Document text
AD
CONTRACT NO: DAMD17-90-C-0090
TITLE: MIXED-BED, ION EXCHANGE DEVICE FOR WATER PURIFICATION
PRINCIPAL INVESTIGATOR: Michael A. Taylor, Ph.D.
CONTRACTING ORGANIZATION: Sepratech
2131 Las Palmas Dr., Suite A
Carlsbad, California 92008
REPORT DATE: October 1, 1990
TYPE OF REPORT: Phase I Final Report
PREPARED FOR: U.S. ARMY MEDICAL RESEARCH AND DEVELOPMENT COMMAND
Fort Detrick, Frederick, Maryland 21702-5012
DISTRIBUTION STATEMENT: Approved for public release;
distribution unlimited
The findings in this report are not to be construed as an
official Department of the Army position unless so designated by
other authorized documents.
93-00804 ,
REPORT DOCUMENTATION PAGE
Form Approved
OMB No 0704 01 88
Public reposing Duroen for this collection of information is estimated to average 1 nour oer resoorse including tne time tor reviewing instructions searcning e«<stmg data sources
gathering ana maintaining the data needed, ano completing ana reviewing the collection ot information Sena comments reqaramg this buraen estimate or any otner asoea of this
collection of information, including suggestions tor reducing this Duraen to Washington HeadQuarters Services. Directorate tor information ODerations ana Reoorts. 12 IS Jefferson
Davis Highway. Suite 1204 Arlington. V A 22202-4302. and to the Office of Management and Budget Paperwork Reduction Project (0704-0188). Washington DC 20503
1. AGENCY USE ONLY ( Leave blank) 2. REPORT DATE 3. REPORT TYPE ANO OATES COVERED
1 October 1990 Phase I Final (4/2/90-10/2/90)
4. TITLE AND SU8TITLE 5. FUNDING NUMBERS
Mixed-Bed, Ion Exchange Device for Water Contract No.
Purification DAMD17-90-C-0090
6. AUTHOR(S)
Michael A. Taylor, Ph.D.
65502A
3P665502M802 . BA. 166
WUDA346137
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
Sepratech
2131 Las Palmas Drive, Suite A
Carlsbad, California 92008
8. PERFORMING ORGANIZATION
REPORT NUMBER
9. SPONSORING I MONITORING AGENCY NAME(S) AND ADDRESS(ES)
U.S. Army Medical Research & Development Command
Fort Detrick
Frederick, Maryland 21702-5012
10. SPONSORING ' MONITORING
AGENCY REPORT NUMBER
12a. DISTRIBUTION AVAILABILITY STATEMENT
12b DISTRIBUTION CODE
Approved for public release;
distribution unlimited
14. SUBJECT TERMS
Ion Exchange; Deionization; Water Purification;
RA II; SBIR; Phase I; Water supply
IS. NUMBER OF PAGES
16. PRICE CODE
18 SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT
OF THIS PAGE I OF ABSTRACT I
nclassif ied
NSN 7540-01-280-5500
Standard Form 298 (Rev 2-89)
P»»\C'rO*0 t* .INS: Std 23* 8
24S 02
APPENDIX A
U.S. DEPARTMENT OF DEFENSE
SMALL BUSINESS INNOVATION RESEARCH (SBIR) PROGRAM
PROPOSAL COVER SHEET
Failure lo till in all appropriate spaces may cause your proposal lo be disqualified.
TOPIC NUMBER: DAMD1 7-90-C-0 0 90 _
PROPOSALTITLE: Mixed-Bed, Ion Exchange Device for Water Purification
FIRM NAME: _ Sepratech _
MAIL ADDRESS: 2131 Las Palmas Dr., Ste. A
CITY:
Carlsbad ,
STATE: CA ZIP: 9 2009
PROPOSED COST: PHASE I OR II: I PROPOSED DURATION; 6
PROPOSAL - IN MONTHS —
BUSINESS CERTIFICATION:
► A /a you a small business as described in paragraph 2.2?
► Are you a minority or small disadvantaged business as defined in paiagraph 2.3?
► Are you a woman-owned small business as described in paragrapn 2 u?
► Will you permit the government to disclose the information on Appendix B, if your proposal does not result
in an award, to any party that may be interested in contacting you for fu-ther information or possible
investment?
YES
0
□
□
0
□
0
0
□
► Has this proposal been submitted to other US government agency/agencies; or DoD components, or other
SBIR Activity? If yes, list the name(s) of the agency, DoD component or other SBIR office in the spaces to
the left below. If it has been submitted lo another SBIR activity list the Topic Numbers in the spaces to the
right below:
0
► Number of employees including all affiliates (average for preceding 12 months)
PROJECT MANAGER/PRINCIPAL INVESTIGATOR CORPORATE OFFICIAL (BUSINESS)
NAME: _ Michael A. _ Taylor , Ph ._D . NAME: Mark Sizelove _
TITLE: _ Product Development _ 1 • 1 _ President
TELEPHONE: (619 ) 438-5233 _ TELEPHONE: ( 61 9 ) 438-5233 _
For any purpose other than to evaluate thr- piooovnl. t'-.i:. tints except Append x A and B shall not be disclosed outside the Government
and shall not be duplicates, used or disclosed in whole or in part, provided thm :t a contract is awarded to this proposer as a result ot or in
connection with the submission oi this data, tha Government sh.il: have :he right to duplicate, use or discloso the data to the extent
provided in the funding agreement. This restriction does not limit :!>_■ Government's right to uso information contained in the data if it is
obtained from another source without restriction. The data subtect 'a this restriction is contained on the pages of the proposal listed on the
line below.
PROPRIETARY INFORMATION:
DISCLOSURE PERMISSION STATEMENTS: All data on Appendix: A iclo.tsapic. /\JI data on Appendix B, of an awarded contract are
also releasablo. -
- _ /ft?
