Modulation of Defects in Semiconductors by Facile ... - ACS Publications

Article pubs.acs.org/IC

Modulation of Defects in Semiconductors by Facile and Controllable Reduction: The Case of p‑type CuCrO2 Nanoparticles Tengfei Jiang,* Xueyan Li, Martine Bujoli-Doeuff, Eric Gautron, Laurent Cario, Stéphane Jobic,* and Romain Gautier* Institut des Matériaux Jean Rouxel (IMN), Université de Nantes, CNRS, 2 rue de la Houssinière, BP 32229, 44322 Nantes, Cedex 03, France S Supporting Information *

ABSTRACT: Optical and electrical characteristics of solid materials are well-known to be intimately related to the presence of intrinsic or extrinsic defects. Hence, the control of defects in semiconductors is of great importance to achieve specific properties, for example, transparency and conductivity. Herein, a facile and controllable reduction method for modulating the defects is proposed and used for the case of p-type delafossite CuCrO2 nanoparticles. The optical absorption in the infrared region of the CuCrO2 material can then be fine-tuned via the continuous reduction of nonstoichiometric CuII, naturally stabilized in small amounts. This reduction modifies the concentration of positive charge carriers in the material, and thus the conductive and reflective properties, as well as the flat band potential. Indeed, this controllable reduction methodology provides a novel strategy to modulate the (opto-) electronic characteristics of semiconductors.

1. INTRODUCTION In solid-state materials, defects interrupt the regular arrangement of atoms, and thus induce deviations in the chemical formula, change in local environment of chemicals, etc. All these factors will perturb the ideal electronic structure with creation of localized levels within the gap, change in the dispersion of the energy bands in the k-space, and furthermore strongly impact the physical properties of the material.1 At very first sight, it could be deemed that defects will deteriorate the physical characteristics of an ideal compound. On the contrary, the defects are often beneficial, and the interest of the material’s applications can be directly correlated to their existence in a controlled concentration and state. For example, extrinsic and intrinsic defects are often associated with the luminescence of solids.2 In a similar way, acceptors and donors levels, introduced via appropriate dopants, determine the specific electronic properties of semiconductors.3 These comments can be expanded to many other domains such as catalysis,4 sensors,5 mechanics, solar energy conversion,6 etc. Thus, constructing and engineering the defects in solid-state materials is of great importance. Our strategy developed here consists in the modulation of defects by capturing of oxygen via an oxygen getter (e.g., CaH2) at moderate temperature under dynamic vacuum. More importantly, the CaH2 can work at low temperature, which is beneficial for maintaining the crystal structure of the materials. To the best of our knowledge, there is no investigation on tuning the defects of transparent conducting oxides (TCO) by this method. In this communication, the soft reduction method © XXXX American Chemical Society

is used to control the defects and therefore the evolution of the optical and electronic properties, and flat band potentials of CuCrO2 nanoparticles are investigated when the reduction temperature changes. CuCrO2 is a p-type semiconductor with a delafossite structure, which is a layered material built upon infinite [CrO2] layers of [CrO6] octahedral interspaced by Cu in oxygen linear coordination. This material can be regarded as a TCO, especially when doped with Mg,7,8 with potential applications for optoelectronics, catalysis,9 and p-type dyesensitized solar cell (p-DSSC),10−13 etc. Delafossite materials can crystallize in the 2H or 3R allotropic forms that differ only in the stacking of the layers (Figure S1 in Supporting Information).14 From ab initio calculations,15,16 the electronic structure of the stoichiometric CuCrO2 is illustrated in Figure 1. The top of valence band originates from Cu 3d states with O 2p orbitals centered at lower energy, and the bottom of conduction band is mainly built upon 4s and 4p orbitals of Cu and Cr. The ground state of Cr3+ (4A2g, t2g3 configuration) is within the valence band, while the excited states (4T2g and 4T1g, t2g2eg1 configuration) are interspaced between the valence and conduction bands. Thus, the green color could be associated with 4A2g → 4T2g (2.1 eV) and 4A2g → 4T1g (2.8 eV) transitions at Cr sites.17 Such an assignment would be in agreement with Cr−O interatomic distances, which are almost identical in CuCrO2 (1.987−1.989 Å) and Cr2O3 (1.983 Å, a well-known Received: May 13, 2016

