United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Cincinnati OH 45268
EPA 600 2 79 181
August 1979
Research and Development
Coupled  Transport
Systems for
Control of Heavy
Metal Pollutants

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further deveJopment and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health  Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has  been assigned  to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate  instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution-sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                            EPA-600/2-79-181
                                            August 1979
         COUPLED TRANSPORT SYSTEMS

              FOR CONTROL OF

          HEAVY METAL POLLUTANTS
                    by

               W.C. Babcock
               R.W. Baker
               D.J. Kelly
               J.C. Klelber
               H.K. Lonsdale

            Bend Research, Inc.
            Bend, Oregon 97701
           Grant No, R804682-01
             Project Officer

             George F. Weesner
    Metals and Inorganic Chemicals Branch
Industrial Environmental Research Laboratory
          Cincinnati, Ohio 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
    OFFICE OF RESEARCH AND DEVELOPMENT
   U.S. ENVIRONMENTAL PROTECTION AGENCY
          CINCINNATI, OHIO 45268

-------
                                DISCLAIMER
     This report has been reviewed by the Industrial Environmental Research
Laboratory, Cincinnati, U.S. Environmental Protection Agency, and approved
for publication.  Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute endorse-
ment or recommendation for use.

-------
                                 FOREWORD
     When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on
our health often require that new and increasingly more efficient pollution
control methods be used.  The Industrial Environmental Research Laboratory-
Cincinnati (lERL-Ci) assists in developing and demonstrating new and im-
proved methodologies that will meet these needs both efficiently and
economically.

     The study described in this report was undertaken to demonstrate the
applicability of a new separation method to the recovery of metal values
from plating rinse waters.  The process is referred to as coupled transport.
In this brief, four-month study, it was demonstrated that coupled transport
membranes can be used to separate copper, nickel, and chromium from these
rinse waters and to "chemically pump" these metal ions against very large
concentration gradients.  High recoveries of metal values can thereby be
effected, producing recycle streams sufficiently concentrated for direct
return to the plating bath.  The economics of the process appear favorable
relative to existing technology.

     For further information on this subject, contact the Industrial
Pollution Control Division, Metals and Inorganic Chemicals Branch.
                                              David  G.  Stephan
                                                  Director
                                Industrial  Environmental  Research Laboratory
                                                 Cincinnati
                                     iii

-------
                                  ABSTRACT
     This report describes a process for separating and concentrating heavy
metals from electroplating rinse waters.  Metal ions can be "chemically
pumped" across a coupled transport membrane against large concentration
gradients by allowing the counterflow of a coupled ion such as hydrogen ion.
 The  process  is  carried  out within a microporous membrane containing
 within its pores an  organic, water immiscible  complexing agent.  The
 complexing agent acts as  a shuttle, picking up metal ions on one side
 of the membrane, carrying them across the membrane as a complex, and
 preserving electrical neutrality by carrying hydrogen ions in the opposite
 direction.

      The  importance  of  coupled transport is its high selectivity and flux.
 High selectivity derives  from the use of specific complexing agents.  High
 flux is possible because  these are actually liquid membranes with diffu-
 sivities  many times  greater than those  in solid membranes.
                                    iv

-------
                                CONTENTS

Foreword                                                             iii
Abstract                                                              iv
Acknowledgement                                                       vi

    1.  Introduction and Summary                                       1
    2.  Principles of Coupled Transport Membranes                      2
              Co-transport                                             2
              Counter Transport                                        4
              Applicability of Coupled Transport Membranes
                                        to Plating Wastes              5
    3.  Experimental                                                   7
              Distribution Coefficients                                7
              Membranes and Permeability Apparatus                     7
              Analytical Methods                                       9
              Reagents                                                 9
    4.  Results                                                       11
              Distribution Coefficients                               H
              Demonstration of Coupled Transport                      H
              Effect of Diluents on Flux                              15
              Simulation of Actual Plating Rinse Conditions           15
                          General Considerations                      15
                          Chromium                                    20
                          Copper and Nickel                           22
    5.  Economics                                                     29
    6.  Future Work                                                   31

Appendices
    A.  Steady State Concentration Factors                            32
    B.  Exponential Decay                                             34
                                   v

-------
                              ACKNOWLEDGMENT
     Our work on coupled transport membranes began under a contract with the
Bureau of Mines (Contract No. H0252066), in which we are examining the
hydrometallurgical recovery of copper and other metals from low grade ores.
Some of the data reported in Figures 4, 5, 6b, 7, and 9 of the present
report were obtained in the course of that ongoing, parallel work.
                                    VI

-------
                                  SECTION I

                           INTRODUCTION AND SUMMARY

     This is the final report on Phase  I of American Electroplaters' Society
Project No. 43, entitled "Coupled Transport Systems for Control of Heavy
Metal Pollutants in Metal Finishing Solutions".  It covers the period from
October 22, 1976 - January 21, 1977.

     The goal of this program is to demonstrate the applicability of coupled
transport membranes to the processing of electroplating rinse waters.  These
waters are a serious pollution problem.  However, they also have an intrinsic
value in the metals they contain.  Recovering these metal values and re-
cycling the rinse water has been a long standing challenge.  Of considerable
promise in this regard is a new membrane process that we refer to as coupled
transport membrane processing.

