ELSEVIER
`
`Chemical Engineering Journal 85 (2002) 137-146
`
`Chemical
`Engineering
`Journal
`www.elsevier.com/locate/cej
`
`The role of electrochemistry and electrochemical
`technology in environmental protection
`
`L.J .J. Janssen, L. Koene*
`Department of Chemical Engineering and Chemislly, Electrochemical Technology, Process Development Group,
`Eindhoven University of Technology, PO Box 513, NL-5600 MB Eindhoven, The Netherlands
`
`Received 27 November 2000; accepted 2 March 200 I
`
`Abstract
`
`For decades electrochemical technology has contributed successfully to environmental protection. In addition, environmental electro(cid:173)
`chemistry is capable of giving a huge contribution to further reach this goal by (1) application and improvement of existing technology,
`and (2) research, development and implementation of new technology. There is a tendency in the metal finishing and metal processing
`industries to redesign processes as closed loop systems. Therefore, we think especially the purification of dilute heavy metal ion-containing
`process liquids needs much of our attention. Electrochemical processes relevant to this purpose are considered and examples of (possible)
`industrial applications are discussed. Moreover, a critical evaluation of the usefulness of different types of electrodes is given. Generally,
`the two-dimensional electrode is the most favourable electrode configuration: the static flat electrode at high flow rates and in combination
`with mass transfer promoters, and the rotating disc electrode at high peripheral velocity. In both cases turbulent flow condition at the
`electrode are preferred to enhance the mass transfer. For redox systems where both the reactant and the product are soluble, the porous
`electrode is favourable due to its high specific surface area. © 2002 Elsevier Science B.V. All rights reserved.
`
`Keywords: Electrochemical processes; Regeneration of metal; Effluent treatment; Ion exchange assisted electrodialysis; Electroplating and surface finishing
`industry
`
`1. Introduction
`
`Electrochemical technology contributes in many ways to
`a cleaner environment and covers a very broad range of
`technology. Examples include generation of energy using
`(new) batteries [I] and fuel cells [2], selective synthesis of
`(organic) chemicals [3], removal of impurities from process
`liquids, air and soil, recycling of process streams [4] (e.g.
`from printed circuit industry) and sensors [SJ.
`During the last two decades, a special research field, viz.
`environmental electrochemistry has been developed. Envi(cid:173)
`ronmental electrochemistry [6-17] involves electrochemical
`techniques or methods to remove impurities from gases, liq(cid:173)
`uids and soil to prevent or minimise environmental pollution.
`In particular, emission to the atmosphere, discharges of
`pollutant into waters and disposal of solids to land sites
`have to be minimised. The best way to attack the problems
`is at their source, viz. mainly processes in the chemical
`and related industries. Environmental electrochemistry can
`contribute a great deal in reaching this goal. Especially, the
`purification of dilute heavy metal process liquids has been
`
`• Corresponding author. Tel.: +31-40-247-3784; fax: +31-40-345-3762.
`E-mail address: 1.koene@tue.nl (L. Koene).
`
`exhaustively investigated, since many heavy metals are very
`toxic and cause great environmental damage. Moreover, the
`reuse of metals needs more attention to prevent the disposal
`of heavy metal solid compounds in land sites.
`The removal of heavy metals especially from dilute liquids
`is extensively discussed in this paper. Some attention is also
`paid to the purification of dilute gases and the removal of
`organics.
`Dilute solutions to be purified are characterised by the:
`
`• Nature of polluting inorganics (heavy metal ions) and or(cid:173)
`ganics.
`• Concentration range of polluting species; mostly lower
`than 1000 ppm.
`• Nature and the concentration of supporting electrolyte
`mostly lower than 1000 ppm and in some cases the con(cid:173)
`centration is some orders of magnitude higher.
`• Specific conductivity of solution; mostly lower than
`1 n-lm- 1.
`
`Electrochemical removal of inorganics is obtained by
`direct electrolysis where only the reduction reaction is of
`interest, electrodialysis, and ion exchange assisted electro(cid:173)
`dialysis. To remove organics electrochemically from dilute
`solutions indirect electrochemical oxidation may be applied
`
`1385-8947/02/$ - see front matter© 2002 Elsevier Science B.V. All rights reserved.
