Electrochemical and Solid-State Letters, 8 共1兲 A52-A54 共2005兲
A52
1099-0062/2004/8共1兲/A52/3/$7.00 © The Electrochemical Society, Inc.
Electrochemical Reactions in a DMFC under Open-Circuit
Conditions
Q. Ye, T. S. Zhao,*,z H. Yang, and J. Prabhuram
Department of Mechanical Engineering, The Hong Kong University of Science and Technology,
Clear Water Bay, Kowloon, Hong Kong, China
Gas evolution, principally consisting of carbon dioxide and hydrogen, was observed in the anode flow field of a direct methanol
fuel cell 共DMFC兲 running under open-circuit conditions and at low oxygen flow rates. This finding is contrary to conventional
wisdom that electrochemical reactions cease as an external load is removed. The mechanism leading to this peculiar phenomenon
is explained theoretically and confirmed experimentally.
© 2004 The Electrochemical Society. 关DOI: 10.1149/1.1836111兴 All rights reserved.
Manuscript submitted July 17, 2004; revised manuscript received August 28, 2004. Available electronically November 29, 2004.
Direct methanol fuel cells 共DMFCs兲 have been envisaged as potential power sources for electric vehicles and portable applications
because of their high energy density and simpler operating system.1
Nevertheless, issues such as poor methanol electro-oxidation kinetics and methanol crossover through the polymer electrolyte membranes 共PEMs兲 from anode to cathode remain as obstacles to deploy
DMFCs for their practical applications. It has been found that using
the Nafion type membranes nearly 30-40% of methanol undergoes
crossover, which causes not only a depolarization effect on the cathode, but also a decrease in the utilization efficiency of fuel.2-4 Under
open-circuit conditions, the methanol crossover is mainly caused by
diffusion as a result of the concentration difference between anode
and cathode.
When used as portable power sources, DMFCs are preferred running under air breathing mode without the help of oxygen pumping
and air blowing devices to achieve compact and simplified system
design and low parasitic power cost.5 In this case, oxygen shortage
and severe flooding problems are usually encountered on the cathode, which reduce not only the open-circuit voltage 共OCV兲 but also
cell performance at high current densities. Qi et al.6 reported that
the OCV of a DMFC declined as the airflow rate decreased in the
temperature range of 60 to 80°C.
During the process of testing DMFCs, we have repeatedly observed gas evolution in the anode flow field when the cell was maintained under open-circuit conditions and at low oxygen flow rates.
This finding is something contrary to conventional wisdom that no
gas evolution normally occurs on the anode because electrochemical
reaction ceases under open-circuit conditions. We found that the
onset of this gas evolution was accompanied by a decline in OCV. In
this paper, we report on experimental evidence for this peculiar phenomenon and discuss the underlying mechanism leading to the
phenomenon.
Experimental
Membrane-electrode assembly (MEA).—An MEA having an active area of 4.0 ⫻ 4.0 cm was fabricated in house employing two
single-side ELAT electrodes from E-TEK and a Nafion membrane
115. Both anode and cathode electrodes used carbon cloth 共E-TEK,
Type ‘A’兲 as the backing support layer with 30% PTFE wet-proofing
treatment. The catalyst loading on the anode side was 4.0 mg cm⫺2
with unsupported 关Pt:Ru兴 Ox 共1:1 a/o兲, while the catalyst loading on
the cathode side was 2.0 mg cm⫺2 using 40% Pt on Vulcan XC-72.
The final MEA was formed by hot pressing at 135°C and 5 MPa for
3 min.
Single cell assembly and flow visualization.—The MEA was inserted into an in-house made single cell fixture. To visualize twophase transport phenomena on both the anode and cathode sides of
* Electrochemical Society Active Member.
z
E-mail: metzhao@ust.hk
the cell, the cell fixture was made of transparent poly methyl methacrylate 共PMMA兲, while both current collector plates were made of
perforated stainless steel. In this study, a single serpentine channel,
having 1.0 mm in channel width and 1.5 mm in rib width, was
formed on both the anode and cathode sides.
