- Email: [email protected]
thaw cycled polymer electrolyte fuel cells
Current Applied Physics 10 (2010) S62–S65
Contents lists available at ScienceDirect
Current Applied Physics journal homepage: www.elsevier.com/locate/cap
Analysis on the freeze/thaw cycled polymer electrolyte fuel cells Gu-Gon Park a,*, Soo-Jin Lim a, Jin-Soo Park b, Sung-Dae Yim a, Seok-Hee Park a, Tae-Hyun Yang a, Young-Gi Yoon a, Chang-Soo Kim a a b
Fuel Cell Research Center, Korea Institute of Energy Research, 102, Gajeong-ro, Yuseong-gu, Daejeon, Republic of Korea Department of Environmental Engineering, College of Engineering, Sanmyung University, 300 Anseo-dong, Dongnam-gu, Cheonam, Chungnam Province 330-720, Republic of Korea
a r t i c l e
i n f o
Article history: Received 3 December 2008 Accepted 8 July 2009 Available online 13 November 2009 Keywords: Fuel cell PEFC MEA Durability Freeze Thaw
a b s t r a c t Enhancement of system durability at the sub-freezing temperature is one of critical issues for the commercialization of polymer electrolyte fuel cell (PEFC) applications. In this work, effects of residual water at the sub-freeze condition on the gas diffusion layer (GDL)/membrane-electrode assembly (MEA) were investigated for the successful cold start-up of PEFC. The performance changes of MEAs were observed by 300 times of freeze/thaw (F/T) cycles with well designed single cell. The post analysis for the ( 30 to 70) °C F/T cycled MEA were conducted. The gradual decreases of I–V performance were observed after the F/T cycle number of 70. About 0.4 mV/cycle of degradation at 1 A/cm2 could be obtained at the controlled operating conditions. The main cause of performance decrease was in the weakened adhesion between the electrodes and membrane interface. The SEM images as well as cell resistance changes directly supported the degradation reasons of cell performance in the freeze condition. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction The polymer electrolyte fuel cells (PEFCs) which have high energy conversion efficiencies and environmentally friendly properties are getting much attention in recent years. For the commercialization, cost reduction and durability problems are major challenging points [1]. In PEFC, water management is one of vital issues. The operating conditions of dry as well as flooding can terribly deteriorate both the performance and the durability of PEFC. Especially, flooding phenomena of cells should be avoided during high current range operation. Water management can be considered in the two categories. One is for the ordinary operating condition and the other is for the start-up/shut-down process. Particularly start-up problem at the sub-freezing temperature is one of critical points for the successful commercialization of PEFC vehicles. Park et al. reported about the drying and flooding effects on the cell performance as well as the gas diffusion layer (GDL) design effects on the water management [2,3]. Oszcipok et al. presented the results on the cold start-up measurements under the isothermal condition of 10 °C [4]. Moore et al. showed four sets of cold start mode and its results [5]. Kim et al. observed the F/T effects on the MEA and GDL in a
* Corresponding author. Fax: +82 42 860 3104. E-mail address: [email protected] (G.-G. Park). 1567-1739/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.cap.2009.11.043
view of material itself [6–8]. Mukundan et al. reported that an increase of cell resistance caused by the breakaway of catalyst in MEA after a number of freeze/thaw cycles is a main reason for performance degradation [9]. In other tests of Mukundan et al., the polarization curves during freeze/thaw cycling experiments using different GDLs were investigated. A paper type GDL showed inferior durability to a cloth type GDL. The paper type GDL showed severe degradation in the mass transport region after 40 cycles, whereas the cloth GDL showed little performance loss even after 100 cycles [10]. In this work, by controlling the cell temperature from 30 °C to 70 °C, the performance as well as property changes of cells were investigated. 2. Experiments 2.1. Freeze/thaw cycles The single cell was controlled from 30 °C to 70 °C by intended ramping rates. For the reference, another single cell was evaluated with the same procedure in the temperature range of (30–70) °C. After the activation process of cell, I–V performance was evaluated. Before conducting the freeze/thaw cycle tests, the both sides of electrodes were purged with nitrogen to remove water in cells. In this work, the soaking time of the cell was 30 min after reaching 30 °C. The cell resistance also was monitored simultaneously with the temperature cycles.
