Evaluation of selected solid adsorbents for passive sampling of atmospheric oil and natural gas non-methane hydrocarbons

Figures & data

Figure 1. Schematic of adsorbent cartridges. (a) Adsorbent cartridge dimension shown with front opening on the left. (b) Schematic of internal components of a cartridge loaded with 540 mg of adsorbent. The glass fiber filter disk, adsorbent, glass fiber filter disk, gauze, and spring (overlapping circles) follow from left. (c) Identical to (b), but with half the mass of adsorbent (270 mg) and a larger spring. Regardless of the adsorbent bed mass, the distance from the front opening to the adsorbent bed was 1.5 cm. The distance from the end opening to the adsorbent bed varied based on the mass and packing of the adsorbent bed. (d) Relative volume of the cartridge occupied by 27, 54, 135, 270, and 540 mg of adsorbent.

Table 1. Pearson correlation coefficients and p-values of cartridge and canister sampling during the summer and fall 2014 field campaigns. Canister and cartridge median relative standard deviations (RSDs), and median relative standard deviation ratios (RSDRs; cartridge/cans) from the Fall 2014 (n = 10) campaign are shown

	Summer 2014		Fall 2014
Compound	Pearson Correlation	P-Value	Pearson Correlation	P-Value	Canister RSD	Cartridge RSD	RSDR
Ethane	0.212	0.117	−0.380	0.278	0.011	0.413	42
Propane	0.855	0.000	0.925	0.000	0.011	0.043	5.2
i-Butane	0.672	0.000	0.884	0.001	0.013	0.062	4.5
n-Butane	0.931	0.000	0.925	0.000	0.009	0.089	14
i-Pentane	0.909	0.000	0.865	0.001	0.014	0.088	9.0
n-Pentane	0.740	0.000	0.879	0.001	0.008	0.099	18
n-Hexane	0.280	0.042	0.215	0.550	0.029	0.055	3.7
n-Heptane			0.879	0.001	0.097	0.301	3.1
Benzene	0.241	0.085	−0.308	0.386	0.020	0.153	7.8
Toluene	0.477	0.007	0.832	0.003	0.030	0.185	6.2
Ethylbenzene	0.136	0.333	0.071	0.846	0.025	0.134	12
o-Xylene	0.302	0.026	0.809	0.005	0.029	0.124	4.3

Table 2. Adsorbent cartridge types tested. Multiple cartridges of each unique adsorbent type and mass were assembled

Adsorbent Type	Mass Adsorbent (mg)	Approximate % of Tube Occupied by Adsorbent
Carboxen 1000	27	5
Carboxen 1000	54	10
Carboxen 1000	135	25
Carboxen 1000	270	50
Carboxen 1000	540	100
Carboxen 1016	520	100
Tenax GR	320	100
Carboxen 1000/1016	285/255	50/50

Figure 2. Blank chromatograms from cartridges loaded with (A) Carboxen 1000, (B) Carboxen 1016, (C) Tenax GR, and (D) a combination of Carboxen 1000 and 1016. Masses vary among adsorbents; 6.75 cm of each cartridge was occupied with adsorbent. The y-axis scale is identical for all panels.

Table 3. Comparison of Carboxen 1000, Carboxen 1016/1000 multi-stage adsorbent cartridges, and Tenax-GR cartridges. Analytical System Responses (ASRs) are the instrument response divided by the mixing ratio of the standard (ppbC) and the sample volume with relative standard deviations (RSD) for n = 5

	Carboxen 1000		Carboxen 1016/1000		Tenax-GR
Compound	ASR	RSD	ASR	RSD	ASR	RSD
Ethane	2.13	1.4%	0.388	39%	0.069	35%
Propane	6.61	0.24%	6.79	12%	0.194	41%
i-Butane			6.05	4.3%	0.197	14%
n-Butane	6.57	0.26%	6.15	3.0%	0.329	13%
iso-Pentane			5.98	3.8%	0.791	2.0%
n-Pentane	6.24	0.35%	5.63	3.6%	1.47	2.4%
Methylcyclopentane			18.6	4.5%	3.76	3.4%
2-Methylpentane			6.14	7.4%	2.03	39%
3-Methylpentane			6.07	5.1%	2.50	3.5%
Hexane	5.73	0.41%	5.88	2.1%	4.98	2.0%
n-Heptane			5.91	1.4%	5.93	1.1%
Benzene	5.99	1.1%	12.6	19%	18.3	12%
Octane			5.71	1.0%	5.89	1.1%
Toluene	5.41	0.94%	7.77	21%	7.02	7.3%
Ethylbenzene	2.87	5.9%	6.73	3.2%	6.53	3.1%
o-xylene	3.26	4.2%	6.32	1.3%	6.46	6.0%

