Document jm5G9d3V831gB2B3koYk7XY2k

BUSINESS CONFIDENTIAL PROJECT REPORT LATEX AND VINYL RESINS GAS CHROMATOGRAPHIC DETERMINATION OF RESIDUAL VINYL CHLORIDE MONOMER it* AUTHORSi Work by: Supervisor: A. E. Gabany, Jr. (2) H. T. Bias A. E. Gabany, Jr. R. A. Bleidt (3) DAT Ei November 22, 1974 PROJECT Mai 326A25 PILE Nai 20069 SUMMARY A gas chromatographic analysis has been developed for the measurement of loe levels of vinyl chloride monomer in vinyl latexes. A TERGITOL E-35 column capable of separating vinyl chloride from acetaldehyde is defined which offers an improved definition of vinyl chloride in vinyl chloride-vinyl acetate containing polymer latexes and resins. Due to interferences from seemingly inherent trace impurities found in tetrahydrofuran, this method has an apparent 1-ppm minimum limit of detection for residual vinyl chloride monomer in latexes and resins. The gas chromatographic method of analysis is attached as Appendix A. INTRODUCTION Residual vinyl chloride monomer in vinyl resins and films is currently being analyzed by gas chromatography using a method developed by Messrs. H. T. Bias and R. A. Bleidt (1). As is generally the case vith most good analytical methods of analysis, requests are soon entertain d to analyze similar products not designed for the analysis. The case in point was the analysis of residual vinyl chloride monomer in vinyl chloride based polymer latexes containing vinyl acetate monomer and to a lesser extent vinyl chloridevinyl acetate copolymer solution resins. From the outset, th analysis appeared reasonable; however, these products con sistently had higher vinyl chloride results than anticipated. Even after various chemical treatments designed to react or remove the monomer were applied, the results remained high. At this point Mr. T. R. Smith of the Latex Program Group began questioning the results concerning the integrity of the observed vinyl chloride chromatographic peak. From this inquiry it was soon determined that acetaldehyde, a hydrolysis by-product from vinyl acetate, was the culprit. RESEARCH AND DEVELOPMENT DEPARTMENT CHEMICALS AND PLASTICS UNION CARRIDE CORPORATION SOUTH CHARLESTON, WEST VIRGINIA UCC 061305 BUSINESS CONFIDENTIAL -2- 326A25 The gas chromatographic column being used was unable to separate vinyl chloride from acetaldehyde and therefore both were being reported as vinyl chloride. Needless to say a revision of the existing gas chromatographic procedure was initiated. The tedious and dil- * igent"search was begun for a column which would separate vinyl chloride from acetaldehyde and yet be amenable to the samples. Eventually, a TERGITOL E-35 column emerged as the most success* ful candidate. This report summarizes our findings in determining vinyl chloride in vinyl chloride-vinyl acetate containing polym r latexes and resins. The final method of analysis is included. DISCUSSION As previously stated, we are currently analyzing residual vinyl chloride monomer in vinyl resins and films by the gas Chromatographic method of Messrs. H. T. Bias and R. A. Bleidt (1). This method was designed to handle vinyl resins, powders and films by dissolving each in a solvent prior to a gas chromatographic analysis. Tetrahydrofuran was chosen as the solvent not only for its superior dissolution but also because it was supposedly free of trace impurities which might interfere with low vinyl chloride detection. When we were first approached to analyze vinyl latex suspensions for vinyl chloride monomer, it seemed a natural that we try the existing method. It appeared superficially that all of the details had been worked out and supposedly, no new problems should arise. This was later shown to have been wish ful thinking as we were soon confronted with an interference from acetaldehyde. The existing column was not