full papers Quantum Dots

Bandgap Tuning with Thermal Residual Stresses Induced in a Quantum Dot Eui-Hyun Kong, Soo-Hyun Joo, Hyun-Jin Park, Seungwoo Song, Yong-June Chang, Hyoung Seop Kim, and Hyun Myung Jang*

Lattice distortion induced by residual stresses can alter electronic and mechanical properties of materials significantly. Herein, a novel way of the bandgap tuning in a quantum dot (QD) by lattice distortion is presented using 4-nm-sized CdS QDs grown on a TiO2 particle as an application example. The bandgap tuning (from 2.74 eV to 2.49 eV) of a CdS QD is achieved by suitably adjusting the degree of lattice distortion in a QD via the tensile residual stresses which arise from the difference in thermal expansion coefficients between CdS and TiO2. The idea of bandgap tuning is then applied to QD-sensitized solar cells, achieving ≈60% increase in the power conversion efficiency by controlling the degree of thermal residual stress. Since the present methodology is not limited to a specific QD system, it will potentially pave a way to unexplored quantum effects in various QD-based applications.

1. Introduction A bulk material should have specific properties regardless of its size, but size-dependent properties are more often than not observed in nanoscale. For example, lower dislocation contents and reduced mobility of dislocations in smaller particles could result in pinned dislocations due to the enhanced dislocation-grain boundary interaction,[1] which helps the yield strength to approach the ideal value of the perfect single crystal.[1–3] Greater surface tension of a nanocrystallite compared to bulk materials results from weak dilation with the increase in surface area,[4] which tends to decrease the melting Dr. E.-H. Kong,[+] S.-H. Joo, S. Song, Dr. Y.-J. Chang, Prof. H. S. Kim, Prof. H. M. Jang Department of Materials Science and Engineering and Division of Advanced Materials Science Pohang University of Science and Technology (POSTECH) Pohang 790–784, Republic of Korea E-mail: [email protected] H.-J. Park National Institute for Nanomaterials Technology (NINT) Pohang University of Science and Technology (POSTECH) Pohang 790–784, Republic of Korea [+] Present address: Frontier Development Group, Samsung Cheil Industries, Samsung Materials Research Complex (SMRC), Suwon 443-803, Republic of Korea DOI: 10.1002/smll.201400392

3678

wileyonlinelibrary.com

point of materials.[4,5] In addition, size-dependent features of the thermal conductivity[6,7] are attributed to the change in the density of phonon states and phonon-boundary scattering.[8,9] Perhaps, the most prominent example of the size-dependent properties is the increase in the optical bandgap of a semiconductor with decreasing particle size including other factors (the number of defects,[10] the degree of oxidation[11] in QDs, etc.) owing to the quantum-confinement effect in nanoscale.[12] Residual stresses are those stresses that exist in a structure or component after all external loads have been removed. One of the known ways to improve the fatigue resistance and the ductility of a bulk metal is by using the shot peening process that does induce residual stresses.[13–15] The origin of residual stresses may be traced to thermal, mechanical, and chemical causes. In this paper, residual stresses caused by thermal processing, such as quenching, were used to tune the bandgap of quantum dots. Thermal residual stresses arise due to differential thermal expansion causing non-uniformed plastic deformation in a homogeneous body[16] or due to the difference in the thermal expansion coefficient between heterogeneous materials[17] when materials are heated and then suddenly cooled. Normally, high residual stresses induced in a material with a low elastic modulus and high yield strength can cause a high residual elastic strain in the lattice. Type II-VI bulk semiconductors with a high elastic modulus (∼GPa) and low yield strength (∼MPa) thus induce very low residual stress, which is insufficient to deform the lattice.[18,19] In addition,

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

small 2014, 10, No. 18, 3678–3684

Bandgap Tuning with Thermal Residual Stresses Induced in a Quantum Dot

for larger particles composed of many atoms including the bulk, the lattice-distortion effect is relatively negligible owing to a smaller fraction of the constituent atoms located at the interfacial region. On the contrary, small nanometer-sized particles (105 dm3 mol−1 cm−1) was evaluated by adopting the Tauc model[29] using the optical properties presented in Figure 5, instead of using PL spectra. The Tauc model is repn resented by ahv ∝ κ (hv − E g ) , where hv is the photon energy, Eg is the optical bandgap energy, and k is the photon-energydependent constant. The power-law exponent, n, depends small 2014, 10, No. 18, 3678–3684