SIGNATURE OF PRINCIPAL INVEST. 'GAT OH ”ue d'jNATM.V'. OFCORPORATeBygrJisS OFFI
OFFICIAL DATE
Nothing on inis par, > prop- -t information 'data
' - • * V Nr,
APPENDIX B
U.S. Of- PART ME NT CF OFFENSE
SMALL BUSINESS INNOVATION RESEARCH (SBIR) PROGRAM
PROJECT SUMMARY
TOPIC NUMBER: r,AMD1 7-90-C-0090 _ _
PROPOSAL TITLE. MjXed-Bed. Ion Exchange Device for Water Purification
FIRM NAME: Sepratech
PHASE I or II PROPOSAL I
Technical Abstract (Limit your abstract to 200 words with no classifiGd'or^lrophetarylnformation/data^™^^^™™1™
I . PROJECT SUMMARY
The purpose of this contract was to determine the feasibility of develop¬
ment and subsequent production of a small ion exchange device for water
purification in a field setting. The device specifications included: a
capacity of 1 g of NaCl ; flow rates of 20-25 or 200-300 ml/min; removal of
dissolved solids to less than 1 mg/ml; and operation in any orientation
without channel formation.
The work performed included selection of a mixed-bed, ion exchange resin
combination after determination of working capacities, physical characteris¬
tics, and resistance to temperature stresses.
Prototype housing design included analysis of: 1.) device patents, 2.)
flow patterns, 3.) bed volume changes, 4.) connection requirements, and 5.)
housing materials; Construction included development of: 1.) volixne ccmpen-
sating frits, 2.) inlet and outlet covers, 3.) component design, 4.) weld
site engineering, and 5.) manufacturing procedures.
Two device designs were used to construct fully functional prototypes
with working capacities of 6.5 and 1.0 grams of NaCl. These devices were ef¬
fective in any orientation without evidence of channel formation at flow
rates of 275 and 25 ml/min respectively.
Anticipated Benefits/Potential Commercial Applications of the Research or Development
Validation of the prototype indicated that these devices exceeded the
specifications necessary for production of ultra-pure water in any setting.
This device could be used as a means of sampling or isolation for any ap¬
plication where elimination of degradative forces could increase sampling ac¬
curacy .
List a maximum of 8 Key Words that describe the Project.
Ion Exchange _ _
Deionization
Water Purification
Nothing on this pags it classified or proprietary information/data
Proposal page No. 2
I I . DETAILED PROJECT OBJECTIVES
A. PRIMARY OBJECTIVE
The primary objective of this project was to determine the feasibility of
designing and constructing a Mixed-Bed, Ion Exchange device with a capacity
of 1 gram of Sodiim Chloride frcm source water containing 10 mg/ml of dis¬
solved salts at flow rates up to 300 ml/min. The device must be equally ef¬
fective in any orientation without the formation of channels within the
separatory bed, and must maintain a uniformity of flow to maximize capacity.
Included in the primary objective was the construction of working prototypes
to exemplify the functional potential and possibilities of mass production of
these devices.
B. OBJECTIVES ACCOMPLISHED TO ACHIEVE THE PRIMARY GOAL
1. Ion Exchange resin selection by:
a. Analysis of resins by:
1. ) Collection of manufacturers published and unpublished data on
resin character i st i cs ,
2. ) Analysis of Resin data,
3. ) Determination of the cost effectiveness of resins, and
4. ) Selection of a group of resins for further testing.
b. Testing of resins by:
1. ) Obtaining test samples of resins, and
2. ) Testing of resins for:
a. ) Working capacity;
b. ) Resistance to temperature stresses at:
(1.) High temperatures, and
(2.) Freezing and thawing; and
c. ) Physical changes in separatory bed volune following ex¬
posure to dissolved salts.
2. Design the device by:
a. Determination of whether patented Sepratech, Ion Exchange device
designs and technologies could be effectively incorporated into a
device designed for field use in the purification of RO pretreated
water, and
b. Designing the device housing by:
1.) Determination of an appropriate flow design within the housing
through :
a. ) Analysis of existing patent designs for determination of
the applicability of use in the proposed device,
b. ) Determination of the flow designs necessary to achieve the
specified flow rates,
c. ) Determination of the separatory bed volune necessary to
achieve a working capacity sufficient to meet the water
volune production specifications, and
d. ) Selection of appropriate connections for use in field set¬
tings and to allow adequate flow rates to achieve the
specified volunes of water output; and
DTIC QUALITY INSPECTED 6
uiocr x t i u 1 1 on/
Availability Codoa
Avail and/or
Olat Spoolal
A I I
□ □
2. ) Selection of an appropriate housing material by:
a.) Analysis of the physical and chemical characteristics of
various potential housing materials based upon:
(1.) Impact strength,
(2.) Resistance high and low temperature stress,
(3. ) Relative cost,
(3.) Ease of manufacturing,
(4.) FDA approval for use in contact with food and medical
devices based upon levels of extractibi 1 ity , and
(5.) Resistance to physical and chemical stresses as¬
sociated with use in a field setting, i.e. ozone and
ultraviolet irradiation.
3. ) Determination of an effective weld site design adequate to
withstand the postulated pressures generated within the device
during use and exposure to temperatures reached during
autoclaving; and
c. Design the housing components, including:
1. ) Selection and development of an effective frit material which
compensated for changes in separatory bed volune during opera¬
tion, and
2. ) Selection and acquisition of effective inlet and outlet covers
capable of temperature resistance without welding to the hous¬
ing ports or loosening due to differences in expansion coeffi¬
cients.
3. Construct prototype devices by:
a. Selection of materials for prototype construction,
b. Preparation of engineering drawings of the prototype,
c. Manufacture of prototype weld sites, and
d. Manufacture of prototype housings.
4. Test manufacture of prototype devices to determine effective methods of
assembly and construction.
a. Determination of the most effective welding machine parameters to weld
together the housing components, and
b. Determination of the most effective procedures to fill housing
uniformly without gaps, channels, or pockets.
5. Performance of validation testing of the prototype working capabilities;
including:
a. Working flow rate potential during use as the separatory bed changes
in volune,
b. Working capacity, and
c. Demonstration of the ability of the device to remain effective in any
orientation by determination of:
1. ) Working capacity of prone devices
2. ) Working flow rates in the prone position
3. ) Absence of channel formation in devices in the prone position
4
III. WORK CARR I ED OUT
A. ION EXCHANGE RESIN SELECTION
Analysis of available information on Ion Exchange resins
The initial phase of this project was the selection of the Ion Exchange
Resins for preparation of the separatory bed. To select the most effective
resin, Ion Exchange resin manufacturers were contacted to request published,
technical data on any resins with potential for use in water purification.