capable of separating acetaldehyde from vinyl chloride and as a result, we were reporting erroneously high vinyl chloride results. The choice of a column which would give the necessary separation of vinyl chloride and acetaldehyde was 20-ft x 1/8-in 20 percent TERGITOL E-35 on 60/80 mesh Chromosorb-W AW. This column permitted a lower column temperature, 70C, to achieve maximum separation and yet allowed the components to elute as reasonably sharp symetrical peaks with a minimum of tailing (See Figure 3 for a typical scan). Another criterion for the final accepted method concerns the limit of detection or sensitivity. The current method has a minimum detectable limit for vinyl chloride of 1.5 ppm. Ideally a new method should meet or surpass that which is already available. However, to Improve on a method at these levels, inherent problems associated with the gas chromatographic method begin to emerge. ucc 061306 BUSINESS CONFIDENTIAL 3- 326A25 First, the method is limited with respect to the amount of sample injected into the chromatograph. The current method uses 2 microliters of a 10 percent solution in tetrahydrofuran. This could probably be increased to 20 percent; however, this would in all likelihood be limited to certain samples. Also the injected sample could be increased to approximately 10 micro1itersj however, saturation of the 'detector could occur. Secondly, these latex samples are approximately 50 percent water. Even though the flame ionization detector is non-responsive to water, there is an unbalance noted on the recorder when water enters the flame. The column must be tai lored to selectively elute water either before or after the desired component depending on the complexity of the analysis. Thirdly, there is always inherent trace impurities associated with the solvent. To date we have not located a supply of tetrahydrofuran which is free of these trace impur ities. Attempts have been mmbe mm `'^clean-up" the solvent with relatively little success. Generally with this approach new impurities are introduced vhlchmre more detrimental than the original. Another method havAmfsaMme limited utility is the introduction of scrubber coluinm^cosnected in series with the analytical column. In this member the impurities are absorbed by the scrubber column or reached in a manner to alter their affinity for the analytical oJfeiwm.^ Fourthly, comes the assignment of the observed peak. The assumption that a peak of a'^ertain retention time on the gas chromatogram is a specifie~ammtmcr*al is risky. In the ppm level the actual confirmation by ewmventional techniques is essentially nonexistent. Mams spectrometer-gas chromatograph interfaced techniques are beiag examined however at the present time a reliable procedure has not been worked out. With the gas chromatographic analysis suffice it to say that the minimum detectable limit is the maximum Lpejjt. present if the peak is correctly assigned. The finaA-'meuda adopted for the low ppm vinyl chloride determination in polymer latexes and resins was designed with full consideration of these points. The sample size was increased 10-fold; however a modification was made in the helium carrier gas route during the analysis. A Bendix 8-port sliding plate valve was installed to provide a column bmekflush mode. Shown in Figure 1 is the flow schematic adopted. In this manner, the carrier flow gas can be reversed in the column at any time to backflush those components still in the column of which there is no Interest of detecting. Consequently the large volume of solvent Injected into the column never reaehes the detector. The backflush valve is switched just after vinyl chloride is detected ucc 061307 BUSINESS CONFIDENTIAL -4- 326A25 and is allowed to remain in that position for 10 minutes to purge the column. Returning