STUDY OF THE ELECTRICAL CONDUCTIVITY OF GRAPHITE FELT EMPLOYED AS A POROUS ELECTRODE

STUDY OF THE ELECTRICAL CONDUCTIVITY OF GRAPHITE FELT EMPLOYED AS A POROUS ELECTRODE

E.O. Vilar^*, N.L. de Freitas, F.R. de Lirio and F.B. de Sousa

UFPB/CCT- Dep. Eng. Química , Av. Aprigio Veloso, 882, Cid.Universitária, 58100-000 Campina Grande PB. Phone [phone] . Email : [e-mail]

(Received: October 20, 1997; Accepted: August 4, 1998)

Abstract - The objective of the present work is to study the variation of the electrode distribution potential under electrical conductivity variation of graphite felt RVG 4000 ( Le Carbone Lorraine) when submitted to a mechanical compression. Experimental and theoretical studies show that this electrical conductivity variation can changes the electrode potential distribution E(x) working under limiting current conditions. This may occur when graphite felt is confined in an electrochemical reactor compartment or simply when it is submitted to a force performed by an electrolyte percolation in a turbulent flow. This investigation can contribute to the improvement of electrochemical cells that may use this material as an electrode. Finally, one modification is suggested in the equation that gives the electrode potential distribution E(x) - E(0). In this case the parameter L (thickness in metal porous electrodes) is substituted for Lf = Li (1-j), where j corresponds to the reduction factor of the initial thickness Li.

Keywords: Electrical conductivity, graphite felt, electrode potential distribution.

INTRODUCTION

For effluent treatment (removing noble metals or pollutants) [Tricoli et al (1993), Oren and Soffer(1983)] or in organic electro-organic synthesis, the basic requirements for materials acting as an efficient electrode can be described as follows:

- High chemical stability.

- Large specific surface area.

- Good fluid permeability.

- High electrical conductivity and structural continuity.

- Cheap and easily available.

Among the materials that are suitable for all these requirements, we can mention some that are conductors of electricity and that are made of carbon fiber, for example: carbon or graphite felt, tissues and RVC (Reticulated Vitreous Carbon). Large specific area and permeability that result from the high interstitial space between the fibers characterize these materials. They are chemically inert and they have very good electrical conductivity. Some articles consider graphite felt more adequate owing to its high permeability because of a more random distribution of the fibers when compared to a graphite tissue [Oren and Soffer (1983)]. Others consider RVC [Strohl and Curran (1979); Wang (1981)] an efficient electrode material, especially when it is used as an analytical sensor. Seemingly, its industrial utilization is restricted to specified assembling and operating conditions.

Nowadays, several research laboratories use graphite felt as an efficient working electrode (cathode) because of its physical-chemistry properties. Since this material is easily deformable, it is not considered that this property may affect the perfomace of an electrochemical cell.

The influence of electrode electrical conductivity (s ) and electrolyte solution (k) on the potential distribution E(x) was studied theoretically and experimentally by Gaunand and Coeuret (1978), making use of a fixed-bed electrode made of graphite microspheres. In this case, (s ) depends on granular contact resistance only, that in turn depends on the function of the arrangement of these particles in the bed.

In the present work, the variation of the (k/s ) ratio will depend on graphite fiber bed compactation or degree of deformation . This subject is important, in some users, such as electroreduction of metal ions present in diluted solution or in organic electrosynthesis. In these cases, the potential distribution may influence the reaction selectivity involved in a given process.

THEORETICAL ASPECTS

Consider a porous electrode, as shown in Figure 1, under a diffusive control mode. It is known that current density variations, of electrode di_e and of solution di_s in the axial electrode direction, are given in equation 1 as follows [Coeuret and Storck (1984)]:

Figure 1
: Distribution of the ion specie concentration A and current density of solution is.

(1)

where ` kd corresponds to the average mass transfer coefficient, n _e corresponds to the electrons involved in the reaction, Á is the Faraday constant, ae is the specific area per unit total volume (empty volume + solid volume ), and C_A (x) is the concentration distribution of the active species of ions through the X coordinate.

Potential Distribution

Considering Ohm’s Law (i_s = - kÑ j _s e i_e = - s Ñ j _e), the potential distribution in the x direction dE(x) is given by the difference between the porous matrix potential j _e and the solution potential j _s that occupy the empty spaces of this matrix, or:

(2)

where k and s correspond to the electrical conductivity of the electrolyte and porous matrix, respectively, and L is the thickness of matrix.

It is known that from a mass balance in the differential element of volume the concentration distribution Ca(x) is expressed as follows in equation (3):

(3)

where V is the percolation average speed of the electrolyte, and CAE is the initial concentration of the electrochemical species. Combining the equations 1, 2 and 3 we have the following 2nd class differential equation:

(4)

The solution of this differential equation [Coeuret and Storck (1984)] gives the potential distribution E(x) through the porous matrix . The boundary conditions for solving equation 4 are the following:

The solution for equation 4 is consequently:

(5)

where XA corresponds to the fractional conversion of the ions, and it = ie + is is the total current density in the system. This equation shows that the potential distribution E(x) is a function of the electrical conductivity of electrolyte solution K and of the electrode electrical conductivity s .