`PII: S 1 3 8 5 - 8 9 4 7 ( 0 I ) 0 0 2 I 8 - 2
`
`Tennant Company
`Exhibit 1021
`
`

`

`138
`
`L.J.J. Janssen, L. Koene/Chemical Engineering Journal 85 (2002) 137-146
`
`Cmin
`Co
`
`Cs
`
`D
`
`E
`f
`F
`
`Nomenclature
`specific electrode area (A/V) (m- 1)
`electrode area (m2)
`bulk concentration of selected species
`(molm-3)
`inlet concentration of selected species
`(molm-3)
`minimal concentration (moJm-3)
`outlet concentration of selected species
`(molm-3)
`concentration of selected species at
`electrode surface (mo! m-3)
`diffusion coefficient of selected species
`(m2s-I)
`electrode potential (V)
`rotation rate of the electrode (rps)
`constant of Faraday (96 485 C mo1- 1)
`current density (A m-2)
`typical minimum effective current density
`(Am-2)
`limiting current density (A m-2
`)
`mass transfer coefficient (ms- 1)
`standard reaction rate constant (m s- 1)
`apparent reaction rate constant (-)
`constant (1 mol 1-"•PP m3"'•PP-3 )
`depth of a 3D-electrode (m)
`molecular mass (kg mot- 1)
`number of electrons in electrode reaction
`(-)
`normalised space velocity (s- 1)
`superficial liquid velocity (m s- 1)
`cell volume (m3)
`charge of cation (-)
`charge of anion (-)
`
`ieff, min
`
`n
`
`Sn
`Us= Us
`V
`
`ON
`µ,
`
`Greek letters
`apparent reaction order(-)
`aapp
`enhancement factor of the mass transfer
`y
`(-)
`thickness of Nernst diffusion layer (m)
`mobility of ion in the bulk solution
`(m2y-ls- 1)
`spacetime yield (kgm-3 s- 1)
`current efficiency
`current efficiency for anion
`current efficiency for cation
`
`p
`</J
`</Jan
`</Jcat
`
`• No cheap and inert electrode material is available, e.g.
`Pb corrodes slowly, and then the water will be poisoned
`by Pb-ions. Recently, barium-doped diamond electrodes
`[18] have been used; it seems these are very inert, but
`expensive.
`• Charge per molecule organic species is very high.
`• Cheap alternatives are available, e.g. adsorption by active
`carbon and thereafter burning and oxidation by H2O2 or
`03 in or not in combination with UV.
`
`Dilute gases contain small quantities of pollutants
`( 500-5 000 ppm). Moreover, many pollutants poisoning the
`air, for instance CO, NOx, NI-13 and I-12S are very difficult
`to oxidise or reduce on an electrode.
`The main purpose of this paper is a critical review of
`the real possibilities of electrochemistry and electrochemi(cid:173)
`cal technology to remove heavy metals from waste liquids,
`especially those from electroplating and surface finishing
`industry. First, direct electrolysis (Section 2) is extensively
`discussed with its relevant process parameters. In this sec(cid:173)
`tion results from a theoretical one-dimensional reactor model
`are discussed (Section 2.3). In addition, a review is given of
`different three-dimensional and two-dimensional electrodes
`(Section 2.4). In the next sections the following interesting
`electrochemical processes are discussed: indirect electroly(cid:173)
`sis (Section 3), electrodialysis (Section 4), and ion exchange
`assisted electrodialysis: a hybrid process (Section 5).
`
`2. Direct electrolysis
`
`The key parameters for an electrolysis in the removal of
`impurities are the current density i, current efficiency¢, and
`spacetime yield ( or normalised space velocity). The cell volt(cid:173)
`age is generally not a dominant parameter in environmental
`electrochemistry. To apply direct electrolysis in the indus(cid:173)
`try, the effective current density ieff must be higher than a
`fixed minimum, mainly determined by investment costs. A
`rough-and-ready ieff will be larger than l0Am-2 , where
`ieff = </Ji and i is based on the electrode surface area pro(cid:173)
`jected perpendicularly to the direction of the electrical cur(cid:173)
`rent through the solution.
