Identification of species in a wet flue gas desulfurization and

May 1, 1984 - Identification of species in a wet flue gas desulfurization and denitrification system by laser Raman spectroscopy. David. Littlejohn, S...
0 downloads 0 Views 1MB Size
Environ. Sci. Technol. 1904, 18, 305-310

(19) Gschwend, P. M.; Hites R. A. Geochim. Cosmochim.Acta 1981,45, 2359-2367. (20) Readman, J. W.; Mantoura, R. F. C.; Rhead, M. M.; Brown, L. Estuarine, Coastal Shelf Sci. 1982, 14, 369-389. (21) Wakeham, S. G.; Schaffner, C.; Giger, W. Geochim. Cosmochim. Acta 1980,44, 415-429. (22) Youngblood, W. W.; Blumer, M. Geochim. Cosmochim. Acta 1975, 39, 1303-1314. (23) Laflamme, R. E.; Hites, R. A. Geochim. Cosmochim.Acta 1979,43,1687-1696. (24) Prahl, F. G.; Carpenter, R. Geochim. Cosmochim.Acta 1979, 43, 1959-1972. (25) Grimmer, G.; Bohnke, H. 2. Naturforsch. C: Biosci. 1977, 32C, 703-711. (26) Dasch, J. M. Enuiron. Sci. Technol. 1982, 16, 639-645. (27) US. Bureau of the Census, Census of Housing (1940-1970).

(3) Prahl, F. G.; Bennett, J. T.; Carpenter, R. Geochim. COSmochim. Acta 1980,44, 1967-1976. (4) Barrick, R. C.; Hedges, J. I.; Peterson, M. L. Geochim. Cosmochim. Acta 1980,44, 1349-1362. (5) Wakeham, S. G.; Farrington, J. W.; Gagosian, R. B.; Lee,

C.; DeBaar, H.; Nigrelli, G. E.; Tripp, B. W.; Smith, S. 0.; Frew, N. M. Nature (London) 1980,286, 798-800. ( 6 ) Prahl, F. G. Ph.D. Dissertation, University of Washington, Seattle, WA, 1980. (7) Brassel, S. C.; Eglinton, G. Pergamon Ser. Enuiron. Sci. 1980, 3, 1-22. (8) Thompson, S.; Eglinton, G. Geochim. Cosmochim. Acta 1978, 42, 199-207. (9) Barrick, R. C.; Hedges, J. I. Geochim. Cosmochim. Acta 1981,45, 381-392. (10) Barrick, R. C. Enuiron. Sci. Technol. 1982, 16, 682-692. (11) Ebbesmeyer, C. C.; Barnes, C. A. Estuarine Coastal Mar. Sci. 1980, 11, 311-330. (12) Curl, H. C., Jr. Office of Marine Pollution Assessment Annual Report, Seattle, WA, 1981, pp 1-228. (13) Dexter. R. N.: Anderson. D. E.: Quinlan. E. A.: Goldstein, L. S.; Strickland, R. M.;’Pavlou,‘S. P.; Clayton, J. R., Jr.; Kocan, R. M.; Landolt, M. NOAA Technical Memorandum, OMPA-13, NOAA/OMPA, Boulder, CO, 1981, pp 1-435. (14) Wang, P.; Ph.D. Dissertation, University of Washington, Seattle, 1955. (15) Hardy, J. T.; Crecelius, E. A. Enuiron. Sci. Technol. 1981, 15, 1103-1105. (16) Baker, E. T.; Milburn, H. Continental Shelf Res. 1983,1, 425-435. (17) Gardner, W. D. J. Mar. Res. 1980, 38, 41-52. (18) Hamilton, S. E.; Bates, T. S.; Cline, J. D. Enuiron. Sci. Technol. 1984, 18, 72-79.

“Housing Characteristics for States, Cities, and Counties”; U.S. Government Printing Office: Washington, DC; Vol. 1, Part 49, Washington. (28) Laflamme, R. E.; Hites, R. A. Geochim. Cosmochim.Acta 1978,42, 289-303. (29) Lee, R. F.; Gardner, W. S.; Anderson, J. W.; Blaylock, J. W.; Barwell-Clarke, J. Enuiron. Sci. Technol. 1978, 12, 832-838. (30) Bennett,J. T. Ph.D. Dissertation,University of Washington, Seattle, WA, 1980.