     In the next section, the principles of coupled transport membranes are
described.  The bulk of this report is concerned with an experimental study
of these membranes and their application to plating wastes.  Initially,
some range-finding studies were performed to find membranes and conditions
that would be suitable for processing dilute solutions of chromium, copper,
and nickel.  The dependence of the transmembrane flux of these metal ions
on such conditions as pH and concentration was measured in laboratory perm-
eation cells.  We then carried out a series of experiments in which the
conditions that would exist in a counter-current coupled transport concen-
trator were simulated.  These results were used to carry out an economic
analysis of the process for chrome-plating rinse waters.  The results
showed that the capital costs of such a coupled transport unit would be
substantially lower than the annual savings realized in chromium recovery.
Favorable economics are also predicted for copper plating rinse waters.
Less favorable were the projected economics with nickel wastes because of
the lower nickel fluxes observed to date.

-------
                                 SECTION II

                  PRINCIPLES OF COUPLED TRANSPORT MEMBRANES

      In a coupled transport membrane,  the  flow of  the permeant of interest
 is coupled to the flow of some second  species.   Under the right conditions,
 the flow of this second species can force  the permeant of interest to flow
 against its own concentration  gradient thereby,  for example,  separating this
 species from other species and concentrating it  as  well.   These membranes
 can thus be considered a kind  of selective chemical pump.

      More specifically,  the coupled transport membranes being developed at
 Bend Research consist of a suitable liquid complexing agent constrained by
 capillary forces to the pores  of a microporous membrane.   Thus, these are
 essentially liquid membranes.   The complexing agent is specific for the
 permeant of interest under certain conditions.   The desired permeant is
 complexed at one interface between the membrane  and an external solution.
 The complex diffuses across the liquid membrane  to  the downstream interface
 where the reaction is reversed because of  some appropriate change in con-
 ditions.   The free complexing  agent then diffuses back across the membrane
 where it picks up more of the  desired  permeant.  The complexing agent thus
 acts as a shuttle,  carrying both the desired permeant and some second,
 coupled species across the membrane.

      In the work described here,  the coupled species is the hydrogen ion.
 This ion can either cross the  membrane in  the same  direction as the metal
 ion of interest (a case  we refer to as "co-transport") or opposite to the
 metal ion (referred to as "counter-transport").  Because of the requirement
 to maintain electroneutrality  across the membrane,  the hydrogen ion must
 move in the same direction as  the metal ion if the  metal is present in
 anionic form,  e.g.  chromium as the chromate ion.  On the other hand, with
 free metal ions such as  copper and nickel  in solution, the flow of hydrogen
 ions must be counter to  the flow of metal  ion.   These two cases are dis-
 cussed below.

 CO-TRMISPORT
           Consider  the case in which a coupled transport membrane separates
 a  dilute  aqueous solution of chromic acid  at low pH from a less acidic
 solution  of  sodium  chromate.   The membrane contains a complexing agent  for
 the  chromate,  R,  which could be a water immiscible  tertiary amine,  for  ex-
 ample.  These  conditions are depicted  schematically in Figure la.   At the
 left-hand membrane-solution interface,  the chromate will partition  into
 the amine phase  in  the membrane where  one  molecule  of chromate complexes
with  two  molecules  of  chromate,  forming the neutral complex (RH)2Cr04.   This
complex is soluble  only  in the amine phase.   The complex will diffuse down
its concentration gradient to  the opposite or downstream interface.   There,
because of the higher  pH,  the  complex  dissociates,  freeing  the  chromate  to
                                     2

-------
        Low pH» dilute aoueous
2R
  
     CpO
        ,,(opg)
Microporous mem-
brane containing
complex ing agent R
High pH concentrated aqueous
solution  of CrOu-
                                               K
                                                      CPO
                                                 2R
                    (orf.)
                                                         ,,,   ,
                            CO-TRAMSPORT
    Low or moderat^ pH, dilute
    solution of H  ions
              + 2RH(org)
               R2H
-------
 the aqueous  phase  and  regenerating the neutral amine,  R.   The amine then
 diffuses  down  its  own  concentration gradient toward the left-hand interface
 (or "feed" side  of the membrane)  where it combines with more chromate and
 the cycle is repeated.   The equilibrium existing across the aqueous-organic
 phase  boundaries is this:

             2R(org) +  2H+(aq)  + CrO^(org')^=^(RH)2Cr04(org) .           (1)


 At the feed  interface,  the hydrogen ion concentration  is high and the re-
 action is forced to the right.  At the downstream interface, the reaction
 is shifted to  the  left because of the higher pH.   Thus, the amine is a
 carrier.  Hydrogen and chromate ions permeate the membrane in the same
 direction, from  left to right  in  Figure la.   Note that an increase in H* con-
 centration on  the  left-hand side  of the membrane or, equivalently, a de-
 crease in H+ concentration on  the right-hand side will favor the left-to-
 right  flux of  chromate.   (The  situation is not quite this simple because of
 the equilibrium  that exists in solution between CrO^ and C^Oy,  but the
 essential result is the same).

     With a  suitable pH gradient,  chromate can be "chemically pumped"
 against its  own  substantial concentration gradient.  It can be shown in fact
 (see Appendix  A) that  in the co-transport case the steady state  concentration
 ratio  of  the divalent  chromium anion across  the membrane depends on the
 hydrogen  ion concentration ratio  in the following way:
                                                                          (2)
where  the  subscripts  o  and  £  refer  to  the  feed  and  downstream solutions,
respectively  (i.e.  the  left-hand  and right-hand  solutions  in  Figure la).
Thus,  a  pH difference across the  membrane of  two units  should produce  a
chromate ion  concentration  ratio  of  10,000, with the less  acidic  solution
being  more concentrated in  chromium.  The chromate ion  will diffuse from the
dilute solution  into  the concentrated solution until the condition  indicated
in equation (2)  is  satisfied.

