Pressure-Induced Structural Phase Transition of Co-Doped SnO2 Nanocrystals

1. Introduction

Nano-scale materials are having a great impact in human life. They are changing dental medicine, healthcare, and human life more profoundly than several developments of the past. Nanomaterials research has also become a striking area of research due to their intriguing chemical and physical properties. Nanomaterials are preferred over their bulk counterpart due to enhancement in catalytic, optical, magnetic, and electrical properties [1,2,3]. The nature of pressure-induced structural phase transitions, elastic properties, and transition pressure in nanomaterials are quite different compared to the bulk materials [4,5,6,7,8,9]. Tin dioxide (SnO₂) is a technologically very important material and a widely used wide bandgap semiconductor. It finds applications in the field of solar cells [10], ultraviolet photodetectors [11], short wavelength light-emitting diodes [12], spintronic applications [13], gas sensors [14], and lithium-ion batteries [15].

At ambient conditions SnO₂, which is also known as stannic oxide, crystallizes in a crystal structure isomorphic to tetragonal rutile (space group P4₂/mnm, Z = 2). In this structure two Sn ions are six-fold coordinated by oxygen atoms, which are shared with adjacent SnO₆ octahedra as shown in Figure 1(a). In the past, it has been reported that SnO₂ undergoes the following sequence of pressure-induced structural phase transitions: rutile-type phase → CaCl₂-type phase → PbO₂-type phase → fluorite-type phase [16]. Earlier high-pressure (HP) X-ray diffraction (XRD) measurements [16,17], theoretical ab-initio calculations [18,19,20,21,22], and HP Raman scattering [23] on bulk as well as nanocrystalline samples indicate the occurrence of a second order-structural phase transition between 7-14 GPa, from the tetragonal rutile-type to the orthorhombic CaCl₂-type structure (space group Pnnm), which is represent in Figure 1(b)). In the CaCl₂-type structure, the Sn atoms are also octahedrally coordinated by oxygen atoms, forming chains of edge-sharing octahedra along the c-axis of the structure, while perpendicular to the c-axis two octahedra are linked by sharing one corner. However, there are few exceptions for the rutile to CaCl₂-type transition. For example, high-pressure XRD measurement carried out on nanocrystalline 5-nm sized samples [24] and 8-nm sized samples [25,26] indicated a direct tetragonal rutile to cubic fluorite first-order structural phase transition beyond 18 GPa. The transition pressure was also found to be increased with decreasing particle size [26]. Even doping has shown alteration in the high-pressure behavior, elastic properties, and optical properties. For example, in Fe-doped nanoparticles (size 18 nm and 10 at % doping) of SnO₂ shows an increase in the transition pressure and bulk modulus [27] whereas V-doped nanoparticles (size 10 - 30 nm and 5 - 12.5 at % doping) of SnO₂ shows a decrease in the bulk modulus with increasing concentration of the dopant [28]. Hence, it is very interesting to know how the particle size, doping concentration, and type of doping ion are altering the high-pressure behavior and elastic properties of SnO₂, and further systematic investigations are required in this direction. In the current contribution, we are reporting HP synchrotron XRD studies up to 20.9 GPa on Co-doped nanocrystalline SnO₂ (crystallite size 15 nm and 10 at % doping) to contribute to the understanding of the effect of doping on the transition pressure and elastic properties. Such kind of study is not reported earlier.

2. Experiments

Cobalt-doped SnO₂ nanoparticles were prepared by the wet chemical co-precipitation method following the procedure described by Ferrari et al. [29]. To implement the cobalt doping we used cobalt chloride from Sigma-Aldrich. The nanoparticles were characterized at ambient-conditions by means of powder XRD confirming the formation of single-phase rutile-type nanoparticles with unit-cell parameters a = 4.732(1) Å and c = 3.185(1) Å, which are consistent with single crystal X-ray diffraction studies on bulk SnO₂ [30,31]. The average particle size of 15 nm was determined from XRD using the Scherrer equation [32]. The cobalt concentration (10 at %) was determined from Energy-dispersive X-ray spectroscopy. HP powder XRD measurements were carried out at the MSPD-BL04 beamline of ALBA synchrotron using a monochromatic beam of wavelength 0.4642 Å. The beam was focused down to a 20 μm full-width-at-half-maximum spot. A Rayonix charge-coupled device detector was used to collect XRD patterns. The detector was calibrated using LaB₆ as standard. Two-dimensional diffraction rings obtained from the detector were transformed into one-dimensional diffractograms using Dioptas. Measurements were performed with a membrane diamond-anvil cell (DAC) equipped with 500-μm diameter diamond anvils. We used a stainless-steel gasket pre-indented to a thickness of 40 μm with a centered hole of 200 μm in diameter. As pressure-transmitting medium we used a 16:3:1 methanol–ethanol–water mixture, which remains quasi-hydrostatic up to 10 GPa [33]. The pressure was determined using the equation of state (EOS) of copper (Cu) [34] with a precision of ±0.05 GPa. The structural analysis was performed by employing the Rietveld technique using Fullprof [35].

