The micro-scale analysis of the micro direct methanol fuel cell

The micro-scale analysis of the micro direct methanol fuel cell

Energy 100 (2016) 10e17 Contents lists available at ScienceDirect Energy journal homepage: www.elsevier.com/locate/energy The micro-scale analysis ...

2MB Sizes 2 Downloads 26 Views

Energy 100 (2016) 10e17

Contents lists available at ScienceDirect

Energy journal homepage: www.elsevier.com/locate/energy

The micro-scale analysis of the micro direct methanol fuel cell Zhenyu Yuan, Jie Yang*, Xiaoyang Li, Shikai Wang College of Information Science and Engineering, Northeastern University, Shenyang 110819, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 June 2015 Received in revised form 18 January 2016 Accepted 20 January 2016 Available online xxx

In this paper, the behavior of micro-scale effect in a micro direct methanol fuel cell (mDMFC) is investigated through both simulations and experiments. A model is built to describe the methanol distribution in the diffusion layer with different feature sizes. In addition, the dynamic movement of a single CO2 bubble is also simulated to study the two-phase characteristics in the micro channels with various aspect ratios. Furthermore, a metal-based transparent mDMFC with the active area of 0.64 cm2 is designed and fabricated to evaluate the two-phase flow characteristic as well as the corresponding cell performance. The experimental results reveal that when the feature size scales down to 0.6 mm and 0.4 mm, the peak power density of 27.1 mW cm2 and 26.3 mW cm2 are achieved at room temperature. Finally, the effect of adjusting channel aspect ratio is experimentally investigated to improve the inner convection transport and the cell output, and the results are well in agreement with the simulation. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Micro direct methanol fuel cell Feature size Micro scale

1. Introduction With the rapid development of Micro Electro Mechanical System (MEMS) technology, mDMFC has been considered as a promising power candidate for portable applications due to its high energy density, high efficiency and environmental-friendly [1e4]. The miniaturization of portable electronic devices requires the power supply unit to be scaled down proportionally. The volume of micro direct methanol fuel cell is determined by the size of current collector, in which the most critical part is the inner flow channel [5e7]. The main function of the flow channel is to ensure the uniform fuel supply for reaction region and to maintain the equilibrium between the fuel distribution and the current collection [8e10]. Therefore, significant attentions have been devoted to the effect of inner channel on mDMFC performance recently. Park et al. [11] studied four different types of serpentine flow-channel geometry structures, and the corresponding current density distributions and the polarization were observed. Wong et al. [12] investigated the effect of the anode flow field design on the performance of mDMFC. The experimental results indicated that the serpentine flow field exhibited higher cell voltages than that of the parallel flow field at high current densities. Based on the single serpentine flow channel, Xu et al. [13] presented a convectionenhanced serpentine flow channel. The results indicated that the

* Corresponding author. Tel./fax: þ86 24 83683832. E-mail address: [email protected] (J. Yang). http://dx.doi.org/10.1016/j.energy.2016.01.057 0360-5442/© 2016 Elsevier Ltd. All rights reserved.

novel flow field created larger pressure difference between adjacent flow channels, and enhanced in-plane forced flow through the electrode porous layer. However, previous studies focused only on fixed cell dimension and lack in-depth theoretical analysis about the influence of the scale-down effect on the cell performance. Under the operating condition, the decrease of cell feature size would introduce micro scale effect and affect the motion state of two-phase flow, which has a great influence on the output performance. Therefore, it is essential to conduct a comprehensive study on both simulation and experiment to fully explore the relationship between the feature size and inner transport characteristics. The aim of this paper is to reveal the methanol/CO2 convection and the mDMFC performance by applying the flow channel with different feature sizes and aspect ratios. A two-phase, threedimensional cell model coupled with mass/momentum transports is established. In this model, the scale-down effect on inner transport and the channel aspect ratio on gas dynamic are numerically defined. In addition, a 0.64 cm2 stainless steel-based mDMFC is fabricated using laser-cut technology, at which the effects of feature size and aspect ratio on cell performance are experimentally investigated to verify the simulation. The results from a series of experiments show that the mDMFC behaviors are influenced by the cell feature size in a complex manner. An optimized structure with the cell feature size of 0.6 mm and the channel aspect ratio of 2:1 is then determined to enhance the methanol mass transport efficiency, CO2 emission rate and the output performance.