COUNTER  TRANSPORT
           Consider  next the situation depicted in Figure Ib.   Here,  the
metal  ions  exist  as cations and because we  are interested  in  copper and
nickel,  we  have  shown them  as  divalent.  The  complexing agent in  the membrane
in this  case  could  be an oxime, for  example,  denoted in the Figure  as  RH.
Again, two  moles  of oxime will react  with one mole of the  metal ion, and the
equilibrium existing  at the membrane-solution interfaces is:


                M"H"(aq)  + 2RH(org)^=±:2H+(aq)  + R2M(org) .             (3)


The metal-oxime complex is  again  soluble only in the organic  phase.  At the
left-hand interface,  the equilibrium is shifted  to the  right,  because  of the

-------
low concentration of H+ ions, and the metal-oxime complex partitions  into
the liquid membrane.  The reverse reaction occurs at the downstream or right-
hand interface.  Thus, metal ions are again chemically pumped from left to
right in the figure while hydrogen ions flow the other way.

     It can be shown  (see Appendix A) that the steady state ratio of  di-
valent metal ion concentrations across the membrane is given by:
                                                                          (A)
The metal ions can again be pumped against a substantial concentration
gradient but in this case  the concentrated metal ion solution will be at the
lower pH.

     While we have not attempted to derive it here, it can also be shown
that the same pH conditions that favor high concentration factors across
the membrane also favor high metal ion flux across the membrane.  Thus, in
the co-transport case, metal ion flux will be enhanced by increasing the
hydrogen ion concentration gradient in the same direction.  In the counter-
transport case, metal ion  flux is enhanced when the hydrogen ion concen-
tration gradient is increased in the opposite direction.

 APPLICABILITY OF COUPLED TRANSPORT MEMBRANES TO PLATING WASTES

          Coupled transport membranes could be assembled into a concentrator.
The way in which such a concentrator would be applied to the treatment of
plating wastes is illustrated in Figure 2.  The rinse water containing metal
ions would be fed to one side of the membrane concentrator.  The metal ions
would be transported across the membrane by the coupled transport process.
Under favorable conditions, the metal could be concentrated in the unit to
produce a solution of sufficient concentration to feed back to the plating
bath, while the rinse water, almost totally stripped of its metal values,
would be returned to the rinse bath.  Alternately, a partially depleted rinse
water could be discharged.

     Under favorable conditions, then, the plating and rinse operation could
be run closed-loop.  No metal is wasted and no make-up water is required for
the rinse bath.

-------
PLATING
   BATH
                                                               RINSE
                                                                BATH
                          r
                                     RAG OUT
                                                    \
                                                        V
                    COUPLED  TRANSPORT CONCENTRATOR
CONCENTRATE
                                                                                    DILUTE
                                                                                     RINSE
                                                            RINSE
                                                            EFFLUENT
                        T
                                                                            DISCHARGE
           Figure 2.   Simple  schematic  of coupled-transport  concentrator as applied to
                      plating rinse  water effluent.

-------
                                 SECTION III

                                EXPERIMENTAL

     Two sorts of experiments were performed in this reporting period.  In
the first of these, we examined  suitable complexing agents for chromium,
copper, and nickel.  This search was performed by measuring distribution
coefficients of the metal ions between organic liquids  (known to complex
these ions) and aqueous solutions at known pH.  An ideal complexing agent
is one that is highly water immiscible and which forms a very stable complex
at a given pH.  The complex should also be highly soluble in the complexing
agent but insoluble in water.  Finally, the complex should be unstable at
some higher or lower pH.  Having identified suitable complexing agents, we
then prepared liquid membranes from them and carried out a number of perm-
eation experiments to measure the metal ion flux across the membrane and its
dependence on pH and other conditions.

     The second type of experiment was designed to simulate the conditions
that would exist in an actual coupled transport membrane concentrator.  The
purpose of these experiments was to obtain -data representative of what might
be achieved in practice so that  an economic assessment of the process could
be made.

DISTRIBUTION  COEFFICIENTS
          These were obtained simply by shaking the complexing agent for two
minutes with an aqueous solution of the metal ion at a known pH in a sep-
aratory funnel at room temperature.  After allowing the phases to separate,
a sample of the aqueous solution was removed and analysed for metal ion.  The
concentration in the organic phase was then obtained from the known volumes
and mass balance considerations.  Acid or base was then added to the aqueous
phase and the process was repeated at a new pH.

MEMBRANES  AND PERMEABILITY  APPARATUS
          Coupled transport membranes were prepared by simply immersing a
microporous membrane in the liquid complexing agent of interest.  The liquid
was rapidly sorbed into the pores bv capillary action.  For our initial
studies, we chose Celanese Celgard ®2400 as the microporous membrane.  It is
made of polypropylene and is highly resistant to acids.

     Permeation experiments were carried out in standard two-compartment
glass cells shown in Figure 3.   The membrane was mounted without a gasket
between the half cells, held together by a standard ball-joint clamp.  The
volume of each half cell was 100 cnH, and the membrane area was 20 cm  .  The

-------
cells were supplied with stirrers to avoid concentration polarization ef-
fects.  The cells were mounted in a constant temperature bath.  All perm-
eation measurements were performed at 30 ± 1°C.
             MEMBRAN
              Permeation Cell
                                                     STIR MOTORS
                           an       na    na
            r
               CONSTANT  TEMPERATURE  BATH
              Permeation Cells in Water Bath
        Figure 3.   Permeability apparatus for testing coupled transport
                   membranes.