Although all strengths of resins were investigated, it was felt that a Mixed-
Bed combination of hydrogen forms of strong acid and free base or hydroxide
forms of strong base exchangers provided the highest degree of dissociation,
therefore, were most appropriate for this application. The greater degrees on
resin dissociation associated with strong ion exchange resins enabled in¬
creased, irreversible ionic interactions between resins and contaminating
ions within the feedwater. Another reason for the preference of a Mixed-Bed
combination of strong acid, cationic and strong base, anionic exchangers was
the predetermination that disposability was preferable over regeneration.
The manufacturers contacted included: Syborn Incorporated; Dow Chemical
Corp.; Bio-Rad Laboratories, Inc.; Rohm & Haas Co.; Mallinkrodt, Inc.; Pierce
Chemical Co.; Signa Chemical Co.; Aldrich Chemical Co.; J.T. Baker Chemical
Co.; Applied Separations Inc.; Benson Polymeries Inc.; ES Industries Inc.; EM
Science; Alltech Associates, Inc.; Pharmacia LKB Biotechnology, Inc.;
Macherey-Nagel ; and Baxter Scientific Products.
The data obtained from ion exchange resin manufacturers were evaluated
for the following criteria:
1. Compliance of resin composition with Code of Federal Regulations
21:173.25 for use in preparation of food and medical materials,
2. The highest thermal resistance (for autoclaving and freeze-thawing) ,
3. The highest capacity,
4. The greatest flow potential,
5. The particle size,
6. The greatest resistance to fracture,
7. The least response to osmotic shock, i.e. swelling when saturated with
dissociated ions, and
8. The relative cost per volume.
Selection of resins for further testing
Following completion of the analysis of all available information on in¬
vestigated resins, the ion exchange resins listed in Table 1. were selected
for further testing.
Obtain resin samples for further analysis
Seventeen test resin samples were obtained after persistent requests from
5 of the listed manuf acturers . Two samples of particular interest were not
received until the project term was near completion. Ecotec, Inc. provided
separate cationic and anionic exchange resins. These resins were of interest
because they were the smallest sized particle of all industrial resins. In
spite of the reduction in particle size the capacity was purported to be un¬
changed. Sybron provided the mixed-bed resin, lonac NM-201/SG. This resin
combination had very recently been brought to market. This was described as
Sybron 's highest capacity, nuclear grade of mixed-bed resin.
5
Table 1. Ion Exchange Resins Selected for Further Testing
Composition
Manufacturer
Resin
Mixed Bed Resins
Dow Chemical
Rohm & Haas
Rohm & Haas
Syborn
Syborn
Dowex MR- 3
Amber lite MB-1
Amber lite IRN-150
lonac NM-60/60
lonac NM-201/SG
Strong Acid Cationic
Exchanger
Bio-Rad Lab
Dow Chemical
Rohm & Haas
Rohm & Haas
Syborn
Ecotec
AG 50X8
Dowex HCR-S
Amber lite iR-120
Amberlite IRN-77
lonac C-267/SG
Cationic Resin
Strong Base Anionic
Exchanger
Bio-Rad Lab
Dow Chemical
Rohm & Haas
Rohm & Haas
Syborn
Ecotec
AG 1X8
Dowex HCRW 2
Amberlite IRN-78
Amber lite 1 RA- 400
lonac ASB-1P
Anionic Resin
Resin Testing
Analysis of the data provided by resin manufactures indicated that, as
postulated in the proposal application, additional testing was necessary to
select a resin combination capable of meeting the requirements associated
with use in field settings. The parameters described were tested for the
defined reasons.
The following tests were performed on the resin samples:
Determination of the working capacity of selected resins
The initial capacity testing was designed to be performed under working
conditions with fluid passing through a closed vessel rather than in a static
condition, as in a slurry.
The standard bead size for resins used in industrial applications, most
commonly range from 16 to 50 mesh size or roughly 0.3 to 1.18 nrm. These were
relatively large particles. Because of their size, the amount of interstitial
spaces between the particles was also relatively large. As the system size
(particularly bed height) decreased the likehood of the fluid phase passing
through the systems without encountering attractive forces associated with
the resin proport i onate 1 y increased. Thus, the capacity of separatory bed was
dependent upon the dimension of the housing as well as the flow rate through
the system. These characteristics were even more significant in smaller
devices with minimal bed height. Therefore, the working capacity under
defined conditions were unique to each design and must be determined empiri¬
cal ly.
One way of maximizing the capacity of the separatory bed was through the
use of smaller resin particles. This was our initial reason for investigating
resins carmonly used in biotechnology related applications where must smaller
6
fluid volute were involved.
Determination of resistance to temperature stresses
Resistance to high temperatures
In order to assure optimal use of these devices, they needed the longest
shelf- life possible, therefore, must be free of any potential pyrogens. To
assure non- contamination during storage these devices were required to be
sterilized during manufacture. Therefore, the components .and resins must be
able to withstand the temperatures and pressures associated with autoclaving
or an alternative means of decontamination. Potential problems resulting from
exposure to elevated temperatures included: changes in the strength of the
housing and the housing weld, loss of Ion Exchange bed capacity, and degrada¬
tion of the Ion Exchanger support. This testing was particularly important
for hydroxyl forms of anion exchangers, because of their sensitivity to
diminished capacity at elevated temperatures .
Resistance to freezing
In field settings, potential exposure to temperatures below freezing
would be inevitable, therefore, housing and resin testing included exposure
to the stresses associated with freezing and thawing. The potential problems
associated with frigid temperatures included increased brittleness of the
housings and the fracture of the particles of the separatory bed. There was
also the potential that the working apacity of the separatory bed would be
diminished following freezing.
Determination of the physical changes reins during use
As the dissociated ions from the fluid phase were removed by the Ion Ex¬
changers, the support matrix compresses. This results in diminished bed
volume. In order to provide adequate compensation for the bed volume changes
to prevent channel formation, the amount of bed volume change must be deter¬
mined under working conditions. Since the bed volune changes were a function
of the working capacity, these characteristics were also dependent upon the
device dimensions, the flow rate, and the test conditions. These parameters
must be determined empirically.
B. DESIGN THE DEVICE HOUSING
Analysis of the applicability of Sepratech Ion Exchange patents
Because of the similarity in the specifications for the Ion Exchange
device solicited in this project and the claims of Sepratech patented design
of Ion Exchange devices, it was felt that the patented designs were poten¬
tially applicable to the project. However, it was necessary to analyze the
designs to determine what modifications were necessary to reach the proposed
requirements. Specifically, the flow rates were to be increased, therefore,
it was necessary to determine whether the closed system could sustain the
defined flow rate. It was also necessary to define whether the backpressure
generated within the flow distribution chamber was adequate to maintain a
uniformity of flow in larger devices.
