Document XRLevBgE22KLD1w160rM9p1R4

ETHYL CORPORATION RESEARCH AND DEVELOPMENT DEPARTMENT BATON ROUGE, LOUISIANA 70 82 1 A DETERMINATION OF THE NON-MIGRATION LEVEL FOR VCM FROM PVC BOTTLES TO FOOD-SIMULATING SOLVENTS Authors: J. L. Cannaday G. A. Daniels G. C. Gaeke Date: June 7, 1977 VC4046 A DETERMINATION OF THE NON-MIGRATION LEVEL FOR VCM FROM PVC BOTTLES TO FOOD-SIMULATING SOLVENTS SUMMARY This report summarizes the results of our five-week and 5 8-day storage tests with water and 50$ aqueous ethanol at 120F in PVC bottles with seven different initial levels of vinyl chloride monomer in the bottles. The proposal for this experimental program was sent to the FDA Hearing Clerk by letter of January 19, 19 77. CONCLUSIONS With analysis by headspace and FID, both water and 50$ aqueous ethanol contents show no statistical difference in extractants in the VCM region of the three (different) low level bottles which contained less than 0.09 ppm VCM. The 0.09 ppm VCM contents with both water and 50$ aqueous ethanol show a significant ir crease in the VCM region of the chromatograms. These correspond to about 0.0006 ppm VCM with water and 0.0008 ppm VCM with 50$ aqueous ethanol after 58 days at 120F. With analysis by GCMS, the sensitivity for VCM in water is 0.0016 ppm and in 50$ aqueous ethanol is 0.0 05 ppm. VCM is not confirmed by. GCMS in the water and 50$ aqueous ethanol contents of the 0.23 ppm VCM bottles. A non-migration level of 0.04 ppm VCM is established which corresponds to the 95$ confidence level that the areas of the chromatogram blips in the VCM region for the contents in a PVC bottle shall not be greater than with no VCM. BOTTLE PREPARATION This study was conducted on bottles prepared from compounds containing the following VCM levels: Compound 1 2 3 4 5 6 7 <0.01 ppm 0.01 ppm 0.02 ppm 0.09 ppm 0.23 ppm 0.8 ppm 1.9 ppm VC4047 2. Four batches of resin plus stabilizer were prepared with target VCM values of <0.01, 0.4, 1.0, and 2.0 ppm. The <0.01 and 0.4 levels were pre pared by using special laboratory techniques. The 1.0 and 2.0 resins were obtained by addition of VCM. The stabilizer was added to the resins to allow removal of VCM without degrading the resin. Table I shows the analyses for the resin plus stabilizer combinations. The analytical method is attached as Appendix I. Analyses in Table I are reported as VCM equivalent to the total area in the 138 to 146 second region of the chromatogram. VCM without inter ferences shows a sharp peak at 141 seconds. Interferences in the VCM region which are not VCM convert this sharp peak into blips which are impossible to distinguish at the 0.01 ppm VCM level and difficult to distinguish at the 0.4 ppm level. All areas and VCM equivalents in this report were obtained by hand integration of the chromatogram peaks. We found that machine integration was meaningless for some of the analyses because of broadened peaks, doublets, and side peaks. Low level resin plus stabilizer and powder blend and all compound pellets and bottles analyses were hand integrated. We believe that the very low level resin plus stabilizer contains less than 0.005 ppm as VCM. Analytical results are shown for two different batches of low level resin plus stabilizer as a second batch was prepared to provide sufficient material for intermediate blends. Additional additives for our 8237 bottle compound were added to the resin plus stabilizer combinations and these were given similar processing in a labora tory intensive mixer to prepare powder blends and pellets with the same heat history. The target 1.0 and 2.0 ppm VCM levels show about one-half the equivalent VCM in the powder blend as in the resin plus stabilizer. The low level VCM shows an increase which is obviously due to non-VCM components because of the non-typical peaks that are observed. The powder blends were then pelletized. The pellets were analyzed by solution in DMAC. This analytical procedure is attached as Appendix II. It is apparent that sensitivity for VCM at low levels is completely lost with the pelletized compound as the apparent VCM for the low level compound is obviously essentially all non-VCM components by comparison with the resin plus stabilizer from which it was prepared. The target 1.0 and 2.0 ppm VCM compounds show reasonable agreement with the powder blends. Bottles prepared from these four compounds were also analyzed by the solution method. Additional increases in non-VCM is indicated by these chromatograms except for the target 2.0 ppm VCM bottles. The intermediate VCM level bottles were prepared by blending the target 0.4 ppm VCM compound (assumed as 0.4 ppm VCM) with the low level compound for 0.02 and 0.04 ppm VCM blends and the target 1.0 ppm VCM compound (assumed as 0.84 ppm VCM) for the 0.1 ppm VCM hlend. Two bottles of each of the <0.01, 0.02, 0.04, 0.1, 0.4 and 0.8 ppm VCM compound were sent to Dr. Breder. VC4048 3. (In routine production, VCM levels for bottle compound are measured and controlled on powder blend. At very low levels, it is obvious these measured values will be in error on the high side because of the noted interferences.) Analysis of the stabilizer shows peaks at 138 and 146 seconds. The chromatograms for resin plus stabilizer for the <0.01 ppm VCM level show double peaks corresponding to the interfering components in the stabilizer. Chromatograms and tapes for this are included in Appendix III. Equivalent VCM shown in Table I for resin plus stabilizer is calculated from the double peak area. The major part of this area is in the 138 and 146 second region. The estimate of less than 0.005 ppm VCM is based upon the assumption that about 80# of the area is represented by the non-VCM components. VCM losses appear consistent in converting resin plus stabilizer to compound at the nominal 1.0 and 2.0 ppm VCM levels. If this proportionate loss is applied to the nominal 0.4 ppm VCM level, estimated compound