I
                                                                   PB85-200038
       Removal and Recovery of Fluoborates and
       Metal Ions from Electroplating Wastewater
       New Jersey Inst. of Tech., Newark
       Prepared for

       Environmental Protection Agency, Cincinnati, OH
       May 85
   DJS.

-------
                                                          EB85-20Q038
                                                EPA/600/2-85/054
                                                Hay 1985
REMOVAL AND RECOVERY OF FLUOBORATES  AND METAL IONS FROM

               ELECTROPLATING HASTEWATER
                           by
         John W.  Liskowltz,  Vincent N. Cagnatl,
            Terrance Hunter  and Ray Haralson
           New Jersey  Institute of Technology
                   Newark,  NJ  07102
                   Grant  No. R 804710
                    Projec* Officer
                   Mary K. Stinson
  Organic  and  Inorganic Chemicals and Products Branch
    Hazardous  Waste Engineering Research Laboratory
                Cincinnati, Ohio   45268
    HAZARDOUS  WASTE ENGINEERING RESEARCH LABORATORY
           OFFICE  OF RESEARCH AND DEVELOPMENT
         U.S.  ENVIRONMENTAL PROTECTION AGENCY
                CINCINNATI, OHIO  45268
             "NATIONAL TECHNICAL
             INFORMATION SERVICE
                        Of CO*«8CI
                       o. »*. am

-------
                                  TECHNICAL REPORT DATA
                           (Rear ma Imurucnont on rt« rrvtnt t*lort compttnnfi
  •E»ORT NO.
  EPA/600/2-85/054
                                                         3   "'NT
 . TI , LE -NO SUBTITLE
      REMOVAL AND RECOVERY OF FLUOBORATES AND METAL
      IONS FROM ELECTROPLATING WASTEUATER
             t. REPORT OATE
                Mav 1985
             •. PERFORMING ORGANIZATION CODE
7 AUTHORlS)
      John  W.  Liskowitz, Vincent N. Cagnatl,
      Terrance Hunter and Ray Haralson
                                                         t. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION '.-MS ANO AOOREU

      NEW JERSEY INSTITUTE OF TECHNOLOGY
      323 HIGH STREET
      NEWARK,  NEW JERSEY 07102
             10. PROGRAM cLcMENT NO.
               1BB610
             II. CONTRACT CRANT NO.
               R 804710
 12 <1»ONSO«»ING AGENCY NAME ANO ADDRESS
   HAZARDOUS WASTE ENGINEERING  RESEARCH  LABORATORY
         OFFICE OF RESEARCH AND DEVELOPMENT
        U.S. ENVIRONMENTAL PROTECTION AGENCY
                CINCINNATI, OH  A5268
             13. TYPE OP REPORT ANO PERIOD COVERED

               FINAL    	
             14. SPONSORING AGENCY CODE
               EPA/600/12
 is. SUPPLEMENTARY NOTES
16. ABSTRACT
      Two  separate methods for the treatment of fluoborate wastewater frost the electron
plating of tin.  solder, copper and nickel stripping were investigated.   These involved!
specific  ion  flotation and electrodialysis to treat dilute  waste  streams from single
tank  rinsing  or  concentrated ^astewaters fron counter-current  or  series  rinsing,
respectively.
      The  fluoborate ion was found to bind with an alkylamine acetate by  displacement
of the acetate group and can be removed from dilute waste stream  either  by air flo-
tation or  ultrafiltration.  Ultrafiltration provided greater rates  of removal than air
flotation.  The  surfactant can be recovered for reuse  hy electrolysis.
      A new high  density low porosity graphite anode that is resistant to the corrosive
properties  of fluoboric acid electrolyte was developed for  electrodialysis treatment
of the tin, solder and copper fluoborate containing wastewaters.  Fluoboric acid
electrolyte was  used to rrevent contamination of the products  with  sulfate ion.   Re-
coveries of the  metal  ions were in excess of 902 with  the stannous  ion being preferen-
tially concentrated in the product and with the stannic ion remaining preferentially
in the feed.  In general , recoveries of the fluoborate ions were  in excess of 80*.-
                               KEY WORDS ANO DOCUMENT ANALYSIS
                 DESCRIPTORS
                                            b.lOENTIFISRS/OPEN ENDED TERMS  Ic.  COSAT* Field/Croup
 specific ion flotation, electrolysis,
 ultrafiltration,  electrodialysis
 fluoborates, surfactants,
 copper, nickel, tin, and
 solder fluoborate plat-
 ing  wastewaters
13 B
13 DISTRIBUTION STATEMENT
 Release to Public
                                             19 SECURITY CLASS iTha Rtforti
                                             Unclassified
                                                                       21 NO. Of "AGES
                                90
2O. SECUR'TY CLASS iTtia pagll
Unclassified
IPA Form 2220-1

-------
                                   DISCLAIMER


     The Information  In this document has been funded wholly or 1n part by
the United States Environmental Protection Agency under Grant No. 804710
to the Kew Jersey Institute of Technology.   It has been subject to the
Agency's peer and administrative review, and it has been approved for publi-
cation as an EPA document.  Mention ,of trade names or commercial products
does not constitute an endorsement or recommendation for use.
                                       ii

-------
                                    FOREWORD


     Today's rapidly developing and changing technologies and Industrial
products and practices frequently carry with them the increased generation
of solid and hazardous wastes.  These materials, if improperly dealt with,
can threaten both,public health and the environment.  Abandoned waste sites
a.nd accidental releases of toxic and hazardous substances to the environment
also have important environmental and public health implications.  The
Hazardous Waste Engineering Research Laboratory assists in providing an
authoritative and defensible engineering basis for assessing' and solving
these problems.  Its products support the policies, programs and regulations
of the Environmental Protection Agency, the permitting and other responsi-
bilities of State and local governments and the needs of both large and
small businesses in handling their wastes responsibly and economically.

     This report describes a laboratory process to remove and recover
fluoborates and metals from platinu rinsewaters.  Several methods were
investigated including specific ion flotation, ultrafiltration, electrolysis,
and electrodialysis.  The results indicate that each method has value
depending on the particular requirements in any specific case.  This report
will be useful to EPA's regulatory program (Effluent Guidelines Division) and
to the industry itself in arriving.at meaningful and achievable discharge
levels.  For further information, please contact the Alternative Technologies
Division of the Hazardous Waste Engineering Research Laboratory.


                                       David G. Stephan, Director
                             Hazardous Waste Engineering Research Laboratory

-------
                                 ABSTRACT




     The study  conducted at the New Jersey  Institute of Technology, Newark,




New Jersey was  concerned with development of  two  separate methods for




treatment of fluoborate-containIng  wastewater from  electroplating of tin,




solder, copper  and nickel stripping.




     The first method  was based on  the  specific ion flotation principle




which involves  the removal of specific  ions from  dilute wastewater through




binding with a  surfactant, followed by  flotation  or ultrafiltration.  This




part of the investigation involved  the  evaluation of structurally different




commercially available surfactants  to determine the type, structure,




mechanism and conditions which govern the formation of a fluoborate-




surfactant complex.  Also, air flotation  and  ultrafiltration were




evaluated for removal  of the complex from the wastewater.  Methods for




recovery of the surfactant were examined.




     The fluoborate was found to bind with  a  commercially available alkyl-




amine acetate type surfactant which reduces the fluoborate concentration In




rinse wastewaters from  100 mg/1 of  fluoborate  to  7-15 mg/1 of fluo-




borate.  Actual plating operation rinse wastewaters containing 100 mg/1 of




fluoborate were used in the study.   Dltrafiltration followed by electro-




lysis provided  the shortest  treatment time  with' recovery of the




surfactant.




     The second method  was electrodiaiysis.   Here t/e major effort was to




find a suitable anode.   Electrodialysis was found feasible for treatment





                                    -iv-

-------
of wasCestrearns containing plating chemical concentrations s* 1000 mg/1




using a high density low porosity graphite anode.  Electrodialysis can




reduce the plating chemical concentrations in the wastestreams to about




100 mg/1.




     The specific ion flotation process either used separately or in the




combination with electrodialysis caa be useful and effective for closed




loop treatment 'of fluoborate-containing vastewaters from electroplating




operations,.  Reagent recovery in speciclc ion flotation and fluo bo rate and




metals recoveries in electrodialysis can be achieved in addition to pollu-




tion control.  However, both methods need further development to make them




commercially suitable for treatment of fluoborate-containing wastestreams.




     This report was submitted in fulfillment of Grant No. R - 804710 by




the New Jersey Institute of Technology under sponsorship of the U.S.




Environmental Protection Agency.  This report covers the period October 1,




1976 to December 1, 1979 and wo-k was completed as of October, 1979.
                                    -v-

-------
                                 CONTENTS




                                                                  Page




      FOREWORD	ill




      ABSTRACT	   lv




      CONTENTS	   vl




      FIGURES	vlil




      TABLES   	    x




  I.  INTRODUCTION'	    1




        Ion Flotation	    4




        Electrodialysis  	    6




  II. CONCLUSIONS	   10




        Specific Ion Flotation 	   10




        Electrodialysis  	   11




III.  RECOMMENDATIONS	   13




 IV.  EXPERIMENTAL	   15




        Ion Flotation	   15




        Surfactants	   15




        Equipment and Procedures 	  ,16




        Fluoborate Analysis  	   19




        Micelle Formation,	   19




        Ultrafiltration  	   19




        Simple Cell	   19




        High Volume Cassette	   22





                                    -vi-

-------
                                                                 Page




       Breaking the Fluoborate-Surfactant Complex . ,	    25




       Addition of Excess Acetic Acid	    25




       Electrolysis 	    25




       Electrodialysis  	    27




 V.  RESULTS AND DISCUSSION	    33




       Ion Flotation	    33




       Factors Influencing the Removal of Fluoborate




       by the Surfactant	    34




       Process Parameters 	    44




       Removal of Fluoborate from Plating Rinse-Waters  ....    45




       Recovery of Surfactant 	    47




       Ultrafiltration  	   '47




       Electrodialysis	    52




       Evaluation of Different Types of Anodes  	    54




       Electrodialysis Result 	    57




       Treatment of Tin Fluoborate Rinse-Waters	    57




       Treatment of Solder Fluoborate Rinse-Waters  	    61




       Treatment of Copper Fluoborate Rinse-Waters  	    69




       Rates of Mass Transfer	    75




       Analysis of Electrolyte  	    76




VI.  REFERENCES	    78
                                 -vii-

-------
                                  FIGURES




Number




  1    A Schematic of A Simple Electrodlalysls Cell	     7




  2    Flotation Apparatus   	    17




  3    Schematic of A Simple Ultrafiltration Cell 	    20




  4    Single Pass Flow System	    23




  5    Recirculating Flow System  	    24




  6    Schematic of Electrolysis Set-Up	'	    26




  7    Schematic of an Electrodlalysls Stack  	    28




  8    Spacer on Graphite Anode to Define S Shaped Flow Pattern .    30




  9    Electrodlalysls Unit Feed Reservoir and




       Electrolyte Reserv6ir  	    31




 10    Treatment of Dilute Sodium Fluoborate Solution




       with Avmac C	    35




 11    Fluoborate Removal from Solder Rinse-Water




       Dependence on Surfactant Molecular Size  	    41




 12    Fluoborate Removal from Solder Rinse-Water




       Dependence on Surfactant Degree of Saturation  	    42




 13    Time Reaction of Fluoborate with Surfactant	    43




 14    Removal of Fluoborate from Plating Bath




       Rinse-Waters Using Armac T 	    46




 15    Electrolysis of Surfactant Fluoborate Concentrated



       Solution from Treatment of Solder Bath Rinse-Water ....    49






                                  -viii-

-------
Number                                                            Page




 16    Photograph of Corroded Platinized Titanium Anode 	    53




 17    Photograph of Corroded Nickel Anode  	    56




 18    Electrodialysis Treatment of Tin Fluoborate




       Rinse-Waters, Feed Cation Concentration  	    58




 19    Treatment of Tin Fluoborate




       Rinse-Waters, Feed Anion Concentration 	    60




 20    Electrodialysis Treatment of Tin Fluoborate




       Rinse-Water, Product Cation Concentration  	  .  .    62




 21    ..lectrodialysis Treatment of Tin Fluoborate




       Rinse-Water, Product Anicn Concentration 	    63




 22    Electrodialysis Treatment of Solder Fluoborate




       Rinse-Waters, Feed Cation Concentration  	    66




 23    Electrodialysis Treatment of Solder fluoborate




       Rinse-Waters, Feed Anion Concentration 	    67




 2A    Electrodialysis Treatment of Solder Fluoborate




       Rinse-Waters, Product Ion Concentration  	    68




 25    Electrodialysis Treatment of Copper Fluoborate




       Rinse-Waters, Fejd Anion Concentration	    72




 26    Electrodialysis Treatment of Copper Fluoborate




       Rinse-Waters, Feed Anion Concentration 	    73




 27    Electrodialysis Treatment of Copper Fluoborate




       Rinse-Waters, Product Ion Concentration  	    74
                                   -ix-