~I

Received for review February 28,1983. Accepted September 26, 1983. This work has been supported by the National Oceanic and Atmospheric Administration’s Section 202 Long-Range Effects Research Program. ContributionNo. 577 from the Pacific Marine Environmental Laboratory.

Identification of Species in a Wet Flue Gas Desulfurization and Denitrification System by Laser Raman Spectroscopy David Littiejohn and Shih-Ger Chang” Lawrence Berkeley Laboratory, University of California, Berkeley, California 94720

Raman spectra have been obtained for a number of intermediates and products of the reaction of nitrite ion with bisulfite ion including hydroxylaminedisulfonate, hydroxylaminemonosulfonate, hydroxylamine, aminetrisulfonate, aminedisulfonate, and sulfamate ions. This is the first direct observation of all stable species in this reaction system. The dynamics of these species have been investigated at several nitrite and bisulfite concentrations that resemble those found in realistic power plant flue gas wet scrubbers.

Introduction Power plant flue gases typically contain hundreds or thousands of parts per million (ppm) of SO2 and several hundred ppm of NO,, mostly in the form of NO. Several wet processes (1,2)for simultaneous removal of SO2 and NO, from power plant flue gases are based on injecting a gas-phase oxidant such as O3into the flue gas to selectively oxidize the relatively insoluble NO to the more soluble NO2. The flue gas is subsequently passed to a N02/SOz absorber. The major fraction of the absorbed NO, in this type of processes has been found to be in the form of nitrogen-sulfur complexes ( I - I O ) , which are the compounds produced in the reaction between nitrite and bisulfite ions.

Some of the reactions that can occur in this system are illustrated in Figure 1 (2-4). The intermediate species nitrososulfonic acid (H0,SNO) has never been identified, but its existence ( I O ) has been inferred from the behavior of the nitrite ion-bisulfite ion reaction. The abbreviations used in the figure are HADS for hydroxylaminedisulfonate, HAMS for hydroxylaminemonosulfonate, HA for hydroxylamine, ATS for aminetrisulfonate, ADS for aminedisulfonate, and SA for sulfamic acid. Wet simultaneous desulfurization and denitrification processes have been demonstrated to be highly efficient in SO2 and NO, removal (1,2); however, these processes have not reached the commercial stage because they are uncompetitive economically, based on the knowledge and designs available a t present. One of the problems in studying this type of process to determine an optimum design and operating condition of scrubbers is the difficulty in quantitatively detecting all the species present under a wide range of solution conditions. This paper demonstrates that laser Raman spectrometry (LRS) can be a promising technique for chemical analysis of species involved in this system. The advantages of LRS include a rapid and simple analysis procedure, simultaneous and unambiguous identification of a large number of species in solutions, and the ability to repeatedly study a solution over a period of time to observe change in concentration of species involved in the system. Unlike infrared spectra,

0013-936X/84/09 18-0305$01.50/0 0 1984 American Chemical Society

Environ. Sci. Technol., Voi. 18, No. 5, 1984

305

HSOHSOHSOHW HO SKO.>(HO s ) NOH&(HO si N Nitrous'acid N i t r O s ~ S U l f o n i C Hydro.&l$mine Amine t ? i s a l f o n i c acid disulfonic acid acid

3 .

NO

+

HSC4

species

HSC;

Hun HSn....l HNC2 HSC; \ + H Vv cC (HO S)NHOH (HC S ) N H O H ~ N H NH(HS0 ( H Sl aa O Hyp~i'izd~t, H yyd rda xr yal axmyi nle a m i n e Amins ,acid monasulfonic a c i d disulfonic acid \

Hqy " Y"i:\ .-

HHhC3 NC3 N2

+

+

H 0 H20

Table I. Raman Shifts and Relative Molar Intensities of Species Studied in This Investigation

N*O/H\

HSCi !H*

N*O

HSCi

N H OH Hydro?iylamine

NO,-

+

\

raman shift, relative molar cm-' intensitya

HSOq /Ht

NO,N,O,,-

HSOq

e2 so,,-

(NH );HS02!id *Suliamic

N,O

NH+ + H S O ~

N~

+

HSO~

Flgure 1. Reactions that occur in the nitrogen oxyacid-sulfur oxyacid system.

HS0,S0,Z-

s,o,z-

water does not interfere significantly in Raman spectra in the region where the compounds of interest have peaks.