VCM content is 0.23 ppm. Heat history was the same for these three compounds, so this appears to be a reasonable assumption. Blending of this compound with the <0.01 ppm VCM compound would then result in 0.01 and 0.02 ppm VCM rather than the target values of 0.0 2 and 0.04 ppm VCM. MIGRATION DATA Tables II and in show the analyses for the water and the 50# aqueous ethanol contents from the PVC bottles after five weeks storage at 120.F. Tables IV and V show similar data after 58 days storage at 120F. Triplicate analyses from three bottles at each VCM level are shown. Analyses are also reported for a fourth bottle for six VCM levels with water after five weeks storage because of inconsistencies between the analyses from the three bottles. The following table shows the differences between the average areas of the chromatogram blips for the four high-VCM level bottles and the baseline areas (three low-VCM level bottles) in terms of VCM as a fraction of the VCM that is predicted for migration to the bottle contents with allowance for VCM loss in bottle blowing. Areas are converted to VCM by division by 5.2 for water and 1.8 for 50# aqueous ethanol. Initial VCM Level Fraction of Predicted Migration Water 50# Aqueous Ethanol Five Weeks 58 Davs Five Weeks 58 Davs 1.9 ppm 0.8 0.23 0.09 0.84 0.56 0.84 1.30** 0.83 0.49 0.63 0.69 0.83 0.10 * 0.93 0.86 0.39 0.27 0.38 * The average total area for this bottle is less than that from Compound 1 bottles, but is within experimental error for these data. ** The value greater than unity is probably a result of experimental error with this lowlevel VCM in the contents. VC4049 4. The water results are observed to be reasonably consistent. The correction for the 0.23 ppm VCM bottle appears to provide about the same fraction migration as for the two higher level bottles. The 50# aqueous ethanol results are not as consistent as for the water. Chromatograms for typical bottle content analyses are provided in Appendix HI. STATISTICAL ANALYSIS OF DATA Tables VI to IX summarize the statistics for the water and 50# aqueous ethanol contents data for both the five-week and the 58-day experiments. Outlier bottle contents were rejected as shown. In all cases the extractable level in Compounds 2 and 3 are not significantly greater than the extractable level for Compound 1. The test for statistical significance is made by asking the following question: Is the extractable level for Compound X greater than the extractable level fbr Compound 1? Associated with this question is the Null hypothesis: The extractable from Compound X is not larger than the extractable from Compound 1 A value of 0.0 5 or smaller for the probability of the Null hypothesis is taken as evidence that the extractable level from Compound X is greater than the extractable level from Compound 1. The extractable level for Compound 4 is clearly greater for water and marginally so for the 50# aqueous ethanol for both the five-week data and the 5 8-day data. In all cases, except for 50# aqueous ethanol at five weeks, the extractable levels for Compounds 5 to 7 are greater than for Compound 1. Compounds 5 and 6 for 50# aqueous ethanol at five weeks appear to be anomalous as Compound 5 is less than Compound 1, while Compound '6 is only marginally greater. An analysis of variance of the observed values of area units indicated that both the variance between analyses within bottles and the variance between bottles are approximately proportional to the square of the level of the observation in the water data at five weeks and the 50# aqueous ethanol data at past five weeks and 58 days. This suggests a logarithmic transformation of the data. The log-transformed data for both sets of 50# aqueous ethanol data and the water data at five weeks were subjected to analysis of variance to determine the variance between analyses and the variance between the bottles on a log-scale. A Bartlett test of the variance shows that these variances on a log-scale are drawn from a homogeneous population, thereby allowing pooling of the individual varianc s For the five-week data, the variances between bottles have a total of 14 degrees of freedom for each series, while the variance within bottles have 42 d grees of VC4050 5. freedom for each series. As the variance between bottles is the major contributor to the total variance, the tests of the various means is made with the pooled standard deviations using 14 degrees of freedom. The 50$ aqueous ethanol data at 58 days has but 12 degrees of freedom for the variance between bottles. The analysis of variance for the water data at 58 days shows two differences from the other three sets of data. 1. The variance is constituted entirely by differences in analysis rather than differences between bottles. 2. Hie variances are not drawn from a homogeneous population for the various compounds, whether the data are used directly or a log-transformation is used. A two-sample t-test for the difference of the mean values, assuming that each analysis is independent, has been used to compare Compounds 2, 3, 4, and 5 with Compound 1. VCM NON-MIGRATION LEVEL A review of the statistical data presented in Tables VI-DC shows that VCM was not detected as a significant difference in the level of extractable from Compound 1 in the cases of Compounds 2 and 3. Compound 4 shows differences which are marginally significant for the 50$ aqueous ethanol and more significant for water. It is our best estimate that Compound 2 contains about 0.01 ppm VCM and that Compound 2 contains aboui 0.02 ppm VCM, while Compound 3 con tains about 0.09 ppm VCM. The data, then places the detectable limit of VCM as occuring with a VCM content of the compound ranging between 0.02 and 0.09 ppm V CM . A more refined estimate of the limit of detection can be made by con sidering the t values which were obtained in the statistical studies. The critical t value for a 9 5$ probability