-------
     The pH and metal ion concentration were periodically determined in the
course of a run by removing small samples  from the cell, through sampling
ports not shown in the figure, and replacing with fresh solution at the same
pH.  Metal ion fluxes were obtained from the concentration-vs.-time data
after correcting for sampling.

ANALYTICAL METHODS
          Chromate concentrations have been measured spectrophotometrically
at 350 nm.  The pH of the samples was first raised to 12 to insure that the
dichromate ion would not interfere.

     Copper and nickel analyses have been  made with a Perkin-Elmer model 290
atomic absorption spectrophotometer.  With this instrument, both copper and
nickel can be measured in aqueous and organic solutions at concentrations as
low as 15 ppm with a precision of ± 2%.

REAGENTS

          The complexing agents studied were Alamine 336, a tertiary amine
from General Mills; LIX 63, LIX 64N, and LIX 70, all oximes from General
Mills; and Kelex 100, a substituted quinoline from Ashland Chemical.  All of
these are common industrial liquid ion exchangers.  Their chemical structures
are presented in Table I.

     In many permeation experiments, these complexing agents were diluted
with a compatible, inert carrier, principally to reduce their viscosity.
The diluents used were Solvesso from Exxon and kerosene.  In some cases, a
small amount (2 vol%) of isohexadecyl alcohol from American Hoechst was
also added to the diluted complexing agents in order to prevent the formation
of a separate phase.(1)  Apparently, this  agent promotes the solubility of
the metal ion complex in the organic complexing agent.
(1)  C.J. Lewis, "Liquid Ion Exchange in Hydrometallurgy", in Recent De-
       velopments in Separation Science Vol. II, N. N. Li  (Ed.), CRC Press,
       Cleveland, Ohio 1972.

-------
                                    Table I


                 COMPLEXING AGENTS  FOR COPPER,  NICKEL,  AND
                                  CHROMIUM
Kelex  100
 (Ashland Chemical)
LIX  63
(General Mills)
           OH   CH2-CH3

CH--CH9-CH-CH-C-CH-(CH?)--CH
  j   ^  I     ,,       £• J
                                            (CH2)3 N-OH

                                             CH
LIX 64N, a mixture of 10%
LIX 63 and 90% LIX 65N
(General Mills)
LIX 70
(General Mills)
Alamine 336
(General Mills)
CH3-(CH2)
                                                                 + LIX 63
                                    10

-------
                                  SECTION IV

                                    RESULTS


DISTRIBUTION COEFFICIENTS
            The distribution coefficient is  defined here as:

               K _ Concentration of  metal in the organic phase
                   Concentration of  metal in the aqueous phase

  Distribution coefficients for copper with  several complexing agents  are  pre-
  sented in Figure 4 as a function of  the pH of the aqueous phase.   The  com-
  plexing agents were diluted to a 10  vol% solution in kerosene in  all cases.

       Similar data for nickel are presented in Figure 5.   Note the hysteresis
  effect.   The data connected by the upward-pointing arrows were taken while
  nickel was being loaded into the organic phase,  i.e.  on increasing pH, while
  the downward-pointing arrows connect data  taken while nickel was  being strip-
  ped from the organic phase.  The hysteresis  effect was reflected  in  the  perm-
  eation experiments as well, as is  discussed  below.

  DEMONSTRATION OF COUPLED TRANSPORT
            The results of a coupled transport experiment with Cr03 are  pre-
  sented in Figure 6a.   The complexing agent was Alamine 336  diluted to  40
  vol% with Solvesso.   The feed side of the  permeation cell initially  con-
  tained 50 ppm Cr03 while the downstream side was buffered at either  pH 7 or
  11,  and it initially contained no  Cr03.  Clearly,  chromium  was transported
  in  the direction of decreasing hydrogen ion  concentration.   Because  the  two
  cell compartments are of equal volume, we  can see that beyond the point
  where the downstream concentration reached 25 ppm,  the chromium was  diffusing
  uphill.   The flux was higher at the  higher downstream pH as is predicted from
  the analysis presented in Part II.   Almost complete removal of chromium  from
  the feed solution was observed in  the course of  the experiment when  the
  downstream pH was 11, and it is predictable  that the same effect  would have
  occurred at a downstream pH of 7 at  some longer  time.

       Coupled transport of copper was demonstrated using a 10% solution of
  Kelex 100 in kerosene as the liquid  membrane.   The results  are presented
  in  Figure 6b.   The initial feed solution contained either 0 or 100 ppm
  copper at pH 2.5,  while the downstream solution  was initially at  pH  1.0
  with either 3.0 or 9.3 wt% copper, the latter being an essentially satu-
  rated solution of copper sulfate.  In Figure 6b, the feed concentration  is
  plotted  against time.  In all cases, the copper  was transported counter  to
                                      11

-------
             100
              10
Distribution
 Coefficient
            1.0
            0.1
           0.01
                                     D
                                                  10% LIX 63
_L
               0       1.0     2.0     3.0     4.0     5.0

                                   pH of Aqueous Phase
               6.0    7.0
                                                 ^^^^
     Figure 4.  Distribution coefficients of  Cu   vs.  pH for several
                liquid  ion exchangers dissolved in kerosene.
                                     12

-------
          10,000
           1,000
             100
Distribution
Coefficient
   10
              0.1
             0.01
                           246

                             pH of Aqueous Phase
Figure 5.
Distribution coefficients for nickel vs.  pH in 10%  L1X 64N
in kerosene (—) and 10% Kelex 100 in kerosene (	).
                                13

-------
                   SO
     Downstream CrO  ,0
      Concentration
         (ppm)
                   20
                  10
                                         Downstream  pH 11
                   pH  7
      Initial conditions:  50  ppm
      CrO, in feed, 0 ppm downstrean
                                     20       30       MO
                                           Time  (min)
                        50
                                60
 Figure  6a.   Downstream chromate concentration vs.  time.
                120
                100
                 80
        Feed
     Concentration  60
        (ppm)
   Initial conditions:  feed pH
   2.5,  Cu concentration 0 or  100
   ppm;  downstream pH 1.0, Cu
   concentration 3.0% or 9.3%.
                         1000
3000
  Time (min)
Figure 6b.   Concentration of copper  in the feed vs. time.