the valve to its original position pre pares the instrument for the injection of another sample. Acet aldehyde, which elutes approximately one minute after vinyl chloride, can be determined by delaying the valve switching until after acetaldehyde is detected on the recorder. Various attempts were made to remove certain inherent trace impurities in the tetrahydrofuran solvent but to no avail. Finally it was decided that the method could not be tailored to detect less than a 1-ppm vinyl chloride. If a ''clean" solvent is obtained the method should provide data to "tenthtf'of a ppm vinyl chloride. Shown in Figure 4 are scans of the solvent impurities along with a latex sample which calculated to contain 1.8 ppm vinyl chloride. Notice that certain other unidentified peaks are also present in the sample which intensify the aolveat impurities. Shown -in Table I are some typical results obtained on samples of UCAR 1000, UCAR 508, MILT 29. and MILT 31. CONCLUSIONS A TERGIT0L E-35 column capable of separating vinyl chloride from acetaldehyde has been used in a gas chromatographic method to improve the definition of vinyl chloride in vinyl chloride-vinyl acetate containing polymer latexes and resins. The method also uses a Bendix 8-port slid ing plate valve to allow a backflush mode in the system. ACKNOWLEDGMENT Messrs. J. L. Marchio and A. L. Monday contrib uted to this project with their advice and installation of the Bendix 8-port sliding plate valve. BIBLIOGRAPHY 1) Bias, H. T. and Bleidt, R. A., Vinyl Resins. Gas Chromatographic Method for the Determination of Low Levels of Vinyl Chloride Monomer, Project Report, R and D, South Charleston, W. Va., File No. 19072, January 2, 1974. NOTEBOOK REFERENCE 15AEG54 Manuscript dated- 11/15 74 Date typed: 11 22 74 vl Attachments: 1 Appendix A 1 Table 4 Figures ucc 061308 BUSINESS CONFIDENTIAL 326A25 FIGURE 4 GAS CHROMATOGRAM OF TRACE IMPURITIES IN TETRAHYDRQFURAN AS RELATED TO LOW PPM VINYL CHLORIDE IN LATEXES V in y l ch lo rid e Retention time, minutes if 3 UCC 061309 business confidential 326A25 APPENDIX A nt v; DETERMINATION OP RESIDUAL VINYL CHLORIDE MONOMER IN VINYL LATEXES for igi no h & Vi ecorde; 1 PURPOSE A procedure for determining small amounts of residual vinyl chloride monomer in vinyl*latex is ov* j xv i described. The latex suspension is dissolved in a-,suitabl { solvent and is analyzed using a gas chromatographic technique to utilizing a backflush mode. The method is applicable for us^.*agCat.s with vinyl chloride-vinyl acetate copolymer latexes and wi ^.ch ca) resins, vinyl chloride-vinyl acetate-ethylene terpolymer *tain c latexes and other vinyl latexes and resins in whicfcnther ar :nt no volatiles which interfere in the determination. The method resolves acetaldehyde from vinyl chloride and can he quantitated with proper calibrations. 5.. ^ rc> -2 EQUIPMENT AND REAGENTS J '1 nd N a) Gas Chromatograph, Hewlett-Packard (F and M) Model 5750 ^ . or equivalent, equipped with a hydrogen flame lonizatiOif detector. f. 1 b) Backflush valve, Bendix 8-port sliging plate valVe,or ' equivalent. ,'r a c) Tetrahydrofuran, reagent grade. " d) Balance with accuracy to 0.10 gram. ' 1 3 INSTRUMENT PARAMETERS Instrument Column Backflush valve Tempera tures Column Injection port Detector Carrier gas Flows Helium carrier gas a) inject b) backflush Hydrogen Compressed air Sample size Hewlett-Packard (F and M) Mod 1 5750 or equivalent, equipped with hydrogen flame ionization detector 20 feet x 1/8-inch stainless steei_ , -~s tubing packed with 20 percent l .