S63
G.-G. Park et al. / Current Applied Physics 10 (2010) S62–S65
2.2. Electrochemical analysis
3. Results and discussion
Cell performance was evaluated by I–V characteristics after F/T cycles. Impedance and cyclic voltammogram were monitored to check the changes of cell resistances and electrochemical properties, respectively. The degree of polymer electrolyte membrane was checked by the chronoamperometry method.
The degradation of cell performance with temperature cycles are seen at Fig. 1. No significant I–V performance degradations could be observed for the purged cell till about 100 times F/T cycles. After 100 times of F/T cycles, the performance decrease occurred with the number of thermal cycle. But it’s difficult to distinguish the main reason of degradation by F/T cycles from the normal operating conditions. Even in the normal operating conditions, the cell performance can be decreased. To know the portion of cell degradation effected by the normal operation, temperature cycle tests from 30 °C to 70 °C were conducted. Fig. 1c shows that even if there was no freeze condition during whole operation, cell performance could be decreased from the number of temperature cycle at 200 times. The cyclic voltammogram and cell resistance were checked with I–V performance simultaneously. Fig. 1 shows that there is no direct relationship between the performance declines and the cyclic voltammogram changes. In the (b) and (d) of Fig. 1, much changes in voltammogram could be observed even in the case of no I–V performance changed. Fig. 2a shows that the electrochemical surface area (ECSA) decreasing rate may be affected mainly by the evaluation numbers of I–V performance rather than other factors. It’s clearly observed
2.3. Single cell test conditions A single cell was constructed with a MEA which has reinforced membrane, carbon papers (SGL carbon group), EPDM gaskets. The temperature of cell and humidifiers were maintained at 70 °C and 65 °C, respectively. During I–V test, the hydrogen utilization was 75% and the air utilization was 50%. The whole tests were conducted at the atmospheric pressure condition. 2.4. SEM analysis For the post analysis, the surfaces and interface of the MEA and GDLs were investigated by SEM images. Especially the interface between the electrodes and membranes was cautiously observed.
1 .0
3
0 .9 2
0 .8 0 .7
0 .4 0 .3 0 .2 0 .1 0
200
400
Current
Current [A]
Voltage [V]
0 .5
0 .0
1
Fr e sh Fre e ze /th aw c yc le 1 0 tim e s Freeze/thaw cycle 20 times Freeze/thaw cycle 30 times Freeze/thaw cycle 40 times Freeze/thaw cycle 50 times Freeze/thaw cycle 60 times Freeze/thaw cycle 70 times Freeze/thaw cycle 80 times Freeze/thaw cycle 100 tim es Freeze/thaw cycle 150 tim es Freeze/thaw cycle 200 tim es Freeze/thaw cycle 300 tim es
0 .6
0
Fresh Freeze/thaw cycle 10times Freeze/thaw cycle 20times Freeze/thaw cycle 30times Freeze/thaw cycle 40times Freeze/thaw cycle 50times Freeze/thaw cycle 60times Freeze/thaw cycle 70times Freeze/thaw cycle 80times Freeze/thaw cycle 100times Freeze/thaw cycle 150times Freeze/thaw cycle 200times
-1
-2
600
800
1000
1200
1400
-3 0.0
0.2
Density [mA [ /cm 2 ]
0.4
0.6
0.8
1.0
1.2
Cell Potential [V]
(a)
(b)
1.0
2.0
0.9
1.5
0.8
1.0 0.5
0.6
Fresh Fre eze/th aw 1cycle Freeze/thaw 2cycle Freeze/thaw 3cycle Freeze/thaw 4cycle Freeze/thaw 5cycle Freeze /thaw 10cycle Freeze/thaw 20cycle Freeze/thaw 60cycle Freeze/thaw 100cycle
0.5 0.4 0.3 0.2 0.1 0.0
Current [A]
Voltage [V]
0.7
0
200
400
Current
600
0.0 Fresh Freeze/thaw 1cycle Freeze/thaw 2cycle Freeze/thaw 3cycle Freeze/thaw 4cycle Freeze/thaw 5cycle Freeze/thaw 10cycle Freeze/thaw 20cycle Freeze/thaw 40cycle Freeze/thaw 60cycle Freeze/thaw 100cycle
-0.5 -1.0 -1.5 -2.0
800
1000
1200
Density[ mA/cm 2 ]
(c)
1400
-2.5 0.0
0.2
0.4
0.6
0.8
1.0
1.2
Cell Potential [V]
(d)
Fig. 1. The temperature cycle effects on the I–V performance and the cyclic voltammogram in PEFCs. (a) I–V performance for the purged cell (F/T cycle: 30 °C to 70 °C), (b) Cyclic voltammogram for the purged cell (F/T cycle: 30 °C to 70 °C), (c) I–V performance for the purged cell (temperature cycle: 30–70 °C), (d) Cyclic voltammogram for the purged cell (temperature cycle: 30–70 °C).
