TECHNICAL NOTE
www.rsc.org/jem | Journal of Environmental Monitoring
Transfer of CO2, N2O and CH4 to butyl rubber (polyisobutylene) septa during storage† ´ ric van Bochoveb Se´bastien F. Lange,a Suzanne E. Allaire*a and E Received 21st December 2007, Accepted 8th April 2008 First published as an Advance Article on the web 28th April 2008 DOI: 10.1039/b719729b Greenhouse gases are more sampled than ever because of environmental interests. Gas samples are often inserted into vials with gas tight butyl rubber septa before concentration analysis. Little is known on the global transfer property of butyl rubber septa for CO2, N2O and CH4. Sorption kinetics were measured by injecting CO2, N2O or CH4 into glass vials with either one of four butyl rubber septa types and stored during 90 days. CO2 and N2O concentrations decreased during storage depending upon septa type and initial concentration, with the highest linear rate being 0.023 for CO2 and 0.0015 mg L3 day1 for N2O. When a low concentration was injected, CH4 concentration changes over time were small and did not differ between septa types. Sorption isotherms were measured using nine concentrations and stored during 45 days. CO2 sorption isotherms ranged from 0 to 3.7 103 m3 m2 and N2O from 0.3 to 1.4 103 m3 m2. Examples of errors associated with the use of these butyl rubber septa are given.
Introduction With the event of the Kyoto protocol and environmental awareness, greenhouse gases are now frequently sampled for concentration analysis related to agricultural, environmental, atmospheric, and industrial research or assessment. The samples are often stored in vials sealed with septa and brought to the laboratory for concentration measurements. To make a meaningful measurement, interference between samples and sample vials should be avoided; or if it can not, it should be minimized, consistent and quantifiable. Thus, for concentration measurements, septa are chosen not only for their availability, in relation to the analytical instrument and cost, but also for their low gas sorption, dissolution, and permeability. Previous studies have tested the reliability of vial seals made of different materials for gas measurement.1 Among a large range of materials used for septa, butyl rubber septa seem to have lower gas diffusion, sorption or production of gases.2–4 But some authors have found influence of rubber butyl septa on gas concentration.4–7 Since then, many studies have used butyl rubber septa for gas proofing their vials with interest in N2O,8 CO2,9–11 or CH4.12 Even if some authors have found an influence of septa on gas concentration, most of their studies did not indicate if corrections were made for losses or changes in concentration during storage (9 out of 10), or for actual storage time before analysis (9 out of 10), and did not describe their septa other than stating they were made of rubber butyl septa (95%). In addition, there is a Horticultural Research Center, Universite´ Laval, 1980 Hochelaga, Que´bec, G1K 7P4, Canada. E-mail:
[email protected]; Fax: +1 (418)656-7871; Tel: +1 (418)656-2131 b Agriculture and Agrifood Canada, Sainte Foy, 2560 Hochelaga, Que´bec, G1V 2J3, Canada † Electronic supplementary information (ESI) available: CO2, N2O and CH4 concentrations as a function of time for all four septa. See DOI: 10.1039/b719729b
This journal is ª The Royal Society of Chemistry 2008
little information available concerning transfer properties of CO2, CH4 and N2O on polyisobutylene.13,14 The goal of this study was to measure a global transfer property (combination of sorption, dissolution, permeation, and diffusion) of four gray butyl rubber septa and their kinetics for CH4, CO2, and N2O. No attempt is made to differentiate between the processes. Only the combined effect is measured and is referred to in this paper as sorption.
