Kuramite Cu3SnS4 and Mohite Cu2SnS3 Nanoplatelet Synthesis

Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI

Article 3

4

2

3

Kuramite CuSnS and Mohite CuSnS Nanoplatelet Synthesis using Covellite CuS Templates with Sn(II) and Sn(IV) Sources Yang Liu, Maixian Liu, Deqiang Yin, Wei Wei, Paras N. Prasad, and Mark T. Swihart Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.6b05428 • Publication Date (Web): 30 Mar 2017 Downloaded from http://pubs.acs.org on March 31, 2017

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

Kuramite Cu3SnS4 and Mohite Cu2SnS3 Nanoplatelet Synthesis using Covellite CuS Templates with Sn(II) and Sn(IV) Sources Yang Liu,†,§ Maixian Liu,‡,§ Deqiang Yin,† Wei Wei,§ Paras N. Prasad‡,§ and Mark T. Swihart†,§,* †

Department of Chemical and Biological Engineering, ‡Department of Chemistry, and §Institute for Lasers, Photonics, and Biophotonics, University at Buffalo, The State University of New York, Buffalo, New York 14260, United States

ABSTRACT: Colloidal synthesis of copper tin sulfide (CTS) nanocrystals has been of great interest due to their potential for use in solution-processed photovoltaics containing only earth-abundant, low-toxicity elements. Post-synthetic incorporation of Sn into copper sulfide is an effective means of producing CTS, and Sn content can be controlled to some extent by the amount of Sn precursor provided. However, the oxidation (valence) state of Sn before and after doping and the effectiveness of Sn(II) vs. Sn(IV) precursors remain topics of significant debate. Here, we demonstrate a step-growth method for preparing monodisperse copper tin sulfide (CTS) nanoplatelets (NPls). We show that Sn2+ ions, but not Sn4+ ions, can convert covellite to Cu3SnxS4 NPls (x ≤ 1) without addition of any reducing or copper-extracting agent. In this case Sn2+ reduces the disulfide bond in covellite and the crystal structure evolves from covellite to kuramite. When dodecanethiol (DDT) is added prior to Sn, only Sn4+ ions, and not Sn2+ ions, are incorporated, ultimately producing mohite Cu2SnS3. Here, DDT can reduce the disulfide bonds and extract Cu from the lattice. When djurleite Cu2-xS was used as the template, only Sn4+, and not Sn2+ could be incorporated without use of DDT, and the extent of Sn incorporation was very limited, perhaps only filling Cu vacancies and not displacing any Cu from the lattice. Together with previous studies of CTS preparation from Cu2-xS using Sn4+ and DDT, these results provide a flexible array of methods of achieving a desired final CTS composition, crystal structure, and morphology. The ability to produce two different crystal phases of CTS, with corresponding differences in band gap and other properties, from the same covellite template while retaining the template size and morphology could be particularly valuable.

Introduction The family of copper chalcogenide-based semiconductors, which includes copper indium sulfide/selenide1-2 (CIS), copper tin sulfide3-7 (CTS), copper indium gallium sulfide/selenide8-9 (CIGS), and copper zinc tin sulfide/selenide10-15 (CZTS) is of great interest for use in thin film solar cells. Preparation of these materials as colloidal inks can allow low-cost fabrication by solution-phase printing and coating processes. To date, the efficiency of CZTS-based solar cells has reached 12.6%.11 CTS is an intermediate crystal that can be further doped with zinc to make CZTS. CTS nanocrystals (NCs) can, themselves, serve as an absorber for thin-film photovoltaics, with a nearoptimal band gap around 1.43eV.16 Further, CTS, is composed of relatively low-cost, naturally abundant and nontoxic elements.6-7, 16 Thus, it is of great interest both as an intermediate to formation of CZTS and for its own potential applications in solar cells and related optoelectronic devices. Properties of compound semiconductors, including the copper chalcogenides, can be tuned by varying the cation fraction in mixed cation materials such as CTS. Two alternatives are available for preparation of such materials: (1) direct synthesis of ternary materials using multiple cation precursors simultaneously;2-3 and (2) multistep synthesis in which NPs of a binary material are first prepared, and then a cation exchange process is used to introduce the second cation.6 In some cases, copper chalcogenides have been used as templates for complete cation exchange to produce an entirely different metal chalcogenide from which all copper cations have been depleted.17 Direct synthesis of ternary or quaternary materials from multiple precursors in one pot often suffers from stoichio-

metric imbalance and formation of secondary phases due to differences in precursor reactivities. Cation exchange can circumvent this by stepwise replacement of cations. Adjustment of reactivity of one reagent is much easier than managing the reactivity of multiple precursors simultaneously.18 The cation ratio in the product can also be easily manipulated by the amount of substitutive cation provided. For copper sulfide, the process can also be tuned through use of copper-extracting agents such as tri-octyl phosphine (TOP) and dodecanethiol (DDT). The strong affinity of such reagents for Cu+, combined with the modest stability of the soft copper-chalcogenide bond, facilitates exchange of Cu for metals that form stronger bonds with chalcogenides and/or are more weakly complexed by these agents.18 This has been demonstrated for making CdS from copper sulfide.19 Choosing appropriate copper-extracting ligands is crucial in cation exchange. Although copper sulfide doping with Sn has been previously studied, valence selectivity has been greatly debated for converting copper chalcogenides into copper tin chalcogenides. Recent research suggested Sn4+ can more effectively insert into the CuX substrate, due to its smaller size relative to Sn2+. Starting from berzelianite Cu2-xSe, use of Sn4+ precursors allowed production of CTSe NCs, while use of Sn2+ produced heterostructures of Cu2-xSe and SnSe.20 The effectiveness of Sn incorporation, however, may vary with the stoichiometry and crystal phase of the starting copper chalcogenide nanostructures. Starting from NCs of different crystal phase may also produce final alloyed NCs of different crystal phases. Here we focus on using covellite CuS as the host lattice, because few examples of cation insertion or exchange starting from covellite CuS have been reported,21 and because covellite has the unique structural feature of potentially reactive disul-