G.-G. Park et al. / Current Applied Physics 10 (2010) S62–S65
Normalized ECSA
S64
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
o
-30 ~ 70 C F/T Cycles o 30 ~ 70 C F/T Cycles
0
50
100
150
200
Number of Temperature Cycles (a) 2000 Temperature cycles
Temperature[ oC]
80
Resistance
60
1500
40 20
1000
0 -20
500
-40 -60 -80 0
200
400
600
800
Resistance [mohmcm2 ]
100
0 1000
Time [min]
(b) Fig. 2. Changes in electrochemical surface area (ECSA) and cell resistance for the freeze/thaw cycled MEAs. (a) Changes in normalized ECSA of MEAs and (b) changes in cell resistance during the F/T cycles ( 30 °C to 70 °C) from the cycle number 41– 50.
in Fig. 2b that one of the main evidence of performance degradation is increase of cell resistance during F/T cycles. No distinctive changes of cell resistance were observed at the initial state of F/T cycles (e. g. 1–10 times). As the number of F/T cycles passes to 50 times, the changes of cell resistance was detectable. Fig. 3 shows the freeze/thaw effects on the physical property of MEAs that were assembled and kept in a single cell. The interfaces between electrodes and membrane were severely damaged in Fig. 3b. In the sub-freeze temperature experienced MEA, much severe physical destruction could be observed. Fig. 3. SEM images of MEAs cross-sections. (a) Fresh, (b) freeze/thaw 300 cycles: 30 °C to 70 °C, and (c) temperature cycles 300 times: 30–70 °C.
4. Conclusions The main reasons of performance degradation of PEFCs in the freezing condition were tried to find out. Two types of temperature cycling tests were conducted. The first set was the F/T cycle ( 30 °C to 70 °C) with gas purged during shut-down process. The second was exactly the same process to the first set except the range of cycle temperature of 30–70 °C. The results showed that it’s difficult to directly correlate the changes in cyclic voltammogram, ECSA, H2 crossover rate and so onto the I–V performance degradations. The clear evidence of performance degradation is the increase of cell resistance with the number of F/T cycles. The physical destruction of the interface between electrodes and membrane also were observed.
Acknowledgments This work was supported by the Ministry of Knowledge Economy, Industry and Energy, Republic of Korea (2004-N-FC12-P-01). References [1]. [2] G.-G. Park, Y.-J. Sohn, T.-H. Yang, Y.-G. Yoon, W.-Y. Lee, C.-S. Kim, J. Power Sources 131 (2004) 182. [3] G.-G. Park, Y.-J. Sohn, S.-D. Yim, T.-H. Yang, Y.-G. Yoon, W.-Y. Lee, K. Eguchi, C.-S. Kim, J. Power Sources 163 (2006) 113. [4] M. Oszcipok, D. Riemann, U. Kronenwett, M. Kreideweis, M. Zedda, J. Power Sources 145 (2005) 407.
G.-G. Park et al. / Current Applied Physics 10 (2010) S62–S65 [5] [6] [7] [8]
M. Sundaresan, R.M. Moore, J. Power Sources 145 (2005) 534. S. Kim, C. Chacko, R. Ramasamy, M.M. Mench, ECS Trans. 11 (11) (2007) 577. S. He, S.H. Kim, M.M. Mench, J. Electrochem. Soc. 154 (10) (2007) B1024. S. Kim, B.K. Ahn, M.M. Mench, J. Power Sources 179 (2008) 140.
