EPA-600/2-76-197
October 1976
Environmental Protection Technology Series
NEW MEMBRANES FOR
TREATING METAL FINISHING EFFLUENTS BY
REVERSE OSMOSIS
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental 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-600/2-76-197
October 1976
NEW MEMBRANES FOR TREATING
METAL FINISHING EFFLUENTS
BY REVERSE OSMOSIS
by
Robert J. Petersen
Kenneth E. Cobian
Midwest Research Institute
Minneapolis, Minnesota 55406
Grant No. R-803264-01-0
Project Officer
Donald L. Wilson
Industrial Pollution Control Division
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 Labora-
tory, U. S. Environmental Protection Agency, and approved for publication. Ap-
proval 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 endorsement or recommendation
for use.
ii
-------
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
(IERL-Ci) assists in developing and demonstrating new and
improved methodologies that will meet these needs both
efficiently and economically.
This report is the result of a nine-month study on "New
Membranes for Treatment of Metal Finishing Effluents by Reverse
Osmosis" which was completed as of June 1975. These results
included a successful long-term demonstration (2360 hours) of
the performance of NS-100 reverse osmosis membranes for treat-
ment of extreme pH electroplating wastes (pH 1 acid copper and
pH 13 zinc cyanide rinse water effluents). Feasibility of this
new membrane to commercial applications in electroplating
installations was thus shown.
This project was one of several projects undertaken by
IERL-C and the American Electroplaters' Society to demonstrate
new techniques for purifying metal finishing waste water, a
source of much water pollution throughout the country. This
report will be especially interesting to individuals in the
plating industry who are compelled by law to meet rather
stringent effluent guidelines within the near future, and to
individuals involved in industrial waste water research.
For further information on this subject contact the Metals
and Inorganic Chemicals Branch, Industrial Pollution Control
Division.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
iii
-------
ABSTRACT
Long-term reverse osmosis tests were conducted with electroplating wastes on
a new membrane referred to as NS-100. This membrane consists of a polyurea
layer, formed by the reaction of tolylene diisocyanate with polyethylenimine,
deposited on a porous polysulfone support film. The membranes were tested as
liners within I/2-inch diameter fiber glass tubes. A total of 2360 hours of
continuous reverse osmosis operation was achieved, 1220 hours on pH 1.2 acid
copper rinse water and 1140 hours on pH 12.8 alkaline zinc cyanide rinse wa-
ter. The membranes exhibited remarkable chemical stability during exposure
to these two pH extremes. Copper and zinc rejections were generally greater
than 99 percent, while cyanide rejections were typically 96 percent or great-
er. Membrane fluxes were in the range of 18 to 24 liters per square meter
per hour (11 to 14 gfd) for acid copper, but only 8 to 15 l/m2-hr (5 to 9
gfd) for zinc cyanide at 41.4 bars (600 psig) and 25°C. Rejection organics
(including brighteners) was 60 to 78 percent for acid copper and greater than
95 percent for zinc cyanide. NS-100 membranes did not reject sulfuric acid.
A modified membrane, NS-101, demonstrated twice the permeate flux of NS-100
toward zinc cyanide baths, but cyanide rejections were low at 90 percent.
The serviceability of these membranes toward these pH extremes was adequate-
ly demonstrated in this test series. Difficulties of producing reproducible,
high-flux tubular membranes were not fully resolved in this study. Thus, in
the tubular configuration, this membrane is not yet in a stage of development
for on-site demonstrations.
iv
-------
TABLE OF CONTENTS
Page
SECTION I
CONCLUSIONS 1
SECTION II
RECOMMENDATIONS 3
SECTION III
INTRODUCTION 4
Background 4
Summary of Previous Work with NS-100 Membrane 5
Current Research Program 6
SECTION IV
EXPERIMENTAL PROCEDURES 8
Polymers 8
NS-100 Membrane 8
NS-101 Membrane 8
Tube Cast Membranes 8
Reverse Osmosis System 12
Reverse Osmosis Testing 16
Test Duration 16
Freesure 16
Concentration 16
Membrane Evaluation 19
SECTION V
PROGRAM RESULTS 20
Optimization of NS-100 Tube Fabrication 20
Nonoptimized Membranes: Performance
Towards Plating Solutions 20
Optimization of NS-100 Membranes 22
NS-101 Membrane Fabrication 24
Long-Term Membrane Performance Toward Acid
Copper Plating Bath Rinse Water 25
Membrane Rejection 25
Membrane Flux 28
Effect of Feed Concentration on Membrane
Performance 28
Plating Solution 28
Summary of Results: Acid Copper Test 30
-------
TABLE OF CONTENTS
(Continued)
Long-Term Membrane Performance Toward Zinc
Cyanide Plating Bath Rinse Water 31
Membrane Rejection 35
Membrane Flux 36
NS-101 Membranes 36
Discussion 36
Summary of Results: Alkaline Zinc Cyanide
Test 37
SECTION VI
REFERENCES 38
SECTION VII
APPENDICES
APPENDIX A. Fabrication Procedure for
Tubular NS-100 Membranes for
Reverse Osmosis 40
APPENDIX B. Individual Membrane Performance
Data with Acid Copper Plating
Bath Rinse Waters and Feed
Analyses 50
APPENDIX C. Individual Membrane Performance
Data with Alkaline Zinc Cyanide
Plating Bath Rinse Waters and
Feed Analyses 55
VI
-------
FIGURES
Figure 1 Schematic Representation of NS-100
Membrane 9
Figure 2 Idealized Structure of Polyethylenimine
Crosslinked with Tolylene 2,4-Diisocyanate 10
Figure 3 Idealized Structure of Polyethylenimine
Crosslinked with Isophthaloyl Chloride 11
Figure 4 Flow Diagram for Reverse Osmosis Test
Loop 12
Figure 5 Photograph of Reverse Osmosis System
Used in Long-Term Studies, Showing the
Control Panel, Flowmeter, and Throttle
Valve for System Pressure Control. 13
Figure 6 View of the Reverse Osmosis Board Showing
Eight Tubes Connected in Series, with
Product Water Collection Line Attached. 14
Figure 7 View of Reverse Osmosis System Showing
Feed Reservoir, Heat Exchange Reservoir,
and Refrigeration Unit. 15
Figure 8 Longitudinal Section of a Fiber Glass
Reverse Osmosis Tube with NS-100 -
Polysulfone Membrane Support Composite 17
Figure 9 Photograph of Tubular Polysulfone Support
Liner, Abcor Fiber Glass Tube with End
Fittings, and Fiber Glass Tube Enclosed in
a Tygon Sleeve. 18
Figure 10 Plot of Reverse Osmosis Performance of
NS-100 Tubular Membranes Toward Acid
Copper Rinse Water 29
Figure 11 Effect of Acid Copper Rinse Water
Concentration on NS-100 Flux and Copper
Rejection. 30
Figure 12 Plot of Reverse Osmosis Performance of
NS-100 Tubular Membranes Toward Zinc Cyanide
Rinse Water. 34
Figure 13 Plot of Reverse Osmosis Performance of
NS-101 Tubular Membranes Toward Zinc
Cyanide Rinse Water 35
Figure Al General Outline of NS-100 Reverse Osmosis
Tube Fabrication 41
Figure A2 Cylindrycal Oven 49
vii
-------
TABLES
Table I Rejection Evaluation Techniques for Reverse
Osmosis Membrane Performance 19
Table 2 Effect of TDI Concentration on the Performance
of Tubular NS-100 Reverse Osmosis Membranes
with Zinc Cyanide and Acid Copper Plating
Rinse Waters 21
Table 3 Optimization of Tubular NS-100 Membranes with
One-percent Sodium Chloride Feed 23
Table A Reverse Osmosis Performance of Tubular NS-101
Membranes with One-Percent Sodium Chloride Feed 25
Table 5 Initial and Final Performances of Tubular NS-100
Membranes with Acid Copper Plating Solution 26
Table 6 Average Performance Data for NS-100 Tubes
During the Acid Copper Test 27
Table 7 Initial and Final Performances of Tubular
NS-100 Membranes with Alkaline Zinc Cyanide
Plating Solution 32
Table 8 Average Reverse Osmosis Performance of NS-100
Tubes During Zinc Cyanide Long-Term Test 33
Table Al Apparatus and Reagents for Tube Fabrication 42
Table Bl Acid Copper Feed Analysis 50
Table B2 Individual Membrane Performances with Acid
Copper Plating Bath Rinse Water 51
Table Cl Zinc Cyanide Feed Analysis 55
Table C2 Individual Membrane Performances with Alkaline
Zinc Cyanide 56
viii
-------
ACKNOWLEDGEMENTS
This program was sponsored through a grant from the U. S. Environmental Pro-
tection Agency with Donald L. Wilson, Project Officer. The American Electro-
platers' Society (AES), Inc., was the grantee co-sponsor of this program,
with J. Howard Schumacher of the AES serving as Project Manager (AES Project
No. 36). The support and assistance of the AES is hereby gratefully acknowl-
edged.
The authors are grateful for the cooperation of Mr. Court Platt, Precious
Metals Platers Incorporated, Mr. Roger Murnane, Superior Plating, Inc., and
Mr. William Cashin, Honeywell, Inc. for providing actual plating baths used
in this program.
This report was submitted in fulfillment of the requirements of Grant No, R-
803264-01-0 of the U. S. Environmental Protection Agency.
ix
-------
SECTION I
CONCLUSIONS
The NS-100 membrane (formerly NS-1) was shown to be an excellent membrane for
potential industrial use in the recycle of rinse water and plating chemicals
in acid copper and zinc cyanide electroplating lines. This was demonstrated
through 2360 hours of continuous operation with 1.27-cm tubular membranes,
half at pH 1.2 (acid copper) and half at pH 12.8 (zinc cyanide). NS-100 mem-
branes demonstrated stable salt rejection performance during this period, show-
ing greater than 99 percent rejection of metals (copper, zinc) and 96 to 99
percent rejection of cyanide ion.
Test conditions were severe in that actual plating baths diluted to one-tenth
full strength were used in these tests, a far higher concentration than pre-
sent in conventional rinse baths. Since most membrane surfaces in a potential
reverse osmosis installation would experience milder conditions, membrane life-
times of at least 3000 hours, and probably up to 5000 hours would be a reason-
able expectation based on these test results.
Water permeation rates through tubular NS-100 membranes were lower than de-
sired, based on previous studies with flat sheet NS-100 membranes. Flux rates
were 18 to 24 liters per square meter of membrane per hour (11 to 14 gallons
per square foot of membrane per day) for acid copper rinsewater, but only 8
to 15 l/m2-hr (5 to 9 gfd) for alkaline zinc cyanide rinse water. Continu-
ous flux decline with time was evident, which could be restored significantly
by osmotic cleaning. It was concluded that this flux decline was due in part
to formation of "dynamic" membranes on the NS-100 membrane surface. No at-
tempts were made to periodically clean membrane surfaces; industrial use of
cleaning aids (detergents, osmotic cleaning) should lead to a higher level of
flux values than observed in this program.
Two experimental NS-101 membranes (made with isophthaloyl chloride rather than
tolylene diisocyanate) demonstrated twice the flux of NS-100 membranes toward
alkaline zinc cyanide (about 27 l/m2-hr, or 16 gfd). Zinc rejections were
greater than 98.5 percent, but cyanide rejections were low at 90 percent. Op-
timization efforts on NS-101 membrane fabrication could lead to suitable salt
rejection characteristics.
At the beginning of the program, optimization of the fabrication procedure for
1.27-cm tubular NS-100 reverse osmosis membranes succeeded in doubling membrane
flux performance based on initial and final test comparisons. Unfortunately,
attempts to fabricate NS-100 tubes late in the program by the optimized pro-
cedure gave low flux membranes with extremely high salt rejections. In fact,
-------
throughout the program, efforts were hindered by a significant variability in
the performance of individual tubes fabricated at various intervals. Variabil-
ity was concluded to arise both from a very narrow set of acceptable fabrication
parameters and from as yet unknown factors contributing to manufacturing nonuni-
formity. This membrane system, at least in the form of tubular membranes, was
judged to be not yet ready for on-site demonstrations. Additional experimental
studies on tube fabrication, with emphasis on the NS-101 modification, seem
necessary as a next step.
-------
SECTION II
RECOMMENDATIONS
The objectives of the program were threefold: 1) to optimize the fabrication
process for NS-100 tubular reverse osmosis membranes to provide optimum per-
formance; 2) to demonstrate sustained performance of NS-100 membranes toward
highly acid and highly alkaline metal finishing waste waters; and 3) to pro-
vide sufficient data for preliminary engineering design of a possible field
demonstration unit.
This program successfully demonstrated the stability and performance of NS-
100 membranes to both types of plating wastes, acidic and caustic. Unfortu-
nately, the optimization studies on the fabrication process did not lead to
routinely reproducible, high-flux membranes. Thus, despite the great promise
of this membrane system for field trials, on-site demonstrations would be pre-
mature at this time without a better understanding of tube fabrication param-
eters.
