Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)
Carbon Monoxide Removal Using Nickel Catalyst for PEM Fuel Cell Ahmad Nafees1, Saleem Ur Rahman2 and S.M. Javaid Zaidi2 1. Chemical Engineering Program, The Petroleum Institute, Abu Dhabi, UAE 2. Department of Chemical Engineering, King Fahd University of Petroleum and Minerals, Dhahran, SA Abstract: Essentially pure H2 is required as fuel for the proton exchange membrane fuel cells. The presence of even a few ppm of CO in H2 can severely poison the anode catalyst resulting in drastically reduced performance of the fuel cell. In a conventional fuel processing system, a large and complex CO gas removal system is required to bring down CO content to acceptable levels. The concept of CO removal by electrochemical filter was proposed recently by Lakshmanan et al [1], which promise to significantly reduce the size and complexity of the CO clean up system. One of the major limitations of the electrochemical filter is the use of expensive Pt catalyst. Therefore, it is highly desirable to replace it with non-noble metal catalyst. In this study, adsorption and electrooxidation of CO on Ni/Raney-Ni catalyst was studied to explore the possibilities of replacing the Pt/Pt-Ru catalyst by Ni/Raney-Ni catalyst. Experimentally it has been demonstrated that CO preferentially adsorb and electrooxidizes on Ni/Raney-Ni catalyst, making it a suitable catalyst for electrochemical filter application. Keywords: PEMFC, Ni/Raney-Ni catalyst, CO adsorption, CO electrooxidation, electrochemical filter enhance the performance of a fuel cell system (i.e., reformer plus fuel cell). It is an electrochemical device that preferentially oxidizes CO over H2 by using pulse potential control before the reformate gas enters the fuel cells. During the off-portion of the pulse adsorption of CO is preferred, while oxidation of CO is preferred during the on portion of the pulse. It uses a conventional PEMFC as a flow reactor for continuous preferential oxidation of CO over H2 [1]. In its present form, it uses a Pt-Ru bifunctional catalyst, which might not be the best choice of the catalyst. This work was undertaken to explore the possibilities of replacing the Pt-Ru catalyst with a non-noble metal catalyst. The objective of this study was to demonstrate the electrochemical removal of CO from H2 by using Ni/Raney-Ni catalyst. The objectives can be briefly divided in two parts. First part mainly deals with the study of CO adsorption and electrooxidation at a planer Ni electrode while second part focuses on the study of CO removal by using Raney-Ni catalyst.
1. INTRODUCTION Proton exchange membrane fuel cells (PEMFC) operating on pure H2 exhibits excellent energy and power characteristics. The application of pure H2 as a fuel in automotive and residential applications has many limitations primarily due to problems associated with the onboard storage and slow refueling [2]. In addition, the distribution infrastructure of the H2 fuel is very limited. Hence, onboard generation of H2 from easily available hydrocarbon fuels is gaining popularity. A conventional fuel processing system usually consists of a reformer, a high temperature and a low temperature water gas shift reactors (HTS & LTS) and a preferential CO oxidation reactor (PrOx) [3]. The gas exiting from reformer, commonly referred as reformate is a mixture of nitrogen (40 –50%), hydrogen (35 – 45%), CO2 (10 –20%), CO, water vapor and traces of other gases [4]. Unfortunately, such high CO content of the fuel can severely limit PEMFC performance due to the poisoning of the anode catalysts. Smooth operation of the fuel cell requires the level of CO in the fuel to be less than 20 ppm. Conventionally two stages of water gas shift and a PrOx reactor is employed to reduce the CO content of the reformate to tolerable levels. This approach has many limitations as size of the cleanup stages is more than an order higher than reformer and fuel cell stack combined. In addition PrOx reactor has poor selectivity for CO oxidation over H2 oxidation. Therefore, new efficient methods of CO removal from reformate is need of the hour.