HS0,c10,HADS

Experimental Section

HAMS

The Raman system used in this study utilized a Coherent CR-2 argon ion laser and a Spectra Physics 165 krypton ion laser as excitation sources and a Jobin-Yvon Ramanor HG-2 double monochromator. Signals from a cooled RCA 31034 photomultiplier tube were digitized by Princeton Applied Research 1140A photometer and sent to a Nicolet 1280 Raman data processing system. Spectra were generally obtained by using the 488.0 nm argon ion line as the exciting light. Samples were illuminated by a single pass of the laser beam. The laser power on the sample was typically 0.3 W at 488.0 nm, 0.4 W at 514.5 nm, and 0.1 W a t 647.1 nm. It was necessary to synthesize HADS, HAMS, ATS, and ADS in order to obtain reference spectra of these compounds. HADS was synthesized from nitrite ion and bisulfite ion by the method of Rollefson and Oldershaw (11). HAMS was prepared by the hydrolysis of HADS using Oblath's (10) modification of See1 and Degener's method (12). ATS and ADS were synthesized from nitrite ion and bisulfite ion by the procedures given by Sisler and Audrieth (13). The prepared compounds were stored in a vacuum desiccator over P206at 5 OC until use. All other chemicals used were standard reagent-grade compounds. Samples containing bisulfite and sulfite ions were prepared by degassing the solution to which the sodium sulfite or sodium metabisulfite was to be added on vacuum line, adding the salt under an argon atmosphere, and again degassing the solution after addition. This was done to prevent oxidation of the ions into sulfate ion. The solution was then transferred to a 8-mm of diameter Pyrex tube under an argon atmosphere and sealed off. The sample used for obtaining the reference spectrum of NzO was prepared by degassing the solution on a vacuum line and adding 1atm of NzO over the solution. The solution was allowed to equilibrate with the N20 for about 2 h and then transferred to a Pyrex tube and sealed off. Because of the limited solubility of NzO in water, the NzO concentration in the sample was 0.025 M. Quantitative measurements (14)by LRS can be achieved by including a known amount of a reference compound in the samples so the other compounds can be compared with the reference compound. The reference also allows correction of the indicated Raman shift for any system errors. Sulfate ion was used in the samples with which the relative molar scattering intensities were determined. (Relative molar intensity = (ion peak height/ion molarity)/(981-cm-l sulfate ion peak height/sulfate ion molarity).) Peak heights were used as a measure of the concentration of 306

Environ. Sci. Technol., Voi. 18, No. 5, 1984

HA (pH < 7) HA (PH > 9) ATSADS SA (pH 3 )

818 1240 1331 1050 69Zb 1115b 1383b 1285 967 1023 -455 981 1055 1050 935 700 1084 -420 760 1058 1004 918 1097 1084 1063 1049

-

a SO,*- 981-cm-' line = 1.000. (15). C N o value obtained,

0.053 -0.025 0.125 0.95 weakb weakb strongb -0.18 0.12 0.10 -0.07 1.00 -0.9 0.05 1.06 -0.20 1.43 -0.13 -0.08 0.48 0.21 0.09 0.10 0.056 C

0.41

Rauch and Decius

species in solution in this study.

Results and Discussion The Raman shifts and relative molar intensities for the compounds investigated are given in Table I. The hyponitrite ion (Nz022-)was not studied, but Raman shifts obtained by Rauch and Decius (15)are given. The Raman shifts and relative molar intensities of nitrate ion, nitrite ion, and sulfate ion are in good agreement with those obtained by Miller (16) and Marston (17). The Raman shifts for ATS, ADS, SA, and HA obtained are in good agreement with those obtained by other workers (18-21). The strong Raman lines for ATS, ADS, and SA between 1040 and 1100 cm-l have been assigned to the S-0 stretch (18,19). While HADS and HAMS have not been previously studied by Raman spectroscopy, the similarity of their structures to ADS and SA, respectively, indicates that the Raman line due to the S-0 stretch in these compounds should have similar Raman shifts. We attribute the lines a t 1084 cm-l for HADS and 1058 cm-l for HAMS to the S-0 stretch. Reference Raman spectra of HADS, HAMS, HA, ATS, and SA are shown in Figure 2. The strong line at 980 cm-1 is the sulfate ion reference line and is off scale in all the spectra. The HAMS and HA spectra were obtained with one scan, the HADS and SA spectra were obtained with two scans, and the ATS and ADS spectra were obtained with four scans. All spectra taken in this study were obtained with a scan speed of 5 cm-I/s and a slit setting giving a resolution of 4.5 cm-l. The concentration of ATS and ADS were both 0.013 M for these spectra. The low concentrations are due to the limited solubilities of the potassium salts of these compounds. No value for the relative molar intensity for sulfamic acid (SA) was obtained a t very low pH since no work was planned for those conditions. Knowledge of the Raman spectra of the potential product compounds of the reaction between nitrite and