S65
[9] R. Mukundan, Y.S. Kim, F. Garzon, B. Pivovar, ECS Trans. 1 (8) (2006) 403. [10] R. Mukundan, Y.S. Kim, T. Rockward, J.R. Davey, B. Pivovar, D.S. Hussey, D.L. Jacobson, M. Arif, R. Borup, ECS Trans. 11 (1) (2007) 543.
Contents lists available at ScienceDirect
Current Applied Physics journal homepage: www.elsevier.com/locate/cap
Analysis on the freeze/thaw cycled polymer electrolyte fuel cells Gu-Gon Park a,*, Soo-Jin Lim a, Jin-Soo Park b, Sung-Dae Yim a, Seok-Hee Park a, Tae-Hyun Yang a, Young-Gi Yoon a, Chang-Soo Kim a a b
Fuel Cell Research Center, Korea Institute of Energy Research, 102, Gajeong-ro, Yuseong-gu, Daejeon, Republic of Korea Department of Environmental Engineering, College of Engineering, Sanmyung University, 300 Anseo-dong, Dongnam-gu, Cheonam, Chungnam Province 330-720, Republic of Korea
a r t i c l e
i n f o
Article history: Received 3 December 2008 Accepted 8 July 2009 Available online 13 November 2009 Keywords: Fuel cell PEFC MEA Durability Freeze Thaw
a b s t r a c t Enhancement of system durability at the sub-freezing temperature is one of critical issues for the commercialization of polymer electrolyte fuel cell (PEFC) applications. In this work, effects of residual water at the sub-freeze condition on the gas diffusion layer (GDL)/membrane-electrode assembly (MEA) were investigated for the successful cold start-up of PEFC. The performance changes of MEAs were observed by 300 times of freeze/thaw (F/T) cycles with well designed single cell. The post analysis for the ( 30 to 70) °C F/T cycled MEA were conducted. The gradual decreases of I–V performance were observed after the F/T cycle number of 70. About 0.4 mV/cycle of degradation at 1 A/cm2 could be obtained at the controlled operating conditions. The main cause of performance decrease was in the weakened adhesion between the electrodes and membrane interface. The SEM images as well as cell resistance changes directly supported the degradation reasons of cell performance in the freeze condition. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction The polymer electrolyte fuel cells (PEFCs) which have high energy conversion efficiencies and environmentally friendly properties are getting much attention in recent years. For the commercialization, cost reduction and durability problems are major challenging points [1]. In PEFC, water management is one of vital issues. The operating conditions of dry as well as flooding can terribly deteriorate both the performance and the durability of PEFC. Especially, flooding phenomena of cells should be avoided during high current range operation. Water management can be considered in the two categories. One is for the ordinary operating condition and the other is for the start-up/shut-down process. Particularly start-up problem at the sub-freezing temperature is one of critical points for the successful commercialization of PEFC vehicles. Park et al. reported about the drying and flooding effects on the cell performance as well as the gas diffusion layer (GDL) design effects on the water management [2,3]. Oszcipok et al. presented the results on the cold start-up measurements under the isothermal condition of 10 °C [4]. Moore et al. showed four sets of cold start mode and its results [5]. Kim et al. observed the F/T effects on the MEA and GDL in a
* Corresponding author. Fax: +82 42 860 3104. E-mail address: [email protected] (G.-G. Park). 1567-1739/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.cap.2009.11.043
view of material itself [6–8]. Mukundan et al. reported that an increase of cell resistance caused by the breakaway of catalyst in MEA after a number of freeze/thaw cycles is a main reason for performance degradation [9]. In other tests of Mukundan et al., the polarization curves during freeze/thaw cycling experiments using different GDLs were investigated. A paper type GDL showed inferior durability to a cloth type GDL. The paper type GDL showed severe degradation in the mass transport region after 40 cycles, whereas the cloth GDL showed little performance loss even after 100 cycles [10]. In this work, by controlling the cell temperature from 30 °C to 70 °C, the performance as well as property changes of cells were investigated. 2. Experiments 2.1. Freeze/thaw cycles The single cell was controlled from 30 °C to 70 °C by intended ramping rates. For the reference, another single cell was evaluated with the same procedure in the temperature range of (30–70) °C. After the activation process of cell, I–V performance was evaluated. Before conducting the freeze/thaw cycle tests, the both sides of electrodes were purged with nitrogen to remove water in cells. In this work, the soaking time of the cell was 30 min after reaching 30 °C. The cell resistance also was monitored simultaneously with the temperature cycles.