that a compound has a higher level of extractable than Compound 1 ranges from 1.76 to 1.78 for the experiments presented in this report. These t values are shown in Figure 1 for the comparisons from Compounds 2-5 with Compound 1. The intersection of a line through these t values with the critical t value defines the concentration of VCM in PVC compound where VCM should be detected in the extract. The t values describe reasonably smooth curves for 50$ aqueous ethanol at 58 days and for water at five weeks. As would be expected, scatter occurs at the lowest levels. The 50$ aqueous ethanol data at five weeks fall below the data for 50$ aqueous ethanol at 58 days, except for the point at 0.09 ppm VCM. The data for water at 58 days are VC4051 6. scattered and fall well above the values for water at five weeks. The major reason for these larger t values is the greatly reduced variance between bottles for this set of data. The analyses of variance suggests 0 for the standard deviation between bottles for the 5 8-day data as compared to 0.1268 (on a In scale) for the five-week data. The variance between analyses is about the same for both sets of water data. Whether this improved replication is due to the additional time of extraction or other causes is not known. A limit of detection for the 58-day water data can be considered a low estimate. The values obtained from Figure 1 are given below; VCM Concentration Experiment________________________________ in PVC for Solvent____________ Time Detection Water 50# Aqueous Ethanol Water 50# Aqueous Ethanol Five Weeks ' Five Weeks 58 Days 58 Days 0.07 >0.09 0.04 0.09 It is interesting to note that if the variances established for the fiveweek data for water were used for the 58-day data, the t value curves would be essentially identical. This would yield a value of 0.07 for water at 58 days. The data given in Figure 1 suggest that VCM is more readily detected in water than in 50# aqueous ethanol in this type of experiment. Water storage at 120F for 58 days corresponds to about 1.2 5 years at 72F. The VCM non migration level of 0.04 in PVC bottles thus represents a conservative level for all applications. GCMS ANALYSES GCMS analyses were obtained at Spectrix Corporation, Houston, Texas. The apparatus was a Supelco 1.5 meter by 2 mm ID glass column with Carbopack C ( 60-80 mesh) coated with 0.2# Carbowax 1500 and a Finnegan Model 3200 gas chromatograph mass spectrometer. Standards analyses showed a sensi tivity of 0.0016 ppm in water and 0.005 ppm in 50# aqueous ethanol. Analyses of the contents from the 0.23 and 0.09 ppm VCM bottles did not show VCM. A report of this work is attached as Appendix IV. VC405? TABLE I PVC ANALYSES; PPM VCM* (Includes total 138-146 second material) Target VCM Level <0.01 <0.01 0.4 1.0 2.0 Resin Plus Stabilizer 0.014** 0.026** 0.44 1.61 3.85 Powder Blend 0.052 0.067 0.39 0.85 2.22 Compound Pellets 0.50 0.41 0.83, 0.70 1.89 o cr>o o CO 03 Bottles 0.64 0.90 0.94 1.86 * Values shown are average of triplicate analyses. ** Detailed study of the chromatogram indicates the VCM portion is <0.005. VC4053 TABLE II WATER CONTENT ANALYSES Five-Week Storage All values are area measurements as products of peak heights and width at half heights. Areas of secondary peaks are added where these occur. Actual Values Bottle 1 2 3 4 Compound 1 15.7, 15.3, 20.6, 20.3, 19.3, 20.6, 22.3, 20.5, 15.6 20.8 20.5 21.0 (<0.01 ppm VCM) rejected Bottle 1 2 3 4 Compound 2 23.1, 22.3, 22.0 19.2, 18.6, 19.6 21.0, 20.7, 20.7 26.3, 25.7, 26.1 (0.01 ppm VCM) rejected Bottle 1 2 3 4 Compound 3 18.9, 21.2, 19.8, 21.5, 22.1, 22.8, 26.4, 27.4, 21.9 22.2 23.0 25.8 (0.02 ppm VCM) rejected Bottle 1 2 3 4 Compound 4 19.7, 22.6, 27.7, 25.7, 24.0, 23.1, 27.8, 28.5, 26.3 25.5 22.7 28.7 (0.09 ppm VCM) rejected Bottle 1 2 3 4 Compound 5 23.7, 24.8, 23.8, 29.1, 35.1, 35.4, 33.3, 31.6, 25.3 25.7 32.8 31.8 (0.23 ppm VCM) Bottle 1 2 3 4 Compound 6 29.0, 27.6, 15.2, 15.0, 42.6, 38.0, 42.2, 40.8, 35.2 17.5 41.1 40.8 (0.8 ppm VCM) rejected rejected Bottle 1 2 3 Compound 7 106.4, 102. 6, 83.7 90.0, 90.9, 90.8 88.0, 87.7, 95.3 (1.9 ppm VCM) VC4054 TABLE III 504 AQUEOUS ETHANOL CONTENT ANALYSES Five-Week Storage Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Area values as defined in Table II. Compound 1 21.5, 21.3, 21.2, 22.2, 19.0, 18.0, 22.0 21.3 18.0 Actual Values (<0.01 ppm VCM) Compound 2 22.1, 19.9, 19.5, 19.9, 20.4, 20.9, 20.6 18.5 21.1 (0.01 ppm VCM) roo o N>o 00 Compound 3 19.1, 19.0, 19.8, 20.9, 19.1 20.2 19.2 (0.02 ppm VCM) Compound 4 21.6, 22.4, 23.0, 24.5, 24.0, 24.2, 20.9 24.2 23.2 (0.09 ppm VCM) Compound 5 20.1, 20.9, 20.3, 19.1, 20.3, 20.4, 19.9 20.6 21.0 (0.23 ppm VCM) Compound 6 23.2, 24.2, 21.6, 21.5, 24.3, 22.5, 23.8 21.6 25.0 (0.8 ppm VCM) Compound 7 80.4, 86.2, 56. 5, 56.3, 68.3, 68.8, 81.1 61.2 65.8 (1.9 ppm VCM) VC4055 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 TABLE IV WATER CONTENT ANALYSES 58-Day Storage Area values as defined in Table II. Compound 1 16.8, 19.1, 17.8, 18.1, 18.4, 18.0, 18.3 17.9 18.0 Actual Values (<0.01 ppm VCM) . Compound 2 16.4, 18.1, 16.8 17.9, 17.9, 17.6 19.5, 14.5, 17.9 (0.01 ppm VCM) Compound 3 18.5, 17.0, 17.0, 18.8, 18.3, 17.9, 19.0 17.8 18.2 (0.02 ppm VCM) Compound 4 21.3, 21.7, 24.0, 21.6, 20.8, 21.3, 21.3 23.0 21.8 (0.09 ppm VCM) Compound 5 27.1, 30.0, 26.2, 31.8, 30.8, 23.7, 24.3 26.9 22.6 (0.23 ppm VCM) Compound 6 44.1, 42.6, 26.8, 26.6, 8.6, 11.1, 40.9 26.0 9.5 Compound 7 116.1, 117. 4, 86.3, 87. 8, 117.9, 113. 