Figure 2. The Rietveld profile fit for tetragonal rutile-type phase at 0.6 GPa. The asterisk (*) is the (111) diffraction peak of copper used to determine pressure. The vertical bars indicate the calculated positions of diffraction peaks of the sample.

Figure 2. The Rietveld profile fit for tetragonal rutile-type phase at 0.6 GPa. The asterisk (*) is the (111) diffraction peak of copper used to determine pressure. The vertical bars indicate the calculated positions of diffraction peaks of the sample.

3. Results and Discussion

Figure 2 shows the powder XRD pattern at the lowest pressure measured in the DAC (0.6 GPa) including the Rietveld refinement. The diffraction pattern can be undoubtedly assigned to the rutile-type structure. In the pattern there is an extra weak peak denoted by the asterisk symbol (*) which is assigned to the copper grain used as pressure marker. The unit-cell parameters at 0.6 GPa are a = 4.728(1) Å and c = 3.184(1) Å. The goodness of fit parameters are Rwp = 6.16 %, Rp = 3.4 %. Figure 3 shows XRD patterns of nanocrystalline Co-doped SnO₂ at representative pressures.

There are no noticeable changes in the diffraction patterns up to 13 GPa, other than peaks shift to higher angle due to the contraction of unit-cell parameters. All the diffraction peaks up to 13 GPa could be successfully indexed to the low-pressure (LP) rutile-type phase. In addition to the peaks assigned to SnO₂, a peak due to copper (marked by *), used to determine pressure, can be identified in all the x-ray diffraction profiles. We have observed a systematic shift in all the diffraction peaks to higher 2θ due to lattice compression. At 14.2 GPa we have observed few discernible changes in the diffraction profile, which are indicative of a phase transformation. The cell metric is no longer consistent with P4₂/mnm symmetry and another crystallographic phase with symmetry Pnmm occurs, corresponding to orthorhombic CaCl₂-type structure. In particular, at the pressure of 14.2 GPa, the broadening of the (101) and (111) diffraction peaks of rutile along with the splitting of the (210) diffraction peak were observed. The diffraction profile could not be well indexed to tetragonal rutile-type phase. On the other hand, the orthorhombic CaCl₂-type structure gives the correct positions of diffraction peaks as indicated by unbiased refinements. Furthermore, this structural phase transition agrees with earlier investigations on rutile SnO₂ [16,17,18,19,20,21,22].

On further increase of pressure, the orthorhombic CaCl₂-type structure continues stable till 20.9 GPa, which is the highest pressure in this investigation. On release of pressure, we have found that the phase transition is reversible. In the experiment, we have not observed any coexistence of phases. This and the group-subgroup relationship between space groups P4₂/mnm and Pnnm suggest that the transition could be of second-order in nature, which agrees with conclusions from previous XRD measurements on doped and un-doped nanocrystalline SnO₂ [27]. However, as we will explain below, at the phase transition, we have detected a volume discontinuity, which undoubtedly supports a first-order nature for the phase transition in Co-doped SnO₂. Notice that up to now this is the only SnO₂ nanomaterial where such a volume discontinuity has been detected. We speculate it could be related to changes in the strong anti-ferromagnetic super-exchange interaction between Co ions that exist in Co-doped SnO₂ [36]. Another hypothesis to explain the abrupt volume collapse is the influence of structural defects that are induced by Co-doping [37]. Additional studies are needed to fully understand the distinctive behavior of Co-dopen SnO₂. The Rietveld refinement of the XRD patterns profile measured for the HP phase at 14.2 GPa is shown in Figure 4. In addition to the peaks from the sample and Cu, we observed a weak contribution of the gasket, which makes the peak (200) of the HP phase to become asymmetric. The unit-cell parameters of the HP phase at 14.2 GPa are a = 4.675(2) Å, b = 4.572(3) Å, and c = 3.151(1) Å. The goodness of fit parameters are Rwp = 9.84 % and Rp = 7.79 %.

