- Email: [email protected]
Nonlinear phenomena in the electrochemical oxidation of sulfide
Electrochemistry Communications 7 (2005) 1471–1476 www.elsevier.com/locate/elecom
Nonlinear phenomena in the electrochemical oxidation of sulfide Jiamin Feng a, Qingyu Gao a
a,*
, Liangqin Xu a, Jichang Wang
a,b,*
College of Chemical Engineering, China University of Mining and Technology, Xuzhou, Jiangsu Province 221008, PR China b Department of Chemistry and Biochemistry, University of Windsor, 401 Sunset Ave., Windsor Ont., Canada N9P 3P4 Received 6 September 2005; received in revised form 6 October 2005; accepted 7 October 2005 Available online 8 November 2005
Abstract The electro-oxidation of sulfide on a Pt electrode is found to exhibit both N-NDR (negative differential resistance) and HN-NDR (hidden negative differential resistance) types of oscillations, making it the first system known to be capable of supporting both types of nonlinear instabilities without changing electrolyte compositions. Six distinct oscillation windows are observed when the external current is adjusted as the control parameter. In addition, temperature also exhibits subtle influences on the observed reaction behavior, where varying the temperature from 30.0 °C results in more complicated oscillatory phenomena. Ó 2005 Elsevier B.V. All rights reserved. Keywords: Electrochemical oscillations; Nonlinear instabilities; Sulfide; Platinum electrodes
1. Introduction Understanding the onset of non-equilibrium phenomena in electrochemical reactions has far reaching impacts [1–5]. The observation of Turing-type patterns on electrode surfaces [5], for example, points to a prospective exploitation of pattern-forming mechanism to manufacture structured electrodes similar of those being developed for the (bio)sensors, where wavelength of the pattern could be conveniently manipulated by changing reaction conditions such as temperature, concentrations of electrolytes and values of the applied potential/currents. Based on the electrochemical impedance spectra (EIS), Strasser and co-workers [6] and Krischer [7] have classified electrochemical oscillations into four categories. According to their methods, the appearance of an N-shaped potentiostatic curve indicates that the studied system belongs to class-III, NNDR (negative differential resistance), or class-IV, HNNDR (hidden negative differential resistance) oscillators. *
Corresponding authors. Tel./fax: +86 516 399 5758 (Q. Gao); Tel.: +1 519 253 3000x3540; fax: +1 519 973 7098. (J. Wang). E-mail addresses: [email protected] (Q. Gao), [email protected] (J. Wang). 1388-2481/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.elecom.2005.10.004
Due to their importance in both industrial productions and scientific researches, oxidations of sulfur-containing species have garnered a great deal of interests in the last two decades [8–21]. The oxidation could take place by reacting with various oxidants such as chlorite, bromate and hydrogen peroxide or via an electrochemical method. Nonlinear phenomena including both simple and complex oscillations and bistabilities have been seen in the studies of electrochemical oxidations of sulfur compounds [8–10]. For example, Chen and Miller [8] recently reported potential oscillations in the electrocatalytic oxidation of sulfide on a microstructured Ti/Ta2O5–IrO2 electrode. In this study, we report that the electro-oxidation of sodium sulfide on a platinum electrode is capable of exhibiting both N-NDR and HN-NDR types of oscillations. As shown in the following, various new phenomena including six distinct oscillation windows are observed. 2. Experimental procedure The working electrode is a polycrystalline platinum disk electrode with a diameter of 2.0 mm (CH Instruments INC., USA) and the counter electrode is a Pt wire. A HgjHg2SO4jK2SO4 electrode is used as the reference
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476 16
a
12
j / mA cm-2
electrode. The working electrode is polished with fine alumina powder (0.05 lm) before its usage and is then activated in 1.0 M H2SO4 solution by cyclically varying the applied potential between 0.2 and 1.2 V (vs. SCE) at a rate of 1.4 V/s for 30 min. The Pt electrode is rinsed repeatedly with de-ionized Milli-Q water (Millipore) after the activation process. All experiments are carried out with a CHI-660A Electrochemical Station (CH Instruments), except that the electrochemical impedance spectroscopy (EIS) is measured with a IM6e Electrochemical Workstation (Zahner-elektrik Gmbh Co. KG). Reactions take place in a single cell reactor thermostated through a circulating water bath (±0.1 °C, Polyscience Instrument). The volume of the reactor is 40.0 ml. The cylindrical reactor has an internal diameter of 5.5 cm. The three electrodes form an equilateral triangle with a length of 1.8 cm. The resistance of solution in the electrochemical cell is about 25 X. Experiments performed with deaerated electrolyte solution produced the same results as reported in this study. All reagents used in this study are analytical grade. Na2S and H2SO4 solutions are prepared with de-ionized water. The concentration of H2SO4 solution is determined by titrating against a standard Na2CO3 solution.
