Selenium in the atmosphere

Jan 2, 2010 - quantitatively in terms of their slightly different, and often well understood, chemical properties. Stable isotopic abun- dance measure...
0 downloads 0 Views 290KB Size
COMMUNICATIONS

Selenium in the Atmosphere Yoshikazu Hashimoto' John W. Winchester2 Massachusetts Institute of Technology, Cambridge, Mass.

rn Selenium was determined by neutron activation in 22 samples of snow and rain collected during the 1964-1965 winter and in seven air samples collected during the following spring in Cambridge, Mass. Sulfate was determined gravimetrically or turbidimetrically in most of the samples. In these samples and in well water and Cambridge tap water analyzed for comparison, the selenium concentration averaged 0.2 Mg. per liter of water or per 200 cu. meters of air, and the average value of the weight ratio Se/S = 1 X loT4.Approximately the same ratio is commonly found in igneous and sedimentary rocks, sulfide ores, and fossil fuels, but the sea water ratio is one thousandth as large. This information implies mainly a terrestrial, including pollution, source for selenium as well as sulfur in the atmosphere of the city of Cambridge.

T

he very rapidly growing literature o n sulfur in the atmosphere (Junge, 1963) reflects its importance, both as a hazardous pollutant and as a significant natural component. I n contrast, the literature on atmospheric selenium (Gassman, 1918; Suzuki, 1959) is very sparse, although its toxicity has been recognized (Cerwenka and Cooper, 1961). I n geochemical studies of trace substances there is considerable value in measuring the concentrations of closely related elements, for differences in abundances may be interpreted quantitatively in terms of their slightly different, and often well understood, chemical properties. Stable isotopic abundance measurements have been especially successful and have revealed details of geochemical processes in the atmosphere, hydrosphere, and lithosphere which could not be seen by measurements of single elements. Different elements with similar properties can be studied in the same way, and the halogen components of sea salt have already been extensively investigated (Winchester and Duce, 1967). A study of the element pair sulfur and selenium may reveal more about the atmospheric behavior of both elements than a study of either Present address, Department of Applied Chemistry, Keio University, Koganeishi, Tokyo, Japan. Present address, Department of Meteorology and Oceanography, University of Michigan, Ann Arbor, Mich. 338 Environmental Science and Technology

element alone. This paper is a report of our first determinations of atmospheric selenium. Experimental Classical determination of Se by precipitation and turbidimetry (Robinson et al., 1934) is sensitive only to 10 Mg, anc' is not suitable for the present investigation. Colorimetric determination by 3,3 '-diaminobenzidine (Hoste, 1952 ; Cheng, 1956; Dye, et al., 1963) and determination by neutron activation (Bowen and Cawse, 1963) both offer much greater sensitivity, and activation analysis offers possibility of chemical yield control during analysis by radioactivity measurement. The greatest neutron activation analysis sensitivity is obtained by measuring 17.5-second Seii"', but the short half life demands special equipment not available to us. High sensitivity is also obtained by measuring 18.6-minute Se81, but interference by radioactivity of 18-minute BrS0 makes its measurement difficult. Adequate sensitivity is obtained after several hours of irradiation by measurement of 120-day Se'j, and 0.01 ,ug. of Se is a practical sensitivity limit when gamma radiation is measured with a scintillation spectrometer. The long half life permits the analysis of a large number of samples and standards in a single neutron irradiation. Approximately 1 liter of rain o r snow melt water was collected o n polyethylene, filtered on filter paper to remove soot, acidified with "03, evaporated to 20 ml. a t 40 millibars and 50" C. in a n evacuated flash evaporator, and reduced to 0.1 ml. in a Teflon beaker by gentle heating In a dust-free enclosure. Tests were made of the procedure using carrier-free 120-day SeT5tracer (Hashimoto and Winchester, 1967) to select conditions such that no Se would be lost in these steps. The samples were then transferred to quartz ampoules for neutron irradiation. Air samples of approximately 100 cu. meters were taken by passing air at about 1 cu. meter per hour either through a n aqueous bubbler or through a 1-micron pore diameter Millipore filter. The aqueous solutions were processed in the same way for irradiation, but the Millipore filters were irradiated directly without processing. Neutron irradiations of u p t o 14 hours were carried out in the MIT reactor a t a flux of approximately 2 X lOI3 n/sq. cm. sec. in the vertical facility. Three samples and standard solutions sealed in quartz were contained in each of eight aluminum vials arranged lengthwise in the reactor, and small differences in neutron flux over the irradiation region were accounted for by the arrangement of the standards. Three weeks