S63
G.-G. Park et al. / Current Applied Physics 10 (2010) S62–S65
2.2. Electrochemical analysis
3. Results and discussion
Cell performance was evaluated by I–V characteristics after F/T cycles. Impedance and cyclic voltammogram were monitored to check the changes of cell resistances and electrochemical properties, respectively. The degree of polymer electrolyte membrane was checked by the chronoamperometry method.
The degradation of cell performance with temperature cycles are seen at Fig. 1. No significant I–V performance degradations could be observed for the purged cell till about 100 times F/T cycles. After 100 times of F/T cycles, the performance decrease occurred with the number of thermal cycle. But it’s difficult to distinguish the main reason of degradation by F/T cycles from the normal operating conditions. Even in the normal operating conditions, the cell performance can be decreased. To know the portion of cell degradation effected by the normal operation, temperature cycle tests from 30 °C to 70 °C were conducted. Fig. 1c shows that even if there was no freeze condition during whole operation, cell performance could be decreased from the number of temperature cycle at 200 times. The cyclic voltammogram and cell resistance were checked with I–V performance simultaneously. Fig. 1 shows that there is no direct relationship between the performance declines and the cyclic voltammogram changes. In the (b) and (d) of Fig. 1, much changes in voltammogram could be observed even in the case of no I–V performance changed. Fig. 2a shows that the electrochemical surface area (ECSA) decreasing rate may be affected mainly by the evaluation numbers of I–V performance rather than other factors. It’s clearly observed
2.3. Single cell test conditions A single cell was constructed with a MEA which has reinforced membrane, carbon papers (SGL carbon group), EPDM gaskets. The temperature of cell and humidifiers were maintained at 70 °C and 65 °C, respectively. During I–V test, the hydrogen utilization was 75% and the air utilization was 50%. The whole tests were conducted at the atmospheric pressure condition. 2.4. SEM analysis For the post analysis, the surfaces and interface of the MEA and GDLs were investigated by SEM images. Especially the interface between the electrodes and membranes was cautiously observed.
1 .0
3
0 .9 2
0 .8 0 .7
0 .4 0 .3 0 .2 0 .1 0
200
400
Current
Current [A]
Voltage [V]
0 .5
0 .0
1
Fr e sh Fre e ze /th aw c yc le 1 0 tim e s Freeze/thaw cycle 20 times Freeze/thaw cycle 30 times Freeze/thaw cycle 40 times Freeze/thaw cycle 50 times Freeze/thaw cycle 60 times Freeze/thaw cycle 70 times Freeze/thaw cycle 80 times Freeze/thaw cycle 100 tim es Freeze/thaw cycle 150 tim es Freeze/thaw cycle 200 tim es Freeze/thaw cycle 300 tim es
0 .6
0
Fresh Freeze/thaw cycle 10times Freeze/thaw cycle 20times Freeze/thaw cycle 30times Freeze/thaw cycle 40times Freeze/thaw cycle 50times Freeze/thaw cycle 60times Freeze/thaw cycle 70times Freeze/thaw cycle 80times Freeze/thaw cycle 100times Freeze/thaw cycle 150times Freeze/thaw cycle 200times
-1
-2
600
800
1000
1200
1400
-3 0.0
0.2
Density [mA [ /cm 2 ]
0.4
0.6
0.8
1.0
1.2
Cell Potential [V]
(a)
(b)
1.0
2.0
0.9
1.5
0.8
1.0 0.5
0.6
Fresh Fre eze/th aw 1cycle Freeze/thaw 2cycle Freeze/thaw 3cycle Freeze/thaw 4cycle Freeze/thaw 5cycle Freeze /thaw 10cycle Freeze/thaw 20cycle Freeze/thaw 60cycle Freeze/thaw 100cycle
0.5 0.4 0.3 0.2 0.1 0.0
Current [A]
Voltage [V]
0.7
0
200
400
Current
600
0.0 Fresh Freeze/thaw 1cycle Freeze/thaw 2cycle Freeze/thaw 3cycle Freeze/thaw 4cycle Freeze/thaw 5cycle Freeze/thaw 10cycle Freeze/thaw 20cycle Freeze/thaw 40cycle Freeze/thaw 60cycle Freeze/thaw 100cycle
-0.5 -1.0 -1.5 -2.0
800
1000
1200
Density[ mA/cm 2 ]
(c)
1400
-2.5 0.0
0.2
0.4
0.6
0.8
1.0
1.2
Cell Potential [V]
(d)
Fig. 1. The temperature cycle effects on the I–V performance and the cyclic voltammogram in PEFCs. (a) I–V performance for the purged cell (F/T cycle: 30 °C to 70 °C), (b) Cyclic voltammogram for the purged cell (F/T cycle: 30 °C to 70 °C), (c) I–V performance for the purged cell (temperature cycle: 30–70 °C), (d) Cyclic voltammogram for the purged cell (temperature cycle: 30–70 °C).
