that n increases with decreasing S:Fe loading in Figure 5b is qualitatively at odds with any simple explanation based on total MBI concentration. The latter decreases with decreasing S:Fe loading (likely due to increased growth rate), in contrast with the observation of heavier doping at lower S:Fe loadings.

(S:Fe loading of 1.69, left panel, Figure 4a−c, and S:Fe loading of 2.37, right panel, Figure 4d−f). As in Figure 3, the surfacedominated regime is shaded. Figure 4a,d first plots the resistivity (log scale), showing the sign change in hightemperature dρ/dT with S:Fe loading, as expected from Figure 3a. Figure 4b,e then shows the 1/T dependence of the Hall electron density, n, plotted on a log scale. As expected from the Introduction, all Hall signals in the bulk conduction regime in the crystals studied here are negative, indicating unintentional n-doping. Note the n(300 K) value of ∼2 × 1015 cm−3 at S:Fe loading of 2.37 (Figure 4e), increasing by 2 orders of magnitude to ∼2 × 1017 cm−3 at S:Fe loading of 1.69 (Figure 4b), again consistent with increased VS doping. In terms of quantitative analysis of n(T), under the assumptions of nondegenerate statistics and negligible compensation, we expect n(T ) =

NCND −ΔE /2kBT e 2

(1)

where NC is the effective density of states in the conduction band, ND is the donor density, ΔE is the donor activation energy, and kB is Boltzmann’s constant.73 In many semiconductors, however, particularly compounds, significant compensation is likely from impurities and/or nonstoichiometry-related defects. Under such conditions n(T ) =

NC(ND − NA ) −ΔE / kBT e NA

(2)

where NA is the (compensating) acceptor density.73 Significantly, there is a factor of 2 difference in the exponents in eqs 1 and 2, generating a twofold difference in extracted ΔE values. Consistent with recent work,9 we assume the compensated case and thus use eq 2; ΔE values below would thus be twice as large if negligible compensation were assumed. In support of this assumption, note that high-purity crystals at high S:Fe loading have ∼1 ppm Co (Table 1) and that Co is a known shallow donor in pyrite.35−38 Full ionization would then produce an expected electron density of ∼8 × 1016 cm−3 due to Co alone, well above the lowest measured n(300 K) ≈ 3 × 1015 cm−3, suggesting that compensation must be present. As discussed below, the behavior of the mobility with doping also provides evidence of compensation. Fe vacancies, among other defects, are one possible source. As illustrated by the solid line fits to eq 2 (Figure 4b,e), simple activated behavior is indeed observed at high T in both S:Fe loading limits, yielding ΔE values of 17 and 220 meV, respectively. The donor activation energy thus decreases approximately 10-fold on decreasing S:Fe loading from 2.37 to 1.69, again consistent with VS acting as donors in pyrite. A potential break to a lower activation energy process at lower T occurs in both cases (Figure 4b,e), but further analysis is hindered by proximity to the crossover to surface conduction at 200−300 K. We note that the extraction of ND is in principle also possible from n(T) (using eq 2 with free electron estimates of NC from me*) but is complicated by the lack of knowledge of NA, as well as possibly degenerate transport at low S:Fe loadings. Analysis below thus focuses on extracted ΔE values. Figure 4c,f shows the Hall mobility, μ(T), computed from ρ(T) and n(T). Both S:Fe loadings result in qualitatively typical behavior for moderately doped semiconductors, μ first increasing on cooling from 400 K (due to phonon scattering), reaching a maximum value (μmax) at some T (Tmax) and then G

DOI: 10.1021/acsami.9b01335 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 5. Summary of electronic transport properties of pyrite FeS2 single crystals vs S:Fe loading during growth. Crystals from high-purity FeS2 precursor powder are on the left (panels (a)−(d), green points), with crystals from standard FeS2 powder on the right (panels (e)−(h), red points). In all cases, the data points result from the averaging of at least two to three results from each growth; error bars represent standard deviations. Shown are the S:Fe loading dependence of (a, e) 300 K absolute value of the temperature (T) derivative of resistivity (ρ) (open circles denote positive (metallic-like) dρ/dT, closed circles negative (insulating-like) dρ/dT); (b, f) 300 K Hall electron density (n); (c, g) activation energy (ΔE) from Arrhenius analysis of n(T); and (d, h) peak Hall mobility (μ) in the ∼175−400 K range. Dashed lines are guides to the eye.