To arrive at a field deiaonstration phase, it is recommended that further re-
search first be directed toward the membrane fabrication process. This task
would specifically involve obtaining higher flux membranes in a reproducible
manner. To achieve this goal, efforts would center on the NS-101 modification,
which has been found in related seawater desalination work to demonstrate even
greater durability than the NS-100 membrane. When this problem of membrane
tube nonuniformity and inadequate flux is solved, the on-site demonstration
phase can be realistically recommended.
-------
SECTION III
INTRODUCTION
Midwest Research Institute, through its North Star Division, has completed the
research program "New Membranes for Treatment of Metal Finishing Effluents by
Reverse Osmosis" under Grant No. R-803264-01-0 from the U. S. Environmental
Protection Agency (EPA) with the American Electroplaters* Society, Inc., as the
grantee organization. This program was the third phase of a study initiated at
North Star in 1970 on the development of reverse osmosis membranes for the
treatment of metal finishing waste waters. This study was designed to help
meet the needs of the metal finishing industry through improved pollution con-
trol and conservation of valuable materials.
During the first phase of this study, the treatment of metal finishing waste
waters by reverse osmosis was shown to be feasible (1). A number of cellulos-
ic membranes, both commercially available and improved derivatives synthesized
at North Star, were demonstrated to be capable of treating various metal finish-
ing effluents. The second phase of this study consisted of the fabrication and
testing of membranes found most promising in the first phase into tubular con-
figurations, and the development of new, noncellulosic, second generation mem-
branes for improved metal finishing waste treatment by reverse osmosis (2). The
third phase of this study, described in this report, consisted of a long-term
test of a very promising reverse osmosis membranes, NS-100, against highly acid
and highly alkaline feed solutions to demonstrate its unique level of chemical
resistance and sustained performance, as a prelude to commercial utilization.
Background
The metal finishing industry has an ever-growing problem in controlling and
eliminating the discharge of wastewater pollutants. The wastes that cause the
problems include rinse waters from metal electroplating solutions and from
acidic and alkaline cleaning and pickling solutions. The rinse water is a
constantly flowing process stream generally too voluminous to impound economic-
ally, yet concentrated enough to be damaging if released to the environment
without treatment. If discharged into the environment without treatment, these
rinse waters can pollute our natural resources, inhibit or destroy biological
activities in the natural environment and in biological sewage treatment pro-
cesses, and adversely affect materials of construction. Specific examples of
detrimental effects include the toxicity of heavy metals and cyanides to vari-
ous forms of aquatic life (3), the deleterious effect of copper and chromium
on biological sewage treatment processes (because of their toxicity to the
microflora) (4), and the corrosive effects of acids and bases on sewer lines
and metal and concrete structures (5,6).
-------
Several methods presently exist whereby waste waters containing cyanide and
metal ions can be treated for clean-up. Many of these techniques are aimed
toward the destruction and/or removal of the contaminating species from the
water. This task is often accomplished by the addition of chemicals to the
effluent stream to convert the undesirable constituents to either a less harm-
ful state or a state whereby it can be effectively removed. Although these
techniques are effective in improving the quality of the water effluent, they
can introduce solid waste disposal problems. Such is the case in the precipi-
tation of potentially harmful and difficult-to-handle metal hydroxide sludges.
An attractive alternative to existing techniques in the treatment of metal
finishing waste waters is reverse osmosis because it offers an opportunity
to reclaim valuable chemicals from the process stream as well as to purify
water for recycling purposes. *n theory, all waste discharge would be en-
tirely eliminated. The savings realized in reduced water consumption and
recovered chemicals can be credited against capital and operating costs for
the treatment systems. Reverse osmosis can be used in combination with other
existing methods to increase their treatment efficiencies. For example, it
can be used to treat water from a continuous cyanide destruction process for
recycling back to the plant operations, or it can reduce the metal ion con-
centration prior to an ion exchange treatment process (which would then act
as a polisher).
Several researchers (7-12) have examined the technical and economic feasibil-
ity of treating various waste water streams from metal finishing operations
by reverse osmosis. Computations, based on laboratory test results, have
shown this process to be economically viable for treatment of nickel plating
streams (8,10,12). Obviously, the degree to which reverse osmosis can be
adapted to a recycling process in a plating operation must be determined in-
dividually, on a case-by-case basis.
Summary of Previous Work with NS-100 Membranes
The polymer currently used most often as a membrane for reverse osmosis is
cellulose acetate. Reid and Breton (13) originally showed that this material
had excellent potential as a reverse osmosis membrane. Loeb and Sourirajan (14)
later developed the process for fabricating asymmetric membranes for cellulose
acetate, now used extensively. Second generation polymer membranes ideally
suited for reverse osmosis should exhibit hydrophilicity and ease of membrane
formation similar to cellulose acetate, and at the same time should have great-
er structural rigidity, resistance to chemical degradation and mechanical dur-
ability. The new nonpolysaccharide membrane, designated NS-100, which was
originally developed as a seawater desalination membrane under contract to the
Office of Saline Water (now part of the Office of Water Research and Technology,
U. S. Department of the Interior), has shown considerable promise as a second-
generation reverse osmosis membrane for metal finishing effluents.
In general, polysaccharide membranes, such as cellulose acetate, have been found
suitable for the reverse osmosis treatment of metal finishing waste solutions
only at pH's from 4 to 8. Many metal finishing effluents, however, are strongly
acidic or alkaline (i.e., acid copper pH M.; zinc cyanide pH >11). Reverse os-
mosis membranes comprised of nonpolysaccharide polymers would offer greater
-------
chemical resistance to a wide variety of metal finishing waste solutions. This
new nonpolysaccharide membrane (NS-100), showed promise for high chemical re-
sistance during the program funded by the Office of Saline Water. Thus, the
NS-100 membrane was applied to metal finishing waste solutions during the
second phase of this effort.
During Phase II, reverse osmosis tests were performed on flat-sheet and
tubular NS-100 membranes using several types of electroplating rinse water
solutions (copper and zinc cyanide, acid copper, chromic acid and Watts nickel).
The test solutions were either actual plating solutions diluted to one-tenth
full strength or simulated feed solutions that contained one-tenth the average
solute concentration of their respective plating baths. High solute rejection
of 99.8 to 99.9 percent were observed for copper, zinc, and nickel metal ions
in all cases, during test periods ranging from 210 to 540 hours. High
cyanide rejections (95.5 and 98.7 percent) were also observed for zinc and
copper feed solutions, respectively.
Membrane degradation occurred when testing chromic acid waste water at
pH 1.5. A substantial increase in flux accompanied by a decrease in rejection
was observed within 7.5 hours of testing.
It was concluded at the end of that study that the NS-100 membrane possessed
outstanding characteristics which merited further attention for treating
metal finishing effluent waste waters. First, it is chemically resistant to
both low and high pH extremes (pH 0.5 to 13.0). In this respect, NS-100
membranes surpass all commercial reverse osmosis membranes. With the exception
of chromic acid at Ph 1.5, high membrane performances for NS-100 membranes
have been reported for acid copper (pH 0.5), Watts nickel (pH 4.0), copper
cyanide (pH 11.8), and zinc cyanide (pH 12.9) plating rinse waters. Cellulose
acetate membranes, on the other hand, are operational only in the 2.5 to 7.0
pH range. The commercial polamide (DuPont) can withstand the high pH of
cyanide solutions (pH 12), but fail in the acid region below pH 4. Thus, the
NS-100 membrane greatly extends the operational pH range for reverse osmosis.
It may be possible to treat plating wastewaters without pH adjustments.
Second, the membrane could be fabricated into the tubular configuration (1.2 cm
I.D. fiber glass-epoxy tubes lined with membrane) which demonstrated its
potential for scale-up development. Third, the organic rejections for NS-100
membranes (15) exceed those of aromatic polamides membranes (16,17) and far
outstrip the performance of cellulose acetate membranes (18,19,20). Therefore,
organic additives to the plating solutions would be less likely to damage the
membrane or interfere with a recycling system. Fourth, the membrane could be
operated successfully at temperatures of up to 55°C with zinc cyanide feed
solutions.
Current Research Program
The primary objective of the current program was to demonstrate the practical-
ity of the NS-100 membrane in reverse osmosis treatment of metal finishing
waste waters through long-term operation with acid copper and zinc cyanide
baths on a time scale comparable to industrial usage.
-------
Specific objectives of this effort werej to modify the fabrication process
for NS-100 tubular reverse osmosis membranes to provide optimum performance
toward metal finishing waste waters; to demonstrate sustained performance
capabilities of the NS-100 membranes on acid and alkaline rinse waters;
and to perform preliminary engineering design studies for a possible field
demonstration unit utilizing data from these tests.
-------
SECTION IV
EXPERIMENTAL PROCEDURES
Polymers
NS-100 Membrane
A schematic diagram of the NS-100 membrane fabrication process is shown in
Figure 1. The actual barrier film consists of an alkyl-aryl polyurea formed
by the interfacial reaction of tolylene diisocyanate (TDI) with the surface
of a film of polyethylenimine (PEI) adsorbed onto a microporous support lay-
er (polysulfone). The chemistry of the membrane is illustrated in Figure 2,
The performance of the membrane is highly dependent on the thickness and
density of the PE1-TDI barrier zone.
NS-101 Membrane
This NS-101 membrane also consists of a microporous polysulfone support film
coated with PEI; however isophthaloyl chloride (IPC) is used as a crosslink-
ing agent instead of TDI. Figure 3 illustrates schematic representation of
the PEI-IPC crosslinked polymer network.
Tube Cast Membranes
The NS-100 membranes were fabricated in tubular form for use in 1.27-cm diam-
eter commercial reverse osmosis tubes (obtained from Abcor, Inc.). The micro-
porous polysulfone liner was prepared in the following manner. A 1.41-cm-I.D.
stainless steel tube was filled with a 15 percent solution of Union Carbide
P-3500 polysulfone resin in dimethylformamide (DMF). The tube was drained,
and a 1.39-cm-diameter aluminum bob was passed through the tube to provide a
uniform film of casting solution on the inside wall. The coated tube was then
lowered mechanically into 1 percent aqueous DMF in a smooth, continuous motion,
gelling the polysulfone coating. The seamless polysulfone tube was removed and
soaked in fresh water for 30 minutes or more.
The NS-100 membrane was fabricated by immersing the seamless polysulfone tube
in an aqueous solution containing 0.67 percent PEI by weight (Tydex 12, Dow
Chemical Company) for 5 minutes. Upon removal from the PEI solution, the tube
was immersed in 0.5 percent TDI in hexane for 1/2 to 1-1/2 minutes, then air-
dried.
-------
PEI IN WATER
TDI IN
HEXANE AND
HEAT CURE
c
SURFACE OF
POLYSULFONE
SUPPORT FILM
PEI COATING
PEI COATING ' PEI-TDI
CROSSLINKED REACTED
BY HEAT ZONE
Figure 1. Schematic Representation of NS-100 Membrane
-------
I
N
H
N-H-NH-+
c=o
K-H-N-
N-C
C f\ rts^
^y ^y V^
N'H
NH
o
NH
o
NH-CN
xNH
NH-c'=0
NH
H
CH2CH2 GROUPS REPRESENTED
BY
Figure 2. Idealized Structure of Polyethylenimine
Crosslinked with Tolylene 2,4-Diisocyanate
10
-------
0
-N -H-N -H-N -H- A-H-N -
NH
'=0
N
-N
c=o
N-
O
NH
NH
N
N
CH2CH2 GROUPS REPRESENTED
BY -4H-
Figure 3. Idealized Structure of Polyethylenimine
Crosslinked with Isophthaloyl Chloride
11
-------
It was then inserted into an insulated cylindrical oven equipped with zone
heating controls and thermocouples (see diagram in Appendix A) for heat cur-
ing. It was subsequently pulled into a microporous, polysulfone-coated,
1.27-cm-diameter Abcor fiber glass tube. End seals were effected by means
of rubber grommets. A detailed fabrication procedure for the NS-100 tubular
membranes is presented in Appendix A.
Reverse Osmosis System
The pilot-scale reverse osmosis test loop used in this program is illustrated
schematically in Figure 4. The system was instrumented and equipped as follows:
1. A 115-liter brine reservoir.
2. A Moyno pump, Model 3RA-8-20, equipped with a magnetic starter.
The pump had a rated capacity of 14 1pm (3.7 gpm) at 41.4 bars
(600 psig) and a pressure range of 0 to 55 bars (0 to 800 psig).
3. Low- and high-pressure safety switches to disconnect the power
from the magnetic starter via a relay. An operator must reset
the starter.
4. A 115-liter constant-temperature bath to maintain the selected
temperature of the feed. The refrigeration system was an air-
cooled type and had a capacity of a 1/3-horsepower compressor.
5. Fittings and gauges for connection of standard commercial tub-
ular-membrane modules. Lines were provided for returning pro-
duct water to the feed reservoir to permit continuous operation.