For infinitely conducting metal matrix (s >>k and k/s ® 0) the solution of equation 4 is simplified in equation 6 below:

(6)

For metallic electrodes, the value of (s ) remain constant and creates an equipotential matrix. The present work considers the relation ( k/s ) starting from two basic assumptions:

(1) For porous carbon electrodes, as in the case of materials made of fiber graphite, k/s ¹ 0.

(2) The electrode will not be used for metal electrodeposition in which ( s ) increases progressively still maintaining the condition of k/s tending to 0.

Additionally the thickness ( L ) in equation 5 must be changed (in the case of compressible porous electrodes ), by equation 7 :

L_f = Li ( 1 - j ) (7)

where L_f corresponds to the final thickness, L_i is the initial thickness and j the reduction factor 0 < j < 1.

EXPERIMENTAL METHODS

Figure 2 shows the experimental setup diagram used to study the electrical conductivity of graphite felt. The procedure employed for this verification was developed using graphite felt (RVG 4000, average porosity` e = 96%, thickness L_i =1.1 cm, higher purity > 99% of carbon, electrical conductivity s =2.85 Ohm^-1m^-1, density = 2.25kg.m^-3) placed between two nickel plates (chosen because of high electrical conductivity of nickel so that it does not interfere with the measurements). It was used a Testometric Micro 350, a power supply (5mA DC) and a plotter CG-Scientific Instruments. The experiments were performed at a temperature of 26 + 1ºC.

Figure 2
: Esquematic diagram of experimental device : 1- Plotter; 2- Power Supply DC; 3-Testing compartment, 4-Compression device-Testometric Micro 350.

Figure 3: a. Graphite felt situated between two nickel plates (A=25x10 ^-4m²). b. Graphite felt located between two nickel discs and confined to a plastic cylinder (A=12.6x10^-4 m²).

Graphite felt was investigated using two configurations as shown in Figures 3a and 3b. In both case, the graphite felt was submitted to a constant compression rate (0.5mm/min) and, simultaneously, changing in the electrical resistance R (Ohm) was recorded as a function of time and compression force (kgf) applied by the testing device. The plotter was calibrated to a minimum resistance (R) of 0.0 Ohm and a maximum of 250 Ohm.

In the first case (Figure 3a- unconfined case), lateral deformation of the material is permitted as a result of compression (this case is similar to axial pressure applied by a liquid in a turbulent regime). In the second case ( Figure 3b-confined ), graphite felt was placed inside of a rigid polymer tube. In this case, it was create additional inner tensions in the porous matrix due to compression forces from tube walls.

To verify the influence of the electrical conductivity variation of this material over the potential distribution, it has been done theoretical comparative calculations using a complete solution, equation 5, and a simplified solution, equation 6. Considering a conversion XA =0.9 and a typical conductivity of electrolyte in a porous matrix k = 10 Ohm^-1m^-1 (NaOH, 1.0N), these equations (5 and 6) were simplified and expressed as a new function Z(x), as giving in equation 8 [Gaunand and Coeuret (1978)]:

(8)

If we compare the equation (8) with (5) and (6), we can verify that Z(x) depends only on the relations k/s and XA.

The variation of the material conductivity (s ) was determined by as a function of the applied compression force [N.m^-2].

The calculation of the average porosity variation with the compression was done by means of the expression:

(9)

where mgraphite corresponds to the electrode mass, r graphite is its density, L and A are thickness and transversal area, respectively.

RESULTS AND DISCUSSION

Conductivity Variation s

Experimental results have confirmed the hypothesis that for small mechanical compression variations in the material, non-negligible variations in the electric conductivity will exist. Typical results are shown in table 1 and figures 4 and 5.

In both situations the results show a significant increasing in the initial electrical conductivity (s = 2.85 Ohm^-1m^-1) with mechanical compression. In the confined graphite felt case, this increasing is approximately 5.5 times higher than the unconfined graphite felt case for almost the same compressive force. This fact can be explained by the greater compactness reached by fiber carbon and consequently increasing in the electrical contact points. It observed in both cases that in the beginning of the experiments, the electrical conductivity is higher than 2.85 Ohm^-1m^-1. This occurs due to a slight initial compression suffered by the material between the two nickel plates. This compression is necessary for an adequate electric contact.

Table 1
: Electrical conductivity variation of the graphite felt with applied compression force

Appl. Force ( 103 N/m² )	s n ( Ohm-1m-1 )	Appl. Force ( 103 N/m² )	s c ( Ohm-1m-1 )
1,2	6,4	6,8	60,2
3,5	8,7	10,2	78,1
7,0	11,5	17,0	98,0
11,8	15,2	23,7	117,6
17,2	18,4	34,0	142,8
25,1	21,3	51,0	172,4
36,1	25,0	71,3	208,3
51,0	28,6	101,8	250,0

s _n - Unconfied ; _{s c} – Confined

Figure 4:
Electrical conductivity variation s n of graphite felt with compression for unconfined electrode.

