EPA-600/3-76-089
                                             August 1976
REACTION KINETICS OF OZONE WITH SULFUR COMPOUNDS
                        by
     Ronald E. Erickson and Lei and M. Yates
              Chemistry Department
              University of Montana
             Missoula, Montana 59801
              Srant Number R-800655
               Joseph 0. Bufalini

     Gas Kinetics and Photochemistry Branch
   Environmental  Sciences Research Laboratory
  Research Triangle Park, North Carolina 27711
      U.S. ENVIRONMENTAL PROTECTION AGENCY
       OFFICE OF RESEARCH AND DEVELOPMENT
   ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
  RESEARCH TRIANGLE PARK, NORTH CAROLINA  27711

-------
                                 DISCLAIMER
     This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views
and policies of the U.S. Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or recommendation
for use.
                                       ii

-------
                                  ABSTRACT
     This report presents data on the reaction between ozone and several
sulfur compounds which are air pollutants.  The compounds of interest are
sulfur dioxide, dimethyl sulfide, methanethiol and dimethyl disulfide.

     The rate of the reaction between ozone and dissolved sulfur dioxide
is strongly pH dependent.  This is so because sulfite ion reacts extremely
rapidly (second order rate constant four orders of magnitude higher than
that of bi-sulfite ion).  These results suggest that under some conditions
atmospheric oxidation of sulfur dioxide may involve ozone.

     Dimethyl sulfide was found to reaction extremely rapidly with ozone
in the gas phase but reproducible kinetic data were not obtained.

     Stoichiometric and yield data from the reaction between ozone and
methanethiol or dimethyl disulfide in aqueous solution indicate those
reactions to be complex mechanistically, although one product, methane
sulfonic acid is predominant.
                                      iii

-------
                                  CONTENTS


Abstract                                                              111
Tables                                                                 vi
Acknowledgments                                                       vii

I    Conclusions                                                        1
II   Recommendations                                                    2
III  Introduction                                                       3
IV   Experimental Methods                                              14
V    Discussion and Results                                            30
References                                                             58

-------
                                   TABLES
Number                                                                Page
 I     Gaseous Sulfur Loss Balance Sheet for Southern Kraft
         Mill A	  5-6
 II    Products of Ozonolysis of Dlsulfides	   13
 III   Sulfur Solution Preparation	   15
 IV    Buffer Solutions	   16
 V     Mean Rate Constants	   31
 VI    Equations for Calculations of Specific Rate Constants	  35-36
 VII   Specific Rate Constants	   37
 VIII  Rate Constants Ozone Bisulfite (Penkett)	   41
 IX    Ozonation of Thiols:Stoichiometry	   47
 X     Ozonation of Disulfides:Sto1ch1ometry	   53
                                      vi

-------
                              ACKNOWLEDGEMENTS
     Much of the experimental work 1n this report was accomplished by
three graduate students, Robert L. Clark, Robert Moody, and Lawrence
Schmidt, of the University of Montana's chemistry department.   Technical
assistance was also afforded by David McEwen.
                                     vi i

-------
                      SECTION I

                     CONCLUSIONS
     The first significant scientific finding of the research
carried out under this grant is that the rate of ozonation
of sulfur IV in aqueous solution is controlled, almost
exclusively, by the concentration of sulfite ion.  Sulfite
reacts about four orders of magnitude faster than bisulfite
ion and at least two million times faster than aqueous sulfur
dioxide.
     However it is the absolute magnitude of the specific
rate constant for the ozone-sulfite ion reaction which is of
significance to the possible role which ozone might play in
the atmospheric oxidation of sulfur dioxide.  Our calculations
show that under certain specified conditions, the most crucial
of which is the presence of liquid water, ozone may play an
important role in the atmospheric phase of the biogeochemical
cycling of sulfur.
     Methane thiol and dimethyl disulfide also react extremely
rapidly with ozone in aqueous solution to yield methane sul-
fonic acid along with several minor products.  The reaction is
more complicated, in a mechanistic sense, than the sulfur IV-
ozone reaction which is a simple second order reaction.
     Gas phase kinetic studies for the reaction between ozone
and dimethyl sulfide show the reaction to be extremely rapid
and complex (mixed order) but too few data were obtained for
meaningful  comment.

-------
                      SECTION II
                     RECOMMENDATIONS

     The rate constants for the reaction  between  ozone  and  sul
fur IV in solution which were obtained  in this  study  should
be considered in any model  for the atmospheric  oxidation  of
sulfur dioxide.
     Since ozone is highly  reactive with  many  subtances,
including all negative ions which  are oxidizable,  and many
positive ions in their lower oxidation  states,  it  would
be necessary to  know both the concentrations  of such  ions
in atmospheric water and their rates  of reaction  with ozone
in order to establish a reasonable model.  Such data  are  not
available.

-------
                      SECTION  III

                      INTRODUCTION
GENERAL
     Our original goal was to measure the rates of the
reaction of ozone with sulfur dioxide and several malodorous
sulfur compounds (mercaptans, sulfides, disulfides and
hydrogen sulfide).  Determinations were to be made in the
gas phase for dimethyl sulfide and methyl mercaptan and in
water for sulfur dioxide, methyl mercaptan and hydrogen
sulfide. Several serious experimental problems and a shortening
of the project period (from three to two funded years)
resulted in a more modest set of achievements.  We report
here on a complete kinetic study of sulfur dioxide in aqueous
solutions, on stoichiometric and product data of dimethyl
mercaptan with ozone in water, and on the gas phase reaction
of dimethyl sulfide with ozone.

SULFUR DIOXIDE-OZONE
     The importance of sulfur dioxide in air pollution
episodes is well documented.   Recent reviews have covered
                   1234
the subject so well    '  '  that only a few references which
are directly pertinent to the possible reaction of sulfur
dioxide with ozone will  be discussed here.
     The reaction of ozone with sulfur dioxide has sparked
some debate.   Gregor and Martin's finding of sulfuric acid
in the oxidation of hydrogen  sulfide  (presumably through
the further oxidation of sulfur dioxide) was not confirmed
by Cadle when both reactants  were in low concentration in
the gas phase .   Pitts recommended in the "Technical  Report
on Air Quality Criteria  for Photochemical Oxidants" by the
State of California Department of Public Health that the
ozone-sulfur dioxide reaction be studied in the presence of
water .   Cadle suggested that such a study would be fruitless

-------
but did comment that a study of the ozonation of sulfurous  acid
                                    o
in fog droplets might be appropriate .
     One very interesting study of the  reaction of ozone with
                                                    g
sulfur dioxide was carried out by Espenson and Taube .   Equation
1 describes the general  stoi chiometry of the reaction in water
while the most intriguing discovery was that the sulfate
formed contains up to two oxygen atoms  from ozone.
                           H?0
equation (1)     S0  + 0    *   HSO   +  H  +  0
                   2    3                       2

     While the work described in this report was in progress,
two important papers were published on possible interactions
between ozone, sulfur dioxide,  and moisture.  In the first,
Cox and Pankett observed that both sulfur dioxide and ozone
decay on the walls of containers, the rate increasing rapidly
                                                        te
                                                        11
with increasing humidity  .   In the second Pankett reported
the rate of reaction of bisulfite ion with ozone at pH 5
Since his work is directly comparable to ours, we will reserve
discussion of that paper until  later in this report.

DIMETHYL SULFIDE-OZONE
     Dimethyl sulfide is a malodorous, water insoluble sulfide
whose ozone oxidation products, dimethyl sulfoxide and dimethyl
sulfone are relatively odorless and water soluble.  Although the su
fide is not normally found in appreciable amounts in polluted
air, its presence is definitely noticeable where Kraft pulp
mills operate.  Akamatsu's studies show that the recovery
process losses plus blow gas emissions constitute the major
sources although not the only source of a series of mal-
                         1 2
odorous sulfur compounds.    Some idea of kinds of compounds
found, the variability of concentrations, and the extent
of the problem for dimethyl sulfide can be found in Table 1,
taken from Akamatsu.

-------
4
C
V
r
e c
0 «

*W •£

dP f
C
» C
H C
(0 4
i t X
4J ^
orf



o

0)

§
rrt
W
n)
tn
«w
O


o














CJ
o

0
en
D
)
•i
3
)
9

4
J
J
» .
•t
5
> ;


W"
a:

tf !
K


rrt
8:


en
w :



CM
O






















0 OOOH 0 OO 0
o H cs m o vo H r- o
• •••• • •• •
m H *M *M m o H CM o
00 O
H

H nvoon o> CMCO vo
p«. CM vo r* in H moo CM
in OOOH o oo o
CM en
^a1 Hinvo^j* H Hr— i o o
o o o T* ** o oo o r^
o oooo o oo H n
H VOVOC4T H I^H VD O
O O H  M-l O
 H
vo oOrf o oo vo u ro
n) CM
0)
C -H
o  0) to 19 ,.^ H^CW 4JC-H JJU-I
>-(> -HQJflJrt («-4,X>i^4-)4J^4')V!'O>M (QU-IP
>OC 0) £OC 4JOIO
OM(dOJ «J
p4 CJ WJ CD OQ PI fjQ CQ
 04

•8
•O

-rl
 r
 I
ca

-------
C
<
1
e ;
o <
M I
ii t
*W
r<
OP t
«
(
%
rH f
id c
-P 4
0"-
b _f
t^ »•
f


C

iH
to
0)
•_J
i
CO
id
rn
".
«M
O

e
ti
w
o
fa
•'




0)
u
M
2
o
(0
1) '
j ;
•4
3 •
5
Q ;
^ •
3
a •
u ;
•4 '
3
j
5-
•4
•i
K
CO
rn
3



«
w

a
2
r*|


W
CN
jTl
M*
CN
O
(0











o ooooo o o
O» OCMrHfNin m tN
rH O in 00 (N O C4 00
^P ro .




vo rHtfCNinvo a\ n
O4 rHr-OOO CN O
in OOiH^fO O »H
in omino rH o
rH OOrHOO O O
O O O O CN O ' O O



O CNOCOOCN O O
O O O CO T* O O O
O OOOrHO O 0
o r»vomoiH iH CN
m ocNOino o o
o ooooo o o


•H CNTTrHon r- iH
M) OCNOrHO CN O
m ooooo o iH
.0 o in |. i i i. i
CN 0 rH « 1 1 I I
• •ill 1 1
rH 0 0 1 • 1. 1 | I
O* V< I -H
C O -H rH
>i-H ^ -P rH M-i C (fl fi
>-l> 0) id -rH -HOMO
OrHj^ 4J ^O'O'H-P'H
> O C 0) O -H -P 3 -P
O W (d C Q) ^ (XrH O id 0) id
OW4JrH frOOlrHriJO^N
0)-rt ^j.HrH>ld
« Q U5 Q « pq E-«



