CHEMSUSCHEM FULL PAPERS DOI: 10.1002/cssc.201402279
Low temperature Hydrogen Reduction of High Surface Area Anatase and Anatase/b-TiO2 for High-Charging-Rate Batteries Edgar Ventosa,*[a] Anna Tymoczko,[a] Kunpeng Xie,[b] Wei Xia,[b] Martin Muhler,[b] and Wolfgang Schuhmann[a] There are several strategies to improve the electrochemical performance of TiO2 as negative electrode material for Li-ion batteries. Introducing oxygen vacancies through hydrogen reduction leads to an enhancement in electrical conductivity. However, this strategy does not improve the low lithium-ion mobility. Herein, we show that by decreasing the temperature of hydrogen annealing the improved lithium-ion mobility of high-surface-area TiO2 and b-TiO2 can be combined with the enhanced electrical conductivity of oxygen deficiencies. An-
nealing at only 275–300 8C in pure hydrogen atmosphere successfully creates oxygen vacancies in TiO2, as confirmed by UV/ Vis spectroscopy, whereas the temperature is low enough to maintain a high specific surface area and prevent b-to-anatase phase transformation. The hydrogen reduction of high-surfacearea anatase or anatase/b-TiO2 at these temperatures leads to improvements in the performance, achieving charge capacities of 142 or 152 mAh g 1 at 10C, respectively.
Introduction Recently, titanium dioxide (TiO2) has received increasing attention as negative electrode material for lithium-ion batteries (LIBs) for which high charging rates and extra safety are required.[1] The low lithium-ion mobility within TiO2 together with the poor electrical conductivity are still challenges that limit the electrochemical performance of this material. Several strategies have been proposed to tackle each of these two issues individually. The low lithiumi-ion mobility has been addressed by (i) decreasing the length of the lithium-ion diffusion pathway, for example, by using mesoporous[2] or nanostructured[3] materials; or by (ii) facilitating the lithiumi-ion mobility, for example, by using less dense structural phases such as bTiO2.[4] On the other hand, the conductivity has been improved by (i) coating TiO2 with carbon[5] or RuO2,[6] (ii) linking TiO2 with carbon nanotubes (CNTs)[7] or graphene[8] networks, or (iii) doping TiO2.[9] The latter approach can be divided into two categories: (i) doping with foreign elements such as niobium,[9a,b] tin,[9c] or nitrogen;[9d,e] and (ii) introducing frozen native defects such as oxygen vacancies.[9e,f] Creating oxygen-deficient TiO2 x is a very versatile approach, which in principle can be employed for TiO2 obtained by any type of synthesis because oxygen deficiencies are introduced through a post-synthesis treatment. The improvement in electrochemical performance [a] Dr. E. Ventosa, A. Tymoczko, Prof. W. Schuhmann Analytische Chemie—Elektroanalytik & Sensorik Ruhr-Universitt Bochum Universittsstr. 150, 44780 Bochum (Germany) E-mail: [email protected] [b] K. Xie, Dr. W. Xia, Prof. M. Muhler Laboratory of Industrial Chemistry Ruhr-Universitt Bochum Universittsstr. 150, 44780 Bochum (Germany)
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
resulting from oxygen deficiencies has been proposed to be more significant than that of typical foreign doping (such as doping with nitrogen).[9e] The electrical conductivity can be improved by introducing oxygen deficiencies; however, the problem of the low lithium ion mobility remains unchanged. In fact, exceeding certain levels of oxygen vacancies is detrimental for the lithium ion diffusivity because they represent defects that hinder the mobility of lithium ions.[9f] Therefore, the approach of introducing oxygen deficiencies requires combination with other strategies to further improve the electrochemical performance. Specifically, oxygen deficiencies have to be combined with nanostructuring or the b-phase of TiO2 if both limitations, electrical conductivity and lithium ion mobility, are to be tackled. Oxygendeficient TiO2 x is typically obtained through hydrogen reduction at 450–500 8C.[9e, f] Nanostructured TiO2 with high specific surface areas, for example, 300 m2 g 1, or b-phase TiO2 are not stable at those temperatures. The small nanoparticles tend to sinter and the metastable b-phase transforms into the anatase phase. Therefore, the challenge is to introduce oxygen deficiencies while maintaining the high specific surface area or preventing the transformation of the metastable phase. Herein, we propose the annealing of high-surface-area anatase (300 m2 g 1) or anatase/b-TiO2 (composite consisting of anatase and b-TiO2) at temperatures of only 275–300 8C in a pure hydrogen atmosphere to obtain oxygen-deficient TiO2 x while maintaining the high surface area or the b-phase in an attempt to simultaneously tackle the two major limitations of TiO2 as negative electrode material for LIBs.
ChemSusChem 2014, 7, 2584 – 2589
2584
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org
Results and Discussion Oxygen-deficient TiO2 drogen reduction
x
obtained by low temperature hy-
It should be noted that hydrogen reduction of anatase is usually carried out at annealing temperatures of 450–550 8C;[9e,f] thus, we use the term low temperature reduction for temperatures ranging 250–300 8C. Anatase with a specific surface area of 300 m2 g 1, referred to as pristine TiO2, was annealed at 250, 275, and 300 8C in pure hydrogen for 4 h. The obtained samples are referred to as TiO2@250_4 h, TiO2@275_4 h, TiO2@300_4 h, respectively. Another sample was annealed at 300 8C for 24 h, which is referred to as TiO2@300_24 h. X-ray diffraction (XRD) was used to evaluate phase transformations potentially occurring during the annealing in hydrogen. Figure 1 a shows the XRD patterns of the five samples. All patterns revealed only the reflections derived from the anatase phase, confirming that no anatase/rutile transformation occurred. Indeed, the only difference among all patterns was the sharpness of the main reflection of anatase at 268, which was slightly lower for pristine TiO2. This indicates a slight increase in the crystallite size after annealing. Figure 1. (a) XRD patterns and (b) specific surface areas (SSAs) of pristine The particle size or, more precisely, the specific surface area anatase, TiO2@250_4 h, TiO2@275_4 h, TiO2@300_4 h, and TiO2@300_24 h. of TiO2 is of great importance for the electrochemical per[3] formance. On the one hand, the distance from the surface to the core drastically affects the ability to store charge at high C-rates. Because of the slow lithium-ion mobility within TiO2, the entire bulk of the particle can only be reached at fast charging rates when the distance from the surface to the core is short. On the other hand, the particle size also affects the specific charge at moderate C-rates, for example, from C/10 to 1C, because lithium-ion mobility becomes extremely low at a lithium-ion content higher than Li ~ 0.55TiO2.[3c] Consequently, storing lihium ions beyond Li ~ 0.55TiO2 can only be achieved near the surface and thus higher specific charge capacities are obtained with increasing specific surface area. The specific surface areas of all annealed samples were determined using the Brunauer–Emmett–Teller nitrogen adsorption method (BET; Figure 1 b). As expected, the values decrease with increasing annealing temperature and annealing time; however, the specific surface area of all samples remained well above 200 m2 g 1 even after annealing for 24 h, confirming that TiO2 retains a high surface area after annealing in pure hydrogen at 300 8C. The change of the color of the samples revealed changes in their electronic structures. Figure 2 a shows a photograph of all samples demonstrating a gradually more yellowish and darker color with increasing annealing temperature and annealing time. This trend is consistent with results previously reportFigure 2. (a) Photographs of pristine anatase, TiO2@250_4 h, TiO2@275_4 h, TiO2@300_ ed for hydrogen annealing of TiO2 at a temperature 4 h, and TiO2@300_24 h. (b) UV/Vis reflectance spectra of pristine anatase, TiO2@275_4 h, of 450 8C.[9f] UV/Vis spectroscopy has previously been and TiO2@300_24 h. The inset in (b) is a magnification showing only the UV/Vis spectra used to confirm the presence of oxygen vacancies.[9e] of pristine anatase and TiO2@275_4 h. 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2584 – 2589