-------
                                  TABLES




Number                                                             Page




  1    Effects of pH Removal en Fluoborate Ion	    37




  2    Fluoborate Removal Using Stepwise Addition of




       Armac C Surfactant to Sodium Fluoborate 	    39




  3    Removal of Surfactant with Low Through-Put




       Ultrafiltration Experiment  	    50




  4    Results of Recycled Retentate Experiment,




       Millipore Cassette  	 •    31




  5    Changes in Percent Stannous Cation in the Feed



       and Product with Time During the Electrodialysis



       Treatment of Tin Fluoborate. Rinse-Mater	64




  6    Changes in the Percent Stannous Ion in Feed




       and Product with Time During the Electrodialysis



       Treatment of Solder Fluoborate Rinse-Water   	    70




  7    Rate of Mass Transport of Cations and Fluoborate




       from Tin, Solder, Copper Fluoborate Rinse-Water 	    77
                                    -x-

-------
                                I.  IHTRODUCTIOH










     A  recognition of the detrimental effects of Industrial pollution on our




environment has surfaced in recent years.  Along with this awareness came




federal regulations calling for reduction of industrial discharges into the




environment.  The metal finishing Industry is an example of an industry,



which will need to develop new technology in order to meet the discharge




requirements under these regulations.




     The metal finishing Industry in the United States includes approxi-




mately  twenty thousand facilities, the majority of which are associated with




the automotive, electronic and jewelry industries.  These facilities en-




compass both job-and captive-shops.  The major operations performed at




metal finishing facilities Include cleaning and pickling, annealing, case




hardening, polishing, buffing, immersion plating, electroplating, phosphatlng,




conversion coating, oxidizing, painting, electropainting and anodizing.




     The wastewater produced  in metal finishing operations is  mainly generated



in two ways.   Concentrated  wastes  come from the dumping of solutions



which have become used or fouled.   This Is especially true of  cleaning,




stripping, passivating and anodizing solutions.  The dilute wastes, which




are the larger volume wastes, come from the rinse-waters.  These rinses wash




off the process solution that has adhered to the surface or was entrapped




in crevices due to the shape  of the processed piece.



     Presently, the most common used procedure for treating metal finishing



wastewaters include the following operations:
                                      -1-

-------
      .  Separation of Oil and Grease




      .  Oxldatlve Destruction of Cyanides




      .  Reduction of Chrcmatea




      .  Neutralization of Acids and Alkalies




      .  Precipitation of Metal Hydroxides




      .  Disposal of Sludge




      Problems arise mainly with the disposal of sludge because the precipitates




present a  potential leaching problem when they are  disposed of on land.




Alternatives to the precipitation process is the recovery and recycle of




wastes  from their point of generation,  or the substitution of toxic process




chemicals  with less objectionable ones.




      Among the more attractive recovery techniques  are reverse osmosis, ion




exchange,  evaporation,  ultraflltration  and electrodialysis.  'In specific




areas,  these processes  have been applied to the recovery of metals from




plating rinse-waters, or the concentration of the rinse-waters for the reuse



In the  plating bath (1).




      Substitution of process chemicals  is practical only when the replacement




does not compromise the quality of the  finished product. One such substition




Is the  use of fluoborate to replace cyanides as a conducting salt in plating




baths.   Fluoborate has  been found to be an excellent carrier ion which gives




a uniform,  bright, well-thrown covering.   Also, fluoborate is much less




toxic than cyanide, and therefore,  provides for a safer plating room.  For




these reasons,  many platers of cadmium,  zinc,  tin,  lead, solder, copper,




nickel  and iron are replacing their cyanide baths with fluoborate b-ths.




     Commercial fluoborate solutions are  presently  available for the plating




of copper,  indium,  iron,  lead,  nickel,  tin and  their alloys.  Also,  fluoboric




                                      -2-

-------
 acid Is used in various pretreatnent operations,  such as stripping and clean-




 ing.  The concentration of these rinse-waters is  variable, depending upon



 the rinsing technique which is employed.   Dilute  rinses would be typical of




 single tank rinsing,  while the concentrated rinse-water would be attributed




 to multiple tank countercurrent or series rinsing.



      The fluoborate ion is an ion in which boron  is covalently saturated by



 fluorine.  This ion forms salts in which  there is a true cation with no co-




 valent bonding to the anion.   Fluoborate  is a very small, tightly bound



 tetrahedron with a uniform charge.  It apparently does not polarize (2).



      Fluoborate by Itself Is relatively non-hazardous.  However, it will



 hydrolize (3)  in water yielding boric acid and fluoride in accordance with




 the following  equilibrium reaction:



      HBF4 + 3H20 -^  B(OH)3 + 4HF





      The presence of  this resulting fluoride in our receiving waters is un-




 desirable.



      There are presently no specific discharge limitations on fluoborate.




 However,  when  a wastestream  is analyzed  for fluoride by the approved method



 (BelLack Distillation), any fluoborate present will be hydrolized yielding



 Inflated fluoride concentrations.  For each fluoborate ion present in a



 sample the test will  show four fluoride ions.   This gives a false Indication




.of fluoride concentration and can show a  National Pollution Discharge



 Elimination System Permit (NPDES) violation where none exists.




      Few processes are known for the removal of fluoborate from plating




 rinse-wa-.ers.   It is  a small tight molecule and is  not easily rejected by



 membrane processes such as reverse osmosis.  Presently, there is no known ion




 exchange  resin  which  will efficiently remove fluoborate frontsolution. Battele
                                     -3-

-------
 Memorial Institute in their January 1974 draft of the Development  Document




 for Limitations for>Electroplating Point Sources, suggested the hydrolysis




 of  the  fluoborate to fluoride, followed by lime precipitation as a possible




 treatment.   This however, results in the production of a sludge that must




 be  disposed in secure landfills.




     Vaccum evaporation is currently being used as a means of recycling




 stannous fluoborate rinse waters  back into the plating tank as make-up




 solution.   Although it provides a closed-loop treatment system, problem*.




 such as precipitation of stannic oxide which inhibits the evaporation are




 encountered during the evaporation process.  It is also an energy  intensive




 operation.



     Since  suitable technology for the treatment of rinse-water from




 fluoborate  plating baths is lacking, this investigation was undertaken.




 Two separate processes, ion flotation and electrodrolysis, were investigated




 as  a possible close-loop treatment of the dilute and concentrated  fluoborate




 containing  wastestresms from the  electroplating industry.



 Ion Flotation




     Flotation processes for use  in separation of solid particles  frrji




 liquids  is  an old,  well-established process in waste removal (4,5)  as  rell



 as  in the process industries (6).  For waste removal tr  flotation,  the




 mechanism is to add a surface active agent to the wastewater and mix it



 well.  This allows the surfactant to adhere to the particles in the water.



 Foaming  through the introduction  of air brings tne surfactant-solids




 combination to the surface even though they are more dense than water.



When an  air bubble contacts the surfactant-particle combination, the




 weight of the particle is balanced by the buoyancy of the affixed air






                                      -4-

-------
bubbles, and the particle floats.




     Ion flotation is a relatively recent process.  The field has been




explored by Sebba (7) in South Africa and Grieves (8) in the United




States.  The concept of ion flotation is different from particle flotation in




that a complex is formed between the surfactant and the dissolved ion. In




participate flotation, the surfactant adheres to the particle, but  In waste




streams such as from metal finishing operations, there are no particles.




The surfactant, therefore, must chemically react with the ion which is to




be removed.




    Most of the work done in the ion flotation field has been concerned




with removing metal cations from solution with anionic surfactants.




However, Grieves (9) has investigated the use of ion flotation for the




removal of  the anionic chromate ion, but' there has been no research into




the removal of the fluoborate ion using this process.  Thus, the use of




specific ion flotation for the treatment of fluoborate plating bath




rinse-waters containing dilute concentration of fluoborate was




investigated.




     The part of the investigation dealing with dilute fluoborate rinse-waters




was concerned with 1) determining the type and structure of the surfactant




that favors the binding of the fluoborate ion with the surfactant, 2) under-




standing the mechanism of binding and those parameters such as pH, ratio of




the amount of surfactant to fluoborate ion, and contact time which can




influence the binding of the fluoborate ion to the surfactant, 3) identifying




those process parameters such as air feed rate, air bubble size, rir diffuser




location, inlet feed direction and mixing time which can Influence the optimum




removal of fluoborate ion, and 4} evaluating this technology for the removal
                                     -5-

-------
 of fluoborate ion from the rinse-waters resulting from nickel stripping,




 solder plating, tin plating and copper plating operations.




      Also, the means to recover the surfactant and fluoborate ion for reuse




 in the treatment process and plating baths, respectively, was examined.




 Ultrafiltration was evaluated, as an alternate to flotation for rapid




 separation of Che fluoborate-surfactant complex from the rinse-waters.




 ElecCrodialysis




      Eleccrodialysis is a membrane process which can be used for the separa-




 tion, removal or concentration of ionized species in aqueous soluCions.  The




 above is accomplished by using an electromotive force to selectively trans-




 port ions Chrough an ion-exchange membrane.  These ion-exchange membranes




 are permeable Co either cations or an ions,1 but not to both.  The membranes




 are thin sheets of ion-exchange material reinforced by a synthetic fabric




 backing.  The resin matrix is usually copolymerized styrene divinylbenzene,




 the exchange capacity being imparted by sulfonic acid groups for cation ux-




1 change niemoranes and quaternary ammonium groups for anion exchange membranes.




      A simple electrodialysis system consists of an anode and a cathode,




 separated by an anion permeable membrane near the anode and a cation




 permeable membrane near the cathode.  Thus a cathode chamber, an anode




 chamber, and a center chamber are formed.  When the electric charge is




 applied, anions pass from the center chamber to the anode chamber while




 cations pass from the center chamber to the cathode chamber.  Therefore,




 the concentration of salt in the center chamber is decreased.  The appli-




 cation of a simple electrodialysis cell in the separation of a potassiutc




 sulfate solution is shown in Figure 1.