ACS Paragon Plus Environment


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

Page 2 of 9

fide bonds in its crystal structure. Stabilizing ligands with a variety of functional groups have been employed to promote cation insertion and exchange. We selected DDT for its low toxicity, stability in air, and good performance in other cation exchange reactions.18 Here, we focus on the cation valence-dependent conversion of covellite nanocrystals to ternary CTS materials, starting from relatively large covellite nanoplatelets. We found the ability of Sn2+ or Sn4+ to incorporate into copper sulfide was determined by the host lattice. Covellite accepts only Sn2+, eventually converting to kuramite CTS. However, this result was reversed in the presence of DDT, which produced mohite upon incorporation of Sn4+. Only Sn4+ was effective for incorporation of Sn into djurleite (Cu31S16). We attribute this cation selectivity to the disulfide bond in the covellite nanocrystal. The different phases of final product reflect two different cation incorporation mechanisms.

Preparing djurleite (Cu2-xS) NPls: We followed our previously-reported method for preparing Cu2-xS djurleite NPls.22 Briefly, S-OA precursor was prepared by mixing 1 mmol sulfur powder with 10 mL OLAM and heating at 120°C for 20– 30 min. 12 nm Cu2-xS NPls were produced by injecting 5 mL S-OA into 10 mL Cu(I)-OLAM solution at 140°C. The temperature was held at 133°C for about 2 min. Incorporation of Sn into Cu2-xS followed the same procedure used for CuS. Incorporation of Sn into Cu2-xS NCs: To incorporate Sn into Cu2-xS, in a typical synthesis, 0.167 mmol (58.3 mg) of SnCl4 or (31.5 mg) of SnCl2 was dissolved in 10 mL OLAM and then heated to 160 °C. Cu2-xS containing 0.5 mmol Cu was injected into Sn-OLAM and aged for 20 min. To purify the NPls, 10 mL ethanol was added to the reaction solution, followed by centrifuging at 5000 rpm for about 1 min.

Experimental Section

Transmission electron microscopy (TEM) images were obtained using a JEOL JEM-2010 or an FEI Tecnai F20 G2, in both cases at an accelerating voltage of 200 kV. TEM grids were prepared by dropping a dilute NC dispersion onto a carbon-coated copper or nickel TEM grid and drying in air. Size distributions were obtained from TEM images by measuring at least 100 (in most cases at least 150) NCs from multiple TEM images from different areas of the grid, using Nanomeasurer v. 1.2. STEM imaging was carried out on a FEI Tecnai F20 200KV STEM with a Fischione HAADF STEM detector using samples prepared on Ni TEM grids. Powder X-ray diffraction (XRD) measurements were carried out using a Rigaku Ultima IV diffractometer with a Cu Kα X-ray source. Samples were prepared by drop-casting concentrated NC dispersions onto glass slides. Elemental analysis of NCs by energy-dispersive X-ray spectroscopy (EDX) was obtained using an Oxford Instruments X-Max 20 mm2 EDX detector within a Zeiss Auriga scanning electron microscope (SEM). UV-vis-NIR spectra of NC dispersions in chloroform were taken on a Shimadzu UV3600. X-ray photoelectron spectra (XPS) were acquired on a Phi Versa Probe 5000 with an Al Kα X-ray source.

All chemicals were used as received. Oleylamine (70%, OLAM), tin (II) chloride (SnCl2, reagent grade, 98%), tin (IV) chloride pentahydrate (SnCl4, reagent grade, 98%), oleic acid (technical grade, OA 90%) and 1-dodecanethiol (DDT, ≥ 98%) were purchased from Sigma-Aldrich. CuCl2·2H2O, CuCl, sulfur powder, toluene, and ammonium sulfide (21.2 wt% in water) were purchased from Fisher Scientific. Preparing CuS Nanoplatelets (NPls): For the synthesis of ~55 nm hexagonal CuS NPls, 1.5 mmol of CuCl2·2H2O was dissolved in a mixture of 10 mL OLAM and 10 mL toluene, then heated to 70°C. 1.5 mL ammonium sulfide solution (21.2% in water) was injected into the Cu-OLAM to form CuS NPls. After 120 min, 10 mL ethanol was added to destabilize the NPl dispersion, followed by centrifugation at 4000 rpm for 1 min. The NPls could be redispersed in organic solvents including chloroform, hexane and toluene. Toluene dispersions were used in subsequent experiments reported here. Incorporation of Sn into CuS NPls: In a typical synthesis of kuramite NPls using Sn2+ precursor without DDT, 0.167 mmol (31.67 mg) SnCl2 was dissolved in 10 mL OLAM and then heated to 160 °C. 0.5 mL of CuS NPl dispersion containing 0.5 mmol Cu was injected into Sn-OLAM and aged for 20 min. To purify the NPls, 10 mL ethanol was added, followed by centrifuging at 4000 rpm for about 1 min. Different SnCl2 amounts were used to synthesize Cu3SnxS3+x NCs (x ≤ 1) with different compositions. To test the incorporation of Sn4+ into CuS, in a typical synthesis, 0.167 mmol (58.3 mg) of SnCl4 was used to replace SnCl2, while the other experimental conditions and procedures were kept the same as those for the synthesis of kuramite NPls. For synthesis of mohite Cu2SnS3 NPls, 0.167 mmol (58.3 mg) of SnCl4 was added to a mixture of 8 mL OLAM and 2 mL DDT, and then heated to 160 °C, followed by injecting 0.5 mL of CuS NPl dispersion into Sn-OLAM-DDT and aging for 20 min. To purify the NPls, 10 mL ethanol was added, followed by centrifuging at 4000 rpm for about 1 min. To test the incorporation of Sn2+ in the presence of DDT, 0.167 mmol (31.67 mg) SnCl2 was used to replace SnCl4, while the other experimental conditions and procedures were kept the same as those for the synthesis of mohite NPls

Characterization

Figure 1. (a) TEM image of covellite CuS NPls with diameter of ~55.5±2.8 nm. (b) XRD pattern of covellite CuS NPls. (c) top view and (d) side view HRTEM images of covellite CuS NPls with thickness of ~4.0 nm.