2585
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org
The reduction of TiO2 leads to an increase in absorption between 400–2500 nm[10] due to free electrons (a continuum at wavelengths > 825 nm), local Ti3 + centers (a band at 620 nm), and oxygen vacancies (bands at 1060, 441, and 486 nm). Figure 2 b shows the UV/Vis reflectance spectra of pristine TiO2, TiO2@275_4 h, and TiO2@300_24 h. The significant increase of the absorbance in the visible light region for TiO2@300_24 h confirms the presence of oxygen vacancies. The absorbance in the visible light region slightly increases from pristine to TiO2@275_4 h (inset of Figure 2 b) indicating that some oxygen vacancies are formed even during annealing at 275 8C for 4 h. Therefore, the amount of oxygen vacancies can be tuned by the annealing conditions at low temperatures. Electrochemical performance as negative electrode for high C-rate Li-ion batteries The influence of low-temperature hydrogen annealing on the electrochemical performance of TiO2 as negative electrode in LIBs was evaluated for two types of materials: (i) high surface area anatase and (ii) anatase/b-TiO2. Oxygen-deficient anatase (TiO2 x) with high specific surface area To understand the changes observed in the electrochemical performance of samples subjected to different annealing conditions, the implications of the applied strategy should be considered. The envisaged goal is a combination of enhanced electrical conductivity provided by oxygen deficiencies with short surface-to-core distances due to a high specific surface area (300 m2 g 1). Consequently, during annealing in hydrogen two effects are introduced, namely, changes in the specific surface area and the oxygen-deficiency level, both influencing the electrochemical performance of the resulting TiO2 sample. On the one hand, the electrochemical performance of TiO2 decays with decreasing surface area,[3b,c] but on the other hand, the performance improves with increasing level of oxygen deficiencies until a certain concentration is reached at which maximum performance is achieved. A further increase in oxygendeficiency level leads to a decay of battery performance because the high level of defects hindering lithium-ion mobility outweighs the performance improvement due to enhanced electrical conductivity.[9f] Five samples, that is, pristine TiO2, TiO2@250_4 h, TiO2@275_ 4 h, TiO2@300_4 h, and TiO2@300_24 h, were investigated as negative electrode material, and Figure 3 a shows the potential profiles of all samples for the first cathodic cycle. Three regions can be identified in the potential curves, which are related to the mechanism of lithium-ion accommodation in anatase.[3c] An initial solid-solution domain (region A for potentials above the plateau in Figure 3 a), in which lithium ions are accommodated without any structural changes, occurs at more anodic potentials. Above a certain lithium-ion concentration in TiO2, a phase transformation from tetragonal to orthorhombic takes place, which results in a potential plateau (region B for the plateau at ca. 1.8 V in Figure 3 a). The phase transformation is 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 3. (a) Potential profiles of the 20th cathodic cycle; (b) specific charge capacities (anodic and cathodic cycles in hollow and solid symbols, respectively) upon cycling at 0.5C, 1C, 3C, 10C and 30C; and (c) specific charge capacities at 10C of pristine anatase, TiO2@250_4 h, TiO2@275_4 h, TiO2@300_ 4 h, and TiO2@300_24 h.
complete at Li ~ 0.55TiO2. The mechanism of further lithium-ion intercalation is still under debate. In any case, intercalation beyond Li ~ 0.55TiO2 only occurs in the near-surface region as lithium-ion mobility becomes extremely hindered, resulting in a sloping curve at potentials below the phase transformation (region C for potentials below the plateau in Figure 3 a). The length of region A and C highly depends on the specific surface area of TiO2. To evaluate the results shown in Figure 3 a in depth, the potential profile of pristine TiO2 was recorded for comparison. For TiO2@250_4 h, the loss in capacity in region C was mainly responsible for the lower charge capacity compared with pristine TiO2, which indicates that the decrease in surface area during annealing at 250 8C dominates the changes in the electrochemical properties. For TiO2@275_4 h, the overall capacity exceeded that of pristine TiO2. The decrease in surface ChemSusChem 2014, 7, 2584 – 2589
2586
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org
area and the introduction of oxygen vacancies affect the performance of TiO2 oppositely. In case of TiO2@275_4 h, the effect of the introduced oxygen deficiencies outweigh that of the decrease in surface area. An analogous interpretation applies to TiO2@300_4 h, although it suffered a slightly larger loss in region C. For TiO2@300_24 h, the shortening of region C is remarkable, which led to lowest specific charge of all samples. Figure 3 b illustrates the specific charge obtained at different C-rates (0.5C, 1C, 3C, 10C, and 30C equaling 0.168, 0.336, 1.008, 3.360, and 10.080 A g 1, respectively) as a function of the number of cycles. The capacity retention over the first 20 cycles at 0.5C was improved for TiO2@275_4 h and TiO2@300_4 h compared with pristine TiO2. This improvement is most likely due to the enhanced electrical conductivity, which assists in maintaining good electrical contact among all particles in the electrode. For C-rate capability, an annealing temperature of 250 8C was clearly not sufficient to introduce a significant amount of oxygen vacancies, and the decrease in the surface area led to reduced C-rate capabilities. An annealing time of 24 h at 300 8C created an excess of oxygen vacancies and/or decreased the surface area too much, leading to the lowest C-rate capability of all samples. Annealing at 275– 300 8C for 4 h, however, introduced a beneficial level of oxygen vacancies while simultaneously a high surface area was retained. Thus improved C-rate capabilities were obtained compared with pristine TiO2. To facilitate visualization of C-rate capabilities, the specific charge capacities obtained at 10C are plotted as a function of annealing conditions (Figure 3 c). The charging rate of 10C was chosen over 30C to avoid interference from the impact of lithium-ion diffusion in the electrolyte, which might occur for materials with excellent performance at high rates. Taking the capacity of pristine TiO2 as reference, an initial decrease in capacity was observed for TiO2@250_4 h due to the decrease in surface area. At a slightly higher annealing temperature of 275 8C, a significant improvement was observed, which can only be attributed to introduced oxygen deficiencies. Increasing the annealing temperature to 300 8C did not further improve the performance. Actually, the capacity of TiO2@300_4 h was lower than that of TiO2@275_4 h, which can be attributed to the fact that the level of oxygen deficiencies exceeded the optimal value and/or further formation of oxygen deficiencies did not outweigh the decrease in surface area. Prolonging the annealing time to 24 h resulted in a considerable lower capacity. It is difficult to conclude whether exceeding the optimal amount of oxygen vacancies or the decrease in the surface area was mainly responsible for the drastic decay in capacity after annealing for 24 h. Nevertheless, the high surface area of 212 m2 g 1 for TiO2@300_24 h, which was similar to the value of 229 m2 g 1 obtained for TiO2@275_4 h. points towards an excess of oxygen vacancies as the predominant reason for the low capacity of TiO2@300_24 h.