      In practical electrodialysis installations (10), there are ten to






                                       -6-

-------
Cathode
Cation-
Pettieable
Membrau?
     i
     I
     i
                                  Anion-
                                  Permeable
                                  Membrane
                  H
                      i
                        >
       K2S°4
                                 30
                                   -
                                                           Anode
FIGURE  1. A SCHEMATIC OF A SIMPLE ELECTRODIALYSIS  CELL
                              -7-

-------
hundreds  of compartments between one pair of electrodes.  These units are




referred  to as electrodialysis stacks.   These stacks  include specially




designed  spacers to separate a cationic and anionic impermeable membrane and




allow flow between adjacent membranes.   The feed  solution is distributed and




the  concentrate collected by two internal hydraulic Channels; one for feed and




one  for concentrate.   \ cell pair includes the following components:




                          .  Cation Selective Membrane




                          .  Feed Spacer




                          .  Anion Selective Membrane




                          .  Concentrate  Spacer



Passing a direct current through the stack causes ions to migrate across the




permeable membranes from the feed channel and collect in the concentrate



channel.




     The  most  frequently used application of electrodialysis is in the



desalination of  brakish water (.11) .   Some recent  investigations relating to




the metal finishing industry include the recovery of nickel and copper from




plating rinse  waters  and closed-loop control of cyanide rinse-waters (11,




12).  However, none have examired    the use of  electrodialysis as a means




of treating fluoborate plating rinse-waters.




     The  feasibility  of using electrodialysis as  a means of treating




concentrated tin,  solder and copper  fluoborate plating bath rinse-waters for




recovery  of the  metal and fluoborate ion as well  as reuse of the feed as



rinse-water was  investigated.




     This  study  also  involved the development and testing of an inexpensive




new anode which  could be used in an  electrodialysis unit and be resistant to




the corrosive  properties of fluoborlc acid electrolyte.  The development of






                                     -8-

-------
this anode allowed fluoboric acid to be used as an electrolyte to match the




anions present in the fluoborate waste stream.




     The use of the traditional anodes such as the platinized titanium




anode was unable to meet the corrosive nature of the fluoboric acid




electrolyte.  The use of this anode in an electrodialysis unit would




require the use of less corrosive electrolytes such as sulfuric acid which




can result In the contamination of the product with undesirable anions;
                                     -9-

-------
                              II.  CONCLUSIONS
     The specific ion  flotation process used either separately or in




combination with electrodialysis appears to be feasible for the close-loop




treatment of waste rinse-waters resulting from solder, tin, nickel and copper




plating operations.  The combined use of both processes provides for recovery




of plating bath chemicals at  concentrations that approach plating bath




strength with a treated effluent suitable for reuse as rinse-water.




Specific Ion Flotation




1.  Specific ion flotation using a commercially available alkylamine acetate




surfactant is feasible for the treatment of dilute rinse waste streams from




solder, tin, nickel and copper, plating operations to provide a product that




can be reused as rinse-water.




2.  Fluoborate anion concentration in the solder, tin and nickel fluoborate




rinse wastestreams can be reduced to a concentration of 7 mg/1.  Fluoborate




anion in the copper plating bath rinse waters can be reduced to a




concentration of 15 mg/1.




3.  The removal of fluoborate anion  by the surfactant is dependent on




replacement of the acetate group on  the surfactant by the fluoborate anion.




Removals are inhibited by anions such as chloride, which fonra stronger




acids than fluoborate  acid.




4.  Fluoborate anion removals are enhanced by an increase in the molecular




weight of the surfactant and  by acidic conditions, which increase the




ionization of acetate  groups  on the  surfactant.






                                     -10-

-------
5.  Excess surfactant is required to make up for the acetate groups that are




lost to the replacement reaction because of micelle formation which occurs




at surfactant concentrations of 12 mg/1.




6.  The replacement of the acetate group on the surfactant by the fluoborate




anion occurs within one minute.  The rate limiting step in this process is




the mixing time required to achieve a complete mix upon addition of




surfactant to a fluoborate  wastestream by the laboratory system used in




this study.




7.  The fluoborate-surfactant complex can be removed from solution by




aeration.  Air bubble size, air diffuser location and inlet feed direction




with respect to bubble rise, does not influence the rate of removal of the




fluoborate-surfactatit complex.




8.  Increases in the air feed rate decreases the concentration of the




fluoborate-surfactant complex in the foam by removing more liquid with the




foam.




9.  Ultrafiltration with recycling of the retentate provides a four times




greater rate of removal ox the fluoborate-surfactant complex from the rinse-




water than aeration.




10.  Electrolysis is suitable for splitting the fluoborate-surfactant




complex to recover the surfactant.  The surfactant is concentrated at the




cathode and can be dissolved from the cathode in the acetate form with




concentrated acetic acid.




Electrodialysis




1.  Electrodialysis appears to be feasible for the treatment of concentrated




fluoborate-containing rinse waste waters.  It provides a product that appro-




 aches plating bath chemical concentrations and an effluent that can be
                                     -11-

-------
further treated by the specific  ion flotation process.




2.  The use of a high density, low porosity graphite anode in an




electrodialys:.s unit provides greater resistance to chemical corrosion and




a greater cuixent density  for a  given applied voltage than the more




commonly usfed platinized titanium  anode.



3.  The graphite anode's resistance to chemical corrosion permitted the




highlv corrosive fluoboric acid  to be used as an electrolyte tu match the




anion in the feed.



4.  The rate of transfer of stannous ion and lead  ion from the feed to the




concentrate is significantly greater than the rate of transfer of stannic




ion from the feed to the concentrate.  This difference in the rate of




transfer of the cations minimizes  the build-up of  undesirable stannic ion




in the plating baths from  reuse  of the electrodialysls product.
                                     -12-

-------
                             III.  RECOMMENDATIONS
     The recovery of placing bath chemicals and the closed-loop treatment




using electrodlalysls with newly developed graphite anode and specific




ion flotation of rinsewater containing fluoborate anlons from solder, tin,




copper and nickel plating operations has been shown to be feasible.  There are




a number of advantages in using electrodialysls and specific ion flotatir.. for




the,treatment of rinsewater containing fluoborate.  Electrolysis treatment




preferentially concentrates the desirable stannous ions in the product




stream from a mixture of stannous and stannic cations usually  present in




the rinsewaters.  The product contains cation and anion concentrations




that approach plating bath strength.  The development of the graphite anode




which is resistant to the corrosive nature of fluoboric acid provides an




anode which Is approximately one-fortieth the cost of the commonly used




platinized titanium anode and allows the fluoborate electrolyte to be used




to match the fluoborate anions present in the rinsewaters.  The use of an




electrolyte in the electrodialysls unit containing anions> that match the




anlons in the feed is preferred to avoid contamination of the product




produced from electrodialysis treatment.  In addition, the surfactant used




in the specific ion flotation treatment of rinsewaters containing low




concentrations of fluoborate can be recovered along with the fluoborate




anion by electrolysis.




     In view of the above, the commercial feasibility of using electro-




dialysis in combination with specific ion flotation to> provide a closed-
                                      -13-

-------
loop treatment  system should be established on a pilot scale.  The




engineering  parameters and costs associated with the design, assembly and




operation of both  systems must be det ermine •!.  Also, this pilot scale




effort should identify the methodology  that will allow rapid mixing of the




fluoborate rinse-waters with the alkylamine acetate surfactant solutions at




concentrations  that  do not favor miceUeformation.  In addition, the long




tena stability  of  electrodialysis membranes (longer than 60 hours) toward




the fluoboric acid electrolyte should be  evaluated since frequent replace-




ment of these membranes can lead to unacceptable operating costs.  The




substitution of  the graphite anode after a period of time should not



significantly contribute to the operating costs of the electrodialysis




unit since Its  replacement cost is approximately one-fortieth of that




required to  restore the'commonly used platinized titanium anode.
                                     -14-

-------
                              IV.  EXPERIMENTAL
Ion Flotation




Surfactants




     A number of different type surfactants were  evaluated in this




investigation.  A  literature search revealed that there were a number of




different  type  surfactants that may bind  to the fluoborate anion'.




However, there  was no  Information to indicate  the type of structure that




would allow the surfactant fluoborate complex  to  be removed from




solution either by air flotation or ultrafiltration and that would




readily allow the  surfactant to be separated from the anion for reuse.




     A number of cacionic surfactants were sslected that differed in




size, degree of saturation and structure.  The cationic surfactants




were evaluated  because of the attraction  that  could be expected in



solution between a positively charged surfactant molecule and a




negatively charged fluoborate ion.  Several non-ionic and anionic




surfactants were investigated to determine if  it would be possible to




bind both the metal and the fluoborate to the  surfactants for removal



from solution.




     The cationic surfactants evaluated were Amine 0, Amlne C, Amine S,




Amlne T,  Ammonyx 220, Ammoayx T, Armac C, Armac T, Armac 180, Armac H,




Armac 8D, Duomac T, Atlas G-3634A and Finazoline T.  The Clba-Geigy Amlne 0,




Amine C,  Amine S and Amlne T surfactants are heterocyclic tertiary amines.
                                    -15-

-------
 The Ammonyx 220 and Ammonyx T obtained from Onyx are an alkyl dimethyl



 benzyl ammonium chloride and a cetyl dimethyl benzyl ammonium chloride,



 respectively.   The Armac C, Armac T, Armac 18D, Armac HI,  and Doumac T
                               i


 surfactants which were obtained from Armak are an alkylamine acetate



 (12-15 carbon  atoms), an alkylamine acetate (16-18 carbon  atoms),  a



 distilled octadecane amine acetate (18 carbon atoms), a hydrogenated



 alkylamine acetate (16-18 car,bon atoms), and an alkylamine diacetate,



 respectively.   The Atlas G-3634A surfactant obtained from  Id is a



 quaternary ammonium derivative.  The Finetex surfactant, Tinazoline  T,


 is  an  amlnoethylimldazolene.



     The  nonionlc surfactants used in this study,  AmmonyxCDO,  Onyxol 336



 and Neutronyx,  were obtained  from Onyx and are an alkyl amido propyl



 dimethyl  amine  oxide, a lauric acid and an alkjlphenol polyglycol ether



 with etbylene oxide,  respectively.



     The  anionic  organic phosphate surfactants used in this  study  were



 Dextrol OC-60,  Dextrol OC-80,  Dextrol OC-90, Oextrol OC-105  and Dextrol



 OC-110 and were obtained from Dexter.



 Equipment and Procedures



     The  specific ion flotation experiments were 'carried out  in a modified



recirculation bath system (see Figure 2)  designed  to  simulate an actual



treatment process.  The recirculation was required  because the length of



 time required for the removal  of the surfactant  by  air  flotation was much



longer than could be  accomodated in  a single flow-through tank.  The



equipment consisted of an eight liter holding tank, a two liter reaction



vessel, an eighty liter per hour recirculation pump, and a twelve liter



foam collector.   The  air for flotation was introduced through a porous




                                     -16-

-------
                  Foam Collector
Reactor
     Air Diffuser
                     Level Control
Air In
                       /
                    Surfactant
                      _v	
                                      Holding
                                       Tank
FIGURE 2.  FLOTATION APPARATUS
              -17-

-------
stone diffuser,  the  air was controlled by a Brooks R-2-15 rotometrer.