0
o
o
o
•H



M3
in
•
CM
rH
in o
^r in
CN O\
iH


r* o
CN . rH
CN 00
rH
o o
CN VO
• *
rH en



• •
rH 0
rH
rH
id
4J
OX «W
•P O O \*
rH Id 3
id «w 0) B  -H 
-------
     The first reported ozonation of organic sulfides was in
                 13
1933, when Maneck   ozonized benzyl sulfide in CC1. and
reported intltlal reaction products to be the sulfoxide
and sulfone.  Prolonged ozonation oxidized these initial
products to sulfuric acid, carbon dioxide, and benzoic acid.
     In 1942, Bohme and Fischer   ozonized several organic
sulfides, including dimethyl sulfide, in chloroform.  A
nearly quantitative yield of the sulfone was obtained for
the compounds Me2S, Et2$, (ClCH2CH2)2S, MeS0, EtSCH20,
and (0CH2)2S.  That the sulfoxide was formed as a precursor
to the sulfone was shown by the isolation of benzyl sulfoxide
from the sulfide when an insufficient amount of ozone was
used.   For some sulfides, i.e., chloromethyl ethyl sulfide,
the sulfoxide was the final  product.
     While studying the possibility of the use of ozone as
a titrimetric agent for quantitative determination of
                                                              1 5
olefinic unsaturation in petroleum fractions, Boer and Kooyman
ozonized several sulfides to determine the effect of their
presence in the olefinic mixtures.  The dialkyl sulfides were
reported to consume from ^  to - mole of ozone per mole of
                         i*     <•
sulfur compound.  Benzothiophene consumed exactly one mole
of ozone, while thiophene and dibenzothiophene did not inter-
fere with the reaction.
     In 1954, Bateman and Cunneen   studied the autoxidabi-
1ity of several  monosul fides at temperatures ranging from 45°-
75°C and found that saturated alkyl  and phenyl  sulfides
absorb no oxygen during  24 hours at 75°.   Similarly inert
were the benzyl  and diphenylmethyl substituted sulfides.
     While the benzyl sulfides undergo photo-oxidation, the
dialkyl  sulfides oxidized only when catalyzed by aa1-
azobis-isobutyronitrile.  The diphenyl  sulfides were inert
to all  autoxidation.

-------
     Barnard   found that monosulfide-ozone reactions in solution
at -25°C yielded not only the sulfoxide and sulfone but
other products as well.   For the reaction:

     R1  •  S •  R" + 203    ->•      R" •  S02-   R" + 202

less than the  theoretical amount of ozone was needed for
a nearly quantitative yield of the sulfone.  The ozonation,
run just to completion,  of di-n-butyl  sulfide gave a 100%
yield of the sulfone "smelling of butyraldehyde and butyric
acid and reacting acid to litmus". The reactivity of the
sulfides indicated that  the most easily oxidized sulfides
required the least ozone.
     Since saturated alkyl sulfides and several aryl sulfides
are inert to molecular oxygen without a catalyst,  Barnard
speculated that the ozone must either catalyze the sulfide-
oxygen oxidation or produce, after loss of an oxygen atom,
activated oxygen molecules some of which participate in
the oxidation  of the sulfide or sulfoxide.
     Barnard's ozone uptake curves indicated that oxidation
occurs in two  stages to  give first the sulfoxide and then
the sulfone.  By following the ozonolysis of cyclohexyl
methyl sulfide with infrared spectroscopy,  he was able to
show that at least 98-99% of the sulfide was converted to
the sulfoxide before any sulfone was detected.  This indicates
the rate of sulfoxide formation is 50-100 times faster than
that of the sulfone.
                                      1 8
     More recently, Maggiolo and Blair  found that ozone
reacted with dialkyl and diaryl sulfides according to the
equation:

              °             °

-------
When the reaction was run in a nonpolar solvent, it was
found to be stoichiometric for the aliphatic and aromatic
sulfoxides and for the aliphatic sulfones.
       ..  S  .-  CH2-0
  10. 7g
 (.0499)
   2.33g
  (.0485)
                                CHC
                                20°C

                                             m    .
                                        11. 3g
                                       (.0491)

0-CH
CH2-
                                                CH2-0
 10. 7g
(.0499)
    4.71g
   (.0981)
                                         12. Og
                                        (.0487)
                                           20
                                1 9
     Horner, Schaefer and Ludwig    also found that sulfides
react in ethyl  chloride with ozone according to Maggiolo
and Blair's equation to give sulfoxides and sulfones in 80-
100% yield.
     Six years  later, work by Hughes, McMinn,  and Burleson
lent still  further support to Maggiolo and Blair's stoichiometry
by obtaining stoichiometric yields of sulfoxide and sulfone
when Bis (B-Hydroxyethyl ) sulfide  was ozonized at 100°C.
     Research was carried out in this laboratory on the ozona-
tion of dimethyl  sulfide in methylene chloride at -78°C.
Nuclear magnetic  resonance analysis of aliquots taken during
the reaction revealed dimethyl  sulfone formation after only
1% of the reaction was completed.   While this  would seem to
indicate a  DMSO/0- reaction considerably faster than that

-------
for DMS and ozone, it is also possible,  due to poor mixing
in the reaction vessel,  that pockets  of  high ozone concentra-
tion could occur and that the sulfone resulted from the slow
oxidation of the DMSO trapped in these areas.   Product
analysis indicated an uptake of 1.86  moles of ozone per mole
of sulfide for conversion to the sulfone.
     Not until 1966 was  there any work done on the vapor
phase ozonation of dimethyl  sulfide.   In that year, pollution
                                                          21
control work resulted in two studies  by  Akamatsu,  et.  al.,
on the removal of dimethyl sulfide from  pulping processes
by vapor phase ozonation.  This work  showed that dimethyl
sulfide oxidized with oxygen containing  7% ozone gave
75% dimethyl sulfoxide and 25% dimethyl  sulfone.  Another
                   1 p
project of Akamatsu   showed that an  85% yield of the  sulfoxide
could be obtained when dimethyl sulfide  and oxygen containing
20% ozone were fed into  a reactor in  a 3:7 ratio at 30°C.
          22
     Hales   studied the gas phase kinetics of hydrogen sulfide
with ozone.  The results of that study have only an indirect
bearing on the reaction  of ozone with dimethyl sulfide, but
the methodology seemed to be particularly useful to the
study we planned to carry out.

THIOLS AND DISULFIDES-OZONE
     The first ozone oxidation of a thiol  was reported in  1933
                I O
by Martin Maneck  .  He  ozonized ethanethiol and obtained
ethanesulfonic acid.  The conversion  of thiols to sulfonic
acids has also been observed by several  other authors.
                     23
Asinger and coworkers   used ozonation of thiol to prepare a
number of long chain aliphatic sulfonic  acids in yields from
70 to 100%.  Barnard   reported quantitative conversion of
thiophenol to benzenesolfonic acid.  The conversion of iso-
butanethiol to isobutanesulfonic was  accomplished with a 90%
                 24
yield by Erickson
                         10

-------
     Both Barnard and Erickson have determined the stoichio-
metry involved.  Barnard   found that 1.6 moles of ozone per
mole of thiophenol were absorbed when the reaction was carried
out by adding ozonized oxygen to a CC1, solution of the thiol
                 o                                         24
maintained at -20 C.  Working in methanol at -78°C Erickson
found a stoichiometry of 1.6:1 when an ozone-oxygen stream
was used and a stoichiometry of 3.3:1 when an ozone-nitrogen
stream was used for the conversion of isobutanethiol  to
the corresponding sulfonic acid.
     Disulfides and thiolsulfonates have been reported as minor
                                 13                           24
products of the reaction.   Maneck   found both.  In Erickson's
work the disulfide was formed, but the thiolsulfonate was not
identified.  However, all  the minor products were not identified
Erickson also found methyl isobutanesulfonate, a product
that must have arisen through a reaction with the solvent
                                                        I O
(CHoOH).  An ozonation of thiophenol performed by Maneck
yielded no sulfonic acid.   The only products were the cor-
responding disulfide, thiolsulfonate and disulfone.  Maneck
also noted that continued ozonation of either ethanethiol
or thiolophenol resulted in  an attack of the carbon-sulfur
bonds and the formation of sulfuric acid.
     The mechanism of the ozone-thiol reaction has hardly been
     dered.  Maneck   offe
he gave no proof for them.
considered.   Maneck   offered the following equations, but
     EtSH + 03    -»•      EtS03H                        (22)
     2EtSH +0    •*      EtSSEt + H90                  (23)
                           0       *
     EtSSEt + 20  ->      EtSSEt                        (24)
                           0
At that time thiolsulfonates were believed to have a disulfoxide
structure and the thiolsulfonate was reported as a disulfoxide.
     Two studies of the ozone-thiol reaction in the gas phase
have been carried out.   The use of ozone to control Kraft
                            11

-------
                                               1 2
pulp mill  odors has been considered by Akamatsu  .   He found
that methanethiol  was converted to an odorless, water-soluble
compound.
     A kinetic study of the ozone-ethanethiol  reaction in
the gas phase was  performed by Kirchner,  Kastenhuber and
       25
Biering  .   Concentrations  were in the parts per  million
range and  under those conditions the reaction  proceeded with
carbon-sulfur bond cleavage.   The kinetics were determined
by competitive reactions with hexene and  isobutene.   At
                                    1
30°C a rate constant of 2.4+0.8 X 10  I/mole sec  was found.
     A series of disulfides has been ozonized  by  Barnard .
The reactions were carried  out in carbon  tetrachloride at
-25°C.  Sulfonic anhydrides and thiolsulfonates were the major
products.   Disulfones were  also formed in small quantities.
His results are summarized  in Table II.   Dimethyl and diben-
zyl disulfide were atypical.   Dibenzyl disulfide  reacted
by giving  carbon-sulfur bond  cleavage.  Dimethyl  disulfide
gave abnormal product ratios.  He proposed that the  reaction
proceeds according to the following scheme:
DC CD 	
Ko


J/
RS


°?
3

' ^^
nc.p
~-~-^_
^^.
I
0
RS>1SR
0-0=0
	 *• DC CD
••• Ko-oK — — — — —
"0-0=0

^ 1/2 RS02SR +
-^Q
o pen c P

^-^^ RSOSOR


0
	 j fA C.R
0-0-0
R
n c . C .
	 Ko . O . 	
o-ozo
+ -
1/2 RSSR


3 \pcn en R





                                                     R
                                                  R:
                                                 Kb
';)
                                                     R

                                                     \J
                                                     I
                                                 RS02OS02R
                                                    1
                                12

-------
                             TABLE II
              PRODUCTS OF OZONOLYSIS OF BISULFIDES
                                                  17
R 03 Absorbed
(moles)
Ph 2.6
p-Cl C6H4 2.9
n-Bu 2.6
Me 2.0
CH2Ph 5
Sulfonic Thiolsulfonate
Anhydride (%) (%)
90 6
84 10
80 10
39 50
__
Homer et.  ajk     have ozonized some of the same disulfides as
Barnard.   The reactions were done in ethyl chloride and they
obtained  different results.  Homer obtained an 82% yield
of benzyl  phenylmethanethiolsulfonate from dibenzyl disulfide.
Diphenyl  disulfide also gave its thiolsulfonate, but only in
23% yield.   The product from ozonation  of di-n-butyl disulfide
decomposed  and  was not identified.
                               13

-------
                      SECTION  IV

                  EXPERIMENTAL METHODS
       OZONE-SULFUR DIOXIDE AQUEOUS PHASE KINETICS
INSTRUMENTAL METHOD
     We determined the rate of ozonation of total  sulfur
species present in buffered aqueous solutions of sulfur dioxide
at variable pH by a stopped flow technique.  The stopped
flow system used was an Aminco-Morrow stopped flow
spectrophotometer with associated ultraviolet source and
a Biomation transient recorder Model  610 with a Cathode
Ray Tube display.  The transient recorder storage  was then
fed into a Bausch and Lomb VOM-10 recorder.  This  enabled
the signal collected over a few milliseconds to be recorded
on a 10 second virtual time base for a permanent record,
rather than photographing the CRT display.  The log mode (0-1
Absorbance units on the oscilloscope and transient recorder)
was used on all runs.  Five traces were recorded with each
run and three runs were made with each pair of reactant
solutions.