-LOW LEVEL SAFETY SWITCH^PRESSURE GAUGES
-RESERVOIR (114 LITERS)
PUMP
MOYNO 3R-8-20
TTTT
220 VAC
30
STARTER -|~\LOW-
WWWWVW\/NA> DOC
SAFETY
RELAY
PRESSURE^
SWITCH
j
TWO-FOOT R.O. TUBES
TEMPERATURE ^FLOWMETFR
CONTROL ^FLOWMETEK
BATH
NOTE: ALL HIGH PRESSURE
TUBINGS FITTINGS -AISI 316
Z V
PRESSURED P'NTCR
CONTROL NLIttx
Figure 4. Flow Diagram for Reverse Osmosis Test Loop
12
-------
6. A high-pressure filter to prevent contamination of the system
pressure control valve.
7. A Hoke needle-type throttle valve for controlling system pres-
sure.
8. A flowmeter to indicate system flow (Brooks rotameter).
9. A Weis thermometer for monitoring feed water temperature.
Photographs of this system in operation are shown in Figures 5, 6, and 7,
Figure 5. Photograph of Reverse Osmosis System Used
in Long Term Studies, Showing the Control
Panel, Flowmeter, and Throttle Valve for
System Pressure Control.
13
-------
Figure 6. View of the Reverse Osmosis Board Showing
Eight Tubes Connected in Series, with
Product Water Collection Line Attached.
14
-------
Figure 7. View of Reverse Osmosis System Showing
Feed Reservoir, Heat Exchange Reservoir,
and Refrigeration Unit (Reverse Osmosis
Pump Located Inside Plywood Sound Shield
in Foreground).
15
-------
The construction of the membrane/fiber glass tube used in this test line are
illustrated in Figures 8 and 9. Figure 8 depicts a longitudinal section of a
fiber glass reverse osmosis tube with a polysulfone support liner. Figure 9
is a photograph showing the polysulfone liner, Abcor fiber glass tube with end
fittings, and the fiber glass tube enclosed in the Tygon product water collec-
tion sleeve.
Reverse Osmos i s Testi no.
The conditions used to measure the reverse osmosis performance of the tubular
membranes during long-term testing were 41.4 bars (600 psig) pressure and 7.0
1pm (1.8 gpm) feed flow rate at a temperature of 25°C.
The feedwaters used in the long-term tests were actual plating solutions of
acid copper and zinc cyanide, each diluted to approximately one-tenth full
strength. Two acid copper baths were tested, one from Precious Metals Platers,
Inc., Hopkins, Minnesota, and the other from Superior Plating, Inc., Minneapolis,
Minnesota. Two zinc cyanide baths were tested. A mid-range zinc cyanide bath
was obtained from Superior Plating, Inc. and a low-range zinc cyanide plater
solution was obtained from Honeywell Inc., Golden Valley Plant, Golden Valley,
Minnesota.
Test Duration. 1200 hours of continuous operation with each plating
solution. Any trends in membrane performance would be noticeable in 1200 hours
of testing. If no apparent deterioration or degradation were observed during
this period ot time, it is likely that the membranes would perform satisfacto-
rily for at least 2000 to 3000 hours. Assuming that a reverse osmosis puri-
fication unit would be operated 16 hours per day, 5 days a week, the expected
lifetime for an NS-100 membrane would be 3-1/8 months for each 1000 hours of
operation.
Pressure. An operating pressure of 41.4 bars was chosen for this pro-
gram as a reasonable commercial operating range. The fiber glass-supported
tubular membranes could actually be operated at pressures up to 105 bars (1500
psig). However, added power costs for high pressure pumping begin to outweigh
water throughput improvements at about 55 bars (800 psig). Below 41.4 bars,
savings in pump costs are lost through decreased operating efficiency of the
membrane system.
ConcentratiOH. Actual plating baths (acid copper and alkaline zinc
cyanide) were diluted to 10 percent of full strength. This concentration rep-
resented a 10- to 100-fold higher level than existed in plating rinse waters.
However, it was considered essential in this study to demonstrate the membrane's
ability to withstand concentrations which would be encountered in an actual
permeate-concentrate recycling situation. These bath concentrations, at one-
tenth full strength, thus represented a severe test of the membrane system.
The use of actual plating baths was desirable to determine the effect, if any,
of organic bath additives on the membrane that would be encountered in a pilot
demonstration facility.
16
-------
RUBBER GROMMET
END SEAL
STAINLESS STEEL
BUSHING
NS-IOO MEMBRANE
POLYSULFONE SUPPORT
LINER
POLYSULFONE FILLER
BRASS BUSHING
FIBERGLASS TUBE
Figure 8. Longitudinal Section of a Fiber Glass Reverse
Osmosis Tube with NS-100 - Polysulfone Membrane
Support Composite
17
-------
Figure 9. Photograph of Tubular Polysulfone Support Liner (left),
Abcor Fiber Glass Tube with End Fittings (center), and
Fiber Glass Tube Enclosed in a Tygon Sleeve (right).
18
-------
Membane Evaluation
Water flux measurements were carried out by measuring the flow rate of the
purified water stream from a tubular reverse osmosis unit.
Rejection measurements were made using standard analytical methods as indi-
cated in Table 1. Assays were performed on the permeate from each tube and
the the feed during each analysis. The rejection was calculated as the per-
cent of the total chemical content in the feedwater returned by the membrane.
The percent rejection, R, is defined as
p - C (feed) - C (permeate)
K ~ C (feed) x 1UU
where C represents the concentration of the species being measured. More de-
tailed feedwater make-up and analyses are given in the appropriate sections
of the report.
TABLE 1. REJECTION EVALUATION TECHNIQUES FOR REVERSE
OSMOSIS MEMBRANE PERFORMANCE
Constituent Method/Equipment
Zinc Atomic Absorption/Techtron AA 120*
Copper Atomic Absorption/Techtron AA 120
Cyanide Orion specific ion electrode/Orion
digital pH meter model #701 (21)
Titration (modified Liebig method
using silver nitrate) (22)
Total Organic TOG Beckman Analyzer/Model #915
Carbon
Total Dissolved Gravimetric (23)
Solids
Acidity/Basicity Orion Digital pH Meter/Model #701
- ~ . . . . - -
Zinc atomic absorption standard solution contained sodium
cyanide and sodium hydroxide as background in the ratio of
the stock plating solution (Zn:NaCN:NaOH was 1:2.4:3.5).
19
-------
SECTION V
PROGRAM RESULTS
The results of this experimental program are divided into three separate parts;
optimization of NS-100 tube fabrication, long-term membrane performance test
results toward acid copper plating bath rinse water, and long-term membrane
performance toward zinc cyanide plating bath rinse water.
Optimization of NS-100 Tube Fabrication
In earlier work, tubular NS-100 membranes with fluxes of as high as 27 1/m2-
hr (16 gfd) at salt rejections of 99 percent were obtained. These results,
however, were more the exception than the rule. Most tubes gave fluxes in the
range of 8.5 to 14 l/m2-hr (5 to 8 gfd). It was reasoned that the perform-
ance of the membrane was highly dependent on the thickness and density of the
PEI-TDI barrier zone (see Figure 1). Therefore, the optimization study was
focused on factors which may have an effect on the thickness and density of
this layer in order that high flux membranes could be consistently fabricated.
Three fabrication variables were examined in the optimization study: the con-
centration of the TDI solution in the interfacial polymerization step, the
time of exposure to the TDI solution, and the degree of heat cure employed.
The first two factors determined the thickness of the alkyl-aryl polyurea
barrier layer on the underlying PEI layer. The third factor affected the
density of both the barrier layer and the underlying PEI layer.
Nonoptimized Membranes: Performance Towards Plating Solutions
An initial set of tubular NS-100 membranes were fabricated using conditions
representative of earlier tube fabrication work. These tubes were tested un-
der reverse osmosis conditions with an actual zinc cyanide plating solution
diluted to one-tenth its strength, and with an actual acid copper plating
solution, also diluted to one-tenth its strength. Table 2 lists the data
obtained from these tests.
The data in Table 2 illustrated the starting point for this optimization ef-
fort. Membrane flux values were 11 to 19 l/m2-hr (6.7 to 11.0 gfd) for the
acid copper bath, and 5.6 to 10 l/m2-hr (3.3 to 6.2 gfd) for the zinc cyanide
bath. These fluxes were thus both rather low and rather variable. The higher
flux rate for the acid copper bath was due to two effects: lower osmotic
strength in the acid copper bath vis-a-vis zinc cyanide bath, and some swell-
ing and opening of the crosslinked PEI matrix by salt formation between the
acid and the PEI amine groups. Rejections of zinc, copper, cyanide, and total
organic carbon were very good. There was fair-to-good rejection of sodium
hydroxide, but apparently no rejection of acid.
20
-------
TABLE 2. EFFECT OF TDI CONCENTRATION ON THE PERFORMANCE OF TUBULAR NS-100 REVERSE OSMOSIS
MEMBRANES WITH ZINC CYANIDE AND ACID COPPER PLATING RINSE WATERS
Tube
Number
338-T-4
338-T-5
338-T-6
338-T-12
TDI
Concentration
in Hexane
(percent)
0.50
0.50
0.50
0.50
Membrane Performance
Flux***
(l/m2-hr)
6.6
10.0
10.5
5.6
Zinc Cyanide Test*
Zinc
Rejection
(percent)
>99.9
99.6
99.5
>99.9
Cyanide
Rejection
(percent)
98.2
95.8
97.6
99.5
TOG
Rejection
(percent)
98
96
97
99
Permeate
PH
12.0
12.4
12.1
11.3
Acid Copper Test**
Flux***
(l/m2-hr)
12.2
18.5
18.7
11.4
Copper
Rejection
(percent)
99.8
98.8
99.6
>99.9
Permeate
PH
1.2
1.1
1.2
1.2
**
***
21-hour test on l/10th actual zinc cyanide bath, 749 ppm (0.10 oz/gal) Zn, 790 ppm (0.11 oz/gal)
CN, 1840 ppm TOG, pH 13.2.
20-hour test on l/10th actual acid copper bath, 5025 ppm (0.67 oz/gal) Cu, 28 ppm TOC, pH 1.1.
multiply by 0.59 to convert to gfd.
-------
It appeared from the very high rejections of metals and cyanide that milder
membrane fabrication conditions could probably be exercised. The preferred
optimization approach, then, was to reduce the membrane fabrication parame-
ters of time, concentration, and heat cure with the objective of maximizing
membrane permeate flux while minimizing loss of solute rejection character-
istics.
Optimization of NS-100 Membranes
Subsequent optimization studies used 1 percent sodium chloride feed instead
of the actual zinc cyanide and acid copper solutions. This change in the
testing procedure expedited the membrane optimization task, since the sodium .
chloride rejection could be conveniently determined by conductivity measure-
ments. This allowed more tubes to be fabricated and tested in a snorter
period of time. A membrane exhibiting high performance for sodium chloride
would undoubtedly yield comparable results with the plating rinse water feeds.
The sodium chloride optimization data are presented in Table 3; 30 tubes
were fabricated under varying conditions in this effort.
Examination of membrane heat cure conditions indicated best results at a
heat cure temperature of 98°C for 5 minutes. At higher temperatures, mem-
brane flux values dropped rapidly; at lower temperatures, salt rejection
fell, indicating an incomplete cure of the PEI-TDI layer.
Examination of the effect of exposure time of the PEI-coated polysulfone
tubes to the TDI reactant indicated that exposure periods of 15 seconds were
sufficient to produce a good PEI-TDI reaction product layer. Attempts to
lower the TDI concentration in hexane to less than 0.5 percent by weight re-
sulted directly in significant losses in solution rejection properties.
A minimum set of fabrication conditions thus appeared to involve exposure of
a PEI-coated polysulfone tube to a 0.5 percent TDI solution for 15 seconds,
followed by air drying and heat curing at 98°C for 5 minutes. This minimum
set of conditions was arbitrarily altered to include a 30-second dip in TDI
solution rather than a 15-second dip. In examining the mechanics of smooth-
ly immersing and removing a tubular membrane from a dip tank, a 15-second
time period appeared too short and too strict an operational variable to
control uniformly.
Under these conditions tubes could be fabricated with fluxes between 14 to
20 l/m2-hr (8 to 12 gfd) at 98 to 99 percent salt rejection, tested against
a 1 percent sodium chloride feed. This was a considerable improvement over
the 4.2 to 9.3 l/m2-hr (2.5 to 5.5 gfd) fluxes obtained at the start of the
program with the same feed.