The surfactant was introduced into  the suction line  of  the pump using



a tventy milliliter  hypodermic syringe after  fluoborate plating bath




solutions were added and  thoroughly mixed  in  the  system.  The syringe




was used because even slow mixing in the pump caused foam to be




generated in  the holding  tank during the makeup.



     The system  was  'cleaned by washing all tanks  and tubing, and



thoroughly rinsing them with  delonlzed water  after each experiment.  The




system was then  charged with  eight  liters of  deionized  water.  The




desired quantity of  sodium fluoborate or plating  bath solutions was added




to the deionized water and the solution recirculated for 36 minutes to




assure thorough  mixing.  A series of tests in which  the resulting dilute




rinse-waters  solutions were circulated through the system for different




lengths of time  up to 24  hours indicated that the test  solutions were




completely mixed within 36 minutes.   The desired  amount of surfactant,




which had been dissolved  in 200 ml  of deionized water,  was injected into




the system at a  rate of 8 ml  per minute.  The system was again



recirculated  for  another  36 minutes  to insure  complete  mixing of surfactant




solution with the fluoborate  solutions.  Air was  applied, and the




solutions were .recirculated continuously unti.  there was no further



generation of foam at  the liquid  surface in the reactor.  The foam was then



removed as formed.
                                      -18-

-------
Fluoborate Analysis



     Analysis of the solution for the fluoborate ion was carried out using



the series 92 and series 93 Orion fluoborate specific ion probes.  The 93



series, which was determined to be the more sensitive probe, was used to



analyze the solutions that contained no surfactants.  The surfactant was



observed to have an adverse effect on the probe's membrane.  Therefore, the
  (


series 92 probe was used to measure the fluoborate concentration in the



presence of the surfactant because its membranes could be replaced.  It was



determined that the membrane in the 92 series probe had to be replaced after



the measurement of 10 solutions containing fluoborate ion and surfactant.



Micelle Formation



     The critical micelle concentration was determined by conductivity



measurement.  A sharp decrease in the plot of conductance -versus the square



root of the surfactant concentration indicated the concentrations at which



micelles begin to form.



Ultrafiltration



a.  Simple Cell



     The equipment used for initial evaluation of ultrafiltration for removal



of the fluoborate surfactant complex from solutions was the Millipore
                            t


Corporation's 47 mm stirred cell, catalogue number XX 42 047 10.  The



membranes were Milliporefs Pellicon membranes, PSAC type.



     The two membranes evaluated were the 1,000 and 10,000 nominal molecular



weight limit (nmwl) membranes.  The rnnwl is a rough guide to the size above



which mrst molecules are efficiently retained by that membrane.



     The siaple, stirred cell, which is depicted in Figure 3, is a cylindrical



plastic unit designed to hold membrane discs for the ultrafiltratlon of small





                                    -19-

-------
                            Pressure = 238 cm.  of  Hg.
                  Feed Solution
             I   Stirring Bar
      Magnetic
      Stirrer
Teflon Ring
Membrane
•Perforated Spacer
                            Filtrate-
FIGURE 3.  SCHEMATIC OF A SIMPLE  ULTRAFILTRATION CELL
                        -20-

-------
fluid volumes.  The cell barrell is polycarbonate with silicone 0-rings


for seals at the base and cap.  The capacity of the cell is 80 ml, the cell

                                                                  f\
takec a 47 mm membrane disc which has a filtration area of 10.5 cm .


     In the evaluation of the two different membranes, the set-up


that is shown in Figure 3 was used.  Seventy-five millillters of solution


was placed in the cell.  The pressure applied to the feed solution was


maintained at 238 cm. of Hg.  The cell was mounted on a magnetic stirrer


and stirred by means of a magnetic stirring bar inside the cell.  This


minimizes polarization occurring at the membrane.  The filtrate was


collected in test tubes from the plastic tube which was inserted in the


base of the cell.


     Initial experiments were run using a solution of 1500 mg/1 of alkyl-


amine acetate surfactant and 100 mg/1 of fluoboratj anion.  During the


evaluation of the 10,000 and 1,000 nmwl membrane a dye test was used to


Indicate the presence of alkylamine acetate surfactant in the filtrate.


The dye test provided qualitative evidence of the presence or absence of


alkylamine acetate in the filtrate solution by visual inspection.


     Bromophenol blue, which was the dye that was used, gave a rough


quantitative indication of the surfactant In solution.  When the surfactant


was present, a blue color was' formed in the solution.  When no surfactant


was present the solution remained clear and colorless.  The deepness of


the blue Increased as the concentration of surfactant increased.  There-


fore, the dye test was a good preliminary indication of the passage of the


surfactant fluoborate complex through the ultrafiltration membrane.
                                    -21-

-------
      The surfactant concentration in the ultrafiltratlon filtrate solution




 was analyzed using a Dorhmann Envirotech DC-52D Carbon Analyzer.   The




 presence of fluoborate was observed to have no effect on this analysis.



 b.   High Volume Cassette




      Further evaluation of ultrafiltration for the removal of the fluoborate-




 surfactant complex from solutions was carried out using the 100 nomimal




 molecular weight limit membrane in a Millipore High Volume Cassette  System.




 The cassette system is 25 cm wide x 23 on. deep x 30 cm.  high.  Membranes  for



 this system have a filtration area of 465 cm2.




      The systec 'vas run utilizing a variable speed tubing pump  to transport




 pressurized solution to the cassette's feed port.  Pressure at  the feed




 port was regulated by the pump speed.  A needle  valve was placed at the




 retcntate port in order to maintain pressure over the entire sesahrane area.




 Using this system, two types of experiments were run:   they were  single pass




 flow and recirculating flow as depicted in Figures 4 and  5, respectively.




 During both experiments the same feed solution containing 1500  mg/1  of




 alkylamlne acetate surfactant and 100 mg/1 of fluoborate  was used.




      Operating pressures were maintained at 155 en of Hg  at the feed port




 and 109 cm of Hg at the retentate port, during the single pass  flow  experi-




 ment.   Due to the viscosity of the feed solution, only 5.7  ml of  filtrate  vas




 collected per liter of feed.   In order to remove more fluid from  the sample



 than can be accomplished in one pass, the recirculating flow system  was



 investigated.




      During the recirculating flow experiment the retentate was run  back




 into the sample vessel.   The sample vessel was mounted on a magnetic stirrer




and the solution was  stirred  to maintain  a  homogeneous  feed.  The pressures
                                    -22-

-------
Pump
                                   Valve
Sample      Retentate
                                                                        Filtrate
                 FIGURE 4.  SINGLE PASS FLOW SYSTEM

-------
                 Pump
                                  Cassette -
Sample
Filtrate
                             FIGURE  5.  RECIRCULATING FLOW SYSTEM

-------
were maintained at 155 cm. of Hg at the feed port and 114 on. of Hg at the


retentate port.


     The filtrates from both the single pass flow and the recirculating flow


experiments were analyzed for TOG to determine the surfactant-fluoborate


complex removal.


Breaking the Fluoborate-Surfactant Complex


a.  Addition of Excess Acetic Acid


     In an attempt to reverse the equilibrium of the surfactant-fluoborate


complex and break the complex, the addition of excess acetic acid was


examined.  A one liter solution of 100 mg/1 of fluoborate and 1500 mg/1


of alkylamine acetate surfactant was used.  Concentrated acetic acid was


pipeted into this solution at an initial volume of 0.5ml.   The solution was


stirred utilizing a magnetic stirrer.  The concentrated acetic acid was added


in increments of 0.5 ml.  After each addition of acid, the concentration


of fluoborate anion was monitored using the Orion 92 series electrode


system.


b.  Electrolysis


     Electrolysis was also investigated as a means for separating the


surfactant-fluoborate complex.  A simple electrolysis cell was set up as


shown in Figure 6.  The components were a pyrex beaker, a spiraled platinum


anode, a copper cathode, and a direct current (D.C.) power supply.


     The D.C. power supply was kept  constant at 45 volts (12 ripple).  This

                                                                       ^
provided a current density at the cathode of approximately .004 Amps/cm .


     A surfactant-fluoborate solution of 1500 mg/1 of alkylamine surfactant


and 100 mg/1 of fluoborate  anion was added to the system and then the power


supply was turned on.  Using the 92  series fluoborate specific ion electrode,
                                    -25-

-------
Platinum
Anode
Copper
Cathode
                      Surfactant-Fluoborate
                            Solution
             FIGUKE 6.  SCHEMATIC OF ELECTROLYSIS  SET-UP
                                -26-

-------
the fluoborate concentration was determined periodically.  These concentra-


tions were then plotted versus time to determine the rate of disassoclation


of the complex.


     Using the above system, disassociatlon of the following four solutions


were examined:


     1.  Solution A: surfactant (1500 mg/1) + sodium fluoborate (100mg/l);


                     stirred for 90 minutes.


     2.  Solution B: Same constituents as Solution A; stirred for 8 hours.


     3.  Solution C: surfactant (1500 mg/1) + lead and tin fluoborate,


                     fluoborate anion = 100 mg/1; stirred overnight.


     4.  Solution D: surfactant (1500 mg/1) + lead and tin fluoborate,


                    fluoborate anlon = 100 mg/1 + brighteners;


                    stirred overnight.


2.  Electrodlalysis


     Elrctrodialysis was investigated here as a means of concentrating


fluoborate plating rinse-waters for recycle back to the plating bath.


All plating bath rinse-waters were generated by diluting plating bath


soluticns with deionized water to desired metal cation and fluoborate


anion concentrations.  A Micro Pore (Hanover, Mass.) electrodialysis stack
                                                                     i

was used to carry out the electrodialysis experiments.  The stack consisted


of thirteen cell pairs and was arranged as illustrated in Figure 7.  Each


cell pair consisted of a Neosepta strongly acidic cation permeable (C66-5T)


and strongly basic anion permeable (AFN) membranes (Tokuyima Soda Co Ltd.,


Japan).


     The size of the membranes and electrodes was 55.9 cm x 30.5 cm.  Appro-

 ximately 63 percent of the area of the membranes was used for ion transfer.
                                    -27-

-------
I
N>


ncentrat






I






V
•a

1
Feed







Cathode-


L



i
f \
.. t It
p
A I C


A;

I
1
<-

P



A C

'C+
A\

i




-




*
A'




?~

Flows
Reservoir
r
A 1 C

H
\


i
<

^>



^
A"




k


L
1 f
A C

h
\


1
X

>>



A

£+
AV

1














4


^ k
r



-Anode






f7 Flows

-M










A = Anion Selec
Membrane
C = Cation Jfelec
Membrane
A = Anion
.
C = Cation



„
Electrolyte
Reservoir
I



                                                             Cell-Pair
                         FIGURE 7.  SCHEMATIC OF KA ELECTRODIALYSIS STACK

-------
The spacers and membranes delineated an S-shaped flow pattern which incor-


porated the advantages of the more common sheet flow and tortuous path


designs (see Figure 8).  Alternate cells were connected to internal feed or


product manifolds.  An additional anion permeable membrane was used at the


cathodes to prevent migration of the cations into the electrolyte and being


reduced at the cathode.1  The ends of the stack were terminated with the two
 i

electrodes.  The cathode was stainless steel 316 while the anode which was


originally platinized titanium metal but had to be replaced because of


corrosion problem.  A hydraullcally Isolated electrode rinse solution


consisting of a mixture of hydrofluoboric acid and sodium fluoborate was


recirculated from an electrolyte reservoir.