6, 115.4 89.3 112.3 (0.8 ppm VCM) rejected rejected (1.9 ppm VCM) rejected VC4056 TABLE V 50j AQUEOUS ETHANOL CONTENT ANALYSES 58-Day Storage Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Bottle 1 2 3 Area values as defined in Table II. Compound 1 18.8, 19.4, 19.7, 20.0, 18.0, 17.4, 19.3 19.1 17.4 Actual Values (<0.01 ppm VCM) Compound 2 19.3, 19.5, 19.8, 19.8, 19.6, 19.5, 19.4 18.2 20.1 (0.01 ppm VCM) Compound 3 18.4, 18.4, 9.2, 9.5, 18.5, 19.3, 17.5 8.5 17.7 (0.02 ppm VCM) rejected Compound 4 20.3, 20.7, 18.3, 19.2, 22.2, 20.6, 20.7 17.9 21.6 (0.09 ppm VCM) Compound 5 21.3, 22.0, 22.2, 22.8, 21.3, 20.3, 21.1 21.0 19.3 (0.23 ppm VCM) Compound 6 24.3, 24.2, 24.8, 26.1, 26.8, 26.3, 25.3 26.2 27.3 (0.8 ppm VCM) Compound 7 84.9, 78.6, 65.2, 65.8, 88.2, 85.3, 80.9 62.7 85.1 (1.9 ppm VCM) rejected VC4057 Compound 1 2 3 4 5 6 7 Compound 2 3 4 5 6 7 TABLE VI STATISTICAL DATA - WATER Five-Week Storage Bottles Included Area Units Average of Ln Avera ere (Area Units) 2, 3, 4 1, 2, 3 1, 2, 3 2, 3, 4 1, 2, 3, 4 3, 4 . 1, 2, 3 20.656 20.800 21.489 25.967 29.3 67 40.917 92.822 3.0273 3.0326 3.0657 3.2532 3.3690 3.7108 4.52 80 Comparisons vs. Compound 1 Degrees of Freedom Difference Ln (Area Units) t*** 14 0.0053 0.050 14 0.0384 0.361 14 0.2259 2.126 14 0.3417 3.438 14 0.6835 5.753 14 1.3207 12.429 Probability of Null Hypothesis^ 0.4912 0.3715 0.0266 0.0020 <0.0001 <0.0001 * t values of greater than about 1.76 indicate that the extraction from the compound is statistically significantly greater than from Compound 1. ** Values less than 0.05 indicate statistically significantly greater extraction than from Compound 1. t values are calculated from the following estimates for the standard deviation of Ln (area units) obtained by pooling all the analyses: Between analyses Between bottles 0.0510 0.1268 VC4058 TABLE VII STATISTICAL DATA - 50^ AQUEOUS ETHANOL Five-Week Storage Compound 1 2 3 4 5 6 7 Area Units Average of Ln Average (Area Units) 20.500 20.322 19.778 23.111 20.289 23.067 69.389 3.0173 3.0106 2.9845 3.13 89 2.9991 3.13 74 4.22 89 Compound 2 3 4 5 6 7 Comparisons vs. Compound 1 Degrees of Freedom 14 14 14 14 14 14 Difference Ln (Area Units) -0.0067 -0.0328 0.1216 -0.0182 0.1201 1.2116 t*** -0.097 -0.474 1.756 -0.263 1.735 17.499 Probability of Null Hypothesis 0.5258 0.6699 0.0518 0.5911 0.0538 <0.0001 * t values of greater than about 1.76 indicate that the extraction from the compound is statistically significantly greater than from Compound 1. ** Values less than 0.05 indicate statistically significantly greater xtraction than from Compound 1. t values are calculated from the following estimates for the standard deviation of Ln (area units) obtained by pooling all the analyses: Between analyses Between bottles 0.0324 0.0827 VC4059 TABLE VP! STATISTICAL DATA - WATER 58-Day Storage Compound 1 2 3 4 5 6 7 Bottles Included 1, 2, 3 1, 2, 3 1, 2, 3 1, 2, 3 1, 2, 3 1 1, 3 Area Units Average 18.044 17.400 18.056 21.867 27.044 42.533 115.450 Comparisons vs. Compound 1 Compound 2 3 4 5 Approximate Degrees of Freedom 10 10 13 8 Difference (Area Units) -0 644 0 012 3 823 9 000 t*** -1.275 0.036 9.787 8.159 Probability of Null Hypothesis** 0.8842 0.4860 <0.0001 <0.0001 * t values greater than about 1.8 indicate that the extraction from the compound is significantly greater (in the statistical sense) than from Compound 1. ** Values less than 0.05 indicate statistically significantly greater extraction than from Compound 1. Analysis of variance for these data indicates that the variance betwe n analyses within bottles contributes essentially all the variance. These variances differ between compounds so that pooling is not indicated. A two-sample t-t st has been used to compare Compounds 2,3,4 and 5 with Compound 1. Each analysi has been assumed to be independent even though replication within bottles was used. This procedure is justified by the observation that the varlanc is con tributed by differences between analyses rather than differences between bottl s. Statistical data have not been made to compare Compounds 6 and 7 with Compound 1 as the peak areas ar much greater. VC4060 Compound 1 2 3 4 5 6 7 Compound 2 3 4 5 6 7 TABLE DC STATISTICAL DATA - 504 AQUEOUS ETHANOL 58-Day Storage Bottles Included 1, 2, 3 1. 2, 3 1, 3 1, 2, 3 1, 2, 3 1, 2, 3 .1,3 Area Units Average of Ln Average (Area Units) 18.789 19.467 18.300 20.167 21.2 56 25.700 83.833 2.9321 2.9678 2.9064 3.0017 3.0555 3.2457 4.4281 Comparisons vs. Compound 1 Degrees of Freedom Difference Ln (Area Units) t*** 12 0.0357 0.875 12 -0.0257 -0.284 12 0.0696 1.706 12 0.1234 3.024 12 0.3136 7.685 12 1.4960 33.199 Probability of Null Hypothesis' 0.2044 0.5996 0.0583 0.0039 <0.0001 <0.0001 * t values of greater than about 1.78 indicate that the extraction from the compound is statistically significantly greater than from Compound 1. ** Values less than 0.05 indicate statistically significantly greater extraction than from Compound 1. t values are calculated from the following estimates for the standard deviation of Ln (area units) obtained by pooling all the analyses: Between analyses Between bottles 0.02 87 0.0465 VC4061 FIGURE I t Values for Extraction From PVC Compounds VCM CONCENTRATION IN PVC COMPOUND, PPM VC406? Appendix I Determination- of Residual Vinyl Chloride Menomer In Polvvinvi Chloride Resins and Compounds (Headspace Analyzer Method) 1. Scope This method describes a procedure for the determination of ppm (w/w) levels of residual vinyl chloride monomer (VC1) in polyvinyl chloride (PVC) resins and compounds by a special automated technique (headspace analysis). Powders (resins and compound powder blends) and also pell ts or finished products (compounds) may be analyzed using one or the other of the two procedures described below. The limit of detection is approximately 0.005 ppm VC1 for powders and 0.03 ppm for pellets. 