A

DOI: 10.1021/acs.inorgchem.6b01169 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Characterization. The X-ray diffraction (XRD) patterns were characterized by a D8-Bruker X-ray diffractometer with Cu Kα radiation (λ = 0.154 18 nm). Transmission electron microscopy (TEM) observations, as well as electronic diffraction patterns, were obtained with an H9000NAR-Hitachi electron microscope working at 300 kV and room temperature. The UV−vis reflectance spectra were recorded on a Lambda 1050 PerkinElmer spectrophotometer. X-ray photoelectron spectroscopy was performed in Kratos Nova X-ray photoelectron spectrometer. The Fourier transform infrared spectra (FTIR) were performed on a Bruker Vertex 70 spectrometer. Electrochemical Impedance Spectroscopy Measurements. CuCrO2 powder was pressed under 100 bar to generate a hard pellet. Carbon paste was painted on one side to make the back contact between CuCrO2 pellet and copper wire. Then, the pellet was embedded in resistant epoxy glue. Before electrochemical impedance spectroscopy (EIS) measurement, the surface of CuCrO2 pellet was polished by SiC paper to get a mirrorlike smooth surface. EIS measurements were performed with an electrochemical workstation (SP-300, Biologic Sciences Instruments). The electrolyte is 1 M LiClO4 aqueous solution (pH ≈ 9.4), and a platinum electrode and a saturated calomel electrode (SCE) served as counter and reference electrodes, respectively. The impedance spectra were performed under dark conditions with alternating current amplitude of 10 mV.

Figure 1. Schematized electronic structure of a hypothetical stoichiometric CuCrO2 material. (red ○) Holes at the top of the valence band.

green pigment). In both phases, Cu 3d orbitals strongly hybridize with O 2p orbitals, which facilitates the migration of holes resulting from a slight Cu off-stoichiometry or a slight O overstoichiometry.18 Because of the nonstoichiometry, a slight amount of CuII in CuCrO2 served as hole source and causes the p-type conductivity. The presence of Cr4+ instead of Cu2+ as charge compensator has been definitely rebutted on the basis of band structure calculations.19,20 If the concentration of CuII is large enough, a strong absorption in the visible and infrared (IR) range can be expected due to intraband transitions resulting in a black color with a plasma frequency that shifts toward high energy as the square root of the hole density. Clearly, transparency and conductivity appear as antagonist and directly related to the hole concentration in the material. Thus, modulating the amount of defects in CuCrO2 to find a balance between the transparency and p-type conductivity, two antagonist properties, becomes of major interest. First, CuCrO2 samples were prepared by a modified hydrothermal method of Xiong et al.11 Second, reduction of the as-obtained materials was initiated in a pyrex tube under dynamic vacuum with CaH2 as reducer at the hot point. Practically, the hydride was introduced in a glass vessel wellseparated from the CuCrO2 sample to avoid contact and hinder pollution (more details in Supporting Information). Compared with other methods, the interest of a getter resides in its capacity to trigger/accelerate the release of oxygen due to the formation of highly stable CaO or Ca(OH)2 phases that drive the reduction.21 Herein, relatively low temperatures were selected (T ≤ 320 °C) to preserve the crystal structure of the pristine materials and to make the gradual release of oxygen more controllable.21

3. RESULTS AND DISCUSSION Prepared materials were analyzed by XRD (Figure 2). The sample prior to reduction contains both 2H (space group:

Figure 2. XRD pattern of CuCrO2 nanoparticles before and after reduction at 250, 300, and 320 °C for 2.5 h.