We interpret these data in terms of a deep donor state due to VS, located ∼225 meV below the conduction band minimum, as in Figure S5a (Supporting Information Section D). In the low VS density limit, at high S:Fe loading, where interdonor interactions are negligible, ΔE is thus 225 meV. As the VS concentration increases when S:Fe loading is decreased, the individual donor states broaden into a donor band, increasing the Fermi energy, decreasing ΔE, and approaching the ΔE → 0 limit expected at the IMT (where dρ/dT becomes positive). The nature of this IMT is discussed further in Section 3.3. A similar trend to Figure 5c is found for standard purity crystals (Figure 5g), albeit with a high S:Fe loading ΔE of

We thus contend that increased VS density at reduced S:Fe loading is the only simple interpretation of these data. As shown in Figure 5f, a similar trend in n(300 K) is observed in standard purity crystals but with larger overall values and lower dynamic range, due to additional impurity doping. Figure 5c,g shows the S:Fe loading dependence of the activation energy, ΔE, from n(T). In high-purity crystals (Figure 5c), ΔE saturates at ∼225 meV at high S:Fe loading, in good agreement with recent reports on high-quality unintentionally doped FeS2 crystals.9,10 As the S:Fe loading is decreased below ∼2.4, however, ΔE decreases by over an order of magnitude, reaching 17 meV at a S:Fe loading of 1.69. H

DOI: 10.1021/acsami.9b01335 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces ∼100 meV, significantly lower than high-purity crystals. As discussed in the Supporting Information (Section E), there are several possible explanations for this. One is that the additional MBI n-doping in standard purity crystals shifts the Fermi level during growth, modifying the enthalpy of formation of charged VS and thus the VS concentration at a given S vapor pressure. This would alter the correspondence between S:Fe loading and VS concentration, making direct comparisons between Figure 5c,g invalid. Another possibility is that the MBIs modify the energy of the VS states. Although this would require a ∼100 meV shift (from ∼225 to ∼100 meV) with an MBI concentration of only ∼500 ppm, this may be possible through the formation of defect complexes. Alternatively, Ni, the highest contributor to the MBI concentration in standard purity crystals, and the most likely to be an n-dopant (by balancing Ni•• Fe formation with electron doping), could also be a deep donor in pyrite, generating states ∼100 meV beneath the conduction band minimum (see the Supporting Information Section D, Figure S5b for a schematic). This could result in the saturation of ΔE at ∼100 meV in Figure 5g, as discussed in the Supporting Information Section E, including Figure S6. Prior work on Ni substitution in FeS2 is inconclusive, however, featuring ρ(T) or n(T) behavior that suggests increasing ΔE with Ni concentration, counter to expectations.36,75 While further work examining the impact of Ni in high-quality FeS2 single crystals is clearly worthwhile, this uncertainty has no impact on the main conclusions here: VS are deep donors in pyrite, and their concentration can be systematically varied to tune n-doping over a wide range. Figure 5d,h shows that in both purities of crystals, μmax increases (up to ∼1000 cm2 V−1 s−1) as the S:Fe loading is decreased. As noted in the discussion of Figure 4, this is interpreted in terms of increased n-doping due to VS at low S:Fe loading, leading to the improved screening of scattering from a background of charged impurities (including compensating acceptors), leading to decreased ionized impurity scattering and improved mobility. Finally, we comment briefly on the saturation of essentially all transport parameters plotted in Figure 5 in the high S:Fe loading limit. The simplest interpretation of such data is that VS concentrations have been suppressed to the lowest achievable levels in our CVT growth. As returned to below (Section 4), defect management to yet further suppress doping in this regime would ultimately be desirable in pyrite. 3.3. Electronic Transport: Nature of the Insulator− Metal Transition. The observation of phonon-limited bulk ρ(T) and an approach to an IMT in heavily VS-doped FeS2 (e.g., Figure 3), albeit obscured from full view by surface conduction at low T, is worthy of further comment. IMTs in semiconductors doped with shallow donors/acceptors, such as Si:P, Si:B, Ge:As, or GaAs:Si, have been studied for decades.76,77 The IMT in semiconductors doped with deep donors/acceptors, however, is more challenging, both experimentally and theoretically. The spatial extent of donor wave functions, ad, can be simply approximated as ad ≈ ℏ/ (2me*Ed)1/2,77 where me* is the electron mass (0.45− 0.49me)49,78 and Ed is the donor energy. Energies up to 225 meV, as observed here, result in very low ad ≈ 6 Å, far smaller ̈ application of than typical Bohr radii of shallow donors. Naive the Mott criterion for the critical density for the IMT, nc ≈ (0.2/ad)3,79 then yields high nc ≈ 1019−1020 cm−3. Until relatively recently, such deep-donor-induced semiconductor IMTs had, in fact, not been achieved, as these concentrations

