Hydrogen Flame Emission Spectrophotometry in Monitoring Air for

Marcela Burguera , Stanley L. Bogdanski , Alan Townshend , David J. Knowles. C R C Critical ... T.Howard McGee , Ralph E. Weston. Chemical Physics Let...
0 downloads 0 Views 512KB Size
Hydrogen Flame Emission Spectrophotometry in Monitoring Air for Sulfur Dioxide and Sulfuric Acid Aerosol SIR: A review of recent literature on flame emission photometry by Gilbert (1) illustrates the large potential usefulness of this analytical principle. By optimization of experimental parameters, it is possible to increase considerably the sensitivity of the flame emission technique of detecting SOz and H2S04 airborne droplets. Optimization of burner design and HP:air ratios and flow rates are discussed. Present detection limits and experimental analytical curves are given. EXPERIMENTAL

Apparatus. The apparatus setup shown schematically in Figure 1 was used in this study t o determine t h e emission spectrum of SO, and to develop a n optimum burner design. T h e Bausch and Lomb grating monochromator (Serial No. DD 3734) has a range from 200 to 700 mp and a scanning speed of 667 mp/minute. An RCA 1 P21 photomultiplier tube was supplied with power by a Kepco Model ABC 1500M regulated d.c. power supply having a n adjustable range from 0 to 1500 volts and 0 to 5 ma. A Hewlett-Packard Model 425 A d.c. microvolt-ammeter with a range from 10 x 10-12 to 3 x amp. measured the photomultiplier tube current and drove a Heathkit recorder by Daystrom. The optical filter spectral transmission was determined with the use of a Cary Model 14 spectrophotometer. A Neptune Products, Inc. Model Yo. 4-K Dyna-Pump with Kel-F diaphragm and check valves supplied air to the burner. The burner tip, housing, and optical-burner assembly as illustrated in Figure 2 were constructed in our laboratories and are similar to those used by the Illinois Institute of Technology Research Institute (3) in a pentaborane monitor. The burner tip, shielded by a 1-inch diameter borosilicate-glass tube, consisted of a l/*-

inch diameter stainless steel tube inserted inside a 1/4-inch-diameter stainless steel tube with both tubes held in place by a stainless steel tee with appropriate fittings. Reagents. hlatheson prepurified grade of hydrogen fuel was used. Because no response could be obtained from room air, no special precautions were taken t o obtain purified air for establishment of base lines or t o measure responses from known concentrations of SO2 for calibration purposes. The SO, used was Natheson anhydrous grade of 99.98% purity. For t h e laboratory flame apparatus calibration and accurate establishment of other SO2concentrations, a modified West-Gaeke method of SO, analysis as reported by Hochheiser (2) was used. The reagents used in these analyses were all reagent grade. &SO4 aerosol concentrations were determined by the electrical conductivity of solutions of samples collected either in a sonic velocity liquid impinger or a Casella cascade impactor. Procedure. Using the apparatus illustrated in Figure 1, no emission was detectable from a n unshielded Hz flame burning air containing SO, in concentrations as high as 10 p.p.m. (v./v.). When a borosilicate-glass tube 1 inch in diameter by 3 inches in length was placed around the burner tip and held in place by a rubber stopper through which the burner tip had been inserted, a visible blue glow outside the combustion zone of the flame was emitted (Figure 3a and b ) a t SOz concentrations less than 1 p.p.m. With spectral and flow rate data obtained from use of the apparatus illustrated in Figure 1, a light-tight burnerdetector housing (Figure 2) was designed and built. This unit incorporated all associated gas handling and electronic equipment used in the spectral studies. A ll/,-inch-square narrowpass optical filter with a peak transmission at 402 mp was chosen because it, of all the immediately available

narrow-pass filters, came closest to the SO2 peak emission intensity a t about 380 mp. To reduce the time to show a stable response to SOz, the apparatus was alternately exposed to room air for 2 minutes and to SO, for 2 minutes. Peak responses a t the end of the 2minute SO, exposure are reported as instrument responses in calibrating the apparatus over the range of 0.1 to 1.3 p.p.m. SO2in air. An attempt to use the apparatus to selectively measure concentrations of SO2 or airborne droplets of when both were present in the same chamber gave marginal quantitative results for H2SO4a t a concentration of 0.17 mg. per cubic meter in the presence of 0.53 p.p.m. SO,. The sample was drawn for 3 minutes through a l-inchdiameter glass filter-holder containing a Gelman type A glass-fiber filter (no support screen) to measure SO, concentration without particulate H2S04; room air was passed through the system for 2 minutes to re-establish the base line. The sample then was drawn through a tube containing no filter for 2 minutes to measure the combined concentration of SOz and &Sod; room air was again passed through the system to establish the base line before repeating the sequence. To determine the concentration of H2SO4, the difference in response of two successive samples was measured, first with filter and then L5ithout filter. With only HzS04 mist in the chamber, the instrument was operated in the same manner as when SOz was measured. RESULTS A N D DISCUSSION

Spectral Distribution. The upper curve of Figure 4 illustrates t h e emission intensity spectral distribution of SO, in the shielded Hz-air flame with H2 introduced by t h e central burner tube as in Figure 2b. As is illustrated in thiq graph and con-

8 AND L MONOCHROMATOR

FLAME

1

TEFLON SPACER IP21 PHOTOMULTIPLIER

a T

NEEDLE VALVES\E

I

= FLOW METERS I

&R