-------
the hydrogen  ion  concentration gradient.   In these experiments,  the steady
state concentration  factor was attained from both directions,  i.e.  both the
0 ppm and  the 100 ppm copper feed solutions reached a final concentration of
about 20 ppm  when the downstream compartment was 3.0 wt% Cu.   The steady
state copper  concentration factors were in  the range  1500-2000 in these
two cases.

     A series of  steady  state  concentration factors for  copper were measured
as a function of  the  pH  difference across the membrane.  The results are
presented  in  Figure  7.   Note that at a  pH difference  of  less than 2 units,
copper was concentrated  against  its own concentration gradient by about
4000-fold.  In fact,  the slope of the line  in this essentially log-log plot
is 2, which is just  the  value  predicted by  equation  (4).

EFFECT  OF DILUENTS ON FLUX
           The diluent added to the complexing agent can  be expected to have
two effects on the flux  of metal ions across coupled  transport membranes.
First, because the diluent displaces some of the  complexing agent, less of
the metal  ion complex will be  formed, and we could expect a lower metal flux
as a result.  Opposing this effect, however, is the fact that the diluents
are lower  molecular  weight compounds and hence are much more fluid at room
temperature than  the complexing  agents.  The reduced  viscosity of the liquid
membrane in the presence of diluent should  enhance the diffusivity of the
metal ion  complex (according to  the Stokes-Einstein equation) and hence in-
crease the flux.  These  two effects thus run counter  to  one another.

     The effect of the concentration of complexing agent in the diluent on
the flux of metal was measured.   Typical results  for  chromium are presented
in Figure  8.  The complexing agent was  Alamine 336 and the diluent was
Solvesso.  The feed  solution contained  5000 ppm chromic acid, while the
downstream solution  was  maintained essentially free of chromium and at pH 7
throughout the experiment.   A  maximum occurred in the flux at about 40%
Alamine, apparently  the  result of the opposing effects described above.
Because of this maximum, we carried out our further chromium experiments at
this concentration of complexing agent.

     Similar  data are plotted  in Figures 9  and 10 for copper and nickel,
respectively.  The liquid  membrane used in  the copper experiments was Kelex
100 diluted with  kerosene;  for the nickel experiments, the Kelex was
diluted with  Solvesso.   Flux maxima occurred in both  cases, and we again
carried out our further  experiments at  the  complexing agent concentrations
at which the maxima  occurred.

     It is noteworthy that the maximum  flux in the chromium experiments was
3-4 times  the maximum observed with either  copper or  nickel.

SIMULATION OF ACTUAL PLATING RINSE CONDITIONS
General  Considerations.                 in Figure  2 we indicated in a general
way how a  coupled transport membrane concentrator might be applied to the
treatment of plating  rinse waters.  An  ideal system would be one that could
be run closed-loop,  i.e. with  no discharge  of metal ions and with complete


                                     15

-------
recycle of the purified rinse water.  This ideal situation will generally
not prevail in a coupled transport membrane concentrator, because there  is
always a flow of coupled ion across the membrane along with the flow of
metal.  In co-transport, the co-permeating species  (or another ion used  to
      •p
      o
      C
      o
      •H
      •M
      •P
      C
      0)
      O
      C
      O
      o
      0)
      ex
      (X
      o
      o
        5000
         2000
        1000
          500
          200
         100
50
20
          10
             o.o
            0.5
O Feed solution maintained
   at pH 2.5


A  Downstream solution
   maintained at pH 1.0
1.0
1.5
2-0
                                     ApH
                                     U£SJ.J.

 Figure  7.   Steady state copper concentration factor vs.  pH cifference across


            the membrane.
                                    16

-------
  Flux
yg
  2
cm -min
                       20       40       60       80

                         % Alamine.in Solvesso
Figure 8.   Cr03  flux vs.  concentration of Alamine 336 in

           Solvesso.
                              17

-------
Copper2Flux
(jjg/cm -min)
                        20       40       60
                                % Kelex 100
80      100
   Figure 9.  Copper  flux vs. concentration of Kelex 100 in kerosene.
                               18

-------
                 12
                 10
(yg Ni/cm -min)  6
                                     o
                                             o
0       20
                                              60
                                       Kelex  100
80      100
      Figure 10.   Ni  flux vs.  concentration of Kelex 100 in  Solvesso.
                                 19

-------
neutralize it) would add  to  the  concentrated  stream.   Before  this  stream
could be fed back to the  plating bath,  this contaminant would have to  be
removed, in general.  In  counter-transport, the  coupled ion contaminates
the rinse water feed from which  the metal  ion was  removed.  These  contam-
inants may be simple to remove,  but it  should be recognized that in general
only one of the two streams  emanating from the concentrator could  be reused
without some further treatment.  In this section,  we describe a series of
experiments designed to simulate the conditions  that might exist in such a
concentrator.

