Raman Scattering Study of La_2/3-xLi_3x_1/3-2xTiO₃ : Relationship between Spectra Parameters and Amount of Lithium

Raman Scattering Study of La_2/3-xLi_3x1/3-2xTiO₃ : Relationship between Spectra Parameters and Amount of Lithium

H. M. Omanda,1,2*, H. Gnanga1, P. Soulounganga1, R. Ondo Ndong1, A. Eyaa-Mvongbote1,

Z.H. Moussambi Membetsi1 And Alain Bulou2

1Laboratoire Pluridisciplinaire des Sciences (LAPLUS), Ecole Normale Supérieure BP 17009 Libreville, Gabon.

2Laboratoire de Physique de l'Etat Condensé (UMR 6087 CNRS), Université du Maine, Av. O. Messiaen, 72085 LE MANS Cedex 9, France

Abstract

The Raman-active lattice modes in lithium lanthanum titanates La2/3-xLi3x1/3-2xTiO3 (LLTO) at room temperature were investigated using a Raman spectrometer of limiting resolution 0.005 cm-1. Particular attention was paid to search a correlation between Raman parameters (positions, linewidth and peaks intensity spectra) and x content. The characterization of LLTO by Raman spectroscopy made it possible to identify Eg modes predicted by the groups theory. The attribution of the modes A1g and B1g could not be done in a precise way because the spectra profile obtained is complex. The variations study of the frequencies, the linewidths and the amplitudes according to x content shows a correlation as regards the vibrations frequencies. We observes a growth of the frequency for the weak concentrations up to x=0.10, then a decrease for higher rates. This behavior is similar to that observed for the variation of conductivity according to the lithium rate.

Keywords: Lithium Ion conduction, Raman spectroscopy, group theorical, vibration modes

1. Introduction

   The series of lithium lanthanum titanates La2/3-xLi3x1/3-2xTiO3 (LLTO), solid solution in the perovskite type ABO3, have received a considerable attention since several years. Belous and al.[1] and Inaguma and al.[2] have shown the high ionic conduction of LLTO. It is owing to the lithium ion mobility and vacancies which induce a structural distortion. In addition, it has been reported that lithium lanthanum titanates show high ionic conductivity as high as 10-3 S cm-1 at room temperature [2]. Therefore a particular interest was related to LLTO because of their

   potential applications in electrochemical devices such as high-energy lithium ion batteries, electrochromic systems, supercapacitors, and electrochemical sensors[3,4]. In the electrochemical devices, LLTO can be used as electrolyte material in the solid-state batteries[5, 6] or as protecting thin coating against the exothermic phenomena in lithium ion batteries with liquid electrolyte [7].

   From a structural point of view, the research undertaken on the LLTO remains discussed as well range of composition x of the solid solution as crystallographic structure [8-12]. We retain for our study the existence of the solid phase of type perovskite ABO3 in the range of composition of 0.06< x <0.14 with a crystallographic structure of group of space P4/mmm [9, 11].

   In a general way we can note that the structure of La2/3-xLi3x1/3-2xTiO3 drifting of a structure perovskite of the lacunar type ABO3, with representing a vacancies, varies according to composition x, but as of the heat treatment as the material could undergo during its development. Several techniques of syntheses of LLTO are reported by the literature among which deposits in thin layer [2, 8-15].

   In order to understand the relationship between the Raman spectra parameters and the x composition of La2/3-xLi3x1/3-2xTiO3, the measurement were leaded on seven samples of LLTO in the range 0.06<x<0.14.

2. Experimental and theorical analysis

   Preparation of La2/3-xLi3x1/3-2xTiO3 sample: LLTO powder samples were obtained using the procedure of Fourquet et al.[9]: the compositions corresponding to these series were synthesized from stoichiometric amounts of La2O3, TiO2, and Li2CO3 (high purity grade). The reagents were mixed and pressed into pellets and first heated in air for 4h at 850°C in

   platinum boats. After regrinding and repressing, the samples were heated three times for 10 h at 1150°C and were allowed to cool in furnace (naturally down to 200°C).

   Structural data: Structural data: The structure of

   La2/3-xLi3x1/3-2xTiO3 series has been studied by Fourquet et al.[9]. They have shown that, at room temperature, all samples of the series can be indexed in a primitive tetragonal cell deriving from the perovskite- type ABO3 structure, with a c/2a distortion decreasing for high lithium content. The space group of primitive tetragonal of LLTO series was assumed to be Z=2. The unit cell parameters are a=ap=3.87 and c=2ap.