TEMPERATURE REGULATION
     A large (4 gal.) water bath was used with a cooling coil,
heating element and thermostat.
     For the 16°C runs the temperature was maintained with
ordinary cold tap water circulating in the bath.  The flasks
containing the reactant solutions were brought to  16° C
before a run was made.  Dry air was pushed through the obser-
vation cell area to avoid condensation on the quartz observation
cell.
     Temperatures were noted in the bath, before and after
entering the cooling area.  The fluctuation was never greater
than 0.2°C between the inlet and the outlet of the cooling block,
                           14

-------
SULFUR REACTANT SOLUTION PREPARATION
     The sulfur solutions were made from sodium bisulfite
(NaHS03) at these pH levels:  0.620, 2.50, 3.55, 3.22, 4.02.
Weighing solid sodium bisulfite was also the procedure used
for all four of the 16°C runs; pH:  0.59, 2.13, 2.09, 3.74.
     For the other pH levels (-.301, 1.20, 1.71, 2.11, 2.80,
3.12) S02 gas was bubbled into the buffer solution.

TABLE III SHOWS APPROXIMATE FIGURES OF THE AMOUNT (MLS) NaHS03
ADDED TO A LITER OF BUFFER

                       TABLE III

     pH        Molarity NaHS03          Ml NaHS03

0.620               .1  M                     5 ml

2.50                .1  M                     2 ml

3.55                .01  M                    1 ml

4.02                .001 M                   5 ml

3.22                .001 M                   >1 ml

     The concentration  of sulfur dioxide was determined by
the absorbance of the sulfur solution in the stopped flow
apparatus.   Wave lengths and extinction coefficients were
270 nm (e = 461) and 276 nm (e = 500).   The concentrations
calculated at the two different wavelengths were always
within 3-4% of each other.
     Knowing the sulfur dioxide concentration it was possible
to calculate the concentrations of bisulfite and sulfite ions
by proper manipulation  of the known equilibrium constant
                            15

-------
expressions.

OZONE REACTANT SOLUTION
     Ozone solutions were prepared by bubbling 03 into the
proper buffer with nitrogen from ozone saturated silica
gel  at -78°.   The concentrations were determined at wavelengths
of 250 nm and 276 nm.  Extinction coefficients were 2430 and
2150 respectively.

BUFFER SOLUTIONS
     In the pH range of 2-5 mixture of 0.10m H3P04 and 0.10m
KH2P04 solutions were used.  The actual  pH was found with a
Beckman pH meter using the expanded scale.  The pH meter
was zeroed in with known solutions in the proper pH ranges.

TABLE IV SHOWS THE PERCENTAGE COMPOSITIONS OF THE BUFFER SOLU-
TIONS
                       TABLE IV
                    BUFFER SOLUTIONS
           APPROXIMATE PERCENTAGES (0.1M EACH)

                                        % KH2P04
4                        2                   98

3.5                      5                   95

3                        12                  88

2.5                      50                  50

2                        90                  10
For pH levels less than 2 0.1-1M H2S04 was used,


                         16

-------
       OZONE-DIMETHYL SULFIDE GAS PHASE KINETICS

INSTRUMENTATION
     The stoichiometric data were obtained using a Packard
Model 7300 gas-liquid chromatograph equipped with dual flame
ionization detectors, dual electrometers, and a Vidar 6300
                                                            i
Autolab digital integrator.  The column employed was a 5' X — "
glass 5% GE XE 60 on Chrom Q.  Triple-distilled butyl
benzoate was utilized as an internal standard for determination
of product yields.
     Product indentif ication was achieved by comparison of
product g.l.c. retention times with those of known compounds
under identical conditions and by mass spectroscopy using a
Varian Mat III GS/MS system.

APPARATUS
     A Welsbach T-408 electric discharge ozonator was used
to produce the ozone.  The ozone flow rate was determined
by passing the ozone-oxygen (nitrogen) stream through a
potassium iodide sol ution for a known time and titrating the
solution, on acidification, with a .005 N solution of sodium
thiosul fate.
     The equation for the flow rate is:
               V =  milliliter of NapS-O-, used
               N =  normality of Na^S^Oo used
               t =  time in minutes that 0~ was passed into
                    KI trap
     In this study,  the carrier gas was simply passed through
the Welsbach ozonator, split to decrease the ozone concentra-
tion,  brought to the desired temperature of either 25°C of
                           17

-------
35° C and Introduced Into the mixing chamber.
     The ozone stream was split in order to decrease concen-
trations to those desired.   Attempts to decrease ozone
concentrations to those required by decreasing the current
on the ozonator yielded inconsistent ozone flow rates.
     After the ozone stream was split,  the flow rate of the
resulting stream was measured by a Fischer and Porter
"Tri Flat" rotometer.  The  nitrogen flow rate was determined
by a Gilmont Model R795 rotometer.
     Dimethyl sulfide was introduced into the nitrogen stream
by a .1 ml Hamilton gas-tight syringe drived by a Sage model
341 syringe pump.  Flow rates for the DMS were varied from
.0003 to .00006 ml/min.
DMS-03/02 Reactor System
     Before the DMS was injected, the nitrogen stream was
brought to the desired temperature in a constant-temperature
water bath.  In order to achieve complete DMS vaporization
and to prevent the liquid DMS from splashing on the walls,
the injection chamber was expanded, the diameter of the glass
tubing carrying the nitrogen was decreased and the end was
splayed.  After injection,  both the N2/DMS and the 03/02
streams entered a styrofoam constant temperature cabinet,
passed through I1 and 17' of glass tubing respectively,
came together in the gas mixing chamber and then entered
into the reactor.  The mixing chamber was designed according
                        22
to Male's specifications   and consisted of a Teflon  collar
mounted within a stainless  steel sleeve.  The 03/02 stream
passed into the sleeve and  proceeded into the reactor through
four .5 mm radial holes drilled into the collar.  The boundaries
between the collar and the  sleeve were kept gas-tight using
Vitron "0" rings.  The reactor was constructed of precision-
bore Pyrex tubing having an inside diameter of 1/8" + .0002"
                          18

-------
and had an inside volume of 1.988 mill mters.  The reactor
temperature was controlled by means of a water jacket.
Figure 1 diagrams the reactor system.  Extreme care was taken
in cleaning the reactor and associated glassware.  The pro-
                                                        op
cedure followed for cleaning the reactor was that Male's  .
     The reaction products were collected by means of a freeze-
out trap consisting of a glass spiral connected to an adapted
10 ml pear-shaped two neck flask.  Products were then removed
from the trap by warming to room temperature and washing
three times with three milliliter portions of methylene chloride.
A 1 ml volumetric flask was attached to the bottom of the
flask in order to obtain accurate 1 ml volumes of CI^Clp-
product solutions on evaporation.  The trap was immersed
in a Dewar containing a solid-liquid slush of n-pentane.
Two other solid-liquid slushes were tried; however, tests,
proved that product loss resulted when either the dry ice/
acetone or the  isooctane/1iquid nitrogen slushes were used.
Since the amount of product was small, the 10 ml of methylene
chloride needed to wash out the trap resulted in product con-
centrations too low to be analyzed.  Vapor pressure data for
DMSO and CH2C12 were obtained from Karchmer's   The Analytical
Chemi stry of Sulfur and Its Compounds and from the Handbook
                        27
of Chemistry and^ Physics   respectively.  Extrapolation of
Karchmer's data yielded a vapor pressure for DMSO of less
than .1 mm of Hg at 0°C.  A value of .0896 mm of Hg was obtained
                                p O
using the values of Jose et.  al.     Methylene chloride's
vapor pressure at this temperature was approximately 129 mm
of Hg.  Since methylene chloride's vapor pressure was almost
1500 times that DMSO's, evaporation without product loss
appeared quite feasible as a  means of obtaining product
concentrations which could be analyzed.   This evaporation
was carried out by removing the 10 ml flask from the rest of
                          19

-------

o
<0 4J
o 2 ^
N 0)
° g
O
>1
•H
04
O.
3
CO
a
<

f
[

W 4J
K

^oOOOOiT'


j
[^
j

v
f
J
c
^
I
J
<0 n)
Constant Temperature Freez
Cabinet TC
20

-------
the CHpClp-washed trap and attaching a condenser and a
and a nitrogen source.  This was then placed in an ice bath
and nitrogen was passed over the solution until a volume
of 1 ml was reached.
     In order to determine the ozone uptake of the reaction,
ozone concentrations were measured before and during the
reaction by means of a KI sampler attached to the freeze-
out trap.  It was found that a normal glass frit would hold
varying amounts of I2, causing inaccuracy in the determination
of the Oo concentration.  It was not possible to divert the
ozone stream in order to wash the frit, since this caused
a pressure change and a resulting fluctuation in the ozone
flow rate.  Studies indicated that it took as long as an
hour to restablish the prior CU flow.  This problem was
solved by exchanging the frit with a glass tip consisting
of a tapered end and four side openings for a 5-way dispersion
of the gas.
     Trapping all of the DMS in the freeze-out trap proved
impossible.  Therefore, the DMS entered the KI solution when
the ozone stream was analyzed.  DMS reacts with I9 forming
              26
DMSO and DMSOp   thereby consuming the !« used for ozone con-
centration determination.  Tests determined that adding starch
solution to the KI
          2KI + 03 + H20      -»•         I2 + KOH
          CH3SCH3 + I2        +         CH3SCH3
          CH3SCH3 -I2 + H20   -»•         CH3SCH3  +  2HI
             0               slow
sample while analyzing for ozone eliminated this problem
since the starch complexed immediately with the I2 on its
                          21

-------
formation, thus preventing the reaction with dimethyl  sulfide.
Water was placed in the sampler between KI  analysis  in order
to maintain a constant pressure head.

