Optimal parameters for synthesizing single phase

Journal of Alloys and Compounds 574 (2013) 278–282

Contents lists available at SciVerse ScienceDirect

Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom

Optimal parameters for synthesizing single phase spinel-type Co2SnO4 by sol–gel technique: Structure determination and microstructure evolution J.A. Aguilar-Martínez a,b,⇑, M.A. Esneider-Alcala a, M.B. Hernández b, M.I. Pech Canul c, S. Shaji b a Centro de Investigación en Materiales Avanzados, S.C. (CIMAV), Alianza Norte No. 202, Parque de Investigación e Innovación Tecnológica (PIIT), Nueva Carr. Aeropuerto Km. 10 Apodaca, Nuevo León 66600, Mexico b Facultad de Ingeniería Mecánica y Eléctrica, Universidad Autónoma de Nuevo León, San Nicolás de los Garza, Nuevo León 66450, Mexico c Centro de Investigación y de Estudios Avanzados del Instituto Politécnico Nacional, Unidad Saltillo. Av. Industria Metalúrgica No. 1062, Parque Industrial, Ramos Arizpe Coah. 25900, Mexico

a r t i c l e

i n f o

Article history: Received 12 February 2013 Received in revised form 3 May 2013 Accepted 21 May 2013 Available online 30 May 2013 Keywords: Spinels Sol–gel X-ray techniques Thermal analysis

a b s t r a c t Powder samples of spinel Co2SnO4 were successfully synthesized by the sol–gel technique at three different temperatures: 900, 1300 and 1400 °C for 1, 5 and 8 h. XRD patterns and SEM images provided evidence of the structural and morphological evolution during the formation of the spinel phase. Thermal analysis confirmed the reaction pathway in which the spinel is formed by the reaction between SnO2 (formed from SnCl4) and CoO which in its turn is produced during the decomposition reaction of Co3O4 [formed from Co(NO3)2] into CoO and O2. Determination of the quantitative phase at each combination of temperature and time as well as the lattice parameters was possible by the Rietveld refinement. The optimal parameters for the complete formation of Co2SnO4 are 1400 °C for 5 h. It is worth mentioning that at 1300 °C for 1 h, the percentage of the spinel obtained is considerably high (94.7%). Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction In the continual research of ceramic materials for varistor applications, numerous authors have put special interest in the ceramic systems based on tin oxide (SnO2) exploring different dopants and processing technologies in order to obtain improved electrical properties as well as to gain a deeper understanding of the phenomena involved in the synthesis stage [1–10]. As it has been frequently reported, cobalt oxide (Co3O4) has been normally chosen as the preferred dopant to promote densification of SnO2 ceramics [11–14]. Moreover, it has been described that Co2SnO4 compound is usually found in the final microstructure of the ceramic material (as the result of the imperfect reaction between the raw materials), typically segregated near the grain boundaries of tin oxide [12,15,16], though the role it plays on the microstructure and electrical performance of the varistor ceramic has not been explained in detail yet. Furthermore, to the best of the authors’ knowledge, there is no established commercial route for the production of Co2SnO4, though some

⇑ Corresponding author at: Centro de Investigación en Materiales Avanzados, S.C. (CIMAV);Alianza Norte No. 202, Parque de Investigación e Innovación Tecnológica (PIIT), Nueva carr. Aeropuerto Km. 10 Apodaca, Nuevo León 66600, Mexico. Tel.: + 52 81 11560805; fax: +52 81 11560820. E-mail addresses: [email protected], [email protected] (J.A. Aguilar-Martínez). 0925-8388/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jallcom.2013.05.146

researchers have synthesized it as anode material for Li-ion batteries [17–21]. Thus, there are strong reasons for synthesizing Co2SnO4 as a single phase. The reported characteristics of Co2SnO4 indicate that it has a spinel-structure. The general formula of oxide spinel is AB2O4, and the distribution configuration is represented as IV (A1xBx)VI(B2xAx)O4. Normal spinels have x = 0, whereas inverse spinels have x = 1; any distribution is possible between the extremes [22]. The unit cell contains eight unit formulas, has cubic symmetry, space group Fd3m, and a cell edge close to 8 Å [23]. Solid state reaction synthesis of Co2SnO4 samples has proved to result in poor cycle stability due to larger particle sizes and wider size distributions [21]. Sol–gel method has demonstrated to be an efficient synthesis technique for ceramic oxide compounds due to its potential advantages over the traditional solid-state reaction and co-precipitation methods for achieving a homogeneously mixing of the component cations at atomic scale, lowering the synthesis temperature quite effectively. These advantages were considered when it was decided to follow this route for obtaining Co2SnO4. The results of a previous study on this compound reported a wider range of temperature to obtain the Co2SnO4 phase [24]. In the current research, Co2SnO4 has been prepared by the sol–gel process to thoroughly study the evolution during the formation of this compound as well as to define the amount of spinel obtained at each condition of temperature and time. At first glance, the percentage of spinel obtained is low but as temperature rises