I/

h

h

A)

1300

1200

1100

1000

900

BOO

HAYS

t d

II

d

I\

ADS

h I;

SA

1200

1300

I100

1000

900

1900

1200

1100

1000

900

800

1 so0

1200

1100

1000

900

BW

800

R4MAN SHIFT ( C Y - ' )

Flgure 2. Raman shifts obtained for some nitrogen-sulfur compounds. A sulfate ion peak at 981 cm-' is used as a reference.

bisulfite ion allowed us to determine the concentrations of intermediates and products as a function of reaction time for various reaction conditions. Mixtures of various ratios of nitrite ion and sulfite ion were prepared, and the Raman spectra of the mixtures were obtained. The initial concentrations of the ions used were (A) 0.10 M NOz- + 0.50 M HSO,, (B) 0.26 M NOS- + 0.50 M HSO,, (C) 0.51 M NOz- 0.50 M HS03-, and (D) 1.02 M NOz- + 0.50 M HSO,. Spectra of the four solutions taken about 24 h after mixing are shown in Figure 3. All of the spectra represent four scans of the solutions. The spectrum of mixture A in Figure 3A shows the presence of sulfite ion (S032-)a t 970 cm-l as a shoulder on the 980-cm-l peak, sulfate ion at 980 cm-l, bisulfite ion (HS03-) a t 1025 cm-l, disulfite ion (Sz0,2-) a t 1050 cm-l, HADS a t 1085 cm-l, and ATS a t 1100 cm-l. Since bisulfite ion and disulfite ion exist in equilibrium in solution, the 1 0 5 0 - ~ m peak - ~ is attributed to disulfite ion, and not to nitrate ion. The spectrum of mixture B, shown in Figure 3B, is less complicated, with sulfite ion a t 970 cm-l, HADS a t 1085 cm-', and ATS a t 1100 cm-'. Sulfate ion is much less prominent in the spectrum of mixture B. The spectra of mixtures C and D are similar to that of mixture B, except for increasingly strong nitrite bands a t 820 and 1240 cm-l and a increasingly weak ATS band. The 1240-cm-l band of nitrite ion is broad and overlaps with the 1330-cm-l nitrite band. Parts A and B of Figure 4 illustrate the spectra of mixtures of HADS plus sulfate ion and HAMS plus sulfate ion, respectively, over a range of 500-2500 cm-l. The HADS peak a t 700 cm-l and HAMS peak at 760 cm-l are useful in determining the presence of HADS and HAMS

+

RAMAN SHIFT

(CV-')

Flgure 3. Raman spectra of nltrite ion-bisulfite ion mixtures taken 24 h after mixing. (A) 0.10 M NO2- 4- 0.50 M HS03-, (B) 0.26 M NOz0.50M HS03-, (C) 0.51 M NOz- 0.50 M HS03-, and (D) 1.02 M NOz- 0.50 M HS03-.

+

+

when compounds with peaks that overlap the other peaks may be present. Parts C and D of Figure 4 are spectra of mixtures B and C over the range 500-2500 cm-l. Figure 4A represents two scans, Figure 4B represents one scan, and parts C and D of Figure 4 represent four scans. Those confirm that the 1085-~m-~ peak is due to HADS, and not ADS, by the presence of the 7 0 0 - ~ m -HADS ~ peak. No unassigned peaks were observed in any of the spectra over the range 500-2500 cm-l. The peak a t 1650 cm-l is due to water. The b a n spectrum of water does not interfere appreciably in the 600-2500-cm-' region, although it does have a strong band over the range of approximately 3000-3500 cm-'. From the spectra, it can be seen that the pH has increased as a result of the reaction between bisulfite ion and nitrite ion. In Figure 3A, some of the bisulfite ion has been converted to sulfite ion. In figure 3B-D, all the bisulfite ion has been converted into sulfite ion. This is due to the reaction producing HADS: NO2- + 2HS03(-0$)2NOH OH-

-

+

The magnitude of the change in pH is proportional to the Environ. Sci. Technol., Vol. 18, No. 5. 1984 307

I=

A

so4

ll

I\ ~,,,.7

1 I\ R

'I

1?.1.,,

Flgure 5. Raman spectrum of a 1.02 M nitrite ion-0.50 M sulfite ion mixture taken 1 week after mixing.

so4

R A W N SIiIF'?