`The effectiveness of electrolysis for dilute solutions is
`strongly determined by the mass transfer of the species to
`be removed, the effective electrode surface area and the oc(cid:173)
`currence of electrode side reactions.
`
`2.1. Mass transfer
`
`in some particular cases. Direct electrochemical oxidation
`methods are industrially of very little or no interest. The
`reasons for their uselessness can be:
`
`• Oxidation of organics at an electrode to CO2 and H2O
`is mostly a very slow electrochemical process and prac(cid:173)
`tically impossible.
`
`Liquid convection, diffusion and migration can determine
`mass transfer. The liquid convection directed parallel to the
`electrode surface determines the thickness of the Nemst
`diffusion layer 8N; outside this layer the concentrations of
`species can be assumed to be constant. A schematic repre(cid:173)
`sentation of a concentration profile of a species reacting on
`the electrode surface is given in Fig. I.
`
`

`

`L.J.J Janssen, L. Koene / Chemical Engineering Journal 85 (2002) 137-146
`
`139
`
`c't
`
`Cb----,-----.:;,-,.----(cid:173)
`,,
`"
`,''
`'
`'
`
`o----------
`x(cid:157)
`
`0
`
`Fig. I. A schematic representation of a concentration profile of a species
`reacting on the electrode surface.
`
`If the concentration of the supporting electrolyte is a factor
`IO higher than the concentration of the selected species to
`be oxidised or reduced, the migration of the selected species
`can be neglected. Its transport rate is then determined by
`diffusion and in stationary state given by Fick's first law:
`
`i = nFD (de)
`dx x=O
`
`or
`
`i =nFDC\~Cs)
`
`or
`
`(])
`
`(2)
`
`(3)
`
`where km= D/oN.
`The limiting diffusion current density is
`
`Cb
`iL = nFD- = nFkmlb
`ON
`In the absence of suppo1iing electrolyte migration as well
`as diffusion determine the transport of selected species. For
`solutions containing exclusively a I: I valency electrolyte
`the mass transfer coefficient of the cation is given by km =
`
`(4)
`
`2D /ON. Generally, km = y D / ON, where y > 1 and y de(cid:173)
`pends on the composition of solution. For a zA:z3 electrolyte
`in the absence of supporting electrolyte for the migration of
`the cation the factor y is given by [ 19]
`
`y =I+ 1::1
`
`(5)
`
`where ZA is the charge of the cation and z8 the charge of the
`anion as explained in Vetter [ 19). For instance, assuming a
`dilute nickel sulphate solution contains Ni OH+ and HSO4 -
`species only, it follows that y = 2.
`Very different configurations of electrodes are applied in
`environmental electrochemistry. They can be divided into
`two groups, viz. the two-dimensional (2D-electrode) and
`the three-dimensional (3D-electrode) electrode. These elec(cid:173)
`trodes can be used under static as well as moving conditions.
`ln addition, turbulence promoters are used, in particular at
`the static 2D-electrodes. The mass transfer rate constant km
`is also determined by the solution convection caused by, e.g.
`forced solution flow through or along the electrode, rotation
`of electrode and gas bubble evolution on electrode.
`Interesting electrode configurations in the environmental
`electrochemistry are given in Table I. To norm the mass
`transfer rate constant km, its value is given for a 1 x 1 x
`0.01 m 3 cell with a projected electrode surface of 1 x 1 m 2 .
`The flat-plate electrodes are placed in the I x 1 planes at a
`distance of 0.0 I m and the 3D-electrode fills the whole cell.
`Estimated km values (8, 10, 13,20- 24] are given for workable
`high mass transfer rate conditions. Moreover, for each elec(cid:173)
`trode configuration a minimum concentration, Cmin, where
`the electrolysis can be carried out on an industrial scale, is
`shown in Table I. The minimum concentration Cmi n is based
`on km given in this table, ieff, mi n is IO A m-2 and 11 = 2.
`The current density ieff, min is a typical minimum effective
`value proposed to use in industry.
`Mass transfer to a membrane is similar to mass transfer
`to an electrode. This situation occurs in conventional elec(cid:173)
`trodialysis cells. This type of electrochemical technology
`often applied in industry and is well described in (8, IO).