often exceed equilibrium solubility limits.80,81 Progress with nonequilibrium processing using techniques like ion implantation, pulsed-laser melting, and rapid thermal treatment has changed this, however, realizing “hyperdoped” semiconductors.81−84 Si, for example, can be doped beyond equilibrium concentrations with deep (e.g., 250−480 meV) double donors S and Se, inducing an IMT.81,84 The IMT is understood to occur when interdonor interactions broaden the donor band to the point of overlap with the conduction band, pushing the Fermi energy above a mobility edge, as can be envisioned for FeS2 with Figure S5a (Supporting Information). In this context, results presented in this work potentially take on broader interest. If an IMT in pyrite can indeed be induced via doping with 225 meV deep donors due to VS, this would constitute a unique case where such a transition is achieved with a native defect, incorporated during growth (as opposed to postgrowth nonequilibrium processing). This could potentially provide facile access to an important materials physics issue that is otherwise difficult to study. Detailed investigation of the IMT requires low T transport measurements, however, which could only be achieved in pyrite FeS2 through the successful mitigation of surface conduction or via some means to make direct contact with the bulk. This stands as a challenge in the field.

4. CONCLUSIONS The above results provide compelling evidence that VS are deep donors in pyrite FeS2, that they are responsible for unintentional doping of single crystals, and that they enable wide-range control over n-doping. As noted in the Introduction, however, theoretical studies have concluded (i) that VS may not be effective n-dopants49,50 and (ii) that they may have formation energies large enough to preclude equilibrium concentrations of ∼1020 cm−3 at the current growth temperatures.11,16,49,50 With respect to (i), we emphasize that further DFT work on VS in FeS2 is clearly warranted. Some of the current authors are in fact engaged in such a study,85 considering not only isolated VS but also VS clusters, which are possible in pyrite46,86 and could form donor states. Other VS-containing defect complexes also warrant further study, as the results presented here cannot discount VS complexes vs isolated VS. This is also important with respect to (ii), as VS clusters/complexes could have lower formation enthalpies than isolated VS. Incorporation of VS during CVT growth of pyrite may also be controlled by a surface formation enthalpy lower than bulk46,50,51 (ΔHVS below ∼1 eV is required for the concentrations observed in this work) and could be a nonequilibrium incorporation process due to slow diffusion.61 Such discussion highlights another important question, which is what the results presented here imply for unintentionally doped pyrite films, which can have much larger electron densities than single crystals. Additional work to establish whether thin-film doping can also be explained by VS, as suggested by the universal μ−n relation in crystals and films, is clearly needed. Ultimately, the control of n-doping in the manner shown here is also needed in thin films, and this work lays a foundation for such an effort. Another important point is that even in crystals, additional work to further reduce the density of VS and other defects in FeS2 is desirable, in order to obtain lower electron densities. The n values in this work, for example, still significantly exceed the intrinsic carrier density. I

DOI: 10.1021/acsami.9b01335 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

ACS Applied Materials & Interfaces



Nevertheless, the 300 K n values below 1016 cm−3 obtained in this work are in the range appropriate for solar cells, demonstrating, in agreement with other recent single-crystal work, that this is possible in pyrite. The mobilities obtained in this range are up to 100 cm2 V−1 s−1, which should also be sufficient; this value could additionally be improved by the further reduction in background doping discussed above. Achieving all of this in thin films, where pyrite PV is most appealing, stands as a major challenge. In summary, thoroughly characterized high-quality pyrite FeS2 single crystals have been shown to exhibit increased bulk Hall electron density and mobility, decreased bulk resistivity, and approach to an insulator−metal transition when grown at progressively decreased S vapor pressure. These changes are uncorrelated with impurity concentrations and can only be simply interpreted in terms of a sulfur-vacancy-related deep (∼225 meV) donor. Wide-range and controlled n-doping of FeS2 is thereby realized, simply by varying growth conditions. This work thus contributes substantially to the elucidation and control of doping in pyrite, a critical limiting factor in its development as a low-cost, earth-abundant, nontoxic photovoltaic material.