   The figure 1 shows a schematic view of LLTO and in the table 1 crystallographic data and symmetry point atom in the structural model in the space group P4/mmm are listed.

   Figure 1: Structure crystallographic of

   La2/3-xLi3x1/3-2xTiO3 in the space group P4/mmm.

   Table 1. Structural Model of La2/3-xLi3x1/3-2xTiO3 in the Space Group P4/mmm[9]

   Atoms	Sites	Symmetries	coordinates
   La1+Li+	1a	D4h	0, 0, 0
   La2+Li+	1b	D4h	0, 0, 1/2
   Ti	2h	C4h	1/2, 1/2, z0.25
   O1	1c	D4h	1/2, 1/2, 0
   O2	1d	D4h	1/2, 1/2, 1/2
   O3	4i	Cv 2h	0, 1/2, z0.25

   According to this model, the La3+ ions are centered in the A-cage of the perovskite structure formed by eight TiO6 octahedra. On the other hand, if the La3+ ion location is perfectly defined, the location of the Li+ ions has been a matter of debate.

   This structure is characterized by a distribution of the vacancies and cation La3+, Li+ on the two sites 1a(0, 0,

   0) and 1b(0, 0, 1/2). It has been shown that the La3+ distribution evolves with x content, that the 1a site is favorable to the La3+ ions, and that the introduction of the Li+ ions increases the cationic disorder on site A. The maximum occupancy of La3+ is observed for a value of x=0.08, corresponding to the composition with the maximum distortion of the tetragonal unit cell. This model does not take account of Li+ and vacancies, it is assumed that they partially occupy both 1a and 1b sites. The TiO6 octahedrons are distorted along the c axis. Raman Spectroscopy Analysis: The Raman spectra were recorded on the powder of LLTO using a DILOR Z24 spectrometer. An argon laser (Coherent Innova 90.3) was used for the excitation: radiation of the wavelength 514.5 nm was used. The spectral width (full-width at half-maximum) was less than 6 cm-1 and the error in the line positions was less than 1cm-1. The measurements were made at room temperature under a microscope (in the back-scattering geometry) with a X50 objective with a long working distance. This was performed on micrometric samples (typically less than 0.1mm in length) fixed on a goniometer head and suitably orientated for polarization analysis.

   Enumeration of vibration modes: Using the structural data and the crystallographic tables of the space group [18], we reported on the table 2 the symmetries vibration modes. The mechanical representation is written:

   Mechanical=2A1g 6A2uB1gB2u3Eg7Eu

   As Mechanical=Acoustic+Optical and according to the character table of the P4/mmm grou, we have

   Acoustic=A2uEu, so the optical modes are noted:

   Optical=2A1g5A2uB1gB2u3Eg5Eu

   The active vibration symmetries modes into Raman spectroscopy are A1g, B1g and Eg. We only observed:

   Raman=2A1gB1g3Eg

   Table 2. The symmetry modes in La2/3-xLi3x1/3-2xTiO3.

   Atoms	Symmetries vibration modes
   La1(1a)	A2u Eu
   La2(1b)	A2u Eu
   O1(1c)	A2u Eu
   O2(1d)	A2u Eu
   O3(4i)	A1g A2u B1g B2u 2Eg 2Eu
   Ti(2h)	A1g A2u Eg Eu

   We can see that only 6 modes are predict in Raman spectroscopy including 3Eg doubly degenerated modes, 2A1g modes and one B1g mode. We can notice, as shown in the table 2, that these modes do not give account of the vibrations of the La and Li atoms in this structure, but only Ti and O3 atoms. La, Li and the vacancies represented by La1 and La2 in the structure are related to inactive modes in Raman spectroscopy.

3. Results and discussion

The Raman spectra of LLTO are represented in figure

2. We have reported on the table 3 the Raman shift for different x values.

For x=0.07, it appears four significant peaks located at 139.14, 236.04, 317.6 and 524.9 cm-1. These peaks are allotted to the vibration modes of the LLTO. Other peaks appear to 567.52 and 781.12 cm-1 after applying of the Fit-Peak program.