2. Outline of the Method a) 'Powders: Resins and compound powder blends are analyzed as solids. A weighed (4.0 g) sample is placed in a 25 ml vial and sealed with a serum stopper. The sample is equilibrated for one hour at (above the glass transition temperature of FVC). A sample of the vapor above the sample (headspace) is then injected into a gas chromatograph. By use of an automated turntable and sequencer up to 30 samples may be prepared and then analyzed without additional operator attention. Standard gas samples are analyzed along with the powder samples to calibrate the instrument. Peak areas are determined with an automatic integrator. From the calibration standards the concentration of VC 1 in the headspace is determined and the residual VC 1 monomer content of the resin is calculated. b) Pel lets: Pelletized compounds and finished products (cut into small pieces) must be dissolved to determine residual VCl monomer content. A weighed (0.5*0.7 g) sample is placed in a 25 ml vial containing 10.0 ml of DMAC (N,N-dimethylacetamide) as solvent. The vial is then sealed with a serum stopper and the samples are warmed C'' 50C) and shaken occasion ally until the PVC is in solution (4-6 hrs). The vials are then placed in the water bath equilibrated at 90 for one hr. A sample of the vapor above the solution (headspace) is then injected into the gas chromatograph. Standard samples (VCl dissolved in DMAC) are analyzed concurrently with the samples. Peak areas are determined with an automatic integrator. Calculation of the residual VCl monomer content of the sample is straightforward, based on the headspace VCl concentration of the standards, the sample, and the sample weight. 3. Apparatus a) Headspace Analyzer: Perkin Elmer Model F-42, equipped with a flame ionization detector, thermostated sample turntable, backflush assembly, and automatic timing sequencer. b) Sample Vials: 25 ml complete with butyl rubber septum caps and aluminum caps. VC4063 2. c) Column: 3' x 1/8" SS plus 6' x 1/8" SS both packed with 0.2% Carbowax 1500 on Carbopak C. The two columns are connected in s ries (31 column closest to injection port) with the backflush "T" between them. d) Recorder: 1 mv, l5"/hr. e) Integrator: Spectra Physics System I (Perkin Elmer Model 1). f) Syrinoes. Gas Tight: 10, 50, 100, 500 ni and 1, 2, 5 nil. g) Ordinary Laboratory Glassware. 4. Reagents and Materials a) Vinyl Chloride: 99.9+% pure. b) Vinyl Chloride in Nitrogen Standards: 10 ppm (v/v), 50 ppm (v/v), 100 ppm (v/v), 1000 ypm (v/v). Available from Supelco, Inc. c) N, N-dimethylacetamide (DMAC): reagent grade. d) Nitrogen: Zero grade carrier gas. e) Hydrogen: high purity for FID. f) Air: high purity for FID. 5. Instrument Operating Conditions a) Water Bath: 90.0C+0.1* b) Gas Flow Controllers N2 (reference col): 10 ml/min (-- 1 Kg (cm2) N2 (analytical col): 10 ml/min 1.2 Kg/cm2) N2 (backflush)11 10 ml/min (-- 0.8 Kg/cm2) H2: 25-30 ml/min (2.0 Kg/cm2) Air: 250-300 ml/min (3.2 Kg/cm2) c) Gas Chromatograph Conditions: Oven: 55C Injector: I00C Detector: 150C Dosing Needle: 100C d) Sequencer (Programmer) Controls Backflush: On P: 30 sec (internal setting) 1: 3 sec A: 0.8 min BF: 8' (I0X internal setting) S: 1' VC4064 3. e) Integrator Controls fV: 5 (peak width) SS: 15 (slope sensitivity = 1.2 p,v/sec) FP: 5 (fused peaks) BL: 2 (baseline) T1: 60 (initial delay) T4: 250 (stop time) MA: 0 (minimum area) Program: 3 (external standard) 6. Calibration a) Powders (resins and compound powder blends only). Calibrate the analytical system using gaseous vinyl chloride in nitrogen standards as follows: Take two or three empty vials, close each with a septum stopper and seal with an aluminum cap. Insert a 3" needle (22 gauge) and a short (1/2", 22 gauge) needle through the septum of the vial. Connect (metal tubing) the long (3") needle to a cylinder containing the vinyl chloride standard (10-100 ppm VC1 in N2) and slowly flush (15-25 ml/min) the vial with the standard for 4-6 min. After the vial has been filled with the VC 1 standard remove first the 3" needle and then the 1/2" needle from the vial. Repeat to fill additional vials. Label the vials containing the standards and use within 24 hours. To calibrate the instrument, take two or three vials each of four or more standards and insert in the F-42 turntable (at operating temperature). Allow the standards to come to thermal equilibrium (one hr) and then analyze using the instrument conditions shown i r. section 5* Calculate the calibration factor as follows: F(P) - C x 0.0194 A where: F(p) = calibration factor for powder samples C = Concentration of VC1 in the gas standard in ppm (v/v). A Area of the VC1 peak 0.0194 = Factor which relates vapor VC 1 concentration (ppm) to original VC1 concentration in powder. Based on 4 g sample, 25 ml vial, 90.0C and Henry's law relation. (See section 8.) Average the calibration factor for the number of standards analyzed (results should agree within + 5% relative). Recalibration should be performed on a regular basis (weekly or biweekly). If the daily standard (section 7) varies more than + 10% from the correct value then the analytical system should be recalibrated. VC4065 4 b) Solutions: (pellets and finished products - may also be used for powders with reduced sensitivity). Calibrate using vinyl chloride in DMAC solution standards as follows: Fill a 100 ml volumetric flask to the mark with DMAC at room temperature. Seal the flask with a septum stopper and chill in an ice bath (about 30 min). Using a gas tight syringe, take a sample of pure vinyl chloride gas from a small cylinder. Inject exactly 1.0 ml of pure VC 1 gas into the cold DMAC contained in the volumetric flask. Mix thoroughly and allow the solution to warm to room temperature. Since 1.0 ml of VCl weighs 2550 p.g (25C, 760 mm Hg) the solution contains 255 Hg VCl/10 ml DMAC. Remove the septum stopper from the volumetric flask. Pipet 10.0 ml of the 255 M-g VCl/10 ml DMAC standard into a second 100 ml volumetric flask; dilute to the mark with pure DMAC and mix. This solution contains 25.5 p>9 VCl/10 ml DMAC. In the same manner, prepare standards containing 2.55 p.g VCl/10 ml DMAC and 0.26 y,g VCl/10 ml DMAC. By means of a 10.0 ml pipet, transfer 10 ml aliquots of each standard to a series of properly labeled 25 ml vials. Seal th vials with septum stoppers and aluminum caps. These DMAC solution standards in sealed vials are stable for at least two months under ordinary conditions (room tempe ra t u re). To calibrate the instrument, take a set (4) of standards and insert in the F-42 turntable (at operating temperature). Allow the samples to come to thermal equilibrium (one hour) and then analyze them using the conditions described in section 5* Calculate the calibration factor as follows: F(S) ' where: F($) = calibration factor for solution W = weight of VCl in solution (p,g VCl/10 ml DMAC) A = Area of VCl peak Average the calibration factor for the number of standards analyzed (results should agree with + 5% relative. Recalibration should be performed on a regular basis (weekly or biweekly). If the daily standard (section 7) varies more than + 10% from the correct value then the analytical system should be recalibrated. 