Figure 4. Computed band structures of CdS for three different (111) spacings. The band structure (a) with the (111) spacing of 0.334 nm corresponds to the chemically crystallized CdS QD prepared at 25 °C by the SILAR process. On the other hand, the band structures (b,c) correspond to CdS QD quenched at 500 °C and at 900 °C, respectively. The corresponding (111) spacings determined by TEM observations are 0.339 nm for (b) and 0.349 nm for (c). The insets of Figure 4a-c show schematic crystal structures of the cubic CdS with the following markings: green spheres for Cd atoms, yellow spheres for S atoms, a blue plane for the (111) plane, and a red arrow for the (111) spacing.

on the transition type, where n = 1/2 for a direct bandgap transition and n = 2 for an indirect bandgap transition. The absorption coefficient (α) can be determined by the transmittance. This Tauc model, which was originally proposed for

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.small-journal.com

3681

full papers

E.-H. Kong et al.

Figure 5. Optical properties and the Tauc plots of CdS/TiO2/FTO electrodes under different conditions of the heat-treatment. a) Absorbance, b) transmittance, c) reflectance spectra, and d) Tauc plots, illustrating the bandgap tuning of the CdS-based quantum dots.

single-component semiconductors,[29] has been successfully extended to multicomponent semiconducting systems.[11,30–32] Since CdS is a well-known direct bandgap semiconductor,[33] its optical bandgap can be determined by plotting (αhv)2 as a function of the photon energy (see Figure 5d for detail). The optical bandgap of pure TiO2 is estimated to be 3.22 eV according to the Tauc model (see the arrow in Figure 5d). This agrees well with the direct interband transition of anatase TiO2 at 3.2 eV.[34] The bandgap of the CdS QDs grown on the TiO2 photoanode is significantly dependent on the heat-treatment condition: from 2.74 eV for no calcination (25 °C) to 2.49 eV for quenching at 900 °C (Figure 5d). The quenched films show a direct transition at a significantly reduced energy, which seems to be closely related to the lattice distortion caused by thermal residual stresses in a given QD (Figure 3b,c). Although there exists a large discrepancy between the DFT values and the experimentally deduced Eg

values, both results show the same trend of the (111)-spacingdependent optical bandgap. It is well documented that standard DFT exchange-correlation functionals, such as the local density approximation (LDA) and the GGA exploited here, notoriously underestimate Eg of insulators owing to the existence of a derivative discontinuity in the Kohn-Sham energy with respect to the number of electrons.[35] The novel method of the bandgap tuning developed in the present study was then applied to quantum-dot-sensitized solar cells (QDSCs) as an application example. The photovoltaic experiments were conducted to evaluate the performance of QDSCs. For this purpose, nanocrystalline TiO2 films were prepared according to the cell-fabrication procedures described in the Experimental Section. Herein, we have applied the idea of the nm-scale lattice distortion in a CdS QD to the QDSCs. The photocurrent density–voltage (J–V) characteristics of the devices are presented in Figure 6a.

Figure 6. Photovoltaic responses of the CdS-QDSCs with CdS/TiO2 photoanodes under different conditions of the heat-treatment. a) J–V characteristics and b) IPCE curves.

3682 www.small-journal.com

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

small 2014, 10, No. 18, 3678–3684

Bandgap Tuning with Thermal Residual Stresses Induced in a Quantum Dot

Table 1. Summarized photovoltaic parameters of the CdS-QDSCs with CdS-based QDs quenched at different temperatures. Jsc [mA cm−2]

Voc [mV]

FF [%]

η [%]

Area [cm2]