`An electrodialysis stack containing a great number of anion
`
`Table I
`Interesting electrode configurations and corresponding transfer coefficients and minimum concentrations at industrially accepted conditions [8, 10,13,20-24)"
`
`Electrode
`
`Flat-plate electrode
`RCEb
`
`High-porosity 3D-electrode
`Porous electrode (RVC; I 00 ppi)
`
`High mass transfer
`conditions (m s- 1)
`Ils = I
`U,= 10
`
`11 5 = 0.[
`
`Low-porosity 3D-electrode
`11 5 = 0.1
`Packed bed electrode (2 mm particles)'
`11 5 = 0.01
`Fluidised bed electrode (0.5 mm beads)
`' Cmin is based on km given. ielT. min is IO A m- 2 and 11 = 2.
`b Diameter and height 0.5 m.
`' Rods with 1 mm diameter and 1.8 mm length.
`
`km (m s- t)
`
`Cmin (mol m- 3)
`
`l x 10-5
`I x 10- 4
`
`I x 10- 2
`
`2 X 10-4
`6 X 10-J
`
`5
`5 X ]0-I
`
`5 X 10-J
`
`5 X 10- 4
`I x 10- 2
`
`

`

`140
`
`L.J.J. Janssen, L. Koene I Chemical Engineering Journal 85 (2 002) J 3 7-146
`
`and cation membrane pairs is common. The concentration
`of supporting electrolyte must be low. This means that the
`solution practically contains only the cation to be removed
`and anions. The rate of mass transfer to a membrane is com(cid:173)
`parable to that of a flat-plate electrode.
`
`2.2. Spacetime yield and normalised space velocity
`
`An important process design parameter, correlated with
`the specific investment costs, is the spacetime yield p of an
`electrochemical cell [25]:
`i</>M
`P = Ge nF
`
`(6)
`
`in which ae is the specific electrode area (A/V), i the current
`density, <f> the current efficiency, Mthe molecular mass, and
`n the number of electrons in electrode reaction. Combining
`Eq. (6) and the formula for the limiting current density,
`Eq. (4), yields an important formula for the design of an
`electrochemical reactor for the purification of wastewater:
`
`(7)
`
`This formula for electrochemical reactor design shows that
`for a certain metal concentration Cb the product of spe(cid:173)
`cific electrode area ae and mass transfer coefficient km
`should be as high as possible. It should be noted, how(cid:173)
`ever, that the spacetime yield of a reactor is dependent of
`the wastewater properties. Therefore, Kreysa [26,27] intro(cid:173)
`duced the normalised space velocity to characterise reactor
`performance:
`
`Sn = ----- log
`l<f>
`(Cj - c0 ) VnF
`
`(Ci)
`-
`c0
`
`(8)
`
`where c; and c0 are the concentrations of the reactant at,
`respectively, the inlet and the outlet of the cell. This figure
`of merit means the volume of wastewater (m3) for which the
`concentration of the key reactant can be reduced by a factor
`10 (c0 = 0. lCj) during 1 s in a reactor volume of 1 m3.
`
`2.3. Potential and current density distribution
`
`In cells with two flat-plate electrodes the electrodes are
`normally placed parallel to each other, while in cells contain(cid:173)
`ing a rotating electrode a static counter cylinder electrode
`is concentrically placed around the working electrode. This
`means that under industrial conditions, where the bulk of
`solution has a practically uniform composition, the potential
`and current distributions over each electrode are unifonn.
`For a 3D-electrode, the potential E and the current density
`i change continuously with increasing depth of electrode Le.
`The potential and the current density are strongly related to
`each other. The E - Le and i - Le curves depend on many
`parameters.
`Portegies Zwart [20] has calculated the potential and
`dimensionless current density profiles in a packed bed elec(cid:173)
`trode for the Fe3+ reduction, where its rate is determined
`
`by both mass transfer and kinetic parameters. He has used
`a theoretical one-dimensional reactor model. Moreover,
`he has calculated the total dimensionless cathodic current
`density as a function of a great number of parameters, viz.
