Water Adsorption Capacity of the Solid Adsorbents Tenax TA, Tenax

1 lists the amount of water vapor in atmospheres at 10, 20, 30, and 40 °C and 20, .... difference between these two brands is that the Carbotrap adso...
0 downloads 0 Views 816KB Size
Anal. Chem. 1995, 67, ~ ~ - 4 3 8 6

Water Adsorption Capacity of the Solid Adsorbents Tenax TA, Tenax GR, Carbotrap, Carbotrap Cy Carbosieve SIII, and Carboxen 569 and Water Management Techniques for the Atmospheric Sampling of Volatile Organic Trace Gases Detlev Helmig* and Lee Vierling National Center for Atmospheric Research, Boulder, Colorado 80307-3000

The solid adsorbents Tenax TA, Tenax GR, Carbotrap C, Carbotrap, Carbosieve SIII, and Carboxen 569, all of which are commonly used for sampling of volatile organic compounds (VOCs) from the atmosphere, were tested for their capacity to trap water vapor by a new experimental approach. Instead of performing breakthrough measurements, water uptake was determined gravimetrically until saturation. Test atmospheres at 10, 20,30, and 40 "C and 20,40,60,80, and 100%relative humidity (RH) in a temperature (T)- and humidity-controlledclimate chamber were collected onto sampling cartridges containing beds of the above-listed adsorbents. This technique allowed us to obtain information on the dependency of the water uptake on sampling volume, T,and RH and to deduce strategies for water management. Tenax TA, Tenax GR, Carbotrap, and Carbotrap C showed low water trapping of generally less than -2-3 mg of water/g of adsorbent under the highest humidity conditions tested. In contrast, the carbon molecular sieve type adsorbents Carboxen 569 and Carbosieve SI11exhibited substantially higher water trapping capacity, with up to -400 mg of water/g of adsorbent under the highest RH conditions tested. For Carboxen 569 and Carbosieve SIII, water saturation occurs relatively slowly; the water saturation point is reached only after the sampling of volumes that are much higher than typical volumes collected in ambient VOC sampling. For both of these adsorbents, it was shown that the amount of water adsorbed at a given temperature depends mainly on the RH levels. For a mdtibed adsorbent cartridge, the overall water adsorption could be calculated reasonably from the expected contributions of the individual components. Possible measures for minimizing the amount of water trapped are, for example, collecting small sampliig volumes, minimizing the amount of adsorbent in the sampling tube, moderate heating of the adsorbent tube during sampling, or including a dry purge step before thermal desorption. The analysis of volatile organic compounds (VOCs) in atmospheric samples usually includes sample concentration, separation of VOCs by gas chromatography (GC), and detection by sensitive GC detectors such as a flame ionization detector (FID), electron capture detector (ECD), or mass spectrometer (MS). Sample 4380 Analytical Chemistry, Vol. 67, No. 23, December 7, 7995

concentration requires cryogenic fi-eezeout of nonpermanent gases or selective trapping of VOCs onto solid adsorbents. A common interference in the analysis procedure arises from the simultaneous trapping of water vapor in the sample concentration step. Water can cause various problems in the concentration step by accumulating as ice during cryogenic preconcentration. It has also been noted that atmospheric humidity can reduce the adsorption efficiency for organic compounds during sampling onto solid ads or bent^.'-^ Other potential problems during analysis include possible loss and chemical transformations of organic trace gases in the waterhce matrix. GC injection and separation of atmospheric VOCs is commonly performed at subambient oven temperatures with cryogenic oven programming. Freezeout of water on a prefocusing trap or on the GC capillary column during cryogenic oven cooling can plug the trap or the column and interrupt the carrier gas f l ~ w . ~A- large ~ water background can also cause retention time shifts and pose problems during GC detection, in particular with EC detection, photoionization detection (PID), atomic emission detection (AED) in the hydrogen or oxygen measurement mode, and MS. Extinguishing of the FID flame during water elution has also been notedS6 Assuming a spherical geometry of an ice plug inside a capillary column, the theoretical minimum water amounts that would cause a column to clog for 0.53, 0.32, and 0.23 mm i.d. capillary columns are 77.9, 17.1, and 6.4 pg (nL) of water, respectively. In practice, however, larger amounts of water are expected to accumulate before column clogging occurs because the water plug will freeze out as an elongated volume rather than a perfect sphere. Table 1 lists the amount of water vapor in atmospheres at 10, 20, 30, and 40 "C and 20, 40, 60, 80, and 100%relative humidity 0, respectively, at standard p r e s ~ u r e .The ~ water vapor pressure (P, in mbar) at these conditions can be calculated as (1) Fabbri, A; Crescentini, G.; Mangani, F.; Mastrogiacomo, A R.; Bruner, F. Chromatographia 1987,23, 856-860. (2) Wood, G. 0. Am. Ind. Hyg. Assoc. /. 1987,48,622-625. (3) Ciccioli, P.; Cecinato, A; Brancaleoni, E.; Frattoni, M.; Liberti, A J. High Resolut. Chromatogr. 1992,15, 75-84. (4) Des Tombe, K; Verma, D. K; Stewart, L; Reczek, E. B. Am. Ind. Hyg. Assoc. 1991,52,136-144. (5) Sturges, W. T.; Elkins. J. W. J. Chromatogr. 1993,642, 123-134. (6) Helmig, D.; Schwarzer, N.; Steinhanses, J. J. High Resolut. Chromatogr. 1990,23,849-851. (7) Nelson, G. 0. Controlled Test Atmospheres. Principles and Techniques; Ann Arbor Science Publishers: Ann Arbor, MI, 1971: pp 221. 0003-2700/95/0367-4380$9.00/0 0 1995 American Chemical Society