25 °C

5.6 ± 0.15

484 ± 1.6

40.7 ± 1.01

1.1 ± 0.02

0.223

500 °C

6.9 ± 0.23

501 ± 1.2

39.4 ± 1.14

1.4 ± 0.01

0.231

700 °C

7.9 ± 0.20

546 ± 1.1

39.3 ± 1.02

1.7 ± 0.03

0.234

900 °C

8.2 ± 0.09

559 ± 1.2

38.3 ± 0.92

1.8 ± 0.02

0.242

In addition, the detailed photovoltaic parameters, that is, open-circuit voltage (Voc), fill factor (FF), short-circuit photocurrent density (Jsc), and photovoltaic conversion efficiency (η) for the devices, are summarized in Table 1. A noticeable enhancement in Jsc is obtained by the quenching at a high temperature (≥500 °C; Figure 6a), which stems mainly from the improved absorbance values. As shown in Figure 6b, the quantum efficiency (IPCE) is also higher in the QDSCs with distorted lattices over the whole spectral range, as compared with that of the reference cell. The increased Jsc contributes to ≈60% increase in the power conversion efficiency (Table 1). In view of this, the present lattice-distorted CdS QDs markedly improve optical properties, which facilitates an important part for the development of high-performance QDSCs. The bandgap energy can also be estimated by linear extrapolation of the edge from the IPCE spectrum.[36] However, the bandgap obtained by the IPCE spectrum is not expected to be exactly consistent with that evaluated by the Tauc plot (Figure 5d). In the case of the IPCE spectra obtained using a complete photoelectrochemical cell, the absorbed photons are not totally transferred to photocurrent owing to a variety of charge-recombination mechanisms, which possibly causes some discrepancies in the estimated bandgap between the IPCE spectra and Tauc plots. Contrary to this expectation, as summarized in Table 2, the bandgap energies calculated by the Tauc plots (Figure 5d) agree with those estimated by using the photocurrent onsets in the IPCE curves (Supporting Information Figure S6)[37] except for the SILAR processed CdS QD (on TiO2) at 25 °C. At the present stage, we do not have any clear reasoning behind the observed discrepancy in the bandgap between the IPCE spectra and Tauc plots for the chemically crystallized CdS QD (at 25 °C) and thus it is a subject of future investigations.

3. Conclusions We have demonstrated a novel way of the bandgap tuning in a CdS QD system by exploiting a suitable quenching process. Table 2. Comparison of the bandgap energies obtained by linear extrapolation of the edges from the IPCE spectra with those by the Tauc plots. 25 °C

500 °C

700 °C

900 °C

Tauc model (Figure 5d)

452 nm (2.74 eV)

484 nm (2.56 eV)

493 nm (2.51 eV)

497 nm (2.49 eV)

Current onset at IPCE (Figure 6b)

465 nm (2.67 eV)

485 nm (2.56 eV)

495 nm (2.50 eV)

499 nm (2.48 eV)

small 2014, 10, No. 18, 3678–3684

In a few nm scales, the lattice distortion in a QD induced by tensile residual stresses significantly improved optical properties (e.g., absorbance values) in the visible wavelength region. The tunable optical bandgap (from 2.74 eV to 2.49 eV) was achieved by tailoring the degree of lattice distortion in a CdS QD grown on a TiO2 particle. Applying the idea of bandgap tuning to QD-sensitized solar cells, we are able to achieve ≈60% increase in the power conversion efficiency. We believe that the present methodology can be extended to various QD systems, which will potentially pave a way to unexplored quantum effects in the QD-based applications.