`three-dimensional cathode thickness, standard electrochem(cid:173)
`ical rate constant ks, mass transfer coefficient, potential
`driving force, apparent exchange current density, and con(cid:173)
`centration levels of Fe2+ and Fe3+. There are two extremes,
`viz. the one where potential and current density profiles are
`determined by the kinetic parameters as well as the spe(cid:173)
`cific resistance of solution, and the other by the local mass
`transfer as well as the specific resistance of solution. It has
`been found that the solution conductivity particularly plays
`a very important role and strongly determines the effective
`thickness of a three-dimensional roll if the electrochemical
`kinetic parameters are of interest.
`Experiments were carried out to reduce ferric ions in a sul(cid:173)
`phuric acid solution using an undivided GBC-reactor (Gas
`diffusion electrode packed Bed electrode Cell) consisting of
`a gas diffusion anode fed by hydrogen and a packed bed
`cathode of small carbon rods [20,28] (see Fig. 2).
`The reduction reaction of ferric at the packed bed elec(cid:173)
`trode is
`
`(9)
`
`and the oxidation reaction of hydrogen at the hydrogen gas
`diffusion anode is
`
`(10)
`
`The macroscopic reduction rate in the GBC-reactor can be
`described using an empirical reaction rate equation with
`apparent reaction parameters:
`r = KkappCaapp
`
`(11)
`
`This equation combines the microscopic effects of mass
`transfer and electrochemical kinetics on the reduction
`
`porous backing
`gas diffusion
`anode
`
`active Jaycr gas
`diffusion nno&
`
`porous cathode
`
`D liquid
`
`electrolyte flow
`
`X
`
`Fig. 2. Schematic presentation of the .divided GBC-reactor configuration.
`
`

`

`L.J.J. Janssen, L. Koene/Chemical Engineering Journal 85 (2002) 137-146
`
`141
`
`1.0
`
`0.8
`
`0.6
`
`I ---0.
`
`~
`0.4
`Ci
`
`0.2
`
`II
`
`a
`
`C
`
`I
`
`! III
`:
`:N
`
`0·Po- 1'10-13I0·1'J.0· 1110·1010·• 10·• 10·1 10-< 10-5 10-•
`k5 /m s· 1
`
`Fig. 3. Calculated apparent reaction order, aapp, as a function of the
`standard electrochemical reaction rate constant at different mass transfer
`rates. Roman numbers indicate different operating regimes (see text). km:
`10-7 (a), 10-5 (b), 10° (c) m s- 1 (km= kmo = kmRl-
`
`reaction into a macroscopic apparent reaction rate constant
`kapp and an apparent reaction order Uapp· The parameter
`K is introduced as a constant, which has a value of 1 and
`a dimension of mo1 1-aapp m 3"app-3, so that kapp has the
`standard dimension of m s- 1. Thus K only maintains the
`dimensional integrity of Eq. (I 1).
`Portegies Zwart [20] has linked the results for the
`one-dimensional reactor model with experimentally ob(cid:173)
`tained parameters using the aforementioned empirical re(cid:173)
`actor rate equation. He distinguished four zones in the
`Uapp - k5 figure (see Fig. 3) for aapp, viz. Uapp = I at
`ks < 10- 13 m s- 1, where the reduction process is kinetically
`controlled (zone I), 0.85 < Uapp < 1 at ks > 10- 13 m s- 1,
`where the reduction process is mainly detem1ined by mass
`transfer (zone II), 0.45 < aapp < 0.85atks > 10- 13 ms- 1,
`where the reduction process is determined by kinetics,
`mass transfer and ohmic effects (zone III), where aapp
`shows a ve1y complex behaviour, and aapp < 0.45 at ks >
`10- 13 m s- 1 (zone IV), where very high mass transfer co(cid:173)
`efficient and standard electrochemical rate constant occur.
`In zone IV, only the region of the 3D-electrode next to
`the gas diffusion anode is still active. It is very unlikely that
`this operating regime occurs in practice. For a kinetically
`limited process and a mass transfer limited process it is
`found that aapp = 1, while kapp is, respectively, equal to
`the electrochemical rate constant for the reduction reaction
`or to the mass transfer coefficient. It has been concluded
`that the operating regime of the reactor can be identified
`using the values of Uapp and kapp using the theoretical model.