     Other related supplies Included pumps for both the feed and electrolyte


solution, reservoirs for 114 liters of feed and 38 liters of electrolyte


collecting vessels for the concentrated products, and an exhaust over the


electrolyte reservoir to remove electrolyte gases, a rectifier and


necessary piping and valving.  The complete installation is shown in Figure 9.


     Initiall>, the electrolyte was recirculated through the cathode and anode


cells at a rate of  133 1/hr., while the feed solution was recirculated


through the feed cells at a rate of 456 1/hr.   The D.C. power supply was


then turned on and the current was brought up to 10 amperes by slowly


increasing the voltage across the electrodes.  The flows were regulated by


gate valves to maintain a pressure of 52 cm. of Hg at both the electrolyte


and feed inlets.  Also, the temperature of the solution at the feed outlet


was monitored in order to prevent possible membrane damage due to the over-
                                                         i
heating of the stack.  At the end of each experiment, the flows were run for


2-3 minutes after the current was shut off.  Also,  between experiments the
                                     -29-

-------
FIGURE 8.   SPACER ON GRAPHITE ANODE TO DEFINE S-SHAPED  FLOW PATTERN
                                -30-

-------
FIGURE 9.  ELECTRODIALYSIS UNIT FEED RESERVOIR (tank on right)




             AND ELECTROLYTE RESERVOIR (lower middle right)

-------
 stack was flushed with distilled water until no metal or fluoborate could



 be  detected.




      Feed,  product and electrolyte samples were collected initially,  at




 regular Intervals during each run and at the completion of each  run and




 analyzed for  metal concentrations, fluoborate concentrations and pH.   The




 pH  was monitored to indicate whether there was a significant transfer of




 hydrogen ions from the feed to the product.   The electrolyte was analyzed




 to  detect metal ion breakthrough frons the concentrate to the electrolyte  at



 the cathode.




      The concentration of total tin, copper  and lead  were analyzed  after




 serial dilution with fluoboric acid solutions using a Varian Techtron




 Atomic Absorption Spectrometer (Model 1200).   The serial dilutions  were




 carried out to reduce the metal ions to levels that could be measured.




 Serial dilution of the samples with fluo boric acid solution at the  same pH




 as  the samples being analyzed was found to prevent precipitation of metal




 prior  to  analysis.




     Stannous Ion concentrations were determined according to Method 155




 (14) by tltration by 0.1 N  and 0.01 N iodine  solution to  a blae  iodine-




starch  complex end point.   The fluoborate ion concentration was  calculated




using  the Orion Model 93 series Pacific ion  electrode.
                                    -32-

-------
                        V.  RESULTS AHD DISCUSSION
1.  Ion Flotation




     This part of the Investigation was concerned with identifylug tbe type




of surfactant which would bind to the fluoborate ion.  A number of cationic,




anionlc and non-ionic surfactants were initially evaluated using sodium




fluoborate solutions.  The cationic surfactants were primary, secondary and




tertiary amines attached to various organic radicals.  The non-ionic surfac-




tants were the oxide of an aromatic amlne, an aliphatic acid and acyclic




ether.  The anionlc surfactants were of the organic phosphate type.



     Significant fluoborate removal was observed with only two of the




cationic surfactants, Duomac T and Armac C.  Approximately 55 percent




removal of the fluoborate was achieved with Armac C, the better of the two




surfactants, during the Initial surfactant screening.  All of the other




surfactants tested showed no significant removal of the fluoborate Ion.




The Duomac T and Armac C differ from the other surfactants examined In that




both surfactants contain acetate groups.  Apparently, the presence of the




acetate group Is required for the removal of the fluoborate by specific




ion flotation.




     Fluoborate ion removal could be increased from 55 to 132 percent by




Increasing the mole ratio of Armac C to fluoborate from 1.0 used in the




initial screening to 3.0.  A series of tests were performed on solutions




containing different initial concentrations of sodium fluoborate and




varying the specific mole ratios of Armac C to fluoborate ion to determine






                                    -33-

-------
the concentrations  tc  vhich Axmac  C could reduce the fluoborate Ion.  The




results indicate  that  At Initial fluoborate concentrations of 100 mg/1 and




below, the Armac  C  can reduce  the  fluoborate  concentration to approximately




18 mg/1, using a  surfactant to fluoborate mole ratio of 3:1 (see Figure 10).




When the initial  fluoborate concentration is  raised to 150 gm/1, the Armac




C reduced the fluoborate concentration to only 48 mg/1 using the same mole




ratio of 3:1.  These reoults appear to be due to the fact that not all of the




surfactant required to provide a mole ratio of 3:1 to treat a fluoborate




concentration of  150 mg/1 can  be dissolved.   An extremely turbid solution




resulted.  Thus,  even  though a 3:1 mole ratio of surfactant to fluoborate




ion was prepared  not all of the surfactant was available to react with




fluoborate ion.



Factors Influencing the Removal of Fluoborate by the Surfactant




     The results  of the initial screening of  the surfactants suggested that




the acetate group on the surfactant molecule  is Involved in the removal of




fluoborate ion.   In order to test  this further, factors which Influence the




removal of the acetate ion from the surfactant were examined.  Sodium




acetate was added to the Axmac C surfactant so as to reduce the ionization




of the acetate group.   A 100 mg/1  of Armac C  acetate solution when added to




the fluoborate solution to provide a 3:1 mole ratio of surfactant to fluo-




borate resulted in  only 63 percent removal of fluoborate ion.  When no




sodium acetate was  added,an 82 percent removal of fluobcrate was obtained.




These results Indicate that the binding of the fluoborate  requires the




removal of the acetate group from  the surfactant through Ionization of




the surfactant to form a cation and an acetate ion.



     If the above is the case, an  Increase in acidic conditions should also





                                     -34-

-------
    70
    60
    50
eo
5
u
o
.a
§
    30
    20
    10
Starting fluoborate concentration

   £  150 mg/1
   O  100 mg/1
   O   50 mg/1

                       25 mg/1
                         I
                      I
I
                  1.5         2.0         2.5         3.0

                      Mole ratio surfactant to fluoborate

            Figure 10.  TREATMENT OF DILUTE SODIUM FLUOBORATE
                                 SOLUTION WITH ARMAC C
                                                   3.5
                                 -35-

-------
favor the removal of fluoborate  by an acetate surfactant since acidic

conditions would favor the formation of the surfactant cation through removal

of the acetate ion from solution in the i'orm of acetic acid.  A reduction In

the final pH of the surfactant-fluoborate solution was found to favor tbe


removal of the fluoborate from solution by the Armac C.  The use of fluoboric

acid Instead of sodium fluoborate In the mole ratio of 3:1 surfactant tc

fluoborate decreased the final pH from 6.2 down to 5.0.  The removal of

fluoborate vas Increased from about 87 percent up to 92 percent (see Table 1).


     The' Influence of other anions which form stronger acids than fluoboric

acid on the removal of the fluoborate by the surfactant uas examined.  Hydro-


chloric acid was added to separate solutions containing the 3:1 mole ratio of

surfactant to fluoborate 'ion in  amounts sufficient to provide final solutions


that contained 50 mg/1 and 100 rag/1 of hydrochloric acid.  Fluoborate ion

removals of only 70 percent and  58 percent were obtained In the solutions

containing the 50 mg/1 and 100 mg/1 of hydrochloric acid, respectively.  Tbjus,

the presence of anions which form stronger acids than fluoboric acid appar-

ently inhibit the removal of the fluoborate ty the surfactant.

     The above results indicate  that the fluoborate ion reacts with the
                   i
surfactant by replacing the acetate group on the surfactant since it forms a


stronger acid than acetic acid.  Ho waver i maximum removal of the fluoborate

in these experiments is achieved with a surfactant to fluoborate mole ratio


of 3:1.  The use of the 3:1 mole ratio Indicates that 3 acetate groups are

Involved In the removal of one fluoborate ion since each Armac C surfactant

molecule contains only one acetate group.  The need for excess surfactant


suggested that all acetate groups are unavailable for replacement by the

fluoborate ion.


                                     -36-

-------
TABLE 1.  Effects of pH on Removal of Fluoborate Ion

mg/1 Fluoborate as
HBF^ NaBF4
100
75
50
40
30
25
0
0
25
50
60
70
75
100
Initial
3.7
3.8
4.1
5.4
7.1
6.7
7.4
pH
Surfactant
6.0
6.0
6.0
6.0
6.2
6.1
6.0
Final
5.0
5.6
5.8
6.2
6.2
6.1
6.3
Z Removal
92
92
89
86
83
87
87
                        -37-

-------
     The results of three atepwlae additions, maintaining a mole ratio of 1:1




of surfactant to fluoborate remaining in the resulting solution after




removal of the surfactant-fluoborats complex reduced the amount of surfac-




tant required to remove a specific number of molecules of fluoborate.




     Only 2.06 millimoles (554 mg) of total surfactant red-ices 1.15 milli-




moles of fluoborate (100 mg)  in one liter to 0.21 millimoles (18 mg) by




stepwise addition of decreasing amounts of surfactant (see Table 2).  The




fluoborate which reacted with the surfactant was removed by aeration after




each addition of surfactant.  In contrast, a single addition of 3.45 milli-




moles (810 mg) of surfactant  was required to reduce the 1.15 millimoles of




fluoborate in one liter of solution to the 18 mg.



     Micelles were observed to form at surfactant concentrations above




12 mg/1.  This could limit the number of surfactant acetate groups that



are available for replacement by the fluoborate. 'Since the fraction of the




total surfactant molecules in solution that exists as micelles generally




decrease as the surfactant solution becomes more dilute, the step wise




addition of decreasing amounts of surfactant to a given amount of fluo-




borate would Increase the fraction of total 'acetate available for replace-




ment by the fluoborate ion.   Thus, it appears that step wise addition of very




dilute solutions of surfactants to minimize the micelle formation should




require less surfactant to react with a specific amount of fluoborate.




     In order to further understand the selective fluoborate ion removals




exhibited by the surfactant containing the acetate group, surfactants similar




to Armac C but differing in carbon chain length and degree of saturation were




Investigated.



     The effect of surfactant molecular size on the removal of the fluoborate





                                      -38-

-------
TABLE 2.  Fluoborate Removal Using Stepwise Addition of Anaac C Surfactant

                           Co Sodium Fluoborate
        Concentration of
            Annac-C
     ag/1

     310

     160

      84
millimoles


   1.15

   0.60

   0.31
                Concentration
                of Fluoborate
ng/1


100

 56

 27
millimoles
 -'	

   1.15

   0.60

   0.31
Fluoborate Concentration
remaining after aeration

        mg/1

         52

         27

         18
                                    -39-

-------
ion in a solder plating  bath rinse-water containing 100 mg/1 of fluoborate




ion UBS studied using Armac C  (12-15 carIon atoms), Armac T (16-18 carbon




~ corns) and Armac 18D (18 carbon atoms) .  The results show that as the chain




length of the surfactant is Increased  from the Cjjj.is to C18» tne Percent



fluoborate removed is increased from approximately 80 percent up to 97




percent when using the surfactant to fluoborate mole ratio of 3:1 (see




Figure 11) .  The treatment of  the solder bath rinse-water with the surfactant




containing the 18 carbon atoms results in only .» mg/1 of the fluoborate ion




remaining in the rinse-water.  Reduction of the fluoboric ion concentrations




in the solder rinse-water to below 3 mg/1 using surfactants with chain




lengths greater than 18  carbon atoms was not possible because of the marked




decrease In solubility exhibited by the surfactants with larger molecular




sizes than Armac 18C.