Determination of an appropriate flow design
Following review of the patented designs, the design of flow through the
7
proposed prototype devices was defined. This was determined by review of the
limits of flow rate, resulting back pressure, bed diameter, and bed height in
relation to the maintenance of the uniformity of flow within the system.
Determination of the housing design adequate for prototype construction
It was necessary to adapt the housing design to provide for ease of con¬
struction and assembly of functional prototypes. The ultimate device housing
was expected to be constructed of two injection molded pieces, ultrasonical ly
welded together. However, the prototype was constructed of five separate
pieces and assembled with four ultrasonic welds.
Determination of the separatory bed volume
Following analysis of separatory bed capacity per bed volume data
provided by manufacturers and determined by preliminary testing, the housing
internal volume and ratio of height to diameter were defined. These dimen¬
sions took into considerations expected differences in working capacities
unique to each device and the test conditions.
Selection of appropriate connections for use in field settings
To assure that this device was attached with the other portions of the IV
water maker securely, aseptically, and easily, an appropriate means of con¬
necting portion of the system was developed.
Selection of an appropriate housing material
Devices dedicated to medical applications in field settings places spe¬
cial requirements on the materials used to construct these devices. There¬
fore, an analysis of the physical and chemical characteristics of various
potential housing materials was performed. This analysis was based upon:
1. Impact strength
2. Resistance high and low temperature stress
3. Resistance to anticipated working pressures
4. Relative cost
5. Ease of manufacturing
6. FDA approval for use in contact with food and medical
devices based upon levels of extract ibi 1 ity.
7. Resistance to physical and chemical stresses associated with
use in a field setting, i.e. ozone and ultraviolet ir¬
radiation.
Determination of an effective weld site design
Construction of prototypes and the final product required that a weld
site be designed, constructed, and validated. This weld site must be capable
of withstanding at least 4 times the working pressures postulated to occur
within the device. The design had to take into consideration the materials
characteristics an housing dimensions to assure weld uniformity and strength.
To construct prototypes of this weld site it was necessary to have the
cylinders and tops machined to close tolerances (+/- 0.001 inches). Depending
upon the material used this could be very difficult, particularly if extruded
cylinders were used in prototypes construction. Extruded materials vary con¬
siderably in dimensions.
8
C. DESIGN THE HOUSING COMPONENTS
Selection and development of an effective frit material
Since the separatory bed volume was expected to diminish cons i deraf 1 y
during operation, the volune compensating frit material must be designed to
meet or exceed this decreased volume. This portion of the device design wr_s
critical to prevent the formation of channels within the separatory bed. In
addition the frit must have sufficient backpressure to force fluids entering
the device to the periphery, prior to passing through the frit. This enhances
the uniformity of flow within the device. However, the backpressure was not
to be excessive in order to minimize the total backpressure within the sys¬
tem. The frits must be composed of materials with FDA/USP approved levels of
extractibi 1 ity and must have resistance to 121 degrees centigrade.
Selection and acquisition of effective inlet and outlet covers
Device use in field settings recjired that the inlet and outlets be
covered securely to prevent contamination. The covers must also be easily
removable. This required that they were composed of dissimilar materials.
Like materials have a tendency to bind and even weld when exposed to elevated
temperatures. The covers had to have compatible coefficients of expansion
since temperature variation could result in the loosening of covers.
D. CONSTRUCT PROTOTYPE DEVICES
After the housing design was decided drawings were prepared for use by
the contracted machining facilities. Three facilities were contracted to mill
prototype weld sites. This enabled comparative analysis of each facilities
ability to perform the milling of the complete housings to the desired
tolerances. Following selection of a single facility, 75 prototypes were
machined from extruded polycarbonate cylinder and flat stock.
E. TEST MANUFACTURE PROTOTYPE DEVICES
In order complete assembly of functional prototypes it was necessary to
develop effective methods of assembly and construction, including the
fol lowing:
Define the most effective welding procedures
For selection of the machine facility for prototype milling and the op¬
timization of ultrasonic welds on prototypes the welding procedures had to
defined. The ultrasonic weld of the housings components required the
coordination of nine separate parameters. The gain of booster had to be
matched with the frequency of the welding horn. In addition the speed of arm
travel had to be matched with the weld time, the pretriggering of the we id
cycle, the hydraulic pressure of the welding arm, the welder triggering pres¬
sure, and the end of weld shut-off. It was necessary to determine the
majority of these settings empirically through progressive adjustments,
weld inspection, and pressure testing.
Determination of the effective cylinder filling procedures
For effective function the housing had to be filled uniformly without
gaps, channels, or pockets. Several basic methods and nunerous variations
9
were tested, including use of dry or wet (slurried) particles.
F. VALIDATE THE COMPLETED PROTOTYPE FUNCTIONAL POTENTIAL
Determination of the working flow rate
The working flow rate and associated backpressures in various orienta¬
tions were determined for both sized prototypes with several different
separatory beds. This testing was monitored as the separatory bed volume
changed to monitor for potential breakthrough due to channel formation.
Determination of the working capacity
The working capacity is various orientations was determined for both
sized prototypes with several different separatory beds. This testing was
monitored as the separatory bed volume changed to monitor for potential
breakthrough due to channel formation.
IV. RESULTS OBTAINED
A. RESIN OOST ANALYSIS
Cost Analysis of Resins
The cost of the resins ccnrmonly used for industrial applications ranged
from $82 to $84 per cubic foot for anionic exchangers and $218 to $226 per
cubic foot for cationic exchangers. Performed mixed exchangers were frcm $151
to $165. Resins of the same support composition i.e. styrene linked divinyl-
benzene, used in laboratory-scale systems were priced at $45 to $52 per 100
grams. This resin, swelled at ionic saturation had a bed volume of 150 ml per
100 grams. Therefore, the price per cubic foot calculated to be roughly $9800
per cubic foot. The later resins are also of considerably more fragile struc¬
ture, therefore, these resins were not felt to be appropriate for this ap-
pl ication.
