Hostname: page-component-745bb68f8f-s22k5 Total loading time: 0 Render date: 2025-02-11T10:27:06.320Z Has data issue: false hasContentIssue false
  • English
  • Français

An investigation into the temperature phase transitions of synthesized lithium titanate materials doped with Al, Co, Ni and Mg by in situ powder X-ray diffraction

Published online by Cambridge University Press:  25 August 2020

X. van Niekerk Open the ORCID record for X. van Niekerk [Opens in a new window] ,
E. E. Ferg ,
C. Gelant  and
D. G. Billing
X. van Niekerk*
Affiliation:
Department of Chemistry and uYilo, e-Mobility Technology Innovation Programme, Nelson Mandela University, P.O. Box 77000, Port Elizabeth6001, South Africa
E. E. Ferg
Affiliation:
Department of Chemistry and uYilo, e-Mobility Technology Innovation Programme, Nelson Mandela University, P.O. Box 77000, Port Elizabeth6001, South Africa
C. Gelant
Affiliation:
Department of Chemistry and uYilo, e-Mobility Technology Innovation Programme, Nelson Mandela University, P.O. Box 77000, Port Elizabeth6001, South Africa
D. G. Billing
Affiliation:
School of Chemistry and DST/NRF Centre of Excellence in Strong Materials, University of Witwatersrand, Private Bag 3, Johannesburg2000, South Africa
*
a)Author to whom correspondence should be addressed. Electronic mail: xandri.vanniekerk@mandela.ac.za
Rights & Permissions [Opens in a new window]

Abstract

Li4Ti5O12 (LTO) and its doped analogues Li4Ti4.95M0.05O12 (M = Al3+, Co3+, Ni2+, and Mg2+) were synthesized and characterized using in situ PXRD to monitor the phase transitions during the sol–gel synthesis of the spinel material. These results are complimented by thermogravimetric analysis, which illustrates the decomposition of the materials synthesized, where the final LTO products are seen to form at approximately 550 °C. The material has an amorphous structure from room temperature, coupled with a crystalline phase which is speculated to be H2Ti2O5·H2O. This crystalline phase disappears at 250 °C, with the material still in the amorphous state. The crystalline LTO phase starts at approximately 550 °C, with anatase co-crystallizing with the spinel phase. Rutile appears at 600 °C and co-crystallizes with the final product at 850 °C, where anatase is no longer seen. The rutile impurity remains present after cooling the material to room temperature, and results indicate that prolonged heating at 850 °C is required to reduce the rutile content. Rietveld refinement of diffraction patterns show that the unit-cell parameter increases with increasing temperature, coupled with a decrease when cooling the sample. The crystallite sizes follow the same trend, with a significant increase above temperatures of 750 °C.

Keywords

in-situ powder X-ray diffractionAl, Co, Ni and Mg-doped Li4Ti5O12unit-cell parameteramorphous and crystalline precursorsrutile and anatase
Type
Technical Article
Information
Powder Diffraction , Volume 35 , Issue 4 , December 2020 , pp. 233 - 246
Check for updates
Copyright
Copyright © 2020 International Centre for Diffraction Data

I. INTRODUCTION

Spinel Li4Ti5O12 (LTO) is a promising anode material for lithium-ion batteries due to its zero-strain electrochemical capability, inherent safety, and excellent cyclic performance (Nugroho et al., Reference Nugroho, Chung and Kim2014, Wang et al., Reference Wang, Wang, He, Tang, Han, Yang, Wagemaker, Li, Yang, Kim and Kang2015). The material has a face-centred structure with an Fd3m (227) cubic space group. The lithium ions are located at the tetrahedral 8a sites (Figure 1), the lithium and tetravalent titanium ions are randomly dispersed at the 16d octahedral sites and the oxygen ions are located at the 32e sites, with a unit-cell parameter of 8.358 Å (Yi et al., Reference Yi, Jiang, Shu, Yue, Zhu and Qiao2010). Upon lithium insertion during discharge, the lithium ions move from the 8a to the 16c sites forming Li7Ti5O12, a rock salt structure, where all 16c sites are occupied and all 8a sites are empty (Ganapathy and Wagemaker, Reference Ganapathy and Wagemaker2012), as shown in the following reaction:

(1)$${\rm L}{\rm i}_4{\rm T}{\rm i}_5{\rm O}_{12}\lpar {\rm s} \rpar + 3e^-{ + } 3{\rm L}{\rm i}^ + \lpar {{\rm aq}} \rpar \to {\rm L}{\rm i}_7{\rm T}{\rm i}_5{\rm O}_{12}\lpar {\rm s} \rpar \quad 1.55\,{\rm V}\;{\rm vs} .\;{\rm Li/L}{\rm i}^{+}$$

Figure 1. Structure of spinel Li4Ti5O12 showing atoms in the unit cell only (left) and unit cell with the polyhedra (right) where the tetrahedra indicates the 8a sites occupied by Li+ and the octahedra indicates the 16d sites that are occupied by Li+ and Ti4+. The O2− anions occupy the octahedral 32e sites (Deschanvres et al., Reference Deschanvres, Raveau and Sekkal1971; Koichi Momma, Reference Koichi Momma2011).

In an electrochemical cell, the spinel LTO gives a flat charge–discharge curve at 1.55 V vs. Li/Li+ owing to the stability of the Ti4+/Ti3+ couple during capacity cycling (Sun et al., Reference Sun, Radovanovic and Cui2015); however, the high oxidation state (+4) of the titanium in the spinel structure acts as an insulator that effectively results in poor electronic conductivity (≈10−13 S cm−1), which is one of the main problems encountered with this material as it restricts electron transfer for high rate applications (Yi et al., Reference Yi, Jiang, Shu, Yue, Zhu and Qiao2010; Sun et al., Reference Sun, Radovanovic and Cui2015). Various methods have been proposed to improve the conductivity, including carbon surface modification, compositing, and doping the material with metals, non-metals, and metal oxides (Ming et al., Reference Ming, Ming, Li, Zhou, Wang, Jin, Fu, Adkins and Zheng2014; Sun et al., Reference Sun, Radovanovic and Cui2015; Xie et al., Reference Xie, Li, Li, Chen and Qu2015; Ni et al., Reference Ni, Song and Fan2016; Coelho et al., Reference Coelho, Pokle, Park, McEvoy, Berner, Duesberg and Nicolosi2017).