`The model was used to optimise a GBC-reactor used for
`the regeneration of an electochemical machining electrolyte
`[20]. Apart from its apparent practical usefulness, this work
`serves as an illustration of how mathematical modelling can
`be used in electrochemical technology in general.
`The potential profile over a RVC-cathode (reticulated vit(cid:173)
`reous carbon) in a GBC-reactor at the reduction of a 1 M
`H2 S04 solution containing small concentrations of ferric
`and ferrous was detem1ined experimentally using a moving
`Luggin capillary. The potential profile was calculated using
`
`-0.40
`
`-0.45
`
`-- -0.50
`
`>
`2'
`
`-0.55
`
`• •
`75
`93
`l/mA
`U1/ms·1
`0.109 0.03
`c,..,/mo!m·' 11.33 !I.33
`c,...,/molm·1 8.67 8.67
`
`- ;
`i •
`
`0
`
`6i.o
`·
`
`·
`
`0.2
`
`o.4
`
`o.6
`
`o.s
`
`1.0
`
`102 XLuggin / ID
`
`Fig. 4. Experimental cathode electrode potential profiles measured at dif(cid:173)
`ferent reactor current levels, superficial liquid velocities and concentra(cid:173)
`tions of ferric and ferrous ions at T = 313 K. Drawn lines show results
`of numerical model simulations [20].
`
`the one-dimensional reactor model. The results are given in
`Fig. 4. The agreement between the experimental and theo(cid:173)
`retical potential profiles is reasonable [20].
`
`2.4. Different electrode con.figurations
`
`interesting electrode configurations,
`industrially
`The
`given in Table 1, are discussed: first 3D-electrodes and
`then 2D-electrodes. Their advantages, disadvantages, and
`possible applications are presented and some experimental
`results are given. The authors are limited by the scope of the
`article and, also, their knowledge about not all electrodes
`discussed is enough to be able to produce an elegant little
`essay on each. However, a critical review is given. For more
`detailed information and also for application in commercial
`reactors see for instance [7,8, 10,25].
`
`2.4.1. Fluidised bed electrode
`The fluidised bed electrode is a dynamic, JD-electrode. A
`schematic representation of a fluidised bed electrode reactor
`used in practice is given in [ 10]. AKZO Zout Chemie orig(cid:173)
`inally developed it and it was further developed by Billiton
`Research for the extraction and reclamation of metals.
`Advantages are a high mass transfer coefficient and the
`possibility of continuous removal of metal deposit from the
`cell. Disadvantages are a non-uniform potential distribution
`over the bed; dissolution of metal deposit in the unpro(cid:173)
`tected zones of the bed (where the electrode potential of the
`paiiicles rises to the corrosion potential); particle-paiiicle
`and particle current feeder agglomeration and fom1ation
`of inactive zones; and preferential growth near and on the
`separating membrane. Also, metal concentrations below
`2 mo! m-3 are not effectively treated because dissolved
`oxygen will sharply decrease the effectiveness of metal
`removal.
`An industrial application is removal of Cu2+ ions from
`high conductive process liquids [29]. Only metal-metal
`ion couples with Er » Er.H2 can produce a successful
`process.
`
`

`

`142
`
`L.J.J Janssen, L. Kaene I Chemical Engineering Journal 85 (1001) 13 7-146
`
`2.4.2. Low-porosity JD-electrodes
`The low-porosity 3D-electrode can be divided into packed
`bed and moving bed electrodes; in both types of electrodes
`each particle is continuously in contact with other particles
`of the bed. This is the essential difference between a fluidised
`bed electrode and a low-porosity 3D-electrode. In most cases
`the latter has a voidage of about 0.4-0.5.
`The electrically conductive particles of the low-porosity
`3D-electrode can have a broad range of geometry, for in(cid:173)
`stance granules, rods, spheroids, microspheres, and fibres.