that can be transformed into anatase at temperatures as low as 350 8C.[4c] Therefore, obtaining oxygen-deficient TiO2 x and ensuring that the phase transformation does not take place requires hydrogen annealing at temperatures substantially below 350 8C. In the previous section, we showed that hydrogen annealing at 275–300 8C for only a few hours improved the electrochemical performance of high surface area anatase. To extrapolate the benefits of low-temperature hydrogen annealing to other materials, an anatase/b-TiO2 composite was synthesized, which had been previously shown to deliver excellent an electrochemical performance as negative electrode.[9b] Anatase/b-TiO2 was annealed at 300 8C in pure H2 for 2, 4, or 24 h. The samples are referred to as TiO2_B_2 h, TiO2_B_ 4 h and TiO2_B_24 h, respectively. The as-prepared anatase/bTiO2 sample was investigated for comparison. As the synthesis involved calcination at 600 8C, the changes in the specific surface area (105 m2 g 1) for the as-prepared sample should be negligible when annealing at 300 8C. Figure 4 a shows the potential profiles of as-prepared, TiO2_ B_2 h, TiO2_B_4 h, and TiO2_B_24 h. The inset in Figure 4 a depicts the differential potential curve of the as-prepared sample, in which the presence of the anatase and the b-phase is confirmed. The cathodic peak at approximately 1.75 V is attributed to lithium-ion intercalation into the anatase phase, whereas the two cathodic peaks at 1.60 and 1.55 V are assigned to lithium-ion intercalation into the b-phase.[4] The length of all regions A, B, and C in the potential profiles increased for samples annealed for 2 or 4 h, which can be attributed to an improved electrical connection of all particles in the paste electrode. Annealing for 24 h resulted in a significant decrease in the length of both, the region attributed to anatase and the b-phase. The b-phase-to-anatase transformation should increase the length of the anatase region at the cost of the length of the b-phase region, which did not occur. Therefore, the equal loss of charge capacity in all regions indicates that an excess of oxygen vacancies was at least predominantly responsible for the poorer performance of TiO2_B_24 h, similarly as in the case of TiO2@300_24 h, rather than the loss of b-TiO2 by phase transformation. The C-rate capability of the samples is illustrated in Figure 4 b. The capacity of TiO2_B_2 h and TiO2_B_4 h exceed that of the as-prepared sample at all C-rates whereas TiO2_B_24 h exhibited the poorest performance. For better comparison, the specific charge capacity obtained at 10C (3.36 A g 1) was plotted as a function of the annealing time (Figure 4 c). Specific charge capacities of 136, 152, 146, and 122 mAh g 1 were obtained for as-prepared, TiO2_B_2 h, TiO2_B_4 h, and TiO2_B_ 24 h. Evidently, the performance of the as-prepared sample clearly increased after the sample was subjected to 2 h of annealing at 300 8C. Annealing times longer than 2 h resulted in a gradual decay in charge capacity at 10C.
Oxygen-deficient anatase/b-TiO2
Conclusions
x
The diffusivity of lithium ions in b-TiO2 is much faster than that in the anatase phase,[4a] which is very beneficial for application in high charging rate batteries. b-TiO2 is a metastable phase 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Introducing oxygen deficiencies while retaining high specific surface area can be combined to tackle simultaneously the poor electrical conductivity and the low lithium-ion mobility ChemSusChem 2014, 7, 2584 – 2589
2587
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org 300 m2 g 1 in pure hydrogen at 275 8C for 4 h leads to a significant improvement in the electrochemical performance as compared with that of the pristine material. Hydrogen annealing at low temperature can be even more advantageously applied to TiO2 containing the b-phase. Annealing anatase/b-TiO2 in pure hydrogen at 300 8C for 2 h results in an excellent electrochemical performance of the material, which exhibit reversible specific charge capacities of 200, 152, and 130 mAh g 1 at 0.5C, 10C, and 30C (168, 3.36, and 10.08 A g 1). This electrochemical performance clearly exceeds that of as-prepared anatase/b-TiO2 and makes the material promising for applications as negative electrode material for LIBs.
Experimental Section
Figure 4. (a) Potential profiles of the 20th cathodic cycle; (b) specific charge capacities (anodic and cathodic cycles in hollow and solid symbols, respectively) upon cycling at 0.5C, 1C, 3C, 10C and 30C; and (c) specific charge capacities at 10C of as-prepared anatase/b-TiO2, TiO2_B_2 h, TiO2_B_4 h, and TiO2_B_24 h. The inset in (a) is the differential potential curve of as-prepared anatase/b-TiO2 where the (de)intercalation peaks of b-phase and anatase phase are visualized.