P63/mmc, major phase) phases and 3R (space group: R3̅m, minor phase). This result agrees well with previous investigations of Dirk et al.23 reporting that a mixture of 3R and 2H phases was systematically obtained by using an excess of NaOH during the hydrothermal reaction. In contrast, our result is not consistent with those of Xiong et al., who claimed that the pure 3R phase could be obtained.11 XRD patterns of all reduced samples match perfectly with that of pristine CuCrO2. This indicates that the reduction process did not affect the crystal structures of CuCrO2 as well as the relative 2H/3R ratio. Moreover, no additional diffraction peaks assigned to hydroxides were detected suggesting that the CuCrO2 samples were free of impurities. FTIR measurements on the samples (Figure S2 in Supporting Information) led to the same conclusion. Selected area electron diffraction (SAED) and highresolution transmission electron microscopy (HRTEM) experiments were performed on CuCrO2 samples prior to and after reduction at 320 °C. As expected, most of the diffraction rings can be ascribed to the CuCrO2 2H phase (Figure 3a,d). However, very few additional diffraction spots show the presence of CuCrO2 3R phase (Figure S3 in Supporting

2. EXPERIMENTAL SECTION Synthesis of CuCrO2 Nanoparticles. The CuCrO2 nanoparticles were prepared by hydrothermal method.22 Typically, 6 mmol of Cu(NO3)2·3H2O and 50 mmol of NaOH were dissolved in 20 mL of H2O, with magnetic stirring, followed addition of 6 mmol of Cr(NO3)2·9H2O. The mix solution was stirred for 1 h, transferred into a 25 mL Teflon-lined autoclave, and maintained at 240 °C for 12 h. Then, the obtained precipitate was dispersed in 1 M hydrochloric acid for 12 h with stirring to remove the subproduct. The black precipitate was harvested by centrifugation at 12 000 rpm for 20 min, followed by washing three times with water and ethanol. After the washes, the precipitate was dried at 70 °C for 6 h. Controllable Reduction of CuCrO2 Nanoparticles. A relatively mild reduction method is used, which is operating at low temperature by using CaH2.21 In a glovebox, a glass tube was filled with 300 mg of CaH2. Then three vials with 150 mg of CuCrO2 each were inserted. The tube, maintained under vacuum during the whole reduction process, was heated for 2.5 h. CaH2 getter was located in the hot zone at 320 °C. CuCrO2 vials were placed along the tube, and their temperature was measured at 320, 300, and 250 °C. B

DOI: 10.1021/acs.inorgchem.6b01169 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 3. SAED patterns, TEM, and HRTEM micrographs of CuCrO2 nanoparticles before (a−c) and after reduction at 320 °C (d−f). For more clarity, only the most intense rings of SAED patterns are indexed with 2H phase.

reflectance spectra is the very broad absorption band in the IR region that extends up to ∼700 nm (1.77 eV). This band would be associated with free positive charge carriers. The higher the reduction temperature, the lower the concentration of CuII will be, and thus the lower intensity of this IR band. This evolution agrees with the color change of CuCrO2 nanoparticles that shifts continuously from black to dark green, and it suggests a decrease of the plasma frequency (also its associated absorption) going from the unreduced to the reduced CuCrO2 material. Hence, the soft reduction strategy did not impact the crystal structure of CuCrO2 but modified the chemical composition and the charge balance; that is, CuII (3d9) cations were progressively reduced into CuI (3d10), which changed the concentration of holes. To assert this assumption, X-ray photoelectron spectroscopy (XPS) measurements were performed (Figure 5). No position change of the Cr 2p1/2 and Cr 2p3/2 peaks located at 586.0 and 576.0 eV, respectively, is observed along the whole series (Figure 5a). Thus, Cr3+ is maintained during the entire reduction process. Two distinguishable signals can be collected for Cu 2p1/2 and Cu 2p3/2, that is, 951.5 and 952.5 eV and 931.9 and 932.9 eV, assigned to Cu+ and Cu2+ cations in large and small amounts, respectively (Figure 5b). Quantification of the Cu2+/Cu+ ratio turns out to be intricate owing to the low signal of CuII. Nevertheless, we may easily speculate that this ratio decreases when CuCrO2 is reduced. From the O 1s XPS (Figure 5c), three contributions at 529.5, 531.8, and 533.5 eV can be distinguished after Gaussian fitting. The lowest binding energy at 529.5 eV is attributed to the lattice oxygen in CuCrO2. The peaks at 531.8 and 533.5 eV might be related, respectively, to hydroxyl and carbonate groups that passivate the surface. 26,27 The former would be related to the nanostructuration of the CuCrO2 materials that strongly favors the presence of OH− groups. The second would originate from the basic conditions used to prepare the material. On the basis of a rich literature on surface defects, we may also suggest that the peak at 531.8 eV corresponds to the oxygen defects (insertion of O2− ions herein).28−31 Thus, the O 1s peak at