II 8 4
I 0 -0.8
0.0
0.8
1.6
160
b 120
j / mA cm-2
1472
VI V
80
IV
40
III I & II 0 -0.6
0.0
0.6
1.2
1.8
500
2.4
c
Fig. 1(a) presents the linear galvanic voltammogram of 1.0 M Na2S solution, in which the scanning rate equals 0.01 lA/s. The slow variation in current functions as a bifurcation control parameter, leading the system through an oscillatory window. This voltammogram illustrates that as the applied current is increased from 0, the measured potential increases almost linearly until the current density reaches 2.1 mA/cm2, where large magnitude fluctuations in the potential are detected. When the applied current density is increased to above 4.7 mA/cm2, oscillations in the potential disappear. Notably, potential of the sulfide system stays constant at about 0.4 V after moving out of the oscillatory region, despite the current is continuously increased. The second oscillation window appears when the applied current density becomes larger than 7.0 mA/cm2 and ends at 8.0 mA/cm2. From there, further increasing the current density does not induce visible fluctuations in the potential. Fig. 1(b) presents the linear galvanic voltammogram in an extended current range. In addition to the regions I and II seen in Fig. 1(a), four more isolated windows with distinct dynamic properties are observed. For example, although there is no potential fluctuations seen in region III, transient oscillations could still be observed under galvanostatic conditions (see Fig. 2(c)). There is an abrupt jump in the potential when the current density is increased to 60 mA/cm2. Beyond that, the sulfide system stays at the high potential state as the current is increased still. There are three distinct oscillatory regions within this current
j / mA cm-2
400
3. Results and discussion
300 200 100 0 -0.6
0.0
0.6
1.2
1.8
Potential / V Fig. 1. (a) Linear galvanic voltammogram collected at a scan rate of 0.01 lA/s; (b) linear galvanic voltammogram collected at a scan rate of 0.02 lA/s; and (c) linear voltammogram obtained under the scan rate of 0.05 mV/s. Temperature is maintained at 20.0 ± 0.1 °C.
range. It is noteworthy that properties of the galvanostatic oscillations in these six regions are significantly different from each other. Examples of these oscillations are presented in Fig. 2. In Fig. 1(c) the potential is scanned linearly from 0.6 V at a rate equal to 0.05 mV/s. There is no reaction activity until the potential is larger than 0.4 V. The current is then found to increase linearly until a plateau is reached. The increase in the anodic current results from the oxidation of sulfide ions. Remarkably, oscillations in current appear when the applied voltage becomes larger than 0.4 V. The result that oscillations take place around a plateau suggests that mass transportation near the Pt electrode is a limiting factor [7]. The oscillatory phenomenon disappears as the potential is increased to 0.7 V. However, as the potential reaches 1.5 V, oscillatory behavior appears again, exhibiting
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
Potential / V
1.6
a
0.8 0.0 -0.8 0
400
800
1200
Potential / V
1.6
b
0.8 0.0 -0.8
0
50
100
150
Potential / V
200
c
1.8 1.2 0.6 0.0 -0.6
0
20
40
60
80
the second oscillatory window. Two broad peaks can be recognized for the potential between 0.7 and 1.5 V, implying that sulfide undergoes consecutive oxidation reactions. Our cyclic voltammetry experiments further illustrate that no oxygen evolution takes place for the potentials below 0.6 V, suggesting that oscillations could exist in the sulfide system even in the absence of convections induced by oxygen evolution. Despite that the slow scanning rate used in this study significantly prolonged the time period required to complete the CV voltammogram, the amount of sulfide ions consumed during such an extended process is actually small. By assuming that HS/S2 are oxidized to sulfur elements, our calculation shows that less than 3.5% of HS/S2 ions in the 1 M NaS solution are consumed after 55 h. Therefore, the observed transient behavior could be caused by other factors. Fig. 2 presents six typical time series of the electrooxidation of Na2S under the galvanostatic conditions marked in Fig. 1(b): (a) 2.86 mA/cm2, (b) 7.27 mA/cm2, (c) 31.83 mA/cm2, (d) 63.66 mA/cm2, (e) 95.49 mA/cm2, and (f) 127.32 mA/cm2. Same as in Fig. 1, here the concentration of sodium sulfide is 1.0 M and the temperature is kept at 20.0 ± 0.1 °C. Bursting phenomenon occurs within the low current range outlined by region I in Fig. 1(a). An 200
d
1.8 1.2
j / mA cm-2
Potential / V
1600
0.6 0.0 -0.6 0
150 100 50
60
90
120 0
1.6 0.8
750
1500
2250
500
750
3000
b
100 50 0 500
1000
1500
f j / mA cm-2
1.5 0
4
8
12
16
Time / s Fig. 2. Time series of galvanostatic potential oscillations during the electrochemical oxidation of sulfide. The applied currents are (a) 2.86 mA/cm2, (b) 7.27 mA/cm2, (c) 31.83 mA/cm2, (d) 63.66 mA/cm2, (e) 95.49 mA/cm2 and (f) 127.32 mA/cm2. Other reaction conditions are [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C.
2000
c
60
1.7 1.6
1000
150
0
1.8
250
200
j / mA cm-2
Potential / V
e
0.0
Potential / V
a
0
30
2.4
0
1473
40 20 0 0
1000
2000
3000
4000
Time / s
Fig. 3. Time series of anodic current oscillations during the electrochemical oxidation of Na2S. The applied potential values are: (a) 0.50 V, (b) 0.525 V and (c) 0.55 V. Other reaction conditions are [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C.
