Nanostructured SnO₂-TiO₂ films as related to lithium intercalation

S. Yu. Vassiliev¹, A. I. Yusipovich¹, Yu. E. Rogynskaya², F.Kh.Chibirova², A. M. Skundin³, T. L. Kulova³

1. Department of Electrochemistry, Faculty of Chemistry, Moscow State University, Moscow, Russia
2. Research centre "Karpov Institute of Physical Chemistry" ,Moscow, Russia
3. Frumkin Institute of Electrochemistry Russian Academy of Sciences, Moscow, Russia

Abstract

The difference in degradation behavior of titania-doped tin dioxide films is explained by a pronounced effect of the doping level on the film dispersity and fine distribution of titania. Two – three times decrease of nanoparticles size in doped films compared to nanoparticles in SnO₂ film (20 – 30 nm) is revealed by using scanning tunneling microscopy (STM). STM data (measured in ex situ configuration) combined with XRD and Mössbauer spectroscopy analysis confirm that nanoparticles are composed of nanostructured heavily disordered SnO₂ and TiO₂  rutile solid solution or of amorphous phase containing both SnO₂ and TiO₂, the content of the crystalline and amorphous phases being approximately equal.

Keywords: Tin dioxide, lithium intercalation, scanning tunneling microscopy and spectroscopy, nanostructure

Introduction

Nanostructured SnO₂-based compositions well-known as materials for gas sensors [1, 2] were recently found to be also prospective candidates for lithium batteries [3-5]. The initial cathodic polarization of doped tin oxides in solutions of lithium salts results in the irreversible reduction of oxide with formation of metallic tin. Thus formed metallic tin dispersed in oxide matrix demonstrates the reversible absorption-desorption of lithium with formation of intermetallic compound Li_4.4Sn. Correspondingly, irreversible (observed in the course of initial charging) and reversible contributions to capacity can be separated. The size of initial oxide particles is responsible for the values of capacity components and stability of lithium recharging (see, for example, in Ref. [4]). A serious problem results from the gradual increase of the size of metal particles, which, in turn, induces the decrease of reversible capacity in respect to lithium.
An alternative concept of lithium intercalation by tin oxides stresses the role of Sn-O-Li nanoscalephases, not intermetallic compounds. The stability of nanoscale phases in relation to reduction by lithium determines more slow growth of tin crystallities [6]. When some foreign oxides (like B₂O₃, Al₂O₃ or P₂O₅) are inserted to form dispersed amorphous tin-based composite oxide, the best long-term cycling is achieved resulting from high stability in respect to reduction [7].
Nanoheterogeneous titania-doped tin oxide films prepared by thermohydrolytic decomposition of SnCl₂ and TiCl₄ solutions can be considered as prospective charge-accumulating materials because of high conductivity, nanometer-scale size of constituents, and mechanical stability in the course of prolonged cycling in aqueous solutions [1, 2]. For lithium accumulation by the films of this type [8, 9], the reversible capacity up to 650 mA h g^-1 is found, with satisfactory cycling stability.
Structural studies of multicomponent oxide film electrodes (electrocatalysts, etc) prepared by the above mentioned technique show that the reaction of hydrated species in solution followed by heat treatment results in the formation of unusual nonequilibrium phases – nanoheterogeneous solid solutions [10,11]. They have been found to be a mixture of various oxide nanoclusters (RuO₂, TiO₂, IrO₂, SnO₂) of less than 1-2 nm size, which are statistically distributed in amorphous matrix formed by their hydrated precursors.
According to STM, XRD and XPS studies, SnO₂ films containing 10 mol % TiO₂ (ST10), annealed at 450°C consist of globules of 15 – 20 nm size, which in turn have a fine nanostructure [1, 2, 11]. The cores of globules contain crystallites of (Sn, Ti)O₂ rutile solid solution (6–8 nm) pierced with amorphous SnO₂ strips. The similar amorphous SnO₂ layers containing up to 15% SnO cover the globule surface. Such nanostructure in ST10 films is due to the formation of the crystalline SnCl₂ intercalation phase as their precursor, which contains inclusions of polynuclear titanium-tin oxohydroxo complexes.
Some two-three times larger titania-free SnO₂ globules were observed [11]. Hydrous tin oxide, which was formed in the course of condensation of polynuclear species {SnO_x(OH)_6-x}_n has been found [11] to be the precursor of that phase. The latter species result from hydrolysis of [Sn (IV)Cl₄]L₂ (L = OH, H₂O) complexes presenting in initial SnCl₂ solution due to the partial oxidation Sn(II) - > Sn(IV).
The role of TiO₂ in creating high charge storing properties of SnO₂ remains unclear.
The aim of this work was to obtain single phase SnO₂-based materials with a variable TiO₂ content and to study their structure, morphology, and electrochemical properties. At this stage the study is limited to freshly prepared films, providing the basis for further understanding of micro- and nanostructural changes induced by lithiation/delithiation. In this study, we report as observed by using of STM, X-ray diffraction, and Mössbauer spectroscopy.