J.A. Aguilar-Martínez et al. / Journal of Alloys and Compounds 574 (2013) 278–282

279

to 1400 °C for 5 h, the formation of spinel is complete and therefore higher temperatures are no longer needed. XRD patterns confirmed the formation of the phase as well as the percentage of spinel obtained while SEM results provided evidence of the microstructure evolution. 2. Experimental procedure The Co2SnO4 spinel was synthetized by the sol–gel method using stannic chloride pentahydrate SnCl45H2O (Spectrum, 99.90%) and cobalt nitrate hexahydrate Co(NO3) 6H2O (Alfa Aesar, 99.74%) as starting materials. Ethylene Glycol CH2OHCH2 (CTR, 99.96%) was used as chelating agent. Ethyl alcohol (JT. Baker, 99.6%) and distilled water were used as solvents. In order to prepare 1 g of spinel, 1.9389 g Co(NO3) 6H2O and 1.1668 g SnCl45H2O were dissolved separately into a 15 ml mixture of ethanol and distilled water in an equal volume ratio (1:1). The solutions containing the nitrate and the chloride were stirred separately in a beaker at a temperature of 60 °C until a transparent solutions was observed. Then, the solutions were mixed at 90 °C with 470 ml of distilled water in a beaker with moderate stirring for 24 h, promoting the hydrolysis and polymerization. The beaker was covered with aluminum foil with the aim of avoiding water evaporation during the stirring period. After 24 h of stirring, 10 ml of ethylene glycol were added to promote gelation. An hour later, the aluminum foil was removed from the beaker and the temperature was decreased to 60 °C without stirring until gel formation. Samples of the gel obtained were calcined at three different temperatures of 900, 1300 and 1400 °C for 1, 5 and 8 h, using heating and cooling rates of 10 °C/min. After cooling to room temperature, the products were removed from the furnace and prepared for characterization by powder Xray diffraction (XRD) and scanning electron microscopy (SEM). The presence of ceramic phases was determined by X-ray diffraction technique (XRD; PANalytical model Empyrean) with Cu Ka radiation (k = 1.5406 Å) operated at 45 kV, 40 mA and an X’Celerator detector in a Bragg–Brentano geometry. The scans were performed in the 2h range from 10° to 120° with a step scan of 0.016° and 80 s per step in a continuous mode. Structure refinements and phase identification were performed using the X’Pert HighScore Plus software, version 3.0d and the ICDD PDF4 plus database (ICDD International Centre for Diffraction Data, Newtown Square, PA). The products from the sol–gel processing were thermally analyzed in the range 30–1500 °C in a simultaneous TG–DSC–DTA equipment (TA-Instruments model SDT Q600). An amount of 20–30 mg of the samples were placed in a platinum crucible and heated at the rate of 10 °C/min in static air atmosphere, using alumina as reference material. Finally, the morphology of the as-prepared powders was observed by a scanning electron microscope (SEM, FEI – Nova Nano SEM 200). The Rietveld’s method was successfully applied for the determination of the quantitative phase of the all samples. This method is a least squares refinement procedure where the experimental step-scanned values are adapted to calculated ones. The profiles are considered to be known, and a model for a crystal structure available. The weight fraction (Wi) for each phase was obtained from the mathematical relationship for refinement:

Si ðZMVÞi Wi ¼ P ; j Sj ðZMVÞj

ð1Þ

where i is the value of j for a particular phase among the N phases present, Sj is the refined scale factor, Z is the number of formula units per unit cell, M is the molecular weight of the formula unit and V is the unit cell volume. Rietveld refinement has been done by adjusting major parameters: scale factor, flat background, zero-point shift, lattice parameters, orientation parameters, peak width parameters (U, V, W), asymmetry parameter and peak shape. Peak profiles were fitted with the pseudo-Voigt function.

Fig. 2. XRD patterns of the powders synthesized by sol–gel method at (a) 900, (b) 1300 and (c) 1400 °C for 1, 5 and 8 h.

The quality of fitting is judged through the minimization of weighted residual error (Rw), through a Marquardt least-squares program and is defined as:

Rwp

2X 31=2 wi ðIo  Ic Þ2 6 i 7 7 ; X 2 ¼6 4 5 wi Io

ð2Þ

l

The goodness of fit (GoF) is established by comparing Rwp with the expected error, Rexp.