Z. Yuan et al. / Energy 100 (2016) 10e17

2. The model analysis

11

 !  V$ rl;a u l;a ¼ ml;a

(3)

2.1. Model description The size of mDMFC is mainly determined by three factors: the type of the flow field, the dimensions of the channel and the rib, and the thickness of the gas diffusion layer. The micro scale effect on inner mass transfer is characterized by the feature size, denoted as F in Fig. 1. With the same external methanol supply, the linear velocity of methanol solution inside the micro channel is inversely proportional to the feature size, and the corresponding relationship can be described as:

N ¼ Av ¼ nAi v ¼ C

(1)

where N, A, v, n and Ai represent the methanol volume from the external supply, the cross-section of the anode flow channel, the linear velocity, channel number and the cross-section of a single channel, respectively. Therefore, the relationship between the linear velocity and the feature size can be given by:



N N f nAi F

(2)

Based on equation (2), under the same external supply condition, the linear velocity of methanol in anode flow channel is increased when the feature size is scaled down. Therefore, a 3D physical model covering the solution region of anode flow channel, anode gas diffusion layer (AGDL) and anode catalyst layer (AGL) is established in this paper. Fig. 2 shows the calculation domain of three micro-scale designs with different aspect ratios (1:1, 2:1 and 4:1). The diffusion layer is defined as homogeneous porous electrode while the mDMFC is assumed under steady-state. In this model, the effective mass transfer area is 0.8 cm  0.8 cm, the width of the channel and rib are both 0.6 mm, and the open ratio is remained the same for all three designs. In order to better analyze the gas dynamic in the micro channel, a typical L-shape unit was presented as the calculation domain, as shown in Fig 2. The continuity equation suitable for methanol solution can be described as:

where rl;a represents the average density of the liquid phase sub! stance, u l;a is the average velocity, and ml;a is the source term. Similarly, the continuity equation of gas is given by:

 !  V$ rg;a u g;a ¼ mg;a

(4)

where rg;a denotes the average density of the gas phase substance, ! u g;a is the average velocity, and mg;a is the source term of gas phase substances. The liquid/vapor phase material are not produced or consumed in the anode gas diffusion layer because there is no electrochemical reaction. Meanwhile, CO2, electrons and protons are produced in the anode catalyst layer. According to Faraday's law, the source term methanol solution can be expressed as:

ml;a ¼

8 <

0 j s : MMeOH a a  MH O ja sa 2 6F 6F

AGDL ACL

(5)

where ja is the anode oxidation reaction rate and sa is the specific surface area. Similarly, the source term of anode CO2 is defined as:

mg;a ¼

8 <

0

AGDL

: MCO ja sa 2 6F

ACL

(6)

The momentum transport in diffusion layer can be described by Darcy law:

Kkrl ! u l;a ¼ Vpl;a ml

Kkrg ! u g;a ¼ Vpg;a mg

(7)

where K and kr represent the absolute permeability and the relative permeability, and m is the viscosity. kr in Equation (7) can be further modified as:

krl ¼ s3

krg ¼ ð1  sÞ3

Fig. 1. The feature size schematic of DMFC anode flow field.

(8)

12

Z. Yuan et al. / Energy 100 (2016) 10e17

Fig. 2. The schematic of the micro scale effect of mDMFC flow field.

In the anode porous medium, the capillary pressure pc is defined as the difference between the gas pressure and the liquid pressure, which can be expressed as:

pc ¼ s cos qc ðε=KÞ0:5 JðsÞ

(9)

where s represents the coefficient of surface tension, qc is contact angle, and ε is the porosity of the anode porous layer. Based on the ideal gas law, the relationship between CO2 density and pressure can be described as:

rCO2 ¼

  pl þ pref MCO2 RT

(10)

where MCO2 , R and T are molar mass, ideal gas constant and the operating temperature, respectively. The mass transport can be described using the ConvectioneDiffusion equation as follows:

  V$  Deff MeOH VCMeOH þ CMeOH ul;a ¼ SMeOH

(11)