The understanding of the synthetic route to form the spinel Li4Ti5O12 and its doped congeners is an important aspect to consider due to the influence it would have on the final electrochemical properties of the active material. The most commonly used synthetic methods include a range of solid-state reactions, hydro- or solvo-thermal methods, sol–gel synthesis, and spray pyrolysis (Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005; Rai et al., Reference Rai, Gim, Kang, Mathew, Anh, Kang, Song, Paul and Kim2012; Han et al., Reference Han, Ryu, Jeong and Yoon2013; Mosa et al., Reference Mosa, Vélez, Reinosa, Aparicio, Yamaguchi, Tadanaga and Tatsumisago2013; Zhang et al., Reference Zhang, Cao, Huang, Wang, Wu and Cai2013; Li et al., Reference Li, Liang, Wei, Zhu and Qian2014; Priyono et al., Reference Priyono, Syahrial, Herman Yuwono, Kartini, Marfelly and Furkon Rahmatulloh2015; Kuo et al., Reference Kuo, Peng, Xiao and Lin2016). Solid-state reaction methods have predominated due to the ease of mixing the solid material in a given ratio; however, it has limitations in terms of the resultant materials particle size and shape, its uniformity and the required prolonged calcination times (Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005; Yi et al., Reference Yi, Jiang, Shu, Yue, Zhu and Qiao2010). Sol–gel synthesis overcomes many of these limitations by producing a material that is homogeneous with cubic-shaped particles, although it can be more expensive when required to make the material on a large scale (Sun et al., Reference Sun, Radovanovic and Cui2015). Chelating agents such as acetic acid and triethanolamine (TEA) have been used in the sol–gel synthesis of lithium titanium oxide, where their main functions are to coordinate to Ti4+ ions, thereby preventing hydrolysis to Ti(OH)4 (Shen et al., Reference Shen, Zhang, Zhou and Li2003; Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005). This method requires high calcining temperatures around 800 °C for long times to form the final lithium titanium oxide powder, with good homogeneity and uniform, sub-micron-sized particles (Shen et al., Reference Shen, Zhang, Zhou and Li2003; Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005; Mahmoud et al., Reference Mahmoud, Amarilla and Saadoune2015). This makes the development of batteries with this type of material for large-scale applications difficult where its manufacturing cost would be very high when compared to the conventional use of carbon as anodes. An in-depth knowledge of the phase transitions involved in the sol–gel synthesis would be beneficial to optimize the manufacturing process. In order to identify and understand the temperature phase transitions that the material undergoes during the calcination synthesis process, in situ powder X-ray diffraction (PXRD) was shown to be a well-suited technique to study the phase transitions of lithium manganese oxide spinel used as cathode materials in lithium-ion batteries and was employed in this study (Snyders et al., Reference Snyders, Ferg and Billing2016).

This paper focuses on studying the doping effect of Ti4+ ions in LTO with 1% Al3+, Mg2+, Co3+, and Ni2+ ions using the sol–gel method as described by Hao et al. (Reference Hao, Lai, Xu, Liu and Ji2005). This study investigates the phase transitions that take place in the synthetic route of both the undoped material (Li4Ti5O12) and the doped congeners. The percentage phase composition, unit-cell parameters, crystallite sizes (LVol-IB), and strain for the doped and undoped spinel LTO were refined and compared for the patterns obtained for the in situ study from 50 to 850 °C, followed by cooling to 50 °C.

II. EXPERIMENTAL

The undoped (Li4Ti5O12) and doped spinel Li4Ti4.95M0.05O12 (M = Ni, Co, Mg, and Al) materials were prepared by a sol–gel method using triethanolamine as a complexing agent, similar to the methods described by Hao et al. (Reference Hao, Lai, Xu, Liu and Ji2005) and Lin et al. (Reference Lin, Hsu, Ho and Wu2013). Titanium butoxide was dissolved in triethanolamine in a 4:5 ratio in an ethanolic solution to form solution A. For the doped materials, aluminium acetate, magnesium acetate tetrahydrate, nickel acetate tetrahydrate, and cobalt acetate tetrahydrate were weighed out in order to achieve a concentration of 1% (moles in relation to the titanium ion) Al3+, Mg2+, Ni2+, and Co3+, respectively. The acetates were added to these solutions to form the respective materials Li4Ti4.95Al0.05O12, Li4Ti4.95Mg0.05O12, Li4Ti4.95Ni0.05O12, and Li4Ti4.95Co0.05O12 (Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005; Lin et al., Reference Lin, Hsu, Ho and Wu2013). Dopants chosen were based on those reported in the literature with improved material characteristics, which would lead to better electrochemical performance when used in a battery (Wang et al., Reference Wang, Jiang, Xiong, Wang and Jiao2013; Abureden et al., Reference Abureden, Hassan, Lui, Ahn, Sy, Yu and Chen2016; Liang et al., Reference Liang, Cao, Song, Gao, Hou, Guo and Qin2017; Ncube et al., Reference Ncube, Mhlongo, McCrindle and Zheng2018).

Lithium acetate dihydrate was dissolved in a 50:50 ethanol/water solution (solution B) and gradually dropped into the titanium solution until a clear mixture formed. The resulting mixture (solution C) was left to hydrolyze overnight until a gel-based precursor at room temperature formed. The precursor gels were placed in a drying oven at 85 °C for 6 h. Preparation of material for in situ analysis was done by further drying the precursor materials at 2 °C min−1 in a tube furnace until 300 °C was reached. PXRD analysis was done on the cooled room temperature material where its phase composition was determined to still be amorphous. The material was subsequently ground into a fine powder for easier sample mounting in the instrument. The in situ PXRD analysis was done on a Bruker D8 Advance with a Vantec detector, Cu radiation, and Goebel primary beam mirror. An Anton Parr XRK900 cell was used with a macor ceramic sample holder including a platinum (Pt) foil inset. In situ PXRD analysis entailed the heating of the precursor material in air (previously heated to 300 °C) in a chamber where the temperature was gradually increased by 10 °C min−1 from 30 to 850 °C. Upon heating, each diffraction pattern was collected at a fixed temperature every 50 °C intervals. The heating stage was kept isothermally for the duration of the analysis which lasted 10 min. The diffraction scan range was from 5–70° at 0.02° step−1. At 850 °C, the heating stage was kept isothermally for 4 h with a diffraction pattern collected every hour. Subsequently, the heating stage was cooled at 10 °C min−1 to room temperature with a diffraction pattern collected at 100 °C intervals. In situ phase identifications and quantification were done by full pattern Rietveld refinement using Topas v.6 software (Bruker, Reference Bruker2016). This was done by allowing the respective lattice parameters, crystallite sizes (LVol-IB), and strain to refine. Refinement of the unit-cell parameters and microstructural data were performed on patterns which were noticeable above the amorphous halo. The doped metal ions replaced the lithium sites at the 16d positions in the quantification of the doped material. The crystallographic open database (Downs and Hall-Wallace, Reference Downs and Hall-Wallace2003; Grazulis et al., Reference Grazulis, Chateigner, Downs, Yokochi, Quiros, Lutterotti, Manakova, Butkus, Moeck and Le Bail2009; Gražulis et al., Reference Gražulis, Daškevič, Merkys, Chateigner, Lutterotti, Quirós, Serebryanaya, Moeck, Downs and Le Bail2012; Gražulis et al., Reference Gražulis, Merkys, Vaitkus and Okulič-Kazarinas2015; Merkys et al., Reference Merkys, Vaitkus, Butkus, Okulič-Kazarinas, Kairys and Gražulis2016) was used for the respective refinements and included the open database files 1001098 (Li4Ti5O12) (Deschanvres et al., Reference Deschanvres, Raveau and Sekkal1971), 9004142 (rutile) (Meagher and Lager, Reference Meagher and Lager1979), 9009086 (anatase) (Wyckoff, Reference Wyckoff1963), and 1515995 (Li2TiO3) (Dorrian and Newnham, Reference Dorrian and Newnham1969).

Thermogravimetric analysis (TGA) of the precursor materials was done at 10 °C min−1 from room temperature to 800 °C in air using a TA instruments SDT Q600 and analysed using TA Universal Analysis v.4.5 software.

The Micrometrics Tristar II Brunauer–Emmett–Teller (BET) analyser was done to determine the surface area of the synthesized materials heated to 650 and 850 °C, respectively. Samples were degassed with nitrogen overnight at 350 °C.