Materials and methods

The condition of the synthesis of a new series of SnO₂ – TiO₂ films (further designated as STx, where x = 0, 10, and 20 mol % TiO₂ (Fabrication mode differed from previously used in [1,2,8], so the sample denoted ST10 and ST20 are not identical to the samples described in these Refs.)). While the concentration of SnCl₂ solution remained the same as in [1,2,8] (0.067 M), the concentration of TiCl₄ was increased by the order of magnitude and amounted 1 M. It is known that in aqueous hydrochloric acid solution of TiCl₄ (c_Ti = 0.01 – 0.05 mol/l and c_HCl = 0.1-2.0 mol/l) mono- and polynuclear forms of Ti (IV) hydroxyoxo complexes exist for the latter [(TiO)₈(OH)₁₂]⁴⁺ species predominate [12]. The rising of TiCl₄ concentration (0.35 – 1.2 mol/l) (c_HCl = 1 mol/l) resulted in the increase of the number of polynuclear complexes and the degree of their polymerization. These polymeric species appear in essence the nuclea of the hydrated titanium oxide [13]. The solutions were applied layer by layer onto titanium supports pretreated by mechanical polishing and etching in sulfuric acid. The temperature of annealing was 450^oC for doped films, 350^oC and 450^oC for pure tin oxide.
X-ray dispersion analysis confirmed that the content of SnO₂ and TiO₂ in films matched to the nominal composition of Sn and Ti cations in original solution. Besides about 1% of chlorine was found. XPES analysis showed that about 25 % of oxygen atoms present in hydroxogroups, and 5% in aqua groups.
For electrochemical measurements, 1 M LiN(CF₃SO₂)₂ in dioxolane was used as electrolyte (water content below 70 ppm). Compartments of the three-electrode cell were separated with porous membranes (polypropylene PORP). Potentials were measured and are referred below vs. lithium reference electrode. Before starting the experiments, oxide films were additionally dried under vacuum at 120^oC for 8 hours. EL-2 potentiostat designed in the Frumkin Institute, RAS, was used.
For XRD, DRON-3M diffractometer run by PC was used. For the decomposition of intensity profiles to multicomponent spectra PROFITVZ program [14] was used. We could not perform quantitative analysis for the electrodes studied since the lines of titanium support superimposed practically all intensive rutile lines (besides 110) and prevented correct decomposition of mixed maximums. The instrumental broadening b was too small (0.21 degree 2q) to make the corrections in FWHM (Full Width at Half Maximum) value, therefore the latter was considered as b (diffraction broadening).
The samples were studied with a conventional Mössbauer spectrometer. The velocity drive of this spectrometer operates under the constant acceleration mode over 1024 channels. The velocity scale of the Mössbauer spectra was calibrated with a reference spectrum of metallic iron. All measurements were carried out at room temperature. The Conversion Electron Mössbauer Spectroscopy (CEMS) measurements were carried out using a gas detector with He + 5%CH₄ as the flow-gas. In the (CEMS) method internal conversion electrons and Auger electrons associated with the de-excitation of Mössbauer nuclei are detected in the sample under study. Due to the small range of electrons in solids this method is especially convenient for studying subsurface regions of the order of a few hundreds atomic layers. 5 mCi CaSn^119mO₃ was used as g-ray source in this experiment.
For tunneling microscopy experiments, homemade LitScan-2 device with extended spectroscopic facilities was used. Pt-Ir tips (10% Ir) of 0.5 mm diameter were mechanically sharpened. In topographic experiments, bias was 1.3 V (positively polarized tip), and STM current was 0.3 nA. The technique of measuring the local spectra is discussed in detail in [15]. Voltage-current spectra were measured under a pulse mode, in order to avoid the shift of tip location in the course of measurements. Voltage-distance spectra were registered under open feedback loop, with the scan rates not exceeding 0.5 V s^-1. Numerous (100-150) repeated measurements of current-voltage and voltage-distance spectra were carried out for all samples. These data were collected for various points inside the area within the limits of piezomanipulator movement, and also for various surface regions and experimental sets. Good reproducibility of the key characteristic features of the tunneling spectra was found for each certain type of oxide. Most typical spectra without any averaging or pretreatment are presented in this article.