2 Rexp Fig. 1. TG–DTG–DSC curves for the decomposition of the precursor prepared by the sol–gel method at a heating rate of 10 °C/min.

31=2

6 NP 7 7 ¼6 4Xw I2 5 i o

i

;

ð3Þ

280

J.A. Aguilar-Martínez et al. / Journal of Alloys and Compounds 574 (2013) 278–282

and time, Rietveld refinement was carried out. Table 1 shows detailed information obtained after Rietveld refinement of all samples, where good agreement values are reported. Considering the integrated intensity of the peaks as a function of structural parameters only, the Marquardt least-squares procedure was used for minimization of the difference between the observed and simulated powder diffraction patterns. The minimization was carried out using the reliability index parameter, Rwp (weighted residual error), RB (Bragg factor) and Rexp (expected error) [25]. Thus, the Rietveld refinement method from the X-ray patterns was used for lattice parameter determination and phase quantification of the powders calcined at 900, 1300 and 1400 °C. These results, together with the thermal analyses strongly support the fact that the spinel formation starts at 900 °C. Moreover, it has been suggested that in SnO2 ceramics doped with cobalt, like in the current work, the diffusion of a defect associate [CoSn + V0] allows to explaining the formation of the spinel (Co2SnO4) at the surface of the specimens [26,27]. More specifically, this is explained by the concentration gradient of oxygen vacancies between the surface and that of the interior of the sintered body of SnO2. The concentration of oxygen vacancies at the surface is lower due to the outward-diffusion of associate defects. Since the cobalt concentration at the surface increases and exceeds its limit of solubility in SnO2, eventually the spinel precipitates at the surface of the specimen. The evidence of the spinel is confirmed at 1400 °C, corresponding to the totality of the peaks in the diffractogram and phase quantified. As the temperature increases to 1300 °C, the Co3O4 phase is no longer detected by the X-ray diffractometer and though the SnO2 phase is still identified, the intensities of its peaks tend to diminish. When the temperature is further increased to 1400 °C and after 5 h of thermal treatment, the intensity of the crystallization peaks of the Co2SnO4 phase increases until complete crystallization is obtained. According to the pathway previously reported [24], the existence of CoO is essential; however in this experiment the XRD analysis did not reveal its presence. Two possible factors can account for this outcome. The first is related to the kinetics for spinel formation (reaction (6)), which might be occurring rapidly. At 1300 and 1400 °C while the amount of SnO2 gradually decreases with time increase, the quantity of the spinel augments in such a way that at 1400 °C for 5 and 8 h, only Co2SnO4 is detected. Therefore the best parameters to produce single phase spinel are 1400 °C

where Io and Ic are the experimental and calculated intensities, respectively, wi = 1/Io and N are the weight and number of experimental observations, and P is the number of fitting parameters. This leads to the value of goodness of fit:

GoF ¼ v2 ¼

Rwp ; Rexp

ð4Þ

Refinement continues until convergence is reached.

3. Results and discussion 3.1. Thermal analysis The thermal decomposition of the wet gel to synthesize spinel Co2SnO4 is presented in Fig. 1. An endothermic process with a weight loss about 1.46% with DTG maximum at 920 °C is observed, which is related to the in situ formation of CoO from Co3O4 (see Eq. (5)). As soon as the formation of CoO is completed, the reaction between the SnO2 and the CoO takes place to obtain the Co2SnO4 spinel, as it is described in Eq. (6). Below 1100 °C, there is an exothermic weight gain associated to oxygen being adsorbed on the surface of the spinel.

2Co3 O4 ! 6CoO þ O2

ð5Þ

920  C

2CoO þ SnO2 ! Co2 SnO4 ðspinelÞ

ð6Þ

After analyzing the thermal results, different calcining parameters were chosen in order to optimize the conditions needed to obtain the Co2SnO4 spinel. A detailed reaction pathway determination for the formation of the Co2SnO4 spinel has been already reported by the same research group [24]. 3.2. Phase analysis and lattice parameters determination Representative X-ray powder diffraction patterns of samples calcined at 900, 1300 and 1400 °C are shown in Fig. 2a–c. For samples treated at 900 °C, the identified peaks are consistent with the standard cards of Co3O4 (ICDD PDF # 04-003-0984), SnO2 (ICDD PDF # 04-005-5929) and Co2SnO4 (ICDD PDF # 04-008-2461) marked as O, ⁄, d, respectively. However, their amounts were different to some extent, as it can be observed in Table 1. In order to determine lattice parameters and to quantify the phases present in all samples as a function of calcining conditions of temperature

Table 1 Lattice parameters and phase quantification by the Rietveld refinement method for samples calcined at 900, 1300 and 1400 °C for 1, 5 and 8 h. Sample calcining temperature–time

Phase present

900-1

SnO2 Co3O4 Co2SnO4 SnO2 Co3O4 Co2SnO4 SnO2 Co3O4 Co2SnO4 SnO2 Co2SnO4 SnO2 Co2SnO4 SnO2 Co2SnO4 SnO2 Co2SnO4 Co2SnO4 Co2SnO4

900-5

900-8

1300-1 1300-5 1300-8 1400-1 1400-5 1400-8

Co3O4 and Co2SnO4 are cubic; SnO2 is tetragonal.

Lattice parameter (A) a

b

c

a=b=c

4.74 8.09 8.64 4.74 8.09 8.64 4.74 8.09 8.64 4.74 8.65 4.74 8.65 4.74 8.65 4.74 8.65 8.65 8.65

4.74 8.09 8.64 4.74 8.09 8.64 4.74 8.09 8.64 4.74 8.65 4.74 8.65 4.74 8.65 4.74 8.65 8.65 8.65

3.18 8.09 8.64 3.18 8.09 8.64 3.18 8.09 8.64 3.18 8.65 3.18 8.65 3.18 8.65 3.18 8.65 8.65 8.65

90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90

wt (%)

Rwp (%)

Rexp (%)

RP (%)

v2

34.5 31.3 34.2 23.8 24 52.2 23.7 23.7 52.4 5.3 94.7 4.3 95.5 3.4 96.6 2.8 97.2 100 100

1.98

1.94

1.52

1.04

1.96

1.77

1.39

1.21

2.34

1.77

1.63

1.73

3.88

1.79

2.16

4.67

3.87

1.84

2.19

4.42

3.79

1.94

2.16

3.80

4.10

1.85

2.29

4.91

3.59 3.69

2.28 2.08

2.14 2.10

2.48 3.15

J.A. Aguilar-Martínez et al. / Journal of Alloys and Compounds 574 (2013) 278–282

281

Fig. 3. SEM micrographs for powder samples obtained by sol–gel at different temperature and times: For 900 °C; (a) 1 h, (b) 5 h and (c) 8 h; 1300 °C; (d) 1 h, (e) 5 h and (f) 8 h and for 1400 °C; (g) 1 h, (h) 5 h and (i) 8 h.

for 5 h. The second reason could be in terms of the amounts of CoO present in the system, well below the limits of detection of the XRD equipment. Most likely the combination of both factors is responsible for the observed situation, and more detailed investigations would be required in order to elucidate the mechanism involving CoO. The authors are currently working in that direction, as the next phase of the investigation. The microstructure evolution supporting the XRD results is shown in the SEM analysis, Section 3.3. 3.3. Microstructure characterization by SEM The microstructure characteristics of the powder samples obtained after calcining at 900, 1300 and 1400 °C at different calcination times are shown in Fig. 3a–i. A significant difference in particle size and morphology can be observed in SEM micrographs evidently affected by the calcination conditions. At 900 °C (Fig. 3a– c) and accordingly with XRD patterns, the grain size and shape are quite inhomogeneous due to the mixture of the three phases present in the sample. At 1300 °C (Fig. 3d–f), a grain growth is evident as the reaction to form Co2SnO4 takes place. At 1400 °C

(Fig. 3g–i), the formation of the single phase of Co2SnO4 is observed as the grain size is greater, grain boundaries are well defined and the morphology is homogeneous. 4. Conclusions An optimization of the post-synthesis parameters to obtain Co2SnO4 by the sol–gel method, starting from the proposed precursor materials was conducted. Moreover, on the basis of thermal analyses of the sol–gel mixture, a reaction pathway to form Co2SnO4 by the reaction between CoO and SnO2 is put forward. The optimal conditions to obtain homogeneous single-phase 100% Co2SnO4 is 1400 °C for 5 h though a high percentage (94.7%) of Co2SnO4 can be synthesized at 1300 °C for 1 h, as confirmed by XRD analysis and Rietveld refinement. SEM analysis provided strong evidence of the evolution in size and morphology of the grains to complete the formation of Co2SnO4. The proposed route in this work paves the way to incorporate the spinel phase into SnO2 ceramics for varistor applications and study its effect on the microstructure and electrical properties.

282

J.A. Aguilar-Martínez et al. / Journal of Alloys and Compounds 574 (2013) 278–282

Acknowledgments Authors gratefully acknowledge Mr. Alberto Toxqui for assistance in the thermal characterization by TGA/DSC and Ms. Nayeli Pineda Aguilar for her valuable help in the microstructure characterization by SEM. J.A. Aguilar Martínez also thanks the CONACyT, México for his doctoral scholarship. References [1] J.A. Aguilar-Martínez, M.I. Pech-Canul, M.B. Hernández, A.B. Glot, Edén Rodríguez, L. García Ortiz, Rev. Mex. Fis. 59 (2013) 6–9. [2] J. He, Z. Peng, Z. Fu, C. Wang, X. Fu, J. Alloys Comp. 528 (2012) 79–83. [3] H. Bastami, E. Taheri-Nassaj, Ceram. Int. 38 (2012) 265–270. [4] I. Safaee, M.A. Bahrevar, M.M. Shahraki, S. Baghshahi, K. Ahmadi, Solid State Ionics 189 (2011) 13–18. [5] J.R. Ciórcero, S.A. Pianaro, G. Bacci, A.J. Zara, S.M. Tebcherani, E. Longo, J. Mater. Sci. Mater. Electron. 22 (2011) 679–683. [6] J. Fan, H. Zhao, Y. Xi, Y. Mu, F. Tang, R. Freer, J. Eur. Ceram. Soc. 30 (2010) 545– 548. [7] S.R. Dhage, V. Ravi, O.B. Yang, J. Alloys Comp. 466 (2008) 483–487. [8] R. Parra, J.E. Rodríguez-Páez, J.A. Varela, M.S. Castro, Ceram. Int. 34 (2008) 563– 571. [9] M.A.L. Margionte, A.Z. Simões, C.S. Riccardi, A. Ries, F.M. Filho, L. Perazolli, et al., Mater. Lett. 60 (2006) 142–146. [10] W.X. Wang, J.F. Wang, H.C. Chen, W.B. Su, B. Jiang, G.Z. Zang, C.M. Wang, P. Qi, Ceram. Int. 31 (2005) 287–291.

[11] J.A. Aguilar-Martínez, M.B. Hernández, A.B. Glot, M.I. Pech-Canul, J. Phys. D Appl. Phys. 40 (2007) 7097–7102. [12] J.A. Aguilar-Martínez, A.B. Glot, A.V. Gaponov, M.B. Hernández, J. Guerrero-Paz, J. Phys. D Appl. Phys. 42 (2009), http://dx.doi.org/10.1088/0022-3727/42/20/ 205401. [13] P.A. Santos, S. Maruchin, G.F. Menegoto, A.J. Zara, S.A. Pianaro, Mater. Lett. 60 (2006) 1554–1557. [14] G. Brankovic, Z. Brankovic, M.R. Davolos, M. Cilene, J.A. Varela, Mater. Charact. 52 (2004) 243–251. [15] B.Ch. Kim, J.I. Jung, J.H. Lee, J.J. Kim, Solid State Ionics 144 (2001) 321–327. [16] M.L. Moreira, S.A. Pianaro, A.V.C. Andrade, A.J. Zara, Mater. Charact. 57 (2006) 193–198. [17] P. He, Z. Xie, Y. Chen, F. Dong, H. Liu, Mater. Chem. Phys. 137 (2012) 576–579. [18] G. Wang, Z.Y. Liu, P. Liu, Electrochim. Acta 56 (2011) 9515–9519. [19] Y. Qi, N. Du, H. Zhang, P. Wu, D. Yang, J. Power Sources 196 (2011) 10234– 10239. [20] R. Alcántara, G.F. Ortiz, P. Lavela, J.L. Tirado, Electrochem. Commun. 8 (2006) 731–736. [21] P.A. Connor, J.S.T. Irvine, J. Power Sources 97–98 (2001) 223–225. [22] B.A. Giovanni, Am. Miner. 87 (2002) 838–844. [23] K.E. Sickafus, J.M. Wills, J. Am. Ceram. Soc. 82 (1999) 3279–3292. [24] J.A. Aguilar-Martínez, M.I. Pech-Canul, M. Esneider, A. Toxqui, S. Shaji, Mater. Lett. 78 (2012) 28–31. [25] L. Lutterotti, P. Maistrelli, J. Appl. Cryst. 25 (1992) 459–462. [26] R. Metz, D. Koumeir, J. Morel, J. Pansiot, M. Houabes, M. Hassanzadeh, J. Eur. Ceram. Soc. 28 (2008) 829–835. [27] M.F. Casula, G. Concas, F. Congiu, A. Corrias, D. Loche, C. Marras, G. Spano, J. Nanosci. Nanotechnol. 11 (2011) 10136–10141.