7. Procedure a) Powders (resins and compound powder blends only). Weigh 4.00 _+ 0.01 g of R/C resin powder or PVC compound powder blend into a 25 ml vial. Close the vial with a septum stopper and seal with an aluminum cap. Analyze each sample in duplicate by weighing separate aliquots into separat vials. Also prepare an air blank by sealing an empty vial. Place the air VC4066, 5. blank vial in the turntable (at operating temperature) followed by a vial containing a VCl in N2 standard followed by the two sample vials. Addi tional samples may be inserted in the turntable at this time. Make sure that the F-42 operating parameters are set as indicated in section 5* Allow the samples to come to thermal equilibrium (one hour). Start the analysis sequence and analyze the samples (VCL elutes at 103-104 sec). After the samples have been analyzed, insert a needle (1/2", 22 gauge) into each vial to vent the internal pressure before removing the vials from the turntable. b) Solutions (pellets and finished products). Weigh a 0.5-0.7 g (to + 0.0001 g) sample into a 25 ml vial. Add 10.0 ml (pipet) of DMAC to the vial and seal with a septum stopper and aluminum cap. Analyze each sample in duplicate by weighing separate samples into separate vials. Warm the sample vials to ~ 50C and shake occasionally until the PVC has dissolved (4-6 hrs). Samples must be completely dissolved (except for fillers and pigments present in some compounds) before proceeding. Prepare a DMAC blank by pipeting 10.0 ml of pure DMAC into a vial and sealing as outlined above. Additional samples may be prepared in the manner outlined above at the same time. Place the DMAC blank in the turntable (at operating temperature) followed by one or more VCl in DMAC standards. Then insert the sample(s) in the turntable in known sequence. Make sure that the F-42 operating parameters are set as indicated in section 5. Allow the samples to come to thermal equilibrium (one hr). Start the analysis sequence and analyze the samples (VCl elutes at 103-104 sec). After the samples have been analyzed, insert a needle (1/2", 22 gauge) into each vial to vent the pressure before removing the vials from the turntable. A typical chromatogram is shown in Figure 1. 8. Calculations a) Powder (resins and compound powder blender only). The following equation relates the concentration of VCl in the PVC to the headspace concentration: where: cVCl C^s x Pa Ti | Mv Va + kt2J \TMt R CVC1 s concentration of VCl in the PVC before equilibration, ppm (wt/wt). CHS - concentration of VCl in the headspace after equilibration, ppm (v/v) Pa - barometric pressure. mm Hg Ti s ambient temperature, k Mv m molecular weight of VCl (62.5) Vg = volume of gas phase in the vial, ml VC4067 6. m* = weight of PVC sample R gas constant (62360) K = Henry's Law constant for fVC (6.52 x 10 *) T2 ** equilibration temperature, K Under ordinary laboratory conditions, the above may be reduced to the following (based on 4.00 g sample). tyci = CHS x 0*0194 or where: CVCl " F(p) * A -F(P) = calibration factor from section 6 A = area of the VC 1 peak The integrator may be programmed to automatically calculate the concentration of VCl in the PVC (see instruction manual for details). b) Solutions (pellets and finished products). Calculate the VCl content of the PVC sample as follows: where: CVCI F(s)xA s ^VCI concentration of VCl in the PVC sample before equilibrium, ppm (wt/wt). F(s) A calibration factor from section 6. area of the VCl peak S = sample wt, g. The integrator may be programmed to automatically calculate the concentration of VCl in the PVC (see instruction manual for details). 9. Precision and Accuracy For powder samples, the precision has been found to be +5% relative at the 1.0 ppm VCl level. The accuracy depends on the accuracy of the standards used for the calibration (generally +5% relative). The lower limit of detection has been found to be 0.005 ppm VCl. For pellet samples (analyzed as solutions) th precision has been found to be +10% relative at the 1.0 ppm VCl level. The accuracy depends on the accuracy of the calibration standards (generally +5% relative). The lower limit of detection has b en found to be 0.03 ppm VCl. The value for Henry's Low Constant for PVC is taken from A. R. Berens, Polymer Reprints, 197, September, 1974. VC406.8 FIGURE 1 Vinyl Chloride In PVC (DMAC Solution) HEWLETT-PACKARD 9280-0222 V VC4069 Appendix II Determination of Low Levels of Vinyl Chloride in Various Solvents (Headspace Analyzer Method) 1. Scope This method describes a procedure for the determination of low levels (ppb to low ppm range) of vinyl chloride (VCl) in various solvents such as water and dilute aqueous solution, 50% ethanol, and oils by a automated technique (headspace analysis). The lower limit of detection is dependent on the solvent to be analyzed but, in general, ranges from 1 - 2 ppb for water up to 6 - 8 ppb for oils. 2. ' Out! ine of the Method a) A 10.0 fjil sample of the liquid to be analyzed is transferred to a 25 ml vial and sealed with a serum stopper. The sample is then equilibrated for one hour at an elevated temperature (90C for water, aqueous solution, and oils; 65C for 507. ethanol solution). A sample of the vapor (headspace) above the liquid is then in jected into a gas chromatograph equipped with a flame ionization detector. By use of an automated turntable and sequencer up to 30 samples may be prepared and analyzed without additional operator attention. Standard gas samples and standard liquid samples containing known amounts of vinyl chloride are analyzed along with the unknown samples to calibrate the instrument. Calibration may be either based on measurement of peak height with a ruler or peak area measured with an automatic integrator. From the calibration standards the VCl content of the unknown sample is determined. 