The pressure evolution of the lattice parameters of Co-doped nanocrystalline SnO₂ in the tetragonal rutile-type and orthorhombic CaCl₂-type phases is shown in Figure 5. The pressure dependence of the unit-cell volume is reported in Figure 6. We have observed a discontinuity in volume (-ΔV/V ~ 1.0%) at the transition pressure, indicating a first-order nature of the structural phase transition. Using the EOSFIT software [38] the linear compressibility of each phase was calculated. We have observed that in the rutile-type phase SnO₂ the axial compressibilities are highly anisotropic. The a-axis is more compressible than the c-axis, as is evident from the increase of the c/a ratio from 0.673 at ambient pressure to 0.681 at 13 GPa. The linear compressibility of the c-axis is found to be K_c = 1.02(4) × 10⁻³ GPa⁻¹, which is almost half than a-axes compressibility K_a = 1.72(6) × 10⁻³ GPa⁻¹. This anisotropic behavior in compressibility is in quite agreement with earlier reported values of bulk SnO₂ samples [30,39,40] and even for un-doped, Fe-doped, and V-doped nanocrystalline SnO₂ [27,28]. However, the values of axial compressibility obtained in this investigation are slightly higher compared to earlier investigations [27,28,30,39,40]. In case of orthorhombic CaCl₂-type phase the axial compressibilities are found to be highly anisotropic as well, the b-axis is more compressible compared to other two axes. The linear compressibility of all the three axes is found to be as K_a = 1.65(4) × 10⁻³ GPa⁻¹, K_b = 1.90(1) × 10⁻³ GPa⁻¹ and K_c = 1.05(2) × 10⁻³ GPa⁻¹.

The P–V data of the rutile phase of SnO₂ in the tetragonal phase, fitted to the second-order Birch–Murnaghan equation of state, gives the bulk modulus to be B_o = 213(9) GPa with B₀´ = 4. During this fit the bulk modulus B_o and ambient pressure volume V_o are kept as free variables. This value is in a good agreement with previously reported values for bulk [16], and even for un-doped, Fe-doped nanocrystalline SnO₂ [27]. These values are summarized in Table 1. However, in V-doped nanocrystalline samples bulk modulus is found to be decreasing with increasing doping concentrations, in the range of 185-142 GPa [28]. We can conclude that in contrast with V-doping, a 10% doping with Co and Fe does not affect the mechanical properties of SnO₂. A possible reason for it, it could be the fact that V has larger ionic radii than Fe and Co. For the high-pressure phase the bulk modulus obtained is B₀ = 228(9) GPa. The slight increase of the bulk modulus at the phase transition is consistent with the fact that the HP phase is denser than the low-pressure phase due to the volume contraction observed at the transition. However, both bulk moduli agree within error bars; 213(9) and 228(9) GPa. In this regard, the behavior is more similar to the behavior of bulk SnO₂ [16], where no changes are detected in the bulk modulus at the transition (see Table 1), than to the behavior of the bulk modulus for Fe-doped and V-doped nanocrystalline SnO₂ [27,28], where a 15 % increase of the bulk modulus was reported after the transition. At the moment, it is not clear that the different behavior is inherent to the different doping or to difference on non-hydrostatic stresses in experiments [41]. Future studies are needed to clarify this issue. On the other hand, it should be noticed that in the equation of state, the volume at zero pressure (V₀) and B₀ are correlated. Then, large uncertainties on V₀ could affect the value of B₀ [42]. This can be the case of SnO₂, where the phase transition is reversible and therefore no data is available for the HP phase at pressures close to ambient pressure. This means that the determination of V₀ comes from the extrapolation of data measured above 14 GPa, which could lead to large errors in the determination of V₀ which will necessarily propagate to the determination of B₀.

4. Conclusions

In summary, our in-situ synchrotron powder x-ray diffraction study on nanocrystalline Co-doped SnO₂ shows that it undergoes a first-order structural phase transition from the tetragonal rutile to the orthorhombic CaCl₂-type phase at ∼14.2 GPa. The crystal structure of the high-pressure phase is the same high-pressure structure observed in bulk, doped and undoped nanocrystalline samples of SnO₂. We have not observed any coexistence of phases during the phase transition. The compressibility behavior was found to be highly anisotropic in both phases, and the value of bulk modulus for the tetragonal rutile phase is in good agreement with earlier measurements. However, for orthorhombic CaCl₂-type phase there our bulk modulus is 10 % smaller than the same parameter in Fe- and V-doped SnO₂. The high-pressure phase was observed to remain stable till 20.9 GPa. On release of pressure the phase transition is found to be reversible.