Cell test station and test conditions.—Experiments were carried
out in a DMFC test station detailed elsewhere.7 Briefly, 2-M methanol solution was pumped to the cell by a digital HPLC micro-pump
共Series III兲 and the flow rate was fixed at 1 standard cubic centimeters per min 共sccm兲. Dry pure oxygen was fed to the cathode of the
cell, whose flow rate was controlled and monitored by a mass flow
controller 共Omega FMA-7105E兲. Cell operating temperature as well
as the temperature of methanol and oxygen feed were maintained
at 60°C.
Results and Discussion
Similar to previous studies,7,8 we also found CO2 gas bubbles in
the anode flow field of the cell, when it was operated at high current
densities. Surprisingly, however, we still observed gas bubbles in the
anode flow field, when external load was removed and the oxygen
flow rate 共OFR兲 decreased to certain value. Figure 1 displays the gas
bubbles evolved in the anode side channels under the open circuit
condition, when the cell was operating at low OFRs 共15, 12, 11, and
10 sccm兲. When OFR decreased to and maintained at 15 sccm, small
bubbles started to appear near the channel downstream 共see Fig. 1a兲.
With decreasing OFR to 12 sccm, it can be seen from Fig. 1b that
more gas bubbles appeared and they covered about one forth area of
the MEA. At OFR of 11 sccm, the gas evolving area increased and
covered about one third of the MEA, as shown in Fig. 1c. With
further decreasing OFR to 10 sccm, almost two third of the MEA
area was covered by the gas bubbles 共see Fig. 1d兲. When OFR was
maintained to below 10 sccm, we noted that the gas bubble generating rate decreased but the gas bubble generating area became
larger.
The above mentioned gas bubble evolving process was accompanied by a decline in OCV of the cell. Figure 2 shows the variation
in OCV, running with 2-M methanol concentration at 60°C, with
time when OFR decreased. Initially, when OFR was maintained in
the range of 50 to 20 sccm, the cell exhibited a typical and stable
DMFC OCV of around 0.63 V. When decreasing OFR to 15 sccm,
fluctuation in OCV 共ranging from 0.63 to 0.53 V兲 was observed.
With further decreasing OFR to 12, 11, and 10 sccm. OCV became
stable and decreased to 0.48, 0.45, and 0.43 V, respectively. Note
that each image shown in Fig. 1 was taken under a steady state, in
which the OCV remained constant.
Because the gas bubbles on the anode were produced while lowering OFR, the gas evolution has something to do with the concentration distribution of O2 on the cathode. Due to reaction with crossover methanol, a decrease in OFR causes a decrease in the average
concentration of O2 over the entire channel. Because the cathodic
reaction causes a decrease in O2 concentration along the flow chan-
Electrochemical and Solid-State Letters, 8 共1兲 A52-A54 共2005兲
A53
Figure 3. Schematic representation of electrochemical reactions under opencircuit and oxygen-shortage conditions.
Figure 1. Gas bubble evolution on the anode side under typical oxygen flow
rates 共a兲 15, 共b兲 12, 共c兲 11, and 共d兲 10 sccm.
nel toward the exit, a shortage of O2 is likely to occur in the channel
downstream region at lower OFRs. In addition to acting as oxidant
on the cathode, oxygen flow stream sweeps the liquid water, resulted
from the cathodic reaction and permeated from the anode, away to
the channel exit, implying that water concentration becomes higher
but O2 concentration gets lower in the channel downstream. Moreover, water vapor and CO2 , resulting from the reactions on the
cathode, also causes a decrease in O2 concentration in the channel
downstream. For these reasons, once OFR decreases just below a
critical value 共around 15 sccm for the cell in this work兲, serious
flooding occurs in the channel downstream region even under opencircuit conditions. The flooded region on the cathode increases with
decreasing OFR, which can prevent O2 from access to the cathode
catalyst sites. In addition, the crossover of methanol from the anode
also enhances the flooding in this region. In the meantime, in the
upstream region, flooding is unlikely to occur because O2 concentration is relatively high and water concentration is relatively low.