and, thus, achieve improved electrochemical performance of TiO2 as negative electrode material in lithium-ion batteries. Oxygen-deficient TiO2 x can be obtained by annealing stoichiometric TiO2 in pure hydrogen for a few hours at temperatures as low as 275 8C while retaining a high surface area. The annealing of commercial anatase with a specific surface area of 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Two types of TiO2 were investigated: (i) commercial anatase TiO2 with a specific surface area of 300 m2 g 1 obtained from Sachtleben Chemie (Duisburg, Germany) and (ii) anatase/b-TiO2 nanoparticles obtained using a continuous spray-drying process as described previously.[11] Briefly, the latter was obtained by dissolving titanium oxysulfate sulfuric acid hydrate (TiOSO4·H2SO4·H2O, Aldrich) in nitric acid (1 mol L 1) and stirring for 4 h. A bench-top spray dryer (B-290, Bchi) with a two-fluid nozzle was used for fast evaporation of the fluid.[12] The colorless powder samples were collected from the dryer and calcined under flowing synthetic air (20.5 % O2 in He, 100 standard cubic centimeter per minute (sccm) at 600 8C for 1 h. After calcination the samples were thoroughly washed with distilled water and dried in air at 110 8C for 24 h. Annealing was carried out in a horizontal furnace equipped with a quartz tube reactor. Typically, the TiO2 powder sample (1 g) was heated in a quartz boat at a temperature gradient of 10 8C min 1 to the predefined temperature (250, 275, or 300 8C) under flowing helium (99.999 %, 100 sccm). After reaching the desired temperature, the gas flow was switched to pure hydrogen (99.999 %, 100 sccm) and maintained for 2–24 h before cooling down to room temperature. XRD studies were performed using a PANalytical MPD diffractometer (the Netherlands), with CuKa radiation. Static nitrogen physisorption measurements were carried out at 77 K using an Autosorb-1 MP Quantachrome system (USA). Samples were degassed at 200 8C for 2 h before measurements. UV/Vis spectra were recorded in the diffuse reflectance mode using a PerkinElmer Lambda 650 UV/Vis spectrometer (USA), with a Praying-Mantis mirror construction. Electrochemical experiments were carried out using three-electrode Swagelok-type cells assembled in an argon-filled glove box. The working electrodes were prepared using the doctor-blade technique. They consisted of the active material (TiO2), a conductive additive (C65 carbon black, Timcal, Bodio, Switzerland), and a binder (polyvinylidene difluoride, Solef S5130, Solvay, Belgium) in a weight ratio of 75:15:10 pasted on a 12 mm copper disc. The mass loading of the resulting electrode was 1.5–2.0 mg cm 2. Glass fiber (WhatmanGF/D, Sigma–Aldrich, Germany) filters soaked in LP40 electrolyte (Merck, Darmstadt, Germany) were used as separators. Lithium foil was used as counter and reference electrode. Galvanostatic cycling of the assembled cells was carried out using a Bio-Logic VMP-3 (Bio-Logic SAS, Claix, France) in the potential range of 1.0–3.0 V at different current densities (1C is equivalent to 336 mA g 1). All potentials are reported versus the Li/Li + potential. ChemSusChem 2014, 7, 2584 – 2589
2588
CHEMSUSCHEM FULL PAPERS Acknowledgements Financial support of the DFG in the framework of SPP 1473 (WeNDeLIB; SCHU 929/11-1) is acknowledged. Keywords: batteries · lithium · oxygen deficiency · surface area · titania [1] a) M. Wagemaker, A. P. M. Kentgens, F. M. Mulder, Nature 2002, 418, 397; b) J.-Y. Shin, D. Samuelis, J. Maier, Adv. Funct. Mater. 2011, 21, 3464; c) A. G. Dylla, G. Henkelman, K. J. Stevenson, Acc. Chem. Res. 2013, 46, 1104; d) Z. Chen, I. Belharouak, Y.-K. Sun, K. Amine, Adv. Funct. Mater. 2013, 23, 959. [2] a) Y.-G. Guo, Y.-S. Hu, J. Maier, Chem. Commun. 2006, 2783; b) Y. Ren, L. J. Hardwick, P. G. Bruce, Angew. Chem. Int. Ed. 2010, 49, 2570; Angew. Chem. 2010, 122, 2624; c) K. Saravanan, K. Ananthanarayanan, P. Balaya, Energy Environ. Sci. 2010, 3, 939. [3] a) A. S. Aric, P. G. Bruce, B. Scrosati, J. M. Tarascon, W. V. Schalkwijk, Nat. Mater. 2005, 4, 366; b) C. Jiang, M. Wei, Z. Qi, T. Kudo, I. Honma, H. Zhou, J. Power Sources 2007, 166, 239; c) M. Wagemaker, W. J. H. Borghols, F. M. Mulder, J. Am. Chem. Soc. 2007, 129, 4323. [4] a) M. Zukalov, M. Kalbc, L. Kavan, I. Exnar, M. Grtzel, Chem. Mater. 2005, 17, 1248; b) Y. Ren, Z. Liu, F. Pourpoint, A. R. Armstrong, C. P. Grey, P. G. Bruce, Angew. Chem. Int. Ed. 2012, 51, 2164; Angew. Chem. 2012, 124, 2206; c) S. Liu, H. Jia, L. Han, J. Wang, P. Gao, D. Xu, J. Yang, S. Che, Adv. Mater. 2012, 24, 3201.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org [5] F.-F. Cao, X.-L. Wu, S. Xin, Y.-G. Guo, L.-J. Wan, J. Phys. Chem. C 2010, 114, 10308. [6] Y.-G. Guo, Y.-S. Hu, W. Sigle, J. Maier, Adv. Mater. 2007, 19, 2087. [7] a) F.-F. Cao, Y.-G. Guo, S.-F. Zheng, X.-L. Wu, L.-Y. Jiang, R.-R. Bi, L.-J. Wan, J. Maier, Chem. Mater. 2010, 22, 1908; b) E. Ventosa, P. Chen, W. Schuhmann, W. Xia, Electrochem. Commun. 2012, 25, 132. [8] a) D. Wang, D. Choi, J. Li, Z. Yang, Z. Nie, R. Kou, D. Hu, C. Wang, L. V. Saraf, J. Zhang, I. A. Aksay, J. Liu, ACS Nano 2009, 3, 907; b) S. Ding, J. S. Chen, D. Luan, F. Y. C. Boey, S. Madhavi, X. W. Lou, Chem. Commun. 2011, 47, 5780. [9] a) Y. Wang, B. M. Smarsly, I. Djerdj, Chem. Mater. 2010, 22, 6624; b) E. Ventosa, B. Mei, W. Xia, M. Muhler, W. Schuhmann, ChemSusChem 2013, 6, 1312; c) L. Aldon, P. Kubiak, A. Picard, J. C. Jumas, J. Olivier-Fourcade, Chem. Mater. 2006, 18, 1401; d) Z. Wan, R. Cai, S. Jiang, Z. Shao, J. Mater. Chem. 2012, 22, 17773; e) E. Ventosa, W. Xia, S. Klink, F. La Mantia, B. Mei, M. Muhler, W. Schuhmann, Chem. Eur. J. 2013, 19, 14194; f) J.-Y. Shin, J. H. Joo, D. Samuelis, J. Maier, Chem. Mater. 2012, 24, 543. [10] A. A. Lisachenko, V. N. Kuznetsov, M. N. Zakharov, R. V. Mikhailov, Kinet. Catal. 2004, 45, 189. [11] B. Mei, M. D. Sanchez, T. Reinecke, S. Kaluza, W. Xia, M. Muhler, J. Mater. Chem. 2011, 21, 11781. [12] a) S. Kaluza, M. K. Schrçter, R. Naumann d’Alnoncourt, T. Reinecke, M. Muhler, Adv. Funct. Mater. 2008, 18, 3670; b) S. Kaluza, M. Muhler, J. Mater. Chem. 2009, 19, 3914; c) S. Kaluza, M. Muhler, Catal. Lett. 2009, 129, 287. Received: April 9, 2014 Published online on July 8, 2014
ChemSusChem 2014, 7, 2584 – 2589
2589
Low temperature Hydrogen Reduction of High Surface Area Anatase and Anatase/b-TiO2 for High-Charging-Rate Batteries Edgar Ventosa,*[a] Anna Tymoczko,[a] Kunpeng Xie,[b] Wei Xia,[b] Martin Muhler,[b] and Wolfgang Schuhmann[a] There are several strategies to improve the electrochemical performance of TiO2 as negative electrode material for Li-ion batteries. Introducing oxygen vacancies through hydrogen reduction leads to an enhancement in electrical conductivity. However, this strategy does not improve the low lithium-ion mobility. Herein, we show that by decreasing the temperature of hydrogen annealing the improved lithium-ion mobility of high-surface-area TiO2 and b-TiO2 can be combined with the enhanced electrical conductivity of oxygen deficiencies. An-
nealing at only 275–300 8C in pure hydrogen atmosphere successfully creates oxygen vacancies in TiO2, as confirmed by UV/ Vis spectroscopy, whereas the temperature is low enough to maintain a high specific surface area and prevent b-to-anatase phase transformation. The hydrogen reduction of high-surfacearea anatase or anatase/b-TiO2 at these temperatures leads to improvements in the performance, achieving charge capacities of 142 or 152 mAh g 1 at 10C, respectively.