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
1.0
a
0.5
0.0
0
100
200
300
400
1.0
4
b
0.5 0.0 -0.5 0
0
250 KHz
50
100
150
1.0
c
-4
25 mHz -8 -4
-2
0
2
0.6
4
0.5 0.0 -0.5
b
0
0.4
50
100
150
0.8
d
0.2
Potential / V
- Im(ZA) / KΩ
a
-0.5
Potential / V
- Im(ZA) / KΩ
8
oxidation of sulfide and could be potentially used to manipulate the reaction. Fig. 3 presents three time series under static potentials: (a) 0.50 V, (b) 0.525 V, and (c) 0.550 V. Other reaction conditions are the same as these used in the galvanostatic studies, e.g., [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C. In Fig. 3(a), simple oscillations of one-peak-per-period are observed, in which the period of oscillation is about 100 s. When the applied potential is increased, modulations in the oscillation amplitude take place, making the behavior resemble quasi-periodic oscillations. Further increase of the potential leads to irregular oscillations (Fig. 3(c)). It suggests that a sequence of quasi-periodic bifurcations is responsible for the occurrence of chemical chaos. The achievement of an N-shaped potentiostatic curve in Fig. 1(c) indicates that the sulfide system belongs to either
Potential / V
example of typical simple oscillations seen in region II of Fig. 1(a) is presented in Fig. 2(b). Notably, in both Figs. 2(a) and (b), the system appears to stay at a low potential after each excursion. Such a ‘‘quiescent’’ time period could be due to the desired replenishment of fresh chemicals such as S2 and HS onto the electrode surface. We have tested the above hypothesis by stirring the solution after the system reaches the low potential phase. The mechanic disturbance does accelerate the occurrence of a large amplitude excursion; however, if the mixing is too violent, the system may take hours to return to the oscillatory state. Figs. 2(c) and (d) illustrate that only transient oscillations could be achieved within the parameter ranges III and IV. There is an essential difference between the two oscillatory phenomena: Oscillations in Fig. 2(c) end up at a low potential state whereas the reaction in Fig. 2(d) evolves to a high potential state at the end of transient oscillations. It is interesting to point out that the transient behavior undertakes simple, complex and then simple oscillations, resembles the phenomenon frequently seen in a closed homogeneous chemical oscillator. Fig. 2(e) shows that low frequency oscillations exist within the parameter range V. As opposed to oscillations in region II, the system in Fig. 2(e) spends extended time at the high potential state. When the current density is increased to 120 mA/cm2, oscillations of high frequency and small magnitude are observed (see Fig. 2(f)). The above result demonstrates that the external current has dramatic influences on the electro-
Potential / V
1474
250 KHz 0.0
25 mHz -0.2 -0.4
0.0
0.4
0.8
Re( ZA ) / KΩ Fig. 4. Electrochemical impedance spectrum measured at different potential values: (a) 0.52 V and (b) 1.7 V. The frequency was varied from 250 kHz to 25 mHz. Other reaction conditions were [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C.
0.4 0.0 -0.4 0
20
40
60
80
100
120
Time / s
Fig. 5. Influence of reaction temperature on the electrochemical oscillations of sulfide: (a) 28.0 °C, (b) 30.0 °C, (c) 32.5 °C and (d) 40.0 °C. The applied anode current equals 3.82 mA/cm2, and [Na2S] = 1.0 M.
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
N-NDR or HN-NDR oscillators, in which the electrical double layer potential is an autocatalytic variable leading to the negative differential resistance [6,7]. To gain further insight into the nature of the sulfide oscillator, two electrochemical impedance spectra are collected, respectively, within the two oscillation windows shown in Fig. 1(c). The frequency is decreased from 250 kHz to 25.0 mHz. The impedance spectrum measured at 0.52 V has a negative zero-frequency impedance (see Fig. 4a), implying that the system belongs to the N-NDR oscillator. In contrast, a positive zero-frequency impedance is obtained at 1.7 V (see Fig. 4b), suggesting that oscillations occurred here are the HN-NDR type of oscillations. Remarkably, without changing the electrolyte, the sulfide system could be either a N-NDR or a HN-NDR oscillator. We are not aware any other electrochemical reactions have this unique property. Influences of temperature on the galvanostatic oscillations are presented in Fig. 5, in which j = 3.82 mA/cm2 and [Na2S] = 1.0 M. Clearly, the complexity of these spontaneous oscillations does not vary monotonically with the reaction temperature. Increasing the temperature from 28.0 to 30.0 °C results in a transition from burst-like complex oscillations to simple oscillations of one-peak-per-period. Further increasing the temperature, on the other hand, leads to complex oscillations again. For example, at 32.5 °C the oscillation becomes period-doubled 11 modes. Perioddoubled oscillations are obtained when the temperature is adjusted to 40.0 °C. Under the reaction conditions studied here, oscillatory phenomena completely disappear once the reaction temperature is increased to 45.0 °C or higher. 4. Conclusions The electro-oxidation of sulfide on a Pt electrode is found to exhibit extremely rich nonlinear phenomena including quasi-periodic and period-doubling bifurcations. Both galvanostatic and potentiostatic oscillations are observed in this simple system. Remarkably, when current is adjusted as the control parameter, the system evolves through six distinct oscillation windows in which the oscillatory behavior exhibits different dynamic properties. In addition, temperature also exhibits subtle influences on the reaction behavior. Our study further demonstrates that oscillations in the sulfide system can appear even in the absence of convections induced by gas evolutions. The electrochemical impedance spectra indicate that the sulfide system can be either a N-NDR or a HN-NDR oscillator, where the transition depends only on the applied potential. Both N-NDR and HN-NDR types of oscillations are observed earlier in the reduction of H2O2 on a Pt electrode, but the addition of halides is required to induce the transition from N-NDR to HN-NDR oscillations [7,22]. To our knowledge, this is the first electrochemical reaction system exhibiting two types of oscillations without changing electrolytes. For the NDR type oscillators, the double layer potential is a fast auto-catalytic system variable [6,7]. The underlying chemical cause for the negative regulation in this sulfide
1475
system, which may arise from mass-transfer or adsorption of S2 and HS on the Pt, will be explored in the future. As discussed in literature [8,10,23–25], the following anodic reactions are expected to take place in the sulfide system, resulting in the formation and removal of sulfur layer: HS þ OH ! S þ H2 O þ 2e 2
S
ð1Þ
! S þ 2e
ð2Þ
ð3Þ
2HS þ 2OH ! S2 2 þ 2H2 O þ 2e HS þ 9OH ! SO2 4 þ 5H2 O þ 8e S þ HS þ OH ! S2 2 þ H2 O
Sþ
S2 2
!