For x=0.075, we find four well-defined peaks to 139.7, 237.25, 318.27 and 523.1 cm-1 like the x=0.07 sample. On the other hand, very badly defined peaks appeared to 382.31 and 453.83 cm-1, in addition to those to

564.66 and 780.13 cm-1, which are comparable with the preceding case. One also observes a series of peaks around the principal line to about 139 cm-1.

Table 3. Experiment Raman shift for various x content in La2/3-xLi3x1/3-2xTiO3.

The x=0.085 sample, we still find the same profile as previously with this time of the lines located at 140.07, 238.11, 318.34 and 524.92 cm-1. It is noticed that the peaks intensity to 570.24, 774.83 and 455.86 cm-1 have relatively increased compared to the preceding rate.

As in the preceding case the x=0.095 sample for bands are dominating in position 141.11, 238.98, 318.69 and

524.38 cm-1. The peak to 452.77 cm-1 seems to have

taken amplitude. The other peaks observed on the preceding spectra are also present here.

For the x=0.10 compound, the lines account this time for 141.64, 240.01, 320.08 and 525.67 cm-1. One

always notices the same peaks with the positions equivalent as previously to 450.37, 571.4, and 786.24 cm-1.

For the x=0.12 sample, One notices a significant displacement of the positions of the peaks compared to the preceding cases. The significant peaks occupy 143.48, 241.03, 313.97 and 523.33 cm-1 position. We always note the presence of a peak with 453.48 cm-1 whose intensity increased compared to the precedents. It is also noticed that the peak to 313.00 cm-1 moved closer to the peak to 241 cm-1. It is the same observation made for the peak located to 571 cm-1. This peak is practically confused with that to 523.33 cm-1. One still finds the profile observed towards 750 cm-1 that previously.

The x=0.13 sample, we found the typical profile observed until x=0.10. The peaks are allocated to 140.62, 238.02, 316.25 and 523.83 cm-1. We have

peaks to 455.96, 558.95 and 790.91cm-1.

Lattice dynamic studies [19, 20, 21] have showed that the lines located on average at 139, 239 and 523 cm-1 can be assigned to the three Eg modes predicted by the theory of the groups. It is noted that these peaks are well defined for all studied rates. With regard to the other awaited modes, their attributions are discussed because of the broad profiles of the peaks observed. These results are explained by the lattice dynamics[19,20]. It appears indeed that in the case of the Eg modes, the displacement of the ions is done in perpendicular plans with the axis C4 of the structure perovskite of the LLTO. These movements around octahedral TiO6 do not utilize the sites statistically distributed in ions It and Li or gaps. All occurs as if there were a compact stacking in the plan, which returns the frequencies of vibration of quite visible the Eg modes in the spectra with rather fine peaks. The displacements associated with the A1g and B1g modes are done in the direction of the axis C thus implying the A-site which are occupied at random by La and Li ions or vacancies. This observation can explain the fact that the lines related to these modes appear broadened owing to a dispersion of the interactions between the atoms involved in the movements. The A1g vibration modes can be allotted to the one of the lines located on average to 318, 455, and 568 cm-1 and the B1g mode

localization remains more difficult, in accordance with the studies carried out by Laguna et al [19,20].

peak to 139 cm-1

Raman Shift (cm-1)

144

143

142

141

140

139

138

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

321

peak to 317 cm-1

320

Raman Shift (cm-1)

319

318

317

316

315

314

241

240

239

238

237

236

peak to 453 cm-1

peak to 239 cm-1

Raman Shift (cm-1)

Raman Shift (cm-1)

458

456

454

452

450

0,07 0,08 0,09 0,10 0,11 0,12 0,13

x content in LLTO

313

0,07 0,08 0,09 0,10 0,11 0,12 0,13

x content in LLTO

peak to 553 cm-1

526,0

448

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

525,5

Raman Shift (cm-1)

525,0

524,5

524,0

523,5

523,0

522,5

522,0

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

Figure 3: Variation of peaks position versus x content.

peak to 239 cm-1

peak to 139 cm-1

Raman linewidth – FWHM – (cm-1)

Raman linewidth – FWHM – (cm-1)

34 70

32 65

30 60

28

55

26

50

24

22

20

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

peak to 317 cm-1

160

Raman linewidth – FWHM – (cm-1)

140

120

100

80

60

40

20

0,06 ,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

45

40

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

45

peak to 455 cm-1

Raman linewidth – FWHM – (cm-1)

40

35

30

25

20

15

10

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

80

peak to 523 cm-1

Raman linewidth – FWHM – (cm-1)

75

Figure 2: Raman spectra of the La2/3-xLi3x1/3-2xTiO3 70

65

from various x values. 60

55

The Raman spectra parameters are shown to the figures 3, 4 and 5: vibration frequency positions (fig.3), linewidth (fig.4) and intensity of characteristic peaks (fig.5).