As such, the MEA area can be divided into two regions: unflooded
region 共oxygen rich region兲 and flooded region 共oxygen lean region兲, as sketched in Fig. 3. Although the two regions are connected
in parallel by both electrodes and the same electrolyte 共membrane兲,
the lateral proton conduction in the electrolyte is so low that the
flooded and unflooded sections of the cell act as if they were two
independent cells that are connected electrically but not ionically.
This explanation can be further clarified by the fact that the in-plane
electrical resistivity 共5 m⍀ cm兲 of carbon cloth 共E-TEK, Type ‘A’兲
is more than one thousand times smaller than the protonic resistivity
共5.9 ⍀ cm at 80°C兲 of the Nafion membrane.9,10 Under this condition, the flooded regions are expected to act similarly to the
methanol/inert cells 共or electrolytic cells兲, used by Ren et al.11 to
measure the rate of methanol crossover through the membrane. At
the same time, the unflooded region of the MEA still acts as a
normal DMFC 共or galvanic cell兲 and maintains a certain voltage
between two electrodes. In the flooded region, as long as the magnitude of voltage applied on the electrodes is sufficient, the following electrolytic reactions will take place:
CH3 OH⫹H2 O → CO2 ↑ ⫹ 6H⫹ ⫹ 6e⫺ 共 Anodic Reaction兲
⫹
⫺
6H ⫹ 6e → 3H2 ↑ 共 Cathodic Reaction兲
关1兴
关2兴
These methanol electrolytic reactions consume electrical power
and behave just like an internal electrical load connected to the
unflooded region, where the galvanic cell is still operable. Therefore, the galvanic cell in the unflooded region does not stay in the
open-circuit status, although the external load is disconnected. As
evident from Fig. 2, once the gas evolution was initiated, the DMFC
exhibited a potential decline from OCV to a low value, indicating
that the galvanic cell was virtually discharging. Hence, the following
electrochemical reactions occur in the un-flooded region:
CH3 OH⫹H2 O → CO2 ↑ ⫹ 6H⫹ ⫹ 6e⫺ 共 Anode兲
Figure 2. Variation in OCV with different oxygen flow rates on the cathode.
and
3
O ⫹ 6H⫹ ⫹ 6e⫺ → 3H2 O 共 Cathode兲
2 2
关3兴
关4兴
Electrochemical and Solid-State Letters, 8 共1兲 A52-A54 共2005兲
A54
Table I. OCV and gas flow rates in the efflux from anode under
open-circuit conditions.
Oxygen
flow rate 共sccm兲
OCV 共V兲
Total gas
flow rate 共sccm兲
H2 flow rate
共sccm兲
CO2 flow
rate 共sccm兲
tures 共higher than 160°C兲. Therefore, for the cell operated at 60°C in
this work, the possibility of gas evolution due thermal decomposition of methanol is eliminated.
12
11
10
Conclusions
0.48
1.3
0.45
1.3
0.43
1.24
0.74
0.76
0.68
0.56
0.54
0.56
Coexistence of the galvanic and methanol electrolytic reactions
in a single cell has been found to be the reason why gas evolution
occurs in the anode flow channel under open-circuit conditions. Although the whole cell is running under open circuit conditions, there
is still self-discharging current causing methanol consumption and
hydrogen evolution. This self-discharging mechanism also explains
why the variation in OCV depends on oxygen flow rates. Practically
speaking, the generation of hydrogen not only decreases the net
output of electrical power for running cells but also consumes fuel
continuously even under open-circuit conditions. Thus, it is important to maintain adequate and uniform supply of oxygen on the
cathode side when DMFCs are under both running and standby conditions. In addition, produced hydrogen requires additional safety
considerations for DMFC design and operation.