Results and discussion

X-ray analysis revealed that the ST10 and ST20 films contained a single crystalline phase. The decomposition of the profile of each reflection by program PROFITVZ evidenced that only one rutile phase gave the contribution to the intensity profile. Maximums in the mixed films patterns are very weak and diffuse (Fig.1).

Fig.1. The results of the decomposition of (110) reflection to possible multicomponents. a, b and c present the decomposition of (110) reflection for SnO2, ST10 and ST20 electrodes, respectively, and show doublet (a1 and a2 ) for only one rutile phase.

That effect may point to the dispersion of crystallites or/and to the microdistortions (disordered crystal structure), as well as to the presence of amorphous phase. The crystallinity of SnO₂ electrode is much higher. The increase of reflection FWHM with the 2q (Bragg angle) proportional to tg q [16] testifies to the existing of microdistortion in ST10 and ST20 electrodes. This may be connected with disordered structure of mixed phase containing SnO₂ and TiO₂. This fact does not allow to do the correct estimation of crystal size from XRD data. The FWHM of SnO₂ film (amounted 0.854) practically does not change with 2q that points to the effect of crystallites dispersion as the main contribution in the width of lines. In the latter case the average size of crystallites about 11 nm was calculated by means of Sherrer formula [16].
The integral intensities of (110) reflections were 148, 80 and 80 relative units for the films with x = 0, 10 and 20, respectively. However, the heating of plain SnO₂ film up to 450 ^oC resulted in the increase of line intensity of main strong lines (~ 20% for (110), and 40% for (101) and (301), that is the degree of crystallinity increased. All these data permit us to estimate that the content of amorphous phase in mixed films is higher than 50%.
The rutile cell parameters a and c for all three samples are given below in Table 1. The remarkable decrease of a and c in titania-containing films in comparison with undoped one means that some type of solid solution has been formed. The decrease of rutile lattice periods from the values typical for tin oxide my be explained by the lower ion radius of Ti⁴⁺  (0.064 nm) in comparison to the radius of Sn⁴⁺ (0.071 nm)
Refinement of film morphology and determination of Sn atoms oxidation state and their local surrounding have been solved by using of ¹¹⁹Sn MS. The analysis of Mössbauer spectra (the main characteristics of the MS such as chemical shift (d), quadrupole splitting (E), and relative intensity, A (area of spectra) are given in Table 1) leads to the following conclusions.
1. According to the magnitude of relative absorption (~60%), the spectrum of the SnO₂ film (Fig. 2) corresponds to a solid sample with a high degree of crystallinity, while the spectra of the films with x = 10 and 20 can be attributed to solids with a high degree of amorphous SnO₂ due to remarkably lower relative absorption.

Fig.2. Mössbauer spectra of SnO2–TiO2 films. a – SnO2, b – ST10

2. A doublet typical of the Sn atoms in SnO₂ represents the main contribution in all three films (93% – 98%); some increase in the chemical shift, d, in the spectra of mixed films in comparison to SnO₂ film may be associated with a small lengthening of the Sn – O bonds in amorphous SnO₂ due to the change in the s – p bonding.
3. The distinctly remarkable second doublet observed for all spectra is attributed to the partly hydrated SnO localized at the surface. Detailed studies by XPS and inverse photoemission spectra [18, 19] demonstrated that these "SnO" states are present on the disordered surface of SnO₂ or in nanosized SnO₂ crystallites. Disorder of SnO₂ results in the loss of the symmetry centre of the Sn⁴⁺ cation octahedral surrounding, in splitting of Sn5s and Sn5p atomic orbitals and emerging of hybrid 5s – 5p orbitals, which are displayed as Sn²⁺ states. So "SnO" surface states are indicative of tin atoms in SnO₂ disordered grain boundaries. The non monotonous dependence of Sn²⁺ content on titania content will be discussed below.
Analysing together the XRD and MS results one can notice two contradictory phenomena: in spite of the remarkable change of rutile cell periods in the films studied, only a small increase in chemical shift (d) with TiO₂ content is observed.
This contradiction is removed if to take in consideration that the amorphous SnO₂ dominates in mixed films while the crystalline rutile phase presents in less quantity, as evidenced by XRD. In such case the amorphous SnO₂ gives higher contribution into Mössbauer parameters of mixed films than tin oxide involved into rutile solid solution. The higher degree of amorphous state in mixed films compared to SnO₂ film results in the small lengthening of the Sn–O bonds. However that effect is not related to crystalline SnO₂ in solid solutions, that is the influence of titanium oxide on electronic parameters of crystalline SnO₂ was not revealed.
The electrochemical behavior of the films under study was similar to previously observed for titania-doped tin oxide films described in [8, 9]. Galvanostatic charge-discharge curves (first cycle) of the samples with various titania content demonstrate no pronounced difference of capacities, whereas the irreversible capacity is noticeably higher for ST20 sample (Fig. 3). For pure tin oxide (curve 1, Fig. 3) a plateau at ca. 1 V is observed, which corresponds to reduction with formation of tin metal; parallel formation of lower valency oxides also can not be excluded at this potential region [8]. For titania-containing films, similar reduction process takes place in a wider potential region and does not demonstrate any special electrochemical features. This behavior can be explained by both kinetic reasons and more pronounced heterogeneity of doped samples. The role of dispersion can not be clarified at this stage.