This study shows spectra with nearby profiles, but presenting rather significant variations of frequencies up to 10 cm-1 and width of peaks; there seems to be a extremum for concentrations of x around 0.10. With regard to the amplitudes, one obtains less regular evolutions not being able to give an account of specificities of material. This result could be explained by the fact why during the Raman analysis the fields explored by the beam are very varied from one sample to another and by the possible presence of impurities.

50

45

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

Figure4: Variation of linewidth (FWHM) of Raman peaks versus x content

In spite of the observation of evolutions of the frequency and width in middle height, those are not regular it would be too early to connect the frequency of vibration of a sample of LLTO to its concentration. But it is not excluded that a more rigorous work makes it possible to bring to the required result.

60000

50000

40000

30000

20000

10000

0

peak to 139 cm-1

Raman Intensity (a.u)

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

peak to 316 cm-1

Raman Intensity (a.u)

x content in LLTO

7000

6000

5000

4000

3000

2000

1000

0

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

20000

peak to 239 cm-1

18000

16000

Raman Intensity (a.u)

14000

12000

10000

8000

6000

4000

2000

0

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

peak to 453 cm-1

Raman Intensity (a.u)

x content in LLTO

1400

1200

1000

800

600

400

200

0

0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

x content in LLTO

5. References

1. A. G. Belous, V. I. Butko, G. N. Novitskaya, S. V. Polyanetskaya, Y. M. Poplavko and B. S. Khomenko, Electrical conductivity of the perovskites La2/3-xM3xTiO3, Ukrainskii Fizicheskii Zhurnal 31, (1986), 576-580.

2. Y. Inaguma, C. Liquan, M. Itoh, T. Nakamura, T. Uchida,

   H. Ikuta and M. Wakihara, High Ionic Conductivity in Lithium Lanthanum Titanate, Solid State Communications, Vol. 86, No. 10, (1993), 689-693.

3. C. Bohnke, J.-L. Fourquet, Impedance spectroscopy on pH-sensors with lithium lanthanum titanate sensitive material, Electrochimica Acta 48, (2003), 1869-1878.

4. O. Bohnke, The fast lithium-ion conducting oxides

   16000

   peak to 523 cm-1

   Li La

   TiO

   from fundamentals to application, Solid State

   3x 2/3-x 3

   14000

   Raman Intensity (a.u)

   12000

   10000

   8000

   6000

   4000

   2000

   0

   0,06 0,07 0,08 0,09 0,10 0,11 0,12 0,13 0,14

   x content in LLTO

   Figure 5: Variation of Raman intensity of principal peaks versus x content

   4. Conclusion

   In this work, we performed the Raman spectroscopy study of ionic conduction oxides by the ions lithium. Raman spectra of LLTO did not make it possible to

   clearly identify the six modes awaited by the theory of the groups. However, thanks to a study of lattice

   Ionics 179, (2008), 9-15

5. S. Stramare, V. Thangadurai and W. Weppner, Lithium Lanthanum Titanates: A Review, Chem. Mater. 15 (2003), 3974-3990.

6. Yoshiyuki Inaguma, Liquan Chen Mitsuru Itoh and Tetsuro Nakamura, Candidate compounds with perovskite structure for high lithium ionic conductivity, Solid State Ionics 70-71, (1994), 196-202.

7. Mitsuru Itoh, Yoshiyuki Inaguma, Woo-Hwan Jung, Liquan Chen, and Tetsuro Nakamura, High lithium ion conductivity in the perovskite-type compounds Ln1/2Li1/2TiO3 (Ln=La, Pr, Nd, Sm), Solid State Ionics 70-71, (1994), 203- 207.