III. RESULTS AND DISCUSSION

The weight loss with temperature TGA curves for the prepared materials show similarities between the undoped material and the magnesium-, nickel-, and cobalt-doped samples (Figure 2). They all show the main decomposition step between room temperature and ±400 °C of about 45–30%, which can be attributed to some of the water lost from the sol–gel precursor, as well as the loss of organic components. The second weight loss is seen above 400 °C for these materials, which is speculated to be further loss of organic-based materials as well as the oxidation of the final precursor compounds. Li4Ti4.95Al0.05O12 showed anomalous behaviour when compared to the other samples, as multiple weight loss steps were detected and at markedly lower temperatures. This can be seen by plotting the derivative weight loss with temperature, where the first mass loss observed was at approximately 100 °C for the Al-doped material (Figure 3) and can be ascribed to water loss. All five synthesized precursors show that around 550–600 °C no further weight loss from the samples were observed, which agrees with results obtained from the in situ studies, as the final material has been formed at this point, with slight variation in the temperature at which the spinel material is formed.

Figure 2. TGA curves of Li4Ti5O12 and Li4Ti4.95M0.05O12 (M = Al3+, Mg2+, Ni2+, and Co3+).

Figure 3. TGA derivative weight loss of undoped and doped spinel material.

The in situ PXRD three-dimensional graph shows the phase transformation for the undoped material after initial drying at 300 °C (Figure 4). The results show the gradual formation of the spinel as well as the anatase phase from a predominant amorphous phase when heated from 30 to 850 °C. The sample was kept isothermal at 850 °C for 4 h followed by the gradual cooling to a temperature of 50 °C. The metal-doped analogues showed similar in situ results, as can be seen from Figures 5–8.

Figure 4. In situ PXRD VT-scan of Li4Ti5O12 synthesized from the sol–gel precursor.

Figure 5. In situ PXRD VT-scan of Li4Ti4.95Al0.05O12 synthesized from the sol–gel precursor.

Figure 6. In situ PXRD VT-scan of Li4Ti4.95Co0.05O12 synthesized from the sol–gel precursor.

Figure 7. In situ PXRD VT-scan of Li4Ti4.95Mg0.05O12 synthesized from the sol–gel precursor.

Figure 8. In situ PXRD VT-scan of Li4Ti4.95Ni0.05O12 synthesized from the sol–gel precursor.

Selected diffraction patterns of the undoped LTO material are illustrated in Figure 9, which show the transition from the amorphous state through to the final spinel phase. The results show the amorphous material from 30 to 550 °C, with a series of peaks at low 2θ values that appeared in the patterns obtained at 30 [Figure 9(a)] and 200 °C [Figure 9(b)], respectively. These peaks around 9° 2θ disappeared once the sample was heated above 250 °C [Figure 9(c)]. Due to the low intensity of the peaks associated with this particular phase and the presence of the bulk amorphous material, the phase could not be accurately identified and quantified; however, it is speculated to be a type of titanium acid (H2Ti2O5·H2O) which was described in the literature to be a precursor material of typical titanate-based oxides (Song et al., Reference Song, Xu, Li, Wang and Yan2005; Etacheri et al., Reference Etacheri, Kuo, Van der Ven and Bartlett2013). The pattern obtained at 200 °C [Figure 9(b)] matched reasonably well with the diffraction pattern of H2Ti2O5·H2O [PDF 00-047-0124 (ICDD, 2002)] (Song et al., Reference Song, Xu, Li, Wang and Yan2005; Etacheri et al., Reference Etacheri, Kuo, Van der Ven and Bartlett2013).

Figure 9. Selected PXRD patterns of the Li4Ti5O12 gel precursor in staggered plot from the in situ results for (a) 30 °C, (b) 200 °C, (c) 550 °C, (d) 650 °C, (e) 750 °C, (f) 850 °C, and (g) 50 °C.

Titanium(IV) butoxide reacts with triethanolamine to form a stable titanium(IV) intermediate (Figure 10), which in turn reacts with water to form Ti(OH)4 (Sugimoto et al., Reference Sugimoto, Zhou and Muramatsu2003). According to the literature, this intermediate further forms TiO2 when heated (Sugimoto et al., Reference Sugimoto, Zhou and Muramatsu2003). It was suggested that the anatase then reacts with water to give H2Ti2O5·H2O (Figure 10). There are reported cases of TiO2 co-existing with H2Ti2O5·H2O; however, the mechanism of its formation has not been well described in the literature (Haro-González et al., Reference Haro-González, Pedroni, Piccinelli, Martín, Polizzi, Giarola, Mariotto, Speghini, Bettinelli and Martín2011; Arruda et al., Reference Arruda, Santos, Orlandi, Schreiner and Lisboa-Filho2015). Upon further heating, this phase subsequently disappears at temperatures above 250 °C, where the titanate phase of anatase and the spinel are seen to form [Figures 6(c) and 6(d)].

Figure 10. Reaction of titanium(IV) butoxide and triethanolamine to form titanium dioxide (Reaction A), which further reacts with water to form H2Ti2O5·H2O (Reaction B).

At 550 °C, the diffraction peaks for the two phases anatase (30.34%) and spinel Li4Ti5O12 (69.66%) could be reasonably quantified by full pattern Rietveld refinement. Upon further heating, the rutile phase appears (600 °C), with spinel LTO as the major phase (70.73%). The LTO crystalline phase is seen to form from 550 to 750 °C with relatively broad peaks [Figures 9(c) and 9(d)], where the sharper, more distinctive peaks can be seen from 850 °C [Figure 9(e)] onwards. There is a decrease in intensity observed upon cooling to room temperature for the main Li4Ti5O12 peak at around 18° 2θ [Figure 9(f)]; however, the relative amounts of the material remains unchanged throughout the cooling process. Since the crystallite size remains unchanged for the undoped material, the peak intensity decrease can possibly be attributed to sample displacement in the sample holder, due to shrinkage, combined with other cooling effects.

The LTO phase at 650 °C is the characteristic of material with relatively small crystallites based on the broad diffraction peaks observed. This correlates with BET surface area results, which show that material synthesized at 650 °C had, on average, a larger surface area (12.79 m2 g−1) when compared to its 850 °C analogue (0.44 m2 g−1) (Table I). Hence, the material with smaller crystallites would, on average, have a larger surface area. This can also be seen in the scanning electron microscopy (SEM) images that showed the material synthesized at 650 °C (Figure 11) to be significantly more porous than the material made at 850 °C (Figure 12) that showed a conglomeration of sintered particles. Energy-dispersive spectroscopy (EDS) analysis was done to show the uniform distribution of the dopant elements. The respective scans and their elemental mapping are shown in the Supplementary Material.

Figure 11. SEM images of the undoped LTO synthesized at 650 °C at (a) ×190, (b) ×250, (c) ×1500, and (d) ×2500 magnification.

Figure 12. SEM images of the undoped LTO synthesized at 850 °C at (a) ×27, (b) ×220, (c) ×1500, and (d) ×2500 magnification.

TABLE I. Summary of BET results and Rietveld refined parameters from in situ X-ray diffraction results.

