1、Designation: D 5303 92 (Reapproved 2007)An American National StandardStandard Test Method forTrace Carbonyl Sulfide in Propylene by GasChromatography1This standard is issued under the fixed designation D 5303; the number immediately following the designation indicates the year oforiginal adoption or
2、, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This test method covers the determination of traces ofcarbonyl sulfide (COS) i
3、n propylene. It is applicable to COSconcentrations from 0.5 to 4.0 mg/kg (parts per million bymass). See Note 1.NOTE 1The lower limit of this test method is believed to be below 0.1mg/kg, depending on sample size and sensitivity of the instrumentationbeing used. However, the cooperative testing prog
4、ram was conducted inthe 0.5 to 4.0 range due to limitations in preparing commercial testmixtures.1.2 The values stated in SI units are to be regarded asstandard. No other units of measurement are included in thisstandard.1.3 This standard does not purport to address all of thesafety concerns, if any
5、, associated with its use. It is theresponsibility of the user of this standard to establish appro-priate safety and health practices and determine the applica-bility of regulatory limitations prior to use. Specific hazardsstatements are given in Section 8.2. Referenced Documents2.1 ASTM Standards:2
6、D 3609 Practice for Calibration Techniques Using Perme-ation TubesD 4468 Test Method for Total Sulfur in Gaseous Fuels byHydrogenolysis and Rateometric ColorimetryE 840 Practice for Using Flame Photometric Detectors inGas Chromatography3. Summary of Test Method3.1 A procedure is given for removing a
7、 sample from thesample cylinder, separating COS from propylene, detectingCOS, calibrating the detector, quantitating COS content in thesample, and assaying the gas standard. General comments andrecommended techniques are given.3.2 Arelatively large volume of sample is injected into a gaschromatograp
8、h having a single packed column, operated iso-thermally at 10 to 50C, that separates COS from propylene.COS is detected with a flame photometric detector.3.3 Calibration data, based on peak areas, are obtained usinga known gas standard blend of COS in the range expected forthe sample. The COS peak a
9、rea in the sample is measured andthe concentration of COS calculated.3.4 The COS gas standard blend is assayed prior to use forcalibration.4. Significance and Use4.1 In processes producing propylene, COS usually remainswith the C3hydrocarbons and must be removed, since it affectsproduct quality. COS
10、 acts as a poison to commercial polymer-ization catalysts, resulting in deactivation and costly processdowntime.4.2 Accurate gas chromatographic determination of traceCOS in propylene involves unique analytical problems becauseof the chemical nature of COS and idiosyncracies of trace levelanalyses.
11、These problems result from the reactive and absorp-tive nature of COS, the low concentration levels being mea-sured, the type of detector needed, and the interferences fromthe propylene sample matrix. This test method addresses theseanalytical problems and ways to properly handle them to assureaccur
12、ate and precise analyses.4.3 This test method provides a basis for agreement betweentwo laboratories when the determination of trace COS inpropylene is important. The test method permits severalcalibration techniques. For best agreement between two labs, itis recommended that they use the same calib
13、ration technique.5. Interferences5.1 Hydrogen sulfide (H2S) or sulfur dioxide (SO2) can bepresent in the propylene and must be separated from COS. (SeeNote 2.)NOTE 2H2S and SO2are separated from COS with the CarbopackBHT 100 columns or with the Chromosil 300 column.1This test method is under the jur
14、isdiction of ASTM Committee D02 onPetroleum Products and Lubricants and is the direct responsibility of SubcommitteeD02.D0.03 on Propylene.Current edition approved Nov. 1, 2007. Published January 2008. Originallyapproved in 1992. Last previous edition approved in 2002 as D 530392(2002)e1.2For refere
15、nced ASTM standards, visit the ASTM website, www.astm.org, orcontact ASTM Customer Service at serviceastm.org. For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page onthe ASTM website.1Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West
16、Conshohocken, PA 19428-2959, United States.6. Apparatus6.1 Gas ChromatographAny gas chromatograph (GC)equipped with a flame photometric detector/electrometer sys-tem (FPD), as described in 6.2, may be used. A GC/FPDequipped with an output signal linearizer is also permitted.6.2 Detector System, flam