+

after irradiation, Se was distilled from HC1 HBr solution by passing in Bra vapor, and halogens were expelled from the distillate by evaporating in a Teflon beaker in the presence of H N 0 3to 1-ml. volume. Precipitation of the activity with mixed Se(IV) and Te(V1) carriers was accomplished by first ensuring chemical exchange by oxidation with H N 0 3 and H202in the presence of HBr and then by reducing t o the elemental form by hydrazine in HCI. The precipitate was coagulated by heating, filtered o n glass fiber filter paper, dried, and transferred t o a vial for gamma counting in a NaI(T1) well-type scintillation counter and a T M C 400-channel analyzer. Characteristic 0.138-, 0.410-, and 0.265 0.280 0.305-m.e.v. gamma-ray photopeaks provided a n easy identification of Seis radioactivity. The integrated sum of these three peaks, after background subtraction, was calculated for each sample and standard, and Se contents were computed assuming that radioactivity is directly proportional to the amount of Se present. All samples were recounted approximately 6 months later, and the radioactivity levels had decreased by the amount expected from the half life of 120-day Se7j. Most of the samples were also analyzed for sulfur using ASTM gravimetric or turbidimetric determination of BaS04 (ASTM Standards, 1964). Although we consider these results t o be less precise than the neutron activation determination of Se, they are consistent with the literature (Junge, 1963) and are useful for comparative purposes.

+

+

Table I. Atmospheric Se and S Concentrations Location 1 2 1 2 1 2 1 2 1 2 3 2 3 2 3 1 2 Mean

Results and Discussion The analytical results are given in Table I. Most of the precipitation and all of the air samples were collected on the MIT campus located in the middle of the city of Cambridge near Boston. There appears t o be no significant difference between Se and S in samples a t ground level and on the roof of the 90meter high earth science building nor between these and samples taken a t other points in Cambridge. However, samples taken in semirural Topsfield and Boxford and in the smaller city of New Haven suggest that Se concentration is lower in these areas. The average value of the Cambridge precipitation samples is 0.21 pg. of Se per liter and 2.0 mg. of S per liter, making the average weight ratio Se,!S = 1 x In the two Cambridge air samples where sulfur determinations were made, about the same ratio was found, and the concentrations indicate that 1 liter of precipitation is roughly equivalent t o 200 cu. meters of

Date,

Se,

S,

Seis

...

..

3.8 1.2 1.4 2.7 1.3 2.3 2.7 0.9 0.9 1.5 1.4 1.2 2.0

0.4 4.4 3.7 0,5 1.2 1.1 0.4 0.9 0.7 0.7 0.6 0.8

Remarks Falling snow Falling snow Rain ice ice Rain Falling snow Falling snow Falling snow Falling snow Falling snow Falling snow Falling snow Rain Ground snow Falling snow Ground snow Falling snow Falling snow

0.08 0.03 0.12 0.04 0.10

1.5 2.4 0.5 0.7 1.5

0 0 2 0 0

Ground Ground Ground Ground Ground

Feb. 27

0.11 0.09

1.7 2.0

0 7 Tap water 0 5 Well water

11 20 20 27 27 1 1

0.03 0.03 0.11 0.13 0.16 0.06 0.10

...

1964-65

mg.0

Dec. 18 Dec. 18 Jan. 3 Jan. 3 Jan. 10 Jan. 10 Jan. 16 Jan. 16 Jan. 24 Jan. 24 Feb. 22 Feb. 25 March 20 March 20 March 29 March 29 March 29

0.15 0.09 1.40b 0.70 0.13 0.16 0.53 0.52 0.14 0.16 0.25 0.10 0.08 0.06 0.10 0.09 0.10 0.21

4 5 5 6 6

Jan. 31 Feb. 27 March 20 March 20 March 20

1 5

...

1 1 1 1 1 1 1

May May May May May June June

... 3.9 0.5b 2.5

... ... 0.97 ... 0.69

...

X 104 , ,

.

0.2 . .. 2.8

5 1 4 6 7

+ +

Air, Air, Air, 1 4 Air, Air, 0 9 Air, . . Air,

snow snow snow snow snow

bubbler bubbler filter bubbler filter bubbler filter

Sampling locations. I . MIT camnus roof of earth science building. 2. Ground location near 1. 3. Cambridge, near Central Square. 4. Topsfield, Mass. 5. Boxford, Mass. 6. New Haven, Conn., Yale University campus = Micrograms of Se and mg. of S per liter of water or 100 cu. meters of b

air. Excluded from mean.

air. Volume 1, Number 4,April 1967 339

Table 11. Se/S Weight Ratios in Geochemical Materialss Material Se/S X lo4 3 Meteorites 2 Igneous rocks 1 Magmatic sulfides Argillaceous sediments 3 1 to40 Oxidates Limestones 0.7 0.1 Sedimentary iron ores 0.001 Sea water