These results were not as good as was hoped. Flat sheet fabrication studies
at North Star have consistently generated NS-100 membranes with twice this
flux. A basic problem in this system appeared to be differences in micro-
porous polysulfone films cast in 1.27-cm tubes versus those cast as flat
sheets on glass surfaces. An examination of possible changes in tube
22
-------
TABLE 3. OPTIMIZATION OF TUBULAR NS-100 MEMBRANES WITH ONE-PERCENT SODIUM CHLORIDE FEED
Number
of Tubes
Tested
4
3
5
3
6
2
3
2
2
Fabrication Parameter
TDI Concentration
in Hexane
(percent)
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.45
0.40
Heat Cure*
Temperature
(°C)
110
104
98
94
98
98
98
98
98
Exposure Time
to TDI Solution
(seconds)
60
60
60
60
30
15
10
15
15
Reverse Osmosis Performance**
_ Flux7"75 (l/mz-hr) Reiect
Average
Value
7.7
14
16
11
14
16
22
15
13
Range
4.2-9.3
10.5-16
10.5-30
8.1-14
8.1-20
15-17
15-37
12-18
11-15
Average
Value
99.2
97.6
98.2
82
98.5
98.6
93.8
94.1
89.0
Lon (%)
Range
98.8-99.6
95.2-99.1
96.5-98.9
50-98.5
97.7-99.2
98.5-98.6
88.5-97.8
92.5-95.6
86.0-93.4
ro
u>
**
***
All membranes were heat cured 5 minutes in insulated cylindrical oven at their indicated
temperature.
*
Readings were taken after 2 hours of testing.
k
Multiply by 0.59 to convert to gfd.
-------
casting procedures for seamless polysulfone tubes to improve resulting mem-
brane characteristics unfortunately could not be carried out within the
scope of this program.
A second area of concern in this optimization study related to the tight
boundaries for membrane fabrication parameters. One had to work with very
dilute TDI solutions (0.5 percent solids) at very short exposure times (15
to 30 seconds), and encountered sharp dependence of flux properties on the
severity of the following heat cure cycle. The high sensitivity of membrane
performance characteristics to small changes in these parameters leads to
considerably variability in performance between individual membrane tubes.
A partial answer to this problem is the use of isophthaloyl chloride in
place of TDI, as will be described in the following section.
NS-101 Membrane Fabrication
Despite concentrated efforts to optimize the NS-100 membrane system it be-
came evident during this program that tubes of high flux and high salt re-
jections could'-jnot be prepared in a reproducible manner. In the meantime,
related efforts on a reverse osmosis membrane contract with the Office of
Water Research and Technology showed that isophthaloyl chloride (IPC) could
be used in place of TDI and led to considerable flux improvement in flat-
sheet membranes. These modified membranes were designated as NS-101. Late
in this program the decision was made to fabricate a few NS-101 tubes and
apply them toward reverse osmosis recycle of zinc cyanide plating wastes.
Fabrication of tubular NS-101 membranes, consisting of a PEI-IPC layer as
the salt barrier zone instead of a PEI-TDI layer, was accomplished following
the same basic procedure as was developed for the NS-100 membranes. Since
time did not permit a thorough optimization study of this experimental mem-
brane system, two procedural modifications were employed to ensure the for-
mation of a tightly crosslinked PEI-IPC layer. First, a higher concentra-
tion (1 percent in hexane) of the crosslinking agent (IPC) was used to pre-
pare these membranes as opposed to the 0.5 percent TDI solution. Second,
the exposure time of the membrane to the hexane solution was increased from
1/2 to 1 minute.
Specifically, a tubular polysulfone support film was immersed in the aqueous
PEI solution (0.67 percent PEI by weight) for 5 minutes. The membrane was
subsequently exposed for 1 minute to a 1 percent isophthaloyl chloride in
hexane solution, air-dried, and heat-cured at 98°C for 5 minutes.
Table 4 illustrates the reverse osmosis performance of three NS-101 membranes
prepared in this manner with 1 percent sodium chloride. The NS-101 exhibited
improved water fluxes, however with much variation. Salt rejections were
somewhat low, ranging from 90 to 95.5 percent. Although time did not permit
a thorough investigation, it may be possible to improve the reverse osmosis
performance of this system by studying fabrication parameters such as IPC
concentration, reaction times, and heat cure conditions.
24
-------
TABLE 4. REVERSE OSMOSIS PERFORMANCE OF TUBULAR NS-101
MEMBRANES WITH ONE-PERCENT SODIUM CHLORIDE FEED
Number
of Tubes Tested
3
Reverse Osmosis Performance*
Flux (l/m^-hr)**
Average
Value
30
Range
15-31
Rejection (percent)
Average
Value
93.0
Range
90.0-95.5
**
All tests were carried out at A1.4 bars (600 psig), 25°C, 7.0 1pm (1.7 gpm)
flow rate. Readings were taken after 2 hours of testing.
Multiply by 0.59 to convert to gfd.
Long-Term Membrane Performance Toward
Acid Copper Plating Batn Rinse Water
An actual acid copper plating bath, provided by Precious Metal Platers, Inc.,
Hopkins, Minnesota, was diluted to approximately one-tenth of full strength
and used as the feed for a set of eight 2-foot-long NS-100 membrane tubes.
The test was performed for a total of 1222 hours. During this time the feed
solution was changed every 2 weeks and kept fresh by frequent additions of
diluted plating bath. Permeate was recycled back to the feed reservoir, ex-
cept during the sampling periods, to maintain constant feed concentrations.
Samples of permeate were drawn from each tube at frequent intervals and the
flux recorded. Also, pH was measured and the copper concentration determined
on the permeate and feed solution. Total dissolved solids (TDS) and total
organic carbon (TOG) analyses were performed on the permeate and feed at
longer intervals during the study. Because of insufficient weighable resi-
dues in the sample permeate, the TDS analyses were not always reliable. De-
tailed performances of each tube are listed in Appendix B of this report.
Membrane Rejecti on
Table 5 contains performance data for eight 2-foot tubular NS-100 membranes
determined at 24 hours and 1222 hours. Copper rejections were uniformly
above 99 percent in six of eight tubes. Mechanical failure was apparent
for one of the tubes (348-T-37C) which showed about 93 percent copper ion
rejection, and was suspected for the other tube (348-T-34A) which showed
about 97 percent rejection. Measurements were discontinued on 348-T037C
when it became apparent that a mechanical failure had occurred. Inspection
of this tube after the test revealed a large defect at one end, which ac-
counted for its poor performance.
Table 6 illustrates the average performance of each tube during this period.
Tube Number 348-T-43A was particularly noteworthy, exhibiting an average cop-
per rejection of >99.9 percent with an average flux of 22 l/m2-hr (12.8 gfd).
The sulfuric acid was not rejected by the membranes, however. The pH's of the
25
-------
TABLE 5. INITIAL AND FINAL PERFORMANCES OF TUBULAR NS-100 MEMBRANES WITH ACID
COPPER PLATING SOLUTION
Time
(hours)
24
24
24
48
24
24
1222
1222
1222
1009
1009
1220
Measurement
Flux (l/m2-hr)***
Concentration of Copper
(ppm)
Copper Rejection (%)
TDS Rejection (%)
TOC Rejection (%)
PH
Flux (l/m2-hr)***
Concentration of Copper
(ppm)
Copper Rejection (%)
TDS Rejection (%)
TOC Rejection (%)
PH
Tube #348-T-
30B
14
7.40
99.8
99.5
65
1.18
14
14.3
99.7
99.6
52
1.39
43A
23
4.20
99.9
98.7
70
1.17
22
4.70
>99.9
99.7
63
1.41
39F
20
17.9
99.5
99.4
74
1.14
19
12.4
99.8
99.6
67
1.39
22B
34
53.5
96.1
98.3
70
1.07
23
52.0
99.0
98.8
76
1.34
31A
24
27.0
99.3
99.1
74
1.11
21
28.0
99.5
99.2
67
1.35
28A
26
35.0
99.1
99.7
65
1.13
20
34.0
99.4
99.0
65
1.35
34A
28
184
95.3
96.7
70
1.09
22
77.5
98.5
97.1
61
1.34
37C**
30
260
93.4
96.1
78
1.13
Feed
Analysis
3950
1.17
5300
1.40
10
Feed Composition
Copper Concentration .
Total Dissolved Solids
Total Organic Carbon .
pH
3850-5850 ppm (0.514-0.781 oz/gal)
12.4-15.5 g/1 (1.65-2.07 oz/gal)
23 ppm
1.14-1.44
**
***
Tube 348-T-37C failed at 428 hours.
Multiply by 0.59 to convert of gfd.
-------
TABLE 6. AVERAGE PERFORMANCE DATA FOR NS-100 TUBES DURING THE ACID COPPER TEST
Measurement
Average Flux
(1/m -hr)*
Average
Rejection of
Copper (%)
Average
Rejection of
TDS (%)
Average
Rejection
of TOG (%)
Average pH
Number
of
Measure-
ments
15
15
5
3
15
Tube Number
348-T-
30B
14
99.8
99.6
60
1.23
348-T-
22
99.9
99.3
66
1.22
348-T-
39F
19
99.7
99.1
71
1.19
348-T-
22B
28
98.7
98.7
69
1.13
348-T-
31A
22
99.4
99.2
71
1.16
348-T-
28A
22
99.4
99.2
69
1.17
348-T-
34A
34
97.1
96.8
67
1.12
Feed
Analysis
4940 ppm
13.83 g/]
23 ppm
1.19
to
Multiply by 0.59 to convert to gfd.
-------
permeates were essentially identical to the pH of the feed. This may not be
a drawback because acid copper plating operations are normally followed di-
rectly by other acid-based metal finishing operations. Total dissolved
solids determinations confirmed the high copper rejections observed. Rejec-
tion of dissolved organic constituents, including brighteners, was in the 60
to 78 percent range. Comparing this with known rejection characteristics of
NS-100 membranes towards organic compounds (15), it appears the data imply
presence of low molecular weight organic species in the acid copper bath,
such as ethyl alcohol.
Membrane Flux
In Figure 10, the flux and copper rejection for each tube was plotted as a
function of operating time. Membrane flux varied from one tube to the next,
ranging from 14 to 30 l/m2-hr (8.3 to 17.6 gfd) at 24 hours, and 14
to 23 l/m2-hr (8.0 to 13.8 gfd) at the end of 1220 hours. The normal op-
erating range appeared to be in the 20 to 24 l/m2-hr (11 to 14 gfd) range.
After a rapid initial flux decline during the first 100 hours tubes leveled
out to relatively constant flux readings. A flux increase of about 10 per-
cent was observed for all tubes after 648 hours. At this time, due to a leak
in a pipe housing on the pump shaft, the line had been stopped and tubes had
been allowed to stand for 24 hours in contact with de-ionized water. Ap-
parently, the tubes experienced osmotic cleaning during this period since
subsequent water flux values were higher and remained so until almost the
end of the test. These results indicated that flux loss of this system due
to fouling could be restored to a significant degree by flushing with water.
Effect ofr Feed Concentration on Membrane Performance
During the long-term acid copper study, time was taken to gather data on the
effect of higher feed concentrations on membranes flux and rejection. Thus,
after 1077 hours of operation the product return line was disconnected and
the feed solution was allowed to concentrate for 10 hours. Results are shown
in Figure 11, where the average flux and average rejection of copper was
plotted as a function of copper concentration. Average membrane flux de-
creased linearly with increasing copper concentration. The average copper
rejection held at 99.3 percent and was not affected over this concentration
range.
Plating Solution
During the last 70 hours of testing, the feed was changed to a Superior Plat-
ing acid copper rinse solution. The reason for this change was to see if
the membranes would give the same performance with other acid copper solu-
tions which may possibly contain different brightener agents. The Superior
Plating acid copper solution contained Udylite UBAC #1 as the brightener ad-
ditive, whereas the organic additive for the Precision Metals bath was CUE
Bath. Feed analysis of this feed solution revealed approximately the same
concentrations of copper and total organic carbon as the Precious Metal
Platers bath. Performance results were identical for both acid copper baths.
(Data are given in Appendix B.)
28
-------
I«
96
5 30
.»
20
20
18
!16
~ 14
12
10
D-OO
=e=o=
O - TUBE #348-T-43A
D - TUBE *348-T-39F
O - TUBE #348-T-28A
6 - TUBE #348-T-34A
9
O
200 300 400 500 600 700
TIME (HOURS)
900
1000
1100 1200
Figure 10. Plot of Reverse Osmosis Performance of NS-100 Tubular
Membranes Toward Acid Copper Rinse Water. (Tubes 43A,
39F, 28A, 3AA, above, and Tubes 22B, 31A, and 30B, below.)
5 100
36
30
5-
I!*0
16
10
6 -
-O
O - TUBE #3481.31 A
D - TUBE#348-T-22B
- TUBE #348-T-30B
r
100 200 300 400 500 600 700
TIME (HOURS)
800
900
1000 1100
1200
29
-------
R
ON
E
T
)
COP
EJEC
(
O
O
CO
CO
CO
03
x 30
LI. £25
< ' 20
LU 2-
10
16
12
8
TEST CONDITIONS:
PRESSURE
pH
TEMPERATURE
FEED FLOW RATE
- 41.4 bars (600 psig)
- 1.05-1.15
- 25° C.
-7.0LPM (1.84 GPM)
5,000 10,000 15,000
COPPER CONCENTRATION (PPM)
Figure 11. Effect of Acid Copper Rinse Water Concentration
on NS-100 Flux and Copper Rejection
Summary of Results; Acid Copper Test
Results of the reverse osmosis study with acid copper feedwater may be sum-
marized as follows:
1. The NS-100 membrane demonstrated long-term stability in treat-
ing highly acidic pH (1.1 to 1.4) copper rinse water over
1220 hours.