Fig.3. Charging curves of tin dioxide films (first cycle) measured in 1M LiN(CF3SO2)2 in dioxolane. 1 – SnO2, 2 – ST10, 3 – ST20. Current density 80 mA g-1.

Despite of close reversible capacities observed for the first cycle, the decrease of capacity in the course of subsequent cycling depends on titania content (Fig. 4). We are concentrated below on the understanding of micro(nano)structural doping effects which induce so essential difference of degradation behavior.

Fig.4. Cyclic stability of the reversible capacitance. 1 – SnO2, 2 – ST10, 3 – ST20. Current density 80 mA g-1.

Figs. 5b-d present typical STM images of the films in mikrometer scale. Image of titanium support in Fig. 5a is given for comparison. It is easy to observe deep scratches formed in the course of mechanical polishing. Fast oxidation of titanium surface induced by the tip is accompanied by formation of loose poorly conducting oxide. Films under study demonstrate another type of topography, consisting of smooth globules of 200–700 nm size. Globules consist of smaller nanoparticles, which size decreases sharply with the increase of titania content (Fig. 6). Crystal size estimated from STM data for SnO₂ film with taking into account image distortions induced by tip non-ideality [17]) are found to be in agreement with the sizes estimated from XRD data (Table 1). These data indicate the higher dispersion of titania-doped films. The clearness of STM images decreases in the sequence ST0-ST10-ST20, as it can be seen in Fig. 6. Loss of clearness can result from the increase of the content of amorphous phase.

Fig.5. Micrometer scale STM images of titanium support (a) and tin oxide films. b – SnO2, c – ST10, d – ST20.

Fig.6. Nanometer scale STM images of tin oxide films. a – SnO2, b – ST10, c – ST20.

Tunneling spectra (Fig. 7) demonstrate the decrease of films conductivity with titania content, and also more pronounced resistive heterogeneity for higher doping level. Low conductivity of ST20 films makes it difficult to fix an STM tip and decreases the quality of STM images of these films. One-two order increase of the resistance of STM gap at positive currents (i.e. for the highest contributions of film resistance) is found.

Fig.7. Local voltage-current spectra obtained for tin dioxide films. 1 – SnO2, 2 – ST10, 3,4 – ST20. (curves typical for various points of heterogeneous sample).

Separation of air gap and film resistance contributions to the total STM gap resistance presents a special problem. With increase of film resistance, the effect of film on the slope of voltage-current spectra rises sharply. The reason is as follows: at a certain fixed base values of bias and current under equilibrium conditions the resistance of air gap for more resistive films should be much lower, in order to compensate the increased contribution of the film. Correspondingly, we can conclude several-order increase of the specific resistance of the films with titania content. Direct conductivity measurements are complicated by low film thickness and the existence of more conducting support.
Qualitatively, the observed resistance vs. composition dependence agrees well with previous data of direct conductivity measurements of similar films on insulating supports [2].
Increased heterogeneity of local conductivity observed for the films with higher titania content can be referred to more pronounced heterogeneity of chemical/phase composition. For ST20, heterogeneity results in one order difference of the slopes of voltage-current spectra measured in various points (curves 3 and 4 in Fig. 7). For less doped films, difference in local conductivity of various areas is clearly seen in more sensitive voltage-distance spectra (Fig. 8). For certain points these spectra are smooth, with the changes of distance below 5 nm (curve 1 in Fig. 8). Curves of this type correspond to more conductive areas. The spectra of similar shape with close distance values were obtained also for pure tin dioxide. For some other areas (which topographic images usually look more cloudy) voltage-distance spectra are asymmetric and include stepwise features with distance changes of 50 – 100 nm (curve 3 in Fig. 8). We associate this sharp distance increase with the fact that at low voltages the film conductivity is too low for passing the established current. Under constant current mode this situation results in closer approaching of tip and appearance of direct tip-sample contact, with partial penetration of tip into the film. In the framework of this hypothesis the appearance of stepwise features manifests much higher local resistance of the areas under study.