REFERENCES

(1) Shockley, W.; Queisser, H. J. Detailed Balance Limit of Efficiency of p-n Junction Solar Cells. J. Appl. Phys. 1961, 32, 510− 519. (2) Ennaoui, A.; Fiechter, S.; Pettenkofer, C.; Alonso-Vante, N.; Büker, K.; Bronold, M.; Höpfner, C.; Tributsch, H. Iron Disulfide for Solar Energy Conversion. Sol. Energy Mater. Sol. Cells 1993, 29, 289− 370. (3) Schieck, R.; Hartmann, A.; Fiechter, S.; Könenkamp, R.; Wetzel, H. Electrical Properties of Natural and Synthetic Pyrite (FeS2) Crystals. J. Mater. Res. 1990, 5, 1567−1572. (4) Büker, K.; Alonso-Vante, N.; Tributsch, H. Photovoltaic Output Limitation of n-FeS2 (Pyrite) Schottky Barriers: A TemperatureDependent Characterization. J. Appl. Phys. 1992, 72, 5721−5728. (5) Ennaoui, A.; Tributsch, H. Iron Sulphide Solar Cells. Sol. Cells 1984, 13, 197−200. (6) Ennaoui, A.; Fiechter, S.; Jaegermann, W.; Tributsch, H. Photoelectrochemistry of Highly Quantum Efficient Single-Crystalline n-FeS2 (Pyrite). J. Electrochem. Soc. 1986, 133, 97−106. (7) Ennaoui, A.; Fiechter, S.; Tributsch, H.; Giersig, M.; Vogel, R.; Weller, H. Photoelectrochemical Energy Conversion Obtained with Ultrathin Organo-Metallic-Chemical-Vapor-Deposition Layer of FeS2 (Pyrite) on TiO2. J. Electrochem. Soc. 1992, 139, 2514−2518. (8) Wadia, C.; Alivisatos, A. P.; Kammen, D. M. Materials Availability Expands the Opportunity for Large-Scale Photovoltaics Deployment. Environ. Sci. Technol. 2009, 43, 2072−2077. (9) Limpinsel, M.; Farhi, N.; Berry, N.; Lindemuth, J.; Perkins, C. L.; Lin, Q.; Law, M. An Inversion Layer at the Surface of n-Type Iron Pyrite. Energy Environ. Sci. 2014, 7, 1974−1989. (10) Cabán-Acevedo, M.; Kaiser, N. S.; English, C. R.; Liang, D.; Thompson, B. J.; Chen, H.; Czech, K. J.; Wright, J. C.; Hamers, R. J.; Jin, S. Ionization of High-Density Deep Donor Defect States Explains the Low Photovoltage of Iron Pyrite Single Crystals. J. Am. Chem. Soc. 2014, 136, 17163−17179. (11) Yu, L.; Lany, S.; Kykyneshi, R.; Jieratum, V.; Ravichandran, R.; Pelatt, B.; Altschul, E.; Platt, H. A. S.; Wager, J. F.; Keszler, D. A.; Zunger, A. Iron Chalcogenide Photovoltaic Absorbers. Adv. Energy Mater. 2011, 1, 748−753. (12) Sun, R.; Chan, M. K. Y.; Ceder, G. First-Principles Electronic Structure and Relative Stability of Pyrite and Marcasite: Implications for Photovoltaic Performance. Phys. Rev. B 2011, 83, No. 235311. (13) Zhang, X.; Manno, M.; Baruth, A.; Johnson, M.; Aydil, E. S.; Leighton, C. Crossover From Nanoscopic Intergranular Hopping to Conventional Charge Transport in Pyrite Thin Films. ACS Nano 2013, 7, 2781−2789. (14) Zhang, X.; Li, M.; Walter, J.; O’Brien, L.; Manno, M. A.; Voigt, B.; Mork, F.; Baryshev, S. V.; Kakalios, J.; Aydil, E. S.; Leighton, C. Potential Resolution to the Doping Puzzle in Iron Pyrite: Carrier Type Determination by Hall Effect and Thermopower. Phys. Rev. Mater. 2017, 1, No. 015402. (15) Walter, J.; Zhang, X.; Voigt, B.; Hool, R.; Manno, M.; Mork, F.; Aydil, E. S.; Leighton, C. Surface Conduction in n-Type Pyrite FeS2 Single Crystals. Phys. Rev. Mater. 2017, 1, No. 065403. (16) Sun, R.; Chan, M. K. Y.; Kang, S.; Ceder, G. Intrinsic Stoichiometry and Oxygen-Induced p-Type Conductivity of Pyrite FeS2. Phys. Rev. B 2011, 84, No. 035212. (17) Zhang, X.; Scott, T.; Socha, T.; Nielsen, D.; Manno, M.; Johnson, M.; Yan, Y.; Losovyj, Y.; Dowben, P.; Aydil, E. S.; Leighton, C. Phase Stability and Stoichiometry in Thin Film Iron Pyrite: Impact on Electronic Transport Properties. ACS Appl. Mater. Interfaces 2015, 7, 14130−14139. (18) Seefeld, S.; Limpinsel, M.; Liu, Y.; Farhi, N.; Weber, A.; Zhang, Y.; Berry, N.; Kwon, Y. J.; Perkins, C. L.; Hemminger, J. C.; Wu, R.; Law, M. Iron Pyrite Thin Films Synthesized from an Fe(acac)3 Ink. J. Am. Chem. Soc. 2013, 135, 4412−4424. (19) Berry, N.; Cheng, M.; Perkins, C.; Limpinsel, M.; Hemminger, J. C.; Law, M. Atmospheric-Pressure Chemical Vapor Deposition of Iron Pyrite Thin Films. Adv. Energy Mater. 2012, 2, 1124−1135.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.9b01335.