 Chromium.                 In  Figure 11,  we  present  a diagram showing how the
concentrator might operate on a  chrome  plating rinse bath.  The following
assumptions have been made:  the plating bath contains 25 wt% Cr03v2)  and
there is a constant dragout  rate equal  to  2%  of  the flow rate of water
through the rinse baths,  producing a Cr03  concentration in the first rinse
bath of 5000 ppm.(3)  This solution would  be  fed to one end of the mem-
brane concentrator, and depleted of chromium  as  it passed through  the  con-
centrator from left-to-right in  the figure.   To  drive  the chromate ion
across the membrane against  its  concentration gradient, we must either
acidify this feed stream  or  make the downstream  side of the membrane basic.
Because we want to reuse  the treated rinse water,  we do not want to con-
taminate it with acid and hence  we chose to add  sufficient base to the
downstream side of the membrane  to neutralize the  chromic acid that is
transported across the membrane.  The most favorable conditions will be
achieved when the feed and downstream solutions  flow counter  to one another
so that high concentration gradients are established everywhere.

     The concentrated downstream solution  emerging from the concentrator
thus contains Na2CrO^.  If we assume that  a 50-fold concentration  factor
is achieved in the concentrator, the emerging concentrate stream will  con-
tain 41% Na2CrO^, which has  the  same chromium content  as the  plating bath.
However, the chromium would  have to be  recovered from  this stream  as CrOg
before it could be returned  to the plating bath.    This might  be achieved,
for example, by the addition of I^SO^ followed by  fractional  crystal-
lization. (4)  in this case, NaHS04 would become  a  by-product.

     To simulate the changing concentrations  that would exist in an oper-
ating concentrator, we arbitrarily divided the length  of the  concentrator
into four zones of decreasing chromium  concentration.   Within each zone, we
assumed a certain CrOg concentration and calculated the concentrations  of
        and NaOH that would exist within that zone.  These concentrations are
(2)  A.  Logazzo, Metal Finishing Guidebook and Directory, Metals and Plastics
       Publications, Inc.  Hackensack, N.J. 1976.

(3)  E.  McGuire, Crown City Plating Co., El Monte, CA, personal communication.

(4)  J.R.  Partington,  General and Inorganic Chemistry, MacMillan, London
       1954.
                                     20

-------
CrO
      PLAT IN6
        BATH
       2S% CrO.
                                                 RINSB CYCLE
                               5000 ppm CrO
i !
>

'^s—^^^-^N.^^^^yO^s^^y
< 100 ppm Cr03
                                      COUPLED  TRANSPORT  CONCENTRATOR
                                                  IT
*
                                                                                       NdOH
NaHSO,
        Figure 11.  Schematic representation of  coupled  transport concentrator as
                    applied to chromium plating  wastes.

-------
 summarized  in Table II.
                                 TABLE II
ZONAL CONCENTRATIONS IN A SIMULATED COUPLED TRANSPORT MEMBRANE CONCENTRATOR
                                FOR CHROMIUM

                 Zone                       12        34
Feed Cr03 Concentration (ppm)
Downstream NaOH Concentration (M)
Downstream Na2Cr04 Concentration (wt%)
5000
0.0
41
2500
2.5
20
1000
4.0
8
100
5.0
0.0
      A series of experiments  was  then performed  in permeation cells under
 each of these four sets of  conditions,  and  the chromium flux was measured.
 The liquid membrane consisted of  40%  Alamine 336 in Solvesso.   The results
 are presented in Figure 12.   Particularly noteworthy are the very high
 chromium fluxes over most of  the  concentration range and the linear de-
 pendence of flux on the concentration of  Cr03 in the feed.

 Copper and Nickel.                The analogous  situation that would occur
 in the treatment of rinse waters  from the acid plating  of copper or nickel  is
 shown in Figure 13.   Typical  compositions were again taken from Reference
 3.   Here hydrogen ion will  flow counter to  the metal ions,  acidifying the
 depleted rinse water.   This water could thus not be fed directly back to
 the rinse bath, as in the case of chrome  plating rinse  waters,  but it might
 be reused after neutralization in an  ion  exchange column,  for  example.  The
 concentrated downstream solution  emerging from the concentrator could be
 fed directly back into the  plating bath.

      The concentrations of metals and acids in the hypothetical zones of  the
 concentrator are presented  in Table III.  (For copper we arbitrarily se-
 lected five zones instead of  four).   These  were  calculated using the same
 considerations applied to the chromium case,  above.   In the nickel case,  we
 have assumed conditions typical of a  Watts  bath.   The boric acid present  in
 these baths is used  as a mild buffering agent and it would not significantly
 affect the coupled transport  process.

      The simulation  experiments with  copper were carried out using Kelex  100
 as  the complexing agent diluted to  10%  in kerosene.   This  choice was based
 on  the fact that Kelex complexes  effectively with copper at the pH at which
 the concentrator would be operated  (i.e.  pH approximately  1),  whereas the
 other  complexing agents do not.

      The results for copper are presented in Figure 14.  Again,  the copper
 flux is  plotted against feed  copper concentration when  the composition of
 the solutions  is that  shown in Table  IIIA.   These data  produced  the dashed
 line.  The  copper  fluxes  in this  case were  less  than Iug/cm2-hr,  well below
 the  data  in Figure 9 for  the  same Kelex concentration.   However,  this flux
was markedly enhanced  when the pH  of  the  feed solution  was  increased.   The
 solid  line  in  Figure 14 represents  comparable data obtained  when 0.03 M/liter
                                     22

-------
                  180-
                  160-
                 140-
                 120-
       Flux2

