8

Click here to load reader

59657824

  • Upload
    anhthoo

  • View
    212

  • Download
    0

Embed Size (px)

Citation preview

Page 1: 59657824

Authors

Vince Giarrocco and Roger FirorAgilent Technologies, Inc.2850 Centerville RoadWilmington, DE 19808-1610USA

Abstract

An Agilent 6890 Series gas chromatog-raphy system was used to determinetrace (low ppm) levels of hydrocarbonimpurities in high-purity ethylene andpropylene. The gas chromatograph (GC)was equipped with a heated gas samplevalve, split/splitless inlet, and flameionization detector (FID). An Agilent HP-PLOT Al2O3 column was used forseparation of the trace hydrocarbons.Impurity levels below 10 ppm (mole)were easily detected in both ethyleneand propylene.

Introduction

High-purity ethylene and propyleneare commonly used as feedstocks forproduction of polyethylene,polypropylene, and other chemicals.

Trace Level Hydrocarbon Impurities inEthylene and Propylene

Typically, these low molecular weightmonomers are of very high purity(99.9+ percent). However, hydrocar-bons, sulfur compounds, and otherimpurities in feed streams can causesuch problems as reduced catalystlifetime and changes to product qual-ity. Process yields can also beadversely affected. Many impuritieshave been identified as potential cont-aminants (1,2).

Recently, ASTM has proposed severalprocedures to determine trace hydro-carbon impurities in both ethyleneand propylene (3). These methods,currently in the investigation stage,use alumina porous layer open tubu-lar (PLOT) columns. This applicationnote describes the suggested Agilentconfiguration for such methods andillustrates resulting separations ofboth quantitative calibration blendsand actual process samples. Theseproposed methods should be valuablein meeting commercial specifications.

Experimental

All experiments were performed onan 6890 Series gas chromatograph(GC) equipped with a split/splitlessinlet and capillary optimized flame

ionization detector (FID). All gasflows and pressures within the GCwere controlled electronically. Gassample injections were made using anautomated sample valve placed in the6890 valve oven (80 ºC). The gassample valve was interfaced to thecapillary inlet using an aluminumtube (1/8-in.) that jacketed the stain-less steel transfer line (option 860).The inlet was fitted with a split/split-less liner (part no. 19241-60540). Allinjections were made in the splitmode.

A 50-m × 0.53-mm, HP-PLOT Al2O3

“M” column was used for separation.For ethylene analysis only, a 30-m ×0.53-mm, 5-µm HP-1 column wasplaced directly behind the HP-PLOTcolumn. The two columns werejoined using a glass press-fitconnector.

The Agilent ChemStation was used tocontrol the 6890 Series GC and to pro-vide data acquisition and peak inte-gration. The ChemStation wasoperated at a data acquisition rate of10 Hz.

Application

Gas Chromatography

March 1997

Page 2: 59657824

2

Standards for retention time andresponse factor calculation wereobtained from DCG Partnership(Houston, Texas, USA 77061). Sam-ples used for this work were obtainedfrom commercial sources.

Table 1 lists the entire set of equip-ment and conditions.

Results and Discussion

Ethylene

The configuration used for ethyleneanalysis is found in figure 1.

The HP-PLOT Al2O3 column was usedfor hydrocarbon separation. The useof HP-PLOT Al2O3 columns for lighthydrocarbon analyses has been previ-ously described (4). These columnsexhibit excellent separation charac-teristics for the C1 through C5isomers.

The proposed method for ethylenespecifies the use of a second nonpo-lar column placed after the HP-PLOTalumina column to improve the sepa-ration of impurity peaks eluting onthe tail of ethylene. This nonpolarcolumn gains importance for tracelevel analysis, where higher concen-trations of ethylene (99.9 percent andhigher) exhibit increased tailing. Noattempt was made to compare separa-tions without the nonpolar Agilent HP-1 column.