Research Article

Calculations on VS diffusion and decomposition to pyrrhotite; additional chemical and impurity characterization; additional transport data and impurity analysis; and density-of-states schematics (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (E.S.A.). *E-mail: [email protected] (C.L.). ORCID

Jeff D. Jeremiason: 0000-0003-3608-6793 Chris Leighton: 0000-0003-2492-0816 Author Contributions

B.V., W.M., M.M., and J.W. grew the crystals, structurally, chemically, and electrically characterized them, and performed all related data analyses and calculations. J.D.J. was responsible for ICPMS measurements. C.L. and E.S.A. conceived of the study and oversaw its execution. B.V., C.L., and E.S.A. wrote the paper, with input from all other authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the customers of Xcel Energy through a grant from the Renewables Development Fund and in part by the National Science Foundation (NSF) through the University of Minnesota MRSEC under DMR-1420013. Parts of this work were carried out in the Characterization Facility, University of Minnesota, which receives partial support from NSF through the MRSEC program. The authors thank D. Ray, L. Gagliardi, K. Reich, and B. Shklovskii for informative discussions. J

DOI: 10.1021/acsami.9b01335 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces (20) Shukla, S.; Nguyen, H. L.; Boix, P. P.; Koh, T. M.; Prabhakar, R. R.; Mulmudi, H. K.; Zhang, J.; Chen, S.; Ng, C. F.; Huan, C. H. A.; Mathews, N.; Sritharan, T.; Xiong, Q. Iron Pyrite Thin Films Counter Electrodes for Dye-Sensitized Solar Cells: High Efficiency for Iodine and Cobalt Redox Electrolyte Cells. ACS Nano 2014, 8, 10597− 10605. (21) Shukla, S.; Xing, G.; Ge, H.; Prabhakar, R. R.; Mathew, S.; Su, Z.; Nalla, V.; Venkatesan, T.; Mathews, N.; Sritharan, T.; Sum, T. C.; Xiong, Q. Origin of Photocarrier Losses in Iron Pyrite (FeS2) Nanocubes. ACS Nano 2016, 10, 4431−4440. (22) Cabán-Acevedo, M.; Faber, M. S.; Tan, Y.; Hamers, R. J.; Jin, S. Synthesis and Properties of Semiconducting Iron Pyrite (FeS2) Nanowires. Nano Lett. 2012, 12, 1977−1982. (23) Cabán-Acevedo, M.; Liang, D.; Chew, K. S.; Degrave, J. P.; Kaiser, N. S.; Jin, S. Synthesis, Characterization, and Variable Range Hopping Transport of Pyrite (FeS2) Nanorods, Nanobelts, and Nanoplates. ACS Nano 2013, 7, 1731−1739. (24) Lucas, J. M.; Tuan, C.-C.; Lounis, S. D.; Britt, D. K.; Qiao, R.; Yang, W.; Lanzara, A.; Alivisatos, A. P. Ligand-Controlled Colloidal Synthesis and Electronic Structure Characterization of Cubic Iron Pyrite (FeS2) Nanocrystals. Chem. Mater. 