30
-------
2. Copper rejections for six NS-100 tubes were greater than 99
percent during most of the test.
3. High rejections of IDS (greater than 99 percent) were observed
throughout the test for the high copper-rejecting membranes.
Rejection of dissolved organic constituents, including organic
brighteners, was in the 60 to 80 percent range.
4. Sulfuric acid was not rejected by the NS-100 membranes.
5. Normal water flux performance was in the 20 to 24 l/m2-hr
(11 to 14 gfd) range. Large variations in flux were observed
from tube to tube, especially at start-up.
6. Flux decline was minimal after the first 100 hours of testing.
Substantial flux could be restored by osmotic cleaning.
7. Water flux decreased linearly with increasing feed concentra-
tion; however, copper rejection remained constant over the
concentration range studied (5,000-15,000 ppm).
8. The NS-100 membrane was equally effective in treating acid
copper baths from two different plating sources.
Long-Term. Membrane Performance Toward
Zinc Cyanide Plating Bath Rinse Water
An actual zinc cyanide plating bath was provided by Superior Plating, Inc.,
Minneapolis, Minnesota, and diluted to one-tenth full strength. Ten 2-foot
tubes were tested with this solution whose pH was 12.8 at 41.4 bars, 25°C,
7.0 1pm flow rate for 1143 hours. Six tubes (30B, 43A, 39F, 22B, 31A, and
28A) were the same tubes that had already passed 1220 hours exposure toward
acid copper rinse water. Two new NS-100 tubes were prepared and were added
to the reverse osmosis zinc cyanide test after 238 hours to replace two
failed tubes. Two tubes (46A and 47B) contained the modified NS-101 experi-
mental membrane, in which isophthaloyl chloride (IPC) was employed as the
crosslinking agent instead of tolylene diisocyanate (TDI).
Permeate was collected at various times throughout the test. Membrane per-
formance parameters such as flux, permeate pH, cyanide rejection, and zinc
rejection were determined after each measurement. Total dissolved solids
and TOC analyses were performed twice on the permeate and feed, once at the
beginning and once near the end of the test. Detailed performance data for
each tube is presented in entirety in Appendix C.
Data summarizing the reverse osmosis performance of each tube during the zinc
cyanide long-term study are illustrated in Tables 7 and 8. Table 7 illus-
trates the initial and final performances of each tube at 24 and 1143 hours,
while Table 8 presents the average overall performance for each parameter for
the respective tubes during the entire test. In addition to these data, mem-
brane flux and solute rejection data are plotted as a function of time in
Figures 12 and 13. Figure 12 graphically illustrates flux, zinc rejection
and cyanide rejection data for the NS-100 membranes. Plotted in Figure 13
are the flux and rejection performance of the experimental NS-101 membranes
(46A and 47B).
31
-------
TABLE 7. INITIAL AND FINAL PERFORMANCES OF TUBULAR NS-100 MEMBRANES WITH ALKALINE ZINC CYANIDE PLATING SOLUTION
U>
N)
Time
(hours) Measurement
24 Flux U/m2-hr)**
24 Concentration of Zinc (ppm)
24 Zinc Rejection (%)
24 Concentration of Cyanide (ppm)
24 Cyanide Rejection (%)
24 pH
304 TDS Rejection (%)
304 TOC Rejection (7.)***
1143 Flux (l/mz-hr)**
1143 Concentration of Zinc (ppm)
1143 Zinc Rejection
1143 Concentration of Cyanide (ppm)
1143 Cyanide Rejection (%)
1143 pH
1044 TDS (wt.%)
Tube Number 348-T-
30B
(NS-100)
10
110
92.9
170
92.9
12.04
92.9
94.2
8.1
66.3
96.0
159
94.5
11.85
0.1429
1044 TDS Rejection (%) 93.5
43A
(NS-100)
18
2.32
99.9
19.7
99.2
11.54
98.3
98.7
14
1.40
>99.9
19.3
99.3
11.41
0.0365
98.3
39F
(NS-100)
14
20.3
98.7
48.8
97.9
11.73
97.0
97.3
12
11.8
99.3
46.3
98.4
11.58
0.0622
97.2
22B
(NS-100)
29
5.40
99.7
67.6
97.2
12.06
93.8
95.7
20
4.50
99.7
111
96.2
12.01
0.1545
93.0
31A
(NS-100)
18
12.0
99.2
44.9
98.1
11.83
96.4
96.6
15
45.5
97.2
120
95.9
28A
(NS-100)
25
9.50
99.4
54.1
97.7
11.97
95.5
96.3
14
38.5
97.7
129
95.6
11.79 ] 11.92
0.0808]' 0.1506
96.3 93.1
54A*
(NS-100)
7.8
4.50
99.8
66.9
97.6
12.15
92.8
95.3
8.7
4.50
99.7
119
95.9
12.00
53A*
(NS-100)
5.1
1.70
-99.9
24.4
99.1
11.55
97.3
97.9
5.2
1.60
>99.9
22.9
99.2
11.53
0.1382; 0.0537
93.7 I 97.6
46A
(NS-101)
i 24
42.5
97.3
236
: 90.0
12.55
76.2
83.3
23
19.8
98.8
276
90.5
12.56
0.5720J
74.0 ;
47B
(NS-101)
40
10.7
99.3
229
90.4
12.58
73.9
84.8
20
15.3
99.1
291
90.0
12.62
0.6467
70.6
Feed
Analysis*
1550 ppm
"
2380 ppm
12.74
2.621 (v
1250 ppm
--
1650 ppm
--
2910 ppm
12.71
-%)
2.198 (wt.%)
Test Conditions:
Feed Composition
1400-2100 ppm (0.187-0.281 oz/gal)
2120-3430 ppm (0.283-0.458 oz/gal)
2.198-2.621 (wt. percent)
1250 ppm
12.71-12.82
. . Zinc Concentration . .
Cyanide Concentration
Total Dissolved Solids
Total Organic Carbon .
PH
Tubes 348-T-54A and 348-T-53A are fresh NS-100 tubes. The initial data presented for these
testing and final test time was 905 hours.
Multiply by 0.59 to convert to gfd.
Total organic carbon measurements taken after 1044 hours of operation were erroneous due to instrumental problems.
tubes were taken after 66 hours of
-------
TABLE 8. AVERAGE REVERSE OSMOSIS PERFORMANCE OF NS-100 TUBES DURING ZINC CYANIDE LONG-TERM TEST
Membrane
Parameter
Average
Flux
u>
Tubes 348-T-54A are fresh NS-100 tubes. These tubes were added to the line after 238 hours of
testing had been completed oh the other tubes. Data designated by asterisk for these tubes
represents average of six readings.
**
Multiply by 0.59 to convert to gfd.
-------
O - TUBE #348-T-43A
D ' TUBE *348-T-39F
O - TUBE *348-T-2M
A - TUBE #348-T-31A
500 600 700
TIME (HOURS)
1000
1100
1200
- TUBE *34a T63A
O - TUBE «348-T«4A
O - TUBE M48-T-30B
O -TUBE «G48-T-28A
O
BOO 600
TIME (HOURS)
800
900
1000
1100
1200
Figure 12. Plot of Reverse Osmosis Performance of NS-100 Tubular
Membranes Toward Zinc Cyanide Rinse Water
34
-------
700
900
1000
1200
TIME (HOURS)
Figure 13. Plot of Reverse Osmosis Performance of NS-101 Tubular
Membranes Toward Zinc Cyanide Rinse Water
Mentorane Rej ecti on
High rejections of zinc (greater than 99 percent) vere observed for six of
the eight SS-100 tubes (43A, 39F, 22B, 31A, 53A, and 54A) during most of the
test. Cyanide rejections generally ranged between 96 and 99.4 percent, ex-
cept for two tubes (30B and 28A), whose cases will be examined later in this
report. Average rejections of zinc and cyanide at the conclusion of the test
(1140 hours) for the eight standard NS-100 membranes were 98.8 and 96.9 per-
cent respectively.
Membrane rejection of alkalinity followed cyanide rejection in that high re-
jection membranes gave lower pH permeate than low cyanide rejection membranes.
For example, tubes exhibiting cyanide rejections of 99 percent or greater ex-
hibited lower pH readings of 11.5 to 11.6 while other tubes whose cyanide re-
jection were 96 percent demonstrated pH values near 12.0. The overall perme-
ate pH for the standard NS-100 membranes vas 11.8, which indicates an average
alkaline rejection of approximately 90 percent. Trends in the total organic
carbon (TOG) and total dissolved solids (IDS) rejections also followed the
cyanide rejections closely, and were generally above 95 percent. Rejections
of TDS ranged from 92.9 to 98.3 percent for the NS-100 membranes, while TOC
rejections ranged from 94,2 to 98.7 percent. Total organic carbon measure-
ments taken after 1044 hours of operation were erroneous, due to instrumental
problems, and are not included in the data in Table 7.
35
-------
One tube carried over from the acid copper test was apparently damaged at the
start of the test. Tube #30B demonstrated low rejections for cyanide and zinc
during the entire test. Rejections of 95 percent for these two species were
not observed until 300 hours had elapsed. Failure of a second NS-100 membrane
(28A) was observed after a power failure shutdown at 466 hours. This tube had
also been carried forward from the acid copper test. Rejections of cyanide
and zinc for this tube both dropped to 93.3 percent after the incident and
remained low for the rest of the test.
The power failure also caused noticeable changes in several NS-100 tubes.
Most dramatically affected were tubes 31A and 22B. Both tubes exhibited at
this time a sharp decrease in cyanide rejection (from 0.9 to 2.7 percent) and
in zinc rejection. Resulting rejection data corresponded approximately to
their initial readings taken at 1 hour. At the same time a slight increase
in flux was observed. Both tubes quickly recovered their zinc rejections
while only 31A recovered its former cyanide rejection. Tube 31A later lost
its zinc and cyanide rejection performance after 1044 hours of use against
zinc cyanide (and 1220 hours against acid copper). This loss on solute re-
jections was not accompanied by an increase in water flux, and may indicate,
instead, contamination of the permeate by feedwater leakage through an adja-
cent mechanical connection.
Membrane Flux
Permeate flux was considerably lower with alkaline zinc cyanide feed than ex-
perienced with acid copper solution. This reflects, in part, the much higher
solids content of the zinc cyanide bath, and consequently its higher osmotic
pressure. Membrane flux ranged from 5.1 to 29 l/m2-hr (3.0 to 16.9 gfd)
(average 16 l/m2-hr) at 24 hours and from 5.1 to 30 l/m2-hr (3.0 to 11.8 gfd)
(average 12 l/m2-hr) at the end of 1140 hours of testing. The average flux
was considerably lowered by the low fluxes of the two new NS-100 tubes pre-
pared for this test. For unknown reasons, attempts to make tubes of high
flux following the optimum fabrication conditions established earlier in
the program failed.
NS-101 Membranes
The two experimental NS-101 membranes (46A and 47B) demonstrated substanti-
ally improved flux (overall average fluxes of 24 and 29 l/m2-hr (14.2
to 16.9 gfd) respectively) compared to the regular NS-100 membranes. Al-
though these tubes rejected zinc quite well (98.6 to 99.3 percent), cyanide
rejections were not too satisfactory at 91 percent. Rejections of TDS and
TOC were also low at 75 and 87.5 percent, respectively. Permeate pH measure-
ments revealed very low rejections of alkalinity for these membranes. Im-
provement in these performance characteristics would seem possible through
optimization of fabrication variables.
Discussion
The results of this study seem to indicate that a dynamic membrane was formed
on the surface of the NS-100 membranes during reverse osmosis testing. This
layer was apparently active in facilitating high solute rejections, especially
36
-------
that of cyanide. Low flux rates accompanied by high rates of flux decline
(average 15 percent) experienced by the NS-100 membranes also lends substance
to this view. The higher flux membranes were more sensitive to this effect
since they experienced higher rates of flux decline and greater changes in
rejection performance during the first 24 hours of testing. The decrease in
rejection values and concurrent flux increases observed after shut-down were
most likely the result of osmotic cleaning, which would have occurred after
the feed pressure was released. The dynamic layer was apparently removed
when the flow of water was reversed, driven by the normal osmotic pressure
gradient.
The formation of this dynamic layer on the surface of the NS-100 membrane
may arise from two possible sources: deposition of colloidal metal hydrox-
ide impurities, and adsorption of organic additives present in the plating
solution. Analysis on the feed solution during the test revealed total or-
ganic carbon concentrations as high as 1250 ppm (excluding cyanide). The
charged nature (cationic) of the NS-100 membranes may act to facilitate de-
position of organic materials on its surface. This effect would likely be
more significant in a feed stream which is highly concentrated with respect
to charged ionic and organic species, such as the one used in this test.
Summary of Results: Alkaline Zinc Cyanide Test
Results of the reverse osmosis study with alkaline zinc cyanide feedwater
may be summarized as follows:
1. The NS-100 membrane demonstrated long-term stability in treat-
ing this highly alkaline (pH 12.8) zinc cyanide waste solution.