`Two different configurations of electrode reactors with
`3D-electrodes are well known, viz. the flow-through reac(cid:173)
`tor, where the direction of electric current is parallel to
`the direction of the solution flow, and the flow-by reactor,
`where the direction of electric current is perpendicular to
`the direction of solution flow [30]. The latter configuration
`is the most appropriate one in the environmental electro(cid:173)
`chemistry. Schematic diagrams for cells with packed and
`moving bed electrodes are shown in [8]. The advantages,
`disadvantages and relevant industrial applications are given
`next.
`Advantages are a high mass transfer rate, a high degree of
`conversion, and the possibilities of simple and cheap elec(cid:173)
`trode material. Moreover, the effective bed thickness will be
`high at high solution conductivity. For reduction reactions,
`the whole bed can be kept at a potential lower than the cor(cid:173)
`rosion or equilibrium potential.
`Disadvantages are that the effective bed thickness is lim(cid:173)
`ited by solution conductivity and the deposition of solid
`material especially on the bed particles, in particular those
`closest to the counter electrode. In addition, the effective bed
`thickness is limited by the concentrations of the reactants
`and the deposition of solid material causes a 3D-electrode
`to behave like a 2D-electrode. Other disadvantages are:
`electrode dendrite formation on the bed particles closest to
`the counter electrode results in damage to the membrane
`or separator; the range of useful electrode materials is
`low; and that a high volumetric flow rate results in a high
`pressure-drop over the reactor.
`Relevant industrial applications are reduction reactions
`in which the reactant and product are soluble (Fe3+, 0 2,
`Croi-). Only in some specific cases metal ion removal
`and metal recovery may be carried out successfully and eco(cid:173)
`nomically. Unfortunately, the packed bed is generally not
`applicable in small galvanic plants.
`The reduction of ferric ions [31], nickel ions [32], and
`of chromate ions [33] in sulphuric acid solutions have been
`investigated for a GBC-reactor with a hydrogen gas diffusion
`anode and a packed bed of graphite particles. It was found
`that the deposition of nickel in the graphite particle bed
`causes many problems.
`
`2.4.3. High-porosity JD-electrodes
`The high-porosity 3D-electrode has a very high voidage,
`larger than about 90%, and it can be composed of various
`materials, for instance, perforated plates, expanded metal
`
`meshes, felts and foams. Reticulated vitreous carbon (RVC)
`has been used very often because of its open structure; RVC
`grade I 00 ppi has a voidage of about 97%. High voidage is
`important as the effect of the specific conductivity of solu(cid:173)
`tion is much smaller for a high-porosity electrode than for a
`low-porosity electrode. Advantages, disadvantages and pos(cid:173)
`sible applications of high-porosity electrodes are compara(cid:173)
`ble to those for low-porosity electrodes.
`Carbon felt is a porous electrode material used in a number
`of commercial cells [8]. This material has the advantage of
`a high mass transfer coefficient. However, it is sensitive to
`plugging. It is applied in reactors to remove metal and for
`the electrochemical oxidations of organic waste.
`The reduction of ferricyanide ions has been carried out
`successfully at a RVC electrode in a I M KOH solution con(cid:173)
`taining 2 mo! m-3 ferricyanide [20]. It was found that under
`these conditions mass transfer determines the reduction of
`ferricyanide.
`The reduction of chromate in a I M H2SO4 solution
`containing 0.4molm-3 chromate on a RVC electrode in a
`short-circuited GBC-reactor with a hydrogen gas diffusion
`anode is determined by kinetic parameters [32]. It was
`found that gold metal is a more attractive cathode material
`for chromate reduction [20]. A suggestion has been made
`to cover the surface of RVC electrode with a thin layer of
`gold metal (i.e. I µm) [20].
`Continuous research has indicated, however, that a large
`gold electrode was clearly etched in a GBC-reactor fed by
`rinse water from a chrome plating bath (Cr-concentration
`was about 0.9 mo! dm-3). Due to the corrosion of gold under
`these conditions gold is not useful as cathode material. A
`comparative study on the reduction of chromate has shown
`that H2O2 can be used very effectively to reduce chromate
`in very dilute chromate solutions. In addition, it will be
`economically profitable.