G.-G. Park et al. / Current Applied Physics 10 (2010) S62–S65
Normalized ECSA
S64
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
o
-30 ~ 70 C F/T Cycles o 30 ~ 70 C F/T Cycles
0
50
100
150
200
Number of Temperature Cycles (a) 2000 Temperature cycles
Temperature[ oC]
80
Resistance
60
1500
40 20
1000
0 -20
500
-40 -60 -80 0
200
400
600
800
Resistance [mohmcm2 ]
100
0 1000
Time [min]
(b) Fig. 2. Changes in electrochemical surface area (ECSA) and cell resistance for the freeze/thaw cycled MEAs. (a) Changes in normalized ECSA of MEAs and (b) changes in cell resistance during the F/T cycles ( 30 °C to 70 °C) from the cycle number 41– 50.
in Fig. 2b that one of the main evidence of performance degradation is increase of cell resistance during F/T cycles. No distinctive changes of cell resistance were observed at the initial state of F/T cycles (e. g. 1–10 times). As the number of F/T cycles passes to 50 times, the changes of cell resistance was detectable. Fig. 3 shows the freeze/thaw effects on the physical property of MEAs that were assembled and kept in a single cell. The interfaces between electrodes and membrane were severely damaged in Fig. 3b. In the sub-freeze temperature experienced MEA, much severe physical destruction could be observed. Fig. 3. SEM images of MEAs cross-sections. (a) Fresh, (b) freeze/thaw 300 cycles: 30 °C to 70 °C, and (c) temperature cycles 300 times: 30–70 °C.
4. Conclusions The main reasons of performance degradation of PEFCs in the freezing condition were tried to find out. Two types of temperature cycling tests were conducted. The first set was the F/T cycle ( 30 °C to 70 °C) with gas purged during shut-down process. The second was exactly the same process to the first set except the range of cycle temperature of 30–70 °C. The results showed that it’s difficult to directly correlate the changes in cyclic voltammogram, ECSA, H2 crossover rate and so onto the I–V performance degradations. The clear evidence of performance degradation is the increase of cell resistance with the number of F/T cycles. The physical destruction of the interface between electrodes and membrane also were observed.
Acknowledgments This work was supported by the Ministry of Knowledge Economy, Industry and Energy, Republic of Korea (2004-N-FC12-P-01). References [1]
G.-G. Park et al. / Current Applied Physics 10 (2010) S62–S65 [5] [6] [7] [8]
M. Sundaresan, R.M. Moore, J. Power Sources 145 (2005) 534. S. Kim, C. Chacko, R. Ramasamy, M.M. Mench, ECS Trans. 11 (11) (2007) 577. S. He, S.H. Kim, M.M. Mench, J. Electrochem. Soc. 154 (10) (2007) B1024. S. Kim, B.K. Ahn, M.M. Mench, J. Power Sources 179 (2008) 140.
S65
[9] R. Mukundan, Y.S. Kim, F. Garzon, B. Pivovar, ECS Trans. 1 (8) (2006) 403. [10] R. Mukundan, Y.S. Kim, T. Rockward, J.R. Davey, B. Pivovar, D.S. Hussey, D.L. Jacobson, M. Arif, R. Borup, ECS Trans. 11 (1) (2007) 543.