Item DescriptionGas ChromatographG1540A 6890 Series GCOption 112 Split/splitless inletOption 211 Capillary optimized FIDOption 701 6-port gas sample valve and automationOption 751 Valve ovenOption 860 Valve to inlet interfaceColumn • 50-m x 0.53-mm HP-PLOT Al2O3 “M” (part no. 19095P-M25)

• 30-m x 0.53-mm, 5-mm HP-1 (part no. 19091Z-236), used forethylene analysis only

Data AcquisitionG2070AA Agilent ChemStationOperating ParametersInjection port temperature 200 °CDetector temperature 250 °CSplit ratio 10/1 to 50/1 depending on sample FID conditions 30 mL/min hydrogen, 350 mL/min air, nitrogen make-up

(25 mL/min column + makeup)Temperature program • Ethylene: 35 °C (2 min), 4 °C/min to 190 °C

• Propylene: 40 °C (2 min), 4 °C/min to 190 °CInjection volume • Ethylene: 0.5 mL

• Propylene: 0.25 mLColumn flow • Ethylene: 6 mL/min constant flow (10 psi)

• Propylene: 3.5 mL/min constant flow (4 psi)Valve temperature 80 °C

Table 1. Instrument Configuration and Operating Conditions

Figure 1. Valve drawing for impurities in ethylene.

FID

Sample In/Out

Helium

0.5-cc Loop

Split Inlet10/1 to 40/1

50-m x 0.53-mmHP-PLOT Al2O 3 “M”

1

4

6

5

2

3 30-m x 0.53-mm, 5-µmHP-1

Page 3: 59657824

3

Figure 2 shows the chromatogram ofan ethylene calibration blend contain-ing most of the major hydrocarbonimpurities. This sample was analyzedat a split ratio of 10/1. The concentra-tion of most components (except forethane) ranges from 8 to 12 ppm(mole). For this analysis, baselineseparation is achieved for all theimpurities except for propane. Totalanalysis time is approximately 30 minutes. Because this separation ismore than adequate, analysis timecan be reduced by increasing the tem-perature program rate. Based uponconditions used for this analysis,most components can be detected atthe 1-ppm level.

Chromatographic results for twoprocess ethylene samples are given infigures 3 and 4. The sample presentedin figure 3 contains only methane,ethane, and propylene as impurities.Less than 1-ppm methane wasdetected. The ethylene sample infigure 4 shows a high concentrationof methane, with trace amounts ofethane, propane, and propylene.

1. Methane (10 ppm)2. Ethane (219 ppm)3. Ethylene 4. Propane (12 ppm) 5. Propylene (9 ppm),

6. Isobutane (10 ppm)7. n-Butane (10 ppm)8. Propadiene (10 ppm)9. Acetylene (10 ppm)10. t-2-Butene (8 ppm)

11. 1-Butene (8 ppm)12. Isobutylene (9 ppm)13. c-2-Butene (9 ppm)14. 1,3-Butadiene (10 ppm)15. Methylacetylene (9 ppm)

Figure 2. Chromatogram of ethylene calibration blend, split ratio 10/1.

Figure 3. Chromatogram of process ethylene sample, split ratio 20/1.

1. Methane (<1 ppm)2. Ethane (5 ppm)3. Ethylene4. Propylene (113 ppm)

Page 4: 59657824

4

Propylene

The configuration used for propyleneanalysis is illustrated in figure 5. Thisconfiguration is essentially the sameas for ethylene, but without theHP-1 column. The sample volume wasreduced to 0.25 mL. Propylene wassampled in the gas state.

1. Methane (4969 ppm)2. Ethane (18 ppm)3. Ethylene4. Propane (2 ppm)5. Propylene (5 ppm)

Figure 4. Chromatogram of process ethylene sample.

Figure 5. Valve drawing for impurities in propylene.