Table 1. Water Content (in mg) of a 1 L Air Sample as a Function of Temperature and Relative Humidity at 1 bar Atmospheric Pressure

relative humidity (%) T CC)

20

40

60

80

100

10 20 30 40

1.9 3.5 6.1 10.2

3.8 6.9 12.2 20.5

5.6 10.4 18.2 30.7

7.5 13.8 24.3 41.0

9.4 17.3 30.4 51.2

P(T,RH) = RH

x 6.1078 x

lo-' exp[17.27T/(T+ 237.311

(1) with Tin "C and RH in %. It becomes obvious that for many typical ambient sampling conditions, trapping and injection of even < 1% of the sample water content of a 1 L air sample is sufficient to pose serious interferences in the chromatographic analysis process. A number of different approaches have been developed to circumvent the interferences from water enrichment described above: 1. Water vapor can be selectively removed from the sampling flow either prior to the sample concentration or during a sample transfer step. Techniques that have been used include the following: (A) passing the sample flow through a trap containing a drying agent such as K Z C O ~ , ~N- '~~ z C OCaC03,13 ~,~ NaOH,13 NazS04,3 Mg(C104)~,'~-'~ P z O ~ ,Aquasorb '~ ( P 2 0 5 on a vermiculite base) ,j or Ascarite (NaOH on a silicate base);j,I7 (B) use of ion exchange membranes (Nation) for water removal in the sample f10w;3J8-z5 (C) passing the sample flow through a cryogenically cooled freezeout trap and selectively freezing out water at subzero temperature^;^^^^^^^-^* (8) Westberg, H. H.; Rasmussen, R A; Holdren, M. Anal. Chem. 1974,46,

1852-1854. (9) Schmidbauer, N.; Oehme, M. J. High Resolut. Chromatogr. 1985,8, 404506. (10) Schmidbauer, N.; Oehme, M. J High Resolut. Chromatogr. 1986,9, 502505. (11) Martin, R S.; Westberg, H.; Allwine, E.; Ashman, L.; Framer, J. C.; Lamb, B. J. Atmos. Chem. 1991,13, 1-32. (12) Staehelin, J.; Graber, N.; Widmer, H. M. Int. J Enoiron. Anal. Chem. 1991, 43, 197-208. (13) Knoeppel, H.; Versino, B.; Schlitt, H.; Peil, A; Schauenburg, H.; Vissers, H. Proceedings of the First European Symposium on Physico-Chemical Behaoiour of Atmospheric Pollutants, Ispra, 16-18 Oct, 1979; Commission of the European Communities: Brussels, 1980; pp 25-40. (14) Burgermeister, S.; Zimmermann, R L.; Georgii, H.-W.; Bingemer, H. G.; Kirst, G. 0.;Janssen, M.; Emst, W.J. Geophys. Res. 1990,95,20607-20615. (15) Hasegawa, A; Yajima, I. Bunseki Kagaku 1991,40, 489-494. (16) Lijfgren, L.; Berglund, P. M.; Nordlinder, R; Petersson, G.; Ramnas, 0.Int. J Environ. Anal. Chem. 1991,45, 39-44. (17) Greenberg, J. P.; Lee, B.; Helmig, D.; Zimmerman, P. R. J. Chromatogr. 1994,676, 389-398. (18) Foulger, B. E.; Simmonds, P. G. Anal. Chem. 1979,51, 1089-1090. (19) McClenny, W. A.; Pleil, J. D.; Holdren, M. W.; Smith, R. N. Anal. Chem. 1984,56, 2947-2951. (20) Pollack, A J.; Holdren, M. W.; McClenny, W. A.J. Air Waste Manag. Assoc. 1991,41, 1213-1217. (21) Hofinann, U.; Hofmann, R; Kesselmeier, J. Atmos. Environ. 1992,26A. 2445-2449. (22) Thomton. D. C.; Bandy, A RJ. Atmos. Chem. 1993,17, 1-13. (23) Blomquist, B. W.; Bandy, A R; Thornton, D. C.; Chen, S. J. Atmos. Chem. 1993,16, 23-30. (24) Simmonds, P. G.; O'Doherty, S.; Nickless, G.; Sturrock, G. A; Swaby, R.; Knight, P.; Ricketts, J.; Woffendin, G.; Smith, R Anal. Chem. 1995,67, 717-723. (25) Gong Q., Demejian K. L.J. Air Waste Manag. Assoc. 1995,45,490-493.

(D)condensation of the bulk water in the concentrated sample by flowing the heated concentrate through a colder water reservoir at controlled flow and t e m ~ e r a t u r e ; ~ ~ (E) using a water trap unit that relies on the physical condensation of water at ambient or moderate subambient temperatures on glass beads prior to the sample elution onto the GC column (this technique can reduce up to 70 p L of water from the sample stream to an interference-free level);30 (F) use of a recently introduced cyclone water management system that relies on the reduction of the sample water content by expanding the concentrated sample flow in a turbulent vortex tube (the adiabatic cooling causes water vapor to condense onto the chamber surface wall and thereby reduces the sample dewpoint to