Table 4. Linear regression results for analyte peak area (mV min) versus sampling time. Slope values represent analyte uptake rates. Other regression results are Pearson Correlation Coefficients (PCCs), and P-Values for collection of a standard mixture onto 1016/1000 and ‘1000/1016ʹ cartridges in a test atmosphere

	1016/1000			‘1000/1016’
Compound	Slope (x 10^−3)	PCC	P-Value	Slope (x 10^−3)	PCC	P-Value
Ethane	0.04	0.15	0.48	0.1	0.18	0.46
Propane	1.5	0.85	0.00	4.3	0.51	0.02
i-Butane	2.5	0.97	0.00	8.2	0.97	0.00
n-Butane	7.1	0.96	0.00	19	0.97	0.00
i-Pentane	8.9	0.97	0.00	14	0.95	0.00
n-Pentane	9.6	0.96	0.00	14	0.95	0.00
Hexane	2.3	0.94	0.00	3.0	0.90	0.00
Heptane	0.2	0.49	0.01	0.2	0.39	0.09
Benzene	−0.3	−0.06	0.79	−3.6	−0.18	0.46
Toluene	−0.4	−0.05	0.83	1.8	0.49	0.03
Ethylbenzene	0.2	0.27	0.19	0.3	0.22	0.36
o-Xylene	0.2	0.47	0.02	0.1	0.21	0.38

Figure 3. Comparison of diffusive uptake on multi-stage cartridges for passive sampling of a standard gas mixture. Black data correspond to 1016/1000 cartridges (see 2.4.2). Red data are ’1000/1016ʹ cartridges. Note differences in y-axis ranges.

Table 5. Details of adsorbent conditioning experiments

Conditioning Trial #	Oven Temp (°C)	Time (min)	N2 Flow (mL min^−1)
1	250	60	25
2	300	60	25
3	300	10	100
4	300	60	100
5	300	1200	25
6	350	1200	25

Figure 4. Total BTEX mean peak area from three cartridges filled with Carboxen 1000 adsorbent conditioned under different flow rates, temperature, and duration. Symbol shape indicates conditioning time, symbol color indicates nitrogen flow through the cartridge during conditioning, and region of the graph indicates conditioning temperature.

Figure 5. Sampling performance of cartridges with increasing masses of Carboxen 1000. Data are mean values of three determinations. All cartridges were actively loaded by pulling the same volume of a multi-component VOC standard through the adsorbent bed. Peak areas were normalized by the analyte mixing ratio in the standard (ppb) and the analyte effective carbon number (ECN; [Scanion and Willis 1985]).

Figure 6. Mass of analytes collected from a dry test atmosphere for ethane through n-hexane (1st and 2nd column) and benzene through o-xylene (3rd and 4th column). The 1st and 3rd columns show the results from the dry test atmosphere (Dry Bag), and the 2nd and 4th column the results from the humidified test atmosphere (Wet Bag). Results are from sampling on cartridges with increasing amounts of Carboxen 1000 adsorbent over 7 days. Multiple cartridges of each adsorbent bed mass were placed in the bag at the beginning of the experiment. One cartridge of each adsorbent bed mass was removed after 0.5, 1, 3, 5, and 7 days.

Figure 6. (Continued).

Table 6. Summary of H₂O loading experiment. A 200 mL min⁻¹ stream of zero air humidified to different RH % was sampled over three different adsorbent bed masses until the weight gain leveled off. The volume of humidified air and the saturation mass of H₂O loaded are shown for each adsorbent bed mass and RH %

H2O Loading
540 mg Carboxen 1000			135 mg Carboxen 1000			27 mg Carboxen 1000
RH (%)	Sample volume (mL)	Mass of H2O take up (g)	RH (%)	Sample volume (mL)	Mass of H2O take up (g)	RH (%)	Sample volume (mL)	Mass of H2O take up (g)
5	100	0.0002	5	100	0.0001	5	100	0.0008
25	500	0.0021	25	500	0.0002	25	500	0.0006
57	2400	0.0042	57	300	0.0009	57	1200	0.0008
75	47,200	0.12	75	18,200	0.030	75	11,200	0.017
92	29,000	0.12	92	8200	0.030	92	6200	0.018

Table 7. Median uptake rates (n = 30) of cartridges loaded with 27, 54, 135, 270, and 540 mg of Carboxen 1000 adsorbent in dry and humidified simulated environments over the course of 0.5–7 days. Relative Standard Deviation results were calculated from the differentials of the day-to-day data