Four NS-100 membranes demonstrated high zinc and cyanide rejec-
tions for (1140) hours after completing 1220 hours of testing
with acid copper at pH 1.1 to 1.4.
2. Five of the eight NS-100 membranes demonstrated zinc rejections
of greater than 99 percent during most of the test. Two tubes
demonstrated rejections of 99.6 and 99.9 percent during the
entire study.
3. Cyanide rejections were generally greater than 96 percent, with
one tube as high as 99.4 percent.
4. High rejections of organic species were observed and closely
followed the cyanide rejection trend. Rejection of total or-
ganic carbon and total dissolved solids were generally above
95 percent.
5. Alkaline rejection also resembled the pattern set by the cyanide
rejection. High rejecting cyanide membranes yielded permeate
with lower pH values.
6. Typical flux operation range for the NS-100 membrane during the
zinc cyanide study appeared to be in the 8.5 to 15 l/m2-hr
(5 to 9 gfd) range.
7. NS-101 membranes also demonstrated high zinc rejections and at
twice the permeate flux,but cyanide rejections, at about 90 per-
cent , needed improvement.
37
-------
SECTION VI
REFERENCES
1. Nelson, B. R., Rozelle, L. T., Cadotte, J. E., and Scattergood, E. M.;
Use of Reverse Osmosis for Treating Metal Finish-ing Effluents; Final
Report, EPA Program No. 12010 DRJ; U. S. Government Printing Office;
Washington, D. C. (November, 1971).
2. Rozelle, L. T., Kopp, C. V., Jr., Cobian, K. E.; New Membranes for Re-
verse Osmosis Treatment of Metal Finishing Effluents; Final Report, EPA
Program No. 12010 DRH; U. S. Government Printing Office, Washington,
D. C. (December, 1973).
3. Pickering, Q. H., and Henderson, C.; "The Acute Toxicity of Some Heavy
Metals to Different Species of Warmwater Fishes"; Air and Water Poll.
Int. J., 10, 453 (1966).
4. Interaction of Heavy Materials and Biological Sewage Treatment Processes;
Public Health Services Publication No. 999-WP-22 (1965).
5. Shea, J. F., Reed, A. K., Tewksbury, T. L., and Smithson, G. R., Jr.;
A State of the Art Review on Metal Finishing Waste Treatment; Federal
Water Quality Administration; U. S. Department of the Interior; Program
No. 12010 EIE 11/68; Grant No. WPRD 201-01-68 (November, 1968).
6. Dobb, E. H.: "Metal Wastes, Contribution and Effect"; Tech. Proc. Amer.
Eleatroplaters Soc., 45, 53 (1958).
7. Donnelly, R. G., Goldsmith, R. L., McNulty, K. J., Tan, M., "Reverse
Osmosis Treatment of Electroplating Waste"; Plating, 61, 432 (1974).
8. Golomb, A.; "Application of Reverse Osmosis to Electroplating Waste
Treatment, Part I: Recovery of Nickel"; Plating, 57, 1001 (1970).
9. Golomb, A.; "Application of Reverse Osmosis to Electroplating Waste
Treatment, Part II: The Potential Role of Reverse Osmosis in the Treat-
ment of Some Plating Waste"; Plating, 59, 316 (1972).
10. Golomb, A.; "Application of Reverse Osmosis to Electroplating Waste
Treatment, Part III: Pilot Plant Study and Economic Evaluation of Nickel
Recovery"; Plating, 60, 482 (1973).
11. Golomb, A.; "Application of Reverse Osmosis to Electroplating Waste Treat-
ment, Part IV: Potential Reutilization of Chromium Plating Waste by Other
Industries"; Plating, 61, 931 (1974).
38
-------
12. Spatz, D. D.; "Electroplating Waste Water Processing with Reverse
Osmosis"; Product Finishing, 25, 79 (1972).
13. Reid, C. E., and Breton, E. J.; Appl. Polym. Soi., 133, 1 (1957).
14. Loeb, S., and Sourirajan, S.; Advances in Chemistry Series No, 38;
American Chemical Society, Washington (1963).
15. Cadotte, J. E., Kopp, C. V., Cobian, K. E., and Rozelle, L. T.; In Situ-
Formed Condensation Polymers for Reverse Osmosis Membranes: Second Phase;
Office of Saline Water Research and Development Progress Report No. 74-
982; U. S. Government Printing Office; Washington, D. C.; p. 59 (June
1974).
16. Matsuura, T., Blais, P., Dickson, J. M., and Sourirajan, S.; J. Appl.
Polym. Sai., 18, 3671 (1974).
17. Dickson, J. M., Matsuura, T., Blais, P., and Sourirajan, S.; ibid., 19,
801 (1975).
18. Matsuura, T. and Sourirajan, S.; ibid., 15, 2905 (1971).
«
19. Matsuura, T. and Sourirajan, S.; ibid, 16, 1663 (1972).
20. Matsuura, T. and Sourirajan, S.; ibid, 17, 3661 (1973).
21. Frant, M. S.; "Application of Specific Ion Electrodes to Electroplating
Analyses"; Plating, 58, 686 (1971).
22. Stevens, F., Fischer, G., and MacArthur, D.; Analysis of Metal Finishing
Effluents; Robert Draper Ltd. Teddington; p. 12 (1968).
23 U. S. Environmental Protection Agency; Methods for Chemical Analysis of
Water and Waste; 16020 (1971).
39
-------
SECTION VII
APPENDICES
APPENDIX A
Fabrication Procedure for Tubular NS-100
Membranes for Reverse Dsmpsi s
The general procedure for the NS-100 membrane tubular fabrication is out-
lined in Figure Al. The tubular fabrication process is divided into three
basic areas: the modification of Abcor fiber glass tubes to accommodate the
NS-100 membrane, the fabrication of the tubular NS-100 membrane, and the
assembly of the reverse osmosis tube containing the NS-100 membranes.
Each of these steps are discussed in detail in the procedures section, below.
Safety Precautions
Tolylene 2,4-Diisocyanate
Vapors of tolylene 2,4-diisocyanate (TDI) are very toxic. Extreme care must
be exercised at all times to avoid inhalation of TDI fumes. The handling of
this compound should be restricted to a well-ventilated area such as a walk-
in hood.
EPI-CURE 8494 (Celanese)
This composite blend consists of a variety of aromatic amines which are ir-
ritants to the skin as well as being carcinogens. This curing agent should
be handled with protective gloves.
40
-------
MEMBRANE FABRICATION
Cast Polysulfone
Support Liner
FIBER GLASS TUBE
MODIFICATION
Abcor Fiber Glass
Tube Cut Into
24-Inch Sections
End Fittings
Attached to
Fiber Glass Tube
Fiber Glass Tube
Impregnated
With Polysulfone
Residual Solvent
Leached From
Support With Water
Liner Immersed In
Aqueous PEI Solution
Excess PEI Solution
Drained from Liner
Liner Immersed in
TDI/Hexane Solution
TUBE ASSEMBLY
Membrane Inserted
Into Tube
Membrane Inspected
Membrane Sealed
With Rubber Grommet
End Seals
Figure Al. General Outline of NS-100 Reverse Osmosis Tube Fabrication
-------
TABLE Al. APPARATUS AND REAGENTS FOR TUBE FABRICATION
EQUIPMENT
Item
Aluminum bob
Number Dimensions
1 I.D. - 1.33 cm
Aluminum bob 1
Stainless steel 1
tube (inside
polished to near
mirror finish)
EC Mototnatic 1
Motor Control
(Model //E550M)
ED Motomatic D.C. 1
Servo Motor-
Generator (Model
#5503)
Metal tube
Glass tube
Oven
Dry-Air Drier
CONSTRUCTION MATERIALS
Item Number
Fiber glass tube 1
(Abcor, Inc.,
Cambridge, Mass.)
Abcor rubber grommet 2
and plastic inserts
Brass bushing 2
Stainless steel (316)
reducing bushing
Purpose
Impregnating fiber glass
tube with polysulfone
I.D. - 1.39 cm Casting polysulfone liner
Length -91.4 cm Casting polysulfone liner
I.D. - 1.41 cm
Casting polysulfone liner
Casting polysulfone liner
Immersion tank for poly-
sulfone liner casting
PEI and TDI immersion
tanks
Membrane heat cure
Humidity control for poly-
sulfone liner casting room
Purpose
R.O. tube housing unit for
membrane
Membrane end seal with
R.O. tube
Allows attachment of sleeve
for permeate collection
Tube end fittings
1
2
1
1
Length
I.D.
Length
I.D.
Length
- 1.02 m
- 3.49 cm
-91.4 cm
- 3.49 cm
-76.2 cm
Dimensions
Length - 1.52 m
I.D. - 1.35 cm
O.D. - 1.55 cm
Length - 1.27 cm
I.D. - 1.59 cm
O.D. - 2.54 cm
Male NPT Pipe
Size - 1.27 cm
Inside bored out
to 1.59 cm I.D.
42
-------
Item
Tygon tubing
Adjustable hose
clamps
TABLE Al. (Continued)
Number Dimensions Purpose
Length - 54.6 .cm Collects product water
I.D. - 2.54 cm
Fastens permeate collection
sleeve to brass bushing
REAGENTS
Item
Polysulfone (Union
Carbide P-.3500
Tydex 12 (Dow
polyethylenimine)
Epon 828 (Shell)
DER 736 (Dow)
EPI-CURE 8494*
(Celanese)
Tolylene 2,4-
Diisocyanate (TDI)*
N,N-Dimethyl formamide
Hexanes
De-ionized water
Grade
Practical
Reagent
Reagent
Purpose
Liner preparation
NS-100 membrane formation
Epoxy ingredient
Epoxy ingredient
Epoxy ingredient
NS-100 membrane formation
Solvent for polysulfone
Solvent for TDI
Solvent for polyethylenimine
See "SAFETY PRECAUTIONS" Page 41.
43
-------
PROCEDURE
Fiber Glass Tube Modification
Operation
2.
End fittings for the tubes are
made by boring out stainless
steel threaded reducing bushings
(1/2-inch Male NPT pipe size) to
1.59 cm in diameter.
Comments
1. This allows a smooth, snug fit
between the tube and fittings.
Abcor fiber glass tubes (1.55 cm 2.
in diameter) are cut into sections
61 cm in length.
The ends of the tube may have to
be sanded in order to fit the
brass and stainless steel bushings.
3. An epoxy blend for sealing the 3.
metal bushings to the tube is pre-
pared by mixing 7.0 grams Shell
Epon 828, 3.0 grams Dow DER 736 and
4.5 grams Celanese EPI-CURE 8494.
4. The metal bushings are epoxied on- 4.
to the ends of the fiber glass tube
in the manner shown in Figure 8 of
the text. The brass fitting is
placed into position first,
followed by the threaded stain-
less steel bushing.
The epoxy formulation can be mixed
in disposable aluminum trays.
Ingredients are thoroughly mixed
and allowed to stand for about 1
hour at room temperature.
Epoxy solution should be applied
liberally to the area of the fiber
glass tube where the bushings are
to be positioned. It is also
applied to inside surfaces of the
bushings themselves. Special
care must be taken to insure that
the epoxy adhesive is applied to
all surface areas in contact be-
tween the stainless steel bushing
and the fiber glass tube. Excess
epoxy is wiped from the tube before
it sets so it will not interfere
with the positioning of the rubber
gommet end seals later in the
fabrication process. The stainless
steel bushing is allowed to extend
0.32 cm over the end of the fiber
glass tube. This creates a shelf
onto which the rubber grommet end
seal can rest.
-------
Operation
Comments
5.
5. Impregnation of the fiber glass
tube with polysulfone is accom-
plished by plugging one end of
the tube with a rubber stopper
and filling the tube with a
20 percent solution of poly-
sulfone in N,N-dimethyl
formamide (DMF). The tube is
allowed to stand in the vertical
position until the solution
seeps through the walls of the
tube (from 1 to 10 minutes,
depending on porosity of the
fiber glass tube). Once the
polysulfone has seeped through
the walls of the tube over all
areas in the bottom half of the
tube, a rubber stopper is placed
at the top of the tube. The
tube is then inverted. Poly-
sulfone solution is added to
replenish the solution which
seeped out. When the solution
has penetrated all areas of the
tube the rubber stopper is re-
moved from the bottom and the
solution is allowed to drain into
a collection pan. An aluminum
bob 1.33 cm in diameter is
passed through the tube. The tube
is then immersed at a uniform rate
into a de-ionized water quench bath.
[This operation can be performed by
hand.] After soaking 15 minutes
the tube is removed from the first
quench bath and placed in a fresh
de-ionized water bath for 4 hours.
The modified fiber glas tube is
then air-dried and ready for use.
The polysulfone is applied to the
fiber glass tube to provide a
smooth uniform surface of even
porosity. The tube should be
held at a tilted angle so the
solution can be poured slowly
down the side of the tube. This
avoids the entrapment of air
bubbles.
45
-------
Membrane Fabrication
Operation
1.
Comment
2.
3.
4.
Prepare a 15 percent poly-
sulfone in DMF solution.