ð4Þ ð5Þ
S2 3
ð6Þ
Our measurement shows that the pH of 1.0 M Na2S solution studied here is close to the pH of 1.0 M NaOH solution, indicating that HS ions are the predominate species. When the external potential is around 0.5 V, periodic formation and dissolution of sulfur elements on the Pt electrode are observed. The experimental phenomenon could be accounted by the following processes: Pt þ HS þ OH ! PtS þ H2 O þ 2e PtS þ
HS x
! Pt þ
ð7Þ
HS xþ1
PtS þ 8OH ! Pt þ 4H2 O þ
ð8Þ SO2 4
þ 6e
ð9Þ
Notably, process (7) causes the reduction in the size of effective surface of the Pt electrode. Such an effect may have been responsible for the development of NDR. As the potential is increased to 1.5 V, no sulfur precipitate is observed, implying that HS/S2 are oxidized to sulfate as shown in the following: Pt þ 2OH ! PtO þ H2 O þ 2e PtO þ 2OH þ H2 O ! PtðOHÞ4 þ 2e
ð10Þ ð11Þ
ðHS Þads þ 9OH ! SO2 4 þ 5H2 O þ 8e
ð12Þ
Although oxygen is produced in 1.0 M NaOH solution for the potential larger than 0.65 V, we did not see oxygen production in the 1.0 M Na2S solution when the potential was adjusted between 0.8 and 1.8 V. Such a result suggests that competitions in the absorption of OH and of HS/S2 could be the culprit of the NDR. The existence of such rich nonlinear phenomena under such broad reaction conditions makes sulfide system an attractive model system for the study of non-equilibrium behavior in electrochemical reactions. In addition, the deposition of sulfide on the electrode surface may offer a convenient avenue to study electrochemical pattern formation, a subject which has tremendous potentials in the applications in areas of catalysis, the fabrication of (bio)sensors, etc. Acknowledgments This work is supported through NSFC (20103010) and EYPT of China. J.W. would like to thank the financial support from NSERC.
1476
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
References [1] S. Fukushima, S. Nakanishi, K. Fukami, S. Sakai, T. Nagai, T. Tada, Electrochem. Commun. 7 (2005) 411. [2] Q.K. Yu, Y. Miyakita, S. Nakabayashi, R. Baba, Electrochem. Commun. 5 (2003) 321. [3] T.H.M. Housmans, M.T.M. Koper, Electrochem. Commun. 7 (2005) 581. [4] A. Birzu, F. Plenge, N.I. Jaeger, J.L. Hudson, K. Krischer, J. Phys. Chem. B 107 (2003) 5825. [5] Y. Li, J. Oslonovitch, N. Mazouz, F. Plenge, K. Krischer, G. Ertl, Science 291 (2001) 2396. [6] P. Strasser, M. Eiswirth, M.T.M. Koper, J. Electroanal. Chem. 478 (1999) 50. [7] K. Krischer, in: R.C. Alkire, D.M. Kolb (Eds.), Advances in Electrochemical Science and Engineering, vol. 8, Wiley-VCH, Weinheim, 2003, p. 89. [8] A. Chen, B. Miller, J. Phys. Chem. B 108 (2004) 2245. [9] L. Xu, Q. Gao, J. Feng, J. Wang, Chem. Phys. Lett. 397 (2004) 265. [10] H. Helms, E. Schlomer, W. Jansen, Monatsh. Chem. 129 (1998) 617.
[11] M. Alamgir, I.R. Epstein, Int. J. Chem. Kinet. 17 (1985) 429. [12] C.J. Doona, R. Blittersdorf, F.W. Schneider, J. Phys. Chem. 97 (1993) 7258. [13] C.R. Chinake, R.H. Simoyi, J. Phys. Chem. 98 (1994) 4012. [14] M. Orba´n, I.R. Epstein, J. Am. Chem. Soc. 107 (1985) 2302. [15] R.H. Simoyi, R.M. Noyes, J. Phys. Chem. 91 (1987) 2689. [16] G. Rabai, M. Orban, I.R. Epstein, J. Phys. Chem. 96 (1992) 5414. [17] I.R Epstein, K. Kustin, R.H. Simoyi, J. Phys. Chem. 96 (1992) 5852. [18] Q. Gao, J. Wang, Chem. Phys. Lett. 391 (2004) 349. [19] G. Rabai, I.R. Epstein, J. Am. Chem. Soc. 114 (1992) 1529. [20] M. Orba´n, I.R. Epstein, J. Am. Chem. Soc. 109 (1987) 101. [21] G. Ra´bai, M.T. Beck, K. Kustin, I.R. Epstein, J. Phys. Chem. 93 (1989) 2853. [22] Y. Mukouyama, S. Nakanishi, T. Chiba, K. Murakoshi, Y. Nakato, J. Phys. Chem. B 105 (2001) 7246. [23] J. Szynkarczuk, P.G. Komorowski, J.C. Donini, Electrochim. Acta 40 (1995) 487. [24] Y. Wang, H. Yan, E. Wang, J. Electroanal. Chem. 497 (2001) 163. [25] Y.E. Sung, W. Chrzanowski, A. Zolfaghari, G. Jerkiewicz, A. Wieckowski, J. Am. Chem. Soc. 119 (1997) 194.