Alkanes				Aromatics
Compound	RH (%)	Median UR (ng ppm^−1 min^−1)	Rel. Std. Dev. (%)	Compound	RH (%)	Median UR (ng ppm^−1 min^−1)	Rel. Std. Dev. (%)
Ethane	5	0.16	651	Benzene	5	2.71	93
	95	0.45	104		95	1.95	132
Propane	5	1.09	114	Toluene	5	2.41	66
	95	0.21	364		95	0.76	158
n-Butane	5	2.38	84	Ethylbenzene	5	1.76	144
	95	0.49	141		95	0.51	140
n-Pentane	5	2.57	87	m&p-Xylene	5	1.62	111
	95	0.99	93		95	0.57	117
n-Hexane	5	2.63	87	o-Xylene	5	1.74	112
	95	0.74	147		95	0.47	209

Table 8. Comparison of passive sampling uptake rates (in units of mL min⁻¹) for BTEX compounds reported in the literature and results from this study. An “x” in the diffusion cap (Diff. cap) column indicates studies that used a diffusion cap

							Uptake Rates (mL min^−1)
Study	Adsorbent Material	Cartridge Type	Diff. cap	Sampling Duration	Relative Humidity (%)	Temp (°C)	Benzene	Toluene	Ethylbenzene	m&p-Xylene	o-Xylene
Jia and Fu 2017*							0.73	0.65	0.60	0.60	0.60
Mowrer et al. 1996	Tenax-TA	Perkin-Elmer	x	7 d	N/A	−5 to 5	0.41	0.45	0.55	0.55	0.55
Markes 2021	Tenax-TA	Markes		8 h	N/A	N/A	0.41	0.44	0.46	0.42	0.42
Walgraeve et al. 2011a,b	Tenax-TA	Markes	x	7 d	5 to 80	18 to 24	0.27	0.32	0.35	0.36	0.34
Civan, Yurdakul, and Tuncel 2012	Chromosorb 106	Gradko		12 h	50 to 86	−1 to 20	0.39	0.44	0.39	0.40	0.35
Martin et al. 2010	Tenax-TA	Perkin-Elmer	x	14 d	0	20	0.57	0.53			0.52
Martin et al. 2010	Tenax-TA	Perkin-Elmer	x	14 d	80	20	0.54	0.48			0.47
Roche et al. 1999	Tenax-TA	Perkin-Elmer	x	0.5– 24 h	5 to 10	25	0.57	0.51	0.45	0.43	0.43
Hellen et al. 2002	Carbopack-B	Perkin-Elmer		14 d	N/A	−8 to 20	0.68	0.62	0.55	0.55	0.53
ISO 1999	Tenax-TA	Perkin-Elmer		8 h			0.63	0.56	0.50	0.47	0.47
Oury et al. 2006	Tenax-TA	Perkin-Elmer		1 d	50	20	0.40	0.45		0.45
Oury et al. 2006	Tenax-TA	Perkin-Elmer		3 d	50	20	0.32	0.40		0.43
Oury et al. 2006	Tenax-TA	Perkin-Elmer		6 d	50	20	0.27	0.37		0.42
Oury et al. 2006	Tenax-TA	Perkin-Elmer		14 d	50	20	0.20	0.30		0.37
Tolnai et al. 2001	Carbopack-B and Tenax-TA	Perkin-Elmer		7 d	n/a	−16	0.57 to 0.61	0.42 to 0.47	0.51 to 0.54	0.41 to 0.58	0.41 to 0.48
						Mean	0.46	0.47	0.48	0.45	0.47
						Rel. Std. Dev. (%)	36	22	18	17	18
This Study	Carboxen 1000	Perkin-Elmer		0.5– 7 d	5	23	1.02	0.77	0.49	0.45	0.48
This Study	Carboxen 1000	Perkin-Elmer		0.5– 7 d	95	23	0.73	0.24	0.14	0.16	0.13

*based on diffusion theory

Table 9. Linear regression results for paired canister and cartridge sampling with Pearson correlation coefficients and p-values for the 2016 campaign. Canister and cartridge median relative standard deviations (RSDs), and median of the ratio of the RSDs of both methods (relative standard deviation ratios (RSDRs: cartridge RSD/canister RSD))