Fig.8. Local voltage-distance spectra of ST10 sample: representative types of curves observed at various points of the sample.

There are a lot of surface areas for which an intermediate type of voltage-distance spectral behavior is observed (curve 2 In Fig. 8), and correspondingly intermediate values of conductivity can be concluded for these areas.
For ST20 samples, the stepwise behavior of voltage-distance spectra is observed for any point of the surface. Typical distances appear to be up to 1 mm, which value exceeds the vertical resolution of the STM device. This observation obtained with the most resistive samples confirms the interpretation of voltage-distance anomalous behavior discussed above.
The pronounced increase of the total film resistance can result from the changes in conductivity of the crystalline solid solution SnO₂-TiO₂, and also from the increase of contribution of the poorly conducting amorphous component (both effects are assumed to be size-dependent). The specific features of fixing the tip in the gap can be the reason of the change of image nature: the increase of total resistance can lead from the imaging of conductive crystalline component to the imaging of amorphous component and (or) simultaneous imaging of both components. The appearance of steps in the voltage-distance spectra agrees with the assumption about the abovementioned change. The qualitative difference of ST20 STM images as compared to less doped films also supports this assumption. However we can not determine the quantity of amorphous component from STM data.

Based on STM, XRD and MS results we could propose the following model of nanostructured doped films.
The highly dispersed (on STM base) rutile phase observed in mixed films studied in the present work may be classified as a solid solution with short range order or as a nanocomposite containing both SnO₂ and TiO₂ nanoclusters (<1 nm in size). Poorly ordered tin and titanium oxide nanoclusters are randomly distributed in partially hydrated amorphous tin oxide matrix. The latter are pierced by ultra fine pores which contain H₂O molecules and Cl^- ions. The crystalline SnCl₂ intercalation phase has been found to be the precursor of that nanostuctured forms [2, 11]. The quantities of that rutile phase in both mixed films are approximately equal, as limited by the content of Sn(II) anionic chloro complexes in origin solution of tin and titanium chlorides. The concentration of these complexes is constant because of constant c_HCl.
As concerned to the amorphous phase, tunneling spectroscopy evidences the remarkable increase of the resistance with TiO₂ content. For ST20 electrode such increase occurs for any point of the surface (both for globules with crystalline core  and completely amorphous globules). The high resistance is typical for the surface of the nanoparticles composing of tin–titanium oxide solid solution [1, 2]. The high resistance of completely amorphous globules means that amorphous SnO₂ phase of the electrodes is also essentially enriched with TiO₂.
The remarkable increase of the amorphous mixed phase may result from the higher number of Ti (IV) hydroxooxo complexes and higher degree of their polymerisation in initial solution with concentration of TiCl₄ of 1 mol/l. These polymeric species, which in essence appear the nuclea of hydrated titanium oxide, favor the hydrolysis of SnCl₂. The latter process induces the formation of mixed hydrated tin and titanium oxides, which remain in amorphous state up to 450 ^oC.
In conclusion, the existance of two types of precursors of the main components in mixed films was the reason of the different content of Sn²⁺ as observed my MS. The increase of SnO in the ST10 film in comparison to SnO₂ film may be explained by both the dispersion of nanoparticles observed by STM and by the role of precursor of the rutile solid solution (crystalline SnCl₂ intercalation phase). In the course of decomposition of the latter the amorphous SnO₂ layers containing up to 15% SnO cover the globule surface [2]. The increase of Ti component results in the higher content of tin(IV) and titanium hydrated forms [11]. Therefore we can atribute the decrease of SnO content in the film x=20 with the latter effect.

