DEVELOPMENT OF RE6ENERABLE FIBERS
FOR REMOVAL OF
SULFUR DIOXIDE FROM WASTE GASES
by
Ronald W. Fuest
Merlin P. Harvey
November 1968
TECHNICAL REPORT
Covering the period January I to June 30, 1968
Contract No. PH 86-68-74
Prepared for
National Center for Air Pollution Control
Cincinnati, Ohio 45227
UNIROYAL, INC.
Research Center
-------
DEVELOPMENT OF REGENERABLE FIBERS
FOR REMOVAL OF SULFUR DIOXIDE FROM WASTE GASES
by
Ronald W. Fuest
Merlin P. Harvey
November 1968
UNIROYAL, INC.
Research Center
-------
FOREWORD
This report coyers work done during the six months period January 1 to
June 30, 1968 which comprised the initial phase of the contract. The con-
tract has been extended for 12 additional months.
The work was under the administrative direction of Dr. J. S. Lasky.
Dr. R. W. Fuest served as Principal Investigator.
Mr. E. D. Margolin of the National Center for Air Pollution Control
served as Project Engineer.
-------
TABLE OF CONTENTS
ABSTRACT
SUMMARY 1
I. INTRODUCTION 3
II. CHEMICAL BACKGROUND 5
III. SCREENING OF BASIC SORBENTS 11
A. STYRENE/DIMETHYLAMINOPROPYLMALEIMIDE COPOLYMER (SSQ) AND
2-VINYLPYRIDINE/2-METHYL-5-VINYLPYRIDINE COPOLYMER (GIQ) 11
B. POLYETHYLENEIMINE (PEI) 15
C. POLYVINYLPYRIDINE 17
D. AMINOPOLYUREA 17
E. ANION EXCHANGE RESIN 19
F. ST YRENE-METHYLVINYLETHER- DIMETHYLAMINOPROPYLMALEIMIDE
TERPOLYMER (VTQ) 19
IV. FIBERS 21
A. POLYPROPYLENE-POLYETHYLENEIMINE FIBERS 21
B. POLYPROPYLENE-SSQ FIBER 25
V. REGENERATION 27
A. POLYPROPYLENE-POLYETHYLENEIMINE FIBER 27
B. MULTIPLE SORPTION-REGENERATION WITH SSQ 27
VI. SORPTION ISOTHERMS 31
VII. DISTRIBUTION COEFFICIENTS 37
VIII. EFFECT OF FLOW RATE 39
IX. EFFECT OF N02 41
X. EXPERIMENTAL 43
XI. FUTURE RESEARCH PROGRAM 47
XII. REFERENCES 49
-------
LIST OF FIGURES
Figure
No.
1 Sorption of S02 by dry SSQ at 35°C (flow rate - 21.2 ml/min.) 14
2 Sorption of SC>2 by dry aminopolyurea at 35°C. 18
3 Effect of flow rate, PEI level, and draw ratio upon rate of
S(>2 sorption of polypropylene-PEI fiber in the presence of
water vapor at 35°C. 23
4 S0» sorption of undrawn polypropylene- SSQ fiber in the
presence of water vapor at 35°C. 26
5 Sorpt ion-regeneration of SSQ 29
6 Dissociation pressure of SSQ-S02 reaction product 33
7 Langmuir isotherms for dry SSQ at 55°C. and 75°C. 34
8 Log distribution coefficient vs. 1/T for SSQ 38
9 Apparatus for treating sorbent and fiber samples with gas
mixtures 44
LIST OF TABLES
Page
Sorption of S02 by SSQ (g/g) 13
Sorption of S0» by various materials (g/g) 15
S02 sorption of polypropylene fiber containing 9.1% PEI
in presence of water vapor 21
IV Regeneration of polypropylene-PEI fiber 27
V S0_ capacity of dry SSQ 32
-------
ABSTRACT
Several basic nitrogen-containing polymers have been evaluated as SO
sorbents for use in melt-spun fibers. The effect of basicity and amine
nitrogen type upon capacity is discussed. Preliminary work on the effect
of water vapor, temperature, and flow rate upon the capacity of several
candidate materials has been carried out. Two candidate sorbents have been
co-spun with polypropylene to form fibers and their sorption properties
-------
SUMMARY
A. We have selected several polymeric sorbents with which we have had prior
experience in our inhouse programs to improve dyeability of fibers.
These include materials of varying structural types, basicity, and nitro-
gen content. All of the materials evaluated are known to be co-spinnable
with polypropylene.
B. Samples of these materials have been treated in a flow system with SO--
nitrogen mixtures containing, in most cases, approximately 10,000 ppm SO^.
Results with some of these materials are as follows:
A styrene-dimethylaminopropylmaleimide copolymer (SSQ) was
evaluated in detail. It was found that at a temperature of 35°C,
S02 capacity of 0.190 g/g was obtained. This capacity is ~-85% of
the stoichiometric value (0.223 g/g) for the 1:1 complex based on
amine nitrogen content. At higher temperatures (95°C), the
capacity was reduced to 0.033 g/g, approximately 157o of stoi-
chiometric. The presence of water vapor in the gas stream, however,
gave a significant increase in capacity at 95°C (0.075 g/g).
High molecular weight polyethyleneimine (PEI) was also evaluated.
This material was of interest because of its low equivalent weight,
which gives a theoretical capacity for the 1:1 complex of 1.47 g/g.
Capacities as high as 1 g/g have been obtained experimentally in the
presence of water vapor. Reversibility with PEI is poor, however.