The anatase phase disappeared at around 750 °C, with only the spinel LTO and the appearance of the rutile phase. The phase transition of the anatase to rutile phase transition is reported in the literature to be around 750 °C (Khatun et al., Reference Khatun, Anita, Rajput, Bhattacharya, Jha, Biring and Sen2017). Upon further heating, the rutile phase would slowly convert to the spinel phase where the final product, once cooled to 50 °C, would show a typical phase composition of 94.62% LTO and 5.38% rutile, respectively (Figure 13). The three main peaks of Li4Ti5O12 at around 17.5°, 35°, and 42.5° (Figure 13) can be assigned to crystallographic peaks (111), (311), and (400), respectively. There is a slight peak shift observed compared to the literature reported LTO due to sample cooling effects, which results in sample shrinkage and therefore sample displacement (Zhu et al., Reference Zhu, Liu, Yu, Zhang, Zhang, Jiang, Wang, Wang and Dong2019). The three distinct peaks at approximately 27° (110), 36° (101), and 54.5° (211) are ascribed to rutile still present in the material.

Figure 13. Rietveld refinement plot showing the phase composition of Li4Ti5O12 after being cooled to 50 °C. The blue line shows the experimental data and the red line shows the fitted data.

The quantitative phase transitions of the undoped LTO material with temperature can be summarized in Figure 14, showing a steady increase in the spinel phase from its formation at 550 up to 850 °C. Notably, the anatase phase would decrease above 550 °C with the appearance and increase of the rutile phase from 600 to 800 °C. The rutile phase in the sample would subsequently decrease when heated to 850 °C. By keeping the sample isothermally at 850 °C for 4 h, the rutile phase composition decreased by 45% from its initial value at 850 °C (13.32%). With the decreasing rutile content, an increase in the LTO phase is observed at the isothermal stage. This increase is still seen during the cooling process until 750 °C, after which a slight decrease is seen. This indicates that temperatures above 750 °C are sufficient to decrease the rutile content and increase the LTO phase, although 850°C would be more efficient.

Figure 14. Percentage phase composition of the spinel LTO material, rutile, and anatase from 550 to 850 °C, followed by cooling to room temperature.

Figure 15 illustrates the selected patterns from 30 up to 850 °C of the in situ results for Li4Ti4.95Ni0.05O12, where the material was heated to 300 °C prior to in situ studies to effectively dry the sample. The doped LTO material showed the amorphous material [Figure 15(a)] up to about 450 °C, with the potential H2Ti2O5.H2O phase (as described for Li4Ti5O12) also seen in the doped material [Figure 15(b)]; however, the titanium acid is only visible between 100 and 250 °C. Similar behaviour was seen for the other doped materials, as seen from Figures 5–8.

Figure 15. Selected PXRD patterns of the Li4Ti4.95Ni0.05O12 gel precursor in staggered plot from the in situ results for (a) 30 °C, (b) 200 °C, (c) 500 °C, (d) 750 °C, (e) 850 °C, and (f) 50 °C.

The phases of anatase and Li4Ti4.95Ni0.05O12 are seen between 450 and 500 °C [Figure 15(c)], with rutile appearing around 550–650 °C. This corresponds to the TGA results, which show the stable LTO products forming at approximately 500 °C (Figure 2). Anatase disappears upon keeping the material at the isothermal temperature [Figure 15(e)]; however, another phase appears with a relatively strong peak close to 43° 2θ. This peak was speculated to be the monoclinic β-Li2TiO3 structure (Figure 16), which is not uncommon in the synthesis of Li4Ti5O12 (Laumann, Reference Laumann2010; Bhatti et al., Reference Bhatti, Anjum, Ullah, Ahmed, Habib, Karim and Hasanain2016; Li et al., Reference Li, Guo, Ma, Yang, Dong, Yang, Yu, Jinxian and Liu2017). This was seen for all doped material; however, only appeared upon cooling for the cobalt-, nickel-, and manganese-doped lithium titanium oxides (Figure 17). The β-Li2TiO3 was only observed for the doped material, which indicates that this particular sol–gel synthetic route for the doped titanates does not allow for pure doped material to be isolated. This phase appears only upon cooling the material and further investigation should be done to study methods of removing this particular phase, although a mixed phase Li4Ti5O12/β-Li2TiO3 does not have a negative impact on the electrochemical characteristics of the anode material (Wang et al., Reference Wang, Zhou, Dai, Feng, Li and Li2014; Bhatti et al., Reference Bhatti, Anjum, Ullah, Ahmed, Habib, Karim and Hasanain2016).

Figure 16. PXRD pattern of Li4Ti4.95Al0.05O12 at 850 °C after 4 h showing the presence of β-Li2TiO3.

Figure 17. Percentage phase composition of the Mg-doped LTO material, rutile, β-Li2TiO3, and anatase from 500 to 850 °C, followed by cooling to room temperature.

The phase transitions of the doped LTO materials (Li4Ti4.95M0.05O12) show similar behaviour to the undoped LTO with an initial slight decrease in the percentage phase composition of the doped LTO spinel phase between 550 and 750 °C (Figure 18). This is followed by an increase in the spinel phase composition between 750 and 850 °C with a further increase during the 4 h isothermal at 850 °C. The Al-doped spinel shows different behaviour at the isothermal stage, with the β-Li2TiO3 phase forming at this point already, resulting in a lower spinel phase composition compared to the other material. The change in the LTO spinel phase formation during the 850 °C isothermal period can be described by the Johnson–Mehl–Avrami–Kolmogorov (JMAK) kinetic equation, as illustrated in the equation [see Eq. (2)], where P (%) is the percentage phase composition of the doped/undoped LTO phase, t is the time in hours, A is the Avrami exponent, and C is the time-dependent constant.

(2)$$P_{\lpar \percnt \rpar } = 1-e^{{\left({{t \over {\rm C}}} \right)}^{\rm A}}$$

Figure 18. Percentage phase composition of the doped and undoped spinel LTO phase from 550 to 50 °C, showing error bars.

This kinetic relationship is often used to describe crystal growth (Starink, Reference Starink2001; Ahmad et al., Reference Ahmad, Kanaujia, Niu, Baumberg and Vijaya Prakash2014) and can be transformed into its linear form [see Eq. (3)] to solve the constants A and C from the slope and y-axis intercept, respectively.

(3)$$\ln \left({\ln \left({\displaystyle{1 \over {1-P_{\lpar \percnt \rpar }}}} \right)} \right) = {\rm A}\ln t-{\rm A}\ln {\rm C}$$