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


Figure 2. TEM images of CTS NPls produced using (a) 7:1, (b) 5:1, and (c) 3:1 ratio of Cu:Sn. Insets show side-view images from the same samples, and are at the same scale as the corresponding main panel.

Results CuS Nanoplatelet Synthesis. Large nanoplatelets were synthesized by a minor modification our previously reported method.23 Synthesis at 70 °C produced large nanoplatelets with a fixed thickness of ~4 nm. The size of the NPls was controlled either by addition of ammonium sulfide or by the concentration of precursors in the solution. Size changes can be explained by sulfur-induced aggregation and nucleationand-growth theory, respectively.23-24 Figure 1a shows a TEM image of 55.5±2.8 nm covellite CuS NPls. Diameters of hexagons are reported as mean ± one standard deviation, according their excircles. Size-distribution histograms for this and subsequent samples are presented in Fig. S1. X-ray diffraction showed that the crystal phase of the CuS NPls is covellite (ICDD PDF 04-007-1392, figure 1b). The lattice fringes in top-view HRTEM images (Fig. 1(c)) have interplanar spacings of 0.19 nm and 0.33 nm, corresponding to the (110) and (100) planes of covellite, respectively. Side-view HRTEM images (Fig. 1(d)) show a lattice spacing of 0.27 nm, corresponding to (001) planes. The TEM and HRTEM images shows that the and directions lie within the plane of the Cu NPls, while the direction corresponds to the NPl thickness. This is consistent with the sharp (110) peak in the XRD pattern (Fig. 1(b)). As shown previously,23 this synthesis approach constrains growth in the direction while allowing control of the size in the perpendicular directions. I. Reaction without DDT Incorporation of Sn2+ into covellite. We tested the incorporation of Sn into the CuS NPls using precursors with each of the stable Sn oxidation states (Sn4+ and Sn2+) to deduce the dependence of Sn incorporation on valency. We first discuss results obtained using Sn2+ ions. Figure 2 shows typical TEM images of Cu3SnxS3+x NCs (x ≤ 1). Sn2+ was supplied in amounts corresponding to Cu:Sn ratios of 7:1, 5:1 and 3:1, using ~55 nm NPls as templates. The size and shape of the CuS NPls were preserved as Sn was incorporated. Use of appropriate temperature and low concentrations of the OLAMSn complex ensured complete preservation of the morphology of the initial NPls. However, at higher temperature (200°C), changes in the shape of the CuS NPls were observed (Fig. S2 in supporting information). Elemental analysis by EDX showed a linear relationship between the Sn cation fraction (Sn/(Cu+Sn)) in the precursors and the final Sn cation fraction in the CTS NPls (Fig. S3). Combined with the observation that no copper ions remain in solution, either after the NPl synthe-

sis or after Sn incorporation, this implies that the efficiency of Sn incorporation was nearly 100%; virtually all of the Sn supplied as OLAM-Sn2+ was incorporated into the covellite host lattice. Starting from the CuS NPl templates, Cu3SnxS3+x NCs (0 ≤ x ≤ 1) gradually changed from the covellite (CuS) crystal structure to kuramite (Cu3SnS4). The excess sulfur required for this conversion was presumably provided by residual S2- remaining from the CuS NPl synthesis. As shown in figure 3, both crystal phases have a major peak at 2θ ≈ 48°, which changes little with Sn incorporation. The shift of the covellite peak near 31° to ~34° provides the clearest evidence of the phase transformation. We employed STEM HAADF EDX mapping to characterize the composition of individual NPls. Panels (a) and (b) of figure 4 show an STEM HAADF image and an EDX line-scan analysis of kuramite CTS NPls (prepared using a 3:1 Cu:Sn ratio), while panels (c) and (d) of Fig. 4 show the results obtained using an excess of OLAM-Sn2+ (1:1 Cu:Sn ratio), which results in formation of a secondary phase, visible as flowerlike structures. We first discuss the 3:1 case. The EDS linescan of Fig. 4(b) shows that Cu, Sn and S are uniformly distributed across the nanoplatelet. The signals decay on both sides, which suggests that the region outside the platelet is free of these elements. The chlorine signals are due to trace residues of the metal-Cl salts used as precursors. When excess Sn2+ was supplied, beyond the stoichiometric 3:1 Cu:Sn ratio for kuramite, a secondary phase appeared. A sharp new peak near 21.7° appears in the XRD pattern shown in Fig. S4(a). This phase appears as the nanoflowers shown in Fig. S4(b) and Fig. 4(c). There are multiple reports of flower-like tin sulfide NCs, which all agree that SnS NCs tend to form flower-like morphology.25-27 An EDS line scan (Fig. 4(d)) across the nanoflower in Fig. 4(c) shows that the Sn content is much higher than that of CTS NPls. However, the significant Cu signal detected implies that CTS NPls may be attached on or beneath the nanoflower. While the single new XRD peak and limited composition data do not allow unambiguous identification of this new phase, clearly a new Sn-rich phase is formed when excess Sn is added. The covellite NPls exhibit strong NIR LSPR absorbance from 1000 to 2000 nm depending on their aspect ratio.23 As the Sn content of the NPls increased, the LSPR shifted to longer wavelength (red-shifted) and decreased in intensity, an important indication of successful Sn incorporation, accompanied by a reduced cation deficiency and corresponding reduced free carrier concentration.3 Fig. S5a shows that starting