REAGENTS
Nitrogen
     Before entering the reaction system,  the nitrogen was
further dried and cleaned by passing it through a column
packed with a layer of Drierite and silica  gel  and a layer
     o
of 4 A molecular sieves.
Oxygen
     Tests were run on the purity of the oxygen by passing
the gas through a Drierite-si1ica gel  column.  Extended use
indicated a level of purity and dryness sufficient for this
investigation.
Dimethyl Sulfide
     The Baker grade sulfide was dried over anhydrous calcium
sulfate and triple distilled using a 600 mm Vigreux distilling
column attached to a 400 mm Claisen-Vigreaux distilling head.
Dimethyl Sulfoxide
     The Baker analysed DMSO was stirred at 25°C with 5% of
its weight of Darco G-60, filtered, treated with MgC03 to
remove the acidic impurities, stored overnight over Cal^j and
triple distilled through a 400 mm Claisen-Vigreaus column
at approximately 10 mm of Hg.
Methylene Chloride
     The Mallinckrodt Spectr AR grade methylene chloride was
washed with concentrated sulfuric acid followed by dilute
sodium hydroxide and finally water.  The washed material
was allowed to stand overnight over sodium hydroxide pellets
and calcium chloride and was then triple distilled using a
600 mm Vigreux column attached to a 400 mm Claisen-Vigreux
                          22

-------
distilling head.
Butyl Benzoate
     The Matheson Coleman and Bell butyl benzoate was triple
distilled using a 400 mm Vigreux column.
     All chemicals were analyzed by gas chromatography for
purity and stored under nitrogen until needed.
Cal ibration o_f Equipment
     Calibration of several components of the experimental
system was required before runs were made.   These calibrations
are as follows.
Rotometers
     Both rotometers were calibrated using  soap-film flow
meters.   Hales found that on continued use  the sapphire
rotometer floats picked up static electricity rendering them
useless  for accurate flow indicators.   Therefore, the rotometers
were used only as secondary indicators and  soap-film flow
meters were used for actual gas flow rates.
Syringe  Drive
     The syringe pump output was calibrated for several
settings on ml/hr for a .1 ml Hamilton syringe by measuring
the time required for .01 ml of material to be injected.
Five settings were calibrated for dimethyl  sulfide (see
below).
          Setting on ml/hr                   Flow Rate
               9                   3.2593 x  10~4  ml/min

               8                   2.131  x 10"4  ml/min

DMS            7                   1.359 x  TO"4  ml/min

               6                   9.147 x  io"5  ml/min

               5                   6.222 x  IO"5  ml/min
                        23

-------
Determination of Peak Height vs.  Concentration Ratios
     Gas chromatograph peak height ratios were calculated for
DMSO, dimethyl  sulfone,  and butyl  benzoate.   On a mole basis
these ratios came out to be:
     Butyl  Benzoate                1.0000
     DMSO                           .1359
     DMS02                           .1788

Evaporation   Trap
     A known amount of DMSO was added to 10 ml of methylene
chloride and evaporated  in the trap to a volume of 1  ml.
Direct comparison of standard gas  chromatographic peak areas
with those  from the trap indicated a trap efficiency of 99.0
percent.
0.3/0-2 Flow  Rate
     Since  a normal run  lasted approximately three hours,
tests were  run  to determine the consistency of the ozone flow
over this time  period.  Due to the slow response of the ozona-
tor, a two  hour warm-up  period was allowed before samples were
taken.  Results indicated an average percent deviation of
                     22
+2.40 percent.   Hales   conducted  experiments to determine the
extent of ozone decay and found that at 28.5°C the decay
was less that 1% during  a run.  Since our flow rates were
much higher and the reactor diameter smaller, ozone decay
should present  no problems.

 OZONE-DIMETHYL DISULFIDE AND METHANE THIOL OZONATIONS
ANALYTICAL  METHODS
Ozone
     The standard iodide-thiosulfate method was used to deter-
mine ozone  concentrations.  Although there recently has been
some debate concerning the stoichiometry of the ozone-
                         24

-------
                29-31
iodide reaction,      the following is generally accepted.

          03 + H20 + 21"      •+         02 + 20H" +  I2

The iodine was determined by titration with sodium thiosulfate
as shown.

          2S203- + I          *         21- +  S406-

A 100.0 ml aliquot of an aqueous ozone solution was pipetted
into a potassium iodide solution.  The solution was acidified
with 5 ml of 6M HC1 and titrated with 0.005 M sodium
thiosulfate.  Just before the end point, 3 ml of a 0.3% starch
solution was added and the titration was continued to the
starch end point.
     The potassium iodide used was Baker Analyzed and Mallin-
ckrodt AR; both contained less than 0.0003% iodate.  Baker
Analyzed sodium thiosulfate pentahydrate was used.  The crys-
tals were used as  a primary standard after determining that the
values obtained based on the weight of the crystals agreed
within 1.5% of those based on standardization against potassium
dichromate.
Methanethiol
     Solutions of  methanethiol were made by passing the thiol
into water and then cooling to the reaction temperature (0°C).
An accurately measured aliquot of the thiol solution was
transferred to a glass stoppered flask containing 25.00 ml
of a known concentration (approximately 0.15 M) iodine solution,
After forty minutes, the solution was acidified with 5 ml
of 6 M HC1 and the remaining iodine was titrated with a
standard 0.1 M thiosulfate solution to a starch end point.
This method is based on the following reaction.
               2CH3SH  +  I2  -»•    CH3SSCH3  + 2HI
                         25

-------
Me thanes u_1 f on ic Acid
     Potentlometrlc titratlons (Beckman Century SS pH meter)
using standard base solutions as titrant were used to determine
methanesulfonic acid.   The base solutions were standardized
against potassium hydrogen phthalate.   Ozone was found to
interfere with the determination and was removed by bubbling
out with nitrogen before the titration.
     The methanesulfonic acid was identified by converting
it to methanesulfonyl  chloride using thionyl chloride.  About
2 ml  of the suspected  sulfonic acid was isolated from the
reaction mixture  by evaporating the water and refluxing
with  20 ml of thionyl  chloride for four and a half hours.
The thionyl  chloride was distilled and the reaction product
was identified by coinjection gas chromatography and its mass
spectrum.
Sulfuric Acid
     The precipitation of barium sulfate from aqueous solutions
is the basis for  a number of qualitative and quantitative
techniques for sulfate.   Qualitative   tests for sulfuric
acid  were performed using three techniques.  One, a small
portion of the solution  to be tested was added to a barium
chloride solution; the formation of a  white precipitate
indicates sulfate and  the absence of the precipitate indicates
the absence of sulfate.   For very dilute solutions the formation
of a  precipitate  can better be detected using a spectrophotometer
Two,  a small portion of  the solution to be tested was used to
adjust a Spectronic 20 UV - visible spectrophotometer to
100%  transmittance and then a few crystals of barium chloride
were  added.   A decrease  in the transmittance indicates pre-
cipitation.   Three, a  drop of a saturated potassium perman-
ganate solution and three drops of the solution to be tested
were  mixed.   One  drop  of the mixture was placed on a filter
                               26

-------
paper which had been impregnated with barium chloride and
heated at 70°-80° C  for seven to eight minutes.  The filter
paper was then washed with water and 1 N oxalic acid.  If
sulfate was present, the precipitated barium sulfate would
trap potassium permanganate in its crystal structure.  The
trapped potassium permanganate would not be washed away;
                                        33
a pink or purple spot indicates sulfate.
Methanesul f im'c Acid
     Ferric ions precipitate sulfinic acids as shown.

               3RS02H + FeCl3    +      (RS02~)3 Fe*3  + 3HC1

     Qualitative tests for sulfinic acids were carried out by
adding 1  ml of 15% FeCl3 to 2 ml  of the solution to be tested.

OZONATIONS
Ozone-Gas Streams
     An ozone-oxygen stream was passed into a 0°C methanethiol
solution  in a gas washing bottle equipped with a glass frit.
It was found that large quantities of the thiol were removed
with the  exit gases.  Cold traps cooled to -78°C and precip-
itation as its silver mercaptide were used to trap the escaping
thiol.  The exit gases from the reaction vessel were passed
through an aqueous potassium iodide solution.  When ozone
passed through the solution as indicated by the formation
of iodine, the reaction was stopped and the aqueous solution
was extracted with several portions of methylene chloride.
The extracts were concentrated by evaporation of the solvent
and analyzed by g.l.c.  and NMR. The aqueous layer was analyzed
for sulfuric, sulfinic and methanesulfonic acids.
     Dimethyl disulfide was ozonized by a similar method.
Because the disulfide is not very soluble in water, some of
the reactions were run on suspensions of the disulfide in
                             27

-------
water.  Methylene chloride extracts of the resulting solutions
were analyzed by g.l.c. Evaporation of the methylene chloride
yielded a compound that was distilled under reduced  pressure.
The compound was characterized by NMR and IR spectra.   The
aqueous portion was analyzed for sulfuric, sulfinic and methane-
sulfonic acids.
Ozone Solutions
     During the ozonations using gas streams to introduce
ozone, it was found that large quantities of methanethiol
or dimethyl disulfide were lost in the gas streams. There-
fore, in order to obtain quantitative results solutions of
the thiol or disulfide were added to an ozone solution.
     About 4000 g of water was  weighed  and cooled to 0°C,
for use in the preparation of an ozone solution.  Both ozone-
oxygen and ozone-nitrogen streams were employed to make the
solution.  Methanethiol solutions prepared as previously
described were made while the ozone solution was being pre-
pared.  Because of the instability of the solutions, the
following procedure was carried out as rapidly as possible.
An aliquot of the methanethiol solution was added to an iodine
solution for analysis of the thiol.  The initial ozone deter-
mination was started by adding an aliquot of the ozone solution
to a dilute potassium iodide solution.  Then an aliquot
of the thiol was added to the ozone solution and stirred
slowly for approximately a minute to insure complete reaction
and mixing.  Two aliquots of the reaction mixture were then
added to potassium iodide solutions for determination of the
final ozone concentration.  In the initial experiments an
aliquot of the reaction mixture was then used to determine
the methanesulfonic acid.  For reasons described later, it was
necessary to remove the residual ozone before titrating.  The
ozone was removed by bubbling with nitrogen until a small
sample of the reaction mixture would give no color when
                          28

-------
added to a potassium Iodide-starch solution.  The quantitative
work on dimethyl disulflde was done using the same procedure.
The disulfide was used as a primary standard in making up
solutions and was not analyzed.
                          29