Nonlinear phenomena in the electrochemical oxidation of sulfide Jiamin Feng a, Qingyu Gao a
a,*
, Liangqin Xu a, Jichang Wang
a,b,*
College of Chemical Engineering, China University of Mining and Technology, Xuzhou, Jiangsu Province 221008, PR China b Department of Chemistry and Biochemistry, University of Windsor, 401 Sunset Ave., Windsor Ont., Canada N9P 3P4 Received 6 September 2005; received in revised form 6 October 2005; accepted 7 October 2005 Available online 8 November 2005
Abstract The electro-oxidation of sulfide on a Pt electrode is found to exhibit both N-NDR (negative differential resistance) and HN-NDR (hidden negative differential resistance) types of oscillations, making it the first system known to be capable of supporting both types of nonlinear instabilities without changing electrolyte compositions. Six distinct oscillation windows are observed when the external current is adjusted as the control parameter. In addition, temperature also exhibits subtle influences on the observed reaction behavior, where varying the temperature from 30.0 °C results in more complicated oscillatory phenomena. Ó 2005 Elsevier B.V. All rights reserved. Keywords: Electrochemical oscillations; Nonlinear instabilities; Sulfide; Platinum electrodes
1. Introduction Understanding the onset of non-equilibrium phenomena in electrochemical reactions has far reaching impacts [1–5]. The observation of Turing-type patterns on electrode surfaces [5], for example, points to a prospective exploitation of pattern-forming mechanism to manufacture structured electrodes similar of those being developed for the (bio)sensors, where wavelength of the pattern could be conveniently manipulated by changing reaction conditions such as temperature, concentrations of electrolytes and values of the applied potential/currents. Based on the electrochemical impedance spectra (EIS), Strasser and co-workers [6] and Krischer [7] have classified electrochemical oscillations into four categories. According to their methods, the appearance of an N-shaped potentiostatic curve indicates that the studied system belongs to class-III, NNDR (negative differential resistance), or class-IV, HNNDR (hidden negative differential resistance) oscillators. *
Corresponding authors. Tel./fax: +86 516 399 5758 (Q. Gao); Tel.: +1 519 253 3000x3540; fax: +1 519 973 7098. (J. Wang). E-mail addresses: [email protected] (Q. Gao), [email protected] (J. Wang). 1388-2481/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.elecom.2005.10.004
Due to their importance in both industrial productions and scientific researches, oxidations of sulfur-containing species have garnered a great deal of interests in the last two decades [8–21]. The oxidation could take place by reacting with various oxidants such as chlorite, bromate and hydrogen peroxide or via an electrochemical method. Nonlinear phenomena including both simple and complex oscillations and bistabilities have been seen in the studies of electrochemical oxidations of sulfur compounds [8–10]. For example, Chen and Miller [8] recently reported potential oscillations in the electrocatalytic oxidation of sulfide on a microstructured Ti/Ta2O5–IrO2 electrode. In this study, we report that the electro-oxidation of sodium sulfide on a platinum electrode is capable of exhibiting both N-NDR and HN-NDR types of oscillations. As shown in the following, various new phenomena including six distinct oscillation windows are observed. 2. Experimental procedure The working electrode is a polycrystalline platinum disk electrode with a diameter of 2.0 mm (CH Instruments INC., USA) and the counter electrode is a Pt wire. A HgjHg2SO4jK2SO4 electrode is used as the reference
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476 16
a
12
j / mA cm-2
electrode. The working electrode is polished with fine alumina powder (0.05 lm) before its usage and is then activated in 1.0 M H2SO4 solution by cyclically varying the applied potential between 0.2 and 1.2 V (vs. SCE) at a rate of 1.4 V/s for 30 min. The Pt electrode is rinsed repeatedly with de-ionized Milli-Q water (Millipore) after the activation process. All experiments are carried out with a CHI-660A Electrochemical Station (CH Instruments), except that the electrochemical impedance spectroscopy (EIS) is measured with a IM6e Electrochemical Workstation (Zahner-elektrik Gmbh Co. KG). Reactions take place in a single cell reactor thermostated through a circulating water bath (±0.1 °C, Polyscience Instrument). The volume of the reactor is 40.0 ml. The cylindrical reactor has an internal diameter of 5.5 cm. The three electrodes form an equilateral triangle with a length of 1.8 cm. The resistance of solution in the electrochemical cell is about 25 X. Experiments performed with deaerated electrolyte solution produced the same results as reported in this study. All reagents used in this study are analytical grade. Na2S and H2SO4 solutions are prepared with de-ionized water. The concentration of H2SO4 solution is determined by titrating against a standard Na2CO3 solution.