. ~

(CM-')

Flgure 4. Raman Specba of (A) HADS + SO,'-, ( 6 )HAMS + SOZ-, (C) 0.26 M NO; + 0.50 M HSO,. and (D) 0.51 M NO2- + 0.50 M HSO;.

nitrite ion concentration when bisulfite ion is in excess, since nitrite ion will control the amount of hydroxyl ion produced. The occurrence of the hydrolysis reactions will cause a decrease in pH. For example, the hydrolysis of HADS and ATS produce Hf: (-03S)2NOH + H 2 0 (-0,S)NHOH + SO>- + H+ (-O,S),N

+ H20

--

(-0,s)"

+ SO," + H+

The hydrolysis reactions in this reaction scheme are all hydrogen ion catalyzed. The increase of pH due to the reaction between nitrite ion and bisulfite ion was c o n f i e d by monitoring the pH of the solution after mixing the reactants together. Prior to mixing, the bisulfite ion solution was pH 4.1. After the addition of nitrite ion, the solution p H rapidly increased to pH 7.0 or higher. Both the rate of pH increase and the amount of pH increase were proportional to the initial concentration of nitrite ion at a given bisulfite ion concentration. Over a longer period of time, the p H slowly decreased, presumably due to the hydrolysis reactions listed above. As shown in Figure 3, the nitrite ion/hisulfite ion ratio affects the products observed, as would he predicted from the reaction scheme in Figure 1. The HN02:HS03- molar 308 Environ.

Sci. Technol., VoI. 18, No. 5, 1984

ratio for production of HADS is 1:2 in the absence of interfering reactions. When the ratio is higher than this, as is the case of mixtures B, C, and D, most of the bisulfite ion will he consumed in the reaction with nitrite ion, so little will he available for conversion of HADS to ATS. As the nitrite ion concentration increases, it will compete more effectively with HADS for any remaining HSO,, resulting in lower ATS concentrations. Since the initial nitrite ion concentration affects the solution pH, the higher nitrite ion concentration mixtures will become basic more rapidly than lower concentration mixtures. The high pH will result in less hydrolysis of HADS and ATS and relatively lower concentrations of hydrolysis products in the higher nitrite ion concentration mixtures. The compounds observed in spectra obtained from mixtures B, C, and D are what would he expected from the reactions in Figure 1. No other reactions appear to he occurring. No unexpected peaks have been observed. Nitrososulfonic acid would not he expected to he present, since the spectra were not taken immedately after mixing, and nitrososulfonic acid should react fairly quickly under the given conditions. Mixtures B, C, and D all show the presence of sulfite ion and nitrite ion in solution 24 h after mixing, indicating the reaction between them and the reaction between HADS and sulfite ion is fairly slow. The was confirmed hy preparing a solution of 1.02 M nitrite ion and 0.50 M sulfite ion at pH 10. A Raman spectrum obtained shortly after mixing showed no evidence of any reaction products, and a spectrum taken 1week later (Figure 5) showed no change. From this, we conclude that the reaction hetween sulfite ion and nitrite ion o c c w very slowly, if at all. This is consistent with the reaction mechanism proposed by Ohlath et al. (7) and by See1 and Knorre (22) involving bisulfite ion for HADS and ATS production. In mixture A, the nitrite ion:hisulfite ion ratio is much less than 1:2. Under these conditions, the kinetics of the system will he considerably different than in mixtures B, C, and D. In this situation, the amount of HADS produced should he limited by the amount of nitrite ion present. Since bisulfite ion should remain after the nitrite ion has been consumed, it can react with HADS to produce ATS. There should he less hydroxyl ion produced than in the other mixtures, so that the hydrogen ion catalyzed hydrolysis reactions should occur more rapidly. The relative amount of ATS in Figure 3A is much larger than in the other spectra. Sulfate ion is produced only by the hydrolysis reactions, according to the reaction scheme in Figure 1. The sulfate ion peak at 980 em-' is much more prominent in Figure 3A than in the other figures, indicating more hydrolysis has occurred. Since no HAMS is apparent a t 1060 cm-', the sulfate ion must he due to the hydrolysis of ATS to ADS. The 1085-cm-' peak is therefore due to both ADS and HADS. The spectra obtained of mixtures B, C, and D did not

h A)

1300

1200

1100

1000

900

1200

1100

1000

900

R)

1300

c

II

Jh

A

0)

l

1300

'

.