This discharging current, flowing between the flooded and unflooded regions within the same MEA, can be regarded as a selfdischarging current. As indicated in Fig. 3, the coexistence of Reactions 2 and 3 within the same MEA explains why gas evolution took
place on the anode flow channel when external load was removed.
We collected the gases generated in the anode flow channel and
conducted a chromatography analysis. The result confirmed that the
gas mixture principally consisted of CO2 and H2 with small amount
of methanol and water vapor. The formation rates of CO2 and H2
gases were measured by the water displacement method. The measured formation rates of CO2 and H2 gases corresponding to typical
OFRs are listed in Table I. It is seen from Table I that the measured
ratio of CO2 to H2 is about 0.7 at different OFRs, which is larger
than the theoretical ratio of 1:3 determined from the reactions given
by Eq. 2 and 3. The extra CO2 gas might be caused by CO2 crossover from the cathode.12 Although the production of CO2 gas was
substantial, there were only a few of bubbles in the upstream region
as seen from Fig. 1. This can be attributed to the fact that most of the
CO2 produced dissolved into methanol solution. The data in Table I
also shows that the total gas flow rate on the anode is rather constant
with oxygen flow rate. The total gas evolution rate depends not only
on the flooded area on the cathode but also on the open-circuit
potential. As the oxygen flow rate decreases, the flooded area increases by expanding upstream, but the circuit potential decreases.
The competition between these two factors determined that the total
gas evolution rate remained almost the same for the three oxygen
flow rates listed in Table I.
It should be recognized that another possibility of gas evolution
in the anode flow channel is that methanol can be thermally decomposes on the catalyst sites to yield H2 , CO2 , and CO.13 Such methanol reforming reactions, however, happen only at elevated tempera-
Acknowledgments
The work described in this paper was fully supported by a grant
from the Research Grants Council of the Hong Kong Special Administrative Region, China 共project no. HKUST6178/00E兲.
The Hong Kong University of Science and Technology assisted in meeting the publication costs of this article.
References
1. K. Kordesh and G. Simader, Fuel Cells and Their Applications, Wiley-VCH, Weinheim, 共1996兲.
2. D. Chu and S. Gilman, J. Electrochem. Soc., 141, 1770 共1994兲.
3. P. S. Kauranen and E. Skou, J. Electroanal. Chem., 408, 189 共1996兲.
4. C. Lamy, J. M. Leger, S. Srinivasan, in Modern Aspects of Electrochemistry, Vol.
34, J. O’M. Bockris, B. E. Conway, and R. E. White, Editors, p. 53, Kluwer
Academic/Plenum Publishers, New York 共2001兲.
5. D. J. Kim, E. A. Cho, S. A. Hong, I. H. Ho, and H. Y. Ha, J. Power Sources, 130,
172 共2004兲.
6. Z. Qi and A. Kaufman, J. Power Sources, 110, 177 共2002兲.
7. H. Yang, T. S. Zhao, and Q. Ye, J. Power Sources, 139, 1 共2004兲.
8. P. Argyropoulos, K. Scott, and W. M. Taama, Electrochim. Acta, 44, 3575 共1999兲.
9. M. Mathias, J. Roth, W. Lehnert, and J. Fleming, Handbook of Fuel CellsFundamentals, Technology and Applications, Vol. 3, John Wiley & Sons, Ltd.,
Chichester 共2003兲.
10. G. Hoogers, Fuel Cell Technology Handbook, CRC Press LLC, Boca Raton, FL
共2003兲.
11. X. Ren, T. E. Springer, and S. Gottesfeld, J. Electrochem. Soc., 147, 92 共2000兲.
12. A. Heinzel and V. M. Barragan, J. Power Sources, 84, 70 共1999兲.
13. B. A. Peppley, J. C. Amphlett, and L. M. Kearns, Appl. Catal., A, 179, 21 共1999兲.