The calculated Avrami exponent and time-dependent constant values for the synthesized materials are tabulated in Table II, with the sigmoidal curve illustrated in Figure 19 which depict both the calculated and experimental phase formation as a function of time. The predicted kinetic behaviour of the undoped material, as well as the nickel-, manganese-, and cobalt-doped analogues, show that the material should form a highly pure spinel phase if allowed to react up to 24 h at a constant temperature of 850 °C. This suggests that after drying the precursor material in the synthetic route of the lithium titanates to 300 °C, the material should be slowly heated to 850 °C and remain there for a minimum of 24 h to effectively remove any rutile present. Rutile can negatively impact the battery performance due to its influence on lithium-ion diffusion (Syahrial et al., Reference Syahrial, Priyono, Yuwono, Kartini, Jodi and Johansyah2016) and should, therefore, ideally be removed completely. The results are based on an in situ study only and laboratory experiments or large-scale processes could yield different results; however, based on the literature, extensive heating does reduce the rutile content or eliminates it completely (Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005; Mahmoud et al., Reference Mahmoud, Amarilla and Saadoune2015; Sun et al., Reference Sun, Radovanovic and Cui2015). A typical literature reports on calcining such samples at 800 °C and higher for 12–24 h (Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005; Sun et al., Reference Sun, Radovanovic and Cui2015). Some have found that an hour at 900 °C was also sufficient to eliminate the rutile phase in the LTO material (Mahmoud et al., Reference Mahmoud, Amarilla and Saadoune2015). The aluminium-doped material indicated that a fair amount of the rutile phase was still present at the high temperatures and would take a significantly long time to obtain the pure product. The β-Li2TiO3 also appears at the isothermal temperature, which results in much lower pure LTO, therefore for the Al-doped material, the decrease in rutile is shown in Figure 19. This was different from what was reported in the literature (Lin et al., Reference Lin, Hsu, Ho and Wu2013) and the difference could be ascribed to the analytical method used for the particular sample. Li4Ti4.95Al0.05O12 was synthesized using a different sol–gel methodology (Hao et al., Reference Hao, Lai, Xu, Liu and Ji2005) than that reported in the literature, where the literature reported doped material showed no rutile when calcining at 800 °C for 12 h (Lin et al., Reference Lin, Hsu, Ho and Wu2013). The presence of β-Li2TiO3, according to the literature, can effectively be removed by calcining at 900 °C for at least 2 h (Laumann et al., Reference Laumann, Boysen, Bremholm, Fehr, Hoelzel and Holzapfel2011), which explains why it was detected as a minor impurity in the doped material.

Figure 19. The calculated LTO phase formation curves with their experimental data from in situ analysis for undoped, Co-, Mg-, and Ni-doped material. The decreased rutile is shown for the Al-doped analogue on the secondary axis.

TABLE II. Summary of JMAK constants and predicted kinetics.

a Based on the decrease in rutile phase.

The proposed final chemical composition of the doped spinel material can be found in Table III, which illustrates that the final proposed chemical composition of the material differs slightly from the nominal formula (as reported throughout the paper), due to the oxygen vacancy created by the substitution of dopants with lower valences into Ti4+ sites, as required by the charge neutrality condition (Hao et al., Reference Hao, Li, Li, Dai and Wang2008; Wang et al., Reference Wang, Chen, Xu, Lv and Yang2011). The Mg2+ and Ni2+ will result in a larger oxygen vacancy, compared to Al3+ and Co3+, as evident from the proposed formula (Table III). The crystallite sizes (LVol-IB) as determined by the full pattern Rietveld refinement for both the undoped and doped LTOs are graphically illustrated in Figure 20, from 550 °C where the spinel phase was initially identified up to 850 °C, and subsequent cooling to 50 °C. For the undoped LTO, the change in the crystallite size with temperature showed an approximate 0.120 nm °C−1 increase when heating from 550 to 850 °C. A further increase to 72.21 nm was observed when leaving the sample at 850 °C for 4 h. Similar behaviour was seen for the doped LTO analogues; however, by comparison, the aluminium, cobalt, and magnesium-doped spinel material showed significantly higher crystallite sizes over the temperature range. As the material is kept at the isothermal temperature (850 °C), the crystallite size is seen to increase, which suggests that an increase in reaction time results in an increase crystallite size and has been observed in lithium titanates before (Nugroho et al., Reference Nugroho, Chung and Kim2014). Cooling the undoped material shows little to no changes in crystallite size; however, the doped material shows a decrease, which can be due to the combined effect of rapid cooling of the material, sample shrinkage and the lattice constraints that is induced upon doping (Figure 17) (Hatfield et al., Reference Hatfield, Eades and McClellan1971; Karhunen et al., Reference Karhunen, Välikangas, Torvela, Lähde and Lassi2016). During the least-squares refinement process in Topas, the L-strain parameter was allowed to refine, which increased as the LTO formation progresses from 550 to 850 °C. Once the material reaches the isothermal temperature range, the strain parameter refined to a minimal value and was relatively constant until the material was cooled to 50 °C. Similar trends in the unit-cell lattice parameters for Li4Ti5O12- and Li4Ti4.95M0.05O12-doped materials were observed with a slight increase (0.63–0.75%) over the temperature range from 550 to 850 °C. A decrease between 0.085  and 0.11 Å in unit-cell lattice parameters were observed when the samples were cooled to 50 °C (Figure 21). The change in unit-cell parameter directly correlates to the change in temperature (Huynh et al., Reference Huynh, Ha, Nguyen, Nguyen, Le and Man Tran2019), due to the unit-cell expansion or contraction upon heating or cooling. At the constant temperature of 850 °C, a relatively constant unit-cell lattice parameter was observed. It is expected that the unit-cell parameter will decrease upon doping with the Co3+ and Al3+ as it has a smaller ionic radii compared to that of Ti4+ (Park et al., Reference Park, Baek, Park and Kim2014; Guo et al., Reference Guo, Wang, Chen, Xu, Wu and Li2016); which correlates with the results found (Figure 21). The decreasing unit-cell parameter upon doping with larger ions Mg2+ and Ni2+ shows anomalous behaviour as it is expected to increase (Li et al., Reference Li, Zeng, Li and Xu2015; Abureden et al., Reference Abureden, Hassan, Lui, Ahn, Sy, Yu and Chen2016), which can be the result of the difference valences of the titanium ions and the doping cations. The strain parameter determined in this study is relative. It shows that the material upon heating and when undergoing phase transitions, simultaneously undergoes crystallite size growth and stress/strain properties within its crystal domain.

Figure 20. Crystallite size (LVol IB) (nm) of the in situ X-ray diffraction results for the undoped and doped lithium titanium oxide material.

Figure 21. Unit-cell lattice parameter (Å) of the in situ X-ray diffraction results for the undoped and doped lithium titanium oxide material, showing error bars.

TABLE III. Final proposed and nominal chemical composition for doped materials.

IV. CONCLUSION

The in situ PXRD temperature studies of forming the spinel LTO and its doped analogues made from its precursor gel-based materials showed that it was predominantly amorphous from room temperature up to 450 °C, with a small amount of an intermediate crystalline phase that was identified to be a type of titanium acid (H2Ti2O5·H2O). Upon further heating, the intermediate phase disappeared and the formation of anatase and the spinel LTO was observed. Rutile appeared to form around 550 to 650 °C, with sharp, distinct peaks of the spinel phase only seen at 850 °C coupled with the disappearance of the anatase phase. The broader diffraction peaks of the spinel phase were observed around 650 °C and would typically correspond to a material with a small crystallite size. This was confirmed by observing a significant decrease in the material's surface areas that were synthesized at 650 and 850 °C, as well as SEM images of the respective materials, where a conglomeration of the crystals was observed for the material made at the higher temperature.

When maintaining the material isothermally at 850 °C for a number of hours, a decrease of the rutile phase was observed that could be described by a JMAK-type kinetic model. A pure LTO phase could not be isolated, as the rutile co-crystallized with the spinel material, although in minimal amounts after cooling to room temperature. The doped material had an additional titanate phase (β-Li2TiO3) co-crystallizing, which became more pronounced upon cooling the material, which is not uncommon in the synthesis of Li4Ti5O12. Rietveld refinement of the diffraction patterns showed that the unit-cell parameters of the spinel typically increased upon heating to 850 °C, followed by a decrease upon cooling that was in agreement of published unit-cell parameters of the material at room temperature.