Introduction Recently, titanium dioxide (TiO2) has received increasing attention as negative electrode material for lithium-ion batteries (LIBs) for which high charging rates and extra safety are required.[1] The low lithium-ion mobility within TiO2 together with the poor electrical conductivity are still challenges that limit the electrochemical performance of this material. Several strategies have been proposed to tackle each of these two issues individually. The low lithiumi-ion mobility has been addressed by (i) decreasing the length of the lithium-ion diffusion pathway, for example, by using mesoporous[2] or nanostructured[3] materials; or by (ii) facilitating the lithiumi-ion mobility, for example, by using less dense structural phases such as bTiO2.[4] On the other hand, the conductivity has been improved by (i) coating TiO2 with carbon[5] or RuO2,[6] (ii) linking TiO2 with carbon nanotubes (CNTs)[7] or graphene[8] networks, or (iii) doping TiO2.[9] The latter approach can be divided into two categories: (i) doping with foreign elements such as niobium,[9a,b] tin,[9c] or nitrogen;[9d,e] and (ii) introducing frozen native defects such as oxygen vacancies.[9e,f] Creating oxygen-deficient TiO2 x is a very versatile approach, which in principle can be employed for TiO2 obtained by any type of synthesis because oxygen deficiencies are introduced through a post-synthesis treatment. The improvement in electrochemical performance [a] Dr. E. Ventosa, A. Tymoczko, Prof. W. Schuhmann Analytische Chemie—Elektroanalytik & Sensorik Ruhr-Universitt Bochum Universittsstr. 150, 44780 Bochum (Germany) E-mail: [email protected] [b] K. Xie, Dr. W. Xia, Prof. M. Muhler Laboratory of Industrial Chemistry Ruhr-Universitt Bochum Universittsstr. 150, 44780 Bochum (Germany)
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
resulting from oxygen deficiencies has been proposed to be more significant than that of typical foreign doping (such as doping with nitrogen).[9e] The electrical conductivity can be improved by introducing oxygen deficiencies; however, the problem of the low lithium ion mobility remains unchanged. In fact, exceeding certain levels of oxygen vacancies is detrimental for the lithium ion diffusivity because they represent defects that hinder the mobility of lithium ions.[9f] Therefore, the approach of introducing oxygen deficiencies requires combination with other strategies to further improve the electrochemical performance. Specifically, oxygen deficiencies have to be combined with nanostructuring or the b-phase of TiO2 if both limitations, electrical conductivity and lithium ion mobility, are to be tackled. Oxygendeficient TiO2 x is typically obtained through hydrogen reduction at 450–500 8C.[9e, f] Nanostructured TiO2 with high specific surface areas, for example, 300 m2 g 1, or b-phase TiO2 are not stable at those temperatures. The small nanoparticles tend to sinter and the metastable b-phase transforms into the anatase phase. Therefore, the challenge is to introduce oxygen deficiencies while maintaining the high specific surface area or preventing the transformation of the metastable phase. Herein, we propose the annealing of high-surface-area anatase (300 m2 g 1) or anatase/b-TiO2 (composite consisting of anatase and b-TiO2) at temperatures of only 275–300 8C in a pure hydrogen atmosphere to obtain oxygen-deficient TiO2 x while maintaining the high surface area or the b-phase in an attempt to simultaneously tackle the two major limitations of TiO2 as negative electrode material for LIBs.
ChemSusChem 2014, 7, 2584 – 2589
2584
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org
Results and Discussion Oxygen-deficient TiO2 drogen reduction
x
obtained by low temperature hy-
It should be noted that hydrogen reduction of anatase is usually carried out at annealing temperatures of 450–550 8C;[9e,f] thus, we use the term low temperature reduction for temperatures ranging 250–300 8C. Anatase with a specific surface area of 300 m2 g 1, referred to as pristine TiO2, was annealed at 250, 275, and 300 8C in pure hydrogen for 4 h. The obtained samples are referred to as TiO2@250_4 h, TiO2@275_4 h, TiO2@300_4 h, respectively. Another sample was annealed at 300 8C for 24 h, which is referred to as TiO2@300_24 h. X-ray diffraction (XRD) was used to evaluate phase transformations potentially occurring during the annealing in hydrogen. Figure 1 a shows the XRD patterns of the five samples. All patterns revealed only the reflections derived from the anatase phase, confirming that no anatase/rutile transformation occurred. Indeed, the only difference among all patterns was the sharpness of the main reflection of anatase at 268, which was slightly lower for pristine TiO2. This indicates a slight increase in the crystallite size after annealing. Figure 1. (a) XRD patterns and (b) specific surface areas (SSAs) of pristine The particle size or, more precisely, the specific surface area anatase, TiO2@250_4 h, TiO2@275_4 h, TiO2@300_4 h, and TiO2@300_24 h. of TiO2 is of great importance for the electrochemical per[3] formance. On the one hand, the distance from the surface to the core drastically affects the ability to store charge at high C-rates. Because of the slow lithium-ion mobility within TiO2, the entire bulk of the particle can only be reached at fast charging rates when the distance from the surface to the core is short. On the other hand, the particle size also affects the specific charge at moderate C-rates, for example, from C/10 to 1C, because lithium-ion mobility becomes extremely low at a lithium-ion content higher than Li ~ 0.55TiO2.[3c] Consequently, storing lihium ions beyond Li ~ 0.55TiO2 can only be achieved near the surface and thus higher specific charge capacities are obtained with increasing specific surface area. The specific surface areas of all annealed samples were determined using the Brunauer–Emmett–Teller nitrogen adsorption method (BET; Figure 1 b). As expected, the values decrease with increasing annealing temperature and annealing time; however, the specific surface area of all samples remained well above 200 m2 g 1 even after annealing for 24 h, confirming that TiO2 retains a high surface area after annealing in pure hydrogen at 300 8C. The change of the color of the samples revealed changes in their electronic structures. Figure 2 a shows a photograph of all samples demonstrating a gradually more yellowish and darker color with increasing annealing temperature and annealing time. This trend is consistent with results previously reportFigure 2. (a) Photographs of pristine anatase, TiO2@250_4 h, TiO2@275_4 h, TiO2@300_ ed for hydrogen annealing of TiO2 at a temperature 4 h, and TiO2@300_24 h. (b) UV/Vis reflectance spectra of pristine anatase, TiO2@275_4 h, of 450 8C.[9f] UV/Vis spectroscopy has previously been and TiO2@300_24 h. The inset in (b) is a magnification showing only the UV/Vis spectra used to confirm the presence of oxygen vacancies.[9e] of pristine anatase and TiO2@275_4 h. 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2584 – 2589
2585
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org