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

from covellite NCs, with an absorbance peak at 1400 nm, the addition of increasing amounts of Sn induced a red shift of the NIR absorbance band to around 1600 nm. The red-shift of the LSPR is attributed to a decrease in the free carrier concentration with increased Sn content,28 This change in LSPR absorbance is consistent with the transformation from covellite to kuramite. However, the fact that the LSPR absorbance persists even after conversion to kuramite suggests that the NPls remain cation deficient. Kuramite nanoparticles prepared directly in solution were reported not to exhibit LSPR,16 but thin films of copper-deficient kuramite have shown free carrier concentrations of 3×1021 cm-3.7 A similar free carrier concentration would produce NIR LSPR in the NPls. In addition, the UV-vis absorbance band of absorbance of these NPls are recorded in inset of Fig. S5a. The absorbance minimum (trough) between the UV-vis band-to-band absorbance and the LSPR absorbance also red-shifted with increased Sn incorporation, reflecting a decrease in band-gap energy accompanying the conversion from covellite CuS to kuramite Cu3SnS4. However, the overlap between the band-to-band and LSPR absorbances along with prevented us from quantitatively extracting the optical band gap from these absorbance spectra. The indirect band gaps of these CTS were estimated by extrapolating the linear portion of plots of (αhν)2 vs. hν, the photon energy, to α = 0 (Fig S5b). Based on these estimation, the band gap energy of covellite CuS NPls is nearly 2.40 eV. The indirect band gap of CTS NPls estimated in this manner increased from 2.46 eV to 2.90 eV with increasing incorporation of Sn into covellite CuS. Although overlap between the tails of band-gap absorbance and LSPR absorbance, along with possible defect-related absorbance, produce substantial ambiguity and uncertainty in the determination of bandgaps from such optical absorbance measurements, we believe the trend of increasing bandgap with increasing Sn incorporation is genuine.

Page 4 of 9

Figure 4. STEM HAADF TEM images (a,c) and STEM-EDX line scans (b,d) along the indicated lines, for (a,b) Cu3SnS4 NPls prepared using a 3:1 Cu:Sn ratio, and (c,d) NCs prepared using a 1:1 Cu:Sn ratio, which results in formation of a Sn-rich secondary phase.

Incorporation of Sn4+ into covellite. In contrast to Sn2+, we found that Sn4+ could not be incorporated into the CuS NPls under conditions similar to those used for Sn2+ incorporation. Panels (a) and (b) of Figure S6 show representative TEM images of the product of attempts to incorporate Sn4+ into covellite CuS. As shown there, the NPls were etched and separate nanoparticles of a secondary phase are evident. Comparison of XRD patterns of samples before and after the attempted Sn4+ incorporation confirm the presence of a second phase (Fig. 5). The primary diffraction peaks of covellite, corresponding to the (110) and (102) planes, remain clearly present after the Sn4+ addition. Simultaneously, all major peaks expected for berndtite SnS2 NCs [(110), (101), (003) and (103)] appear, clearly indicating the formation of a separate SnS2 phase. The coexistence of covellite CuS and berndtite SnS2 further confirms that Sn4+ cannot be incorporated into CuS without the use of additional reagents that promote Cu extraction or disulfide bond reduction.

Figure 3. X-ray diffraction patterns of Cu3SnxS4 (0≤x≤1), from covellite to kuramite NPls. Magenta, red, and blue curves are from NPls produced using 3:1, 5:1, and 7:1 Cu:Sn2+ ratios, respectively. Figure 5. XRD pattern of CuS NCs treated with Sn4+.


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


Reactions with djurleite copper sulfide. To further understand the mechanism of Sn incorporation into CuS NCs, we also conducted experiments starting from a different host lattice. Djurleite copper sulfide (Cu31S16) nanoparticles, like covellite, have a high concentration of free holes and exhibit LSPR absorbance.29-31 The key difference between djurleite and covellite is the absence of disulfide bonds in djurleite, which implies that Sn incorporation requires actual cation exchange and cannot occur by disulfide bond reduction. Thus, the similarity of between Cu and Sn ions is expected to be more important in this case. Our group previously reported that djurleite Cu2-xS can be prepared by hot injection of oleic acid-S (OA-S) into OLAM-Cu precursor.22 The TEM image in figure 6a shows a representative example of 11.8±0.7 nm Cu2xS NPls prepared by this approach, including a few NPls that are stacked such that they are visible in side-view, confirming their “plate-like” shape. A similar protocol for Sn ion exchange to that used with the covellite NPls was applied to these Cu2-xS NPls. As described in more detail in part 5 of the SI, both relatively air stable Sn2+ and Sn4+ were employed in these experiments. As shown in Fig. S7 in the SI, Sn2+ shows no signs of incorporating into Cu2-xS. Rather, as indicated by both imaging and SAED, a secondary phase of larger, flowerlike structures was formed. This is not surprising, because the radius of Sn2+ (~1.18 Å) is much larger than that of Cu+ (~0.6 Å) and this difference inhibits the substitution of Cu in the lattice by Sn.20 Because the radius of Sn4+ is only ~0.55 Å, it may be better able to participate in the cation exchange reaction. The small size is similar to that of Cu+, allowing it to diffuse more readily in the crystal lattice. After exposing the Cu2-xS NPls to Sn4+, little evidence of formation of secondary phases, e.g., much bigger or smaller NCs, was observed and the NPls retained their initial size and shape. However, STEM HAADF mapping and EDX line profiles showed that very little Sn was present in the Cu2-xS NPls (Figure 6b-6c). Large-area EDS analysis in the SEM (Table S1 in SI) indicated a Sn content of 2.3 at.%. This modest Sn content could result from Sn4+ filling vacancies in the djurleite lattice without replacing any Cu cations or from Sn adsorbed at the NPl surface. The LSPR absorbance (Fig. S8) of the NPls showed little change after exposure to Sn4+, suggesting that relatively few cation vacancies were filled. Previously, Gao et al. reported that orthorhombic Cu3SnS4 nanocubes and nanosheets were prepared using monoclinic Cu31S16 nanocrystals as seeds.6 More broadly, several examples of cation exchange using copper chalcogenide templates (e.g. Cu2-xS32 and Cu2-xSe20) with DDT as a surfactant and copper extraction agent have been reported. Based on hard/soft acid base theory, the thiol in DDT, a soft base, can interact strongly with Cu+, which is a soft acid. The ability of the thiol to help extract Cu+ out of the framework allows Sn4+ to more readily replace it in the cation sublattice. The limited cation exchange observed here suggests that OLAM cannot serve to extract Cu+ from the lattice in a manner similar to DDT. II. Reaction with DDT DDT has been used as a capping ligand for many nanoparticle synthesis processes and, as discussed above, can enhance cation exchange in copper chalcogenides by extracting copper from the lattice. Here, we performed the same experiments described above, using either Sn2+ or Sn4+, with the addition of DDT to test the hypothesis that DDT drives cation exchange