Figure 5:
Electrical conductivity variation s c of graphite felt with compression for confined electrode.

Variation of the Function Z(x)

For the unconfined graphite felt case, Figure 6 shows a variation of the function to a reduced (x/L) coordinate.

In this case, it can be seen that for a matrix much more conductive than electrolyte (k/s ® 0), Z(x) becomes more negative ( E(x) goes to the more cathodic values) as (x/L) increases from 0 to 1. For higher (k/s ) values Z(x) remains positive for every (x/L). This means that part of the electrode, which is situated between E(0), and E(L) (see Figure 1) not work properly under a diffusive control mode.

Figure 7 shows, for the confined graphite felt case, that the Z(x) potential distribution is emphatically more negative because of a greater electric conductivity of the electrode (s ). In this case there are more parts of the porous matrix that operate at cathodic potentials in the range between the potential E(0) and E(1) and is possible to guarantee the operation of the whole porous electrode at the limiting diffusion current.

Figure 6: Z(x) variation for unconfined electrode position. s n =6.4; 15.2 and 28.6 Ohm^-1 m^-1.

Figure 7: Z(x) variation for confined electrode position. s c =60.2 ; 98.0 and 250.0 Ohm^-1 m^-1.

Figure 8: Theoretically average porosity variation as a function of the applied compression force.

Average Porosity Variation

Figure 8 shows the initial porosity variation of the graphite felt (96%) as a function of the applied compression force. It can be verified that for a 27% reduction of its initial thickness, there was only a decrease of 2% of the average porosity approximately. This means that in spite of the electrode volume reduction of this material (associated with the improvement of its electrical conductivity), the effect on its permeability, when the fluid runs on is very small.

CONCLUSION

For the confined graphite case, the potential distribution E(x) remains cathodic for almost of the electrode. This can theoretically guarantee an operation under a more efficient and desirable diffusive control mode. Moreover, it will be more desirable for this material to operate under more elevated electrical conductivity conditions, in which can be reached through an adequate reactor design that compresses the material. It was observed that compactness does not affect its permeability.

NOMENCLATURE

A Surface area, m²

ae Specific area for volume unit, m^-1

CA Ion specie concentration, A [mol/m³]

CAE Ion specie entrance concentration, A [mol/m³].

E Potential distribution in the solution, V

Faraday constant, 96500 C/eqv

ie Electrode current density, A/m²

is Solution current density, A/m²

it Total current density, A/m²

j Reduction factor defined in equation (8)

k Electrolyte conductivity, Ohm^-1 m^-1

Average mass transfer coefficient, m/s

L Electrode thickness, m

Li Initial thickness of the electrode, m

Lf Final thickness of the electrode, m

m mass, kg

R Electrode electrical resistance, Ohm

x Coordinate

XA Conversion rate

Z Parameter defined in equation (7)

Percolation average speed, m/s

Greek letters

r Specific density, kg/m³

s Electrode electrical conductivity, Ohm^-1 m^-1

n _e Number of electrons

j _e Electrode potential, V

j _s Solution potential, V

Average porosity

• Coeuret, F., Introduccion a la Ingeniería Electroquimica, Ed. Reverté (1992).
• Coeuret, F. et Storck, A., Elements de Génie Electrochimique,Ed. TEC DOC Lavoisier (1984).
• Carta, R.; Palmas, S.; Polcaro, A. M. and Tola, G., J. of Appl. Electrochem., Vol. 21 (1991).
• Gaunand, A. and Coeuret, F., Eletrochimica Acta, Vol.23, pp.1197-1203 (1978).
• Ismail, M.I. (editor), Electrochemical Reactor, Ed. Elsevier (1989).
• Oren, Y. and Soffer, A., Electrochimca Acta, Vol. 28, 1649 (1983).
• Strohl, A.N. and Curran, D.J., Analytical Chemistry, Vol. 51, 7 (1979).
• Tricoli,V. ;Vatistas, N. and P. F.Marconi,Journal of Applied Electrochemistry, Vol. 23 (1993).
• Vatista, N.; Marconi, P.F. and M. Bartolozzi, Electrochimica Acta, Vol. 36, 2 (1991).
• Walsh F., Electrochemical Engineering, Ed. ECC, Alresford, Hants (1993).
• Wang, J., Electrochimica Acta, Vol. 26, 12, pp.1721-1726 (1981).

* To whom correspondence should be addressed.

To whom correspondence should be addressed.

Publication Dates

• Publication in this collection
  27 Oct 1998
• Date of issue
  Sept 1998

History

• Received
  20 Oct 1997
• Accepted
  04 Aug 1998