(yg CrO /cm -min)
       O
                 100 _
                            1000
2000      3000      4000      5000
                                  Feed Concentration  (|ig  Cr03/ml)
           Figure 12.  Cr03 flux vs. feed concentration.   All other

                       conditions simulate those in a coupled transport

                       concentrator.  (See Table II).
                                      23

-------
                  P LATINO
                    BATH
                                                                             CYCLE
                                drag out    __
              5* Cu, 0.5 M H2SOM,
                     or
              25% NiCl,,19.5% NisO
N3
                                5% Cu
                            0.5  M H2
                            25%  NiCl
                            19.5% ItiS'
                                   **
I
I
_> _
 1000 c»>» «« ,
 0.01 M~H2SOi, or
 5000 ppm NiCl2
 3900 ppm NiSOi,
 800 ppm K3B04
                                                       COUPLED TRAIMSPOR1 CONCENTRATOR
                    ^CU++,yi+^
                                                                     ^H?
                                                                                                                  Concen-
                                                                                                                    Dilute
                                                                                                                  Discharge
                         Figure  13.   Schematic  representation of  coupled transport concentrator
                                        as applied to  nickel or  copper plating  wastes.

-------
                           Table III


ZONAL CONCENTRATIONS IN A SIMULATED COUPLED TRANSPORT MEMBRANE
               CONCENTRATOR FOR COPPER AND NICKEL
  Zone
                           A.   Copper
Feed Cu (ppm)
Feed H2SO, (M)
Downstream Cu (wt%)
Downstream H SO, (M)
Zone
Feed N1C12 (ppm)
Feed NiSO^ (ppm)
Feed HC1 (M)
Feed H-BO- (ppm)
Feed H2S04 (M)
Downstream NiCl2 (wt%)
Downstream NiSO^ (wt%)
Downstream HC1 (M)
Downstream 113803 (wt%)
Downstream H2SOA (M)
1000
0.01
5
0.5
B.
1
5000
3900
0
800
0
25
19.5
0
4
0.0
500 250 100
0.0175 0.0213 0.0235
2.5 1.25 0.5
0.875 1.06 1.18
Nickel
2 3
2500 1000
1950 780
0.017 0.026
800 800
0.009 0.014
12.5 5
9.75 1.95
0.96 1.38
4 4
0.44 0.70
50
0.0250
0.0
1.25
4
100
78
0.033
800
0.0174
0
0
1.72
4
0.81
                              25

-------
     Flux,2
(yg Cu/cm -min)
                            250      500        750        1000
                           Feed Concentration (yg Cu/ml)
    Figure 14.   Cu flux vs. feed concentration.  Dashed curve:
                conditions simulate those in a coupled transport
                concentrator.   (See Table IIIA)  Solid curve:
                0.03 M/liter NaOH added to feed solution.
                             26

-------
of NaOH was added  to  each  of  the  feed  solutions.   (This is just the amount
of base needed to  neutralize  the  acid  permeating counter to the copper.)

     Nickel removal  from rinse waters  is  similar in principle to copper re-
moval.  However, the  hysteresis observed  in  the distribution coefficient
measurements with  both Kelex  100  and LIX  64N adds  some complexity.  Large
pH gradients would be required to drive the  nickel across the membrane, and,
as can be  seen from  Table  II1B, such large pH gradients would not normally
occur in the concentrator. Furthermore,  addition  of base to the feed solu-
tion to increase the pH gradient  could not be tolerated because nickel
precipitates above pH 6, where the concentrator would operate.  The much
narrower pH differences existing  under these simulated conditions produced
much lower nickel fluxes than we  observed under more favorable conditions.
The nickel fluxes  are plotted against  feed concentration in Figure 15.
                                      27

-------
               1.2
               1.0
               0.8
     Flux2     0.6
(pg Ni/cm -min)
               0.4
               0.2
                             J_
                             500       1000       1500      2000
                             Feed Concentration (pg Ni/ml)
        Figure  15.  Ni  flux vs. feed concentration.  All other conditions
                   simulate those in a coupled transport concentrator.
                   (See Table IIIB).
                                   28

-------
                                  SECTION V

                                  ECONOMICS

      Using the flux data generated under these simulated processing  con-
ditions,  we are in a position to crudely evaluate the economic  feasibility
of  the coupled membrane transport process.   We consider here  only chromium
for the present,  which, because of the large observed fluxes, is the most
favorable case.

      We begin by calculating the membrane area required to process a given
volume of rinse water per unit time.   An effective membrane permeability
can be obtained from Figure 12.   The flux-vs.-concentration curve in Figure
12  can be fitted by a straight line of slope k,  where


                   ,  _ 170 yg/cm2-min   . _„.    3, 2   .
                   k = 	°	— = 0.034 cm /cm -mm.
                        5000 yg/cm

Thus,  k is a kind of "clearance rate", expressing the volume  of solution
that is depleted of chromium per unit area per unit time.  As shown  in
Appendix  B, the concentration of chromium, C,  in a volume, V, of solution
in  contact with a unit area, A,  of membrane  will decay according to  a first-
order rate law:

                             C = C exp(-kAt/V)  ,
                                  o

where CQ  is the initial concentration of chromium in  the  rinse water.  If
we  assume that CQ = 5000 ppm and if we further assume that we want to reduce
the concentration to 100 ppm for return to the rinse  bath, we have that
C/CQ = 0.02 from which we have that V/At = 8.7 x 10~3 cm3/cm2-min.   This,
then,  is  the volume of rinse water that can  be treated per minute by 1 cvr
of  membrane.