Conclusion

Doping of SnO₂ with TiO₂ results in pronounced dispersion of the solid. This effect is induced by 2-2.5 times decrease of nanoparticles forming the films, and also by oxide dispersion in the shorter scale. Nanoparticles consist of disordered nanostructured solid solutions (presented by SnO₂ and TiO₂ nanoclusters of nanometer-scale size) and amorphous phase formed by both SnO₂ and TiO₂. The portion of amorphous phase in the films under study is essentially higher than in electrode materials studied in our previous works [1, 2], and its reason is the change of hydrolysis conditions for initial solutions of salt reagents. Just this extremely high dispersity of titana, which is a component stable to reduction with lithium, can be considered as a reason of better degradation stability with simultaneous keeping high capacity values (typical for nanostructured tin oxides). Direct observation of the increase of tin dioxide dispersion with doping titania content is obtained.
Nanoheterogeneous character of rutile phase observed for the films under study is confirmed by XRD and Mössbauer spectroscopy investigations.
The decrease of particle size is assumed to be a reason of more slow degradation in the course of cycling in lithium-containing solutions (Fig. 3). Actually, the enlargement of metallic tin particles should be very slow when these particles are separated by low conductive (and most probably amorphous) oxide with high titania content. Heterogeneity of doped films confirmed by local tunneling spectroscopy can be considered as a possible reason of the absence of plateau. Both heterogeneity of ohmic drop and equilibrium potentials for certain redox transformations of chemically different areas can be assumed, and these two effects can hardly be separated at this stage.
Interpretation of electrochemical data for highly resistive films is complicated by the uncertainty of potential values, which are distorted by the ohmic drop and possible heterogeneity of potential distribution along the surface. Under these circumstances one can not exclude the selective solvent discharge at already formed metallic fragments, which gives additional contribution to the total charge.

Acknowledgements

The authors are grateful to A.V.Denisov for technical support of STM and spectroscopic studies. The study is supported by RFBR, projects 03-03-32422-a, 02-03-32226-a.

References

1. Tsirlina GA, Roginskaya YuE, Postovalova GG, Vassiliev SYu (1999) Russ J Electrochemistry 35:1385.
2. Postovalova GG, Roginskaya YuE, Zavyalov SA, Galyamov BSh, Klimenko NL (2000) Neorganicheskie materialy (in russian) 36:452.
3. Courtney IA, Dahl JR (1997) J Electrochem Soc 144:2045.
4. Courtney IA, Dahl JR (1997) J Electrochem Soc 144:2943.
5. Zhu J, Lu Zh, Aruna ST, Aurbach D, Gedanken A (2000) Chem Matter 12:2557
6. Goward GR, Leroux F, Power WP, Ouvrard G, Dmovski W, Egami T, Nazar LF (1999) Electrochemical and Solid-State Letters..2:367
7. Idota Y, Matsufuji A, Maekawa Y, Miyasaki T (1997)  Science 276:1395
8. Kulova TL, Skundin AM, Roginskaya YuE, Chibirova FKh (2004) Russ J Electrochemistry 40:484.
9. Kulova TL, Roginskaya YuE, Skundin AM (2004) Russ J Electrochemistry 40:(in press).
10. Roginskaya YuE, Morozova OV (1995) Electrochimica Acta  40:817.
11. Postovalova G, Morozova OV, Galyamov BSh, Lubnin EN, Prutchenko SG, Kozlova NV, Roginskaya YuE (1998) Russian Journal of Inorganic Chem. 43:36.
12. Einaga H (1979) J.Chem.Soc., Dalton Trans. 12:1917.
13. Bragina MI, Bobyrenko YuYa (1972) Russian Jornal of Inorganic Chem. 17:417.
14. Zhurov VV, Ivanov SA (1993) Crystallography Report 42:202.
15. Vassiliev SYu, Denisov AV (2000) Zhurnal tehnicheskoy fiziki (in russian) 70:100.
16. Iveronova VI, Revkevich GP (1972) The theory of X-ray scattering. Ed. by MSU, p.124-125
17. Vassiliev SYu, Pron’kin SN, Tsirlina GA, Petrii OA (2001) Russ J Electrochemistry 37:523.
18. Cox PA, Egdell RG, Harding C, Patterson WR, Tavener PJ (1982) Surf. Sci 123:179
19. Hollamby PC, Aldridge PS, Moretti C, Egdell RG (1993) Surf Sci 280:393
20. Zhu J, Zhonghua Lu, Aruna ST, Aurbach D, Gedanken A (2000) Chem Mater 12:2557

Table 1. XRD and Mössbauer spectra data for SnO₂ – TiO₂ films.

* D_av  — the average size of crystallites determined with the use of FWHM for (110) reflection by Sherrer method. (,  nm [16])

Back