Information). As displayed in Figure 3b,e nanoparticles (with sides of ∼10 nm) and nanorods (∼5 nm × 20 nm) can be observed for both samples. They correspond to CuCrO2 crystals with their stacking axis perpendicular and parallel to the electron beam, respectively. In Figure 3c,f, the 0.57 nm observed d-spacing perfectly agrees with the interlayer distances in the delafossite structure. It is worth mentioning that particle dimensions are not modified by the reducing treatment. The reflectance spectra of CuCrO2 nanoparticles are given in Figure 4. The absorption threshold of CuCrO2 samples appears

Figure 4. (a) Transformed Kubelka−Munk reflectance spectra of CuCrO2 nanoparticles before and after reduction. (b) Photographs of CuCrO2 nanoparticles before and after reduction at 320 °C.

approximately at 400 nm, which agrees with the commonly reported value of 3.2 eV.24 A careful examination also suggests the existence of three absorption bands at 440 nm (2.8 eV), 590 nm (2.1 eV), and 710 nm (1.7 eV). The first two bands could be assigned to Cr3+ in octahedral coordination as discussed above. The last one may originate from Cr3+ in illenvironment (e.g., [CrO5] polyhedra instead of [CrO6]), since this band is indeed better defined once reduction took place. This band could also be associated with Cu2+ on-site transitions as suggested by Shin et al.25 The most attractive feature of the C

DOI: 10.1021/acs.inorgchem.6b01169 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 5. XPS spectra of CuCrO2 nanoparticles before and after reduction at different temperatures. (a) Cr 2p1/2 and 2p3/2; (b) Cu 2p1/2 and 2p3/2; (c) O 1s.

533.5 eV would be assigned to OH− species. This would account for the strong decrease in intensity of these two peaks compared to the one at 529.5 eV under reductive conditions (Table S1 in Supporting Information). On the basis of this speculation, the real chemical formulas of CuCrO2 could be written as Cu1+1−xCu2+xCr3+O2+x/2; that is, an oxygen overstoichiometry would be privileged as suggested in the literature.32 The deinsertion of oxygen from the CuCrO2 host lattice is expected to decrease the concentration of excess hole concentration at the top of the valence band, which will change the position of the Fermi level. To prove this speculation, capacitance measurements were performed on this material in which the flat band potentials (Vfb) were determined from Mott−Schottky plots (more details in Supporting Information). The evolution of the flat band potential and its implications on the electronic band structure is given in Figure 6. The Vfb shifts from 0.50 (vs SCE) to 0.07 V going from the unreduced sample to the reduced one at 320 °C (Figure S4 in Supporting Information). Under soft reductive conditions, the concentration of Cu2+ can decrease in a controllable way. Consequently, depleted levels at the top of the valence band progressively disappear, and the Fermi level is raised in an impressive 0.43 V range, which is a direct consequence of the concentration variation of positive charge carriers resulting also in the color change. The carrier concentration decreases from 3.94 × 1018 cm−3 for CuCrO2 reduced at 250 °C to 1.48 × 1017 cm−3 for CuCrO2 reduced at 320 °C (see Table S2). The sample before reduction is out of this trend, and the reason is that the carrier concentration is dependent on the surface area of the electrode, which is inaccurate for nanoparticles-pressed pellets with different roughness. Nevertheless, we may conclude that the carrier concentration decreases when CuCrO2 is reduced, since the Vfb is independent of the surface area of the electrode. As a consequence, the topochemical reduction protocol with oxygen getters can be viewed as an appropriate method to finely tune