*

'

l

1200

~

"

.

I

'

1100

l

l

l

'

-

1000

'

*

I

900

R A W S SYIFT (CM'l)

Figure 6. Raman spectra of 0.10 M nitrite ion-0.50 M bisulfite Ion mixture. (A) 20 mln after mixing, (B) 2 h after mixing, (C) 24 h after mixlng, and (D) 105 days after mixing.

display much change with time. There appeared to be a small reduction in ATS and HADS peak heights relative to nitrite ion, probably due to hydrolysis or, in the case of HADS,a reaction with sulfite ion. The spectra obtained of mixture A with time changed considerably. The change with time is shown in Figure 6, over the range 900-1300 cm-l. The times after mixing and the number of scans for the spectra are the following: Figure 6A, 20 min and one scan; Figure 6B, 2 h and one scan; Figure 6C, 24 h and four scans; Figure 6D, 105 days and four scans. In the first spectrum in Figure 6A, HADS is prominent at 1085 cm-l. The presence of ATS is indicated by the peak at 1100 cm-l, and HSOs- and S2052-are visible a t 1025 and 1050 cm-'. There is weak evidence for SO3%and SO4" at 960 and 980 cm-l. In Figure 6B, the ATS and sulfate ion peaks have become more prominent. In Figure 6C, the ATS peak is very prominent, the HADS/ADS peak is smaller, the sulfate ion peak is larger, and the sulfite ion peak is still present. In the last spectrum in Figure 6D, the peak due to ATS is absent, the 1085-cm-l peak is due solely to ADS (because of the absence of the 700-cm-l HADS peak), and the 970-cm-' S032-peak is absent. The chemistry that is occurring in the reaction mixture of 0.1 M nitrite ion and 0.5 M bisulfite ion is consistent with the reaction scheme shown in Figure 1. The reaction to produce HADS is fairly rapid, and it was essentially completed by the time the first spectrum was taken. The conversion of HADS to ATS has started as well. The

second spectrum shows that further conversion of HADS to ATS has occurred. Hydrolysis is also occurring, evidenced by the growth of the sulfate ion peak at 980 cm-l. The size of the sulfite ion peak relative to the bisulfite ion peak has increased from Figure 6A, indicating the solution has become more basic from the production of HADS. In Figure 6C, much of nitrogen from the nitrite ion is in the form of ATS, with the rest as ADS and a little HADS. Considerable hydrolysis of ATS has occurred, shown by the growth of the sulfate ion peak. The solution is a t a higher pH than it was originally, as indicated by the presence of sulfite ion as well as bisulfite ion. In the last spectrum, all the ATS has been hydrolyzed to ADS, and hydrolysis of ADS to SA is beginning. The hydrolysis rate of ADS is much slower than that of ATS, so evidence of its occurrence will be much slower in appearing. No reference compound was included in the nitrite ion-bisulfite ion mixture to act as a source of calibration for the other peaks because of potential interference with either the reaction system or overlap with other peaks in the spectra. One potentially useful reference compound is C104-, with a peak a t 935 cm-l; HA is the only species of the observed compounds that reacted appreciably with C104-. Since HA is produced slowly, the C104- concentration should remain at a nearly constant level during the period when the chemistry of interest is occurring. However, the reactions that are involved in the system of nitrite ion-bisulfite ion mixtures are fairly well understood, so that estimates of the concentrations of the species present can be made, knowing the initial reactant concentrations and the reactions that occur. For example, the estimated concentrations of the species in Figure 3 can be obtained from the spectra. In Figure 3A, the approximate concentrations are (ATS) = 0.08 M, (ADS) = 0.01 M, (HADS) 0.01 M, (HSOs-) = 0.05 M, (SO4") = 0.01 M, and @Os2-)= 0.05 M. In Figure 3B, the approximate concentrations are (ATS) = 0.1 M, (HADS) = 0.05 M, (SO(-) = 0.1 M, and (NO;) = 0.3 M. In Figure 3D the approximate concentrations are (ATS) = 0.06 M, (HADS) = 0.1 M, (Sot-) 0.15 M, and (NO;) = 0.8 M. Understanding of the kinetics and intermediates of this system is needed for the development of wet scrubber systems to remove NO, and SO2from flue gases. Previous work on identification of species present in this system have been done by wet analytical techniques (23),which are slow, tedious, and susceptible to interference from other species in solution. Use of LRS permits relatively rapid, simultaneous determination of the species present. I t allows repeated measurement of a sample over time, which can provide useful kinetic information. This study provided simple determination of all the reaction products present. We have confirmed the accuracy of the reaction scheme shown in Figure 1. These results are of great value in developing and improving wet flue gas scrubber systems. Our work indicates that for optimum scrubber operation, the S02:N0, ratio in the flue gases should be large enough so that the ratio of bisulfite ion to nitrite ion in solution is greater than 2. The results also show that the pH of the scrubbing solution should be mildly acidic, so that bisulfite ion is the dominant form of sulfur oxyanion in solution.