3. Apparatus a) Headspace Analyzer: Perkin Elmer Model F-42, equipped with a flame ionization detector, thermostatted sample turntable, backflush assembly, automatic timing sequencer. b) Sample Vials: 25 ml complete with butyl rubber septum caps and aluminum caps. These vials are available from Perkin Elmer to fit the F-42 turntable. c) Column: 5` x 1/8" SS packed with 10% Carbowax 20 M on 60/80 Chromosorb W HP followed by 10' x 1/8" SS packed with 10% OV-17 on 60/80 Chromosorb W HP. The two columns are connected in series (Carbowax 20 M section closest to injection port) with the backflush "P1 between the two columns. VC4070 - 2- d) Recorder: 1 mv full scale, 15"/hour. e) Integrator: Spectra-Physics System I (Perkin-Elmer Model 1). f) Syringes, Gas Tight: 10, 50, 100, 500 p,l and 1, 2, 5 ml. g) Ordinary Laboratory Glassware. 4. Reagents and Materials a) Vinyl Chloride: 99-9+% pure. b) Vinyl Chloride in Nitrogen Standards: 1 ppm (v/v), 10 ppm (v/v), 50 ppm (v/v), 100 ppm (v/v), 1000 ppm (v/v). Available from Supelco Inc. c) Nitrogen; Zero grade carrier gas. d) Hydrogen: High purity for FID. e) Air: High purity for FID. 5. Instrument Operating Conditions a) Water Bath: 90.0 C +0.1 (for water and oil samples) 65.0 C + 0.1 (for 507. ethanol samples) b) Gas Flow Controllers N2 (reference col.): 10 ml/min ( ~ 1 Kg/cm2) N2 (analytical Col.): 10 ml/min ( -- 1.2 Kg/cm2) N2 (backflush): 10 ml/min (^0.8 Kg/cm2) H2: 25.- 30 ml/min (2.0 Kg/cm2) Air: 250 - 300 ml/min (3.2 Kg/cm2) c) Gas Chromatograph Conditions Oven: 75 C Injector: 100 C Detector: 150 C Dosing Needle: 100 C d) Sequencer (Programmer) Controls Backflush: On P: 30 sec. (internal setting) I: 4 sec. A: 1.5 min. BF: 15 min. (internal 1 OX setting) S: 1 min. VC4071 3 e) Integrator Controls PW: 5 (peak width) SS: 15 (slope sensitivity = 1.2 pv/sec) FP: 5 (fused peaks) BL: 2 (baseline) T1: 100 (initial delay) T4: 250 (stop time) MA: 0 (minimum area) Program: none; i.e. print areas and times only. 6. Calibration: Determination of Henry's Law Constant a) Prepare gas standards as follows; Close an empty vial with a serum stopper and seal with an aluminum cap. Insert a 3" needle (22 gauge) through the septum of the vial. Connect this needle with metal tubing to a cylinder containing a VC 1 in nitrogen standard (1 - 1000 ppm). Insert a short 1/2" needle (22 gauge) through the setpum to act as a vent. Slowly flush (15 - 25 ml/ min) the vial with the vinyl chloride standard for approximately 10 min. After the vial has been filled with the VC1 standard, remove first the 3" needle and then the 1/2" needle from the vial. Repeat to fill additional vials with the various standards. At least five vials should be filled with each standard. If the Henry's Law constant for water is being determined, inject 20 - 25 y.1 of distilled water into each vial. If the Henry's Law constant for 507. ethanol or oil is being determined then the gas standards are used dry. Label the vials and use within 24 hours. b) Prepare liquid standards as follows: Add 10.0 ml of the liquid to a vial and seal as above. By means of a microliter syringe, add an appropriate amount of pure vinyl chloride gas to the liquid. A convenient range of concentrations to prepare is from 2 ppb to 10 ppm. Prepare at least five vials of each standard. Label and use within 24 hours. c) Analyze both the gas standards and the liquid standards as described in the Procedure ^Section 7). From the areas (or peak heights) obtained on the gas standards plot concentration vs. peak area. From the area of the VC1 peak obtained on each liquid standard and the calibration curves plotted for gas standards, determine the concentration of vinyl chloride in the headspace above each liquid standard. Calculate the Henry's Law Constant from the relationship on the following page. VC407? -4 K= L T2 MvVg h?r where: K Tt CVC1 " Pa mt = R Henry's Law constant equilibration temperature, K (363 for water and oil; 338 for 50% ethanol). concentration of VC1 in the liquid before equilibration 1 ppm (wt/wt). ambient temperature, K. concentration of VC1 in the headspace after equilibration, ppm (v/v). barometric pressure, mm Hg. molecular weight of VC1 (62.5) volume of gas phase in the vial, ml. weight of liquid, g. gas constant (62360). d) The following values for Henry's Law constants have been obta i ned by the R&D laboratories of the Ethyl Corp., Baton Rouge, LA. KH,0 50%E+0H Ksoya oil 1.5 x 10"6 (wet gas standard) 5 7.1 x 10 -6' (dry gas standard) .6 = 26-8 x 10~" (drY 9as standard) ^mineral oil = 19*5 x 10'6 (drY 9as standard) e) Gas standards must be analyzed with each batch of samples analyzed. f) The determination of Henry's Law constant for each liquid should be repeated whenever extreme accuracy is desired for the analysis of a group of samples or whenever any major operating parameter is changed. 