The first term of the left-hand equation represents transfer term of the methanol diffusion, and the second term is the transfer term of the methanol convection. The above parameters and their values used in the model are listed in Table 1. 2.2. Simulation results and discussion The model is numerically solved by finite element method using COMSOL Multiphysics, with the number of degrees of 152,952. The calculation time is about 8 h on a Inter(R) Core(TM) i5 3.40 GHz PC. The methanol distribution on the diffusion layer with different feature sizes is displayed in Fig. 3. In this figure, the corresponding feature size from bottom to top are 1.5 mm, 1.2 mm, 1.0 mm, 0.8 mm and 0.6 mm. During the operation, the under-rib transport is critical

to the anode-active mDMFC because of the decreasing porosity of this region and the increasing mass transport resistance. Also, the uniform distribution is of great importance to the cell performance. The data illustrated that the larger-rib exhibited non-uniform concentration distribution over the active area, and the methanol concentration under the channels is much higher than that under the ribs. Compared with the cell with a larger feature size, the methanol distribution in diffusion layer becomes more uniform when the feature size is reduced. The mDMFC performance is related to the movement of gaseliquid flow. The investigation of channel structure has a great importance on optimization of the bubble dynamic characteristics. In micro channels, the volume of gas CO2 produced by oxidation reaction increases gradually. Fig. 4 depicts the bubble dynamic under different aspect ratios. The bubble was generated at inlet after T ¼ 0. The red part of the color bar represents the gas CO2, while the blue part represents the methanol solution. In this simulation, the aspect ratio of the channel was set as 1:1, 2:1 and 4:1, and the corresponding total removal time of gas CO2 are 0.08 s, 0.04 s and 0.4 s, respectively. The simulation result indicates that with the same open rate and operation parameters, the optimal aspect ratio is 2:1. The main reason is that at the same supply rate, the methanol linear velocity in the channel with the aspect ratio of 1:1 is less than that in the channel with the aspect ratio of 2:1. As such, the less driving force leads to the longer bubble detachment time. Moreover, bubbles generated from anode oxidation reaction move in the micro channel with a spherical shape. In the channel with the aspect ratio of 4:1, due to surface tension effect, the gas flow momentum at the channel is blocked by the local resistance, thus the mobility of bubbles is reduced. Fig. 5 shows the simulation of the average size of CO2 bubbles. Three different aspect ratios of 1:1, 2:1 and 4:1 were selected during the simulation, and the corresponding CO2 diameter are 3.5  104 m, 2.6  104 m and 8.5  105 m, respectively. The data indicates that larger aspect ratio yield smaller gas size.

Z. Yuan et al. / Energy 100 (2016) 10e17

13

Table 1 Parameters used in the 3D mDMFC model calculations. Property

Value

Working temperature T (K) Gas constant R (J$(mol$K)1) Standard pressure Patm (Pa) Faraday constant F (C$mol1) Molar mass of water MH2 O (kg$mol1) Molar mass of methanol MMeOH (kg$mol1) Molar mass of CO2 MCO2 (kg$mol1) Molar mass of O2 MO2 (kg$mol1) Water density rH2 O (kg$m3) Methanol density rMeOH (kg$m3) Liquid dynamic viscosity ml (kg$(m$s)1) Gas dynamic viscosity mg (kg$(m$s)1) Methanol diffusivity DMeOH (m2$s1) O2 diffusivity DO2 (m2$s1) Surface tension s (N$m1) ref Anode referential current im (A$m2) ref Cathode referential current iO2 (A$m2) Permeability of gas diffusion layer (m2) Permeability of catalyst layer (m2) Porosity of gas diffusion layer Porosity of catalyst layer Anode inlet velocity VMeOH;in (mL$min1) Inlet concentration CMeOH;in (mol$m3) Transfer coefficient on ACL aa Transfer coefficient on CCL ac Anode equilibrium potential Eaeq (V)

300 8.314 1.013  105 96,485 0.018 0.032 0.044 0.032 1000 791.7 0.9  103 14.96  106 105.4163999.778/T 1.775  105  (T/273)1.823 0.0625 94.25exp(35,570  (1/353  1/T)) 0.04222exp(73,200/R  (1/353  1/T)) 1.2  1012 1.5  1014 0.7 0.4 1.0 1000 0.239 0.875 1  (131350  408.22  T)/6/F

Cathode equilibrium potential Eceq (V)

1  (285830  3 þ 489.52  T)/6/F

3. Experimental The transparent mDMFC with the active area of 0.64 cm2 was fabricated to verify the simulation results. The current collectors based on 316L stainless steel were fabricated using laser-cut technology, and the fabrication error was less than 0.01 mm. A NiP/Au metal layer was deposited on the finished current collector to resist electrochemical corrosion and reduce contact resistance. The NiP metal layer is served as the intermediate layer to enhance the

bonding force between the stainless steel and Au metal layer. The thickness of NiP/Au metal layer was measured by XRAY thickness measurement system, as shown in Table 2. The polymethylmethacrylate (PMMA) plate was utilized to fabricate visualization window. The Nafion117 membrane was adopted as the proton exchange membrane, whose electrode assembly is composed of 4.0 mg cm2 Pt/Ru on the anode catalyst layer and 2.0 mg cm2 Pt on the cathode catalyst layer. Finally, the mDMFC assembly was finished with the auxiliary components, as shown in Fig. 6.