2013, 25, 1615−1620. (25) Bronold, M.; Tomm, Y.; Jaegermann, W. Surface States on Cubic d-Band Semiconductor Pyrite (FeS2). Surf. Sci. 1994, 314, L931−L936. (26) Willeke, G.; Blenk, O.; Kloc, C.; Bucher, E. Preparation and Electrical Transport Properties of Pyrite (FeS2) Single Crystals. J. Alloys Compd. 1992, 178, 181−191. (27) Kinner, T.; Bhandari, K. P.; Bastola, E.; Monahan, B. M.; Haugen, N. O.; Roland, P. J.; Bigioni, T. P.; Ellingson, R. J. Majority Carrier Type Control of Cobalt Iron Sulfide (CoxFe1−xS2) Pyrite Nanocrystals. J. Phys. Chem. C 2016, 120, 5706−5713. (28) Yamamoto, A.; Nakamura, M.; Seki, A.; Li, E. L.; Hashimoto, A.; Nakamura, S. Pyrite (FeS2) Thin Films Prepared by Spray Method Using FeSO4 and (NH4)2Sx. Sol. Energy Mater. Sol. Cells 2003, 75, 451−456. (29) Lichtenberger, D.; Ellmer, K.; Schieck, R.; Fiechter, S.; Tributsch, H. Structural, Optical and Electrical Properties of Polycrystalline Iron Pyrite Layers Deposited by Reactive D.C. Magnetron Sputtering. Thin Solid Films 1994, 246, 6−12. (30) Ares, J. R.; Ferrer, I. J.; Sánchez, C. R. Majority Carriers in Pyrite Thin Films: An Analysis Based on Seebeck and Hall Coefficient Measurements. Thin Solid Films 2003, 431−432, 511−513. (31) Willeke, G.; Dasbach, R.; Sailer, B.; Bucher, E. Thin Pyrite (FeS2) Films Prepared by Magnetron Sputtering. Thin Solid Films 1992, 213, 271−276. (32) Le Comber, P. G.; Jones, D. I.; Spear, W. E. Hall Effect and Impurity Conduction in Substitutionally Doped Amorphous Silicon. Philos. Mag. A 1977, 35, 1173−1187. (33) Lewis, A. J. Conductivity and Thermoelectric Power of Amorphous Germanium and Amorphous Silicon. Phys. Rev. B 1976, 13, 2565−2575. (34) Note that in that work the Co concentration was deduced from a certificate of analysis of the Fe precursor, not from measurements on resultant FeS2 crystals. Note also that C and O were not included in the total impurity count, as they were surface-segregated. H was included, at 15 ppm, the concentration determined from bulk (polished) crystals. (35) Lehner, S. W.; Savage, K. S.; Ayers, J. C. Vapor Growth and Characterization of Pyrite (FeS2) Doped with Co, Ni, and As: Variations in Semiconducting Properties. J. Cryst. Growth 2006, 286, 306−317. (36) Lehner, S. W.; Newman, N.; Van Schilfgaarde, M.; Bandyopadhyay, S.; Savage, K.; Buseck, P. R. Defect Energy Levels and Electronic Behavior of Ni-, Co-, and As-Doped Synthetic Pyrite (FeS2). J. Appl. Phys. 2012, 111, No. 083717. (37) Guo, S.; Young, D. P.; Macaluso, R. T.; Browne, D. A.; Henderson, N. L.; Chan, J. Y.; Henry, L. L.; Ditusa, J. F. Charge Transport in Cobalt-Doped Iron Pyrite. Phys. Rev. B 2010, 81, No. 144424.