7. Procedure a) Transfer 10.0 ml of the liquid to be analyzed to a vial, close with a serum stopper and seal with an aluminum cap. For highest accuracy each sample should be analyzed in triplicate. Sample weight may be determined from tables of density or by weighing a seprate 10.0 ml portion of the sample. b) Place the sample(s) in the turntable (at operating tempera ture). Make sure that all operating parameters as described in Section 5 are set on the instrument. VC4073 -5- c) Also place several gas standards In the turntable. d) Allow the samples and standards to equilibrate for one hour. e) Analyze the samples using the automated sequencer. Vinyl chloride elutes at 141 - 142 sec. under the conditions given in Section 5. A typical chromatogram is shown in Figure 1. f) After all of the samples and standards have been analyzed remove the vials from the turntable (relieve the internal pressure in each vial by venting with a 1/2" 22 gauge needle). g) From the data obtained on the gas standards, determine the headspace concentration of VC1 for each sample. 8. Calculations a) Calculate the concentration of VC1 in the liquid sample from the following relationship: CVC1 where: CVC1 CHS Pa Tl Mv mt R K T2 CHS pa Tl Hy Vq mtR concentration of VC1 in the liquid before equili bration, ppm (wt/wt). concentration of VC 1 in the headspace after equili bration, ppm (v/v). barometric pressure, mm Hg. ambient temperature, K. molecular weight of VC1 (62.5) sample weight gas constant (62360) Henry Law Constant (Section 6) equilibration temperature, K (363 for water and oil; 338 for 507. ethanol). VC4074 - 6- b) Under ordinary conditions, most of the terms in the above equation are constant i.e., Pa = 760; Tj = 298; Vg = 13.8 ml; etc. Therefore, the following simplified equations may be used; Water: 507. E+OH: Soya Oil; Mineral Oil: CVC1 C^s * 0.0049 (wet gas standard) CVCl = CHS x 0.0099 (dry gas standard) CVC1 = C^s x 0.029 (dry gas standard) CVCl * Chs x 0.022 (dry gas standard) 9. Precision and Accruacy a) The precision of the analysis is dependent on two measurements. (1) the precision of the gas standards (generally + 57. relative) and, (2) the precision of the determination of Henry's Law Constant. The latter varies from + 207. for water to + 5" or better for 507. ethanol and the oils. The accuracy is also dependent on these same two measurements i.e., preparation of gas standards and liquid standards, and is generally of the same order as the precision. b) The limit of detection is dependent on the lowest concentration of VC1 that can be measured in the headspace ( ~ 0.2 ppm) and the Henry's Law Constant for each liquid. For water the lower limit of detection is therefore 1 - 2 ppb (wt/wt) while for Soya oil the lower limit of detection is 5 - 8 ppb (wt/wt). VC4075 FIGURE 1 Vinyl Chloride In Water VC4076 APPENDIX III VC4077 VC4078 RETENTION TIME OF GAS STANDARDS Water at Five Weeks File No. 1 -5 45 - 49 89 - 93 Retention T1 me, Sec. 140 - 142* 143 - 144 143 - 144 * No. 1 has a retention time of 13 4 seconds. First analysis on startup of analyzer has an anomolous retention time. VC4080 VC4081 VC4082 r' 0 HriWIXTT-PACKAWD D2CQ-022 VC4084 WHmm RETENTION TIME OF GAS STANDARDS 5(H Aqueous Ethanol at Five Weeks File No. 1-5 45 - 49 74 - 78 Retention Time, Sec. 139 - 141 143 - 144 146 - 148 VC4086 RETENTION TIME OF GAS STANDARDS 50% Aqueous Ethanol at Five Weeks CD i LO File No. 1-5 74 - 78 Retention Time Sec. 139 - 141 142 " 144 146 - 148 VC4087 VC4088 ________________ nfwnr IIII s VC4089 4 X[; 4-A 11 ? t 112 127 119' 171 1 dS ii A it L A 5 CP 1 6 *- 3 1'- 2 it jy 1 2d7 4- A -i 6 3b file 1 5 xn ii' 7Ih 112 1 26 1 39 ' 1 72 l 9U A " L / Ovc 193 32 7 25H 43 1329 MI! .. u 1 l I4)- - Etk ! j l-i f !`f i ! I IJ .;i: L 1 t.J * i i>; i.A I - ii$ W cks i1. + ! J- !i to :V Hdp te m r-L-l tm 4- 00- .rT~?l i-i-i-- -A. I l-j- 1 II 1 .L. Hrwi rrr-f'An</''?n o HO-02??, vrir 'i- VC4090 1***1111 bC FlLt I 8'I9IIII Ii ** * li t' A i $II 45 RETENTION TIME OF GAS STANDARDS Water at 5 8 Days File No. 121 - 123 151 - 153 181 - 183 208 - 210 Retention Time, Sec. 142 - 143* 148 - 149 142 - 143 143 - 144 * No. 121 is 146 seconds. This is the first sample of a run. VC4094 - -I /- I I- er VC4095  / t t / f rrt rf19$ I i* vvi '"T T . r \ r i !> , - l VC4096 fr(pm ~T~0~p iI rj-i .. Li-i-i . i .Ll. --J;. tfrH+fl -<>- .4-1 j--*- _Li 'P TT-00- >-l--4 -H-H- 4=: /a - ^ -i a a 163 F ILL I 13 xc tf: 7 lyc 119 136 14 9 199 --- t- r. a 1132 b4L 693 96 266 i --T-p -f-H-r M- -ph1^- f TTt" ;;n: mr /a- 6 v. 1 5 4 FILL i 14 xn i rr iP o>* - 1" |-U1- -+-r TIME A !i A 1 19 13 6 149 1132 f,4 4 95 3 2 JO 1L 1 27 31 T1 4--1--L4-_--J 1 ' 1--4-- '"it t( pii((i*r Xu'l ^_!_r 1------- * * * - rTE+-iHi'-usiSi t"jj&TlLXfPc --CiZ^riirrIzTtILX-! :r r: rrjm: xrr. . Xm w&m a EE: 1,1 1 ,.L 3tft ,m XfU TFhH--j-H-r f-H- 'n TI~+T--Tt--4--1--s--f--; m: TTir ---j-LJ. -U I nt -H-p-rr-rr XJ..LL __ L-1~L !SE J--J- 35 *. i--r --LL+. i-f- _L. --K- 4-- Tt ir J_!__ ~rr i : .i.J _L m n-T -cn- I ... I rr7 -l-- "'I ; ' -i_i_ -a- rr -f- ztEr_"ir -D- --OJ- U-L :T" * r2f2E}. :zz rrii -to- -ro- .li. : )--- ----* T- 1-4 --- -- *- ' j J-Aj1. xt bxixxrt r- U- --o- L4i34J :A tIt ' rt -l: j... Tp :r^L 144 liU> 1 *** ii, Tt: RETENTION TIME OF GAS STANDARDS 5(H Aqueous Ethanol at 58 Days File No. 1 -3 31 - 33 61 - 63 88 - 90 Retention Time, Sec. 142 - 143* 141 - 142 145 - 146 147 * No. 1 has 136-second retention time. First analysis of a run. VC4099 ct VC4100 1-^ I* Tr',1t'- - 1 T^r I 2h 1 4U 172 1 3S 1 4CU 17 it I -I T IT i i -n----- tZ- A -/ A a 43 FILE 1 I 3 -XU IL TIME 112 1 2 >3 140 172 - AREA 4L0 JC4 04 3 27V 16 2* 12-A-i 6 a 44 file 14 XU l If. TIME i12 AKta 43C tr VC4104 Il1 TT1TT V<Z\- .5 .}! - -ULl... .tliti. L-i.i i. r.i :tb t- iti-l ;:tr; -i- i -i~U-L tir IP-t 11a in l4A * *\ I. / jxj fcl 17 K* 13-A -/ /4 /3 - A -I 6 33 FlLt 1 23 XU K t i >-1; 114 1 3u 1 4-1 1o9 174 A ii i A ./ 4 - del; 2V $* 17? 7 i y~ ~-iT-i-p'Tr tt U U U U I1v I' t *--r-rxo t. ! *) ; !.! | ;: J' A*, gg ;: & H L- -- ;_H -j- : r: IT j r;:;: ~H~h~ --l-i. I > ' iKv - -i-Tt-. 'TfiP' i. i -00.- n : : l_i h 4_ h itt*: "HH _ J-Ui-J-! J. jj- iii.' r-f-rp --- --hr- xixT- TiX]. -j h-H- t rn* *7 t --H-r 4 i~hi t SP -<7>- ' .*-1"7. i~r :nq: x: X_L -U- i~r U~L .c!nX-- xtix x --jTT 4X ~T~' T` 1 "r r- -n-i-- "17-'.:x J-L r-t+