Fig. 3. The methanol distribution on catalyst layer with different feature sizes: 1.5 mm, 1.2 mm, 1.0 mm, 0.8 mm and 0.6 mm from bottom to top.

14

Z. Yuan et al. / Energy 100 (2016) 10e17

Fig. 4. The gas dynamic under the different aspect ratios.

Z. Yuan et al. / Energy 100 (2016) 10e17

15

Fig. 5. The average size of CO2 in the channel with different aspect ratios.

Table 2 The test report of XRAY XDL. Parameters

Ni/P

Au

Mean value (mm) Measuring time (s) Standard deviation (mm) C.O.V (%) Test range (mm) The lowest reading (mm) The highest reading (mm)

32.60 30 1.390 4.25 2.58 31.60 34.20

3.31 30 0.141 4.26 0.28 3.16 3.44

4. Results and discussion In order to accurately analyze the micro scale effect on inner gaseliquid transport and cell performance, three operating parameters were remained constant during the experiments, including the flow rate of 1.0 ml min1, the methanol concentration of 1.0 mol L1 and the operating temperature of 20  C. Fig. 7 shows the effect of feature sizes (0.4 mm, 0.6 mm, 1.0 mm and 1.5 mm) on the cell output. It can be seen that the maximum power density of 27.1 mW cm2 and the minimum power density of 13.9 mW cm2 were obtained under the feature size of 0.6 mm and 1.5 mm, respectively. When the feature sizes were 1.0 mm and 0.4 mm, the peak power density were 21.1 mW cm2 and 26.3 mW cm2. The main reason is that the decrease of the feature size results in a more uniform methanol distribution in anode flow field as well as on the surface of the catalyst layer. Meanwhile, it is worth to notice that the increase trend of the maximum output power density gets saturated with the decrease of the anode flow

field feature size. This is due to the fact that although the decrease of the feature size is helpful to enhance the linear velocity of methanol in the anode flow channel, smaller feature size enhances the micro-scale effect inside the cell, which has a great influence on the gaseliquid convective transport inside the mDMFC. This is the reason why a slightly lower performance was obtained for the feature size of 0.4 mm. Besides, smaller feature size normally incurs higher processing cost. Therefore, the feature size of 0.6 mm is the optimal selection. The effects of channel aspect ratio on cell output were also studied and the results are summarized in Fig. 8. During the experiment, the single serpentine flow field with 50% open ration was used. It can be seen that while the aspect ratio increases from 1:1 to 4:1, the peak power density experiences a nonlinear trend of initially going up and then dropping off. A maximum output power density of 25.6 mW cm2 is achieved with the aspect ratio of 2:1. On the contrary, the fuel cell exhibits the lowest power density of 19.5 mW cm2 when the aspect ratio is 4:1. The trend in Fig. 8 could be further validated by the visual observation of CO2 bubbles shown in Fig. 9. The CO2 emission dynamics in the different channel were detected by CMOS High Speed Imaging System, featured 120 dB dynamic range, 1024  1024 resolution, and 50 frames per second capturing rate. The improved performance with the aspect ratio of 2:1 is primarily attributed to the fact that at the same operating conditions, the CO2 emission velocity in the micro channel is much higher than that in the channel with other aspect ratios. As calculated from the figure, the average time of gas slug pass through the channel of 2:1 and channel of 1:1 are 0.43 s and 0.90 s respectively. The results of Fig. 9 also indicates that when the aspect ratio is 4:1, the bubbles inside

Fig. 6. Picture of the air-breathing mDMFC and the current collector.

16

Z. Yuan et al. / Energy 100 (2016) 10e17

Fig. 7. The IeV and IeP performance of different feature size.

Fig. 8. Power density curves of the mDMFC with different aspect ratio.

the anode flow field mainly appears as large and long slugs, which greatly reduces the effective contact area between the methanol solution and the diffusion layer. When the aspect ratio decreased from 4:1 to 2:1, the gas bubbles became gradually smaller and shorter, resulting in the increase of the effective membrane

electrode assembly (MEA). Moreover, this difference was more significant with the increase of current density due to the fact that the quantity of CO2 was gradually increased with the increase of the operating current. The quantity of CO2 can be expressed using the following equation:

Fig. 9. The CO2 distribution under different aspect ratio.