The reduction of TiO2 leads to an increase in absorption between 400–2500 nm[10] due to free electrons (a continuum at wavelengths > 825 nm), local Ti3 + centers (a band at 620 nm), and oxygen vacancies (bands at 1060, 441, and 486 nm). Figure 2 b shows the UV/Vis reflectance spectra of pristine TiO2, TiO2@275_4 h, and TiO2@300_24 h. The significant increase of the absorbance in the visible light region for TiO2@300_24 h confirms the presence of oxygen vacancies. The absorbance in the visible light region slightly increases from pristine to TiO2@275_4 h (inset of Figure 2 b) indicating that some oxygen vacancies are formed even during annealing at 275 8C for 4 h. Therefore, the amount of oxygen vacancies can be tuned by the annealing conditions at low temperatures. Electrochemical performance as negative electrode for high C-rate Li-ion batteries The influence of low-temperature hydrogen annealing on the electrochemical performance of TiO2 as negative electrode in LIBs was evaluated for two types of materials: (i) high surface area anatase and (ii) anatase/b-TiO2. Oxygen-deficient anatase (TiO2 x) with high specific surface area To understand the changes observed in the electrochemical performance of samples subjected to different annealing conditions, the implications of the applied strategy should be considered. The envisaged goal is a combination of enhanced electrical conductivity provided by oxygen deficiencies with short surface-to-core distances due to a high specific surface area (300 m2 g 1). Consequently, during annealing in hydrogen two effects are introduced, namely, changes in the specific surface area and the oxygen-deficiency level, both influencing the electrochemical performance of the resulting TiO2 sample. On the one hand, the electrochemical performance of TiO2 decays with decreasing surface area,[3b,c] but on the other hand, the performance improves with increasing level of oxygen deficiencies until a certain concentration is reached at which maximum performance is achieved. A further increase in oxygendeficiency level leads to a decay of battery performance because the high level of defects hindering lithium-ion mobility outweighs the performance improvement due to enhanced electrical conductivity.[9f] Five samples, that is, pristine TiO2, TiO2@250_4 h, TiO2@275_ 4 h, TiO2@300_4 h, and TiO2@300_24 h, were investigated as negative electrode material, and Figure 3 a shows the potential profiles of all samples for the first cathodic cycle. Three regions can be identified in the potential curves, which are related to the mechanism of lithium-ion accommodation in anatase.[3c] An initial solid-solution domain (region A for potentials above the plateau in Figure 3 a), in which lithium ions are accommodated without any structural changes, occurs at more anodic potentials. Above a certain lithium-ion concentration in TiO2, a phase transformation from tetragonal to orthorhombic takes place, which results in a potential plateau (region B for the plateau at ca. 1.8 V in Figure 3 a). The phase transformation is 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 3. (a) Potential profiles of the 20th cathodic cycle; (b) specific charge capacities (anodic and cathodic cycles in hollow and solid symbols, respectively) upon cycling at 0.5C, 1C, 3C, 10C and 30C; and (c) specific charge capacities at 10C of pristine anatase, TiO2@250_4 h, TiO2@275_4 h, TiO2@300_ 4 h, and TiO2@300_24 h.
complete at Li ~ 0.55TiO2. The mechanism of further lithium-ion intercalation is still under debate. In any case, intercalation beyond Li ~ 0.55TiO2 only occurs in the near-surface region as lithium-ion mobility becomes extremely hindered, resulting in a sloping curve at potentials below the phase transformation (region C for potentials below the plateau in Figure 3 a). The length of region A and C highly depends on the specific surface area of TiO2. To evaluate the results shown in Figure 3 a in depth, the potential profile of pristine TiO2 was recorded for comparison. For TiO2@250_4 h, the loss in capacity in region C was mainly responsible for the lower charge capacity compared with pristine TiO2, which indicates that the decrease in surface area during annealing at 250 8C dominates the changes in the electrochemical properties. For TiO2@275_4 h, the overall capacity exceeded that of pristine TiO2. The decrease in surface ChemSusChem 2014, 7, 2584 – 2589
2586
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org
area and the introduction of oxygen vacancies affect the performance of TiO2 oppositely. In case of TiO2@275_4 h, the effect of the introduced oxygen deficiencies outweigh that of the decrease in surface area. An analogous interpretation applies to TiO2@300_4 h, although it suffered a slightly larger loss in region C. For TiO2@300_24 h, the shortening of region C is remarkable, which led to lowest specific charge of all samples. Figure 3 b illustrates the specific charge obtained at different C-rates (0.5C, 1C, 3C, 10C, and 30C equaling 0.168, 0.336, 1.008, 3.360, and 10.080 A g 1, respectively) as a function of the number of cycles. The capacity retention over the first 20 cycles at 0.5C was improved for TiO2@275_4 h and TiO2@300_4 h compared with pristine TiO2. This improvement is most likely due to the enhanced electrical conductivity, which assists in maintaining good electrical contact among all particles in the electrode. For C-rate capability, an annealing temperature of 250 8C was clearly not sufficient to introduce a significant amount of oxygen vacancies, and the decrease in the surface area led to reduced C-rate capabilities. An annealing time of 24 h at 300 8C created an excess of oxygen vacancies and/or decreased the surface area too much, leading to the lowest C-rate capability of all samples. Annealing at 275– 300 8C for 4 h, however, introduced a beneficial level of oxygen vacancies while simultaneously a high surface area was retained. Thus improved C-rate capabilities were obtained compared with pristine TiO2. To facilitate visualization of C-rate capabilities, the specific charge capacities obtained at 10C are plotted as a function of annealing conditions (Figure 3 c). The charging rate of 10C was chosen over 30C to avoid interference from the impact of lithium-ion diffusion in the electrolyte, which might occur for materials with excellent performance at high rates. Taking the capacity of pristine TiO2 as reference, an initial decrease in capacity was observed for TiO2@250_4 h due to the decrease in surface area. At a slightly higher annealing temperature of 275 8C, a significant improvement was observed, which can only be attributed to introduced oxygen deficiencies. Increasing the annealing temperature to 300 8C did not further improve the performance. Actually, the capacity of TiO2@300_4 h was lower than that of TiO2@275_4 h, which can be attributed to the fact that the level of oxygen deficiencies exceeded the optimal value and/or further formation of oxygen deficiencies did not outweigh the decrease in surface area. Prolonging the annealing time to 24 h resulted in a considerable lower capacity. It is difficult to conclude whether exceeding the optimal amount of oxygen vacancies or the decrease in the surface area was mainly responsible for the drastic decay in capacity after annealing for 24 h. Nevertheless, the high surface area of 212 m2 g 1 for TiO2@300_24 h, which was similar to the value of 229 m2 g 1 obtained for TiO2@275_4 h. points towards an excess of oxygen vacancies as the predominant reason for the low capacity of TiO2@300_24 h.