-

Acknowledgments We thank Leo Brewer, Robert Connick, and Scott Lynn for helpful discussions, and we appreciate the support and encouragement of Earl Evans, Joseph Strakey, and John Williams. Registry No. NOz-, 14797-65-0; NO3-, 14797-55-8; NzO, 10024-97-2;S032-, 14265-45-3;HS03-, 15181-46-1;S04z-,14808Environ. Sci. Technoi., Voi. 18, No. 5, 1984

308

Environ. Scl. Technol. 1984, 18, 310-318

79-8; SZOs*-, 23134-05-6; HS04-, 14996-02-2;SOP, 7446-09-5; NO,, 11104-93-1; HzO, 7732-18-5; HADS, 36324-19-3; HAMS, 1826517-3; ATS, 72198-01-7;ADS, 64647-46-7; SA, 15853-39-1; HA, 7803-49-8.

Literature Cited Martin, A. E. “Emission Control Technology for Industrial Boilers”;Noyes Data Corp.: Park Ridge, NJ, 1981. Chang, S. G.; Littleiohn,. D.;. Lin, N. H. ACS SvmD. ” . Ser. ,

i 9 8 2 , ~ 0188,127-i52. .

Raschig, F. “Schwefel und Stickstoffstudien“; Verlag Chemie: Berlin, 1924. Latimer, W.; Hildebrand, J. H. “ReferenceBook of Inorganic Chemistry”; Macmillan: New York, 1951; p 208. Oblath, S. B.; Markowitz, S. S.; Novakov, T.; Chang, S. G. J. Phys. Chem. 1982,86,4853-4857. Chang, S. G.; Toossi, R.; Novakov, T. Atmos. Environ. 1981, 15, 1287-1292.

(12) Seel, V. F.; Degener, E. Z. Anorg. Allg. Chem. 1956,284, 101-130. (13) Sisler, H.; Audrieth, L. F. J. Am. Chem. SOC.1938, 60, 1947-1948. (14) Irish, D. E.; Chen, H. Appl. Spectrosc. 1971, 25, 1-6. (15) Rauch, J. E.; Decius, J. C. Spectrochim.Acta 1966,22,1963. (16) Miller, A. G. Anal. Chem. 1977, 49, 2044-2048. (17) Marston, A. L. Nucl. Technol. 1975,25, 576-579. (18) Hall, J. R.; Johnson, R. A.; Shurvell, H. F. J. Raman SDectrosc. 1979. 8. 145-150. (19) Hall, J. R.; Johnson, R. A. J.Raman Spectrosc. 1981,11, 279-287. (20) Raj, A. S.; Muthusubramanian, P.; Krishnamurthy, N. J. Raman Spectrosc. 1981, 11, 127-130. (21) Krishnan, R. S.; Balasubramanian, K. Proc. Indian Acad. Sci., Sect. A 1964,59A, 285-291. (22) Seel, V. F.; Knorre, H. Z. Anorg. Allg. Chem. 1956, 284, 70-89. (23) Sato, T.; Matani, S.; Okabe, T. “Abstracts of Papers”;

Oblath, S. B.; Markowitz, S. S.; Novakov, T.; Chang, S. G.