     The degree of saturation  of the surfactant does not appear to effect the




removals of the fluoborate ion.  The removals of fluoborate ion in solder




bath rinse-water achieved with Armac HT whose percent saturation is 97




percent is comparable ".o that  achieved with Armac T whose percent saturation




is only 58 percent.  This Is observed  for the range of surfactant to fluo-




borate ion mole ratios studied (see Figure 12).



     The reaction of the fluoborate ion with the surfactant occurs within




a relatively short period of time.  The mixing of Armac C with fluoborate in




mole ratio of 3:1 resulted in  the reduction of fluoborate ion from 87 mg/1




to 18 mg/1 within 1 minute (see Figure 13).  These results indicate that




a contact time of one minute between the surfactant and fluoborate is




required for the replacement of the surfactants' acetate group by the




fluoborate ion prior to  removal by aeration.





                                     -40-

-------
100
                             Q  Armac-18D (Cjg)

                             O  Armac-T

                             A  Annac-C
               1234
               '  Molar Ratio of Surfactant to Fluoborate

   FIGURE 11. FLUOBOKATE REMOVAL FROM SOLDER RINSE-WATER
                DEPENDENCE  Oil SURFACTANT MOLECULAR SIZE

-------
    100
     90
O
£~   80

-------
     90
     80
     70
    60
 §.
 o.
•o
41
u
U
a
ai
M

3
Vi
O
Jl
    20
         1
n
                          5                10


                               Time (Minutes)
                                     15
       FTCTRE 13.  TIME OF REACTION OF FLDOBORATE WITH SURFACTANT
                            ••43-

-------
Process  Parameters

      Changes in air feed rate, air bubble size,  air diffuser location,  Inlet

feed  direction, and mixing time on the removal of the surfactant-fluoborate

complex  from solution by aeration were investigated because these  factors can

Influence  the performance of bubble columns (15, 16).

      Varying the air feed rate from 4 cc/sec.  to 30 cc/sec.  using  a  3:1. mole

ratio of Armac C to fluoborate did not improve the rate  of  removals  of

fluoborate and surfactant from solution.   In both cases, time in excess of

20 hours was required to remove all of the surfactant.   However, an  Increase

in the air feed ratios does augment  the  amount  of solution that is  carried

over  in  the foam.   This results in a wetter foam and a reduction in  the

concentration of the surfactant-fluoborate complex in the foam.  For example,

twa foaming operations were carried out at air feed rates of 15 cc/sec. and

30 cc/sec.  were used.   The lower air feed rate provided  a resulting  foam with

a surfactant-fluoborate complex concentration  of 4000 mg/1  whereas the higher

air feed rate provided a foam with surfactant-fluoborate complex

concentration of less  than 1000 mg/1.

      Neither changes in bubble size, air  diffuser location nor inlet feed

directions  with respect to the rising air bubbles were observed to

influence  the removal  of the surfactant-fluoborate molecule  in the bubble

column.  Separate  experiments using 3:1 mole ratio of Armac  C to fluoborate

ion In which 1)  initial air bubble diameters of  0.003 and 0.007 were
                                                      i
generated  in the column,  2)  positioning the air  diffuser location  at 10 cm.

and 30 cm.  below the surface of the liquid In  the columns to  increase the

contact  time between the bubble and the solution,  and 3)  introducing the feed

into  the column  counter-current as well as perpendicular to  the rising air
                                     -44-

-------
bubble did not Improve ths rate of removals of the surfactant-fluoborate




complex.  Times in excess of 20 hours were required to remove ail of the




surfactant for each of the above experiments.  However, these results




Indicate that significant variation can occur in the bubble size, and




contact time between the air bubble and the solution without influencing the




rate of removal of the fluoborate-surfactant complex.




Removal of Fluoborate from Plating Rinse-Waters



     A series of tests were performed on rinse-water containing 100 mg/1 of




fluoborate ion by adding different amounts of Anaac T to different rinse-




waters that are representative of that obtained from solder plating, tin




plating, nickel stripping, and copper plating operations.  Armac T rather




than Armac 18D which provided greater removals of fluoborate Ion was used for




these studies because it is far less expensive.  It was felt that in practice




the slightly better removals achieved earlier with Armac 18D did not warrant




the additional expense.



     The results indicate that Armac T can reduce the fluoborate concentra~




tlon from 100 mg/1 to approximately 7 mg/1 in the solder, tin and nickel




stripping rinse-water using a surfactant to mole ratio of 3:1.  The




fluoborate Ion in the copper plating rinse-water is only reduced to



approximately 15 mg/1 (see Figure 14).  This Is probably due to the fact that




the pH of the copper plating rinse-waters is less acidic than the other




plating bath rinse-water.



     The above results are significant in that they Indicate that the removal




of fluoborate with surfactant is comparable with that which can be achieved




with li>e precipitation of fluoride in wastestreams.  The widely used lime




precipitation of fluoride In uastestreams results in a fluoride ion residual






                                      -45-

-------
      70
  ao
  g
 O
  i
  1
     SO
     SO
     30
     2Q
     10
                                     O  Solder Rinse Waters

                                     O  Tin Rinse Water

                                     O  Nickel Stripping Rinse Water
                                         Copper Rinse Water
                         _L
              JL
       312345

                         Mole  ratios of surfactant to fluoborate


        FIGURE 14.  REMOVAL  OF FLUOBORATE FROM PLATING BATH
                           RINSE WATERS WITH ARMAC T
-46-
                                         i 0'. r".2E
F- - .-B7 :c

-------
of about 8 mg/1 in the wastestream (17).  Thus, where fluoride concen-



trations of 
-------
 alternate more rapid process for removing the surfactant-fluoborate complex.




 Evaluation of the ultrafiltration membranes utilized a low through-put




 pressure cell.  In the cell, experiments were run with 10,000 and 1,000 molecular




 weight cut-off membranes.  Experiments with the 10,000 molecular weight cut-




 off membrane were unsuccessful.  The solution passed right through the



 membrane.




      The 1,000 molecular weight cut-off membrane was effective in filtering




 a solution of plating rinse-water containing fluoborate ions complexed with




 the Armac T surfactant.  Removals of Approximately 79 percent were achieved




 (see Table 3) in the low through-put cell as determined by total organic



 carbon measurements.




      Although the molecular weight of Armac T surfactant is approximately




 330,  it was determined from conductivity that the Armac T forms micelles




 at concentrations greater than 12 ppm.  Therefore, at the concentration of




 surfactant that is required to complex the fluoborate,  it will exhibit an



 apparent molecular weight much greater than 330.




      Experiments utilizing the 1,000 molecular weight cut-off membrane by




 recycling the retentate in a Millipore Cassette ultrafiltration apparatus




 showed favorable results.   Recycling the retentate back into  the ultrafil-




 tration cassette resulted in  a removal of surfactant of 85 percent  (see Table




 4).




      These ultrafiltration experiments produced the same results as air




 flotation.  However,  times In excess of 20 hours  were required to achieve




 maximum removal  of the  surfactant-fluoborate complex from solution using




 air flotation.   The  recycled  ultrafiltration experiments achieved this




separation  in approximately 5 hours.   The ion flotation time  cannot be
                                     -48-

-------
FIGURE 15.  ELECTROLYSIS OF SURFACTANT FLIJOBORATE
            CONCENTRATED SOLUTION FROM TREATMENT
                OF SOLDER BATH RINSE-HATER
                  -49-

-------
TABLE 3.  Removal of Surfactant with a Low Through-Put Ultraflltration




                                 Experiment
Solution
Initial
1st filtrate collected
2nd filtrate collected
3rd filtrate collected
4th filtrate collected
5th filtrate collected
Total
Z Surfactant fluoborate
Volume (ml)
75
1.2
2.6
10.6
43.0
2.4
59.8
initial cone.
I
TOC (rag)
80
0.13
0.56
2.12
10.06
0.54
13.5
- filtrate cone. .„ ..
       retained by membrane     initial cone.
                                  -50-

-------
TABLE 4.  Results of Recycled Retentate Experiment, Hlllipore Cassette
Solution
Initial
1st filtrate collected
2nd filtrate collected
3rd filtrate collected
4th filtrate collected
Total
% surfactant fluoborate
Volume (ml)
1,000
20
50
10
50
130
initial cone.
TOG (ing)
1,073
1.72
8.65
1.65
9.25
21.3
- filtrate cone. lofl „ a.
       retained by membrane    initial cone.
                                   -51-

-------
reduced  by  Industrial scale-up;  however,  the  time required for ultrafUtratlon



can  be minimized  by  Increasing the number of  membranes used In the ultra-



filtration  unit.



Electrodlalysls



     Countercurrent  or series  rinsing  of  products from fluoborate plating



operations  can lead  to fluoborate  concentrations equal to or in excess of



1000 mg/1.  For these highly concentrated wastewaters ion flotation



treatment is not  suitable.  Thus,  electrodialysis was investigated as a means



to pretreat these concentrated rinse-waters to recover and reuse the plating



reagents lost to  rinse-waters  and  reduce  the  plating chemicals in the



effluent to levels suitable for ion flotation  treatment.



     Application  of  electrodialysis for treatment of rinse-water containing



high concentration of plating  reagents required modification of an existing



electrodialysis unit.   Rapid deterioration of the commonly used platinized



titanium anode was encountered when fluoboric acid is used as the electrolyte.



Although this anode  has been successfully used with nickel sulfate plating



bath rinse-waters and a sulfuric acid  electrolyte, it quickly turns black,



cracks, and peels from the titanium metal backing when voltage is applied in



the presence of fluoboric acid electrolyte (see Figure 16, Photograph of



Corroded Platinum Anode).  Apparently, the smaller fluoborate ion can



penetrate the porous  platinum  coating  and corrode the metal bond between



the platinum and   the titanium.



     The matching of  the an ion in  the  electrolyte with that in the rinse-



waters is preferred  to using another anion such as sulfuric acid electrolyte



which is non-corrosive to the  platinized  titanium anode.  The sulfuric acid
                                                                i


electrolyte would result In introduction  of undesirable sulfate anion Into
                                   -52-

-------
FIGURE 16.  PHOTOGRAPH OF CORRODED PLATINIZED TITANIUM ANODE
                          -53-

-------
 the concentrate stream and ultimately the plating bath.




 Evaluation of Different Types of Anodes




      Several different approaches were considered to  obtain an anode that




 would not  corrode In presence of fluoboric acid electrolyte.  These were




 1)  to Increase the thickness of the platinum plated on the platinized




 titanium anode,  2)  to plate a less porous metal such  as gold on nickel




 backing and 3) examine the use of inexpensive conducting materials such



 as  low porcslty high density graphite to which voltage could be applied



 directly.




      An increase In the thickness of the platinum plated on the titanium




 anode from 1.5 micrometers to 3.9 micrometers did not improve the anode's



 ability to withstand the corrosive properties of the  fluoboric acid




 electrolyte.   The platinum cracked and peeled away from the anode within



 an  hour after voltage was applied to the anode ir. the presence of the




 fluoboric  acid electrolyte.




      Gold  when plated on nickel and heated under controlled conditions can



 form  a solid  solution at the interface  between the gold and nickel.  This




 concept was utilized to achieve a bond between the gold and nickel backing




 at  the Interface.   It was anticipated that the solid  solution at the inter-




 face  would resist the peeling and cracking that  was encountered with the



 platinized titanium anode in the presence of the flueboric acid electrolyte.