FID

Sample In/Out

Helium

0.25-cc Loop

Split Inlet10/1 to 40/1

50-m x 0.53-mmHP-PLOT Al2O 3 PLOT “M”

1

4

6

5

2

3

Page 5: 59657824

5

A chromatogram representing thetrace hydrocarbon impurities inpropylene is shown in figure 6. Thissample was analyzed at a split ratio of20/1. The concentration of most impu-rities range from 8 to 20 ppm.Ethylene is present at a higher con-centration level. Most of the impuri-ties in the sample are well separatedusing the conditions described intable 1. Cyclopropane elutes justbefore propylene and is baselineseparated under these conditions.Several of the C4 hydrocarbons eluteon the tail of the high-purity propy-lene. This affects the lower limit ofdetection for these peaks, comparedto those components that are baselineseparated. The remainder of the C4

and C5 impurities are well separated.

1. Methane2. Ethane (10 ppm)3. Ethylene (50 ppm)4. Propane5. Cyclopropane (10 ppm)6. Propylene7. Isobutane (10 ppm)8. n-Butane (7 ppm)9. Propadiene

10. Acetylene11. t-2-Butene (10 ppm)12. 1-Butene13. neo-Pentane14. Isobutylene (9 ppm)15. Isopentane16. c-2-Butene (9 ppm)17. n-Pentane (10 ppm)18. 1,3-Butadiene (9 ppm)

Figure 6. Chromatogram of propylene calibration standard.

Page 6: 59657824

6

For comparison, figure 7 shows theanalysis of a second calibration blendcontaining a higher level of impurities(50 to 1000 ppm).

Figure 8 presents the chromato-graphic results for a high-puritypropylene process sample. Thissample contains only ethane andpropane impurities.

Summary

This application note describes twomethods for analyzing trace hydrocar-bon impurities in ethylene and propy-lene. These methods use a gas samplevalve with split injection, an AgilentHP-PLOT Al2O3 and HP-1 (for ethyl-ene only) column, and an FID. Impu-rities below the 10-ppm mole levelcan be easily quantitated using thesemethods. For some impurities, espe-cially those that are well separatedfrom the large ethylene or propylenepeaks, detection limits were esti-mated to be about 1 ppm.

1. Methane2. Ethane3. Ethylene4. Propane (988 ppm)5. Cyclopropane (100 ppm)6. Propylene7. Isobutane (129 ppm)8. n-Butane9. Propadiene (62 ppm)

10. Acetylene (48 ppm)11. t-2-Butene12. 1-Butene13. Isobutylene15. Isopentane16. c-2-Butene17. 1,3-Butadiene18. Methylacetylene (100 ppm)

Figure 7. Chromatogram of propylene calibration blend containing higher levels ofimpurities.

Figure 8. Chromatogram of process propylene sample.

1. Ethane (82 ppm)2. Propane (4358 ppm)3. Propylene

Page 7: 59657824

7

References1. ASTM Method D 5325, “Standard

Guide for the Analysis of EthyleneProduct,” Annual Book of Stan-dards, Volume 5, ASTM, 100 Bar Harbor Drive, West Conshohocken, PA 19428USA.

2. ASTM Method D 5273, “StandardGuide for the Analysis of Propy-lene Concentrates,” Annual Bookof Standards, Volume 5, ASTM,100 Bar Harbor Drive, West Conshohocken, PA 19428USA.

3. Proposed methods for hydrocar-bon impurities in ethylene andpropylene by gas chromatographyare being investigated underASTM committee D-2,subcommittee D.

4. “Optimized Determination ofC1–C6 Impurities in Propyleneand Ethylene UsingHP-PLOT/Al2O3 Columns,”Agilent Technologies, Inc.Publication (43) 5062-8417E,March 1994.

Page 8: 59657824

Agilent shall not be liable for errors contained herein or forincidental or consequential damages in connection with thefurnishing, performance, or use of this material.

Information, descriptions, and specifications in this publicationare subject to change without notice.

Copyright© 2000Agilent Technologies, Inc.

Printed in the USA 3/20005965-7824E