SUPPLEMENTARY MATERIAL

The supplementary material for this article can be found at https://doi.org/10.1017/S0885715620000494.

ACKNOWLEDGEMENTS

The authors express their gratitude towards the National Research Foundation and Nelson Mandela University for their financial support.

References

Abureden, S., Hassan, F. M., Lui, G., Ahn, W., Sy, S., Yu, A., and Chen, Z. (2016). “Multigrain electrospun nickel doped lithium titanate nanofibers with high power lithium ion storage,” J. Mater. Chem. A 4, 12638–12647.CrossRefGoogle Scholar
Ahmad, S., Kanaujia, P., Niu, W., Baumberg, J., and Vijaya Prakash, G. (2014). “In-situ intercalation dynamics in inorganic-organic layered perovskite thin films,” ACS Appl. Mater. Interfaces 6, 10238–10247.CrossRefGoogle ScholarPubMed
Arruda, L. B., Santos, C. M., Orlandi, M. O., Schreiner, W. H., and Lisboa-Filho, P. N. (2015). “Formation and evolution of TiO2 nanotubes in alkaline synthesis,” Ceram. Int. 41, 2884–2891.CrossRefGoogle Scholar
Bhatti, H. S., Anjum, D. H., Ullah, S., Ahmed, B., Habib, A., Karim, A., and Hasanain, S. K. (2016). “Electrochemical characteristics and Li+ ion intercalation kinetics of dual-phase Li4Ti5O12/Li2TiO3 composite in the voltage range 0–3 V,” J. Phys. Chem. C 120, 9553–9561.CrossRefGoogle Scholar
Bruker, A. (2016). Topas Manual, Version 6.0 (Computer Software) (Bruker Software, West Germany).Google Scholar
Coelho, J., Pokle, A., Park, S. H., McEvoy, N., Berner, N. C., Duesberg, G. S., and Nicolosi, V. (2017). “Lithium titanate/carbon nanotubes composites processed by ultrasound irradiation as anodes for lithium ion batteries,” Sci. Rep. 7, 7614.CrossRefGoogle ScholarPubMed
Deschanvres, A., Raveau, B., and Sekkal, Z. (1971). “Mise en evidence et etude cristallographique d'une nouvelle solution solide de type spinelle Li1+xTi2−xO4 0 ⩽ x ⩽ 0, 333,” Mater. Res. Bull. 6, 699–704.CrossRefGoogle Scholar
Dorrian, J. F., and Newnham, R. E. (1969). “Refinement of the structure of Li2TiO3,” Mater. Res. Bull. 4, 179–183.CrossRefGoogle Scholar
Downs, R. T., and Hall-Wallace, M. (2003). “The American mineralogist crystal structure database,” Am. Mineral. 88, 247–250.Google Scholar
Etacheri, V., Kuo, Y., Van der Ven, A., and Bartlett, B. M. (2013). “Mesoporous TiO2–B microflowers composed of (1 1 0) facet-exposed nanosheets for fast reversible lithium-ion storage,” J. Mater. Chem. A 1, 12028–12032.CrossRefGoogle Scholar
Ganapathy, S., and Wagemaker, M. (2012). “Nanosize storage properties in spinel Li4Ti5O12 explained by anisotropic surface lithium insertion,” ACS Nano 6, 8702–8712.CrossRefGoogle ScholarPubMed
Grazulis, S., Chateigner, D., Downs, R. T., Yokochi, A. F. T., Quiros, M., Lutterotti, L., Manakova, E., Butkus, J., Moeck, P., and Le Bail, A. (2009). “Crystallography Open Database – an open-access collection of crystal structures,” J. Appl. Cryst. 42, 726–729.CrossRefGoogle ScholarPubMed
Gražulis, S., Daškevič, A., Merkys, A., Chateigner, D., Lutterotti, L., Quirós, M., Serebryanaya, N. R., Moeck, P., Downs, R. T., and Le Bail, A. (2012). “Crystallography Open Database (COD): an open-access collection of crystal structures and platform for world-wide collaboration,” Nucleic Acids Res. 40, D420–D427.CrossRefGoogle ScholarPubMed
Gražulis, S., Merkys, A., Vaitkus, A., and Okulič-Kazarinas, M. (2015). “Computing stoichiometric molecular composition from crystal structures,” J. Appl. Cryst. 48, 85–91.CrossRefGoogle ScholarPubMed
Guo, Q., Wang, Q., Chen, G., Xu, H., Wu, J., and Li, B. (2016). “Molten salt synthesis of transition metal oxides doped Li4Ti5O12 as anode material of Li-Ion battery,” ECS Trans. 72, 11–23.CrossRefGoogle Scholar
Han, S. W., Ryu, J. H., Jeong, J., and Yoon, D. H. (2013). “Solid-state synthesis of Li4Ti5O12 for high power lithium ion battery applications,” J. Alloys Compd. 570, 144–149.CrossRefGoogle Scholar
Hao, Y., Lai, Q., Xu, Z., Liu, X., and Ji, X. (2005). “Synthesis by TEA sol–gel method and electrochemical properties of Li4Ti5O12 anode material for lithium-ion battery,” Solid State Ion. 176, 1201–1206.CrossRefGoogle Scholar
Hao, G., Li, N., Li, D., Dai, C., and Wang, D. (2008). “Study on the effect of Li doping in spinel Li4+xTi5-xO12 (0≤x≥0.2) materials for lithium-ion batteries,” Electrochem. Commun. 10, 1031–1034.Google Scholar
Haro-González, P., Pedroni, M., Piccinelli, F., Martín, L. L., Polizzi, S., Giarola, M., Mariotto, G., Speghini, A., Bettinelli, M., and Martín, I. R. (2011). “Synthesis, characterization and optical spectroscopy of Eu3+ doped titanate nanotubes,” J. Lumin. 131, 2473–2477.CrossRefGoogle Scholar
Hatfield, J. D., Eades, J. L., and McClellan, G. H. (1971). Investigation of the Reactivities of Limestone to Remove Sulfur Dioxide from Fuel Gas (Tennessee Valley Authority, Division of Chemical Development, Fundamental Research Branch, Muscle Shoals, Alabama).Google Scholar
Huynh, L. T. N., Ha, C. T. D., Nguyen, V. D., Nguyen, D. Q., Le, M. L. P., and Man Tran, V. (2019). “Structure and electrochemical properties of Li4Ti5O12 prepared via low-temperature precipitation,” J. Chem. 2019, 7.CrossRefGoogle Scholar
ICDD (2002). PDF-2 2004 (Database), edited by Dr. Soorya Kabekkodu, International Centre for Diffraction Data, Newtown Square, PA, USA.Google Scholar
Karhunen, T., Välikangas, J., Torvela, T., Lähde, A., and Lassi, U. (2016). “Effect of doping and crystallite size on the electrochemical performance of Li4TI5O12,” J. Alloys Compd. 659, 132–137.CrossRefGoogle Scholar
Khatun, N., Anita, , Rajput, P., Bhattacharya, D., Jha, S. N., Biring, S., and Sen, S. (2017). “Anatase to rutile phase transition promoted by vanadium substitution in TiO2: a structural, vibrational and optoelectronic study,” Ceram. Int. 43, 14128–14134.CrossRefGoogle Scholar
Koichi Momma, F. I. (2011). “VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data,” J. Appl. Crystallogr. 44, 5.Google Scholar
Kuo, Y. C., Peng, H. T., Xiao, Y., and Lin, J. Y. (2016). “Effect of starting materials on electrochemical performance of sol-gel-synthesized Li4Ti5O12 anode materials for lithium-ion batteries,” J. Solid State Electr. 20, 1625–1631.CrossRefGoogle Scholar
Laumann, A. (2010). Novel Routes to Li4Ti5O12 Spinel: Characterization and Phase Relations (Ludwig Maximilian University of Munich, Munich).Google Scholar
Laumann, A., Boysen, H., Bremholm, M., Fehr, K. T., Hoelzel, M., and Holzapfel, M. (2011). “Lithium migration at high temperatures in Li4Ti5O12 studied by neutron diffraction,” Chem. Mater. 23, 2753–2759.CrossRefGoogle Scholar
Li, N., Liang, J., Wei, D., Zhu, Y., and Qian, Y. (2014). “Solvothermal synthesis of micro-/nanoscale Cu/Li4Ti5O12 composites for high rate Li-ion batteries,” Electrochim. Acta 123, 346–352.CrossRefGoogle Scholar
Li, F., Zeng, M., Li, J., and Xu, H. (2015). “Preparation and electrochemical performance of Mg-doped Li4Ti5O12 nanoparticles as anode materials for lithium-ion batteries,” Int. J. Electrochem. Sci. 10, 10445–10453.Google Scholar