American Chemical Society/Chemical Society of Japan Chemical Congress, Honolulu, HI, April 1979; American Chemical Society: Washington, DC; Abstr. INDE 210.

J. Phys. Chem. 1981,85, 1017-1021. Gomiscek, S.; Clem, R.; Novakov, T.; Chang, S. G. J.Phys. Chem. 1981,85, 2567-2569.

Oblath, S. B.; Markowitz, S. S.; Novakov, T.; Chang, S. G. Inorg. Chem. 1983, 22, 579-583. Oblath, S. B., Ph.D. Dissertation, University of California, Berkeley, CA, 1981. Rollefson, G. K.; Oldershaw, C. F. J. Am. Chem. SOC.1932, 54, 977-979.

Received for review March 7,1983. Accepted November 10,1983. This work was supported by the Assistant Secretary for Fossil Energy, Officeof Coal Research, Advanced Environment Control Division of the US.Department of Energy, under Contract DE-AC03-76SF00098 through the Pittsburgh Energy Technology Center, Pittsburgh, PA.

Trace Organic Compounds in Rain. 1. Sampler Design and Analysis by Adsorption/Thermal Desorption (ATD) James F. Pankow,” Lorne M. Isabelle, and Wllllam E. Asher Department of Environmental Science, Oregon Graduate Center, Beaverton, Oregon 97006

The design and use of a rain sampler with a 0.89-m2 collection surface area are described. The sampler is controlled electronically, provides for the in situ filtration of the sample, and carries out the preconcentration of nonpolar organic compounds by means of cartridges of the sorbent Tenax-GC. The possibility exists for including cartridges of ion-exchanging resin in the sampling train to provide for the preconcentration of organic acids. Analytical results were obtained for 27 compounds by fused silica capillary column gas chromatography with detection by mass spectrometry for four rain events sampled 12 km southwest of Portland, OR, at the Oregon Graduate Center (OGC), and for five rain events sampled in southeast Portland. Mean dissolved rain concentrations for a-hexachlorocyclohexane (a-HCH), naphthalene, acenaphthylene, fluorene, and phenanthrene were 5.9,11,4.7, 3.2, and 24 ng/L, respectively, at OGC. The mean values for the Portland events were 47,72,55,43, and 140 ng/L, respectively. Since dissolved concentrations were measured, the data were also used in conjunction with available Henry’s law constants to estimate the concurrent, local atmospheric levels of these compounds at these sites. Many of the H values of interest are available only near T = 298 K. Therefore, the further understanding of the wet deposition of toxic organic compounds will be facilitated by the direct study of wet deposition as well as by the determination of the temperature dependence of the H values of environmentally interesting compounds. W

Introduction A main accomplishment in the atmospheric sciences during the 1970s has been the development of an appre310

Environ. Sci. Technol., Vol.

18, No. 5, 1984

ciation of the large physical scale of many environmental contamination phenomena. Gaseous SOz emitted at one point may be oxidized in transit to an acidic aerosol and deposited as acidic precipitation hundreds of kilometers from its point of generation (1). Similarly, organic compounds such as the polychlorinated biphenyls (PCBs), the polycyclic aromatic hydrocarbons (PAHs), and pesticides such as the hexachlorocyclohexanes (HCHs) are known to have become distributed throughout the global atmospheric environment (2-7). The scale of this contamination requires the acquisition of a much better understanding of the role played by dry and wet deposition in removing such contaminants from the atmosphere. Our interests have included the study of how the gas and particle precipitation scavanging processes control the fate of atmospheric organic compounds. Such studies are heavily dependent upon the availability of good, temperaturedependent, Henry’s law constant data as well as upon the development of suitable sampling and analysis procedures. Considering that many environmentally important organic contaminants in precipitation occur at very low levels (e.g., of the order of 1-100 ng/L (2-7)), the sampling procedure must provide an adequate volume of artifactfree sample. The simplest type of sampler which has been used is just an open collection container (6,8). If such a device is set out a t the beginning of a storm event and removed immediately at the end, the contaminating effects of dry deposition may be minimized. While the simplicity of this type of sampler is attractive (low cost and minimal chances for sampler-related contamination), the openness of the sampler creates possibilities for contamination and losses if the sample is not analyzed immediately. Other

0013-936X/84/0918-0310$01.50/0

0 1984 American Chemical Society