	2016
Compound	Pearson Correlation	P-Value	Canister RSD	Cartridge RSD	RSDR
Ethane	0.68	0.015	0.031	0.08	4.1
Propane	0.97	0.000	0.024	0.12	8.8
i-Butane	0.56	0.058	0.11	0.12	1.4
n-Butane	0.96	0.000	0.011	0.097	8.3
i-Pentane	0.86	0.000	0.018	0.16	8.7
n-Pentane	0.79	0.002	0.03	0.22	7.7
n-Hexane	0.12	0.71	0.046	0.32	6.9
Benzene	−0.38	0.25	0.060	0.37	5.3
Toluene	0.79	0.011	0.040	0.34	6.2
o-Xylene	−0.73	0.025	0.11	0.88	6.2

Figure 7. Pearson Correlation of canister and cartridge data during the 2014 BCPH field campaign (Carboxen 1016/1000), the winter 2014 comparison study (Carboxen 1016/1000), and the 2016 comparison study (27 mg of Carboxen 1000). Data for some compounds are not available for each campaign. P-Values and RSDRs are available in Table 1 and Table 9.

Figure 8. RSD of paired passive VOC sampling on adsorbent cartridges versus the ambient mixing ratio (from the canister sampling) during the 2014 and 2016 field deployments. Please note the different sampling strategies deployed during the two campaigns (see text).

Scanion, J. T., and D. E. Willis. 1985. Calculation of the flame ionization detector relative response factors using the effective carbon number concept. J. Chromatogr. Sci. 23 (8):333–40. doi:https://doi.org/10.1093/chromsci/23.8.333.

 Web of Science ®Google Scholar

Jia, C. R., and X. Q. Fu. 2017. Diffusive uptake rates of volatile organic compounds on standard ATD tubes for environmental and workplace applications. Environments 4 (87): 1–19.

 Google Scholar

Mowrer, J., P.-A. Svanberg, A. Potter, and A. Lindskog. 1996. Diffusive monitoring of C6—C9 hydrocarbons in urban air in Sweden. The Analyst 121 (9):1295–300. doi:https://doi.org/10.1039/AN9962101295.

 Web of Science ®Google Scholar

Markes. 2021. Listing of uptake rates for axial passive samplers. Accessed March 30, 2021. https://markes.com/content-hub/application-notes/application-note-001.

 Google Scholar

Walgraeve, C., K. Demeestere, J. Dewulf, K. Van Huffel, and H. Van Langenhove. 2011a. Diffusive sampling of 25 volatile organic compounds in indoor air: Uptake rate determination and application in Flemish homes for the elderly. Atmos. Environ. 45 (32):5828–36. doi:https://doi.org/10.1016/j.atmosenv.2011.07.007.

 Web of Science ®Google Scholar

Civan, M. Y., S. Yurdakul, and G. Tuncel. 2012. Improvement of uptake rate equations depending on meteorological conditions for 25 volatile organic compounds. Talanta 99:720–29. doi:https://doi.org/10.1016/j.talanta.2012.07.011.

 PubMed Web of Science ®Google Scholar

Roche, A., R. Thevenet, V. Jacob, P. Kaluzny, C. Ferrari, P. Baussand, and P. Foster. 1999. Performance of a thermally desorbable type-tube diffusive sampler for very low air concentrations monitoring. Atmos. Environ. 33 (12):1905–12. doi:https://doi.org/10.1016/s1352-2310(98)00206-4.

 Web of Science ®Google Scholar

Hellen, H., H. Hakola, T. Laurila, V. Hiltunen, and T. Koskentalo. 2002. Aromatic hydrocarbon and methyl tert-butyl, ether measurements in ambient air of Helsinki (Finland) using diffusive samplers. Sci. Total Environ. 298 (1–3):55–64. doi:https://doi.org/10.1016/s0048-9697(02)00168-7.

 PubMed Web of Science ®Google Scholar

Iso, I. S. O. 1999. Ambient, indoor and workplace air ± determination of concentrations of volatile organic compounds in air, thermal desorption method ± part 2: diffusive sampling method. ISO/DIS, 16017-2.

 Google Scholar

Oury, B., F. Lhuillier, J. C. Protois, and Y. Morele. 2006. Behavior of the GABIE, 3M 3500, PerkinElmer Tenax TA, and RADIELLO 145 diffusive samplers exposed over a long time to a low concentration of VOCs. J. Occup. Environ. Hyg. 3 (10):547–57. doi:https://doi.org/10.1080/15459620600906613.

 PubMed Web of Science ®Google Scholar

Tolnai, B., A. Gelencser, and J. Hlavay. 2001. Theoretical approach to non-constant uptake rates for tube-type diffusive samplers. Talanta 54 (4):703–13. doi:https://doi.org/10.1016/s0039-9140(01)00322-8.

 PubMed Web of Science ®Google Scholar