17、e photometric detector, eithersingle or dual burner design. Noise level must be no more thanone recorder chart division (see 6.5). The signal for COS mustbe at least twice the noise level at the 0.1 mg/kg level. Adiscussion of this detector is presented in Practice E 840. Theelectrometer used with t
18、he detector must have a sensitivity of1012A full scale ona1mVrecorder to achieve optimumdetectability at lowest levels.6.3 ColumnAny column that will effect the completeseparation of COS from propylene and other compoundsnormally present in propylene concentrates, and that is suffi-ciently inert to
19、preclude the loss of COS, may be used.Columns that meet these criteria, and that were used in thecooperative study for this test method, are listed in Table 1.6.4 Sample Inlet SystemAny gas sampling valve or gastight syringe that will permit introduction of up to 5.0 mLto thecolumn, and that will no
20、t cause any loss of COS, is suitable.6.5 RecorderAny strip chart recorder with a full scalerange of 1 mV, a maximum full scale balance time of 2 s, anda minimum chart speed of 0.5 cm/s, may be used.6.6 Data Handling SystemAny commercially availableGC integrator or GC computer system capable of accur
21、atelyintegrating the area (uVs) of the COS peak is satisfactory. Datasystems that will linearize the logarithmic output of the FPDare also satisfactory.6.7 Sample Cylinders, 300 mL capacity or larger, fluorocar-bon lined stainless steel, Type DOT 3E, 12409 kPa (1800 psi)working pressure.7. Reagents
22、and Materials7.1 Air, zero grade.7.2 Carbonyl sulfide (COS), lecture bottle, 97.5 % min.(WarningToxic! See Section 8, Hazards.)7.3 Gas Calibration Blends, 1 to 10 mg/kg COS in eithernitrogen, argon, propylene or a propylene/argon mixture. Theycan be obtained from any commercial supplier or prepared
23、asshown in Appendix X1 or Test Method D 4468.7.4 Gas Sampling Syringe, 0.1, 1.0, and 5.0 mL.7.5 Gas Sampling Valve and Sample Loops, fluorocarbon or316 stainless steel. See Footnote B of Table 1.7.6 Glass Vials, 125 cm.7.7 Hydrogen, pure grade, 99.9 %.7.8 Isooctane (2,2,4-trimethylpentane), sulfur f
24、ree, mini-mum purity 99 mol %. (WarningFlammable! Health Haz-ard.)7.9 Nitrogen or Helium, 99.999 % min.7.10 TFE-fluorocarbon septa and aluminum seals for vials.8. Hazards8.1 Carbonyl sulfide is toxic and narcotic in high concen-trations, and upon decomposition can liberate hydrogen sulfide.Exposure
25、to dangerous concentrations of COS is most likelywhen handling the pure component for preparation of standardblends for assaying the COS calibration gas standards.9. Sampling9.1 Supply samples to the laboratory in high pressurecylinders coated internally with TFE-fluorocarbon, or other-wise speciall
26、y treated to reduce or eliminate loss of COS dueto reaction with the cylinder walls.9.2 The sample cylinder and contents should be at roomtemperature prior to sampling to the chromatograph. Testsamples as soon as possible after receipt.NOTE 3Cooperative studies indicate that the measured value for C
27、OSwill decrease with time.9.3 Place the sample cylinder in a vertical position and useeither of the following two techniques to obtain a vaporizedsample from the container for introduction into the GC.9.3.1 Connect the sample cylinder to the sampling valve onthe chromatograph, using a minimum length
28、 of 316 ss tubing,so that sample is withdrawn from the bottom of the cylinder.Adjust the flow rate from the sample cylinder so that completevaporization of the liquid occurs at the cylinder valve. A flowrate of 5 to 10 bubbles/s through a water bubbler placed at thesample vent is sufficient (see Not
29、e 4). Turn the sampling valveto the “flush” position and flush for approximately 15 s. Shutoff the cylinder valve and allow the pressure to drop toatmospheric.NOTE 4If the flow rate is too fast, warming of the valve can berequired to avoid freezing and to ensure complete vaporization of thesample.9.
30、3.2 Alternatively, obtain a sample with a gas tight syringe.A convenient way to do this is to use flexible plastic tubing toconnect the bottom of the sample cylinder to the water bubblerand then to pierce the tubing with the syringe needle after flowis established.10. Preparation of Apparatus10.1 In
31、stall in the GC according to the manufacturersinstructions any of the columns that meet the criteria in 6.3. Setthe instrument conditions as follows:TABLE 1 Suitable GC Columns and TemperaturesAColumnNumberSize, m 3 mm TubingBTypePacking and OvenTemperature, C10.93 3.78 SS Porapak R, 80/100 Mesh; 47
32、2143 3.78 TFECCarbopack BHT 100, 40/60Mesh; 25,40D31.83 3.78 TFE Carbopack BHT 100; 25,30D413 3.78 TFE Porapak Q, AW, 50/80 Mesh2.4 3 3.78 TFE (Above in Series); 7452.43 3.78 SS Carbopack BHT 100; 476283 3.78 TFE Carbopack BHT 100, 40/60Mesh; 5073.63 3.78 TFE Carbopack BHT 100, 40/60Mesh; 5084.33 3.
