United States
Environmental Protection
Agency
Hazardous Waste Engineering
Research Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-85/055 Aug. 1985
SERA Project Summary
Evaluation of Donnan Dialysis
for Treatment of Nickel Plating
Rinsewater
Edward T. Roach
A program to demonstrate the feasi-
bility of metal salt recovery and pollu-
tion control by ion selective diffusion,
also known as Donnan dialysis, was
conducted on a Watts-type nickel plat-
ing line. The objective of the program
was to demonstrate the Donnan dialy-
sis system for recovery of nickel utiliz-
ing the tubular membrane configura-
tion and to determine the suitable
operating conditions for this equip-
ment.
Several modules, consisting of tubu-
lar cation exchange membrane material
mounted within a shell, were used to
remove nickel ion from plating rinse-
water with accompanying replacement
of the nickel by hydrogen ion (supplied
by sulfuric acid stripping solution). Re-
covery rates were determined at vari-
ous operating conditions in order to
specify the equipment for commercial
use with emphasis on practicality, eco-
nomics, reliability, and limitations.
This Project Summary was devel-
oped by EPA's Hazardous Waste Engi-
neering Research Laboratory, Cincin-
nati, OH, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).
Introduction
Current concern over the presence of
metallic ions in the watershed has
caused the government to restrict and
regulate wastewater discharge prac-
tices. This concern prompted industry
to investigate certain recovery pro-
cesses where the technology is suffi-
ciently advanced to offer payback and
compete with conventional disposal
methods. Three processes depend on
membranes for effecting separation: re-
verse osmosis, electrodialysis, and
Donnan dialysis (also known as ion dif-
fusion separation). Driving force for the
first two processes are respectively
pressure and electrical potential. Driv-
ing force for Donnan dialysis, however,
is the membrane potential related to
concentration differences across the
membrane.
Utilizing the permselective properties
of ion-exchange membranes one can
cause an equilibrium to become estab-
lished between two solutions of elec-
trolytes separated by the membrane. In
the case of the cation exchange mem-
brane used in this project, the anionic
species in the solutions are prevented
from diffusing while the mobile cationic
species will distribute themselves until
at equilibrium:
1/2
1/2
1/Z
Cil/
Cir
= Cjl/
= Ckl/
= k
Cjr
Ckr
Where i, j, and k are cationic species,
z is their respective valence, C is con-
centration, and I and r refer to left and
right sides of the membrane. The driv-
ing force for the cation transfer from
one side to the other is, therefore, the
magnitude of the systems displacement
from equilibrium, and this driving force
can be affected by varying the solution
concentrations. The extent that equilib-
rium conditions will be achieved de-
pends on electrolyte concentration,
membrane properties and operating
conditions.
-------
This program studied Donnan dialy-
sis as a means of concentrating nickel
ions from plating rinsewaters using hy-
drogen ions from dilute sulfuric acid as
the other mobile cationic species.
Procedure
A series of preliminary tests was per-
formed with the unit to determine oper-
ating characteristics and to prepare for
later tests. Simulated nickel rinse solu-
tions were dialyzed at flows between
0.063 and 3.78 l/m (0.017 and 1 gal/m).
Feed solution concentrations, flowing
on the tube side of the membranes,
were varied between 0.025 and 0.164 g/l
(0.0014 Ib/gal.) of nickel. Feed flows
were varied from 0.44 to 2.84 l/hr (0.12
to 0.75 gal./m). Strip solutions of 0.5 and
1 Normal sulfuric acid flowed at rates
from 37.8 to 378 l/m (10 to 100 gal./m)
on the shell side of the membrane. Flux
increased both with increased feed flow
and increased feed (rinse) concentra-
tion. Percent removal was in the 40 to
50% range for all operating conditions
except extremely low feed flow rates
where removal of 75 to 85% was
achieved. Strip concentrations of 0.5
and 1.0 Normal sulfuric acid gave identi-
cal results as long as the acid concentra-
tion was maintained. Also the flow di-
rection, counter-current or concurrent,
had no effect on the removal rates.
After this testing was completed, all
membrane modules were replaced with
newly fabricated perfluorinated ion ex-
change membrane tubular modules.