-------
                      SECTION V

                DISCUSSION AND RESULTS
                 SULFUR-DIOXIDE OZONE

     The various species of sulfur dioxide in aqueous
solution are shown in equation 1.
        H?0                               .        _       ,
(1)  S02 J  S02 -H20  +  H2S03 2 HS03~ + H   £ S03~  +  H2
     It was our initial  assumption that each of the species-
sulfite, bisulfite, and  sulfurous acid (or aqueous sulfur
dioxide) was capable of  reacting with ozone.  To distinguish
among these possibilities we measured the rate of the ozonation
reaction at various hydrogen ion concentrations.  Kinetics
thus determined yield, for any single pH, an overall  rate
constant which may include contributions from one or  more
sulfur species.  Equations 2-4 below indicate the reactions
which might be occurring.
                          ^(SO )
(2)       H2S03  +  03      + 2         H2S04 + 02
                           H20
                          kHSO ~
(3)       HS03" + 03       ->  *         HS04
                           SO =
(4)       S03=+  °3        * 3          S04= + °2
                           H20

     Table V shows the mean rate constant (k/tota-i suifur))
                           30

-------
               TABLE  V
         MEAN  RATE  CONSTANTS
  pH                     k(total sulfur)

0.
1.
1.
2.
2.
2.
3.
3.
4.
301
620
20
71
11
50
80
12
55
02
2
9
4
1
1
3
4
5
9
1
.24
.82
.31
.02
.63
.03
.57
.97
.47
.74
x
x
X
X
X
X
X
X
X
X
1
1
1
1
1
1
1
1
1
1
o3
0
0
0
0
0
0
0
0
0
3
4
5
5
5
5
5
5
6
+
+
+
+
+
+
+
+
+
.89
.48
.35
.21
.27
.37
.42
.53
.66
        16  Determinations
0.59              8.60 x 103  + .23
2.13              1.42 x 105  + .11
3.09              2.13 x 105  + .37
3.74              1.43 x 106  + .28
                   31

-------
variation with pH at 25.0° and 16.0°.  Figure II is a plot
of that data, which shows its regularity in graphical form.
     Rate constants were determined using the following
equation:
               k = [1/a-b] [ln(b(a-x)/a(b-x))] /t
where          k = rate constant
               a = initial ozone concentration
               b = initial sulfur concentration
               x = change in concentration at
               t = time
     Analysis showed the reaction to be second order, first
order in both ozone and total S(IV) species.
     In order to account for the strong dependence of the
rate constant on pH, we have assumed the reaction takes place
at different rates with the three species of S(IV) present due
to the equilibria
     H2S03     ?         H+ + HS03"          K]

     HS03"     £         H+ + S03"2          K2

The fractional distribution of the species can be calculated
from the pH and the equiIibriurn constants for the above
mentioned equilibria.   There seems to be no clear cut choice
for the values to be used with values of K, reported ranging
from 9xlO"3 to 5xlO~2 and K2 ranging from 6xlO"8 to 5xlO~6
at 25°C.
     Our calculations reported here are based on the K,values
of Arkhipova et. al..  35 who report KI  values at 35°, 25°,
and 10° as l.OxlO"2, 1.3xlO"2 and l.SxlO"2 respectively.
The K2 values used are those of Teder who gives values of
1.16x10~7 and 9.3xlO"8 for 25° and 60°.  These values were
                            32

-------
20
18
16
14
12
10
 8
                          FIGURE  II



                  DEPENDENCE OF RATE ON pH
                                33

-------
chosen because they gave a means of calculating values at
temperatures other than those given.
     A series of calculations were made using the values K,
= 1.72xl02 and K2 = 1.02xl07.  Calculated rate constants
showed about the same variance but absolute values were dif-
ferent due to the different species distribution calculated.
     The sulfur species concentrations were converted into
fractions, leading to the equations shown in Table VI. It
was obvious that k^n  was considerably smaller than the
other specific rate constants and adds essentially nothing
to the overall rate of reaction at hydrogen ion concentrations
lower than .1 M.  Its fraction of the total sulfur is there-
fore not included in equations 4-10.
     Equations 3-10 in Table VII  were used to determine
k
    3 an   SO^ (computer program using each equation, one
at a time, with every other equation, and solving for the
rate constants).  Those values were then substituted into
equations 1 and 2 to determine the value of kcn .  Similar
                                               2
procedures were followed for the data determines at 16°.
     Table VII shows the rate constants thus determined.
     Arrheneus treatment reads to the following formulas:

     d[HSO ~]      ,,         11,600
                   1        "
                      exp  "  -    [HS03~] [03-]
        dt                      RT
     d[SO~2]                 10,500         9
     	—  - 1017exp      	    tS03-2]  C03]
        dt                      RT
                             34

-------
                            TABLE VI


      EQUATIONS FOR CALCULATIONS OF SPECIFIC RATE CONSTANTS



  General  Equations

    k    = [Fraction HS03]k] + [Fraction S03]k2 + [Fraction S02]k3
                                 25.0° Runs


1)   -.301    2.24 x IO3 = .0065 kHSO* + 3.8 x IO"10  k$0= +  .993  k$0
                                     3                  «3           t

2)   0.620    9.82 x IO3 = .051 kHSQ-  + 2.48 x 10"8  k$0= +
8)


9)
0)   4.02     1.74 x 106 =  .994  kH$()-   +  1.20  x  10
                                                   -3
                                    '3                  DU3
                                   35
                                    o                   *

3)   1.20     4.31 x 104 = .171 kHSQ-  + 3.14 x IO"7  k$0= +  .
                                    3                   <5

4)   1.71     1.02 x 105 • .402 kHSQ-  + 2.39 x 10'6  ks(J.
                                    3                   3

5)   2.11     1.63 x IO5 = .627 kHSQ-  + 9.38 x 10"6  k$0=
                                    3                   «5

6)   2.50     3.03 x 105 * .805 kHSQ.  + 2<95 x 1Q-5  ^^
                                                        «5

7)   2.80     4.57 x 105 - .891 kHSQ-  + 6.54 x 10"5  k$0=
                                    v                   »

     3.12     5.97 x 105 = .946 kHSQ-  + 1.44 x 10"4  k$(J*
                                    3    .              
-------
                        TABLE VI  (cont'd)
                                  16.0°  Runs
11)   0.59     8.6 x 103 = .058 k,   + 3.15  x  10"8  k?   +  .94  k
                                                   £


12) 2.13       1.42 x 105 =  .683 k,  + 2.76  x  10~5  k2   +  .32  k



13)   3.09     2.13 x 105 =  .951  k]  + 1.64  x  10"4  k£



14)   3.74     1.43 x 106 =  .989 k]  + 7.61  x  10"4  k
                                     36

-------
                      TABLE  VII




                SPECIFIC  RATE CONSTANTS
 j. O c"
at 25




            = 3.1  + 1.1  x 105  M/lsec
     k n=   = 2.2 + 1.6 x 109  M/lsec

      S03


     k      = 5.9 ±     x 102  M/lsec
      so2
at 16°
     kHSQ-  = 1.71 ±  .3 x 105  M/lsec





     kSO=   =1-3  ±  .7 x 109  M/lsec


     k      =0                M/lsec

     Kso2
                              37

-------
from which the reaction rate at any temperature can be
calculated.   The overall  reaction rate will  be the sum of the
two rates determined by calculating the concentration of
each species from the equilibrium constants  and the pH.
     The contribution of the SO,, •  H,,0 species to the rate
is small at  all  pH of interest in the atmosphere.  Our value
               _ 2
of less than 10    x kuc  ~  would make the contribution  of
this species less than    1% of the total  rate at all  pH values
more than 1.7.
                        _2
     Conversely,  the SO,   species becomes a greater than
1% contributor  when its fraction becomes greater than  about
  -6                                    4
10   since its  rate constant is about 10  x that of the HSO.,
species.  This  occurs at about the same pH (1.7) that  the SO,,
species becomes unimportant.  Thus below pH 1.7 we can ignore
sulfite contribution and above this pH we  can ignore the SOp
contribution to the rate.   At pH > 4, the  reaction is  almost
                   _?
entirely due to S03
     Because of the extreme dependence of  rate on pH
(rate approximately doubles for 0.5 pH change), the method
of dividing the overall rate constant into the specific rate
constants for the different species is valuable. The distri-
bution among species is a  function only of the H  concentration
and the ionization constants of the species involved so our
rate constants  can be used to determine an overall rate
constant at any desired ph.
     There were experimental difficulties  in working at the
hydrogen ion concentrations of atmospheric interest.  The
problems, were:
     (1)  Increasing the pH decreases the  amount of sulfur
          dioxide in solution (bisulfite becoming the
          common  sulfur species present under our conditions).
          This  meant that  either we increase total sulfur,
          in order to have relatively constant absorption
                             38

-------
of sulfur dioxide for the determination of initial  concen-
trations or work at lower accuracies with respect to all
concentrations.  We chose to keep relatively high sulfur
dioxide concentrations.   In retrospect (see 2&3) our
choice was wrong.
     (2)  The rates of ozonation increased dramatically with
increasing pH.  This meant that at higher pH values our
kinetic results started  to be limited by the mixing time of
the stopped flow apparatus.  For the 16° runs concentrations
of both ozone and total  sulfur were lowered for higher pH
runs, but the determinations above pH 3 at 25° are  suspect.
     Because of that (after the termination of the  grant)
we lowered the concentration of ozone by a factor of 20 and
the concentration of total  sulfur by a factor of 80 and did
one more determination at pH 3.22.  The observed rate
constant (1.79x10 ) was  almost three times larger than what
we would have expected from extrapolation on Figure n.
The result by itself does not indicate whether we had pushed
the limits of the stopped flow apparatus previously or whether
our decreased accuracy of measuring total sulfur was responsible,
In fact we think it likely that a third factor (3 below)  may
have been important in both that experiment and in  all
kinetic determinations at higher pH.
     (3)  Our basic assumption has been that equilibrium occurs
in the conversion of one sulfur species to another.  Using
the rate constants from  equation 5 it can be shown  that equil-
ibrium would be established rapidly enough so that  ozonation
can occur with each species at most concentrations  used.
However, at high pH and  with very high concentrations of
total sulfur and ozone the rate of formation of sulfite from
sulfurous acid is only about ten times that of its  dis-
appearance from reaction with ozone.  This seems not to
                           39

-------
be a particularly bad problem since at such acidities most
of the total sulfur is already in the bisulfite form.  More
serious is our original assumption that the rate constants
for the second step of the ionization
                   k2 =  104
     HSO "            t            H+ + SO"
                           11             ^
                  k_2 =  10''

has the values shown.  We have found no mention of these
rate constants in the literature and have assumed them from
a) knowing the equilibrium constant and b) guessing that
the recombination of a proton with sulfite ion might have
about the same rate constant as the similar reaction between
sulfate ion and a proton (a diffusion controlled reaction
rate).  At pH above about 2, the SO.,   rate is important.
                                                2
The above assumed rates would indicate that SO,    is being
used up more rapidly than it can be replenished by HS03~
dissociation.  This would result in a trend similar- to that
observed of an apparent relative decrease in the constant
with the increase in pH.
Environmental Implications
     The reason for determining the kinetics of the reaction
between ozone and sulfur dioxide in solution was to discover
whether such a reaction might take place under normal atmo-
sphere conditions.  While this study was in progress, Penkett
determined the rate of the reaction between bisulfite ion and
ozone and reported the following results.
     a)   The reaction is rapid, having a second order rate
          of
               dO.  _
               —-     [3.32+ .11 x 10V sec "'] [Oj [HSO ~]
               dt           "                        J
          at 9.6° and at a pH of 4.65.
                               40

-------
     b)   Using reasonable atmospheric conditions [SO,, in
          atmosphere = 0.007 ppm, 0- - 0.05 ppm, clouds
          contain 0.1 to 1 g liquid water per cubic meter,
          bisulfite concentration in solution = 5 x 10" ]
          he calculates an oxidation rate of 12.6% hr  .
          He notes that his is 70 times the rate of disappearance
          of sulfur dioxide via oxygen.
     Our results are in conflict with those of Penkett on
several counts.  First we note that the pH dependence of
the reaction is such that there seems to be little doubt
that Penkett measured the rate of the sulfite ion-ozone
reaction as a major part of his rate constant.  Since his
ozone solution was unbuffered and at a pH of 4 and his bi-
                                       - 4       - 5
sulfite concentration varied between 10   and 10   the exact
pH (and therefore the exact sulfite ion concentration) must
have been slightly variable.  We have retabulated his rate
constant variations with bisulfite ion concentration in Table
IX.