      The  amount  of chromic acid  recovered from this rinse water is 4900
yg/cm2.   At a price of 60c?/lb (or 13c/gram)  of chromic acid,  the chromic
acid  recovered has a value of $26 per ft2 of membrane per year.  This is
well below the expected capital  cost  of a coupled transport membrane con-
centrating unit.   We believe that these units  could be constructed for a
cost comparable  to that of current  reverse osmosis units, which are  com-
parable in complexity.   These units sell for approximately $10/ft2 of
membrane.   Thus,  the expected payback time for the membrane concentrator,
ignoring  operating costs,  could  be  less than 6 months.
                                     29

-------
     Under favorable circumstances where complete recycle was practiced,
there would be an additional saving in the cost of make-up water for the
rinse bath.   This could amount to a savings of 1000 gallons of water per
ft^ of membrane per year, or approximately $l/ft^ of membrane per year.

     In spite of the lower fluxes observed for copper and the lower cost of
this metal,  the economic picture for copper is also quite favorable.  Only
for nickel is the present economic projection unfavorable, because of the
low fluxes observed to date.
                                    30

-------
                                 SECTION VI

                                 FUTURE WORK
     The work we proposed to do in Phase I of this project is now essentially
completed.

     In Phase II, we propose to design and construct a 50 ft^ mobile mem-
brane concentrator for use in the field.  The membranes would be prepared
in the form of small, microporous polysulfone hollow fibers.  These would
be assembled into a multiple tube-in-shell configuration and ultimately
field tested on plating rinse waters.  The unit will be designed for con-
tinuous, unattended operation.
                                      31

-------
                                 Appendix A

                     Steady State Concentration  Factors
     First, consider the reaction

                      CrO^ + 2H+ + 2R:5=

for which there is an equilibrium constant:
Let us denote the two sides of the membrane by the subscripts o and  $,.
 lere will be no flux of CrO-r across the membrane when |(RH),,CrO, |   =
          ^                 4                               2   4 |o
            ,                                           !•         «J
  Fer
  RH)
Under these conditions, it follows that  _Rj  = [Rn-  Thus,  because  the
equilibrium constant must hold at both membrane-solution  interfaces,  a
steady state is established when
or
                            M
(This development, of course, strictly applies only at pH  >  7.   Below this PH
the chromate-dichromate equilibrium becomes important).

     Next, consider the reaction


                          M"*""1" + 2RH ,    WR M + 2H+
                                     32

-------
                          K
                           eq
As above,  there, wilj, b,e zero flux when

conditions |RH



Thus,
                                     =r
the


 t
or
                                                  , and under these
                                   33

-------
                                Appendix B

              Exponential Decay of  CrO   in a Concentrator

     The removal of CrO« from the feed follows first order kinetics:
                             dt
where k is the slope of the flux-vs. -concentration curve, A is the membrane
area, and V is the volume of the feed solution.  Set  fcrOo] at t=0 equal  to
C  and [CrO,,"| at time t equal to C.  We can then integrate:
                             C dC
or                          ln(C/C ) = -kAt/V


and                         C/C  = exp (-kAt/V) .
                                     34

-------
                                   TECHNICAL REPORT DATA
                            ff lease read Instructions on the reverse before completing)
 REPORT NO.
EPA-600/2-79-181
 TITLE AND SUBTITLE

COUPLED TRANSPORT SYSTEMS FOR CONTROL OF HEAVY METAL
POLLUTANTS
. AUTHOR(S)                                    ~~~~

W.  C.  Babcock, R. ¥.  Baker, D. J. Kelly, J.  C.  Kleiber
             3. RECIPIENT'S ACCESSION NO.
             5. REPORT DATE
               August 1979 issuing date
             6. PERFORMING ORGANIZATION CODE
             8. PERFORMING ORGANIZATION REPORT NO.
i. PERFORMING ORGANIZATION NAME.AND ADDRESS

Bend Research, Inc.
Bend,  Oregon 97701
             10. PROGRAM ELEMENT NO.
               1BB610
             11. CONTRACT/GRANT NO.

               R80^682-01
12. SPONSORING AGENCY NAME AND ADDRESS
 Industrial Environmental Research Laboratory
 Office of Research and  Development
 U.  S.  Environmental  Protection Agency
 Cincinnati, OH   ^5268
             13. TYPE OF REPORT AND PERIOD COVERED
             14. SPONSORING AGENCY CODE
              EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT

       This report describes a process for separating  and concentrating heavy metals
 from electroplating  rinse waters.  Metal ions can  be  "chemically pumped" across a
 coupled transport membrane against large concentration gradients by allowing the
 counterflow of a coupled ion such as hydrogen ion.  The process is carried out within
 a microporous membrane containing within its pores  an organic,  water immiscible
 complexing agent.  The complexing agent acts as a  shuttle,  picking up metal ions on
 one side of the membrane, carrying them across the  membrane as  a complex, and
 preserving electrical  neutrality by carrying hydrogen ions  in the opposite direction.

       The importance of coupled transport is its high selectivity and flux.  High
 selectivity derives  from the rise of specific complexing agents.  High flux is
 possible because these are actually liquid membranes  with diffusivities many times
 greater than those in  solid membranes.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b. IDENTIFIERS/OPEN ENDED TERMS
                          c.  COSATI Field/Group
Electroplating, Industrial Waste Treatment,
Water  Pollution, Membranes, Osmosis, Copper,
Nickel,  Chromium
 Liquid Membranes
 Metal Ion Separation
 Membrane Transport
13B
18. DISTRIBUTION STATEMENT


Release  to Public

EPA Fo,m 2220-1 (Rev. 4-77)    PREV.OUS ED.T.ON is OBSOLETE
 Unclassified
20. SECURITY CLASS (Thispage)
 Unclassified
                           22. PRICE
                                             35
                                  -65T-060/5389

-------