Figure 6. (a) Flat band potential of CuCrO2 nanoparticles before and after reduction at different temperature. (b) Schematic illustration of the Fermi level (EF) shift.

the chemical potential of solid materials. This may really open the door to the adjustment of a material for applications as (photo)catalyst, sensor, etc. The strong sensitivity of the Vfb of delafossite to the reducing atmosphere also might partially explain why the measured open-circuit voltage (Voc) in p-DSSC is commonly lower than the difference between the reduction potential of the redox mediator and flat band potential of the ptype semiconductor. Practically, the thermal treatment of the photocathode might modify the chemical composition of the ptype semiconductor, which decreases its flat band potential, and therefore lead to a lower Voc than the expected one.

4. CONCLUSION In summary, we presented here a general method to modulate the amount of defects by a facile and controllable reduction D

DOI: 10.1021/acs.inorgchem.6b01169 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

(16) Ma, Y.; Zhou, X.; Ma, Q.; Litke, A.; Liu, P.; Zhang, Y.; Li, C.; Hensen, E. M. Catal. Lett. 2014, 144, 1487−1493. (17) Srinivasan, R.; Bolloju, S. AIP Conf. Proc. 2014, 1576, 205−208 (DOI: 10.1063/1.4862021). (18) Kawazoe, H.; Yasukawa, M.; Hyodo, H.; Kurita, M.; Yanagi, H.; Hosono, H. Nature 1997, 389, 939−942. (19) Scanlon, D. O.; Watson, G. W. J. Mater. Chem. 2011, 21, 3655. (20) Okuda, T.; Jufuku, N.; Hidaka, S.; Terada, N. Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72, 144403. (21) Kaur Behrh, G.; Serier-Brault, H.; Jobic, S.; Gautier, R. Angew. Chem. 2015, 127, 11663−11665. (22) Xiong, D.; Xu, Z.; Zeng, X.; Zhang, W.; Chen, W.; Xu, X.; Wang, M.; Cheng, Y.-B. J. Mater. Chem. 2012, 22, 24760−24768. (23) Friedrich, D. Am. J. Nano Res. Appl. 2014, 2, 53−60. (24) Li, D.; Fang, X.; Deng, Z.; Zhou, S.; Tao, R.; Dong, W.; Wang, T.; Zhao, Y.; Meng, G.; Zhu, X. J. Phys. D: Appl. Phys. 2007, 40, 4910. (25) Shin, D.; Foord, J. S.; Egdell, R. G.; Walsh, A. J. Appl. Phys. 2012, 112, 113718. (26) Stoch, J.; Gablankowska-Kukucz, J. Surf. Interface Anal. 1991, 17, 165−167. (27) Cano, E.; Torres, C. L.; Bastidas, J. M. Mater. Corros. 2001, 52, 667. (28) Wang, J.; Wang, Z.; Huang, B.; Ma, Y.; Liu, Y.; Qin, X.; Zhang, X.; Dai, Y. ACS Appl. Mater. Interfaces 2012, 4, 4024−4030. (29) Lei, F.; Sun, Y.; Liu, K.; Gao, S.; Liang, L.; Pan, B.; Xie, Y. J. Am. Chem. Soc. 2014, 136, 6826−6829. (30) Zhang, X.; Qin, J.; Xue, Y.; Yu, P.; Zhang, B.; Wang, L.; Liu, R. Sci. Rep. 2014, 4, 1. (31) Chu, D.; Younis, A.; Li, S. J. Phys. D: Appl. Phys. 2012, 45, 355306. (32) Ma, Y.; Zhou, X.; Ma, Q.; Litke, A.; Liu, P.; Zhang, Y.; Li, C.; Hensen, E. J. M. Catal. Lett. 2014, 144, 1487−1493.

route. As an example for p-type CuCrO2 nanoparticles, the color changes progressively from black to dark green owing to the reduction of CuII (3d9) to CuI (3d10). Meanwhile, the flat band potentials of CuCrO2 nanoparticles change from 0.50 to 0.07 V versus SCE because of the decreasing concentration of excess holes. Thus, the topotactic reduction method provides a promising opportunity to modulate the defects and further the opto-electrical properties of materials (including thin films).