that can be transformed into anatase at temperatures as low as 350 8C.[4c] Therefore, obtaining oxygen-deficient TiO2 x and ensuring that the phase transformation does not take place requires hydrogen annealing at temperatures substantially below 350 8C. In the previous section, we showed that hydrogen annealing at 275–300 8C for only a few hours improved the electrochemical performance of high surface area anatase. To extrapolate the benefits of low-temperature hydrogen annealing to other materials, an anatase/b-TiO2 composite was synthesized, which had been previously shown to deliver excellent an electrochemical performance as negative electrode.[9b] Anatase/b-TiO2 was annealed at 300 8C in pure H2 for 2, 4, or 24 h. The samples are referred to as TiO2_B_2 h, TiO2_B_ 4 h and TiO2_B_24 h, respectively. The as-prepared anatase/bTiO2 sample was investigated for comparison. As the synthesis involved calcination at 600 8C, the changes in the specific surface area (105 m2 g 1) for the as-prepared sample should be negligible when annealing at 300 8C. Figure 4 a shows the potential profiles of as-prepared, TiO2_ B_2 h, TiO2_B_4 h, and TiO2_B_24 h. The inset in Figure 4 a depicts the differential potential curve of the as-prepared sample, in which the presence of the anatase and the b-phase is confirmed. The cathodic peak at approximately 1.75 V is attributed to lithium-ion intercalation into the anatase phase, whereas the two cathodic peaks at 1.60 and 1.55 V are assigned to lithium-ion intercalation into the b-phase.[4] The length of all regions A, B, and C in the potential profiles increased for samples annealed for 2 or 4 h, which can be attributed to an improved electrical connection of all particles in the paste electrode. Annealing for 24 h resulted in a significant decrease in the length of both, the region attributed to anatase and the b-phase. The b-phase-to-anatase transformation should increase the length of the anatase region at the cost of the length of the b-phase region, which did not occur. Therefore, the equal loss of charge capacity in all regions indicates that an excess of oxygen vacancies was at least predominantly responsible for the poorer performance of TiO2_B_24 h, similarly as in the case of TiO2@300_24 h, rather than the loss of b-TiO2 by phase transformation. The C-rate capability of the samples is illustrated in Figure 4 b. The capacity of TiO2_B_2 h and TiO2_B_4 h exceed that of the as-prepared sample at all C-rates whereas TiO2_B_24 h exhibited the poorest performance. For better comparison, the specific charge capacity obtained at 10C (3.36 A g 1) was plotted as a function of the annealing time (Figure 4 c). Specific charge capacities of 136, 152, 146, and 122 mAh g 1 were obtained for as-prepared, TiO2_B_2 h, TiO2_B_4 h, and TiO2_B_ 24 h. Evidently, the performance of the as-prepared sample clearly increased after the sample was subjected to 2 h of annealing at 300 8C. Annealing times longer than 2 h resulted in a gradual decay in charge capacity at 10C.
Oxygen-deficient anatase/b-TiO2
Conclusions
x
The diffusivity of lithium ions in b-TiO2 is much faster than that in the anatase phase,[4a] which is very beneficial for application in high charging rate batteries. b-TiO2 is a metastable phase 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Introducing oxygen deficiencies while retaining high specific surface area can be combined to tackle simultaneously the poor electrical conductivity and the low lithium-ion mobility ChemSusChem 2014, 7, 2584 – 2589
2587
CHEMSUSCHEM FULL PAPERS
www.chemsuschem.org 300 m2 g 1 in pure hydrogen at 275 8C for 4 h leads to a significant improvement in the electrochemical performance as compared with that of the pristine material. Hydrogen annealing at low temperature can be even more advantageously applied to TiO2 containing the b-phase. Annealing anatase/b-TiO2 in pure hydrogen at 300 8C for 2 h results in an excellent electrochemical performance of the material, which exhibit reversible specific charge capacities of 200, 152, and 130 mAh g 1 at 0.5C, 10C, and 30C (168, 3.36, and 10.08 A g 1). This electrochemical performance clearly exceeds that of as-prepared anatase/b-TiO2 and makes the material promising for applications as negative electrode material for LIBs.
Experimental Section
Figure 4. (a) Potential profiles of the 20th cathodic cycle; (b) specific charge capacities (anodic and cathodic cycles in hollow and solid symbols, respectively) upon cycling at 0.5C, 1C, 3C, 10C and 30C; and (c) specific charge capacities at 10C of as-prepared anatase/b-TiO2, TiO2_B_2 h, TiO2_B_4 h, and TiO2_B_24 h. The inset in (a) is the differential potential curve of as-prepared anatase/b-TiO2 where the (de)intercalation peaks of b-phase and anatase phase are visualized.