(38) Guo, S.; Young, D. P.; Macaluso, R. T.; Browne, D. A.; Henderson, N. L.; Chan, J. Y.; Henry, L. L.; Ditusa, J. F. Discovery of the Griffiths Phase in the Itinerant Magnetic Semiconductor Fe1‑xCoxS2. Phys. Rev. Lett. 2008, 100, No. 017209. (39) Luck, J.; Hartmann, A.; Fiechter, S. Stoichiometry and Impurity Concentration in Synthetically Grown Iron Pyrite Crystals and Their Constituents. Fresenius’ Z. Anal. Chem. 1989, 334, 441−446. (40) Fiechter, S.; Birkholz, M.; Hartmann, A.; Dulski, P.; Giersig, M.; Tributsch, H.; Tilley, R. J. D. The Microstructure and Stoichiometry of Pyrite FeS2−x. J. Mater. Res. 1992, 7, 1829−1838. (41) Birkholz, M.; Fiechter, S.; Hartmann, A.; Tributsch, H. Sulfur Deficiency in Iron Pyrite (FeS2‑x) and its Consequences for BandStructure Models. Phys. Rev. B 1991, 43, 11926−11936. (42) Andersson, K. J.; Ogasawara, H.; Nordlund, D.; Brown, G. E., Jr.; Nilsson, A. Preparation, Structure, and Orientation of Pyrite FeS2{100} Surfaces: Anisotropy, Sulfur Monomers, Dimer Vacancies, and a Possible FeS Surface Phase. J. Phys. Chem. C 2014, 118, 21896− 21903. (43) Andersson, K.; Nyberg, M.; Ogasawara, H.; Nordlund, D.; Kendelewicz, T.; Doyle, C. S.; Brown, G. E., Jr.; Pettersson, L. G. M.; Nilsson, A. Experimental and Theoretical Characterization of the Structure of Defects at the Pyrite FeS2(100) Surface. Phys. Rev. B 2004, 70, No. 195404. (44) Schaufuß, A. G.; Nesbitt, H. W.; Kartio, I.; Laajalehto, K.; Bancroft, G. M.; Szargan, R. Reactivity of Surface Chemical States on Fractured Pyrite. Surf. Sci. 1998, 411, 321−328. (45) Nesbitt, H. W.; Bancroft, G. M.; Pratt, A. R.; Scaini, M. J. Sulfur and Iron Surface States on Fractured Pyrite Surfaces. Am. Mineral. 1998, 83, 1067−1076. (46) Herbert, F. W.; Krishnamoorthy, A.; Ma, W.; Van Vliet, K. J.; Yildiz, B. Dynamics of Point Defect Formation, Clustering and Pit Initiation on the Pyrite Surface. Electrochim. Acta 2014, 127, 416− 426. (47) Herbert, F. W.; Krishnamoorthy, A.; Van Vliet, K. J.; Yildiz, B. Quantification of Electronic Band Gap and Surface States on FeS2(100). Surf. Sci. 2013, 618, 53−61. (48) Guevremont, J. M.; Strongin, D. R.; Schoonen, M. A. A. Photoemission of Adsorbed Xenon, X-Ray Photoelectron Spectroscopy, and Temperature-Programmed Desorption Studies of H2O on FeS2(100). Langmuir 1998, 14, 1361−1366. (49) Hu, J.; Zhang, Y.; Law, M.; Wu, R. First-Principles Studies of the Electronic Properties of Native and Substitutional Anionic Defects in Bulk Iron Pyrite. Phys. Rev. B 2012, 85, No. 085203. (50) Krishnamoorthy, A.; Herbert, F. W.; Yip, S.; Van Vliet, K. J.; Yildiz, B. Electronic States of Intrinsic Surface and Bulk Vacancies in FeS2. J. Phys.: Condens. Matter 2013, 25, No. 045004. (51) Zhang, Y. N.; Law, M.; Wu, R. Q. Atomistic Modeling of Sulfur Vacancy Diffusion Near Iron Pyrite Surfaces. J. Phys. Chem. C 2015, 119, 24859−24864. (52) Ganduglia-Pirovano, M. V.; Hofmann, A.; Sauer, J. Oxygen Vacancies in Transition Metal and Rare Earth Oxides: Current State of Understanding and Remaining Challenges. Surf. Sci. Rep. 2007, 62, 219−270. (53) Tufte, O. N.; Chapman, P. W. Electron Mobility in Semiconducting Strontium Titanate. Phys. Rev. 1967, 155, 796−802. (54) Spinelli, A.; Torija, M. A.; Liu, C.; Jan, C.; Leighton, C. Electronic Transport in Doped SrTiO3: Conduction Mechanisms and Potential Applications. Phys. Rev. B 2010, 81, No. 155110. (55) McCalla, E.; Phelan, D.; Krogstad, M. J.; Dabrowski, B.; Leighton, C. Electrical Transport, Magnetic, and Thermodynamic Properties of La-, Pr-, and Nd-Doped BaSnO3−δ Single Crystals. Phys. Rev. Mater. 2018, 2, No. 084601. (56) Ganguly, K.; Ambwani, P.; Xu, P.; Jeong, J. S.; Mkhoyan, K. A.; Leighton, C.; Jalan, B. Structure and Transport in High Pressure Oxygen Sputter-Deposited BaSnO 3‑δ . APL Mater. 2015, 3, No. 062509. (57) Ganguly, K.; Prakash, A.; Jalan, B.; Leighton, C. MobilityElectron Density Relation Probed via Controlled Oxygen Vacancy Doping in Epitaxial BaSnO3. APL Mater. 2017, 5, No. 056102. K