We were unable to locate industrial resin suppliers manufacturing par¬
ticles smaller than the 16 to 50 mesh size, with the single exception of
Ecctec Inc. Their resin sizes were from 100 to 200 mesh. Because their
processes reduce resin size without changing the per particle capacity, the
relative capacity of an equivalent bed volume was postulated to be markedly
increased.
B. RESIN TEST.NG
Capacity testing
The baseline working capacity of resin samples was initially tested in
closed, 5 ml vessels at slow flow rates between 20 to 30 ml/min. Weighed
sample of resin were exposed to source •. ater consisting of ultra-pure water
containing 100 mg/1 of Sodium Chloride. The effluent frcm the test device was
monitored for changes in TD5. The solution was recirculated through the test
sample. Increased TDS was taken to indicate saturation of the Ion Exchange
resins in this test system under the defined conditions. From the volune of
water passing through the device the capacity of the resin was calculated.
The Rohm & Haas, Amberlite IRN-150 and the Sybron, lonac NM-201/SG had
equivalent capacities 23 mg of NaCl per ml of resin at test device saturation
(Table 2.). All other resin samples had lower capacities. The ability to
10
remove dissociated ions increased as the flow rates were decreased.
Table 2. WORKING CAPACITY TESTING
RESIN
BED TYPE GRADE
WORKING CAPACITY
g ions/g resin
g ions/ml
AMBERLITE MB-1
MIXED
0.030
0.020
AMBERLITE IRN-150
MIXED NUCLEAR
0.033
0.022
DCWEX MR-3
MIXED NUCLEAR
0.030
0.020
DCWEX MRS-C
MIXED
0.025
0.017
IONAC NM-60/SG
MIXED NUCLEAR
0.025
0.017
I0NAC NM-201/SG
MIXED NUCLEAR
0.032
0.021
ECOTEC (REGENERATED) MIXED
0.005
0.003
AMBERLITE IR-120
CAT IONIC- NUCLEAR
0.075
0.050
AMBERLITE IRN-77
CAT 1 ON 1 C-H
0.095
0.064
DOMEX HCR-S
CAT 1 ON 1 C~H
0.055
0.037
IONAC C-201/SG
CAT 1 ONI C-H
0.086
0.058
E00TEC
CAT 1 ONI C-H
ND
ND
AMBERLITE IRN-78
ANIONIC-O NUCLEAR
0.065
0.044
AMBERLITE IRA-400
ANIONIC-OH
0.075
0.050
DCWEX SBR-P
ANIONIC-OH
0.053
0.036
IONAC ASB-1P
ANIONIC-OH
0.060
0.040
EOOTEC
ANIONIC-OH
ND
ND
Defining a capacity per unit volume provided what was likely an in¬
herently erroneous basis for comparison. It was clear that the resin capacity
was a relative measure of the mix of particle sizes within a given range, the
regeneration level, and the level of hydration. The particle size in the 16
to 50 mesh range are from 1.18 millimeters to 0.3 millimeters. Any skewing in
the distribution made a significant difference in the bed volume to capacity
ratio. The regeneration level depended upon the manufacturer, however, was
also dependent upon the age of the resin with labile chemistries (hydroxyl
anion exchangers). The level cf hydration affected the particle size, there¬
fore, the bed volume. This of course varied considerably as the dissociated
salts were inmob i 1 i zed . For these reasons the capacity per bed volume was
likely to vary significantly depending upon the conditions.
The working capacity of the Ecotec resin combination was purported to be
roughly 6 times the working capacity of the Amberlite IRN-150 and lonac NM-
201/5G samples. The working capacity was based upon assumptions provided by
the manufacturer , related to equivalents of absorption of metal ions in solu¬
tion.
The variation in manufacturers data and the observed capacities were in
part attributed to the size of the test vessel and test conditions. Under all
manufacturers suggested test systems the bed height was defined as a minimun
of 30 inches tall, with flow rates greater than a gallon/minute (3.8
1 iters/min) .
11
High Temperature Testing
Following exposure of weighed resin samples to 121 degrees Centigrade for
15 minutes in pressurized steam, no observable physical changes were evident
(Table 3.). however, there was seme loss of working capacity. The results of
testing of the separate cation and anion exchangers indicated that no sig¬
nificant loss of capacity occurred among any of the the cationic exchangers.
The capacity loss among anionic and mixed-bed exchangers ranged frem 5 to 15
percent. In addition, since separate samples of all the constituents of the
mixed bed combinations were not available for all samples, it was not pos¬
sible to verify that the loss of capacity was due entirely to the loss of
capacity of the anion exchangers.
Freeze-Thaw Testing
Weighed samples of resins were exposed to -20 degrees Centigrade for 12
hours then allowed to return to rocm temperature. After thawing fractured
resin particles were evident in most of the resin samples (Table 3.). These
fractured particles were of a wide range in sizes. Analysis of the particles
indicated that the fractured particle sizes were always greater than 100
microns. The fractured particles constituted a very small fraction of the
tested vo lime. Among the 16 to 50 mesh samples weighed samples of particles
less than 300 microns was never greater than 0.5% of the bed weight. The
amount of fractured particles observed smong the Ecotec resins was less than
with other resins. No particles less than 75 microns were observable in these
test samples. Since the porosity of the media restraints is roughly 20
microns no fractured articles frem any samples penetrated the restraints.
This parameter was, therefore, not included as a basis for resin selection.
No changes in working capacity were observed among test samples exposed
to below zero temperatures (Table 3.)
Resin Swell Testing
Because each sample of resin had varied degrees of hydration
with various hydration fluids, the swelling characteristics of each resin
were compared between identical, 5 gram samples of resins. One resin sample
was hydrated in excess ultra-pure water containing less than 1.0 mg/1 of to¬
tal dissolved water (TDS). The other weighed sample was hydrated in excess 1
M Sodium Chloride. Both samples were allowed to equilibrate for several hours
with occasional mixing. Thereafter, the bed volume of each sample was detei —
mined. The mean results were listed in Table 4.