Experimental Septa description and vial preparation Four types of rubber butyl septa having 20 mm diameter were compared. The four septa types were lyophilisation rubber butyl stoppers having the following respective properties: (1) three legs chlorobutyl/50 (3-C-W) having 836 mm2 and weighted 2.364 0.012 g (made by Helvoet-Pharma bought from Wheaton cat. 224 100–202, Millville, New Jersey, USA), (2) two legs made of halobutyl (2-H-W) having 526 mm2 surface area in contact with the gas and weighted 2.372 0.013 g (made by Hospira bought from Wheaton cat. 224 100–193), (3) two legs made of chlorobutyl/50 (2-C-W) having 674 mm2 and weighted 2.222 0.021 g (made by Helvoet-Pharma bought from Wheaton cat. 224 100– 194), and (4) two legs made of chlorobutyl (2-C-K) having 749 mm2 and weighted 2.405 0.011 g (made by West Pharmaceutical, bought from Kimble Glass Inc.,Vineland, New Jersey, USA, cat.73828A-21). Surface areas given for each septum correspond to the surface in contact with the gas inside the vials once capped. All septa look similar. They are identified as made of the same materials in the catalogues. The manufacturers did not give any information on their manufacturing process nor on their exact formulation Prior to gas injection, 11.1 mL glass vials (Agilent, 20 mm, 5182–0838) were vacuumed three times at 100 KPa. He (5.0 Ultra High Purity, Praxair, Que´bec, Canada) was injected J. Environ. Monit., 2008, 10, 775–777 | 775
between the first two vacuum events using a vacuum/capping instrument (CRB, Que´bec, Canada, cat. Betsy). Similar procedure was followed in other studies.6 This procedure was made to minimise atmospheric gas concentration in the vials. Once vacuumed, all vials were hermetically closed using this instrument with either one of the septa and cramped with aluminium seals (20 mm, Wheaton, Millville, New Jersey, USA, cat. 224178– 01). Standards and blanks were prepared with the same procedure. Blanks were prepared for each septa type, concentration, time, and replicate. All vials were randomly vacuum and injected to minimize bias. Sorption kinetics Either CO2, N2O or CH4 gas was present in all vials. No gas mixtures were studied. Pure CO2 (CD 4.8RS, Praxair, Que´bec, Canada) was injected into vials at either one of three concentrations: 0.327 mg L1, 1.144 mg L1, or 2.45 mg L1 with gastight syringes (10 or 25 mL, Hamilton Company, Reno, Nevada, USA, cat. 80065, 80265). A 0.653 mg L1 of pure N2O (NS 2.5 AA, Praxair, Que´bec, Canada) and 0.238 mg L1 of CH4 (ME 3.7UH, Praxair, Que´bec, Canada) was injected with 10 mL syringes. All vials were injected with He (HE 5.0 Ultra High Purity, Praxair, Que´bec, Canada) to pressurise all vials at the same pressure (101 KPa). Gas concentration was monitored during 90 days at 15 sampling dates. A set of 3 replicates was prepared for each date, septa, gas, and concentration (740 vials + standards). All vials were prepared and stored in the laboratory at 23 1 C. Additional kinetics tests were performed to detect any short term sorption or emission by the rubber butyl septa. No significant changes in concentration of any of all three gases on any of the septa were detected during the first eight hours. Thereafter, only long term sorption will be discussed. Sorption isotherms The same vials, syringes, gases, and methodology were used as for the first experiment. Nine concentrations covering more than two orders of magnitude for each gas and each septum were injected for calculating sorption isotherms. The concentrations for the CO2, N2O and CH4 were as follows: 0.084, 0.411, 0.737, 1.23, 2.53, 4.17, 8.25, 16.4, and 24.6 mg L1 of CO2. There was no 0.0 mg L1 of CO2 concentration because of a constant residual CO2 concentration in the vials. The concentrations were 0.0, 0.33, 0.65, 1.17, 1.63, 2.44, 4.07, 8.14, and 16.3 mg L1 of N2O, and 0.0, 0.12, 0.24, 0.60, 1.49, 2.98, 5.95, 8.93, and 14.88 mg L1 of CH4. A set of 3 replicates were prepared for each septa, gas,
and concentration (324 vials + standards). The vials were stored in the laboratory at 23 1 C during 45 days. Gas chromatography Concentration measurements were completed on a gas chromatograph (microGC, CP 4900, Varian Inc., Middelburg, Netherlands) equipped with a TCD detector and a 10 m PPQ column (PPQHBF, Varian). Pure He was the carrier gas. The detection limit of the instrument is