II 8 4
I 0 -0.8
0.0
0.8
1.6
160
b 120
j / mA cm-2
1472
VI V
80
IV
40
III I & II 0 -0.6
0.0
0.6
1.2
1.8
500
2.4
c
Fig. 1(a) presents the linear galvanic voltammogram of 1.0 M Na2S solution, in which the scanning rate equals 0.01 lA/s. The slow variation in current functions as a bifurcation control parameter, leading the system through an oscillatory window. This voltammogram illustrates that as the applied current is increased from 0, the measured potential increases almost linearly until the current density reaches 2.1 mA/cm2, where large magnitude fluctuations in the potential are detected. When the applied current density is increased to above 4.7 mA/cm2, oscillations in the potential disappear. Notably, potential of the sulfide system stays constant at about 0.4 V after moving out of the oscillatory region, despite the current is continuously increased. The second oscillation window appears when the applied current density becomes larger than 7.0 mA/cm2 and ends at 8.0 mA/cm2. From there, further increasing the current density does not induce visible fluctuations in the potential. Fig. 1(b) presents the linear galvanic voltammogram in an extended current range. In addition to the regions I and II seen in Fig. 1(a), four more isolated windows with distinct dynamic properties are observed. For example, although there is no potential fluctuations seen in region III, transient oscillations could still be observed under galvanostatic conditions (see Fig. 2(c)). There is an abrupt jump in the potential when the current density is increased to 60 mA/cm2. Beyond that, the sulfide system stays at the high potential state as the current is increased still. There are three distinct oscillatory regions within this current
j / mA cm-2
400
3. Results and discussion
300 200 100 0 -0.6
0.0
0.6
1.2
1.8
Potential / V Fig. 1. (a) Linear galvanic voltammogram collected at a scan rate of 0.01 lA/s; (b) linear galvanic voltammogram collected at a scan rate of 0.02 lA/s; and (c) linear voltammogram obtained under the scan rate of 0.05 mV/s. Temperature is maintained at 20.0 ± 0.1 °C.
range. It is noteworthy that properties of the galvanostatic oscillations in these six regions are significantly different from each other. Examples of these oscillations are presented in Fig. 2. In Fig. 1(c) the potential is scanned linearly from 0.6 V at a rate equal to 0.05 mV/s. There is no reaction activity until the potential is larger than 0.4 V. The current is then found to increase linearly until a plateau is reached. The increase in the anodic current results from the oxidation of sulfide ions. Remarkably, oscillations in current appear when the applied voltage becomes larger than 0.4 V. The result that oscillations take place around a plateau suggests that mass transportation near the Pt electrode is a limiting factor [7]. The oscillatory phenomenon disappears as the potential is increased to 0.7 V. However, as the potential reaches 1.5 V, oscillatory behavior appears again, exhibiting
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
Potential / V
1.6
a
0.8 0.0 -0.8 0
400
800
1200
Potential / V
1.6
b
0.8 0.0 -0.8
0
50
100
150
Potential / V
200
c
1.8 1.2 0.6 0.0 -0.6
0
20
40
60
80
the second oscillatory window. Two broad peaks can be recognized for the potential between 0.7 and 1.5 V, implying that sulfide undergoes consecutive oxidation reactions. Our cyclic voltammetry experiments further illustrate that no oxygen evolution takes place for the potentials below 0.6 V, suggesting that oscillations could exist in the sulfide system even in the absence of convections induced by oxygen evolution. Despite that the slow scanning rate used in this study significantly prolonged the time period required to complete the CV voltammogram, the amount of sulfide ions consumed during such an extended process is actually small. By assuming that HS/S2 are oxidized to sulfur elements, our calculation shows that less than 3.5% of HS/S2 ions in the 1 M NaS solution are consumed after 55 h. Therefore, the observed transient behavior could be caused by other factors. Fig. 2 presents six typical time series of the electrooxidation of Na2S under the galvanostatic conditions marked in Fig. 1(b): (a) 2.86 mA/cm2, (b) 7.27 mA/cm2, (c) 31.83 mA/cm2, (d) 63.66 mA/cm2, (e) 95.49 mA/cm2, and (f) 127.32 mA/cm2. Same as in Fig. 1, here the concentration of sodium sulfide is 1.0 M and the temperature is kept at 20.0 ± 0.1 °C. Bursting phenomenon occurs within the low current range outlined by region I in Fig. 1(a). An 200
d
1.8 1.2
j / mA cm-2
Potential / V
1600
0.6 0.0 -0.6 0
150 100 50
60
90
120 0
1.6 0.8
750
1500
2250
500
750
3000
b
100 50 0 500
1000
1500
f j / mA cm-2
1.5 0
4
8
12
16
Time / s Fig. 2. Time series of galvanostatic potential oscillations during the electrochemical oxidation of sulfide. The applied currents are (a) 2.86 mA/cm2, (b) 7.27 mA/cm2, (c) 31.83 mA/cm2, (d) 63.66 mA/cm2, (e) 95.49 mA/cm2 and (f) 127.32 mA/cm2. Other reaction conditions are [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C.
2000
c
60
1.7 1.6
1000
150
0
1.8
250
200
j / mA cm-2
Potential / V
e
0.0
Potential / V
a
0
30
2.4
0
1473
40 20 0 0
1000
2000
3000
4000
Time / s
Fig. 3. Time series of anodic current oscillations during the electrochemical oxidation of Na2S. The applied potential values are: (a) 0.50 V, (b) 0.525 V and (c) 0.55 V. Other reaction conditions are [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C.