4. Experimental Section Synthesis and Fabrication: TiO2 pastes were screen-printed on the fluorine-doped tin oxide (FTO) glass. A conventional box furnace was used for the thermal-treatment of the films. The TiO2 films were gradually heated under air environment by the following profile: i) 375 °C for 5 min, ii) 450 °C for 10 min, and iii) 500 °C for 15 min. The thickness of the TiO2 film was adjusted to be approximately 6 µm, which was measured using a surface profiler (Tencor, Alpha-Step 500). The successive ionic layer adsorption and reaction (SILAR) technique,[25] known as a modified version of the chemical bath deposition, was employed to synthesize the CdS QDs on the nanostructured TiO2 film. For the deposition of the CdS QDs from the precursor solutions, two separate solutions were prepared: a) 0.1 M Cd(NO3)2 in ethanol and b) 0.1 M Na2S in methanol. The working electrodes were immersed into the Cd2+ solution and the S2− solution successively for 3 min each. After dipping into one of the two solutions, electrodes were rinsed with ethanol and methanol to remove the excess of each precursor. In the present research, the number of the SILAR cycles was chosen to be 4. The quenching process for lattice distortion in a QD follows three steps: i) put samples in as-heated box furnace, ii) stay for several seconds, and iii) take out samples from a furnace and cool down samples in air. The counter electrode was prepared by dripping a Pt solution (Solaronix, Platisol T) on the FTO glass substrate, which was followed by thermal treatment at 500 °C for 30 min. For the cell assembly, a hot-melt 60-µm-thick Surlyn (Solaronix, Meltonix 1170–60) was used as a spacer between the working electrode (CdS/TiO2/FTO) and the counter electrode (Pt/FTO). The polysulphide electrolyte was composed of 2 M S, 0.5 M Na2S, and 0.1 M KCl in the solvent with the methanol/water ratio of 8/2 (by volume). Characterizations: Crystallography of the CdS QDs was studied by employing X-ray diffraction (Rigaku RINT 2000 with Cu Kα ray). For the CdS-sensitized TiO2 samples, high-resolution TEM images were obtained by a TEM (JEOL JEM-2200FS) located at NINT (National Institute for Nanomaterials and Technology in

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.small-journal.com

3683

full papers

E.-H. Kong et al.

Pohang). The differential scanning calorimetry (DSC) analyses were performed under atmosphere with a QMS 403 C (NetzschGerätebau GmbH). The experimental parameters are as follows: Al2O3 (99.99 %) for the reference material, from 30 to 900 °C for the scanning temperature range, 20 °C/min for the heating rate. UV-visible diffuse reflectance and transmittance spectra were recorded with a Perkin Elmer UV-Vis spectrometer (Lambda 750S). The photovoltaic performance was measured under the illumination of a solar simulator (Newport, Oriel class A, 92251A) at one sun (AM 1.5, 100 mW cm−2). The incident-photon-to-current conversion efficiency (IPCE) values were recorded as a function of the wavelength from 400 nm to 600 nm (PV Measurements, Inc). A 75 W Xenon lamp was used as a light source with a monochromator. Calibration was performed with a NIST-calibrated photodiode G425 as a reference (at a chopping frequency of 10 Hz). The monochromatic power density was calibrated using a reference Si photodiode as a standard from the NIST.

Supporting Information Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements This work was financially supported by the Basic Science Research Program (Grant No. 2012R1A1A2041628 and 2013R1A2A2A01068274) through the National Research Foundation (NRF) funded by the Ministry of Education, Science and Technology of Korea.

[1] B. Chen, K. Lutker, S. V. Raju, J. Yan, W. Kanitpanyacharoen, J. Lei, S. Yang, H.-R. Wenk, H.-k. Mao, Q. Wiliams, Science 2012, 338, 1448−1451. [2] Z. W. Shan, G. Adesso, A. Cabot, M. P. Sherburne, S. A. Syed Asif, O. L. Warren, D. C. Chrzan, A. M. Minor, A. P. Alivisatos, Nat. Mater. 2008, 7, 947−952. [3] Z. Wang, X.-D. Wen, R. Hoffmann, J. S. Son, R. Li, C.-C. Fang, D.-M. Smilgies, T. Hyeon, Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 17119−17124. [4] A. N. Goldstein, C. M. Echer, A. P. Alivisatos, Science 1992, 256, 1425−1427. [5] P. Buffat, J.-P. Borel, Phys. Rev. A 1976, 13, 2287. [6] X. F. Liu, R. Wang, Y. P. Jiang, Q. Zhang, X. Y. Shan, X. H. Qiu, J. Appl. Phys. 2010, 108, 054310.