2. EXPERIMATAL DETAILS A planer Ni electrode (> 99.99 %) of 0.3 cm2 surface area was used in the first part of the study. The electrode surface was carefully cleaned with increasing grades emery papers and washed firstly in boiling stone and finally by de-ionized water. The electrode was anodized in 85% H3PO4 solution and washed with de-ionized water. A phosphate buffer solution of 6.86 pH containing buffer salt (Fisher Gram-Pac) of Potassium Phosphate Monobasic/Sodium Phosphate Dibasic was used as electrolytic solution. A Pyrex H-type cell was used for housing working electrode, reference electrode and counter electrode respectively and two gas inlet points. In the second part of the study, Raney-Ni-Al catalyst (Merck-Schuchardt, Germany) having 88% Ni and 12% Al was used. The Raney-Ni electrode was made up of a hollow cylindrical Pyrex glass containing Raney-Ni cata-
Methanation, membrane purification, integrated water shift reactor, sorption process, adsorption and plasma induced reforming are some of the new CO removal method that are at various stages of development [4]. Recently concept of a novel electrochemical filter for CO removal was proposed Lakshmanan et al [1]. It promises to reduce the size and complexity CO cleanup system and Corresponding author: A. Nafees,
[email protected] 61
Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)
Nitrogen voltammogram, during forward scan represents a characteristic peaks at -0.25V, which is due to the Ni oxidation process. During backward scan, the cathodic current starts rising at -0.7V, representing onset of electrochemical activity, which is essentially due to H2 evolution process. Hori et al [7,8] have also observed similar phenomenon. During anodic scan, CO-voltammogram has a characteristic peak at -0.13V in contrast with -0.25V peak without CO exposure. This characteristic peak is due to electrooxidation of adsorbed CO. During cathodic scan, H2 evolution reaction gets delayed and starts at -0.96V due to presence of adsorbed CO. A H2 voltammogram was also obtained by exposing pure H2 gas and the voltammogram represents that H2 evolution starts at -0.7V in agreement with N2 voltammogram, while oxidation occurred throughout the cathodic range. From the comparison of above voltammogram it can be fortified that anodic peak in CO voltammogram is due to adsorbed CO on Ni surface [6].
lyst of approximately 2 mm bed height. The lower end of the working electrode was fritted to facilitate the electrochemical communication between working and counter electrodes during the electrochemical measurements. The experimental procedure and other details have been given elsewhere [5]. A platinum and a saturated calomel electrode (SCE) were used as counter and reference electrode. A potentiostat (Model No. 283, EG&G, PARC USA) driven by a manufacturer software package (Powersuit) was used for the experiments. Nitrogen gas was purged in all the three electrode compartment for at least 30 minutes in order to remove any oxygen containing species from the electrolyte solution. The flow of gas was maintained in such a way to keep the catalyst bed in fluidized state, which removes any active species present at the electrode surface. Subsequently flow of N2 gas at the working electrode was switched to a 1% CO gas mixture. The gas was allowed to flow for enough time to adsorb significant amount of CO at catalyst surface. Afterwards the flow of CO gas was again switched to N2 gas, to remove any dissolved CO from the electrolyte solution. All experiments were performed at room temperature and under N2 blanket.
Hori et al [9] tried to remove the adsorbed CO by electroreduction through repeated cathodic scanning. The electroreduction of CO may not be the best choice for CO removal. It is evident from Figure 1 that CO oxidizes easily at -0.13V. A multiple scan of the CO exposed electrode shows that the oxidation peak in subsequent scans was absent. Therefore it can be inferred that complete CO was oxidized in the first scan itself. A voltammogram of the electrode was obtained after CO oxidation and it was observed that both Ni oxidation and H2 evolution peaks coincide with the N2 voltammogram. This suggests that the entire adsorbed CO was successfully removed from the catalyst. In order to support the CV results, cathodic scans of the Ni electrode was also recorded under N2, CO and pure H2 environments. Figure 2 shows the cathodic scans of the electrode before and after CO oxidation. The detailed mechanism of the CO electrooxidation has been reported by Lakshmanan et al [1]. It is anticipated that similar electrooxidation mechanism will take place on Ni catalyst as well. The amount of the CO oxidation can be calculated by using Faraday’s law.