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
1.0
a
0.5
0.0
0
100
200
300
400
1.0
4
b
0.5 0.0 -0.5 0
0
250 KHz
50
100
150
1.0
c
-4
25 mHz -8 -4
-2
0
2
0.6
4
0.5 0.0 -0.5
b
0
0.4
50
100
150
0.8
d
0.2
Potential / V
- Im(ZA) / KΩ
a
-0.5
Potential / V
- Im(ZA) / KΩ
8
oxidation of sulfide and could be potentially used to manipulate the reaction. Fig. 3 presents three time series under static potentials: (a) 0.50 V, (b) 0.525 V, and (c) 0.550 V. Other reaction conditions are the same as these used in the galvanostatic studies, e.g., [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C. In Fig. 3(a), simple oscillations of one-peak-per-period are observed, in which the period of oscillation is about 100 s. When the applied potential is increased, modulations in the oscillation amplitude take place, making the behavior resemble quasi-periodic oscillations. Further increase of the potential leads to irregular oscillations (Fig. 3(c)). It suggests that a sequence of quasi-periodic bifurcations is responsible for the occurrence of chemical chaos. The achievement of an N-shaped potentiostatic curve in Fig. 1(c) indicates that the sulfide system belongs to either
Potential / V
example of typical simple oscillations seen in region II of Fig. 1(a) is presented in Fig. 2(b). Notably, in both Figs. 2(a) and (b), the system appears to stay at a low potential after each excursion. Such a ‘‘quiescent’’ time period could be due to the desired replenishment of fresh chemicals such as S2 and HS onto the electrode surface. We have tested the above hypothesis by stirring the solution after the system reaches the low potential phase. The mechanic disturbance does accelerate the occurrence of a large amplitude excursion; however, if the mixing is too violent, the system may take hours to return to the oscillatory state. Figs. 2(c) and (d) illustrate that only transient oscillations could be achieved within the parameter ranges III and IV. There is an essential difference between the two oscillatory phenomena: Oscillations in Fig. 2(c) end up at a low potential state whereas the reaction in Fig. 2(d) evolves to a high potential state at the end of transient oscillations. It is interesting to point out that the transient behavior undertakes simple, complex and then simple oscillations, resembles the phenomenon frequently seen in a closed homogeneous chemical oscillator. Fig. 2(e) shows that low frequency oscillations exist within the parameter range V. As opposed to oscillations in region II, the system in Fig. 2(e) spends extended time at the high potential state. When the current density is increased to 120 mA/cm2, oscillations of high frequency and small magnitude are observed (see Fig. 2(f)). The above result demonstrates that the external current has dramatic influences on the electro-
Potential / V
1474
250 KHz 0.0
25 mHz -0.2 -0.4
0.0
0.4
0.8
Re( ZA ) / KΩ Fig. 4. Electrochemical impedance spectrum measured at different potential values: (a) 0.52 V and (b) 1.7 V. The frequency was varied from 250 kHz to 25 mHz. Other reaction conditions were [Na2S] = 1.0 M and T = 20.0 ± 0.1 °C.
0.4 0.0 -0.4 0
20
40
60
80
100
120
Time / s
Fig. 5. Influence of reaction temperature on the electrochemical oscillations of sulfide: (a) 28.0 °C, (b) 30.0 °C, (c) 32.5 °C and (d) 40.0 °C. The applied anode current equals 3.82 mA/cm2, and [Na2S] = 1.0 M.
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
N-NDR or HN-NDR oscillators, in which the electrical double layer potential is an autocatalytic variable leading to the negative differential resistance [6,7]. To gain further insight into the nature of the sulfide oscillator, two electrochemical impedance spectra are collected, respectively, within the two oscillation windows shown in Fig. 1(c). The frequency is decreased from 250 kHz to 25.0 mHz. The impedance spectrum measured at 0.52 V has a negative zero-frequency impedance (see Fig. 4a), implying that the system belongs to the N-NDR oscillator. In contrast, a positive zero-frequency impedance is obtained at 1.7 V (see Fig. 4b), suggesting that oscillations occurred here are the HN-NDR type of oscillations. Remarkably, without changing the electrolyte, the sulfide system could be either a N-NDR or a HN-NDR oscillator. We are not aware any other electrochemical reactions have this unique property. Influences of temperature on the galvanostatic oscillations are presented in Fig. 5, in which j = 3.82 mA/cm2 and [Na2S] = 1.0 M. Clearly, the complexity of these spontaneous oscillations does not vary monotonically with the reaction temperature. Increasing the temperature from 28.0 to 30.0 °C results in a transition from burst-like complex oscillations to simple oscillations of one-peak-per-period. Further increasing the temperature, on the other hand, leads to complex oscillations again. For example, at 32.5 °C the oscillation becomes period-doubled 11 modes. Perioddoubled oscillations are obtained when the temperature is adjusted to 40.0 °C. Under the reaction conditions studied here, oscillatory phenomena completely disappear once the reaction temperature is increased to 45.0 °C or higher. 4. Conclusions The electro-oxidation of sulfide on a Pt electrode is found to exhibit extremely rich nonlinear phenomena including quasi-periodic and period-doubling bifurcations. Both galvanostatic and potentiostatic oscillations are observed in this simple system. Remarkably, when current is adjusted as the control parameter, the system evolves through six distinct oscillation windows in which the oscillatory behavior exhibits different dynamic properties. In addition, temperature also exhibits subtle influences on the reaction behavior. Our study further demonstrates that oscillations in the sulfide system can appear even in the absence of convections induced by gas evolutions. The electrochemical impedance spectra indicate that the sulfide system can be either a N-NDR or a HN-NDR oscillator, where the transition depends only on the applied potential. Both N-NDR and HN-NDR types of oscillations are observed earlier in the reduction of H2O2 on a Pt electrode, but the addition of halides is required to induce the transition from N-NDR to HN-NDR oscillations [7,22]. To our knowledge, this is the first electrochemical reaction system exhibiting two types of oscillations without changing electrolytes. For the NDR type oscillators, the double layer potential is a fast auto-catalytic system variable [6,7]. The underlying chemical cause for the negative regulation in this sulfide
1475
system, which may arise from mass-transfer or adsorption of S2 and HS on the Pt, will be explored in the future. As discussed in literature [8,10,23–25], the following anodic reactions are expected to take place in the sulfide system, resulting in the formation and removal of sulfur layer: HS þ OH ! S þ H2 O þ 2e 2
S
ð1Þ
! S þ 2e
ð2Þ
ð3Þ
2HS þ 2OH ! S2 2 þ 2H2 O þ 2e HS þ 9OH ! SO2 4 þ 5H2 O þ 8e S þ HS þ OH ! S2 2 þ H2 O
Sþ
S2 2
!