in this system. Figure 7a-7b show a representative TEM image and XRD pattern of the product of treating covellite NPls with Sn4+ in the presence of DDT. The TEM image shows that the original size and shape of the NPls was well-preserved, and no separate phases can be found. The corresponding XRD pattern shows that the NPls have been converted to the mohite phase of CTS (ICDD PDF 04-010-5719), for which the standard stoichiometry is Cu2SnS3. This demonstrates that the initial covellite NPls can be converted to two different CTS materials using different combinations of Sn sources and additives. Without DDT, Sn2+ addition produced kuramite, and with DDT, Sn4+ produced mohite, while in both cases the NPl size and morphology was preserved. When Sn2+ was used with DDT, as shown in the TEM images of S9 (a-b) in SI, NPs of a secondary phase were formed, and the corresponding XRD pattern showed the presence of both covellite and herzenbergite SnS. This may be explained by greater reactivity of the thiol with the metal cations, compared to the amine and the ability of DDT to reduce disulfide bonds. The combined ability of DDT to extract Cu and to reduce disulfide bonds reversed the dependence of Sn-incorporation on the valency of the Sn source.

Figure 6. (a) TEM images of djurleite NCs. (b) STEM HAADF image of Sn4+ treated djurleite NCs. (c) STEM EDS linear mapping along the line indicated in (b). The scale bar applies to both (a) and (b).

Discussion The results presented above show that Sn ions of different valences can incorporate into copper sulfide of different stoichiometry and crystal structure, and that this incorporation is influenced by the addition of ligands that can extract Cu from Cu2-xS and that can reduce disulfide bonds in covellite. The impact of Sn incorporation on the free carrier concentration can be tracked by observing the red-shift and decreased intensity of LSPR absorbance.28 The covellite structure can be described by a repeat unit composed of a layer of triangular CuS3 units, sandwiched between two layers of CuS4 tetrahedra (see figure S10 in SI). Each triple layer is then linked to a top and bottom triple layer by disulfide bonds33. The disulfide bonds not only link the CuS triple layers, but also provide two free holes in the valence band per unit cell.34 Covellite thus behaves as a semi-metal with the highest hole concentration among the family of stable Cu2-xS (0 ≤ x ≤ 1) materials. The disulfide bonds also appear to play a key role in the transformation to from CuS to CTS. We propose that Sn2+ can reduce the disulfide bonds and can thus be simultaneously oxidized and incorporated into the covellite structure. Reduction of disulfide bonds decreases the concentration of free holes in the NPls. With the NPl geometry and other parameters fixed, the LSPR energy is approximately proportional to the square-root of the free hole concentration. Thus, reduction of the disulfide bonds produces a red-shift of the LSPR absorbance peak of the NPls. In contrast to Sn2+, Sn4+ ions cannot be further oxidized and therefore cannot reduce disulfide bonds. Addition of



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

Sn4+ to the CuS dispersion, which contains excess sulfur, thus leads to formation of SnS2 as a separate secondary phase rather than incorporation of Sn into the covellite host lattice. Conversely, when Sn2+ is used in the presence of DDT, the Sn2+ is not oxidized to Sn4+ and is not incorporated into the covellite lattice. Instead, it remains in its Sn2+ oxidation state and forms SnS.

Page 6 of 9

cause Cu+ is the only cation present. Other studies suggested the presence of both Cu+ and Cu2+ as (Cu2+)(S22−)(Cu+)2(S2−).3839 Comparing the XPS of kuramite NPls to that of the covellite NPls in our previous study21 suggests that the S is present entirely as S2-in the kuramite NPls, in contrast to its mixed valence in covellite. The observed oxidation states would imply a formula of (Cu+)3(Sn4+)(S2-)4, which does not reflect a balance between the cation and anion valences. The ambiguity of the oxidation states is similar to that of the parent covellite structure, and simply reflects the fact that the description of bonding in these structures, in which the bonds have substantial covalent character, is not perfectly described by strict oxidation states.

Figure 7. (a) TEM image and (b) XRD pattern of CTS NCs prepared using Sn4+ in the presence of DDT.