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b01169. Schemes of synthesis and reduction process, structure of CuCrO2material, FTIR spectra, SAED patterns, O 1s XPS area ratio, Mott−Schottky plots, and table of flat band potential and carrier concentration. (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. (T.J.) *E-mail: [email protected]. (S.J.) *E-mail: [email protected]. (R.G.) Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS For financial support, T.J., M.B.-D., L.C., and S.J. are grateful to the ANR program, POSITIF (No. ANR-12-PRGE-0016-01). REFERENCES

(1) Pacchioni, G. ChemPhysChem 2003, 4, 1041−1047. (2) Zhang, C.; Lin, J. Chem. Soc. Rev. 2012, 41, 7938−7961. (3) Sze, S. M.; Ng, K. K. In Physics of Semiconductor Devices; John Wiley & Sons, Inc, 2006; pp i−x. (4) Polarz, S.; Strunk, J.; Ischenko, V.; van den Berg, M. W. E.; Hinrichsen, O.; Muhler, M.; Driess, M. Angew. Chem., Int. Ed. 2006, 45, 2965−2969. (5) Epifani, M.; Prades, J. D.; Comini, E.; Pellicer, E.; Avella, M.; Siciliano, P.; Faglia, G.; Cirera, A.; Scotti, R.; Morazzoni, F.; Morante, J. R. J. Phys. Chem. C 2008, 112, 19540−19546. (6) Zuo, F.; Wang, L.; Wu, T.; Zhang, Z.; Borchardt, D.; Feng, P. J. Am. Chem. Soc. 2010, 132, 11856−11857. (7) Han, M. J.; Duan, Z. H.; Zhang, J. Z.; Zhang, S.; Li, Y. W.; Hu, Z. G.; Chu, J. H. J. Appl. Phys. 2013, 114, 163526. (8) Barnabe, A.; Thimont, Y.; Lalanne, M.; Presmanes, L.; Tailhades, P. J. Mater. Chem. C 2015, 3, 6012−6024. (9) Amrute, A. P.; Larrazábal, G. O.; Mondelli, C.; Pérez-Ramírez, J. Angew. Chem., Int. Ed. 2013, 52, 9772−9775. (10) Xu, X.; Zhang, B.; Cui, J.; Xiong, D.; Shen, Y.; Chen, W.; Sun, L.; Cheng, Y.; Wang, M. Nanoscale 2013, 5, 7963. (11) Xiong, D.; Xu, Z.; Zeng, X.; Zhang, W.; Chen, W.; Xu, X.; Wang, M.; Cheng, Y.-B. J. Mater. Chem. 2012, 22, 24760−24768. (12) Xiong, D.; Zhang, W.; Zeng, X.; Xu, Z.; Chen, W.; Cui, J.; Wang, M.; Sun, L.; Cheng, Y.-B. ChemSusChem 2013, 6, 1432−1437. (13) Powar, S.; Xiong, D.; Daeneke, T.; Ma, M. T.; Gupta, A.; Lee, G.; Makuta, S.; Tachibana, Y.; Chen, W.; Spiccia, L.; Cheng, Y.-B.; Götz, G.; Bäuerle, P.; Bach, U. J. Phys. Chem. C 2014, 118, 16375− 16379. (14) Marquardt, M. A.; Ashmore, N. A.; Cann, D. P. Thin Solid Films 2006, 496, 146−156. (15) Scanlon, D. O.; Watson, G. W. J. Mater. Chem. 2011, 21, 3655− 3663. E

DOI: 10.1021/acs.inorgchem.6b01169 Inorg. Chem. XXXX, XXX, XXX−XXX