ð4Þ ð5Þ
S2 3
ð6Þ
Our measurement shows that the pH of 1.0 M Na2S solution studied here is close to the pH of 1.0 M NaOH solution, indicating that HS ions are the predominate species. When the external potential is around 0.5 V, periodic formation and dissolution of sulfur elements on the Pt electrode are observed. The experimental phenomenon could be accounted by the following processes: Pt þ HS þ OH ! PtS þ H2 O þ 2e PtS þ
HS x
! Pt þ
ð7Þ
HS xþ1
PtS þ 8OH ! Pt þ 4H2 O þ
ð8Þ SO2 4
þ 6e
ð9Þ
Notably, process (7) causes the reduction in the size of effective surface of the Pt electrode. Such an effect may have been responsible for the development of NDR. As the potential is increased to 1.5 V, no sulfur precipitate is observed, implying that HS/S2 are oxidized to sulfate as shown in the following: Pt þ 2OH ! PtO þ H2 O þ 2e PtO þ 2OH þ H2 O ! PtðOHÞ4 þ 2e
ð10Þ ð11Þ
ðHS Þads þ 9OH ! SO2 4 þ 5H2 O þ 8e
ð12Þ
Although oxygen is produced in 1.0 M NaOH solution for the potential larger than 0.65 V, we did not see oxygen production in the 1.0 M Na2S solution when the potential was adjusted between 0.8 and 1.8 V. Such a result suggests that competitions in the absorption of OH and of HS/S2 could be the culprit of the NDR. The existence of such rich nonlinear phenomena under such broad reaction conditions makes sulfide system an attractive model system for the study of non-equilibrium behavior in electrochemical reactions. In addition, the deposition of sulfide on the electrode surface may offer a convenient avenue to study electrochemical pattern formation, a subject which has tremendous potentials in the applications in areas of catalysis, the fabrication of (bio)sensors, etc. Acknowledgments This work is supported through NSFC (20103010) and EYPT of China. J.W. would like to thank the financial support from NSERC.
1476
J. Feng et al. / Electrochemistry Communications 7 (2005) 1471–1476
References [1] S. Fukushima, S. Nakanishi, K. Fukami, S. Sakai, T. Nagai, T. Tada, Electrochem. Commun. 7 (2005) 411. [2] Q.K. Yu, Y. Miyakita, S. Nakabayashi, R. Baba, Electrochem. Commun. 5 (2003) 321. [3] T.H.M. Housmans, M.T.M. Koper, Electrochem. Commun. 7 (2005) 581. [4] A. Birzu, F. Plenge, N.I. Jaeger, J.L. Hudson, K. Krischer, J. Phys. Chem. B 107 (2003) 5825. [5] Y. Li, J. Oslonovitch, N. Mazouz, F. Plenge, K. Krischer, G. Ertl, Science 291 (2001) 2396. [6] P. Strasser, M. Eiswirth, M.T.M. Koper, J. Electroanal. Chem. 478 (1999) 50. [7] K. Krischer, in: R.C. Alkire, D.M. Kolb (Eds.), Advances in Electrochemical Science and Engineering, vol. 8, Wiley-VCH, Weinheim, 2003, p. 89. [8] A. Chen, B. Miller, J. Phys. Chem. B 108 (2004) 2245. [9] L. Xu, Q. Gao, J. Feng, J. Wang, Chem. Phys. Lett. 397 (2004) 265. [10] H. Helms, E. Schlomer, W. Jansen, Monatsh. Chem. 129 (1998) 617.
[11] M. Alamgir, I.R. Epstein, Int. J. Chem. Kinet. 17 (1985) 429. [12] C.J. Doona, R. Blittersdorf, F.W. Schneider, J. Phys. Chem. 97 (1993) 7258. [13] C.R. Chinake, R.H. Simoyi, J. Phys. Chem. 98 (1994) 4012. [14] M. Orba´n, I.R. Epstein, J. Am. Chem. Soc. 107 (1985) 2302. [15] R.H. Simoyi, R.M. Noyes, J. Phys. Chem. 91 (1987) 2689. [16] G. Rabai, M. Orban, I.R. Epstein, J. Phys. Chem. 96 (1992) 5414. [17] I.R Epstein, K. Kustin, R.H. Simoyi, J. Phys. Chem. 96 (1992) 5852. [18] Q. Gao, J. Wang, Chem. Phys. Lett. 391 (2004) 349. [19] G. Rabai, I.R. Epstein, J. Am. Chem. Soc. 114 (1992) 1529. [20] M. Orba´n, I.R. Epstein, J. Am. Chem. Soc. 109 (1987) 101. [21] G. Ra´bai, M.T. Beck, K. Kustin, I.R. Epstein, J. Phys. Chem. 93 (1989) 2853. [22] Y. Mukouyama, S. Nakanishi, T. Chiba, K. Murakoshi, Y. Nakato, J. Phys. Chem. B 105 (2001) 7246. [23] J. Szynkarczuk, P.G. Komorowski, J.C. Donini, Electrochim. Acta 40 (1995) 487. [24] Y. Wang, H. Yan, E. Wang, J. Electroanal. Chem. 497 (2001) 163. [25] Y.E. Sung, W. Chrzanowski, A. Zolfaghari, G. Jerkiewicz, A. Wieckowski, J. Am. Chem. Soc. 119 (1997) 194.