Covellite can be described as Cu6S6, in which 4 of the Cu ions have tetrahedral coordination and 2 have triangular coordination, while 4 of the S atoms form 2 disulfide bonds, and 2 S atoms are present as sulfide ions. The product of CuS treatment with Sn2+ is kuramite, with a tetragonal structure (space group I4m2). If the cations are fully ordered, then the 2a and 4d positions are occupied by Cu, while Sn is at the 2b position and S is at the 8i position5 (see figure 8). In this structure, all of the atoms are tetrahedrally coordinated. To convert covellite to kuramite, each entering Sn2+ ion must break a disulfide bond and migrate to a suitable positon. Breaking all of the disulfide bonds leads to formation of kuramite CTS NCs. The observed upper limit of tin incorporation of one Sn for every three Cu atoms (two Sn per unit cell) corresponds to the amount required to break the two disulfide bonds per unit cell. The color of the supernatant and the dispersions of NPls before and after conversion from covellite to kuramite (Fig. S11a-b) support the conclusion that the solid-state phase transformation occurred by reduction of disulfide bonds and inertion of Sn ions rather than cation exchange. Any significant concentration of Cu ions in the supernatant would produce a visible blue or green color, which was not observed. Manna’s group has similarly reported that S-S bonds are ruptured by electron donors and foreign metal cations.35 Note that conversion of covellite to kuramite requires incorporation of two additional S atoms per unit cell, along with the addition of two Sn atoms per unit cell, transforming the Cu6S6 unit cell to Cu6Sn2S8. XPS analysis of the kuramite NPls produced by addition of Sn2+ to covellite NPls revealed that copper was present solely as Cu+; no Cu2+ satellites were observed (see figure S12a-c). The Sn valency is predominantly Sn4+. This is consistent with the hypothesis that each Sn2+ provides two electrons to reduce a disulfide bond, while the Sn2+ itself is oxidized to Sn4+. The determination of the valence state of sulfur is more difficult than for copper and tin. The valency of sulfur in covellite remains the subject of debate. Fjellvag et al. found the copper in covellite to be monovalent and concluded that the oxidation states should be (Cu+)3(S22−)(S−).36 Liang et al. suggested that the valency should be (Cu+)3(S2−)(S2−).37 Both of these reports presume that the average valency of sulfur is equal to −1 be-

Figure 8. Schematic illustration of the structural transformation from covellite to kuramite.

When DDT was present, it could either reduce disulfide bonds directly or could extract Cu+ and then the Cu+ could reduce the disulfide bonds. EDX and XRD showed that CTS synthesized from covellite using a combination of DDT and Sn4+ adopted a different crystal phase than that prepared using Sn2+. These results not only show that this approach to producing CTS NPls is flexible and controllable, but also suggests that the transformation mechanisms in the two cases are significantly different. To investigate the effect of DDT, we consider the change of crystal structure upon Sn incorporation. Mohite Cu2SnS3 (with nominal oxidation states (Cu+)2(Sn4+)(S2-)3) can be described as Cu2.667Sn1.333S440. For comparison with covellite (Cu6S6), we can consider mohite as Cu8Sn4S12 and covellite as Cu12S12. Stoichiometrically, the transformation consists of replacing 4 Cu+ cations with 4 Sn4+ cations. In contrast with kuramite formation, in which breaking of disulfide bonds makes space for entering Sn, the formation of mohite involves replacement of Cu with Sn in the covellite framework. In the covellite lattice 1/3 of the Cu in the covellite structure have planar trigonal coordination, and the remaining 2/3 of the Cu has tetrahedral coordination. The more weakly bound trigonally-coordinated copper is presumably more easily extracted and therefore is most likely to be replaced by Sn.

Conclusion In this study, we observed that Sn incorporation into copper sulfide is strongly influenced by the crystal structure of the copper sulfide, the oxidation state of the Sn source, and the presence of an extracting ligand. We exploited this dependence to obtain two different phases of CTS from the same covellite template nanoplatelets (NPls) while retaining the NPl size and shape. Kuramite NPls can be obtained from covellite


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


NPls by utilizing Sn2+ alone. We showed that Sn2+ can reduce disulfide bonds in covellite and incorporate into the lattice without displacing copper cations. Alone, Sn4+ did not incorporate into covellite. However, use of Sn4+ together with DDT produced mohite NPls from covellite NPls. Using Sn2+ along with DDT did not lead to Sn incorporation into covellite, but instead produced a secondary phase (SnS). Results using djurleite Cu2-xS in place of covellite were completely different: use of Sn2+ without DDT led to secondary phase formation, while use of Sn4+ without DDT produced minimal Sn incorporation. This provides both new insights into formation mechanisms for CTS ternary nanostructures and a practical means of producing well-controlled CTS NPls of desired crystal phase.

ASSOCIATED CONTENT Supporting Information. Size distribution histograms for samples shown in Figs. 1, 2, 6, and 7. TEM images of CuS NPls treated at higher temperature. EDS analysis relating Sn incorporation to Sn/(Sn+Cu) in precursors for covellite NPls treated with Sn2+; XRD and TEM of NCs produced using excess Sn2+; UV-vis-NIR absorbance spectra and XPS spectra for Sn2+ treated CuS NCs TEM images for Sn4+ treated CuS NCs; TEM, EDS analysis, and NIR absorbance spectra of Sn4+-treated djurleite (Cu2-xS) NCs; TEM and XRD of CuS NCs treated with Sn2+ in the presence of DDT; crystal structures of covellite and kuramite; photographs of dispersions of covellite and kuramite NPls, the supernatant a of kuramite, and pure OLAM. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected].

ACKNOWLEDGMENT This work was supported in part by the New York State Center of Excellence in Materials Informatics.