33、78 TFE Chromosil 300; 50E9613 3.78 TFE Hayes Sep Q, 80/100 Mesh; 65AThese columns have been tested cooperatively and found suitable for use withthis test method.B316 SS Tubing for columns or connection of sample cylinder to samplingsystem can be TFE lined internally to improve on system stability. T
34、his tubing iscommercially available from chromatography vendors.CTFEHomopolymer of tetrafluoroethylene.DIdentical columns used by different labs at different temperatures.EPropyne (methyl acetylene) can interfere with COS using this column.D 5303 92 (2007)210.1.1 Oven Temperature, as determined by c
35、olumn used,10.1.2 Detector, 100 to 200C, and10.1.3 Injector, 100 to 150C.11. Calibration11.1 Three methods of calibration are permitted. These arethe Standard Sample Method (see 11.2), the Permeation TubeMethod (see 11.3) and Blend Preparation Techniques (see11.4). Obtain a calibration standard acco
36、rding to one of thesemethods, which are described below. Then follow the proce-dure in 12.1-12.4 and the calculations described in 11.5.11.2 Standard Sample MethodPurchase a certified com-mercial calibration sample of 10 mg/kg COS in propylene, orother suitable matrix gas such as nitrogen, argon, or
37、 apropylene/argon mixture. If an inert gas is chosen, the usermust ensure that the column is actually effecting a separation ofCOS and propylene. Establish a calibration curve with thestandard sample using either a gas syringe or different sizesample loops. For example, assume the normal sample size
38、 forthe analysis is 1.0 mL and the calibration range to beestablished is 0.5 to 5 mg/kg of COS. Establish a calibrationcurve by injecting the volumes of a 10 mg/kg standard sampleshown in the first column of the table below. The equivalentconcentration of COS in a 1.0 mL sample would be that shownin
39、 the second column:Standard Sample Equivalent Concentration, COS mg/kg0.05 0.50.10 1.00.20 2.00.30 3.00.40 4.00.50 5.011.3 Permeation Tube MethodRefer to Practice D 3609for directions on using permeation tubes.11.4 Blend Preparation TechniquesTechniques for thepreparation and assay verification of c
40、alibration blends in thelaboratory are described in Appendix X1 and Appendix X2.Also, a technique using a moving piston graduated cylinderapparatus, that is described in the calibration section of TestMethod D 4468, can be used. However, some laboratories havefound that the preparation of such blend
41、s is far from easy, andsuccessful efforts require considerable knowledge and experi-ence.11.5 QuantitationThe flame photometric detector re-sponds logarithmically to the mass of the sulfur present in theflame. Some GC/FPD systems are programmed to linearizelogarithmic data, and with such systems the
42、 output can becorrelated directly with the COS concentration, using a singlepoint calibration. Calculate a calibration factor, F, in accor-dance with (Eq 1) below:F 5 C/A (1)where:F = calibration factor,C = concentration, mg/kg, of COS in this test method, andA = area (uVs) of the COS peak in this t
43、est method.F will be used in (Eq 2) in 13.1.1. However, if a linearizer isnot used, or if the data system does not have a provision tohandle logarithmic output, use the method in 11.5.1 or thealternate in 11.5.2, below:11.5.1 Calculate the nanogram (ng) amounts of sulfur, asdescribed in Appendix X3,
44、 for each injection of the standard,and plot the natural logarithm (1n) of peak area versus the 1n(ng) of sulfur, as illustrated in Table 2 and Fig. 1. The plotshould be a straight line.11.5.2 Alternatively, plot the concentration of COS inmg/kg versus the square root of the peak area. This plot sho
45、uldalso be a straight line.12. Procedure12.1 Using either the gas sampling valve or a gas tightsyringe, as described in 9.3, inject the sample into the gaschromatograph.12.2 Record the response of the FPD on the strip chartrecorder as the COS elutes from the column.12.3 Alternatively, obtain the com
46、puter or integrator outputof COS retention time and peak area.12.4 Obtain duplicate chromatograms of the sample. Fig. 2illustrates a typical analysis using a Carbopack BHT-100column.13. Calculation13.1 Depending on the method of calibration used (seeSection 11), determine the concentration of COS in
47、 the sample.13.1.1 If the system provides a linearized output, determineCOS concentration according to (Eq 2), below:COS, mg/kg 5 F 3 S (2)where:F = calibration factor from (Eq 1), andS = area (uVs) of the COS peak from the sample.13.1.2 If a calibration curve of 1n peak area versus 1n (ng)sulfur wa
48、s used (see 11.5.1), then determine the concentrationof COS as shown in Appendix X1.13.1.3 If a calibration curve of concentration versus logpeak area was used (see 11.5.2), then determine the COSconcentration as follows:13.1.3.1 Calculate the log of the area of the COS peak ofsample.TABLE 2 Example
49、 of COS Calibration DataANOTECOS Standard (3.00 ng S/cm3).Amount of StandardInjected (cm3)Amount ofStandard Injected (ng S)yB(peak area units)3.0 9.0 533.0 9.0 512.5 7.5 392.5 7.5 382.0 6.0 252.0 6.0 231.5 4.5 181.5 4.5 161.0 3.0 101.0 3.0 90.5 1.5 30.5 1.5 4ACorrelation coefficient of fit (r) = 0.9952:m = slope (detector response factor) = 1.4920,y = peak area units,z = nanograms of sulfur as COS injected, andb = intercept = 1.8394.BCalibration equation: y = bzm.D 5303 92 (2007)313.1.3.2 Take the COS concentration directly from thecurve using the log value fr