Z. Yuan et al. / Energy 100 (2016) 10e17

QCO2 ¼

Sactive $MCO2 I T 6pF

(12)

where Sactive represents the MEA area, MCO2 is the molar mass of CO2, I and T are operating current and cell temperature, respectively. 5. Conclusions In this paper, the effects of micro scale on mass transport and output characteristics were conducted by simulation and experimental investigations. Firstly, a multi-physics model coupled with mass/momentum transports was established, and the simulation results based on the model were well fitted with the experimental data. It was discovered from the model that the methanol concentration and the dynamic distribution of CO2 were significantly correlated to the feature size and the channel aspect ratio. Secondly, a transparent, metal-based mDMFC was designed and fabricated using laser-cut technology. Thirdly, the polarization test and visualization test were conducted to verify the model accuracy. Experimental results showed that the feature size and the aspect ratio had complex impacts on cell performance. With the optimal feature size of 0.6 mm and the aspect ratio of 2:1, the maximum power density of 27.1 mW cm2 was scored. Meanwhile, the emission of CO2 gas in the cell patterned with the channel aspect ratio of 2:1 was faster than that with other aspect ratios. The investigation on micro scale effect is practically critical for portable applications of mDMFC. Acknowledgments The work described in this paper was supported by the National Natural Science Foundation of China (No. 61504023 and No. 61372015), Research Fund for the Doctoral Program of Higher

17

Education (No. 20130042120023) and Fundamental Research Funds for the Central Universities in China (No. N140403001).

References [1] Carton JG, Lawlor V, Olabi AG, Hochenauer C, Zauner G. Water droplet accumulation and motion in PEM (Proton Exchange Membrane) fuel cell minichannels. Energy 2012;39:63e73. [2] Carton JG, Olabi AG. Design of experiment study of the parameters that affect performance of three flow plate configurations of a proton exchange membrane fuel cell. Energy 2010;35:2796e806. [3] Falc~ ao DS, Pereira JP, Rangel CM, Pinto AMFR. Development and performance analysis of a metallic passive micro-direct methanol fuel cell for portable applications. Int J Hydrogen Energy 2015;40:5408e15. [4] Falc~ ao DS, Oliveira VB, Rangel CM, Pinto AMFR. Experimental and modeling studies of a micro direct methanol fuel cell. Renew Energy 2015;74:464e70. [5] Carton JG, Olabi AG. Representative model and flow characteristics of open pore cellular foam and potential use in proton exchange membrane fuel cells. Int J Hydrogen Energy 2015;40:5726e38. [6] Li XL, Faghri A. Local entropy generation analysis on passive highconcentration DMFCs (direct methanol fuel cell) with different cell structures. Energy 2011;36:403e14. [7] Ali Doner, Ramazan Solmaz, Gulfeza Kardas. Fabrication and characterization of alkaline leached CuZn/Cu electrode as anode material for direct methanol fuel cell. Energy 2015;90:1144e51. [8] Tafaoli-Masoule M, Bahrami A, Elsayed EM. Optimum design parameters and operating condition for maximum power of a direct methanol fuel cell using analytical model and genetic algorithm. Energy 2014;70:643e52. [9] Na YS, Kwon JM, Kim H, Cho HJ, Song I. Characteristics of a direct methanol fuel cell system with the time shared fuel supplying approach. Energy 2013;50:406e11. [10] Abdelkareem Mohammad Ali, Yoshitoshi Tsukasa, Tsujiguchi Takuya. Verical operation of passive direct methanol fuel cell employing a porous carbon plate. J Power Sources 2010;195:1821e8. [11] Park YC, Chippar P, Kim SK, Lim SY, Jung DH, Ju H, et al. Effects of serpentine flow-field designs with different channel and rib widths on the performance of a direct methanol fuel cell. J Power Sources 2012;205:32e47. [12] Wong CW, Zhao TS, Ye Q, Liu JG. Experimental investigations of the anode flow field of a micro direct methanol fuel cell. J Power Sources 2006;155: 291e6. [13] Xu C, Zhao TS. A new flow field design for polymer electrolyte-based fuel cells. Electrochem Commun 2007;9:497e503.