3684 www.small-journal.com

[7] X. Feng, X. Huang, X. Wang, Nanotechnology 2012, 23, 185701. [8] A. I. Hochbaum, R. Chen, R. D. Delgado, W. Liang, E. C. Garnett, M. Najarian, A. Majumdar, P. Yang, Nature 2008, 451, 163−168. [9] L. H. Liang, B. Li, Phys. Rev. B 2006, 68, 153303. [10] E. Rabinovich, G. Hodes, J. Phys. Chem. C 2013, 117, 1611−1620. [11] E.-H. Kong, Y.-J. Chang, H.-J. Park, H. M. Jang, Small 2013, 10, 1300−1307. [12] A. P. Alivisatos, Science 1996, 271, 933−937. [13] J. O. Almen, P. H. Black, Residual Stresses and Fatigue in Metals McGraw-Hill, New York, USA 1963. [14] J. Lindemann, C. Buque, F. Appel, Acta Materialia 2006, 54, 1155–1164. [15] Y. Zhang, W. H. Wang, A. L. Greer, Nature Mater. 2006, 5, 857−860. [16] B. N. Das, M. S. Mitra, Residual Stress & its Effect on Service. In: Symposium on Industrial Failure of Engineering Metals & Alloys, NML, Jamshedpur, India 1953. [17] T. Ericsson, Advances in Surface Treatments (ed. A. Niku-Lari), Pergamon Press, Oxford, England 1986. [18] T. J. Garosshen, C. S. Kim, J. M. Galligan, Appl. Phys. Lett. 1990, 56, 335−336. [19] R. Balasubramanian, W. R. Wilcox, Mater. Sci. Eng. 1993, B16, 1−7. [20] R. Agarwal, C. J. Barrelet, C. M. Lieber, Nano Lett. 2005, 5 (5), 917–920. [21] B. Mahler, P. Spinicelli, S. Buil, X. Quelin, J.-P. Hermier, B. Dubertret, Nat. Mater. 2008, 7, 659−664. [22] S. Santra, H. Yang, P. H. Holloway, J. T. Stanley, R. A. Mericle, J. Am. Chem. Soc. 2005, 127 (6), 1656–1657. [23] D. R. Hummer, P. J. Heaney, J. E. Post, Powder Diffract. 2007, 352–357. [24] S. Ouendadji, S. Ghemid, H. Meradji, F. E. H. Hassan, Comput. Mater. Sci. 2011, 50, 1460–1466. [25] H. M. Pathan, C. D. Lokhande, Bull. Mater. Sci. 2004, 27, 85−111. [26] R. I. Dimitrov, N. Moldovanska, I. K. Bonev, Thermochim. Acta 2002, 385, 41−49. [27] J. P. Perdew, K. Burke, Y. Wang, Phys. Rev. B 1996, 54, 16533–16539. [28] H. Wang, Y. Bai, H. Zhang, Z. Zhang, J. Li, L. Guo, J. Phys. Chem. C 2010, 114, 16451–16455. [29] J. Tauc, R. Grigorovici, A. Vancu, Phys. Status Solidi (b) 1966, 15, 627−637. [30] J.-W. Lee, J.-H. Im, N.-G. Park, Nanoscale 2012, 4, 6642−6648. [31] T. Kiyonaga, T. Akita, H. Tada, Chem. Commun. 2009, 2011–2013. [32] P. Ardalan, T. P. Brennan, H.-B.-R. Lee, J. R. Bakke, I.-K. Ding, M. D. McGehee, S. F. Bent, ACS Nano 2011, 5, 1495−1504. [33] J. I. Pankove, Optical Processes in Semiconductors Dover Publications, New York, USA 1975. [34] X. Chen, S. S. Mao, Chem. Rev. 2007, 107, 2891−2959. [35] L. J. Sham, M. Schlüter, Phys. Rev. Lett. 1983, 51, 1888–1891. [36] L. M. Peter, D. J. Riely, E. J. Tull, K. G. U. Wijayantha, Chem. Commun. 2002, 1030–1031. [37] G. Zhang, S. Jiang, Y. Lin, W. Ren, H. Cai, Y. Wu, Q. Zhang, N. Pan, Y. Luo, X. Wang, J. Mater. Chem. A 2014, 2, 5675–5681.

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Received: February 12, 2014 Revised: April 11, 2014 Published online: May 15, 2014

small 2014, 10, No. 18, 3678–3684

Bandgap tuning with thermal residual stresses induced in a quantum dot.

Lattice distortion induced by residual stresses can alter electronic and mechanical properties of materials significantly. Herein, a novel way of the ...
3MB Sizes 2 Downloads 3 Views