Filter the solution through a
Seitz filter to remove any par-
ticulate matter.
1.
A polysulfone liner is cast in a
polished stainless steel tube
(1.41 cm in diameter). One
end of the stainless steel tube,
thirty-six inches in length, is
plugged with a rubber stopper.
The filtered polysulfone solution
is poured slowly down the side of
the tube through the open end.
After the solution settles (one
minute) it is drained through
the bottom of the tube. An alu-
minum bob 1.39 cm in diameter
passed through from top to bottom
at the same time. The stainless
steel tube is immediately lowered
at a. uniform rate 10.2 to 15.2 cm/
second (4 to 6 inches/second)into
an aqueous 2-percent DMF bath. The
polysulfone is gelled immediately
as it contacts the water solution.
The freshly prepared polysulfone
liner is allowed to stand in the
aqueous DMF quench bath for
l/2-hout» It is then removed
and thoroughly rinsed by soaking
in de-ionized water for at least
1 hour.
The polysulfone liner is removed
from the stainless steel tube and
immersed for 5 minutes in an
aqueous solution of polyethylen-
imine (PEI) that is 0.66 percent
solids by weight.
2.
3.
4.
Time can be minimized by adding the
polysulfone pellets to hot DMF
while stirring rapidly. The mix-
ture is stirred at approximately
110°C tto 120°C for about 1 hour
(until solution occurs). The hot
solution is easily filtered. If
heated, the polysulfone solution
must be cooled to room temperature
before use.
This step must be carried out under
low humidity conditions. The poly-
sulfone solution can be used
several times. The solution be-
comes cloudy after exposure to
moisture in the air over long periods
of time (1 to 2 days); however,
it can be restored by heating at
110°C to 120°C for about 1 hour
Erratic immersion rates will produce
liner defects. Uniform immersion
rates can be accomplished mechanically
by the use of an EC Motomatic Motor
Control and an EC Motomatic D.C.
Servo Motor-Generator set-up.
It is important that all DMF is
removed from the polysulfone support
film. Residual DMF adversely
affects later fabrication steps.
The operation must be performed on
a wet support film. Once the poly-
sulfone is dried it is not receptive
to water and cannot be adequately
coated by PEI.
46
-------
Operation
Comment
The polysulfone support is
removed from the PEI solution
and excess water allowed to drain
off.
5. The polysulfone support is 5. The support cannot be dried at this
point because the water within the
pores protects the polysulfone
from TDI attack during the sub-
sequent steps. The aqueous PEI
solution is stable for periods
up to 2 to 3 weeks; however,
the solution has to be kept clean
by frequent filtrations.
6. The PEI coated support is immersed 6. Caution must be exercised when
in a 1/2 percent TDI in hexane
solution for 1/2 minute.
7.
8.
9.
Membrane is air dried.
Membrane is inspected for pin
holes by holding it next to a
strong fluorescent lamp in a
dark room and siting through it.
Membrane is heat cured at 98°C
for 5 minutes.
7.
8.
9.
Assembly of NS-100 Reverse Osmosis Tube
1. The NS-100-polysulfone composite 1.
membrane is pulled into the
modified Abcor fiber glass tube
by taping to a wooden dowel
and pulling the dowel through.
2. The reverse osmosis tube is pulled 2.
into a 1-inch diameter Tygon
sleeve (for water collection).
The ends of the sleeve are
fastened to the brass bushings
with adjustable hose clamps.
A small hole (0.64 cm I.D.) is
punched near one end of the tube
to allow drainage of product water.
working with TDI, which is very
toxic. This step should be
carried out under a properly
ventilated hood. Fresh TDI-hexane
solutions must be prepared each day.
In the presence of water TDI reacts
slowly with itself to form an
insoluble urea.
Thorough liner inspection cannot
be accomplished unless the liner
is dry, since pinholes do not
show up when liner is wet.
The cylindrical oven used for
membrane heat cure during this
program is illustrated in Figure A2.
Other oven designs may be equally
practical.
The membrane is pulled into the
tube slowly to avoid damage
resulting from tearing.
47
-------
Operation Comment
A small Tygon hose (0.95 cm
O.D.) is inserted into this hole
and fastened to the sleeve by
solvent fusion with cyclo-
hexanone.
3. The membrane is fastened to the 3. A trace of Vaseline lubricant is
fiber glass tube with Abcor applied to the grommet where
rubber grommet end seals. The contact is made with the membranes.
membrane seal is made when a
plastic expander is inserted
into the rubber piece. Hollow
metal spacers inserted into
stainless steel fitting on
reverse osmosis line hold the
plastic expanders in place.
4. The reverse osmosis tube con- 4. High performance NS-100 membranes
taining the NS-100 membranes will suffer degradation when
is stored in de-ionized water exposed to air for long periods
until testing. of time.
48
-------
COPPER TUBE 3.81 CM ID
NS-100 POLYSULFONE
MEMBRANE
HEATING
33.6 CM ZONE
III
HEATING
38.1 CM ZONE
I
COPPER CONSTANTAN
THERMOCOUPLE
(Type TG-36-ODT)
THERMOCOUPLE
SAUEREISEN 33
INSULATION
THERMOCOUPLE
Figure A2. Cylindrical Oven
49
-------
APPENDIX 8
Individual Membrane Performance Data with Acid Copper
Plating Bath Rinse Maters and Feed Analyses
Test Condition!: Pressure . . .
Tenperature
Feed Flov Rate
Feed
. 41.4 bars (600 pslg)
. 25'C
. 7.0 Ipn
. One-tenth actual acid
copper plating bath
Table Bl. Acid Copper Feed Analysis
Time
(hours)
1.0
26
68
96
144
263
309
335
628
502
602
649
797
890
1009
1077
1081
1088
1152
1222
Copper
Concen-
tration
(ppm)
3850
3950
4950
4450
___
5850
4850
5150
4650
SOOO
...
4700
5500
5750
7500
12600
5300
PH
1.24
1.17
1.18
1.16
-.-
1.16
.
1.14
1.17
1.20
1.27
1.18
1.15
1.13
1.05
1.40
Total
Dis-
solved
Solids
(Wt, %)
1.320
...
1.548
1.244
1.263
-- .
1.420
Total
Organic*
Carbon
(ppm)
23.0
...
...
....
22.5
23.0
18.0
Bath Source
Precious Metal
Platers. Inc.
"
"
"
11
11
"
11
"
"
"
"
"
"
"
"
"
"
ing. Inc.
"
Comments
Shut-down to change line filter
Changed feed
Shut-down due to equipment fail-
ure (22 Hours)
Changed feed
Changed feed
Feed solution allowed to concen-
trate
"
"
Changed eed
*Preciou
agent.
Superior
Metal Platers stock plating soluti<
Plating stock plating solution cont,
.on contained CUE Bath as the organic brightening
ained Udyllte UBAC 11 as the organic brightening agent.
50
-------
TABLE B2.
INDIVIDUAL MEMBRANE PERFOMIANCE5 WITH ACID COPPER PLATINS
BATH RINSE WATER
u
£
1
o
§
348-T-
30B
e
o t:
M
c n
4J
Is
I
1?
k*
3
§
HI
S
H
1.0
24
48
96
263
335
428
502
602
797
1009
1077
1081
L088
1152
1122
X
g
S **
P
15.4
14.1
14.3
14.4
13.7
14.1
13.9
13.6
13.4
13.9
13.3
13.1
12.8
11.2
13.6
i.
57
u 5.
±> &.
n
Is.
ft.
23.2
7.4
7.1
8.8
8.0
5.7
6.5
8.5
8.5
9.2
11.2
22.6
12.7
21.8
14.3
-u
3
99.4
99.8
99.9
99.8
99.9
99,9
99.9
99.8
99.8
99.8
99.8
99.6
99.8
99.8
99.7
n
o.
V
I
1.22
1.18
1.22
1.23
1.23
1.20
1.24
1.24
1.32
1.24
1.21
1.20
1.10
1.39
01
Bjs
O
u
&
0.0063
0.0090
0.0038
0.0092
0.0062
g
H
S
a»^
W
B
99.5
99.7
99.7
99.3
99.6
0
8
51
|£
1
8
8
11
9.0
1
*<
«? tt
a *^
g
65
64
52
50
8
M
&
£
*
Precious Metal
Platers , Inc
"
ir
t«
ii
"
M
u
n
"
ii
11
11
t*
Superior
Plating, Inc.
5
.5
Jg
(9
I
348-T-
43A
S*
H
C (1
H (S
4-1
99.9
>99.9
>99.9
>99.9
>99.9
>99.9
>99.9
99.9
>99.9
>9».9
>99.9
>99.9
>99.9
»
V
CA
fi
4*
1.21
1.17
1.19
1.21
1.21
1.20
1.23
1.23
1.32
1.24
1.21
1.19
1.09
1.41
e~
* .
?u
o
S
0.3172
.
0.0032
0.0138
0.0064
0.0047
5
-U
U
«
^ ^
5ti
(W s^t
n
g
___
98.7
99.8
98.9
99.5
99.7
8
* 'H'
fl EL
fj >5?
B.
7
8
8.5
6.5
|
Cl
n^
f6 *-"
g
~
70
64
63
64
0
u
A
uS
a
u
PraciouB Metal
'later* t Inc.
"
M
II
II
"
M
II
"
"
II
II
II
II
Superior
Plating, Inc,
51
-------
n
Q)
A
£
0)
,0
3
H
348-T-
39 F
a n:
S£
4J
^ *
OT O
0 *
3
03
g
5,
S
«H
H
24
48
96
263
335
428
502
602
797
1009
1077
1081
1088
1152
1222
x
S
Iw
01 *?
*Ts
a> J
| s
B!
20.2
19.9
19. 9
19.2
19.5
19.5
19.2
19.2
19.7
18.7
19.0
18.2
16.4
19.2
a
a B.
u a
C at
CM
17.9
13.0
12 1 0
8.8
6.8
8.0
9.4
10.8
9.7
9.8
12.7
3.2
20.2
12.4
c
*-
S
0! v*
3
O
99.5
99.7
99 . 7
99.8
99.8
99.8
99.8
99.8
99.8
99.8
99.8
>99.9
99.8
99.8
X
a.
u
g
at
&H
1.14
1.17
1 17
1.17
1.18
1.23
1.19
1.27
1.21
1.16
1.16
1.06
1.39
Q
at
| *"
V
Bu
0.0073
0.0056
0.0260
0.0063
...
a
5
S
a?«?
a: *
c/i
&"
99.4
99.6
97.9
99.6
___
g
u ^
4^ cj
a 5.
6 "^
Cu
6
6
7.5
6.5
a
o
H
U
0}
1?M
u
fr-*
74
73
67
64
H
U
0>
M >-»
9
U
97.5
96.1
98.9
98.7
99.2
99.2
99.3
99.0
99.0
99.0
99.0
99.0
99.1
98.9
99.0
X
a
u
a
y
a
41
h
1.13
1.07
1.11
1.11
1.11
1.14
1.16
1.14
1.23
1.15
1.09
1.07
0.98
1.34
en
H
41
4J
a w
g "^
a
0.0226
0.0187
0.0196
0.0167
0.0176
|
iJ
U
U
W
H
98.3
98.8
98.4
99.0
98.8
0
°
0) *-
<9 §.
g ^
a
7
9
5.5
7.5
|
& si
o
H
70
60
76
58
01
u
lJ
a
&
u
Precious Metal
Platers, Inc.
"
"
"
"
"
"
H
II
"
II
..
Superior Platlnj
Inc.
II
52
-------
4)
i
348-T-
31A
Position On
R.O. Board
5
___
CO
Vt
1
1
24
48
96
263
335
428
502
602
797
1009
1077
1081
1088
1152
1222
X
9
Permeate Fl
(l/m2-hr)
28.7
24.4
23.9
23.6
21.1
20.9
20.9
20.0
20.0
22.8
21.2
21.7
20.5
18.0
20.9
I
a
Permeate Co
per (ppm)
80.5
27.0
26.3
36.5
20.3
14.8
17.0
18.8
22.0
23.3
26.8
30.0
38.3
65.5
28.0
a
3
1
£S
3
97.9
99.3
99.5
99.2
99.7
99.7
99.7
99.6
99.6
99.5
99.5
99.5
99.5
99.5
99.5
P3
CL
4)
U
£
1.17
1.11
1.14
1.15
1.19
1.16
1.17
1.17
1.25
1.20
1.12
1.10
1.00
1.35
a
Permeate TD
(Wt. Z)
0.0117
0.0100
0.0084
0.0143
0.0108
§
H
U
S
TS
S'-'
99.1
99.4
99.3
98.9
99.2
__
0
Permeate TO
(ppm)
-_
6
6
7.5
6.0
§
H
U
3
SB
S
74
73
67
67
1
S
Preclcua Metal
Platers, Inc.
11
11
11
11
11
"
11
"
"
u
M
tl
II
Superior Flatini
Inc.
be Number
s
348-T-
28A
c
P
H >
[A O
0
6
9
H
24
48
96
263
335
428
502
602
797
1009
1077
1081
1088
1152
1222
s
f
4-t W
£
26.0
25.8
25.0
22.4
21.9
21.4
21.4
21.1
22.1
21.2
21.5
20.5
18.7
20.5
L
h
4J 0.
It
-------
u
x
«
348-T-
34A
Position On 1
R.O. Board 1
1
"
M
I
Ii
24
48
96
263
335
426
502
602
797
1009
1077
1081
1088
1152
1222
Permeate Flux
(iV-hr)
28.2
27.7
26.8
24.6
24.1
22.4
22.8
21.7
23.9
22.8
23.2
27.2
19.9
21.5
*1
U O.
<9 **
Ss
*
184
142
160
178
126
106
132
112
128
140
183
300
77.5
Cu Rejection
(Z)
95.3
97.1
96.4
97.0
97.0
97.9
97.2
97.8
97.7
97.6
97.6
97.6
98.5
5.
«
*J
!
1.09
1.08
1.10
1.12
1.11
1.12
1.13
1.21
1.14
1.08
1.06
0.96
1.34
ta
!«
4J
1*
£
0,0432
0.0544
0.0411
0.036O
0.0412
TDS Rejection
(Z)
96.7
96.5
96.7
97.1
97.1
-_-
1
31
1*
£
7
7
9.0
5.5
TOC Rejection
(Z)
70
69
61
69
a
1
1
Precious Metal
Platers, Inc.
11
"
11
11
"
n
ti
"
"
"
ii
"
Superior Plttin;
Inc.
n
Tube Number
34B-T-
37C**
Position On I
R.O, Board 1
B
Time (Hours)
1
Iti
48
96
263
335
428
Tea tli
aeabit
Permeate Fluxl
-------
Individual Membrane Performance Data *rtth Alkaline Z1nc Cyanide
Plating Bath Rinse Waters and Feed Analyses
Teat Conditions: Pressure 41.4 bars
Temperature ..... 25*C
Feed How Rat* .... 6.8 1pm
Feed One-tenth actual cine
cyanide plating bath
TABLE Cl. ZINC CYANIDE FEED ANALYSIS
Zinc Cyanide Feed Analysis
(Hour.)
1.0
24
120.5
304
362
466
499
666
827
881
1044
1143
Cyanide
Concen-
tration
(ppm)
2120
2380
2520
2820
3280
3430
2510
2750
2910
Zinc
Concen-
tration
(ppm)
1400
1550
1750
2100
1800
1650
1550
1550
1650
Total
Organic*
Carbon
pH (ppn)
12.80
12.74
12.75
12.74 1250
12.80
12.82
12.80
12.77
12.73
12.71
Total
Dla-
eolved
Solid.
Oft. I)
2.621
2.198
Convent .
Changed feed
Shut-down due to power
failure
Changed feed and line
filter
...
Stock plating aolution contained Enthone Q-540 ae the organic brighten.! bath additive.
55
-------
TABLE C2.
INDIVIDUAL MEMBRANE PERFORMANCES WITH ALKALINE ZINC CYANIDE
PLATING BATH RINSE WATER
1
i-l
O> V
I*
30B
43A
Position on
R.O. Board
1
2
s**
w
Jh
g
H .C
1
24
304
498.6
666.0
881
1044
1143
1
304
498.6
1044
1143
Permeate
Flux
(l/in2J»r)
11.3
9.8
8.8
7.8
8.1
8.0
8.5
8.2
8.1
20.0
18.0
15.9
14.1
14.9
14.4
15.3
14.7
14.3
Permeate
Cyanide
(ppm)
260
170
138
160
150
153
17.5
20.6
18.8
19.3
kz
n
B3
87.7
92.9
95.1
95.1
95.6
94.4
99.4
99.4
99.3
99.3
Permeate
Zinc
(ppra)
172
110
90
65.0
50.0
2.40
1.80
1.35
1.40
8*
I-l1
O
01 §
S3
87.7
92.9
95.7
96.4
96.8
99.9
99.9
>99.9
>99.9
Perneate
pH
12.23
12.04
11.96
11.95
11.88
11.51
11.48
11.41
11.41
Permeate
TOC (ppm)
71.3
**
16.4
**
I
o
01 ^
1-1 w
J«
O *H
94.2
**
98.7
**
at i-*
S "
t°i
v C-
t*
0.1871
0.1429
0.0448
0.0365
I
a
V *-*
3*.
w o
92.9
93.5
98.3
98.3
z
39F
22B
Position onl
R.O. Board
3
4
*~*
CO
Is
1
24
120.5
304
498.6
666.0
1044
1143
1
24
304
498.6
1044
1143
Permeate
Flux
(l/m2-hr) 1
34.7
14.4
12.9
11.5
12.1
11.9
12.3
12.0
33.8
28.7
22.2
22.6
20.7
20.0
Z 0)
<8 -O '->
Hi
& 5
65.1
48.8
42.5.
44.7
58.2
54.3
56.9
46.3
77.7
67.6
65.0
106
103
111
i
y *-*
a; H
«:' ***
96.9
97.9
98.3
98.4
98.2
98.4
97.9
98.4
96.3
97.2
97.7
96.8
96.3
96.2
Permeate
Zinc
(ppm)
20.3
18.0
19.3
14.0
14.3
11.8
11.8
6.64
5.40
5.40
4.90
4.60
4.50
t
98.7
99.0
99.1
99.2
99.1
99.2
99.3
99.5
99.7
99.7
99.7
99.7
99.7
Permeate
PH
11.86
11.73
11.71
11.70
11.69
11.68
11.59
11.58
12.20
12.06
12.06
12.10
12.04
12.01
Permeate
TOC (ppm)
33.4
**
54
**
k~
gJ
97.3
**
95.7
**
01 <~*
S w "
e»
0.0790
0.0622
0.1625
0.1545
i
« ^
"iS
VI O
97.0
97.2
93.8
93.0
56
-------
V 01
§»
H
31A
28A
Position on
R.O. Board
5
6
Time
(hours)
1
24
120.5
304
498.6
666.0
881.2
1044
1143
1
24
me
J
304
498.6
666.0
881.2
1044
1143
Permeate
Flux
(l/m2-hr)
20.7
18.2
16.1
14.5
15.5
14. B
15.9
15.2
15.1
20.7
25.1
t C Q
A3( 7
14.1
15.6
14.9
15.8
14.7
14.4
Permeate
Cyanide
(PPn)
73.5
44.9
40.3
42.3
125
65.3
52.0
52.5
120
62.5
54.1
Aft 9
**O. 1
48.9
221
202
172
137
129
is
si
96.5
98.1
98.4
98.5
96.2
98.1
97.9
98.1
95.9
97.0
97.7
Ml
Ji
98.3
93.3
94.1
93.1
95.0
95.6
Permeate
Zinc
(ppm)
2.60
12.0
10.3
12.3
32.5
12.0
9.6
10.5
45.5
10.0
9.50
7«e
fJ
9.10
120
100
77.5
43.5
38.5
n**
3%
98.1
99.2
99.4
99.4
98.2
99.3
99.4
99.3
97.2
99.3
99.4
QQ d
77 fO
99.6
93.3
93.9
95.0
97.2
97.7
0)
u
11.97
11.83
11.83
11.80
11.80
11.79
11.70
11.73
11.79
12.04
11.97
UQA
T*
11.73
12.11
12.09
12.00
11.95
11.92
"i
u S.
43.0
**
45.9
**
s~
96.6
**
96.3.
t*
S R
0.0946
0.0808
0.1181
0.150*
jig
BS
96.4
96.3
95.5
93.1
I.
0 U
.C -Q
54A
53A
Position onj
R.O. Board
7
B
CO
SS
H 0
*&
64
260.5
428
643
806
905
66
260.5
428
643
806
905
Permeate
Flux
(l/m2--hr)
7.8
8.5
8.4
8.9
8.8
8.7
5.1
5.2
5.1
5.3
5.1
5.2
Permeate
Cyanide
(ppm)
66.9
111
100
94.4
89.7
119
24.4
25.4
25.4
22.4
24.3
22.9
SJJJ
r,
BU
97.6
96.6
97.1
96.2
96.7
95.9
99.1
99.2
99.3
99.1
99.1
99.2
ai
u
Ml
i-8*
&4
4.50
6.10
5.0
4.10
4.30
4.50
1.70
1.60
1.35
1.35
1.30
1.60
|e
\l
99.8
99.7
99.7
99.7
99.7
99.7
>99.9
>99.9
>99.9
>99.9
>99.9
>99.9
Permeate
PH
__.
12.13
12.14
12.09
12.02
12.02
12.00
11.55
11.53
11.56
11.47
11.60
11.53
2?
r
h
___
59.1
**
26.2
**
1
u
V ^i
Ttt
It
___
95.3
**
97.9
**
2 C
|u .
£ ^
____
0.1889
0.1382
0.0697
0.0537
lu
11
...»
92.8
93.7
97.3
97.6
57
-------
jj
46A
47B
Position onl
R.O. Board 1
9
10
CO
II
1
24
120.5
304
498.6
666.0
881.2
1044
1143
1
24
120.5
304
498.6
666.0
881.2
1044
1143
SB?
v ON
es-s
& a
21.7
24.1
25.3
25.6
23.4
24.4
25.0
23.9
23.2
36.0
39.5
39.0
31.6
26.0
23.1
22.1
20.4
20.0
Pemeate
Cyanide
(ppn)
229
236
240
247
299
291
251
251
276
181
229
234
250
320
328
299
260
291
.'.8
.1
89.2
90.0
90.5
91.2
90.0
91.5
90.0
90.8
90.5
91.5
90.4
90.7
91.1
90.2
90.4
88.1
90.5
90.0
in
£ ~
55.5
42.5
31.3
27.5
24.3
23.1
30.3
25.0
19.8
10.3
10.7
14.2
16.0
19.0
15.5
13.0
15.0
15.3
8«
*~t
S -Si
96.0
97.3
98.2
98.7
98.7
98.6
98.0
98.4
98.8
99.3
99.3
99.2
99.2
98.9
99.1
99.2
99.0
99.1
Pemeate
PH
12.51
12.55
12.60
12.53
12.63
12.64
12.58
12.58
12.56
12.57
12.58
12.65
12.63
12.70
12.73
12.69
12.60
12.62
Permeate 1
TOC (ppn)
209
**
190
**
I
s
O *H
H **
83.3
**
84.8
**
S 5
3» .
BBS
£ e
0.6228
0.5720
0.6832
0.6467
I
u
f!
M o
83
76.2
74.0
73.9
70.6
58
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-600/2-76-197
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
NEW MEMBRANES FOR TREATING METAL FINISHING
EFFLUENTS BY REVERSE OSMOSIS
6. REPORT DATE
October 1976
(Issuing date)
6. PERFORMING ORGANIZATION CODE
7.AUTHOHIS)
Robert J. Petersen, Kenneth E. Cobian
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Midwest Research Institute
North Star Division
3100 - 38th Avenue South
Minneapolis, Minnesota 55406
10. PROGRAM ELEMENT NO.
1BB610; 01-01-08A
11. CONTRACT/GRANT NO,
R-803264-01-0
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory - Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
F-tnal
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT Long_term reverse osmosis tests were conducted with electroplating wastes
on a new membrane referred to as NS-100. This membrane consists of a polyurea layer,
formed by the reaction of tolylene diisocyanate with polyethylenimine, deposited on
a porous polysulfone support film. The membranes were tested as liners within 1/2-
inch diameter fiber glass tubes. A total of 2360 hours of continuous reverse osmosis
operation was achieved, 1220 hours on pH 1.2 acid copper rinse water and 1140 hours
on pH 12.8 alkaline zinc cyanide rinse water. The membranes exhibited remarkable
chemical stability during exposure to these two pH extremes. Copper and zinc rejec-
tions were generally greater than 99 percent, while cyanide rejections were typically
96 percent or greater. Membrane fluxes were in the range of 18 to 24 liters per sq.
m. per hr. (11 to 14 gfd) for acid copper, but only 8 to 15 l/m2 -hr (5 to 9 gfd) for
zinc cyanide at 41.4 bars (600 psig) and 25°C. Rejection of organics (including
brighteners) was 60 to 78 percent for acid copper and greater than 95 percent for
zinc cyanide. NS-100 membrandes did not reject sulfuric acid. A modified membrane
NS-101, demonstrated twice the permeate flux of NS-100 toward zinc cyanide baths,
but cyanide rejections were low at 90 percent. Difficulties of producing
reproducible, high-flux tubular membranes were not fully resolved in this study.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Copper, Cyanides, *Electroplating,
Industrial Waste Treatment, Industrial
Water, *Membranes, *0smosis,
*Semipermeability, Water Pollution, Zinc
Polymer membranes,
Reverse osmosis,
Electroplating waste
water
13B, 13H, 13K
g7 DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
69
20. SECURITY CLASS (Thispage)
ONCLA33IFIBD
22. PRICE
EPA Form 2220-1 (9-73)
59
*USGPO: 1977 7S7-OS6/5490 Rejlon 5-11
------- |