     A series  of experiments were carried out in which gold was plated on



nickel strips  and heated for 30 minutes and 45 minutes at temperatures of




1GOQ°F, 1250°F and  1500°F, under a reducing atmosphere of'hydrogen.  Micro-




probe scan of  the Interface  between the nickel and gold revealed that




heating the anode at  1250°F  for 45 minutes resulted In the nickel just
                                     -54-

-------
completely diffusing through the gold to the outer gold surface, whereas a




heating time of 30 minutes resulted In the nickelforming a solid solution




through 80 percent of the gold layer.  This latter strip was selected to be



evaluated for Its resistance to corrosion by the fluoboric acid electrolyte




because only the' gold surface would be exposed to the fluoboric acid




electrolyte.



     Application of a voltage to this gold nickel anode strip to achieve a




current density of .07 amps/cm^ while Immersed in the fluoboric acid




electrolyte resulted in the gold layer peeling from the nickel backing




(see Figure 17).  Apparently, the solid solution between the gold and




nickel at the Interface does not resist the corrosive properties of fluo-




boric acid electrolyte.



     Two graphite anode strips obtained from Union Carbide, Carbon Products




Division, Chicago, Illinois, Grade AGXC and Stackpole, St. Mary, Pa. Grade




LXB were evaluated to determine if they are resistant to the corrosive



properties of fluoboric acid electrolyte under an applied voltage.  Visual




Inspection revealed that neither graphite strips exhibited any deterioration




when Immersed in fluoboric acid electrolyte under applied voltage to provide




a current density of .07 amps/cm2.  As a result, two anodes were fabricated




for use In the electrodialysis unit and their performance evaluated under




actual operational conditions.



     The Union Carbide anode was observed to exhibit appreciable leakage of




the fluoboric acid electrolyte from its sides.  Encasement of this anode In




a plastic coating did not prevent this leakage.



    The use of the Stackpole high density low porosity graphite anode of the




electrodialysis unit proved to be successful.  There was no leakage of
                                    -55-

-------
                       '• r

                         ~
                         r  -* r"9
                     - •


                          -••*» -*
FIGURE 17.  PHOTOGRAPH  OF  CORRODED NICKEL ANODE
                     -56-

-------
 fluoboric acid electrolyte fron the anode and a greater current was achieved




 for a given applied voltage than that obtained with  the widely used platinized




 titanium anode.  The use of the graphite anode In the electrodlalysis unit




 containing 13 Ion pairs of membranes produce 10 amperes of current with an




 applied voltage of 10 volts.  In contrast, the use of the platinized titanium




 anode in the electrodialysls unit containing the 13  ion pairs of membranes




 required 15 volts to produce the same 10 amperes of  current.



 Electrodialysls Result




     In the treatment of concentrated rinse-waters prepared by diluting tin,




 solder and copper fluoborate plating bath solutions  was carried out in the




 electrodialysis unit containing the Stackpole high density low porosity gra-




 phite anode and 13 ion pair membranes.  The voltage  across the anode and




 cathode was adjusted to maintain 10 amperes of current for all the experi-




ments.  The effectiveness of electrodialysis was evaluated for reducing the




 concentrations of metal and fluoborate ions In the rinse-water to levels




 which would allow the effluent to be treated using the ion flotation process,




 while providing a product with concentrations of metal and fluoborate ions




 that would approach that needed in the plating baths.




Treatment of Tin Fluoborate Rinse-Waters




     A volume of 114 liters of tin fluoborate rinse-water containing 2400 mg/1




of stamous ion, 525 mg/1 stannic ion and 6500 mg/1  of fluoborate ion was




treated by electrodialysis.  After 12 hours of operation using 13 ion pair




membranes at .012 amperes/on^ current density, the stannous ion concentration




in the feed was below measurable levels,-  The stannic ion was reduced to 75 ng/1




and the fluoborate ion was reduced to concentrations below 150 mg/1 (see




Figure 18).  The stannous ion is reduced to concentration levels approaching






                                    -57-

-------
i,
T)
I
                                                    6           8
                                                       Time (hours)
10
12
                FIGURE  18.   ELECTRODIALYSIS TREATMENT OF TIN FLUOBORATE HINSE-WATERS,
                                             FEED CATION CONCENTRATION

-------
 zero within  8 hours.   The stannic  ion on the other hand Is removed at  a much


 slower rate  than the  stannous  Ion  from the feed over the initial 8 hours.


 The rate of  removal of stannic ion Increases after removal of  Che stannous


 ion.


      The most rapid decrease in fluoborate concentration in the feed appears


 to  occur within the initial 8  hours of operation.   During this period  the


 fluoborate ion concentration is reduced from 6500'mg/1 down to about 500 mg/1


 (see Figure  19).   This amounts to  a removal of  about 92 percent.   The  last


 four hours of operating the electrodLalysis unit reduces the fluoborate


 concentration from 500 mg/1 down to below 150 mg/1.


      The above results are of  interest because  the;  Indicate that the


 plating  chemicals  in  the feed  can  be  reduced to levels acceptable for  ion


 flotation treatment.   Also,  the stannous ion which is  the desirable ion in


 plating  baths  is preferentially removed from the feed.  Thus,  the product  from


 the electrodlalysis unit  should be  enriched with stannous ion.


      The analysis  of  the  initial product from the  electrodialysis unit


 shows a  30 fold increase  in  the stannous and  fluoboiate ions concentration.
                              I

 The  stannous  ion in the Initial product collected  is about  73,000 mg/1  after


 2 hours  of operation and  decreases  to  17,500  mg/1  (see Figure  20}  as the


 concentration of stannous  ion  in the feed  is  further depleted.  The fluoborate


 anion in the initial product is about  200,000 mg/1 and decrease to  100,000


mg/1  in  six hours of operation  (see Figure 21).  This .decrease is  expected


 since the concentration of a specific  ion  in  the product  is reported (13)


 to be proportional  to that in the feed.  The  stannic ion concentration  in


 the initial product is about 7500 mg/1 and decreases to about 5000 mg/1


 within six hours due to reduction in stannic  Ion in  the feed from 565 mg/1
                                    -59-

-------
                                                          10
                                                        12
FIGURE  19.
                Time (hours)

TREATMENT OF TIN FUJOBORATE RINSB-WATERS,
      FEED ANION CONCENTRATION
                              -60-

-------
down to 500 Dg/1.




     The percent stannous ion in the product is observed to remain essen-




tially well above 80 percent even though the stacnous ion in the feed has




been reduced from 83 percent down to 23 percent (see Table S).




     These results reflect the greater rate of removal of the stannous ion




from the feed than the stannic ion.




     The total amount of product generated in six hours from the treatment




of 114 liters of feed was roughly 4.3 liters.  This amounts to a volume




ratio of product to feed of about 1:27.  The concentration of tin and fluo-




borate in the product was 45 gm/1 of stannous ion, 6.7 gm/1 of stannic ion,




and 140 gm/1 of fluoborate ion.  The stannous and fluohorate ions in the




product represent approximately one-half and one-third, respectively, of the



concentration of stannous and fluoborate ions required for plating bath




strength.  The product collected in the initial six hours of operation




represent a recovery of 95 percent of the stannous ions, 48 percent of the




stannic Ion and 70 percent of the fluoborate ion originally present.




Treatment of Solder Fluoborate Rinse-Waters




     The treatment of solder fluoborate plating bath which contains stannous,




staialc, lead and fluoboric ions in Its rinse-waters was carried out under




the same conditions employed for the treatment of the tin fluoborate plating



bath rinse-waters.  Greater than 97 percent reductions in the stannous ion




and lead ions in the feed was achieved in eight hours.  The time of operation




was not extended beyond these 8 hours because it was shown earlier that the



stannic ion and fluoborate ion concentration in the feed could be reduced to




levels acceptable for treatment by ion flotation.




     The stannous ion is again observed to be preferentially removed from the





                                     -61-

-------
s
X
o
Q)
U
          0
                             Time (hours)

          FIGURE  20. ELECT80DIAL7SIS TBEATKENT OF TIM FLUOBORATE.
                        RINSE-WATER,  PRODUCT CATION CONCENTRATION
                              -62-

-------
in
 o
 e
 o
 41
 4J
 a

 o


 1
 O
 4)
 u
        1.5
       0.5
                              Time  (hours)
          FIGURE 21.   ELECTRODIALJfSIS TREATMENT OF TIN FLUOBORATE

                        RINSE-WATER, PRODUCT ANION CONCENTRATION
                             -63-

-------
TABLE 5.  Changes  In Percent  Stannous  Cation  in  the Feed and Product with Time

          During the Electrodlalysls Treatment of Tin Fluoborate Rinse-Water
                               42                     4-2
      Time  (hours)          ZSn   feed             ZSn   product




           0                    83



          0.5                   83                     85



          1.5                   70                     86



          2.75                  67                     87



          3.75                  45                     89



          4.75                  39                     89



          6.00                  23                     72
                                    -64-

-------
feed.  After eight hours of operation, the stannous ion  is reduced from about


275 mg/1 to below 5 mg/1, whereas, the stannic ion concentration in  the feed


was reduced from  about 185 mg/1 to only about 165 mg/1 (see Figure 22).


Reductions in the lead ion are comparable to the stannous ion.  The  lead  ion


concentration in  the feed was reduced from about 190 mg/1 to below 5 mg/1.


     Significant  reductions in the concentration of fluoborate ion in the


feed was also observed.  During eight hours of operation, the fluoborate  Ion


was reduced from  1920 n>s/l to about 220 mg/1 (see Figure 23).  This  reductica


represents a removal of 90 percent of the fluoborate ion in the feed.  Since


the stannous ion  is again observed to be removed at a much faster rate from


the feed than the stannic Ion, the concentration of stannous ion In  the


product should be significantly higher than the stannic.


     Analysis of  the Initial product collected after the first hour  of


operation reveals the stannous ion concentration to be about 1300 mg/1


whereas, the stannic ion is found to be only 1800 mg/1 (see Figure 24).   A 50


fold Increase In  the stannous ion concentration in the product is achieved


over that originally present In the feed.  The Increase  in the stannic Ion


concentration In  the product represents only a 10 fold increase over that


originally present in the feed.  This Is the same Increase as encountered In


the treatment of  the tin fluoborate rinse.  This greater rate of removal  of
  i

the stannous ion  produces a product where most of the total tin is in the


form of stannous  ion.  The stannous ion in the product averages above 80


percent even though the percent stannous ion in the initial feed is  only


62 percent (see Table 6).


     Analysis for the lead ion in the products reveals that its concentration


is comparable to  that found for the stannous ion (see Figure 24).  These

-------
                                                           5
8
                                        Time (hours)




FIGURE 22.  ELECTRODIALVSIS TREATMENT OF SOLDER FLUOBORATE RINSE-MATER, FEED CATION CONCENTRATION

-------
                                          Time (hours)




FIGURE 23. ELECTRODIALYSIS TREATMENT OF SOLDER FLUOBORATE RINSE-WATER, FEED.ANION CONCENTRATION

-------
20000
     012345            678




                                              Time  (hours)




     FIGURE 2A.   ELECTRODIALYSIS TREATMENT OT SOLDER FLUOBORATE RINSE-WATER, PRODUCT ION CONCENTRATIOH

-------
results are expected since the removal of lead from the solder rinse-water




feed as a function of time is comparable to that achieved with the stannous



ion.