                        Table  VIII
       Rate Constants Ozone-bisulfite (Penkett)

          cone HS00 (M)                 k
                  ^

                    10"5           5.7 x 105
              2.5 x 10"5          3.48 x 105
              5.0 x 10"5          3.46 x 105
                    10"4          3.27 x 105

     If Penkett's initial  pH were 4.65, as he suggests,
then the normal prediction would be that the solution w.th
least buffering capacity (the 10"  solution) would be at
the lowest  pH when mixed with the ozone solution at  pH 4.

                          41

-------
Our initial  prediction would have been that rates  would
increase with increasing bisulfite ion concentration  since
the highest  pH solution would have the highest sulfite
ion concentration.   The trend of Penkett's  data is clearly
in the other direction.  The highest rate is found with  the
lowest initial concentration of bisulfite.
     However when the major discrepancy of  our work with
Penkett's is considered, a possible explanation for his
upward trend in rates with decreasing bisulfite may be
given.  Specifically, our rate equation predicts that the
second order rate constant for Penkett's conditions should
have been 5.5 x 106, using HSQ,"= 10"5 M.  His value  of  5.7
    5
x 10  would  therefore be too low by a factor of about 10,
while his "average" rate constant (3.32 x 10 ) is  about  15
times lower  than what we would predict.
     We believe that Penkett's upward trend  in rates with
decreasing bisulfite concentrations, our own similar  finding,
and the fact that our predicted rate constants are consider-
ably higher  than his experimentally determined values,  have
the same cause.  Specifically, sulfite can  react with ozone
only if it is formed rapidly enough under equilibrium con-
ditions to do so.
     For purposes of comparison, let us assume the same
concentrations as Penkett did:  S09 at 0.007 ppm,  ozone  at
                     3
0.05 ppm, .1 gm per m  liquid droplets of water, pH = 5.0,
and a temperature of 10°C.  Using the activation energies
obtained in  this work we can extrapolate values for k,,cg-
and kQn= of 1 .
^3 i
respectively.
Using the
1 x 105 (M/l
values of KH
r1 sec'1
Mso2 (
Pso2 (
and
aq)
g)
7.4 x 108 (M/l)"
= 2.20 and
                         42

-------
     (H+)  (HSO")
K, = 	 =  .0184 from the work of Johnstone and Leppla
     (SO-  (aq)  )
        £                                          c
(1934) we  arrive at a HSO" concentration of 4 x 10  M
        =                               -7
and a SO,  concentration of about 4 x 10   M.  and an ozone
           -9
cone, of 10   M.  Therefore,

   - dS(IV)        = (1.1 x 105 x 4 x 10"5 + 7.4 x 108 x 4 x 10"7)xlO"9
      dt                    7
                    =  3.  x 10"' (M/l) sec -1

           3               -4
For each M  we have 1 x  10   1 of solution so we now obtain
3 x 10"   moles/m /sec reacted or 1 x 10"  moles/m  hr or
6.4 x 10"6 g/m3 hr.
     The starting concentration of .007 ppm gives 20 mg m~
so 6.4 = .32 or 32% per  hr is  reacted.  Penkett arrived  at
    20
a figure of 12.6% per hr.  This difference is partially due to
the apparently different values used for K,, and K, (He
calculated a concentration of  HSO^ = 1 x 10 ).
Another possible contributing  factor may lie in the precision
of our activation energy and the consequent uncertainty of
the extrapolated values for the specific rate constant.   All
of the above calculations assume that solution  is rapid enough
that equilibrium between the various phases and species is main-
tained.
     Comparing the above rate  with rates published for other
oxidation modes, it must be concluded that, for the cited
conditions (liquid droplets, presence of ozone  at a constant
concentration) ozone caused oxidation is an important contri-
bution to the oxidation of S(IV) in the atmosphere.   For exam-
                              •i /•
pie, Brimblecombe and Spedding   using approximately the same
total  S(IV)  and H20 (1) values that we assume, calculate the
                          43

-------
removal  of about 3% per day by the Fe (III) catalysed

                                              37
oxidation by atmospheric oxygen, Cheng et. aj_.   calculate


a 2%/hr  decrease in sulfur dioxide in natural  fog in the presence

                                         38
of manganese salts and Sidebottom et. al.    show a 1.9%/hr


loss of  sulfur dioxide photochemically at high humidity.
                          44

-------
                  METHANETHIOL-OZONE

Methanethiol-General Results

     When an ozone oxygen stream was passed into a 0°C
aqueous solution of methanethiol, a rapid reaction occurred.
Quantitative analysis was not possible using this system
because large amounts of methanethiol were carried away in the
oxygen stream.  Cold traps were not effective in trapping
all of the escaping thiol, but silver ions were found to
precipitate all of it.   Since the determination of the
amount of thiol precipitated was not practical, an indirect
method of analysis was  attempted.  A known amount of silver
nitrate was used to trap the escaping thiol and the excess
silver was titrated with potassium thiocynate.   Unfortunately,
absorption of silver ions on the precipitated mercaptide
obscured the end point.  Therefore, the reaction was usually
stopped when the thiol  concentration was so low that some
of the ozone passed through the solution without reacting.
     A strongly acidic  solution resulted from the reaction;
however, qualitative tests for sulfuric and methanesulfinic
acids were negative.  The acidic product was isolated and
it was identified as methanesulfonic acid by conversion to
methane-sulfonyl chloride.  Extraction with methylene chloride
revealed that other products were also present.  Dimethyl
disulfide, methyl methanethiosulfonate and methyl methane-
thiosulfinate were identified as minor constituents.  The
disulfide and thiosulfonate were identified by coinjection
gas chromatography and  their mass spectra.  The thiosulfinate
was shown to be present by gas chromatography and nuclear
magnetic resonance resonance spectra.  Continued ozonation
resulted in the formation of sulfuric acid.  The equations
                         45

-------
below summarize the chemical  processes:
     CH3SH
                   slow
                                   CH3SO 'SCH3
                                   CH3S02SCH3         Xtrace)
     The stoichiometry of this reaction was studies by
adding known thiol solutions to known ozone solutions.
The results obtained are given in Table X.   The results
indicate a stoichiometry of slightly less than 2 moles of
ozone per mole of thiol.  No difference in  stoichiometry
was found, whether ozone-oxygen or ozone-nitrogen
streams were used to prepare the ozone solutions.
     Titration of the resulting sulfonic acid was not straight-
forward.  When a sample was titrated immediately following
reaction, the titration curve contained two inflection points.
The first inflection corresponded closely to quantitative
conversion of the thiol to sulfonic acid.  If the solution
was allowed to stand several days before it was titrated,
a single inflection point having a value equal to that of
the second inflection in the original solution was obtained.
It was found that if the excess ozone was removed before titra-
ting, only one inflection point was obtained and it correspond-
ed to near quantitative conversion of the thiol to sulfonic
acid.  The yield of sulfonic acid was taken at the first inflec-
tion point was obtained and it corresponded to near quantitative
conversion of the thiol to sulfonic acid.  The yield of sul-
fonic acid was taken at the first inflection point or from
a titration in which the excess ozone  had been removed.  The
yields obtained are also given in Table X.
     The rate of the reaction appeared to be very fast.

                          46

-------
                              TABLE IX
Run       Moles of 03    Moles  of       03/   Moles  of              %
          Consumed     CH3SH Added    CH-jSH   CH3$03H Produced      Yield
1
2
3
4
5


7
/
8
1.24 x 10"3
1.18 x 10'3
#9.7 x 10"4
#4.3 x 10"4
1.03 x 10'3
Q £ y y 10
J • U A A 1 U

#1.82 x 10'3
6.
6.
5.
2.
5.

O •
1
1 •
1.
30 x
84 x
58 x
38 x
97 x
7 1 v
/ O A
fiK y
U J A
03 x
TO'4
TO'4
ID'4
TO'4
TO'4
IO-4
1 U
TO'4
1 U
TO'3
1
1
1
1
1
1
1
.
1
.97
.75
.74
.80
.73
fifi
• o o
— _
.76
6.
6.
5.
2.
7.


1.
1.
36 x
87 x
54 x
74 x
40 x


71 x
02 x
10
10
10
10
10


10
10
-4
-4
-4
-4
-4


-4
-3
101
100
99.3
115
124


*103
98.5
                                              1.09  x  10"3         *106
          #  -  Solution  made  with  ozone-nitrogen  stream.

