Environ. Sci. Technol. 2001, 35, 771-780
Specific Surface Area of Snow Samples Determined by CH4 Adsorption at 77 K and Estimated by Optical Microscopy and Scanning Electron Microscopy F L O R E N T D O M I N EÄ , * A X E L C A B A N E S , ANNE-SOPHIE TAILLANDIER, AND L O ¨I C L E G A G N E U X CNRS, Laboratoire de Glaciologie et Ge´ophysique de l’Environnement, BP 96, 38402 St. Martin d’He`res Cedex, France
Snow is a divided medium that can adsorb atmospheric trace gases. Evaluating the impact of the snow cover on atmospheric chemistry therefore requires the knowledge of the specific surface area (SSA) of snow. This paper compares the results of three methods used to measure or estimate the SSA of four snow samples: CH4 adsorption at 77 K, optical microscopy (OM), and scanning electron microscopy (SEM, used only on two samples). Within error bars, CH4 adsorption and OM yield similar results on three of the four snow samples. Values for the 4th sample are within a factor of 2. For both samples where CH4 adsorption, OM, and SEM are used, all three methods yield similar results, but CH4 adsorption always has a better accuracy and a much better precision. Thus, despite its ease of use, estimates from OM images are often not accurate enough to monitor the evolution of snow SSA. The main sources of error in the OM method are the difficulty to determine snow crystal thicknesses and to take into account the topography of the snow crystal surface. The combination of CH4 adsorption and OM or SEM can provide useful information on the evolution of both the SSA and the shape of snow crystals. This will be useful to evaluate the respective contributions of adsorption/desorption and sublimation/condensation processes to the impact of the snow cover on atmospheric chemistry.
Introduction Snow can cover up to 50% of land masses (1), and the impact that this highly divided surface cover may have on atmospheric chemistry has been the subject of several recent studies (2-5). Gas-phase measurements above and within the snowpack have indicated that snow can exchange gaseous species such as formaldehyde (2, 3), acetaldehyde and acetone (4), NOx (5, 6), and HNO3 (6) with the atmosphere. The mechanisms of exchange are still unclear but may include adsorption/desorption from the snow surface (2-4), photolysis of a dissolved precursor (5, 6), co-condensation of the trace gas with water vapor with subsequent sublimation of the solid solution formed (6), and solid-state diffusion in and out of ice crystals (7, 8). * Corresponding author phone: (33) 476 82 42 69; fax: (33) 476 82 42 01; e-mail:
[email protected]. 10.1021/es001168n CCC: $20.00 Published on Web 01/17/2001
2001 American Chemical Society
Evaluating the contribution of adsorption/desorption to the above observations requires the knowledge of the surface area (SA) of snow. By surface area, we mean, following Gregg and Sing (9), the area of the snow sample considered that is accessible to gases. This area is expressed in m2 or cm2 for small values. By specific surface area (SSA), we mean the surface area of a given mass of snow. This is expressed in m2/g or here in cm2/g as values of SSA for snow are often small. The determination of the SSA of a snow sample thus requires the knowledge of its mass. Several methods can be used to measure the SSA of snow. Among the most relevant ones is the adsorption of gases, often at 77 K, as this process is similar to the adsorption of atmospheric gases, the topic of interest here. The use of this method to measure snow SSA has been detailed earlier (10). Recent uses include those by Hoff et al. (11), who used N2 adsorption at 77 K to measure the SSA of six snow samples, and by Chaix et al. (12) and Hanot and Domine´ (10), who used CH4 adsorption at 77 K to measure the SSA of a snow sample and the evolution of the SSA of a snow layer, respectively. The disadvantage of these adsorption methods is that they require liquid nitrogen (N2(l)), which is not always available for field studies in remote regions such as the Arctic and Antarctic. Alternative methods include stereology, which requires making serial sections of snow samples that have been previously reinforced by soaking into a liquid that is allowed to solidify and whose images must be analyzed using an appropriate method (13-15). Using a similar principle, another technology is X-ray tomography (16), which can have a resolution of 10 µm if synchrotron radiation is used. The simplest method is probably to analyze optical microscopy (OM) images that allow the determination of some or all the dimensions of snow crystals from which their mass and SA, and hence their SSA, can be obtained. OM can readily be transported anywhere, but its reliability needs to be evaluated as it only provides SSA estimates rather than actual measurements. Several causes of errors can be anticipated: (i) The determination of SSA values representative of snow layers requires the study of many crystals or grains selected without bias. (ii) Microstructures may not be visible. Hanot and Domine´ (10) invoked microstructures of about 3 µm to explain the high SSA of some of their snow samples. Such small microstructures, observed by Wergin et al. (17) using scanning electron microscopy (SEM), can increase SSA manyfolds and are difficult to see with OM. (iii) It is difficult to evaluate the SA of snow crystals having complex shapes. For example, determining the perimeter of dendritic crystals is difficult even if complex image analysis techniques are used, and this is a lengthy procedure (18) that reduces the advantage of OM. (iv) It may be difficult to measure all the dimensions of a snow crystal and therefore its mass: determining the thickness of a dendritic crystal is difficult as it has to stand on its side. Also, the volume of air bubbles reduces the ice mass in an often unknown manner (19). The purpose of this paper is to test whether OM can be used to estimate the SSA of snow samples in replacement of the gas adsorption technique. Four samples of different snow types were studied, and attempts to estimate their SSA by OM were made. The values obtained were compared to those obtained by CH4 adsorption, which is considered to be more reliable because the application of these measurements is to quantify the capacity of snow to adsorb gases. SEM images of two of the four samples have also been obtained to provide SSA estimates and to test for the presence of structures not VOL. 35, NO. 4, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
771
TABLE 1. Snow Samples Characteristics sample no.
date sampled and location
1 2 3 4
Jan 13, 1999, Alps Mar 4, 1999, Alps Apr 14, 2000, Arctic Apr 18, 2000, Arctic
date of snow fall
snow temp (°C)
depth of snow sample density (cm) (g cm-3)
Jan 13, 1999 -4 2 0.086 Mar 4, 1999 -0.4 8 0.10 Apr 13-14, 2000 -27 surface