REFERENCES 1. Sakamoto, M.; Chen, L. H.; Okano, M.; Tex, D. M.; Kanemitsu, Y.; Teranishi, T., Photoinduced Carrier Dynamics of Nearly Stoichiometric Oleylamine-Protected Copper Indium Sulfide Nanoparticles and Nanodisks. J Phys Chem C 2015, 119 (20), 11100-11105. 2. Wang, X.; Swihart, M. T., Controlling the Size, Shape, Phase, Band Gap, and Localized Surface Plasmon Resonance of Cu2–xS and CuxInyS Nanocrystals. Chem. Mater. 2015, 27 (5), 1786-1791. 3. Liu, X.; Wang, X.; Swihart, M. T., Composition-Dependent Crystal Phase, Optical Properties, and Self-Assembly of Cu–Sn–S Colloidal Nanocrystals. Chem. Mater. 2015, 27 (4), 1342-1348. 4. Su, Z. H.; Sun, K. W.; Han, Z. L.; Liu, F. Y.; Lai, Y. Q.; Li, J.; Liu, Y. X., Fabrication of ternary Cu-Sn-S sulfides by a modified successive ionic layer adsorption and reaction (SILAR) method. J Mater Chem 2012, 22 (32), 16346-16352. 5. Dzhagan, V. M.; Litvinchuk, A. P.; Kruszynska, M.; KolnyOlesiak, J.; Valakh, M. Y.; Zahn, D. R. T., Raman Scattering Study of Cu3SnS4 Colloidal Nanocrystals. J Phys Chem C 2014, 118 (47), 2755427558. 6. Yi, L.; Wang, D.; Gao, M., Synthesis of Cu3SnS4 nanocrystals and nanosheets by using Cu31S16 as seeds. CrystEngComm 2012, 14 (2), 401-404. 7. Goto, Y.; Sakai, Y.; Kamihara, Y., Electronic structure and transport properties of Cu-deficient kuramite Cu3−xSnS4. Jpn. J. Appl. Phys. 2015, 54 (2), 021801. 8. Panthani, M. G.; Akhavan, V.; Goodfellow, B.; Schmidtke, J. P.; Dunn, L.; Dodabalapur, A.; Barbara, P. F.; Korgel, B. A., Synthesis of CuInS2, CuInSe2, and Cu(InxGa1-x)Se2 (CIGS) Nanocrystal "Inks" for

Printable Photovoltaics. J Am Chem Soc 2008, 130 (49), 16770-16777. 9. Wu, X. J.; Huang, X.; Qi, X.; Li, H.; Li, B.; Zhang, H., Copper-based ternary and quaternary semiconductor nanoplates: templated synthesis, characterization, and photoelectrochemical properties. Angew Chem Int Ed Engl 2014, 53 (34), 8929-8933. 10. Yu, K.; Carter, E. A., A Strategy to Stabilize Kesterite CZTS for High-Performance Solar Cells. Chem Mater 2015, 27 (8), 2920-2927. 11. Wang, W.; Winkler, M. T.; Gunawan, O.; Gokmen, T.; Todorov, T. K.; Zhu, Y.; Mitzi, D. B., Device Characteristics of CZTSSe Thin-Film Solar Cells with 12.6% Efficiency. Adv Energy Mater 2014, 4 (7), 1301465. 12. Lu, X.; Zhuang, Z.; Peng, Q.; Li, Y., Wurtzite Cu2ZnSnS4 nanocrystals: a novel quaternary semiconductor. Chem Commun 2011, 47 (11), 3141-3143. 13. Zou, Y.; Su, X.; Jiang, J., Phase-Controlled Synthesis of Cu2ZnSnS4 Nanocrystals: The Role of Reactivity between Zn and S. J Am Chem Soc 2013, 135 (49), 18377-18384. 14. Tan, J. M. R.; Lee, Y. H.; Pedireddy, S.; Baikie, T.; Ling, X. Y.; Wong, L. H., Understanding the Synthetic Pathway of a Single-Phase Quarternary Semiconductor Using Surface-Enhanced Raman Scattering: A Case of Wurtzite Cu2ZnSnS4 Nanoparticles. J Am Chem Soc 2014, 136 (18), 6684-6692. 15. Yang, H. R.; Jauregui, L. A.; Zhang, G. Q.; Chen, Y. P.; Wu, Y., Nontoxic and Abundant Copper Zinc Tin Sulfide Nanocrystals for Potential High-Temperature Thermoelectric Energy Harvesting. Nano Lett 2012, 12 (2), 540-545. 16. Lin, X. Z.; Steigert, A.; Lux-Steiner, M. C.; Ennaoui, A., Onestep solution-based synthesis and characterization of kuramite Cu3SnS4 nanocrystals. RSC Adv 2012, 2 (26), 9798-9800. 17. Wang, X. L.; Liu, X.; Zhu, D. W.; Swihart, M. T., Controllable conversion of plasmonic Cu2-xS nanoparticles to Au2S by cation exchange and electron beam induced transformation of Cu2-xS-Au2S core/shell nanostructures. Nanoscale 2014, 6 (15), 8852-8857. 18. Lesnyak, V.; George, C.; Genovese, A.; Prato, M.; Casu, A.; Ayyappan, S.; Scarpellini, A.; Manna, L., Alloyed copper chalcogenide nanoplatelets via partial cation exchange reactions. ACS nano 2014, 8 (8), 8407-8418. 19. Sadtler, B.; Demchenko, D. O.; Zheng, H.; Hughes, S. M.; Merkle, M. G.; Dahmen, U.; Wang, L. W.; Alivisatos, A. P., Selective Facet Reactivity during Cation Exchange in Cadmium Sulfide Nanorods. J Am Chem Soc 2009, 131 (14), 5285-5293. 20. De Trizio, L.; Li, H.; Casu, A.; Genovese, A.; Sathya, A.; Messina, G. C.; Manna, L., Sn cation valency dependence in cation exchange reactions involving Cu2-xSe nanocrystals. J Am Chem Soc 2014, 136 (46), 16277-16284. 21. Wang, Y.; Chen, Z.; Shen, R. G.; Cao, X.; Chen, Y. G.; Chen, C.; Wang, D. S.; Peng, Q.; Li, Y. D., Pd-dispersed CuS hetero-nanoplates for selective hydrogenation of phenylacetylene. Nano Res 2016, 9 (4), 1209-1219. 22. Liu, X.; Wang, X. L.; Zhou, B.; Law, W. C.; Cartwright, A. N.; Swihart, M. T., Size-Controlled Synthesis of Cu2-xE (E = S, Se) Nanocrystals with Strong Tunable Near-Infrared Localized Surface Plasmon Resonance and High Conductivity in Thin Films. Adv Funct Mater 2013, 23 (10), 1256-1264. 23. Liu, M. X.; Xue, X. Z.; Ghosh, C.; Liu, X.; Liu, Y.; Furlani, E. P.; Swihart, M. T.; Prasad, P. N., Room-Temperature Synthesis of Covellite Nanoplatelets with Broadly Tunable Localized Surface Plasmon Resonance. Chem Mater 2015, 27 (7), 2584-2590. 24. Thanh, N. T. K.; Maclean, N.; Mahiddine, S., Mechanisms of Nucleation and Growth of Nanoparticles in Solution. Chem Rev 2014, 114 (15), 7610-7630. 25. Salavati-Niasari, M.; Ghanbari, D.; Davar, F., Shape selective hydrothermal synthesis of tin sulfide nanoflowers based on nanosheets in the presence of thioglycolic acid. J Alloy Compd 2010, 492 (1-2), 570-575. 26. Yousefi, M.; Salavati-Niasari, M.; Gholamian, F.; Ghanbari, D.; Aminifazl, A., Polymeric nanocomposite materials: Synthesis and thermal degradation of acrylonitrile-butadiene-styrene/tin sulfide (ABS/SnS). Inorg Chim Acta 2011, 371 (1), 1-5. 27. Zou, Y. Q.; Wang, Y., Microwave solvothermal synthesis of flower-like SnS2 and SnO2 nanostructures as high-rate anodes for lithium ion batteries. Chem Eng J 2013, 229, 183-189. 28. Luther, J. M.; Jain, P. K.; Ewers, T.; Alivisatos, A. P., Localized surface plasmon resonances arising from free carriers in doped quantum dots. Nat Mater 2011, 10 (5), 361-366. 29. Wagner, J. B.; Wagner, C., Investigations on Cuprous Sulfide.