          *  -  Yield  determined  after  removal  of  excess  ozone
                              47

-------
Observation of the decrease in ozone was followed spectrophoto-
metrically at 285 nm using the stopped flow apparatus.   At
initial thiol concentrations of 7 x 10   M. and ozone concen-
trations of 3 x 10   M the reaction was over in about 0.1
seconds.
     The experimental  titration curves show a double hump
pattern obtained when  an aliquot of the reaction mixture
is titrated in the presence of ozone.  However if the ozone
is removed a standard  titration curve is obtainable.
This pattern, although puzzling at first, proved to have a
simple explanation.   In basic solutions, ozone will react
with methane sulfonic  acid to produce sulfuric acid much
faster than in acidic  solutions.  In acidic solutions,  the
reaction proceeds very slowly.  That is why letting the solu-
tion stand for several days results in a single hump pattern
corresponding to the second hump in a titration with ozone
present.  As the equivalence point is reached during titra-
tions with ozone present, the pH rises allowing sulfuric
acid to be produced  at a reasonable rate.  The pH then  remains
relatively constant  until the ozone has all reacted.  Tests
with methanesulfonic acid showed the same behavior as the
reaction mixtures.  Other tests demonstrated that sulfate
was formed when ozone  was added to basic solutions of methane-
sul foni c acid.
Methanethiol-Errors
     Any quantitative  measurement is subject to some error.
The results given in Table X  exhibit some random error.
The stoichiometry from run #1 and the yield from run #5
are more than two standard deviations from the mean and,
therefore, have been dropped from the respective calculations.
                            48

-------
Doing this one obtains a stoichiometry of 1.74 + 0.04
and a yield of 103% + 6%.  The limits of error given are
one standard deviation.
     This treatment of random errors does not include any
systematic errors.  Although efforts were made to reduce any
error of this type, several sources of possible error exist.
The solutions used in this study were all unstable with
respect to loss of the solute into the air.   In order to
minimize this type of error, initial concentrations were
measured, the reaction was carried out and final concentra-
taions were measured as  rapidly as possible.  Ozone is also
known to undergo thermal decomposition, but this is a slow
reaction compared to the time between samplings (about one
minute).
     Because of the limited solubility of ozone in water,
the solutions used were  very dilute.  Final  ozone concentra-
                        - 5          -4
tions ranged from 5 x 10   to 3 x 10   M.  At these concentra-
tions, end points were not extremely sharp.   This was par-
ticularly a problem in the sulfonic acid determinations where
the steep portion of the titration curve generally occurred
in 1/4 to 1/2 mill il Her.

Methanethiol- Scientific Implications
     The reaction of ozone with methanethiol has been shown
to produce methanesulfonic acid and eventually sulfuric acid.
The conversion of the sulfonic acid to sulfuric acid is
very slow and, for practical purposes, the reaction stops
at the sulfonic acid.  When insufficient ozone was added
directly to a solution of thiol,  dimethyl disulfide, methyl
methanethiolsulfinate and methyl  methanethiolsulfonate
were also detected.  The quantitative results  seem to indicate
that these products were not formed in the reactions carried out
                         49

-------
by mixing solutions.  This is reasonable because, in the
presence of excess ozone, any disulfide and thiolsulfinate
would have been oxidized further.   However, Barnard    has
reported that thiolsulfonates are  resistant to ozonation.
No minor products were  detected from these reactions; but
at such low concentrations, they may have escaped detection.
     These results are  not surprising and are similar to
the results reported for other thiols.   A thiolsulfinate, how-
ever, has not previously been reported  as a minor product
or intermediate in the  ozonation of a thiol. The reason
for this probably involved a difference in the mechanism
of disulfide ozonations in aqueous and  organic solvents.
Previous workers using  organic solvents have not
found thiolsul finates to be products of long-lived  intermediates
in the ozone-disulfide  reaction.  This  evidence,  although
it seems to indicate that the thiolsulfinate is  formed from
the disulfide,  is not sufficient for proof.
     Because thiols are readily converted to disulfides, it
might be expected that  the ozonation of thiols proceeds via
disulfides.  This does  not appear  to be the case for methanethiol
If the reaction involved the disulfide, the overall  rate would
be controlled by the rate of ozonation  of the disulfide or
the rate of conversion  of the thiol to  the disulfide, whichever
is slower.  But methanethiol was observed to react faster
than dimethyl disulfide with ozone.  Therefore,  it must
be concluded that disulfide production  is only a minor side
reaction.
     The stoichiometry  found is similar to those obtained
by Barnard17 (1.6 03:SH) and Erickson (1.6 03:iso-BuSH)
when the ozone-oxygen streams were passed into the  thiol
solutions.
                          50

-------
Methanethiol-Envi ronmental Imp!ications

     The reaction product (CHoSCKH) from the ozone-methanethiol
reaction is odorless, nonvolatile, and water soluble.  These
are all positive attributes that  could be taken advantage
of in designing pollution control  equipment for Kraft pulp
mills.  Although the rate data obtained are in no way complete,
it appears that the reaction is fast enough so that no major
technological difficulties should be encountered in equipment
design.  These are some difficulties in using aqueous phase
ozonations for pollution control.   However, use of ozonized
water in the water scrubbers for  stack gases or direct gas
phase ozonation might be possible.
     The importance of this reaction in the atmosphere is
hard to judge.  In regions where  ozone and thiols are present
in the atmosphere together, some  reaction  no doubt takes place,
However, these reactions are probably occurring in the gas
phase for the most  part.  In plumes from pulp mills, where
suspended water droplets are in high concentration, some
aqueous phase reaction may occur  if ozone is present.
     The situation in the upper atmosphere is similar; if
thiols and ozone are both present, some reaction probably
occurs, but in the gas phase.  Two things should be kept
in mind when dealing with the upper atmosphere:  1) it has not
been shown that thiols are present there, and 2) other path-
ways involving the high energy species would compete for any
thiol present.

               DIMETHYL DISULFIDE-OZONE

Dimethyl Disul fide-General Results

     When an aqueous solution of  dimethyl disulfide was
                          51

-------
ozonized by passing an ozone-oxygen stream into the solution,
the major product formed was methanesulfonic acid.   The acid
was identified by converting it to methanesulfonyl  chloride
followed by coinjection gas chromatography and determination
of its mass spectrum.   Methyl  methanethiolsulfinate was observed
to form in high yield  during the beginning of the reaction.
In one experiment 0.0056 moles of ozone  was  added to 0.011
moles of dimethyl disulfide dissolved in two liters of water.
The only product detected by NMR spectra of  a methylene
chloride extract was methyl methanethiolsulfinate.   The
final pH of the reaction mixture was 5.3.   In some  runs a small
amount of methyl methanethiolsulfonate  was also formed.
No sulfinic acids were detected.  This  is  summarized in equation
                                      °3
     CH3SSCH3 + 03   ->   CH3SO -SCH3  ->      CH3$03H

     The stoichiometry and yields found  for  the disulfide reaction
are given in Table XI.  All titrations  of the sulfonic acids
were performed after the removal of excess ozone from the
solution.
     The rate of this  reaction appeared  to be slower than
                                                       -4
that of the thiol.  At initial concentrations of 1  x 10
M CH3SSCH3 and 6 x 10"  M 03 the reaction was observed to be
complete in about one  half minute.
Dimethyl Pisulfide-Errors
     The mean stoichiometry and yield found  were 3.95 +
0.45 and 94.8% + 7.1%.  Eliminating runs more than  two standard
deviations away from the mean gives 3.89 + 0.07 and 95.6%
+ 4.0%.
     This reaction was carried out using the same procedure
as the thiol reaction, and the limitations of the method
are the same.  One additional  problem was encountered,
                         52

-------
                       TABLE X





Run  Moles of 0.,    Moles of       0, /       Moles of       %
               3                    0


     Consumed       CH3SSCH3     CH3SSCH3    CH3S03H        Yield





                     Added                   Produced

1

2
3
4

c;

6

*2.

*2.
2.
2.

2.

1.

24 x

17 x
68 x
15 x

21 x

90 x

10"

10"
10"
10"

10"

10"
q
o
•3
o
3
3
•3
*}
q
o

5.63

5.63
5.63
5.63

5.63

5.63

x 10

x 10
x 10
x 10

x 10

x 10
-4
*T
-4
~ *T
-4
-4
-4
*T
-A
*T

3

3
4
3

3

3

.97

.84
.77
.82

.92

.38

1

1
1
1

9

1

.16 x

.13 x
.10 x
.04 x

.37 x

.04 x
_ q
10 J
_ q
10 6
10"3
10"3
-4
10 4
q
10 J

103

100
98.1
92.2 -

83.2

92.2
       -  solution made with  an ozone-oxygen  stream.
                             53

-------
however.  The time necessary for completion of the reaction
was not considerably less than the time allowed for reacting.
Although enough time for complete reaction was generally
allowed, it is remotely possible that some reactions were
stopped before completion.   Longer reactions times were
avoided because of the instability of the solution.

Dimethyl Disulfide-Scientific Implications
     There appears to be a  difference in yield of sulfonic
acid between reactions run  with solutions made from ozone-
oxygen and ozone-nitrogen streams.  Considering the limited
amount of data and the large degree of scatter, it is possible
that this apparent difference is not real.  It is probably
best to conclude that ozone will convert dimethyl disulfide
to methanesulfonic acid in  high yield.
     Sulfonic acids have not previously been reported as
the products from the reactions of disulfides with ozone,
but the formation of sulfonic anhydrides was observed in
nonaqueous solvents by Barnard  .  Although water must be
involved at some point in the reaction sequence, the reaction
does not necessarily follow the mechanism proposed by Barnard
with the addition of a hydorysis step at the end.  In fact,
the results of the investigation indicate that a different
mechanism is in operation.  In trying to elucidate the mechanism
of the reaction, Barnard   conducted experiments in which
insufficient ozone for complete conversion was used.  Only
the normal products of complete ozonation were found.
In this study, however, high yields of methyl methanethiol-
sulfinate were isolated when insufficient quantities of ozone
were used.  Apparently in water, the first step of the conversion
to sulfonic acid is the oxidation of the disulfide to the
thiolsulfinate.  Although no sulfonic acid was formed early
in the reaction, quantitative data is needed to show that
                           54

-------
another reaction pathway going through an undetected interme-
diate was not in operation.  The small amounts of thiol-
sulfonate that were detected in some runs probably arose through
minor side reactions.  Known reactions that could possible
be the source are:
     1)   the disproportionation of thiolsulfinate to disulfide
          and thiolsulfonate, and
     2)   the autoxidation of thiolsulfinate.  It is also
          possible that the thiolsulfonate is a direct
          ozonation product.

Dimethyl  Disulfide-Environmental Implications
     Because methanesulfonic acid is the major product of
this reaction as well as from methanethiol, the same advantages
(odorless, nonvolatile, and water soluble) in designing
pollution control equipment are gained.   The slower reaction
rate would make removal of the disulfide more difficult.
     The  opportunity for reactions in the atmosphere are the
same as those for the thiol.

DIMETHYL  SULFIDE-OZONE
     Over a year's work went into the building of gas phase
kinetic system, and many months of obtaining rate data follow-
ed.  Although the data is available  , it is meaningless.
Our original  system apparently measured the rate of reaction
of ozone  with dimethyl sulfide which had coo ed out in the
freeze out trap.  Near the end of the  research period several
new reaction chambers were built and it appeared as if the
reaction  was suitable for study.
     The  last of the reaction chambers was built from non-
precision bore 20 mm (OD) tubing and was 120 cm long (Volume =
305 ml).   The bend into the KI sampler was considerably
less restricted than earlier models.  The following data was
determined.