Reconstruction of the unit was carried
out to accommodate these new mod-
ules as well as to provide greater flexi-
bility in operating conditions. A total of
six Nation"* mass exchanges were sup-
plied by DuPont and installed in two
groupings designated front (F-1, F-2,
F-3, and F-4) and rear (R-1 and R-2).
Piping and valving allowed for choice of
either parallel or series feed flow in the
front exchangers each of which was
fitted with a pressure gauge, sampling
port, cartridge filter and flowmeter. The
rear exchangers were piped for series
flow only. All of these fittings were lo-
cated downstream of the system to
allow for product sampling. The accom-
panying feed side flow schematic (Fig-
ure 1) illustrates the piping and valving
described above. Note that this sche-
matic shows only the feed (rinse) flow
which is always downflow and always
on the tube (inner) side of the Nafion®
membrane. The strip solution (not
shown) is always counter-current and
upflow and always in parallel for the
front four exchangers and in series for
the rear two.
The refitted dialysis unit was installed
at Seaboard to treat the first of three
counterflow rinses that follow the plat-
ing tanks and a spray rinse on a Watts-
type nickel electroplating line. Rinse
water from a second nickel line was also
discharged through the circulating
counterflow tank giving a total rinse
water input to this tank of approxi-
mately 1136 l/hr (300 gal./hr). The
Donnan system was installed on this
first rinse tank, pumping nickel-
contaminated rinsewater through the
Donnan system and returning the prod-
uct (slightly acidic because of the ex-
change of I-T for Ni+ f). The sulfuric acid
strip tank (not shown in Figure 1) con-
stantly recirculated acid stripping solu-
tion through the strip side of the Don-
nan modules and back into the strip
tank. Because of osmotic flow through
the membrane, the volume of a strip-
ping solution slowly increased, necessi-
tating a periodic return of the nickel-
laden stripping solution back to the
nickel bath. The bath could accommo-
date the volume because of evaporative
losses.
The next series of tests was designed
to determine the performance of the ex-
changers and to decide on the most
convenient method for handling the
strip and rinse solutions. Initial opera-
tions were undertaken with parallel feed
flow to the four front exchangers. The
resulting data (Table 1) showed that
both increases in feed concentration
and feed flow rate caused an increase in
the transport rate, recorded as flux,
which at 1136 l/hr (300 gal./hr) was
greater than at 681 l/hr in spite of a
slightly lower feed concentration. These
-eed in b-j
S
<
+++Ffif
s.
1
I
k
n
S-*
p
n
; t
n
rt
4
t
— *.
s-*-
1
^
/:
<
i
h
i
k
5
1
^
t
k
p.
.L
J L
O
t
1
J k
T
•Mention of trademarks or commercial products
does not constitute endorsement or recommenda-
tion for use by the U.S. Environmental Protection
Agency
""" Parallel flow through F, - f
'Series flow through F, - F,
TSeries flow through /?, - R2
figure 7. Feed s/de flow schematic
U Valve
D Filter
O Pressure gauge
S = Sampling port
-------
Table 1. Initial On-Site Results
Feed Flow Rate (l/hr)
681
TI36
454
Average Feed
Concentration (g/l)
0.72
0.66
0.80
Average Nickel
Transport (g/hr>
163
176
210
Average
Flux (g/hr/m2)
2.15
23.7
28.0
variations did not seem to affect trans-
port rate as long as the strip solution
(H2SO4) concentrations remained at
1 Normal or above.
The front four exchangers were next
run in series. These runs showed gener-
ally lower nickel transport. This was ex-
pected since the initial rinse concen-
trations were gradually reduced on
passing through each successive ex-
changer and, as expected, the flux rate
decreased when the nickel concentra-
tion in the feed (rinse) decreased.
Results of further tests showed that
exchanger F-4 was more efficient than
the other three front exchangers. Sub-
sequent investigation showed this tube
was of an improved design that indi-
cated the possibility of upgrading the
total performance of this type of equip-
ment in the future. Additional testing
was performed to assess the effects of
increased acidity in the strip solution, to
evaluate the rear exchangers, which
were hard piped for series flow, and to
investigate the effects of operating the
mass exchangers with the rinse on the
shell side and the strip on the tube side
of the membranes.