     The fluoborate ion concentration in the product collected within the




first hour of operation was found to be about 16,000 mg/1.  This represents




only a 10 fold Increase in the fluoborate ion found in the product over




that present in the Initial feed.  Treatment of the tin fluoborate rinse-




water resulted in a 30 fold increase in the concentration of,f1uoborate




anion in product over that found in the feed.  The reason for smaller




fluoborate anion concentration effect obtained in the product from the




treatment of the solder rinse-water is not clear at this time.



     The electrodialysia treatment of approximately 114 liters of solder




rlnae-water over a period of eight hours produced roughly 3.2 '.iters of




product.  This amounts to a volume ratio of product to feed of about 1:36.



The concentration of stannous ion, stannic ion, lead ion and fluoborate ion




in this product was found to be 6.5 gm/1, Q.6 gm/1, 5.9 gm/1 and 11 gm/1,




respectively.  The concentrations of stannous ion and lead ion in the. product




are about half of the concentrations of stannous and lead ions commonly used



in the solder plating bath.  The concentrations of fluoborate ions in the




product represents about one 40th o£ the fluoborate ion used in the plating




baths. These results represent a 97 percent recovery of stannous Ion and




lead ion,. 88 percent of the fluoborate ion and 11 percent of the stannic ion.




Treatment of Copper Fluoborate Rinse-Waters




     The electrodlalysis treatment of the copper fluoborate rinse- waters were



carried out under the same experimental conditions as the tin and solder




rinse-waters.  The copper ion in the feed was reduced from initially





                                     -69-

-------
TABLE 6.   Changes in Percent Stannous Ion in Feed and Product with Time
         During Electrodialysis Treatment of Solder Fluoborate Rinse-Water
     Time  (hours)           ZSn"1"2 feed             ZSn"*"2 product

          0                     62

          1                     44                      89

          3                     32                      36

          4                     15                      79

          5                      8                      83

          6                      6                      70
                                   -70-

-------
 2500 mg/1 to less  Chan  40 mg/1  lii  approximately  three  hours  (see Figure 25).




 This represents about a 97  percent reduction  In  copper Ion concentration  in




 the feed in three  hours.  In  comparison, a  97 percent  reduction in  tbe




 stannous ion in the  tin fluoborate rinse-water and the solder rinse-water




 required 6 hours and 8  hours, respectively.   A V7 percent reduction in the



 solder rinse-water feed also  required 8 hours.




     The 'greatest  reduction in  the fluoborate ion concentration in  the



 treatment of the copper fluoborate rinse feed was encountered in the initial




 four hours of operation.  The concentration of the fluoborate ion in the




 feed was reduced from 7500  mg/1 to 650 mg/1 during this period  (see Figure




 26).  This reductiou amounts  to 90 percent removal of  the fluoborate ion  in




 the feed.  In comparison, the electrodlalysis treatment of both the tin and




 solder fluoborate  rinse-waters  required 8 hours  to achieve the  -same removal




 percentages.  Apparently, the presence of stannic ion  in the feed of the  tin




 and solder rinse-water   which is removed at a. much slower rate  than the



 stannous and lead  ions  may  inhibit the removal of the  fluoborate ion.




    The analysis of the products collected from  the electrodialysis of




 copper fluoborate rinse-water after about I's  hours of  operation showed a  20



 fold increase In cupric and fluoborate ion over  their  initial concentration.




The cupric and fluoborate ion In the product  va's about 53,000 mg/1  and




 160,000 mg/1, respectively after l*s hours of  operation  (see Figure  27).



     In comparison, the initial stannous and  fluoborate ion product concen-




tration collected from  the electrodlalysis of the tin fluoborate rinse-water




showed a .J fold increase over that present in the initial feed.  The initial



stannous and lead ion product concentration from electrodialysis  of the



solder hath rinse-water showed about a 50 fold increase whereas  the initial





                                     -71-

-------
 3000
2500
                       Time  (hours)

     FIGURE 25.  ELECTRODIALYSIS TREATMENT OF COPPER
                    FLDOBORATE RINSE-WATER,  FEED
                        ANION CONCENTRATION
                       -72-

-------
             123            4
                            Time (hours)

FIGURE 26. ELECTRODIALYSIS TREATMENT 07 COPPER FLBOBORATE
             RINSE-WATER, FEED ANION CONCENTRATION
                            -73-

-------
   18
   16
   14
<=>
BO
e -
   10
o
01

                                                                        u
                                                                        O
                                                                        u
                              Time  (hours)


     FIGURE 27.  ELECTRODIAIYSIS TREATMENT OF COPPER FLUOBORATE

                     RINSE-WATER, PRODUCT ION CONCENTRATION


                               -74-

-------
fluoborate Ion product concentration showed only a 10 fold Increase.



     The electrodialysls of the copper fluoborate rinse-water over a period




of slightly less tnan 6 hours produced roughly 7 liters of product.  This



amounts to a volume ratio of product to feed of 1:17.  The concentratlou of




cupric Ion and fluoborate ion present ;ln this product was found to be about




40 gm/1 and 120 gm/1 respectively.  The cupric concentration In the product




is about 3 times more concentrated than plating bath strength and the fluo-




borate Ion concentration Is about one-third that generally used in a



copper plating bath.




Rates of Mass Transfer



     The results of the treatment of the fluoborate plating bath rinse-




vaters indicate that the rate of mass transfer of the ions actoss a square




centimeter of membrane surface from the feed to the concentrate Is dependent




upon the cations In the rinse-waters and plating bath rinse-water that is




being treated.  The rate of mass transport across one square centimeter of




membrane'surface for the stannous ions from both the tin and solder rinse-




water feed, respectively are comparable (see Table 7).  The lead ion in the



solder rinse-water feed is also removed at approximately the same rate




(1.4 mg/hr. on2) as the stannous ion (1.7 mg/hr. cm2, 1.9 mg/hr. cm ).  The



stannic Ion on the other hand shous a mass transport rate that is some 10




times slower than the rates obtained with the stannous and lead tons.  These




results cannot be explained at this time because the charge on the cation




does not appear to influence the mass rate of transfer.  The stannic cations




with a +4 charge would be expected to move across the membrane at a greater




rate than a cation with a +2 charge.  Yet the copper ion has the same +2



charge as the stannous and lead Ions, but exhibits a 20 tilmes faster rate of





                                    -75-

-------
 transport (41 mg/hr. on2) then either the stannous or lead ions.

      The rate of mass transfer of the fluoborate anlon also appears  to  be

 dependent upon the plating rinse-water.   The rate of mass  transport  exhibi-

 ted  by  the fluoborate anlon in the copper bath rinse-water is  significantly

 higher  than that encountered in either, the tin or solder bath  rinse-water
                                                           <*
 (see Table 7).   The rate of mass transport is 94 mg/hr. cm  in the copper

 rinse-water following in decreasing order by 28 mg/hr.  cm2 and 12 mg/hr. cm2

 in the  tin and solder rinse-water,  respectively.   The difference in mass

 rate of  transport between the fluoborate anlon in the treatment of the  tin

 and  solder rinse-waters cannot be explained at this  time.   Since the rate of

mass  transport  for the stannous and stannic ions  between the tin and solder

rinse-waters are comparable,  one would expect the rate of  mass transport for

the  fluoborace  anlon to be similar.

Analysis  of Electrolyte

     The  electrolysis or hydrolysis of the electrolyte as  evidenced by

changes of pH and fluoborate  concentration was not observed.  Both the pH

and fluoborate  concentration  in the -electrolyte remained constant during

the electrodialysls  treatment  of  the tin,  solder  and  copper fluoborate

rinse-waters.
                                    -76-

-------
TABLE 7.  Rate of Mass Transport of Cations and Fluoborate from the Tin, Solder,



                               Copper Fluoborate Rinse-Waters
                                  Plating Bath Finse-Vaters

                             Rate of Mass Transport,.

                                         (mg/hr. cm*)

           Ions          Tin fluoborate     Solder fluoborate   Copper fluoborate




          Sn*2                 1.7                 1.9



          Sn"**                  .19                 .14



          Pb+2                                     1.4



          Cu+2                                                             41



          BF.~                28                  12                       94
            4
                                   -77-

-------
                                VI. REFERENCES








 1.  "Waste Treatment - Upgrading Metal Finishing Facilities to  Reduce




      Pollution", EPA Technology Transfer Seminar Publication, January, 1974.




 2.  Sharp, D.W.A., "Fluoboric Acids and Their Derivatives", London, England:




     Bullersworth, 1960, pp. 69-80.




 3.  Cagnatl, V.N., R. Haralson, G. Hunter,  J.W.  Llskoultz,  A. Perna, and




     R.  Trattner, "Removal of Fluoborate From Plating Wastewater:  Technique




     and Mechanism", Water - 1977, American Institute of  Chemical  Engineers,



     pp. 309-315.




 4.  Eckenfleder, W.W., Jr., "Industrial Pollution Control", McGraw-Hill




     Book Company, New York, New York,  1966.




 5.  Fair,  G.M.,  J.C. Geyer, and D.A.  Okum,  "Water and Waste Water




     Engineering", Vol. 2, John Wiley  & Sons,  New York, New  York,  1968.




 6.  Shneve,  R.N., "Chemical Process Industries",  3rd Edition, McGraw-Hill




     Book Company, New York, New York,  1967.




 7.  Sebba, F., "Ion Flotation", Elsevler Publishing  Co., 1962




 8.  Grieves, R.B.and G.A. Ettelt, "Continuous Dissolved Air Flotation of




     Hexlvalent Chromium", A.I.Ch.E Vol.  13,  No.6, November  1967.




 9    Grieves,  R.B., "Foam Separation of  Anlons  from Aequous Solutions,




     Selectivity of Cationlc Surfactants", Separation  Science, 19(1), 1975.




10.  Cohan, H.J., "Electrodlalysls Equipment  and  Membranes",  Chem. Eng. Prog.,




     Vol. 52, No. 2, 1961
                                     -78-

-------
11.  Birkett, J.D.."Electrodlalysis - An Overview", Industrial Hater

     Engineering, September, 1977, pp. 6-10.

12.  Eisenmann, J.L., "Recovery of Nickel From Plating Bath Rinse Waters

     by Electrodlalysis", Plating and Surface Finishing, November, 1977,

     pp. 34-38.
                 4
13.  Twiner, S.B., "Investigation of Treating Elnctroplaters Cyanide Waste

     by Electrodialysis", EPA-R2-73-287, December, 1973.

14.  Langford, K.E., and G.E. Parker, "Method 155, Analysis of Electro-

     plating and Related Solutions", Teddington and Draper Publishing Co.,

     1971.

15.  Easa, H.S., and K. Scfaugerl, "Holdup and Backmixing Investigations in

     Concurrent and  Countercurrent Bubble Columns", Chem. Eng. Sci.,

     Vol. 30, 1975

16.  Qulnn, J.A., R.A.  Stelnbrook, and J.L.  Anderson,  "Breaking Bubbles and

     Water-tonAir Transport of Particulate Matter", Chem. Eng. Sci.,

     Vol. 30, 1975

17.  Faust, C.L., "Modifying a Conventional Chemical Waste Treatment  Plant

     to Handle Fluoride and Ammonia Wastes", Plating and Surface Finishing,

     1975,  pp.  1048-1052.
                                    -79-

-------