and, thus, achieve improved electrochemical performance of TiO2 as negative electrode material in lithium-ion batteries. Oxygen-deficient TiO2 x can be obtained by annealing stoichiometric TiO2 in pure hydrogen for a few hours at temperatures as low as 275 8C while retaining a high surface area. The annealing of commercial anatase with a specific surface area of 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Two types of TiO2 were investigated: (i) commercial anatase TiO2 with a specific surface area of 300 m2 g 1 obtained from Sachtleben Chemie (Duisburg, Germany) and (ii) anatase/b-TiO2 nanoparticles obtained using a continuous spray-drying process as described previously.[11] Briefly, the latter was obtained by dissolving titanium oxysulfate sulfuric acid hydrate (TiOSO4·H2SO4·H2O, Aldrich) in nitric acid (1 mol L 1) and stirring for 4 h. A bench-top spray dryer (B-290, Bchi) with a two-fluid nozzle was used for fast evaporation of the fluid.[12] The colorless powder samples were collected from the dryer and calcined under flowing synthetic air (20.5 % O2 in He, 100 standard cubic centimeter per minute (sccm) at 600 8C for 1 h. After calcination the samples were thoroughly washed with distilled water and dried in air at 110 8C for 24 h. Annealing was carried out in a horizontal furnace equipped with a quartz tube reactor. Typically, the TiO2 powder sample (1 g) was heated in a quartz boat at a temperature gradient of 10 8C min 1 to the predefined temperature (250, 275, or 300 8C) under flowing helium (99.999 %, 100 sccm). After reaching the desired temperature, the gas flow was switched to pure hydrogen (99.999 %, 100 sccm) and maintained for 2–24 h before cooling down to room temperature. XRD studies were performed using a PANalytical MPD diffractometer (the Netherlands), with CuKa radiation. Static nitrogen physisorption measurements were carried out at 77 K using an Autosorb-1 MP Quantachrome system (USA). Samples were degassed at 200 8C for 2 h before measurements. UV/Vis spectra were recorded in the diffuse reflectance mode using a PerkinElmer Lambda 650 UV/Vis spectrometer (USA), with a Praying-Mantis mirror construction. Electrochemical experiments were carried out using three-electrode Swagelok-type cells assembled in an argon-filled glove box. The working electrodes were prepared using the doctor-blade technique. They consisted of the active material (TiO2), a conductive additive (C65 carbon black, Timcal, Bodio, Switzerland), and a binder (polyvinylidene difluoride, Solef S5130, Solvay, Belgium) in a weight ratio of 75:15:10 pasted on a 12 mm copper disc. The mass loading of the resulting electrode was 1.5–2.0 mg cm 2. Glass fiber (WhatmanGF/D, Sigma–Aldrich, Germany) filters soaked in LP40 electrolyte (Merck, Darmstadt, Germany) were used as separators. Lithium foil was used as counter and reference electrode. Galvanostatic cycling of the assembled cells was carried out using a Bio-Logic VMP-3 (Bio-Logic SAS, Claix, France) in the potential range of 1.0–3.0 V at different current densities (1C is equivalent to 336 mA g 1). All potentials are reported versus the Li/Li + potential. ChemSusChem 2014, 7, 2584 – 2589
2588
CHEMSUSCHEM FULL PAPERS Acknowledgements Financial support of the DFG in the framework of SPP 1473 (WeNDeLIB; SCHU 929/11-1) is acknowledged. Keywords: batteries · lithium · oxygen deficiency · surface area · titania [1] a) M. Wagemaker, A. P. M. Kentgens, F. M. Mulder, Nature 2002, 418, 397; b) J.-Y. Shin, D. Samuelis, J. Maier, Adv. Funct. Mater. 2011, 21, 3464; c) A. G. Dylla, G. Henkelman, K. J. Stevenson, Acc. Chem. Res. 2013, 46, 1104; d) Z. Chen, I. Belharouak, Y.-K. Sun, K. Amine, Adv. Funct. Mater. 2013, 23, 959. [2] a) Y.-G. Guo, Y.-S. Hu, J. Maier, Chem. Commun. 2006, 2783; b) Y. Ren, L. J. Hardwick, P. G. Bruce, Angew. Chem. Int. Ed. 2010, 49, 2570; Angew. Chem. 2010, 122, 2624; c) K. Saravanan, K. Ananthanarayanan, P. Balaya, Energy Environ. Sci. 2010, 3, 939. [3] a) A. S. Aric, P. G. Bruce, B. Scrosati, J. M. Tarascon, W. V. Schalkwijk, Nat. Mater. 2005, 4, 366; b) C. Jiang, M. Wei, Z. Qi, T. Kudo, I. Honma, H. Zhou, J. Power Sources 2007, 166, 239; c) M. Wagemaker, W. J. H. Borghols, F. M. Mulder, J. Am. Chem. Soc. 2007, 129, 4323. [4] a) M. Zukalov, M. Kalbc, L. Kavan, I. Exnar, M. Grtzel, Chem. Mater. 2005, 17, 1248; b) Y. Ren, Z. Liu, F. Pourpoint, A. R. Armstrong, C. P. Grey, P. G. Bruce, Angew. Chem. Int. Ed. 2012, 51, 2164; Angew. Chem. 2012, 124, 2206; c) S. Liu, H. Jia, L. Han, J. Wang, P. Gao, D. Xu, J. Yang, S. Che, Adv. Mater. 2012, 24, 3201.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org [5] F.-F. Cao, X.-L. Wu, S. Xin, Y.-G. Guo, L.-J. Wan, J. Phys. Chem. C 2010, 114, 10308. [6] Y.-G. Guo, Y.-S. Hu, W. Sigle, J. Maier, Adv. Mater. 2007, 19, 2087. [7] a) F.-F. Cao, Y.-G. Guo, S.-F. Zheng, X.-L. Wu, L.-Y. Jiang, R.-R. Bi, L.-J. Wan, J. Maier, Chem. Mater. 2010, 22, 1908; b) E. Ventosa, P. Chen, W. Schuhmann, W. Xia, Electrochem. Commun. 2012, 25, 132. [8] a) D. Wang, D. Choi, J. Li, Z. Yang, Z. Nie, R. Kou, D. Hu, C. Wang, L. V. Saraf, J. Zhang, I. A. Aksay, J. Liu, ACS Nano 2009, 3, 907; b) S. Ding, J. S. Chen, D. Luan, F. Y. C. Boey, S. Madhavi, X. W. Lou, Chem. Commun. 2011, 47, 5780. [9] a) Y. Wang, B. M. Smarsly, I. Djerdj, Chem. Mater. 2010, 22, 6624; b) E. Ventosa, B. Mei, W. Xia, M. Muhler, W. Schuhmann, ChemSusChem 2013, 6, 1312; c) L. Aldon, P. Kubiak, A. Picard, J. C. Jumas, J. Olivier-Fourcade, Chem. Mater. 2006, 18, 1401; d) Z. Wan, R. Cai, S. Jiang, Z. Shao, J. Mater. Chem. 2012, 22, 17773; e) E. Ventosa, W. Xia, S. Klink, F. La Mantia, B. Mei, M. Muhler, W. Schuhmann, Chem. Eur. J. 2013, 19, 14194; f) J.-Y. Shin, J. H. Joo, D. Samuelis, J. Maier, Chem. Mater. 2012, 24, 543. [10] A. A. Lisachenko, V. N. Kuznetsov, M. N. Zakharov, R. V. Mikhailov, Kinet. Catal. 2004, 45, 189. [11] B. Mei, M. D. Sanchez, T. Reinecke, S. Kaluza, W. Xia, M. Muhler, J. Mater. Chem. 2011, 21, 11781. [12] a) S. Kaluza, M. K. Schrçter, R. Naumann d’Alnoncourt, T. Reinecke, M. Muhler, Adv. Funct. Mater. 2008, 18, 3670; b) S. Kaluza, M. Muhler, J. Mater. Chem. 2009, 19, 3914; c) S. Kaluza, M. Muhler, Catal. Lett. 2009, 129, 287. Received: April 9, 2014 Published online on July 8, 2014
ChemSusChem 2014, 7, 2584 – 2589
2589