The greatest degree of variation in bed volume was seen among the anion
and mixed bed exchangers, which was roughly double that of the cation ex¬
changers. This was somewhat surprising since it was initially postulated that
the greatest variation would occur within the anion exchangers. It was fur¬
ther postulated that the change in bed volume of the mixed bed exchangers
would be roughly the average of the bed volume changes of the single resin
samples. This was proposed since the mixed bed exchangers were composed of
combinations of the individual cation and anion resins tested. It was of note
that the only mixed bed sample not to follow this pattern was the Eco-Tec
mixed bed resin. This sample was produced by combining equivalent weighed
portions of the individual Eco-Tec cation and anion resins. Also of note was
the fact that the weighed samples of the Eco-Tec cation resins had a smaller
bed volume than the other manufacturers resins. This could have been ac¬
counted for by the considerably smaller particle size of the resin. This
12
Table 3. Effects of Temperature Stress on Resin Samples
RESIN BED TYPE GRADE CHANGES
PHYSICAL CAPACITY
TEMPERATURE TEMPERATURE
HIGH LOW PARTICLE % BY HIGH LOW
SIZE WEIGHTS OF ORIG. )
(tm)
AMBERLITE MB-1
MIXED
NC
FINES
>
100
<
0.5
87
NC
AMBERLITE IRN-150
MIXED NUCLEAR
NC
FINES
>
100
<
0.5
95
NC
DOWEX MR-3
MIXED NUCLEAR
NC
FINES
>
100
<
0.5
92
NC
DOWEX MRS-C
MIXED
NC
FINES
>
100
<
0.5
85
NC
IONAC NM-60/5G
MIXED NUCLEAR
NC
FINES
>
100
<
0.5
86
NC
IONAC NM-201/SG
MIXED NUCLEAR
NC
FINES
>
100
<
0.5
92
NC
ECOTEC (REGENERATED) MIXED
NC
NC
>
75
<
0.5
NC
AMBERLITE 1R-120
CATIONIC- NUCLEAR
NC
FINES
>
100
<
0.5
NC
NC
AMBERLITE IRN-77
CAT 1 ON 1 C-H
NC
F 1 NES
>
100
<
0.5
NC
NC
DOWEX HCR-S
CAT 1 ON 1 C-H
NC
FINES
>
100
<
0.5
NC
NC
IONAC C-267/SG
CAT 1 ON 1 C-H
NC
FINES
>
100
<
0.5
NC
NC
ECOTEC
CAT 1 ONI C-H
NC
FINES
>
100
<
0.5
NC
NC
AMBERLITE IRN-78
ANIONIC-O NUCLEAR
NC
FINES
>
100
<
0.5
95
NC
AMBERLITE IRA-400
ANIONIC-OH
NC
FINES
>
100
<
0.5
ND
ND
DOWEX SBR-P
AN 1 ON 1 C-OH
NC
FINES
>
100
<
0.5
85
NC
IONAC ASB-1P
ANIONIC-OH
NC
FINES
>
100
<
0.5
86
NC
EOOTEC
ANIONIC-OH
NC
FINES
>
100
<
0.5
ND
NC
ND = NOT DONE
NC = NO CHANGE; LESS THAN 5% DIFFERENCE
Table 4. Swell Characteristics of Resins Sanrples
Resin
Type
Ultra-pure
Water
1 M NaCl
% Change
Af'BERL 1 TE IRN-77
CATION
6.6
5.75
-12.9
AMBERLITE IR-120
CATION
6.3
5.85
-7.5
IONAC C-267/SG
CATION
6.5
5.9
-9.2
IONAC CFP-110
CATION
6.6
6.0
-9.0
DOWEX HCR-S
CATION
6.5
6.0
-7.7
ECO-TEC
CATION
5.8
5.2
-10.3
AMBERLITE IRN-78
ANION
7.3
5.3
-27 .4
IONAC ASB-1P/5G
ANION
7.4
5.5
-25.7
IONAC A-641
ANION
7.0
5.4
-22.9
DOWEX 5BR-0H
ANION
7.5
5.8
-22.0
ECO-TEC
ANION
7.8
6.6
-15.4
AMBERLITE IRN-150
MIXED
7.3
5.95
-18.5
AMBERLITE MB-1
MIXED
8.4
5.85
-30.4
IONAC NM-60/SG
MIXED
7.9
5.85
-25.9
DOWEX MRS-C
MIXED
8.4
5.85
-22.0
ECO-TEC
MIXED
6.6
5.9
-10.6
13
would suggest that Eco-Tec cation exchanger would have relatively greater ex¬
posed surface area than other cation exchangers. In contrast, the weighed
sample of the Eco-Tec anion exchanger had a greater bed volume than the other
resins, in spite of the smaller particle size. This would suggest that this
resin had a greater degree of porosity than all other anion exchangers. If
the structural integrity were maintained this may facilitate flow rate.
Final Resin Selection
Based upon the results of analysis of manufacturers data, resin testing
and cost analysis, three resin combinations were selected to be incorporated
into prototype devices. These mixed-bed resin combination are Rohm & Haas,
Amber lite IRN-150; the Sepratech prepared Mixed-Bed combination of Ecotec
anion and cation exchangers; and Sybron, lonac NM-201/SG. These resins were
selected because of the relatively low backpressure at working flow rates,
the working capacity in smaller test devices, the capacity following exposure
to elevated temperatures , the resistance to fracturing at low temperatures ,
the absence of impurities associated with the nuclear grade, and the cost.
Although the Ecotec resins did not demonstrate the working capacity an¬
ticipated, it was felt that the particle size and the probability of enhance¬
ment of the working capacity justified further investigation.
C. HOUSING DESIGN
Flow Design
Existing patented designs of similar devices were reviewed with regards
to the design requirements (internal pressure tolerance and potential flow
rate) associated with either of the proposed reverse osmosis purification
systems. This review indicated that existing flow designs and housing pres¬
sure resistance potentials could easily be applicable for use in prototype
devices of the approximate size proposed within this contract.
In the device flow design (refer to Fig. 1-4) the fluid enters through the
inlet and is dispersed to the periphery within the fluid dispersal chamber.
This dispersal is induced by the backpressure inherent in the frit. The fluid
passes through the frit and into the separatory bed to the downstream frit.
Fluid passing through the frit enters the fluid collection chamber and exits
the device via the outlet. By forcing the fluid to the periphery, the fluid
passes through the device in uniformily; meaning the face of the fluid in the
separatory bed is not retarded at the areas away from the inlet and outlet.
The uniformity of flow maximizes the capacity of the separatory bed through
maximal exposure of the resin to the fluid. In addition, the uniformity of
flow enables the the device to be used in any orientation. Flow through the
bed is also maintained as the bed volume contracted by frits designed to ex¬
pand, compensating for volume changes and preventing channel formation.
Housing Design
The prototype housing was decided to consist of three major components, a
cylinder and two end pieces. The resin bed was to be contained within the
cylinder. The resin w as to be prevented from existing the cylinder by
upstream and downstream frits. These frits were also to have the capacity to
compensate for changes in bed volume as the resin bound free ions within the
fluid as it passed through the device. Upstream and downstream of the resin
restraining frits were to be fluid dispersal and collection chambers. These
14
chambers allowed fluid to pass through the device without the generation of
vortices or dead spaces, while maximizing the resin exposure to the fluid.
Entry and exit were to occur through an inlet in the top and an outlet from
the base of the cy 1 i nder .
It was determined that in order for the flow to the distal portions of
the dispersal chamber the height of this chamber should have been extended.
To determine the internal dimensions and resulting flow rates and back pres¬
sures of the prototype cylinders, other factors would have to be defined,
i.e. capacity /bed volume and relative swell /bed volume.
Housing Internal Volume
The results of working capacity test indicated that to achieve a produc¬
tion capacity of 100 liters, the required separatory bed volume for a 16-50
mesh bead ion exchanger, had to be roughly 150 ml and 25 ml at flow rates of
200 to 300 ml/min and 25 ml/min, respectively. This included enough excess
volume to provide reasonable assurance that the proposed production capacity
was met.
The most effective volume compensating frit design constructed, had a ex¬
panded volume sufficient to compensate for a 203> reduction in total bed
volume. If necessary this could have been increased. The nominal height of
this frit was 0.42 inches in the larger prototype device.
The outer dimensions of the housing also had to take into consideration
the thickness of the housing materials for each of the end pieces, 0.125
inches each.+
To meet these criteria, the prototype housings were constructed to be 5.5
inches in height by 1.5 inches in diameter and 1.75 inches in height and 1
inch in diameter.
Attachment Sites
To afford ease of attachment, resistance to contamination, and pressure
resistance, the upstream and downstream attachment sites of the prototype
device were decided to include female Leui — lock fittings centrally located at
both ends. For construction of prototypes; injection molded, fittings from
analogous devices were ultrasonical ly welded to machined prototype tops.
Tests of the flow rate through the attachment site orifices and the as¬
sociated back pressures were less than 0.5 PS I at 345 ml/min therefore were
acceptab 1 e .
Housing Material Selection
Preliminary investigation of the potential housing materials indicated
that either polypropylene or polycarbonate were most appropriate materials
for prototype construction. Polypropylene provided ease of machining,
however, was considerably more difficult to ultrasonical ly weld effectively.
Polycarbonate was more easily welded and was more transparent , but was likely
to be more difficult to machine to close tolerances. Polypropylene or
polycarbonate devices with wall thickness of 0.125 inches could withstand
system pressures up to 150 PS I. The decision was made to use polycarbonate
for construction of all prototypes housing components.
Weld Sites
Four weld sites were necessary (Fig. 5). These sites were designed to be
identical. The weld sites on the cylinder ends and the two end pieces top
15
were decided to be shear welded together. The weld sites had at least 0.035
inches of interference initiated at the contact point by a 45 degree angle.
The tolerances between adjoining outer and inner diameters were milled to
within +/- 0.001 inches to provide overall tolerances of +/- 0.002 inches.
Wall Thickness Necessary for Pressure Resistance
Prototype devices of polycarbonate were constructed of materials 0.125
inches in thickness. This was deigned to be more than adequate to endure the
projected internal pressures encountered within the envisaged system, in ad¬
dition this provided a significant level of impact resistance to the device.
D. COMPONENT DESIGN
Frit Material Selection
Initially, polypropylene was preselected as the exclusive component of
the frit. However, further analysis indicated that considerably greater
volune compensation was required than possible from existing polypropylene
frits. Subsequently, a number of ccmbinations of materials were investigated,
including multiple nylon mesh, cellulose based mesh, ana polypropylene mesh.
In addition the initially tested frits generated a pressure drop 5.2 PSI .
A ccmbination of multiple layers of cellulose, nylon, and polypropylene
were determined to provide sufficient volume compensation while minimizing
the associated pressure to less them 2.0 PSI . The downstream frit of both
device were designed to be 0.145 inches in height. The upstream frit of the
150 ml device was designed to provide a minimum expanded height of 0.775
inches and a compressed height of 0.275 inches.
Port Covers for Leur Attachment Sites
Both polypropylene and polycarbonate covers for male and female Leur- lock
ports. In order to prevent binding of the covers to the housing dissimilar
plastic covers were used. The composition of these covers were approved by
FDA/U5P for medical use.
c n nryr/yryn c i^vt i rs ki
l_ • r r\ v i v i iru \A/iioir\<A/i ivii
Prototype Weld Site Construction
Construction of prototypes progressed in a series of steps. The first
step was to design, construct, optimize, and test the weld site between com¬
ponent parts of the housing. Preliminary prototype weld sites of several
diameters and thicknesses of cylinders were constructed of polycarbonate.
This material was selected because of the ease with which it can be welded.
Because of the relative rigidity of this plastic, it provided seme difficulty
in machining. However, since the envisaged device was proposed to be injec¬
tion molded, it was felt that providing a solution to the welding problems
would easiest to begin with polycarbonate prototypes rather than alternative
plastics, i.e. polypropylene. Weld sites were milled into cylinders and tops
of various diameters. In these prototypes, connections consisted of injection
molded, female Leui — lock fittings were welded to prototype tops. Ultrasonic
welding equipment was also constructed to test weld these housing. This in¬
cluded bases to secure the components in trie proper orientation and welding
horns tuned to the frequencies necessary for each weld.
Test procedures were standardized for the different wall thicknesses and
16
cylinder diameters as well as nine variables associated with the use of the
ultrasonic welder. Following optimization of the equipment and welding proce¬
dures the welded housings were pressure tested with compressed gas. The
defined weld design and procedures withstood internal pressures greater than
100 PSI .
Design and Construction of Prototypes
The two different sized prototype housings were machined from five parts:
a single cylinder, two end pieces, and two female Leur-lock fittings. All
materials were of 0.125 inch thick polycarbonate. The weld sites were milled
to the specifications described in Figure 1. and were within tolerances of
0.002 inches. The same weld procedures were followed as determined in weld
testing.
Selection of Filling Procedures
Preliminary testing indicated that achieving a uniform mixture of mixed-
bed components, particularly with slurried particles, may be rather dif¬
ficult. However, this suggested that filling housings with slurried resins
may lead to separation of the anion and cation exchangers. The procedures
selected included manual filling and compacting to defined pressures of
weighed resin aliquots. The 150 ml prototypes were filled with 100 mg of
resin. The smaller prototypes were filled with 25 mg of resin.
D. PROTOTYPE VALIDATION TESTING
Flow rate
Both the large and small prototype devices were capable of funtioning at
the specified flow rates, roughly 275 and 25 ml /min respectively. There was
considerable variation in the backpressure associated with different
separatory beds and frit configurations.
Flow rate testing indicated that even though the Ecotec resin was the
smallest particle size of all resins samples tested, the backpressure at the
specified flow rates were the least of all resins tested. Because of this
contradiction, it was decided that this resin should be further investigated.
Capacity testing indicated that more complete regeneration of this resin to
the nydroxyi derivative was necessary. Efforts to attain the potential
capacity indicated by the manufacturer were unsuccessful.
Working Capacity Testing
The working capacity of prototypes of various separatory beds are indi¬
cated in Table 5. No evidence of channel formation was observed during any
working capacity testing. The curves of ion removal from source water were
consistently repetitive until the ion breakthrough occurred. The flow rates
and back pressures were consistent, without detectable variations in flow
rate.
17
Table 5. Working Capacity of Completed Prototypes
Separatory
Bed
Volume
(ml)
Flow
Rate
(ml/min)
Source
Water
(mg/1)
Working
Capacity
(g/device)
Back
Pressure
(PSI)
R & H IRN-150
150
275
1000
6.6
5.7
lonac NM-201/SG
150
275
1000
5.2
6.7
Dowex MR-3-C
150
275
1000
5.1
6.0
Ecotec
150
275
N.D.
N.D.
5.1
IR & H IRN-150
25
25
100
1.0
3.0
R & H IRN-150
150
07C
L. »
10
6.5
5.7
R & H IRN-150
oc
4.J
OC
L.*J
10
1.0
2.8
UNIT DESIGN CHARACTERISTICS
Large unit
Solicited Goals: Adaptable to higher flow rate RO pump/filter
combination
Flow Rate: 200-300 ml/min
Capacity: 1 g of NaCl
Source Water: 10 mg/ml TD5
Dimensions: Within 23 x 20 x 8 centimeters
Design Characteristics
Exterior dimensions
Height: 5.5 inches/ 14.0 centimeters
Diameter: 1.75 inches/ 4.45 centimeters
Bed volume: 150 ml
Separatory Matrix
Composition: 100 g Rohm & Haas; IRN-150, Nuclear Grade,
Mixed-Bed Exchangers
Connections: Inlet and Outlet female Leur-lock connectors
Housing Materials: Polycarbonate
Matrix Restraints: Combinations of Cellulose, Polypropylene,
Polyethylene, and Nylon
Functional capability
Weld strength:
Working Capacity: 6.6 grams of NaCl (at 285 ml/min)
Efficacy in altered orientation
Small Unit
Sol icited Goals:
Flew Rate:
Capacity:
Source Water:
Dimensions:
18
20 25 ml/min
1 g of NaCl
10 mg/ml TDS
Within 23 x 20 x 8 centimeters
Design Characteristics
Exterior dimensions
Height: 2.0 inches/ 5.1 centimeters
Dianeter: 1.0 inches/ 2.5 centimeters
Bed volume: 6.6 ml
Separatory Matrix
Composition: 12.5 g Rohm & Haas; IRN-150,
Nuclear Grade, Mixed-Bed Exchangers
Connections: Inlet and Outlet female Leur-lock connec¬
tors
Housing Materials: Polycarbonate
Matrix Restraints: Combinations of Cellulose, Polypropylene,
Polyethylene, and Nylon
Functional Capabilities
Weld strength:
Working Capacity: 1.0 grams of NaCl (at 125 ml/min)
Efficacy in altered orientation
The following factors are all interrelated in a very complex manner and
have a particularly significant effect on small Ion Excnange systems.
5. ESTIMATES OF TECHNICAL FEASIBILITY
Mass production is quite feasible based upon:
1. Completion of two fully functional prototype designs,
2. Demonstrated capabilities which exceed the specified goals for
capacity of Sodium Chloride in solution and functional requirements,
and
3. Completion of proposed device housing designs, including:
a. Injection molded cylinder similar to the prototype
1. ) Radial fins in place of multi-layered mesh to support the
frits and improve fluid collection.
2. ) Either male or female Leur-lock connection to ASMI stand¬
ards
b. Injection molded top
1. ) Fluid collection chamber with minimal space
2. ) Female Leur-lock connection to ASMI standards
The following areas could be improved further in massed produced devices:
1. Obtaining smaller mesh resin particles. Easily solved with more com¬
plete regeneration of existing smaller mesh resins. There is a ten¬
dency for some non- uniformity of flow in prototype device with the
largest particle sizes.
2. The relative density of counterions imrobi 1 ized to individual resin
particles varies significantly enough to constitute a potential
problem, if the water source contains increased concentrations of
dissolved salts. The localized attachment of ions can induce local¬
ized changes in the separatory bed, i.e. formation of pockets of
compacted resins. This was observed only when the source water TD
concentration was 1 gram/ liter or greater.
13
3. Obtaining frits of FDA approved materials capable of with standing
autoclaving for sterilization.
20
Figure 1 . Flow Through A Tradational Column
Conventional Open Column
Face of Fluid Phase
within the device is
retarded at the
periphery
21
Sepratech Column
Figure 2 . Sepratech Flow Design
Face of Fluid Phase
Difference
7K
Cone
Traditional Column
Sepratech DctIc«
/
^ t
— ■
ihhhhh!
»»*»##»««i******-**#***#*#**
V V.V.V.V.V.VAV'V.V.V/
Y y y y y y y
minium
Figure 3 . Bed Vol
Compens
Frit
Expansion
Ours
Traditional
Bed
Compacted
r?
Head Space
Formation
Channel
Formation
Figure 4. Variability of Orientation
Additional
Versatility of Orientation
Ours
Result
- Decreased Capacity
- No Uniformity of
Flow
- No Separation
of Peaks
Figure 5. Prototype Housing Desgin
Q,
(3)