DOI: 10.1021/acsami.9b01335 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces (58) Yahia, J. Dependence of the Electrical Conductivity and Thermoelectric Power of Pure and Aluminum-Doped Rutile on Equilibrium Oxygen Pressure and Temperature. Phys. Rev. 1963, 130, 1711−1719. (59) Weiher, R. L. Electrical Properties of Single Crystals of Indium Oxide. J. Appl. Phys. 1962, 33, 2834−2839. (60) Fan, J. C. C.; Goodenough, J. B. X-Ray Photoemission Spectroscopy Studies of Sn-Doped Indium-Oxide Films. J. Appl. Phys. 1977, 48, 3524−3531. (61) Watson, E. B.; Cherniak, D. J.; Frank, E. A. Retention of Biosignatures and Mass-Independent Fractionations in Pyrite: SelfDiffusion of Sulfur. Geochim. Cosmochim. Acta 2009, 73, 4792−4802. (62) Lambert, J. M., Jr.; Simkovich, G.; Walker, P. L., Jr. The Kinetics and Mechanism of the Pyrite-to-Pyrrhotite Transformation. Metall. Mater. Trans. B 1998, 29, 385−396. (63) Butler, S. R.; Bouchard, R. J. Single Crystal Growth of Pyrite Solid Solutions. J. Cryst. Growth 1971, 10, 163−169. (64) Brostigen, G.; Kjekshus, A. Redetermined Crystal Structure of FeS2 (Pyrite). Acta Chem. Scand. 1969, 23, 2186−2188. (65) Fleet, M. E. The Crystal Structure of a Pyrrhotite (Fe7S8). Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1971, 27, 1864− 1867. (66) Fiechter, S.; Mai, J.; Ennaoui, A.; Szacki, W. Chemical Vapour Transport of Pyrite (FeS2) with Halogen (Cl, Br, I). J. Cryst. Growth 1986, 78, 438−444. (67) Brandon, D.; Kaplan, W. D. Microstructural Characterization of Materials, 2nd ed.; Wiley, 2008. (68) Maxwell, J. A.; Campbell, J. L.; Teesdale, W. J. The Guelph PIXE Software Package. Nucl. Instrum. Methods Phys. Res., Sect. B 1989, 43, 218−230. (69) Bither, T. A.; Bouchard, R. J.; Cloud, W. H.; Dokohue, P. C.; Siemons, W. J. Transition Metal Pyrite Dichalcogenides. HighPressure Synthesis and Correlation of Properties. Inorg. Chem. 1968, 7, 2208−2220. (70) Bayliss, P. Crystal Structure Refinement of a Weakly Anisotropic Pyrite. Am. Mineral. 1977, 62, 1168−1172. (71) Blenk, O.; Bucher, E.; Willeke, G. p-Type Conduction in Pyrite Single Crystals Prepared by Chemical Vapor Transport. Appl. Phys. Lett. 1993, 62, 2093−2095. (72) Voigt, B.; Moore, W.; Walter, J.; Manno, M.; Aydil, E. S.; Leighton, C. unpublished. (73) Ashcroft, N. W.; Mermin, N. D. Solid State Physics; Harcourt College, 1976. (74) Sze, S. M. Physics of Semiconductor Devices; John Wiley & Sons: New York, 1981. (75) Ho, C. H.; Huang, C. E.; Wu, C. C. Preparation and Characterization of Ni-Incorporated FeS2 Single Crystals. J. Cryst. Growth 2004, 270, 535−541. (76) Mott, N. F. Metal-Insulator Transitions, 2nd ed.; Taylor & Francis, 1990. (77) Shklovskii, B. I.; Efros, A. L. Electronic Properties of Doped Semiconductors; Springer: Berlin, 1984. (78) Karguppikar, A. M.; Vedeshwar, A. G. Transport Properties of Thin Iron Pyrite Films. Phys. Status Solidi A 1986, 95, 717−720. (79) Mott, N. F. Metal-Insulator Transition. Rev. Mod. Phys. 1968, 40, 677−683. (80) Carlson, R. O.; Hall, R. N.; Pell, E. M. Sulfur in Silicon. J. Phys. Chem. Solids 1959, 8, 81−83. (81) Winkler, M. T.; Recht, D.; Sher, M.; Said, A. J.; Mazur, E.; Aziz, M. J. Insulator-to-Metal Transition in Sulfur-Doped Silicon. Phys. Rev. Lett. 2011, 106, No. 178701. (82) Younkin, R.; Carey, J. E.; Mazur, E.; Levinson, J. A.; Friend, C. M. Infrared Absorption by Conical Silicon Microstructures Made in a Variety of Background Gases Using Femtosecond-Laser Pulses. J. Appl. Phys. 2003, 93, 2626−2629. (83) Crouch, C. H.; Carey, J. E.; Shen, M.; Mazur, E.; Génin, F. Y. Infrared Absorption by Sulfur-Doped Silicon Formed by Femtosecond Laser Irradiation. Appl. Phys. A 2004, 79, 1635−1641.

(84) Ertekin, E.; Winkler, M. T.; Recht, D.; Said, A. J.; Aziz, M. J.; Buonassisi, T.; Grossman, J. C. Insulator-to-Metal Transition in Selenium-Hyperdoped Silicon: Observation and Origin. Phys. Rev. Lett. 2012, 108, No. 026401. (85) Ray, D.; Voigt, B.; Manno, M.; Aydil, E. S.; Leighton, C.; Gagliardi, L. unpublished. (86) Puff, W.; Birkholz, M.; Balogh, A. G.; Fiechter, S. Vacancy-Type Defects in Iron-Pyrite FeS2‑x. Mater. Sci. Forum 1997, 255−257, 342− 344.

L

DOI: 10.1021/acsami.9b01335 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX