- Email: [email protected]
Phase transition and electrochemical capacitance of mechanically treated manganese oxides
Journal of Alloys and Compounds 414 (2006) 137–141
Phase transition and electrochemical capacitance of mechanically treated manganese oxides Akira Taguchi, Sachio Inoue, Satoshi Akamaru, Masanori Hara, Kuniaki Watanabe, Takayuki Abe ∗ Hydrogen Isotope Research Center, Toyama University, Gofuku 3190, Toyama 930-8555, Japan Received 25 January 2005; accepted 17 February 2005 Available online 1 September 2005
Abstract Mechanical grinding (MG) leads to a sequential phase transition of ␥-MnO2 to form the thermodynamically stable ␣-Mn2 O3 and, subsequently, Mn3 O4 depending on the duration of MG treatment. By MG treatment for 150 h, ␣-Mn2 O3 became a predominant species. A subsequent transition to Mn3 O4 was observed on further MG treatment for 200 h. The particle size of the resultant manganese oxides was reduced to as small as 20 nm after 300 h of MG treatment. The change in the specific capacitance of manganese oxides obtained was monitored by XRD and electrochemical experiments. It was observed that the specific capacitance of MG-treated manganese oxides depends on the amount of ␥-MnO2 left without phase transition. © 2005 Elsevier B.V. All rights reserved. Keywords: Manganese oxides; Phase transition; Mechanical grinding; Supercapacitor
1. Introduction Recently, the development of novel energy storage devices such as batteries and capacitors has been extensively studied in order to improve efficiency and reliability of electric utility systems such as solar-cell reservoirs or power quality regulators for computers, electric and hybrid vehicles, and fuel cell systems [1–4]. The oxide and hydroxide of ruthenium provide a good capacitance in aqueous and non-aqueous electrolyte systems [5,6]. However, the noble metal ruthenium, being expensive, has its own limitations. Hence, the fabrication of substitutional supercapacitive materials for ruthenium (hydro)oxides is gaining importance. Presently, several metal oxides and hydroxides, for example, those of Ni, Co, V, and Mn, are being studied extensively [8–10]. Among these, manganese oxide (MnO2 ) is one of the promising materials for this purpose due to its availability, cost-effectiveness and non-toxicity.
∗
Corresponding author. Tel.: +81 76 445 6933; fax: +81 76 445 6931. E-mail address: [email protected] (T. Abe).
0925-8388/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2005.02.108
The design and synthesis of manganese oxide electrodes is a key field for improving upon its electrochemical properties. In the case of super capacitor electrodes which comprise manganese oxide, it has been widely observed that MnO2 exhibits a rather high performance in comparison to its other oxides such as Mn(OH)2 , Mn2 O3 , and Mn3 O4 [11–13]. For example, amorphous hydrous manganese oxide (a-MnO2 ·nH2 O) fabricated from MnSO4 ·5H2 O solution via an anodic deposition yielded a specific capacitance in the range of 265–320 F/g between 0 to 1.0 V in 0.1 M Na2 SO4 solution [12]. A thin film of MnO2 electrode prepared by the dip-coating of colloidal MnO2 exhibits an excellent specific capacitance value of unto 698 F/g that is comparable to the ruthenium (hydro)oxides [5,7,13]. However, a structural change in the active component of MnO2 causes a loss of the electrochemical activity. This is due to of the significant change in the conductivity of manganese oxides—the conductivity of manganese oxides involving trivalent Mn(III) is rather low as compared to that of MnO2 or amorphous manganese (oxyhydro)oxides involving Mn(IV); this is attributed to the Jahn–Teller distortion of the MnIII O6 octahedral [8,14]. A few reports pertaining to
138
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141
the quantitative investigations of the content and capacitance of a mixture of MnO2 and Mn2 O3 have been published. Much attention has been paid to a mechanical grinding (MG) technique by means of ball milling, which is a powerful tool to reduce the particle size or mix several components to homogeneous dispersion [15–17]. Besides these fine mixing, a phase transition by MG treatment has been reported by Wakamatsu et al. [18] and Kuznetsov et al. [19]. Wakamatsu et al. have shown that rutil and/or brookite were obtained by MG treatment of anatase [18]. The mechanical treatment of -FeOOH caused the phase transition to FeO, via ␣-Fe2 O3 and Fe3 O4 as intermediate phases [19]. Depending on the duration of MG treatment, we observed a sequential phase transition of manganese oxides of ␥-MnO2 to ␣-Mn2 O3 and then subsequently to Mn3 O4 . The resultant manganese oxides are suitable examples to investigate the relationship between the composition of manganese oxides and their capacitances. In this paper, we report a sequential phase transition of manganese oxides by MG treatment, and the change in the specific capacitances of the resultant manganese oxides.
2. Experimental Commercially available MnO2 (Kanto Chemical Co.) was mechanically ground using a planet type ball mill (Fritsch, pulverized 5). In a container (26 mm in diameter and 27 mm in height) with 30 balls (each 4 mm in diameter), 4 g of MnO2 was placed and ground at 800 rpm for sufficient periods of time. Hereafter, these MG-treated samples are denoted as MG-n; where n is the period (in h). The structure and crystalline of the obtained samples were investigated by a powder X-ray diffraction (XRD, Philips, PW1825/00) and transmission electron microscopy (TEM, TOPCON, EM-002B). The manganese oxides prepared by the MG treatment were electrochemically investigated. Manganese oxide that had been treated for a specific time was mixed with a small amount of polytetrafluoroethylene (PTFE), which was used as a binder, and carbon powder (10 wt.%). A few drops of nmethylpyrrolidine were added to the mixture to prepare the sample paste. A Pt plate (10 mm × 10 mm), used as a current collector, was immersed in concentrated nitric acid for 2 h and dried in an oven at 80 ◦ C; subsequently, it was washed with distilled water. The sample paste was spread on the Pt plate, and it was dried at 60 ◦ C for 3 h. The amount of manganese oxides on the composite plate was measured by subtracting the weight of the Pt plate from that of the plate with the sample. In the case of the manganese oxides containing a mixture of ␥-MnO2 , ␣-Mn2 O3 , and Mn3 O4 , the composition of each component was calculated from the relative peak intensity for the most intense peak of each. Cyclic voltammetry (CV) was carried out in 1 M KOH aqueous solution using a threeelectrode system with a Pt counter electrode and a saturated calomel electrode (SCE) under potential control conditions (EG&G, potentiostat Model 263A).
Fig. 1. X-ray diffraction patterns of manganese oxides of (a) untreated (MG0), (b) MG-80 (an offset by 200 cps.), (c) MG-120 (an offset by 400 cps.) and (d) MG-300 (an offset by 800 cps.). Symbols (䊉), () and () correspond to the peaks of ␥-MnO2 , ␣-Mn2 O3 , and Mn3 O4 , respectively.
3. Results The manganese oxide used in this study was ␥-MnO2 (nsutite, JCPDS 4-779 and 17-510), and its XRD pattern is shown in Fig. 1(a). Its particle size could not be estimated due to the very low diffraction intensity and the poor resolution of diffraction signals. The poor crystallinity is probably due to the presence of a small structural defect in ␥-MnO2 [8]. The phase composition and crystalline structure of ␥-MnO2 changed significantly after the MG treatment. As shown by the XRD pattern in Fig. 1(b), new diffraction signals assignable to ␣-Mn2 O3 appeared at 2θ of 32.9◦ , 42.8◦ and 43.6◦ after the MG treatment for 80 h (MG-80). These peaks could be indexed to (2 2 2), (3 2 0) and (3 3 2), respectively; subsequent grinding for 120 h (MG-120) supported this observation. The diffraction signal of the ␣-Mn2 O3 phase (bixbyite, JCPDS 31-825 and 41-1442) intensified and those of ␥-MnO2 almost completely disappeared during the duration of this MG treatment. The particle sizes of the obtained ␣-Mn2 O3 were calculated with the diffraction signals of (2 2 2), (4 4 0), and (6 2 2), and were found to be 16.0, 13.4, and 15.6 nm, respectively, using the Scherrer’s equation. Furthermore, on increasing the duration of MG treatment up to 300 h (MG-300), some diffraction signals corresponding to Mn3 O4 (hausmannite, JCPDS 1-1127) appeared along with ␣-Mn2 O3 signals. Fig. 2 shows a change in the intensities of the diffraction signals of ␥-MnO2 , ␣-Mn2 O3 , and Mn3 O4 , depending on
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141
139
Fig. 2. Change in the X-ray diffraction intensities depending on the duration of MG treatment. Symbols (䊉), () and () correspond to the peaks of (0 2 1) in ␥-MnO2 , (2 2 2) in ␣-Mn2 O3 , and (1 0 1) in Mn3 O4 , respectively.
the duration of MG treatment. In this case, the most intense diffraction signals of each compound at 2θ = 37.5◦ (indexed to (0 2 1) in ␥-MnO2 ), 33.9◦ ((2 2 2) in ␣-Mn2 O3 ) and 18.0◦ ((1 0 1) in Mn3 O4 ) were employed. The signal intensity of the starting material of ␥-MnO2 remained almost constant for the MG treatment duration of 60 h, and then decreased slightly after the MG treatment for 80 h. The diffraction signals of ␣-Mn2 O3 were observed in 80 h, and its signal intensity increased significantly up to 150 h. At this point, ␣-Mn2 O3 became a dominate species. The peak intensity of ␣-Mn2 O3 decreased on further MG treatment, and a new signal corresponding to Mn3 O4 was observed in the MG-200 sample. The peak intensity of Mn3 O4 was increased simultaneously and gradually as that of ␣-Mn2 O3 decreased. The standard enthalpy of formation (f H◦ ) of MnO2 , Mn2 O3 , and Mn3 O4 were −520.0, −959.0 and −1387.8 kJ/mol, respectively [20]. Therefore, it was assumed that the sequential phase transition toward a thermodynamically stable composition occurred due to mechanical grinding. The phase transition toward the thermodynamically stable compounds has been widely observed in several metal oxides, during the MG treatment. For example, Wakamatsu et al. have shown a phase transition of TiO2 from anatase to brookite and/or rutile by ball milling [18]. They revealed that the longer or stronger the milling conditions for the anatase powder the shorter is the duration of structural change. The chemical reduction reaction of -FeOOH to form stable FeO, via ␣-Fe2 O3 and Fe3 O4 as intermediate phases, has also been reported by Kuznetsov et al. [19]. These facts support the possible phase transition of manganese oxides by MG treatment. The mechanism for this reduction reaction is not clear. One possible reason is that the metal components of the balls and container in the ball mill instruments participate in this reduction reaction. The reduction in particle sizes of manganese oxides by MG treatment was investigated by a TEM. Fig. 3 shows the TEM photographs of manganese oxides before and after MG
Fig. 3. TEM images of (a) untreated (MG-0) and (b) MG-300.
treatment. An amorphous bulky particle with a particle size of several ten nm is observed in the sample before MG treatment (MG-0). On the other hand, flake particles with a particle size of less than 20 nm were observed in the MG-300 sample. For example, the dimensions of the particle highlighted in the open circle in Fig. 3(b) were estimated to be 10 and 20 nm in shorter and longer diameters, respectively. These values are consistent with the ones calculated from XRD signals using the Scherrer’s equation (13.4–16.0 nm). Significant reduction in particle size by MG treatment has been widely reported. In our previous report, it was revealed that the particle size of RuO2 could be successfully reduced from 17.0–15.0 nm by MG treatment [15]. These values are fairly consistent with those observed in the present study, suggesting a limitation of particle downsizing by our instrument. Manganese oxides are promising candidates for new capacitive materials as alternatives for those based on ruthenium. Fig. 4 shows the cyclic voltammograms of the composite electrodes of MG-treated manganese oxides and that of the bare Pt plate (used as a current collector) in a 1 M KOH solution. The scan rate of the potential was 20 mV/s. The CV obtained in the MG-0 modified Pt plate with a load-
140
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141
Fig. 4. Cyclic voltammograms of manganese oxides-modified electrode in a 1 M KOH solution. Bold solid line: MG-0; fine solid line: MG-60; broken line: MG-120; dotted line: bare Pt; scan rate: 20 mV/s.
ing amount of 0.36 mg had a rectangular shape, showing the typical behavior of a capacitor. The specific capacitance of MG-0 was calculated from the difference between CVs at the MG-0 modified electrode and at the Pt plate alone, and was found to be 73.7 F/g. The characteristic rectangular CV shape has been maintained in the sample that was MG-treated up to 60 h; the capacitance of MG-60 was estimated to be 64.6 F/g (0.39 mg loading). These curves show no peaks, indicating that the electrode capacitor is charged and discharged at a constant rate throughout the entire voltammetric cycle. However, the values of observed specific capacitances were far smaller than those reported, e.g., 265–320 F/g for amorphous MnO2 -deposited electrode (scan rate of 25 mV/s in 0.1 M Na2 SO4 ) [12]. By further MG treatment, the CV curve of the modified electrode exhibited considerable changes and distortion (Fig. 4). The capacitance of MG-120, corresponding to ␣-Mn2 O3 on the basis of the XRD measurement, was significantly decreased to 18.9 F/g (0.37 mg loading). This value is less than 30% of the initial capacitance observed in the starting material of ␥-MnO2 . In addition, the anodizing current demonstrated by the CV curve is slightly larger than the cathodizing one; consequently, the reversibility in terms of the ratio of anodic and cathodic current has been changed. The structural change of manganese oxides electrode causes a loss of electrochemical activities. In this study, specific capacitances of modified electrodes were examined in various manganese oxides at different durations of MG treatment. Fig. 5 shows a relationship between the specific capacitance and XRD signal intensity of ␥-MnO2 used in this study. In order to estimate the composition of ␥-MnO2 and ␣-Mn2 O3 in the mixture, their relative amounts were calculated from the diffraction peak intensity (I(hkl) ). In this case, the relative intensities were calculated for the most intense signals of MG-0 in ␥-MnO2 and MG-150 in ␣-Mn2 O3 by the following equations. C␥-MnO2 = C␣-Mn2 O3 =
I(0 2 1) I(0 2 1)
in MG-n in MG-0
I(2 2 2) in MG-n I(2 2 2) in MG-150
Fig. 5. Relationship between the specific capacitance of manganese oxidesmodified electrodes and XRD signal intensity of the (0 2 1) peak of ␥-MnO2 . The error bar represents the difference in specific capacitance that arises from an estimation of the composition.
Under ideal conditions, the sum of the fraction of compositions of C␥-MnO2 and C␣-Mn2 O3 should be 1.0; however, it was lower than this value. The difference, which mainly arose from an imprecise definition of equations, is represented as an error bar in Fig. 5. In this figure, a linear decrease in the specific capacitance with a decrease in the diffraction signal intensity of ␥-MnO2 is clearly visible. In other words, the specific capacitance decreases linearly with a decrease in the amount of ␥-MnO2 . Thus, it can be concluded that the content of ␣-Mn2 O3 is directly related to the capacitance of the ␥-MnO2 electrode.
4. Conclusion In conclusion, we have reported the phase transition of manganese oxides depending on the duration of the MG treatment. The starting material ␥-MnO2 began to change to ␣-Mn2 O3 in 80 h, and subsequently transformed into Mn3 O4 by further milling for 200 h. The phase transition sequence of ␥-MnO2 , ␣-Mn2 O3 and Mn3 O4 follows a thermodynamic stability pattern of manganese oxides as predicted by the standard enthalpy of formation. Due to the formation of MnIII O6 octahedral, which has a rather low conductivity in comparison to that of MnIV , the capacitances of manganese oxides decreased linearly with a decrease in the crystallinity of ␥MnO2 .
References [1] M. Winter, J.O. Besenhard, M.E. Spahr, P. Nov´ak, Adv. Mater. 10 (1998) 725. [2] E. Frackowiak, F. B´eguin, Carbon 39 (2001) 937. [3] J.H. Jang, S. Han, T. Hyeon, S.M. Oh, J. Power Sources 123 (2003) 79. [4] A. Nishino, J. Power Sources 60 (1996) 137. [5] W. Sugimoto, H. Iwata, Y. Yasunaga, Y. Murakami, Y. Takasu, Angew. Chem. Int. Ed. 42 (2003) 4092.
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141 [6] C.C. Hu, W.C. Chen, K.H. Chang, J. Electrochem. Soc. 151 (2004) A281. [7] K.-H. Chang, C.C. Hu, J. Electrochem. Soc. 151 (2004) A958. [8] Y. Chabre, J. Pannetier, Prog. Solid State Chem. 23 (1995) 1. [9] M.M. Thackeray, Prog. Solid State Chem. 25 (1997) 1. [10] M. Winter, J.O. Besenhard, M.E. Spahr, P. Nov´ak, Adv. Mater. 10 (1998) 725. [11] R. Ma, Y. Bando, L. Zhang, T. Sasaki, Adv. Mater. 16 (2004) 918. [12] C.C. Hu, T.W. Tsou, Electrochem. Commun. 4 (2002) 105. [13] S.C. Pang, M.A. Anderson, T.W. Chapman, J. Electrochem. Soc. 147 (2000) 444.
141
[14] T. Ohzuku, M. Kitagawa, T. Hirai, J. Electrochem. Soc. 137 (1990) 769. [15] T. Abe, S. Inoue, K. Watanabe, J. Alloys Compd. 358 (2003) 177. [16] G. Sandrock, K. Gross, G. Thomas, J. Alloys Compd. 229 (2002) 299. [17] C. Weidenthaler, A. Pommerin, M. Felderhoff, B. Bogdanovi´c, F. Sch¨uth, Phys. Chem. Chem. Phys. 5 (2003) 5149. [18] T. Wakamatsu, T. Fujiwara, K.N. Ishihara, P.H. Shingu, J. Jpn. Soc. Powder Powder Metall. 48 (2001) 950. [19] P.N. Kuznetsov, L.I. Kuznetsova, V.G. Chumakov, G.A. Moiseeva, Mater. Res. Innovat. 3 (2000) 340. [20] D.R. Lide, H.V. Kehiaian, CRC Handbook of Thermophysical and Thermochemical Data, CRC Press, Florida, 1994.
Phase transition and electrochemical capacitance of mechanically treated manganese oxides Akira Taguchi, Sachio Inoue, Satoshi Akamaru, Masanori Hara, Kuniaki Watanabe, Takayuki Abe ∗ Hydrogen Isotope Research Center, Toyama University, Gofuku 3190, Toyama 930-8555, Japan Received 25 January 2005; accepted 17 February 2005 Available online 1 September 2005
Abstract Mechanical grinding (MG) leads to a sequential phase transition of ␥-MnO2 to form the thermodynamically stable ␣-Mn2 O3 and, subsequently, Mn3 O4 depending on the duration of MG treatment. By MG treatment for 150 h, ␣-Mn2 O3 became a predominant species. A subsequent transition to Mn3 O4 was observed on further MG treatment for 200 h. The particle size of the resultant manganese oxides was reduced to as small as 20 nm after 300 h of MG treatment. The change in the specific capacitance of manganese oxides obtained was monitored by XRD and electrochemical experiments. It was observed that the specific capacitance of MG-treated manganese oxides depends on the amount of ␥-MnO2 left without phase transition. © 2005 Elsevier B.V. All rights reserved. Keywords: Manganese oxides; Phase transition; Mechanical grinding; Supercapacitor
1. Introduction Recently, the development of novel energy storage devices such as batteries and capacitors has been extensively studied in order to improve efficiency and reliability of electric utility systems such as solar-cell reservoirs or power quality regulators for computers, electric and hybrid vehicles, and fuel cell systems [1–4]. The oxide and hydroxide of ruthenium provide a good capacitance in aqueous and non-aqueous electrolyte systems [5,6]. However, the noble metal ruthenium, being expensive, has its own limitations. Hence, the fabrication of substitutional supercapacitive materials for ruthenium (hydro)oxides is gaining importance. Presently, several metal oxides and hydroxides, for example, those of Ni, Co, V, and Mn, are being studied extensively [8–10]. Among these, manganese oxide (MnO2 ) is one of the promising materials for this purpose due to its availability, cost-effectiveness and non-toxicity.
∗
Corresponding author. Tel.: +81 76 445 6933; fax: +81 76 445 6931. E-mail address: [email protected] (T. Abe).
0925-8388/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2005.02.108
The design and synthesis of manganese oxide electrodes is a key field for improving upon its electrochemical properties. In the case of super capacitor electrodes which comprise manganese oxide, it has been widely observed that MnO2 exhibits a rather high performance in comparison to its other oxides such as Mn(OH)2 , Mn2 O3 , and Mn3 O4 [11–13]. For example, amorphous hydrous manganese oxide (a-MnO2 ·nH2 O) fabricated from MnSO4 ·5H2 O solution via an anodic deposition yielded a specific capacitance in the range of 265–320 F/g between 0 to 1.0 V in 0.1 M Na2 SO4 solution [12]. A thin film of MnO2 electrode prepared by the dip-coating of colloidal MnO2 exhibits an excellent specific capacitance value of unto 698 F/g that is comparable to the ruthenium (hydro)oxides [5,7,13]. However, a structural change in the active component of MnO2 causes a loss of the electrochemical activity. This is due to of the significant change in the conductivity of manganese oxides—the conductivity of manganese oxides involving trivalent Mn(III) is rather low as compared to that of MnO2 or amorphous manganese (oxyhydro)oxides involving Mn(IV); this is attributed to the Jahn–Teller distortion of the MnIII O6 octahedral [8,14]. A few reports pertaining to
138
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141
the quantitative investigations of the content and capacitance of a mixture of MnO2 and Mn2 O3 have been published. Much attention has been paid to a mechanical grinding (MG) technique by means of ball milling, which is a powerful tool to reduce the particle size or mix several components to homogeneous dispersion [15–17]. Besides these fine mixing, a phase transition by MG treatment has been reported by Wakamatsu et al. [18] and Kuznetsov et al. [19]. Wakamatsu et al. have shown that rutil and/or brookite were obtained by MG treatment of anatase [18]. The mechanical treatment of -FeOOH caused the phase transition to FeO, via ␣-Fe2 O3 and Fe3 O4 as intermediate phases [19]. Depending on the duration of MG treatment, we observed a sequential phase transition of manganese oxides of ␥-MnO2 to ␣-Mn2 O3 and then subsequently to Mn3 O4 . The resultant manganese oxides are suitable examples to investigate the relationship between the composition of manganese oxides and their capacitances. In this paper, we report a sequential phase transition of manganese oxides by MG treatment, and the change in the specific capacitances of the resultant manganese oxides.
2. Experimental Commercially available MnO2 (Kanto Chemical Co.) was mechanically ground using a planet type ball mill (Fritsch, pulverized 5). In a container (26 mm in diameter and 27 mm in height) with 30 balls (each 4 mm in diameter), 4 g of MnO2 was placed and ground at 800 rpm for sufficient periods of time. Hereafter, these MG-treated samples are denoted as MG-n; where n is the period (in h). The structure and crystalline of the obtained samples were investigated by a powder X-ray diffraction (XRD, Philips, PW1825/00) and transmission electron microscopy (TEM, TOPCON, EM-002B). The manganese oxides prepared by the MG treatment were electrochemically investigated. Manganese oxide that had been treated for a specific time was mixed with a small amount of polytetrafluoroethylene (PTFE), which was used as a binder, and carbon powder (10 wt.%). A few drops of nmethylpyrrolidine were added to the mixture to prepare the sample paste. A Pt plate (10 mm × 10 mm), used as a current collector, was immersed in concentrated nitric acid for 2 h and dried in an oven at 80 ◦ C; subsequently, it was washed with distilled water. The sample paste was spread on the Pt plate, and it was dried at 60 ◦ C for 3 h. The amount of manganese oxides on the composite plate was measured by subtracting the weight of the Pt plate from that of the plate with the sample. In the case of the manganese oxides containing a mixture of ␥-MnO2 , ␣-Mn2 O3 , and Mn3 O4 , the composition of each component was calculated from the relative peak intensity for the most intense peak of each. Cyclic voltammetry (CV) was carried out in 1 M KOH aqueous solution using a threeelectrode system with a Pt counter electrode and a saturated calomel electrode (SCE) under potential control conditions (EG&G, potentiostat Model 263A).
Fig. 1. X-ray diffraction patterns of manganese oxides of (a) untreated (MG0), (b) MG-80 (an offset by 200 cps.), (c) MG-120 (an offset by 400 cps.) and (d) MG-300 (an offset by 800 cps.). Symbols (䊉), () and () correspond to the peaks of ␥-MnO2 , ␣-Mn2 O3 , and Mn3 O4 , respectively.
3. Results The manganese oxide used in this study was ␥-MnO2 (nsutite, JCPDS 4-779 and 17-510), and its XRD pattern is shown in Fig. 1(a). Its particle size could not be estimated due to the very low diffraction intensity and the poor resolution of diffraction signals. The poor crystallinity is probably due to the presence of a small structural defect in ␥-MnO2 [8]. The phase composition and crystalline structure of ␥-MnO2 changed significantly after the MG treatment. As shown by the XRD pattern in Fig. 1(b), new diffraction signals assignable to ␣-Mn2 O3 appeared at 2θ of 32.9◦ , 42.8◦ and 43.6◦ after the MG treatment for 80 h (MG-80). These peaks could be indexed to (2 2 2), (3 2 0) and (3 3 2), respectively; subsequent grinding for 120 h (MG-120) supported this observation. The diffraction signal of the ␣-Mn2 O3 phase (bixbyite, JCPDS 31-825 and 41-1442) intensified and those of ␥-MnO2 almost completely disappeared during the duration of this MG treatment. The particle sizes of the obtained ␣-Mn2 O3 were calculated with the diffraction signals of (2 2 2), (4 4 0), and (6 2 2), and were found to be 16.0, 13.4, and 15.6 nm, respectively, using the Scherrer’s equation. Furthermore, on increasing the duration of MG treatment up to 300 h (MG-300), some diffraction signals corresponding to Mn3 O4 (hausmannite, JCPDS 1-1127) appeared along with ␣-Mn2 O3 signals. Fig. 2 shows a change in the intensities of the diffraction signals of ␥-MnO2 , ␣-Mn2 O3 , and Mn3 O4 , depending on
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141
139
Fig. 2. Change in the X-ray diffraction intensities depending on the duration of MG treatment. Symbols (䊉), () and () correspond to the peaks of (0 2 1) in ␥-MnO2 , (2 2 2) in ␣-Mn2 O3 , and (1 0 1) in Mn3 O4 , respectively.
the duration of MG treatment. In this case, the most intense diffraction signals of each compound at 2θ = 37.5◦ (indexed to (0 2 1) in ␥-MnO2 ), 33.9◦ ((2 2 2) in ␣-Mn2 O3 ) and 18.0◦ ((1 0 1) in Mn3 O4 ) were employed. The signal intensity of the starting material of ␥-MnO2 remained almost constant for the MG treatment duration of 60 h, and then decreased slightly after the MG treatment for 80 h. The diffraction signals of ␣-Mn2 O3 were observed in 80 h, and its signal intensity increased significantly up to 150 h. At this point, ␣-Mn2 O3 became a dominate species. The peak intensity of ␣-Mn2 O3 decreased on further MG treatment, and a new signal corresponding to Mn3 O4 was observed in the MG-200 sample. The peak intensity of Mn3 O4 was increased simultaneously and gradually as that of ␣-Mn2 O3 decreased. The standard enthalpy of formation (f H◦ ) of MnO2 , Mn2 O3 , and Mn3 O4 were −520.0, −959.0 and −1387.8 kJ/mol, respectively [20]. Therefore, it was assumed that the sequential phase transition toward a thermodynamically stable composition occurred due to mechanical grinding. The phase transition toward the thermodynamically stable compounds has been widely observed in several metal oxides, during the MG treatment. For example, Wakamatsu et al. have shown a phase transition of TiO2 from anatase to brookite and/or rutile by ball milling [18]. They revealed that the longer or stronger the milling conditions for the anatase powder the shorter is the duration of structural change. The chemical reduction reaction of -FeOOH to form stable FeO, via ␣-Fe2 O3 and Fe3 O4 as intermediate phases, has also been reported by Kuznetsov et al. [19]. These facts support the possible phase transition of manganese oxides by MG treatment. The mechanism for this reduction reaction is not clear. One possible reason is that the metal components of the balls and container in the ball mill instruments participate in this reduction reaction. The reduction in particle sizes of manganese oxides by MG treatment was investigated by a TEM. Fig. 3 shows the TEM photographs of manganese oxides before and after MG
Fig. 3. TEM images of (a) untreated (MG-0) and (b) MG-300.
treatment. An amorphous bulky particle with a particle size of several ten nm is observed in the sample before MG treatment (MG-0). On the other hand, flake particles with a particle size of less than 20 nm were observed in the MG-300 sample. For example, the dimensions of the particle highlighted in the open circle in Fig. 3(b) were estimated to be 10 and 20 nm in shorter and longer diameters, respectively. These values are consistent with the ones calculated from XRD signals using the Scherrer’s equation (13.4–16.0 nm). Significant reduction in particle size by MG treatment has been widely reported. In our previous report, it was revealed that the particle size of RuO2 could be successfully reduced from 17.0–15.0 nm by MG treatment [15]. These values are fairly consistent with those observed in the present study, suggesting a limitation of particle downsizing by our instrument. Manganese oxides are promising candidates for new capacitive materials as alternatives for those based on ruthenium. Fig. 4 shows the cyclic voltammograms of the composite electrodes of MG-treated manganese oxides and that of the bare Pt plate (used as a current collector) in a 1 M KOH solution. The scan rate of the potential was 20 mV/s. The CV obtained in the MG-0 modified Pt plate with a load-
140
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141
Fig. 4. Cyclic voltammograms of manganese oxides-modified electrode in a 1 M KOH solution. Bold solid line: MG-0; fine solid line: MG-60; broken line: MG-120; dotted line: bare Pt; scan rate: 20 mV/s.
ing amount of 0.36 mg had a rectangular shape, showing the typical behavior of a capacitor. The specific capacitance of MG-0 was calculated from the difference between CVs at the MG-0 modified electrode and at the Pt plate alone, and was found to be 73.7 F/g. The characteristic rectangular CV shape has been maintained in the sample that was MG-treated up to 60 h; the capacitance of MG-60 was estimated to be 64.6 F/g (0.39 mg loading). These curves show no peaks, indicating that the electrode capacitor is charged and discharged at a constant rate throughout the entire voltammetric cycle. However, the values of observed specific capacitances were far smaller than those reported, e.g., 265–320 F/g for amorphous MnO2 -deposited electrode (scan rate of 25 mV/s in 0.1 M Na2 SO4 ) [12]. By further MG treatment, the CV curve of the modified electrode exhibited considerable changes and distortion (Fig. 4). The capacitance of MG-120, corresponding to ␣-Mn2 O3 on the basis of the XRD measurement, was significantly decreased to 18.9 F/g (0.37 mg loading). This value is less than 30% of the initial capacitance observed in the starting material of ␥-MnO2 . In addition, the anodizing current demonstrated by the CV curve is slightly larger than the cathodizing one; consequently, the reversibility in terms of the ratio of anodic and cathodic current has been changed. The structural change of manganese oxides electrode causes a loss of electrochemical activities. In this study, specific capacitances of modified electrodes were examined in various manganese oxides at different durations of MG treatment. Fig. 5 shows a relationship between the specific capacitance and XRD signal intensity of ␥-MnO2 used in this study. In order to estimate the composition of ␥-MnO2 and ␣-Mn2 O3 in the mixture, their relative amounts were calculated from the diffraction peak intensity (I(hkl) ). In this case, the relative intensities were calculated for the most intense signals of MG-0 in ␥-MnO2 and MG-150 in ␣-Mn2 O3 by the following equations. C␥-MnO2 = C␣-Mn2 O3 =
I(0 2 1) I(0 2 1)
in MG-n in MG-0
I(2 2 2) in MG-n I(2 2 2) in MG-150
Fig. 5. Relationship between the specific capacitance of manganese oxidesmodified electrodes and XRD signal intensity of the (0 2 1) peak of ␥-MnO2 . The error bar represents the difference in specific capacitance that arises from an estimation of the composition.
Under ideal conditions, the sum of the fraction of compositions of C␥-MnO2 and C␣-Mn2 O3 should be 1.0; however, it was lower than this value. The difference, which mainly arose from an imprecise definition of equations, is represented as an error bar in Fig. 5. In this figure, a linear decrease in the specific capacitance with a decrease in the diffraction signal intensity of ␥-MnO2 is clearly visible. In other words, the specific capacitance decreases linearly with a decrease in the amount of ␥-MnO2 . Thus, it can be concluded that the content of ␣-Mn2 O3 is directly related to the capacitance of the ␥-MnO2 electrode.
4. Conclusion In conclusion, we have reported the phase transition of manganese oxides depending on the duration of the MG treatment. The starting material ␥-MnO2 began to change to ␣-Mn2 O3 in 80 h, and subsequently transformed into Mn3 O4 by further milling for 200 h. The phase transition sequence of ␥-MnO2 , ␣-Mn2 O3 and Mn3 O4 follows a thermodynamic stability pattern of manganese oxides as predicted by the standard enthalpy of formation. Due to the formation of MnIII O6 octahedral, which has a rather low conductivity in comparison to that of MnIV , the capacitances of manganese oxides decreased linearly with a decrease in the crystallinity of ␥MnO2 .
References [1] M. Winter, J.O. Besenhard, M.E. Spahr, P. Nov´ak, Adv. Mater. 10 (1998) 725. [2] E. Frackowiak, F. B´eguin, Carbon 39 (2001) 937. [3] J.H. Jang, S. Han, T. Hyeon, S.M. Oh, J. Power Sources 123 (2003) 79. [4] A. Nishino, J. Power Sources 60 (1996) 137. [5] W. Sugimoto, H. Iwata, Y. Yasunaga, Y. Murakami, Y. Takasu, Angew. Chem. Int. Ed. 42 (2003) 4092.
A. Taguchi et al. / Journal of Alloys and Compounds 414 (2006) 137–141 [6] C.C. Hu, W.C. Chen, K.H. Chang, J. Electrochem. Soc. 151 (2004) A281. [7] K.-H. Chang, C.C. Hu, J. Electrochem. Soc. 151 (2004) A958. [8] Y. Chabre, J. Pannetier, Prog. Solid State Chem. 23 (1995) 1. [9] M.M. Thackeray, Prog. Solid State Chem. 25 (1997) 1. [10] M. Winter, J.O. Besenhard, M.E. Spahr, P. Nov´ak, Adv. Mater. 10 (1998) 725. [11] R. Ma, Y. Bando, L. Zhang, T. Sasaki, Adv. Mater. 16 (2004) 918. [12] C.C. Hu, T.W. Tsou, Electrochem. Commun. 4 (2002) 105. [13] S.C. Pang, M.A. Anderson, T.W. Chapman, J. Electrochem. Soc. 147 (2000) 444.
141
[14] T. Ohzuku, M. Kitagawa, T. Hirai, J. Electrochem. Soc. 137 (1990) 769. [15] T. Abe, S. Inoue, K. Watanabe, J. Alloys Compd. 358 (2003) 177. [16] G. Sandrock, K. Gross, G. Thomas, J. Alloys Compd. 229 (2002) 299. [17] C. Weidenthaler, A. Pommerin, M. Felderhoff, B. Bogdanovi´c, F. Sch¨uth, Phys. Chem. Chem. Phys. 5 (2003) 5149. [18] T. Wakamatsu, T. Fujiwara, K.N. Ishihara, P.H. Shingu, J. Jpn. Soc. Powder Powder Metall. 48 (2001) 950. [19] P.N. Kuznetsov, L.I. Kuznetsova, V.G. Chumakov, G.A. Moiseeva, Mater. Res. Innovat. 3 (2000) 340. [20] D.R. Lide, H.V. Kehiaian, CRC Handbook of Thermophysical and Thermochemical Data, CRC Press, Florida, 1994.