                           55

-------
Run
               Flow (1/min)
Order uncorrected = 1.314
corr. coef.        =  .995
Run
               Flow (1/rnin)
M(03)xlO(
M(DMS)xlO'
d(0-)/dtxlO'
1
2
3
4
5
6
7
8
1
1
1
1
1
1
1
1
.260
.278
.250
.261
.273
.252
.259
.275
9.
11.
7.
2.
2.
6.
7.
8.
049
362
054
481
724
702
425
797
1 .646
1 .912
1.103
.2779
.3241
1.242
1.273
1 .498
     d(03)/dtx!0-
1
2
3
4
5
6
7
1.265
1.261
1.274
1 .267
1 .272
1 .275
1 .260
1.
•
•
3.
.
2.
2.
462
9874
9773
499
6657
276
303
1.1967
.7842
.7080
1.7399
.5467
1.498
1.646
Order uncorrected = .757
corr. coef.        = .974
Correctee Ozone Order = 1.398
Corr. Coef.            =  .990
Corrected DMS Order   =  .861
Corr. Coef.            =  .969
*    This order was computed assuming a 1:1  stoichiometry
for ozone and DMS.   If the ozone order is assumed to be 1.5
and the DMS  to be 1.0, the rate constant at  25° is approximately
4 10

         moles/1  min.   This might be compared to the rate
                          56

-------
              22
found by Hales   for the gas phase reaction between hydrogen
sulfide and ozone of
     Repetition of the method at 35° yielded inconsistent
results.  Specifically ozone concentrations became highly
erratic.  Whether these problems arose from a faulty ozonator
or the method itself is still unknown.
                          57

-------
                        REFERENCES
 1.    Urone,  P.  and  Schroeder,  W.H.,  "S02  in  the  Atmosphere:
      A Wealth  of Monitoring  Data,  but  Few Reaction  Rate  Studies"
      Env.  Sci.  and  Tech.  3,  436-445,  1969.

 2.    Bufalini,  M.,  "Oxidation  of Sulfur Dioxide  in  Polluted
      Atmospheres -  A  Review"  Env.  Sci.  and Tech.  5,  685-700,
      1971.

 3.    Kellog,  W.W.,  Cadle,  R.D.,  Allen,  E.R.,  Lazrus,  A.L.  and
      Martell,  E.A.,  "The  Sulfur  Cycle"  Science  175,  587-596,
      1972.

 4.    Williams  M.D.  "Review of  Selected  Literature on  Sulfur
      Dioxide  to Sulfate Aerosol  Conversion".

 5.    Gregor,  I.K.  and Martin,  R.L.,  Reaction  Between  Ozonized
      Oxygen  and H9S in the Gaseous Phase  Austral  J.  Chem.,
      14,  462,  196T.

 6.    Cadle,  R.D. and  Ledford,  M.,  The  Reaction  of Ozone  with
      Hydrogen  Sulfide Air and  Water  Pollut.  Int.  J.  10,25,
      1966.

 7.    Pitts,  J.N.,  Environmental  Appraisal:   Oxidants, Hydro-
      carbons,  and  Oxides  of  Nitrogen,  J.  Air  Pollut.  Control
      Assoc.,  19, 658, 1969.

 8.    Cadle,  R.D.Discuss ion,  ibid.  19,  668,  1969.

 9.    Espenson,  J.H.  and Taube, H., Tracer Experiments with
      Ozone  as  Oxidizing Agent  in aqueous  solution;  Inorg.
      Chem.,  4,  704,  1965.

10.    Cox,  R.A.  and  Penkett,  S.A.,  Oxidation  of  Atmospheric
      Sulfur  Dioxide by Products  of the  ozone-olefin  reaction
      Nature,  230,  320, 1971.

11.    Penkett,  S.A.,  "Oxidation of S02  and other  Atmospheric
      Gases  by  Ozone in Aqueous Solution"  Nature,  240, 105-6,
      1972.

12.    Akamatsu,  I.,  Kamishima,  H. and Kimura,  Y.,  Kami-pa
      Gikyoshi,  1968,  22,  406-410,  Deodorization  of Exhaust
      Gas  in  Kraft  Pulping I  Formation  of  Malodorous  Components
      in Kraft  pulping.

 13.   Maneck,  M., Das  Braunkohlen-Archiv.40,  61  (1933).  Desul-
      furization of  brown-coal  tar distillates.

                             58

-------
14.  Bohme, H. and Fischer, H., Ber. .  75B.J310 (1942).
     Uber die Einwirkung von Ozon auf  ThToather.

15.  Boer, H. and Kooyman, E.G., Analy.  C h i m.  Acta, 5, 550
     (1951).   The Use of ozone as a ti trimetric agent for
     the determination of definic unsaturation.

16.  Bateman, L., and Cunneen, J.I., J.  Chem.  Soc, 1596
     (1955).   Oxidation of Organic Sulfides  III-

17.  Barnard, D., J.  Chem. Soc. . 4547  (1957).   Oxidation of
     Organic  Sulfides IX The Reaction  of Ozone with Organic
     Sulfur Compounds-

18.  Maggiolo, A. and Blair, E.A., "Ozone Chemistry and
     Technology" Advances in Chemistry Series  No.  21,
     American Chemical  Society, pp. 200-01.

19.  Homer,  L., Schaefer, H.  and Ludwig, W.,  Ber., 91,  75
     (1958).   Ozonisierung von tertiaren Aminen (Phosphinen
     Und Arsinen) sowie Thioathern und Disulfiden.

20.  Kimura, H. Huges   L.J., McMinn Jr.,  T.D. and Burleson,
     J.C., C.A..  60, 6751 (1964).U.S. 3, 114, 775 (Cl.  260-607)

21.  Akamatsu I, Kami-pa Gikyoshi 22 (4), 200  (1968) and
     Akamatsu I, Kamishima, Kami-pa Gikyoshi 20 (w)
     556 (1966).

22.  Hales, J.,  Doctorate Thesis, University of Michigan,pp.  69-
     70, 1968.

23.  Asinger, F., Ebeneder, F. and Richter,  G., J. Prakt. Chem.,
     2,203 (1955).  Products of the joint action of sulfur
     dioxide  and chlorine on alphatic hydrocarbons  in U.V.
     light VI.

24.  Erickson, R.E.,  unpublished research.

25.  Kirchner, K., Kastenhuber, H. and Biering, L., Chem. Ing.
     Tech. , 43,  626 (1971). Kinetics of  the  reaction between
     mercaptan and ozone in the ppm range.

26.  Karchmer, J.H.,  "The Analytical Chemistry of  Sulfur
     and its  Compounds", Part  II, Wiley  Interscience, New
     York, 1972, p. 764.
                          59

-------
27.   Weast, R.C.,  et.  al . ,  "Handbook  of  Chemistry  and  Physics,"
     45th Ed.,  Chemical  Rubber Co.,  Cleveland,  Ohio,  1964,  p.
     D-103.

28.   Jose,  J.,  Philippe,  R.  and Clechet,  P.,  Bui 1 ,  Soc.  Chim.
     France 2860 (1971).   Etude des  equilibres  liquide-
     vapeur par mesure de  la pression totale  -DMSO.

29.   Boyd,  A.M., Willis,  C.  and Cyr,  R.,  Anal.  Chem..  42.
     670 (1970).  New  Determination  of Stoichiometry  of  the
     lodometric Method For  Ozone Analysis at  pH 7.0.

30.   Hodgeson,  J.A.,  Baumgardner,  R.E.,  Martin, B.E.  and
     Rechme,  K.A.,  Anl.  Chem 43, 1123 (1971).

31.   Kopczynski, S.L.  and  Bufalini,  J.J., Anl .  Chem.,  43,
     1126 (1971).   Some  Observations  on  Stoichiometry  of
     lodometric Analyses  of Ozone  at  pH  7.0.

32.   Ref. 25,  p.481.

33.   Feigl, F., "Spot  Tests  in Inorganic  Analysis,"
     Elsevier  Publishing  Company,  New York,  1958,  pp.  314-
     315.

34.   Ref. 25,  Part  II, pp.  767 and 803.

35.   Arkhipova, G.P.,  Mischenko, K.P., and Flies,  I.E.
     "Equilibrium  during  disolution  of gaseous  sulfur  dioxide."

36.   Brimblecombe  P,   and  Spedding D.J.  Atmospheric  Environment
     8,  937 (1974)   The  Catalytic  Oxidation  of  Mfcromolar
     Aqueous  Sulfur Dioxide.

37.   Cheng R.,  Com M.,  and Frohliger J., Atmospheric  Environ-
     ment 5,987-1008,  1971,  Contribution to  the Reaction Kinetics
     of Water  Soluble  Aerosols and S02 in Air at PPM  Concentrations

38.   Sidebottom H.W.,  Badcock C.C.,  Jackson  G.E.,  Calvert  J.C.,
     Remhardt  G.W., and  Damon E.K.,  Envi ron.  Sci.  Technol.6,  72-79,
     1972,  Photoxidation  of Sulfur Dioxide.

39.   Moody R.,  Thesis, U  of Montana,  1973, "Kinetics  of  the Ozon-
     ation  of  Dimethyl Sulfide in  the Gas Phase" 63  pp.
                               60

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1 Ttff-W/3-76-089
                              2.
                                                            3. RECIPIENT'S ACCESSIOWNO.
4. TITLE ANDSUBTITLE

   REACTION KINETICS  OF OZONE WITH SULFUR COMPOUNDS
             5. REPORT DATE

               August  1976
                                                            6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)

   Ronald E. Erlckson  and Leland M. Yates
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS

   Chemistry Department
   University of Montana
   Missoula, Montana    59801
                                                            10. PROGRAM ELEMENT NO.
                1AA008
             11. CONTRACT/GRANT NO.

                R-800655
 12. SPONSORING AGENCY NAME AND ADDRESS
                                                            13. TYPE OF REPORT AND PERIOD COVERED
   Environmental Sciences Research Laboratory
   Office of Research and Development
   U. S. Environmental  Protection Agency
   Research Triangle  Park, NC  27711
              Final Report
             14. SPONSORING AGENCY CODE

              EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
        The reaction  between ozone, sulfur dioxide, dimethyl  sulfide, methanethiol
   and dimethyl disulfide are reported.

        The rate of reaction between ozone and dissolved sulfur dioxide is
   strongly pH dependent.  These results  suggest that under  some conditions
   atmospheric oxidation of sulfur dioxide may involve ozone.   Dimethyl
   sulfide was found  to react rapidly with ozone in the gas  phase but
   reproducible kinetic data were not obtained.  Stoichiometric and yield data
   from the reaction  between ozone and methanethiol or dimethyl  disulfide in
   aqueous solution indicate those reactions  to be complex mechanistically,
   although one product, methane sulfonic acid is predominant.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                           c. cos AT I Field/Group
  *Air pollution
  *Reaction kinetics
  * Ozone
  *Sulfur dioxide
  *Sulfur organic compounds
                              13B
                              07D
                              07 B
                              07C
18. DISTRIBUTION STATEMENT

   RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
 UNCLASSIFIED
                                                                          21. NO. OF PAGES
69
                                              20. SECURITY CLASS (This page)
                                                UNCLASSIFIED
                                                                          22. PRICE
EPA Form 2220-1 (9-73)
                                              61

-------