`Scale up and design optimisation of a GBC-reactor with
`a RVC cathode and a hydrogen gas diffusion anode were
`carried out by numerical calculations for electrochemical re(cid:173)
`generation of an ECM electrolyte containing 3M NaNO3 +
`small quantities of Cr(VI), Cr(III) and Fe(III). This solu(cid:173)
`tion is used in the processing of an iron-chromium alloy
`[20].
`
`2.4.4. Two-dimensional rotating electrode
`The removal of metal via cathodic deposition is carried
`out industrially with a reactor containing a rotating cylinder
`electrode (RCE). This is usually comprised of an inner ro(cid:173)
`tating cathode in turbulent flow and a concentrically placed
`cylindrical counter electrode. Examples of cells are given in
`[8,10]. The advantages, disadvantages, and relevant indus(cid:173)
`trial applications are as follows.
`Advantages are a practically uniform potential and cur(cid:173)
`rent density distribution and high mass transfer rates. These
`rates are enhanced by rough metal crystallites acting as
`micro-turbulence promoters. A typical enhancement factor
`for mass transfer is about 10 under industrial conditions (see
`
`

`

`L.J.J. Janssen, L. Koene / Chemica/ Engineering Journal 85 (2002) 137-146
`
`143
`
`Walsh in [8]). The mass transfer coefficient can be controlled
`by rotation rate, independent ofliquid flow. In addition, there
`is a small gap between anode and cathode, which also re(cid:173)
`duces the cell volume. The continuous removal of deposit
`is possible. Cascade reactors with cells containing only one
`RCE have been developed.
`Disadvantages are that the fractional conversion per pass
`in a single reactor is normally much less than 0.6 and contin(cid:173)
`uous removal of deposit (metal and/or metal hydroxide) from
`the cylinder leaves particles in the solution and results in the
`re-dissolution of metal. Besides, hydroxide formation can
`occur for some metal ions, e.g. Ni. The diameter of the cylin(cid:173)
`der is limited, particularly for a divided concentric reactor.
`Relevant industrial applications are redox couples with
`soluble compounds, e.g. fe3+, CrO42-, for which it is very
`useful, and noble metal deposition from dilute solutions: Cu
`from 0.5 M H2SO4 [10], Ag from photographic bath [ 10].
`
`2.4.5. Two-dimensional flat-plate electrode
`The flat-plate electrode is the simplest configuration
`and is used in the well-known and simple filterpress cells,
`mostly with forced convection flow. To enhance mass trans(cid:173)
`fer plastic mesh and fluidised glass beads are used. The ad(cid:173)
`vantages, disadvantages and relevant possible applications
`are indicated.
`Advantages are that it is easy to build in the well-known
`and simple filterpress reactor as a three-compartment reac(cid:173)
`tor; a practically uniform distribution of potential and cur(cid:173)
`rent density; easy periodic removal of cathodes covered with
`a metal deposit from an open cathodic cell compartment;
`and reuse of cathode after removal of metal deposit. In ad(cid:173)
`dition, simple turbulence promoters (plastic mesh, fluidi sed
`inert particles) can be used: an enhancement factor for mass
`transfer can be about a factor 10.
`Disadvantages include a relatively low mass transfer
`coefficient, the minimum concentration to be achieved is
`relatively high, viz. about 5 mo! m- 3, and formation of hy(cid:173)
`droxides in solutions with typical metal ions, e.g. Ni. The
`conversion factor per pass for a single reactor is low and
`strongly related to the volumetric liquid flow rate though
`the reactor.
`A possible industrial application is to keep process liq(cid:173)
`uids on a low heavy metal ion concentration of about
`2-10 mo! m- 3 by re-circulation of liquid through the reac(cid:173)
`tor. In some cases a metal hydroxide deposit is formed, its
`adhesion to the electrode is weak.
`The minimum attainable concentration strongly depends
`on the nature of the heavy metal cation to be removed, the
`composition of the supporting electrolyte and in particular
`on the pH and the presence of complexing ions.
`The flat-plate reactor is not used on a large scale, also not
`in the traditional galvanic plating factories. This is probably
`due to the use of closed reactors, where the removal of
`the deposit is very time and labour intensive. To make this
`reactor more accessible an open cathode compartment in the
`filterpress cell