3. RESULTS AND DISCUSSION Main results of this study can be divided into two parts. The first part focuses on CO adsorption and electrooxidation on planer Ni electrode, while second part deals with CO adsorption and electrooxidation on Raney-Ni catalyst. The results of the first part have been reported elsewhere [6]. Only a part of results is reproduced here for understanding of the subject. Characterization of Raney-Ni catalyst is also included in the second part. 3.1. CO adsorption and electrooxidation on planer Ni electrode Figure 1 represents the voltammogram of planer Ni electrode under N2 and CO (1% CO in H2) exposures at 20 mV/s scan rate. The voltammogram were obtained by varying the potential during the forward scan (initial point -0.6V to end point +0.4V) and during backward scans (initial point +0.4V to end point -1.2V).
5 .0 O C t s 2 o O N P C
4 .0
1 4 .0
CO N2
Current Density (A/cm) x 10
-4
1 2 .0
1 0 .0
6 .0
2 .0
1 .0
2
Current Density (A/cm ) x 10
-4
8 .0
3 .0
4 .0
0 .0 2 .0
-1 .0 -1 .3
0 .0
-1 .2
-1 .1
-1 .0
- 0 .9
-0 .8
- 0 .7
-0 .6
- 0 .5
E le c t r o d e P o t e n t ia l ( V v s N H E ) -2 .0
Figure 2: Cathodic scans Ni electrode under different exposure -4 .0
3.2. Characterization of Raney-Ni catalyst The Raney-Nickel catalyst is usually produced by leaching Al from Ni-Al alloy in a hot alkali. During this process H2 gas evolves and gets adsorbed in the micropores resulting in pyrophoric nature of the catalyst. Hence,
-6 .0
-8 .0 -1 .4
-1 .2
-1 .0
-0 .8
-0 .6
-0 .4
-0 .2
0 .0
0 .2
0 .4
0 .6
E le c tro d e P o te n tia l (V v s N H E )
Figure 1: Voltammogram of CO and N2 exposed planer Ni electrode
62
Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)
Figure 3 shows the complete voltammogram of the N2 exposed Raney-Ni electrode under N2 blanket, obtained at 10 mV/sec scan rate by varying the potential from initial potential (-0.4V) to first vertex potential (-1.4V) to second vertex potential (+0.4V) upto end potential (-0.4V). It is evident that at the beginning of the cathodic scan (-0.4V) electrode current is essentially zero representing absence of any activity at the catalyst surface. As electrode is further scanned, cathodic current starts to increase at -0.7V. This cathodic current is essentially due to H2 evolution reaction, which is in agreement with the planer Ni electrode. During the forward scan the anodic current starts to increase at -0.3V and reaches a maximum value at +0.04V. As in the case of the planer electrode, this anodic activity is due to the passivation of the Raney-Ni catalyst. However a major shift of 135 mV in the dissolution potential is observed. This huge shift in the dissolution potential is due to significantly higher surface area of the Raney-Ni electrode as active area of the electrode is many times higher. The observation of the anodic activity during anodic scan is in close agreement to that of the Ni oxidation results reported by Hori et al [7,8].
catalyst needs to be passivated for characterization [10]. In this study the catalyst was passivated by H2O2 oxidation method [10]. The stabilized catalyst was characterized by scanning electron microscope (SEM), x-ray diffraction (XRD), inductive coupled plasma-absorption emission spectroscopy (ICP-AES) and BET surface area to study the structure and active area of the catalyst. Table1: Characteristic values of Raney-Ni catalyst Parameters Particle diameter BET surface area Density Ni content NiO content Mol Weight
Values 25.06 micron 72 m2/g 5.122 g/cm3 88% 10% 58.71
From ICP-AES elemental analysis, the passivated catalyst was found to have Ni(88 %), Al(2.7%), Cu(