In all, approximately 71 completed
runs were performed to assess the per-
formance of this dialysis unit under var-
ious operating conditions. In summary,
the modes of operation were the follow-
ing:
Strip solution on the shell side, and:
• front 4 exchangers in parallel
• front 4 exchangers in series
• investigation of single exchanger
• rear exchangers only
Strip solution on tube side, and:
• front 4 exchangers in parallel
• front 4 exchangers in parallel with
rear exchangers in series
No less than three runs were performed
to establish average flux values. All data
used in the project report, summarized
here, are listed in the appendix of the
full report.
Results and Discussion
An investigation of parallel versus
series flow through the membrane
modules showed significant differences
in performance. In all cases parallel op-
eration gave at least 15% higher nickel
flux values, confirming that the process
is concentration sensitive. Combined
parallel-series operation gave lower
flux values than would be expected and
suggests that series operation should
probably be restricted to polishing ap-
plications while parallel operation is
particularly indicated for higher recov-
ery rate systems.
The most effective operations were
those runs in which the exchangers
were piped in parallel configuration
with a feed concentration of 0.5 g/l
(0.004 Ib/gal.) or greater and with feed
flow rates equal to or greater than 795
l/hr (210 gal./hr). With the exchangers
used in these tests, flux was signifi-
cantly greater when feed was circulated
through the tubes and strip was circu-
lated on the shell side of the membrane
(outside of the tubes).
Applications certainly exist for Don-
nan dialysis treatment of plating rinse-
waters utilizing either tubular mem-
brane modules or plate and frame
multi-cell stacks. The exchange of H +
ions for Nif + ions will produce a slightly
acidic effluent, which must be neutral-
ized in an end-of-pipe treatment system
before discharge. The stripping solution
containing the nickel may, in most
cases, be returned directly to the plating
bath. Donnan dialysis should also be
considered as a treatment complement-
ing other technologies as part of the
overall recovery and pollution control
facility.
Conclusions and Recommenda-
tions
Donnan dialysis is an effective pro-
cess for recovery of nickel ion from elec-
troplating rinse water.
By recirculating rinse water from a
Watts-type nickel plating line through a
Donnan dialysis unit fitted with perm-
selective cation exchange membranes
and utilizing a strip solution of 1 Normal
sulfuric acid as the source for other
cation species, it was possible to re-
move over 6.6 grams of nickel per
square meter of membrane per hour
(0.004 Ib/ft2/hr) from the rinse water
thus prolonging the life of the rinse so-
lution. This process generates a concen-
trated solution with nickel ion levels in
excess of 30 grams per liter (0.25 Ib/gal.)
and is useful to replace the nickel re-
moved from the electroplating bath and
to help control the plating batch pH.
Economic advantages lie in the sav-
ings of raw material cost, the reduction
of conventional disposal costs, reason-
able capital investment costs, and lower
operating costs than the electrical and
high-pressure-driven dialyses pro-
cesses.
Variables affecting the rate of recov-
ery of nickel ion from wastewater have
been throughly examined both in con-
trolled laboratory tests and in actual on-
site operations of the Donnan dialysis
unit. Separation in this unit was accom-
plished using modular hollow tube
cation exchange resin of the perfluoro-
sulfonic acid type. This process also can
be carried out using the conventional
plate or frame configuration, such as is
typical of the electrodialysis multicell.
Plate and frame multicells, while
larger than tubular modules, have been
shown to enjoy long life expectancy in
nickel recovery operations. Such stacks
should be tested in series and parallel
combinations under operating condi-
tions similar to those described herein.
Techniques for nickel recovery by this
method should then be adapted for re-
covering other ionic species.
The Donnan dialysis system gener-
ates a mildly acidic effluent because of
the exchange of H^ for Ni + + that can be
combined with other flows to the end-
of-pipe neutralization system.
The full report was submitted in fulfill-
ment of Grant No. R-804655 by the
Seaboard Metal Finishing Company,
Inc. The report covers the period March,
1978 to June, 1979, and work was com-
pleted as of June, 1981.
-------
Edward T. Roach is with Chemical Recovery Systems, Fall River, MA 02720.
Mary K. Stinson is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of Donnan Dialysis for Treatment of
Nickel Plating Rinsewater, "(Order No. PB 85-200 046/AS; Cost: $8.50, subject
to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Hazardous Waste Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S2-85/055
OC003Z9 PS
?30 S DEAS8CRN STREET
CHICAGO it
u.S. GOVERNMENT PRINTING OFFICE:1985—559-01b/27121
------- |