8. H. Omanda, T. Brousse and D. M. Schleich, New Trends in Intercalation Compounds for Energy Storage, (2002), 483- 487.

9. A. Varez, F. Garcia-Alvarado, E. Moran and M. A. Alario-Franco, Microstructural Study of La0.5Li0.5TiO3, Journal of Solid State Chemistry 118, (1995), 78-83.

   dynamics[19,20], we could allot the 3Eg

   modes to the

   -1

10. J. L. Fourquet, H. Duroy, and M. P. Crosnier-Lopez,

    Structural and Microstructural Studies of the Series

    frequencies located towards 138, 236 and 524 cm . The other modes are not very obvious in the spectra. This result is because in these ionic conducting materials where the no-stoichiometric armature the presence of gaps, the neighborhood of atoms is very variable because of the random distribution of the atoms in the various sites. The analysis of the curves of frequencies, linewidths and amplitudes variations according to rate x of lithium in LLTO, did not give a great satisfaction. However, for the frequencies variations, we observe the same profile i.e. a growth for the weak concentrations to about of x=0.10, then a decrease for higher rates. This behavior can be compared with the variation of conductivity [22]. Only thorough studies will be able to say us some more.

    La2/3-xLi3x1/3-2xTiO3, Journal of Solid State Chemistry 127, (1996), 283294.

11. J. Ibarra, A. Varez, C. Leon, J. Santamaria, L.M. Torres- Martinez, J. Sanz, Influence of composition on the structure and conductivity of the fast ionic conductors La2/3xLi3xTiO3 (0.03×0.167), Solid State Ion 134 (2000) 219-228.

12. Odile Bohnke, Huguette Duroy, Jean-Louis Fourquet, Silvia Ronchetti, Daniele Mazza, In search of the cubic phase of the Li+ ion-conducting perovskite La2/3xLi3xTiO3: structure and properties of quenched and in situ heated samples, Solid State Ionics 149, (2002), 217226.

13. Yoshiyuki Inaguma, Tetsuhiro Katsumata, Mitsuru Itoh, Yukio Morii, Takao Tsurui, Structural investigations of migration pathways in lithium ion-conducting La2/3xLi3xTiO3 perovskites, Solid State Ionics 177, (2006), 30373044

14. Odile Bohnke, Joel Emery, Jean-Louis Fourquet, Anomalies in Li+ ion dynamics observed by impedance spectroscopy and 7Li NMR in the perovskite fast ion conductor (Li3xLa2/3-x1/3-2x)TiO3, Solid State Ionics 158 (2003) 119-132

15. T. Brousse, P. Fragnaud, R. Marchand, D. M. Schleich,

    O. Bohnke and K. West, All oxide solid-state lithium-ion cells, Journal of Power Sources 68, (1997), 412-415.

16. J. K. Ahn and S. G. Yoon, Characteristics of Amorphous Lithium Lanthanum Titanate Electrolyte Thin Films Grown by PLD for Use in Rechargeable Lithium Microbatteries, Electrochemical and Solid State Letters 8, (2005), A75-A78.

17. M. Morales, P. Laffez, D. Chateigner and I. Vickridge, Characterization of lanthanum lithium titanate thin films deposited by radio frequency sputtering on [100]-oriented MgO substrates, Thin Solid Films 418 (2002), 119-128.

18. D. L. Rousseau, R. P. Bauman S. P. S. Porto, Normal mode determination in crystal, Journal of Spectroscopy, vol. 10, (1981), 253-290.

19. M. L. Sanjuán and M. A. Laguna, Raman study of antiferroelectric instability in La(2-x)/3LixTiO3 (0.1<x<0.5) double perovskites Phys. Rev. B 64, 174305 2001

20. M. A. Laguna, M. L. Sanjuá, A. Várez, and J. Sanz, Lithium dynamics and disorder effects in the Raman spectrum of La(2-x)/3LixTiO3, PHYSICAL REVIEW B 66, (2002) 054301-1 – 054301-7

21. B. Sitamtze Youmbi, Serge Zékeng, Samuel Domngang, Florent Calvayrac, Alain Bulou, An ab initio molecular dynamics study of ionic conductivity in hexagonal lithium lanthanum titanate oxide La0.5Li0.5TiO3, Ionics 18 (4), (2012), 371-377

22. O. Bohnke, C. Bohnke, J.L. Fourquet, Mechanism of ionic conduction and electrochemical intercalation of lithium into the perovskite lanthanum lithium titanate, Solid State Ionics 91 (1996) 21-31.