Li, S., Guo, J., Ma, Q., Yang, Y., Dong, X., Yang, M., Yu, W., Jinxian, W., and Liu, G. (2017). “Electrospun Li4Ti5O12/Li2TiO3 composite nanofibers for enhanced high-rate lithium ion batteries,” J. Solid State Electrochem. 21, 2779–2790.CrossRefGoogle Scholar
Liang, Q., Cao, N., Song, Z., Gao, X., Hou, L., Guo, T., and Qin, X. (2017). “Co-doped Li4Ti5O12 nanosheets with enhanced rate performance for lithium-ion batteries,” Electrochim. Acta 251, 407–414.CrossRefGoogle Scholar
Lin, J. Y., Hsu, C. C., Ho, H. P., and Wu, S. H. (2013). “Sol–gel synthesis of aluminum doped lithium titanate anode material for lithium ion batteries,” Electrochim. Acta 87, 126–132.CrossRefGoogle Scholar
Mahmoud, A., Amarilla, J. M., and Saadoune, I. (2015). “Effect of thermal treatment used in the sol–gel synthesis of Li4Ti5O12 spinel on its electrochemical properties as anode for lithium ion batteries,” Electrochim. Acta 163, 213–222.CrossRefGoogle Scholar
Meagher, E. P., and Lager, G. A. (1979). “Polyhedral thermal expansion in the TiO2 polymorphs; refinement of the crystal structures of rutile and brookite at high temperature,” Can. Mineral. 17, 77–85.Google Scholar
Merkys, A., Vaitkus, A., Butkus, J., Okulič-Kazarinas, M., Kairys, V., and Gražulis, S. (2016). “COD::CIF::Parser: an error-correcting CIF parser for the Perl language,” J. Appl. Cryst. 49, 292–301.CrossRefGoogle ScholarPubMed
Ming, H., Ming, J., Li, X., Zhou, Q., Wang, H., Jin, L., Fu, Y., Adkins, J., and Zheng, J. (2014). “Hierarchical Li4Ti5O12 particles co-modified with C&N towards enhanced performance in lithium-ion battery applications,” Electrochim. Acta 116, 224–229.CrossRefGoogle Scholar
Mosa, J., Vélez, J. F., Reinosa, J. J., Aparicio, M., Yamaguchi, A., Tadanaga, K., and Tatsumisago, M. (2013). “Li4Ti5O12 thin-film electrodes by sol–gel for lithium-ion microbatteries,” J. Power Sources 244, 482–487.CrossRefGoogle Scholar
Ncube, N. M., Mhlongo, W. T., McCrindle, R. I., and Zheng, H. (2018). “The electrochemical effect of Al-doping on Li4Ti5O12 as anode material for lithium-ion batteries,” Mater. Today Proc. 5, 10592–10601.CrossRefGoogle Scholar
Ni, H., Song, W., and Fan, L. (2016). “Double carbon decorated lithium titanate as anode material with high rate performance for lithium-ion batteries,” Proc. Natl. Sci. Mater. Int. 26, 283–288.CrossRefGoogle Scholar
Nugroho, A., Chung, K. Y., and Kim, J. (2014). “A facile supercritical alcohol route for synthesizing carbon coated hierarchically mesoporous Li4Ti5O12 microspheres,” J. Phys. Chem. C 118, 183–193.CrossRefGoogle Scholar
Park, J. S., Baek, S. H., Park, Y., and Kim, J. H. (2014). “Improving the electrochemical properties of Al, Zr Co-doped Li4Ti5O12 as a lithium-ion battery anode material,” J. Korean Phys. Soc. 64, 1545–1549.CrossRefGoogle Scholar
Priyono, B., Syahrial, A. Z., Herman Yuwono, A., Kartini, E., Marfelly, M., and Furkon Rahmatulloh, W. M. (2015). “Synthesis of lithium titanate (Li4Ti5O12) through hydrothermal process by using lithium hydroxide (LiOH) and titanium dioxide (TiO2) xerogel,” Int. J. Technol. 6, 555–564.CrossRefGoogle Scholar
Rai, A. K., Gim, J., Kang, S.-W., Mathew, V., Anh, L. T., Kang, J., Song, J., Paul, B. J., and Kim, J. (2012). “Improved electrochemical performance of Li4Ti5O12 with a variable amount of graphene as a conductive agent for rechargeable lithium-ion batteries by solvothermal method,” Mater. Chem. Phys. 136, 1044–1051.CrossRefGoogle Scholar
Shen, C. M., Zhang, X. G., Zhou, Y. K., and Li, H. I. (2003). “Preparation and characterization of nanocrystalline Li4Ti5O12 by sol–gel method,” Mater. Chem. Phys. 78, 437–441.CrossRefGoogle Scholar
Snyders, C. D., Ferg, E. E., and Billing, D. (2016). “An investigation into the temperature phase transitions of synthesized materials with Al- and Mg-doped lithium manganese oxide spinels by in-situ powder X-ray diffraction,” Powder Diffr. 32, 23–30.CrossRefGoogle Scholar
Song, Z., Xu, H., Li, K., Wang, H., and Yan, H. (2005). “Hydrothermal synthesis and photocatalytic properties of titanium acid H2Ti2O5⋅H2O nanosheets,” J. Mol. Catal. A Chem. 239, 87–91.CrossRefGoogle Scholar
Starink, M. (2001). “On the meaning of the impingement parameter in kinetic equations of diffusion controlled reactions,” J. Mater. Sci. 36, 4433–4441.CrossRefGoogle Scholar
Sugimoto, T., Zhou, X., and Muramatsu, A. (2003). “Synthesis of uniform anatase TiO2 nanoparticles by gel–sol method: 3. Formation process and size control,” J. Colloid Interface Sci. 259, 43–52.CrossRefGoogle ScholarPubMed
Sun, X., Radovanovic, P. V., and Cui, B. (2015). “Advances in spinel Li4Ti5O12 anode materials for lithium-ion batteries,” New J. Chem. 39, 38–63.CrossRefGoogle Scholar
Syahrial, A. Z., Priyono, B., Yuwono, A. H., Kartini, E., Jodi, H., and Johansyah, J. (2016). “Synthesis of lithium titanate (Li4Ti5O12) by addition of excess lithium carbonate (Li2CO3) in titanium dioxide (TiO2) xerogel,” Int. J. Technol. 7, 392–400.CrossRefGoogle Scholar
Wang, Z., Chen, G., Xu, J., Lv, Z., and Yang, W. (2011). “Synthesis and electrochemical performances of Li4Ti4.95Al0.05O12/C as anode material for lithium-ion batteries,” J. Phys. Chem. Solids 72, 773–778.CrossRefGoogle Scholar
Wang, W., Jiang, B., Xiong, W., Wang, Z., and Jiao, S. (2013). “A nanoparticle Mg-doped Li4Ti5O12 for high rate lithium-ion batteries,” Electrochim. Acta 114, 198–204.CrossRefGoogle Scholar
Wang, Y., Zhou, A., Dai, X., Feng, L., Li, J., and Li, J. (2014). “Solid-state synthesis of submicron-sized Li4Ti5O12/Li2TiO3 composites with rich grain boundaries for lithium ion batteries,” J. Power Sources 266, 114–120.CrossRefGoogle Scholar
Wang, C., Wang, S., He, Y.-B., Tang, L., Han, C., Yang, C., Wagemaker, M., Li, B., Yang, Q.-H., Kim, J.-K., and Kang, F. (2015). “Combining fast Li-ion battery cycling with large volumetric energy density: grain boundary induced high electronic and ionic conductivity in Li4Ti5O12 spheres of densely packed nanocrystallites,” Chem. Mater. 27, 5647–5656.CrossRefGoogle Scholar
Wyckoff, R. W. G. (1963). Crystal Structures (New York, NY), 2nd ed., pp. 239–444.Google Scholar
Xie, Z., Li, X., Li, W., Chen, M., and Qu, M. (2015). “Graphene oxide/lithium titanate composite with binder-free as high capacity anode material for lithium-ion batteries,” J. Power Sources 273, 754–760.CrossRefGoogle Scholar
Yi, T. F., Jiang, L. J., Shu, J., Yue, C. B., Zhu, R. S., and Qiao, H. B. (2010). “Recent development and application of Li4Ti5O12 as anode material of lithium ion battery,” J. Phys. Chem. Solids 71, 1236–1242.CrossRefGoogle Scholar
Zhang, Z., Cao, L., Huang, J., Wang, D., Wu, J., and Cai, Y. (2013). “Hydrothermal synthesis of Li4Ti5O12 microsphere with high capacity as anode material for lithium ion batteries,” Ceram. Int. 39, 2695–2698.CrossRefGoogle Scholar
Zhu, C., Liu, J., Yu, X., Zhang, Y., Zhang, Y., Jiang, X., Wang, S., Wang, Q., and Dong, P. (2019). “Enhance the electrochemical performance of Li4Ti5O12 with Co doping via a facile mechanical activation strategy,” J. Mater. Sci. Mater. Electron. 30, 5866–5873.CrossRefGoogle Scholar
Figure 0