^is i TERGITOL E-35 on Chromosorb W AW,urn na 60/80 mesh Bendix, 8 port sliding plate valve or equivalent (Figure 1) __ 70C isothermal 100C (maximum) 300C helium 0 . Ga 40 cc/minute 40 cc/minute 20 psi at cylinder head 40 psi at cylinder head 2 microliters (20 microliters if the expected monomer is less than 10 ppm) BUSINESS CONFIDENTIAL -A2- 326A25 Approximate elution times Vinyl chloride 2.8 minutes Acetaldehyde 3.7 minutes 4 SCHEMATIC OF FLOW SYSTEM A schematic of the floe system utilizing the Bendix 8-port slid- ing valve is attached as Figure 1. 5 PROCEDURE Accurately weigh one gram of the latex to be analyzed into a 6-dram vial containing 9 grams of tetrahydrofuran (THF). Cap the vial and shake to effect solution. Inject 2 microlltere of the test solution (20 micro liters if the expected monomer content is less than 10 ppm> into the gas chromatograph using the parameters listed in Section 3. A backflush valve has been incorporated into the system to facilitate the analysis. After the vinyl chloride monomer and or acetaldehyde have eluted from the chromato graphic column, the column is backflushed and vented to the atmosphere thus preventing the solvents from reaching the ' -detector. In'this manner the analysis time per sample is drastically shortened. Component vinyl chloride acetaldehyde Backflush time 10 minutes 10 minutes After the prescribed backflush period, return the valve to its original position. The baseline should be allowed to equilibrate before injecting another sample. Measure the area of the vinyl chloride peak using an acceptable technique (e.g. planimeter, automatic integrator, computer, etc,'*. 6 CALIBRATION Prepare solution mixtures in TOF of varying amounts of vinyl chloride to cover the expected ppm range of concentration. Obtain the gas chromatogram for each sample using the parameters listed in Section 3. NOTE: Use 2-microliter sample size. Prepare a chart plotting the area obtained from the chromatogram of each sample versus the known concentration to establish a relative detector response. See Figure 2 for typical calibration curve. 7 CALCULATIONS Determine the monomer content in the test solution from the calibration chart. Apply the area of the monomer obtained from the chromatographic scan of the test solution to the calibration curve and read the the corresponding concentration in parts per million (ppm1* . [[if the sample size of the test solution injected into the gas chromatograph was 20 microliters, multiply the ppm con centration by a 0.1 dilution factor]) BUSINESS CONFIDENTIAL -A3- 326A25 The concentration value btain d from th calibration curv is total monomer in the solution. In order to convert the data to monomer concentration based on total latex the value must b multiplied by a factor of 10. 8 TYPICAL CHROMATOGRAM A typical chromatogram is attach d as Figure 3. ucc 061312 BUSINESS CONFIDENTIAL 326A25 TABLE I TYPICAL ANALYSES OF VINYL CHLORIDE- VINYL ACETATE LATEX SUSPENSIONS Sample UCAR 1000<a) Reading, Pa. San Francisco, Calif. Carrollton, Texas Atlanta, Ga. Huron, Ohio Chicago, 111. UCAR 508 Somerset Storage UCAR 508 B-1857 (21HOC53) UCAR 508 B-2599 (21HOC105-C) UCAR 508 B-2441 (21H0C94-A) UCAR 508 B-2597 (21HOC105-A) UCAR 508 B-1859 (21H0C55) UCAR 508 (21H0C127-A) UCAR 508 B-2290 (21HOC90-A) VACA B-1107 HILT 29 B-4489 MILT 29 B-4630 HILT 31 B-4564 (19H0C136B) HILT 31 B-5072 HILT 31 B-4901 ppm Vinyl Chlorid 50 113 88 63 88 50 90 75 37 2.8 50 9.6 1.8 7 6 38 100 250 188 38 (a) Samples shipped to Glidden Paint ucc L 061313 BUSINESS CONFIDENTIAL FIGURE I SCHEMATIC OF FLOW SYSTEM UTILIZING THE BENDIX 8-PORT SLIDING PLATE VALVE 326A25 "BACKFLUSH" BUSINESS CONFIDENTIAL 326A25 FIGURE 2 TYPICAL GAS CHROMATOGRAPHIC CALIBRATION CURVE F RE :t o.< 0 >c 1 pi it tY it 31 tv- * s r u e t r & S- t i b a ucc BUSINESS CONFIDENTIAL 326A25 FIGURE 3 typical gas chromatogram of vinyl latexes FOR RESIDUAL VINYL CHLORIDE MONOMER 1 ^ h ; Tl !C De te c to r Response Retention time, minutes is n "1 it IF .01 m - -ag ail" ' 'it . I*1 f :o id -t r* r., es ent jsl the rsjult *P ac at . w u * *e-: 4T ucc 061316