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

The Journal of Chemical Physics 1957, 26 (6), 1602-1606. 30. Ku, G.; Zhou, M.; Song, S.; Huang, Q.; Hazle, J.; Li, C., Copper Sulfide Nanoparticles As a New Class of Photoacoustic Contrast Agent for Deep Tissue Imaging at 1064 nm. ACS Nano 2012, 6 (8), 74897496. 31. Zhao, Y.; Pan, H.; Lou, Y.; Qiu, X.; Zhu, J.; Burda, C., Plasmonic Cu2-xS nanocrystals: optical and structural properties of copperdeficient copper (I) sulfides. J Am Chem Soc 2009, 131 (12), 4253-4261. 32. Akkerman, Q. A.; Genovese, A.; George, C.; Prato, M.; Moreels, I.; Casu, A.; Marras, S.; Curcio, A.; Scarpellini, A.; Pellegrino, T.; Manna, L.; Lesnyak, V., From Binary Cu2S to ternary Cu-In-S and quaternary Cu-In-Zn-S nanocrystals with tunable composition via partial cation exchange. ACS Nano 2015, 9 (1), 521-531. 33. Xie, Y.; Riedinger, A.; Prato, M.; Casu, A.; Genovese, A.; Guardia, P.; Sottini, S.; Sangregorio, C.; Miszta, K.; Ghosh, S.; Pellegrino, T.; Manna, L., Copper Sulfide Nanocrystals with Tunable Composition by Reduction of Covellite Nanocrystals with Cu+ Ions. J Am Chem Soc 2013, 135 (46), 17630-17637. 34. Conejeros, S.; Moreira Ide, P.; Alemany, P.; Canadell, E., Nature of holes, oxidation states, and hypervalency in covellite (CuS). Inorg Chem 2014, 53 (23), 12402-6. 35. Xie, Y.; Bertoni, G.; Riedinger, A.; Sathya, A.; Prato, M.;

Page 8 of 9

Marras, S.; Tu, R. Y.; Pellegrino, T.; Manna, L., Nanoscale Transformations in Covellite (CuS) Nanocrystals in the Presence of Divalent Metal Cations in a Mild Reducing Environment. Chem Mater 2015, 27 (21), 7531-7537. 36. Fjellvag, H.; Gronvold, F.; Stolen, S.; Andresen, A. F.; Mullerkafer, R.; Simon, A., Low-Temperature Structural Distortion in Cus. Z Kristallogr 1988, 184 (1-2), 111-121. 37. Liang, W.; Whangbo, M. H., Conductivity Anisotropy and Structural Phase-Transition in Covellite Cus. Solid State Commun 1993, 85 (5), 405-408. 38. Grioni, M.; Goedkoop, J. B.; Schoorl, R.; de Groot, F. M.; Fuggle, J. C.; Schafers, F.; Koch, E. E.; Rossi, G.; Esteva, J.; Karnatak, R. C., Studies of copper valence states with Cu L3 X-ray-absorption spectroscopy. Phys Rev B Condens Matter 1989, 39 (3), 1541-1545. 39. Kumar, P.; Nagarajan, R.; Sarangi, R., Quantitative X-ray Absorption and Emission Spectroscopies: Electronic Structure Elucidation of CuS and Cu2S. J Mater Chem C Mater Opt Electron Devices 2013, 1 (13), 2448-2454. 40. Chen, X.; Wada, H.; Sato, A.; Mieno, M., Synthesis, electrical conductivity, and crystal structure of Cu4Sn7S16and structure refinement of Cu2SnS3. J. Solid State Chem. 1998, 139 (1), 144-151.


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


Insert Table of Contents artwork here


9

Kuramite Cu3SnS4 and Mohite Cu2SnS3 Nanoplatelet Synthesis

Recommend Documents