Figure 1. Structure of spinel Li4Ti5O12 showing atoms in the unit cell only (left) and unit cell with the polyhedra (right) where the tetrahedra indicates the 8a sites occupied by Li+ and the octahedra indicates the 16d sites that are occupied by Li+ and Ti4+. The O2− anions occupy the octahedral 32e sites (Deschanvres et al., 1971; Koichi Momma, 2011).

Figure 1

Figure 2. TGA curves of Li4Ti5O12 and Li4Ti4.95M0.05O12 (M = Al3+, Mg2+, Ni2+, and Co3+).

Figure 2

Figure 3. TGA derivative weight loss of undoped and doped spinel material.

Figure 3

Figure 4. In situ PXRD VT-scan of Li4Ti5O12 synthesized from the sol–gel precursor.

Figure 4

Figure 5. In situ PXRD VT-scan of Li4Ti4.95Al0.05O12 synthesized from the sol–gel precursor.

Figure 5

Figure 6. In situ PXRD VT-scan of Li4Ti4.95Co0.05O12 synthesized from the sol–gel precursor.

Figure 6

Figure 7. In situ PXRD VT-scan of Li4Ti4.95Mg0.05O12 synthesized from the sol–gel precursor.

Figure 7

Figure 8. In situ PXRD VT-scan of Li4Ti4.95Ni0.05O12 synthesized from the sol–gel precursor.

Figure 8

Figure 9. Selected PXRD patterns of the Li4Ti5O12 gel precursor in staggered plot from the in situ results for (a) 30 °C, (b) 200 °C, (c) 550 °C, (d) 650 °C, (e) 750 °C, (f) 850 °C, and (g) 50 °C.

Figure 9

Figure 10. Reaction of titanium(IV) butoxide and triethanolamine to form titanium dioxide (Reaction A), which further reacts with water to form H2Ti2O5·H2O (Reaction B).

Figure 10

Figure 11. SEM images of the undoped LTO synthesized at 650 °C at (a) ×190, (b) ×250, (c) ×1500, and (d) ×2500 magnification.

Figure 11

Figure 12. SEM images of the undoped LTO synthesized at 850 °C at (a) ×27, (b) ×220, (c) ×1500, and (d) ×2500 magnification.

Figure 12

TABLE I. Summary of BET results and Rietveld refined parameters from in situ X-ray diffraction results.

Figure 13

Figure 13. Rietveld refinement plot showing the phase composition of Li4Ti5O12 after being cooled to 50 °C. The blue line shows the experimental data and the red line shows the fitted data.

Figure 14

Figure 14. Percentage phase composition of the spinel LTO material, rutile, and anatase from 550 to 850 °C, followed by cooling to room temperature.

Figure 15

Figure 15. Selected PXRD patterns of the Li4Ti4.95Ni0.05O12 gel precursor in staggered plot from the in situ results for (a) 30 °C, (b) 200 °C, (c) 500 °C, (d) 750 °C, (e) 850 °C, and (f) 50 °C.

Figure 16

Figure 16. PXRD pattern of Li4Ti4.95Al0.05O12 at 850 °C after 4 h showing the presence of β-Li2TiO3.

Figure 17

Figure 17. Percentage phase composition of the Mg-doped LTO material, rutile, β-Li2TiO3, and anatase from 500 to 850 °C, followed by cooling to room temperature.

Figure 18

Figure 18. Percentage phase composition of the doped and undoped spinel LTO phase from 550 to 50 °C, showing error bars.

Figure 19

Figure 19. The calculated LTO phase formation curves with their experimental data from in situ analysis for undoped, Co-, Mg-, and Ni-doped material. The decreased rutile is shown for the Al-doped analogue on the secondary axis.

Figure 20

TABLE II. Summary of JMAK constants and predicted kinetics.

Figure 21

Figure 20. Crystallite size (LVol IB) (nm) of the in situ X-ray diffraction results for the undoped and doped lithium titanium oxide material.

Figure 22

Figure 21. Unit-cell lattice parameter (Å) of the in situ X-ray diffraction results for the undoped and doped lithium titanium oxide material, showing error bars.

Figure 23

TABLE III. Final proposed and nominal chemical composition for doped materials.

Supplementary material: File

van Niekerk et al. supplementary material

Figures S1-S12

Download van Niekerk et al. supplementary material(File)
File 853.8 KB