A vinylpyridine polymer showed much lower capacity than the alkyl-
amine polymers. This is probably due to its weaker basicity, and
indicates that strong-base polymers are probably necessary for high
capacity.
C. Fifteen-denier, round-cross-section, polypropylene fibers containing
10-12.5 parts per hundred (phr) of high molecular weight PEI have
been spun by conventional melt-spinning techniques without great
-------
believe that considerably higher loadings are achieveable, but this
will require investigation of spinning parameters and development of
specialized spinning techniques.
No attempts to obtain minimum denier or other than round cross-sectional
shapes have been made•
D. Preliminary evaluation of a fiber containing PEI has shown that: capaci-
ties of about 0.184 g S02 per g PEI, i.e., 0.017 g/g on weight of fiber,
are obtained under conditions (35°C, presence of water vapor) where a
capacity of approximately 1 g SO- per g PEI was obtained when the PEI
was supported on a porous mineral material. It should be emphasized
that this fiber does not represent optimum.
E. Regeneration experiments with dry SSQ show that sorbed SO- can be removed
from the polymer by flushing with nitrogen at elevated temperatures.
Work in progress involves carrying samples through several absorption-
regeneration cycles to determine the degree of cumulative irreversibil-
ity. In addition, we are carrying out regeneration studies on fibers.
F. We have shown that the presence of water in the gas stream increases the
SO- capacity of basic nitrogen-containing sorbents. We are in the pro-
cess of adding other typical flue gas constituents (e.g., 0„, NO) to
our synthetic gas streams to determine the effect of these materials
on the sorbtive and physical properties of the sorbents, and to obtain
information on the selectivity of these sorbents toward SO in the
presence of other acidic gases.
G. Dry SSQ has been shown to follow the Langmuir adsorption isotherm for
-------
I. INTRODUCTION
Sulfur dioxide is one of the major air pollutants, and it occurs as a
waste product in a number of industrial operations including the smelting
of ores and the combustion of coal and oil. The deleterious effects of
sulfur dioxide in the atmosphere on plant and animal life, as well as on
materials, has been widely documented.
Considerable effort has been expended in recent years to reduce or
eliminate the discharge of sulfur dioxide into the atmosphere, and it is
recognized that a technique satisfying the following criteria for treatment
of waste gases to remove S0_ is urgently needed:
(1) Low cost, both in materials and construction.
(2) Adaptable for use in combustion units ranging
in size from small industrial furnaces to large
power plants, as well as smelters and other
SO- sources.
(3) Regenerable, with economic sulfur recovery.
Our approach to a solution of this problem involves the use of basic
nitrogen-containing organic polymers for the sorption of sulfur dioxide,
these sorbents being admixed with a melt-spun synthetic fiber to act as a
strong flexible carrier for them.
The fibrous form presents several advantages over particulate sorbent
beds - it is tough, flexible and strong, and it can be fabricated into mats,
felts, knitted fabrics and many other assemblies which lend themselves to
the design of continuous processes. Contact area can be made very high,
depending upon the size and cross-sectional shape of the fiber.
Over the period from January 1, 1968 to June 30, 1968 we have investi-
gated the feasibility of this concept for the Bureau of Disease Prevention
and Environmental Control, Public Health Service, under Contract No. PH 86-
68-74. We have selected and evaluated several basic sorbents which would
-------
co-spinnability, and regenerability, and have obtained quantitative data
regarding the effect of temperature and moisture. Work to determine quali-
tatively the effect of.oxygen and oxides of nitrogen is in progress. Some
-------
II. CHEMICAL BACKGROUND
The consideration of organic amines in either polymeric or small-
molecule form for use as SO- sorbents for treatment of waste gases must
include detailed consideration of the chemistry involved. Although a
review of all potential reactions involving organic amines and flue gas
components would be beyond the scope of this section, we will discuss the
more obvious reactions.
Although the interaction of organic amines with SO- was reported as
early as 1843, one of the first systematic studies was carried out by
(1)
Hill , in which he studied the reaction of aniline with SO-. Aniline
was found to form a 1:1 addition compound with SO-, and the decomposition
pressure of this compound was studied over the range of 3.5° to 50°C.
The dissociation pressure was approximately 0.005 atm. at 3.5° and 0.83 atm.
at 50°C. The heat of dissociation was calculated to be 19,630 cal/mole. In
the presence of water, normal sulfite and acid sulfite salts were formed
, and these products had dissociation pressures markedly less than that
of the anhydrous addition compound. Heating of the anhydrous complex at
100°C for 16 hours in a closed tube gave a colorless solid with no appre-
ciable SO- pressure, indicating that further reaction had occurred to form
a compound in which the SO- was bound in a different manner.
It is possible that the product found here was a thionamic acid or its
(2)
amine saltx .
Later work with homologs of aniline, benzidine, phenylene diamines,
(3)
n-amylamine and n-heptylaminev ' showed that compounds of SO- and amines
were formed at S0_/amine ratios of 1:2 and 1:1. Heats of formation for
several of these, calculated by the Clausius-Clapeyron equation, ranged
-------
amine which were reported as 3,500 cal/mole and 6,400 cal/mole, respectively.
These anomalously low AH values could not be explained.
has studied the reaction of S02 with N,N-dimethyltoluidines and
N,N-dimethylaniline and has found AH values to be in the 21,000-24,000 range,
except for N,N-dimethylaniline (11,600 cal/mole). A similar study was carried
out with trimethylamine^ , which was found to form addition products similar
to those formed by aromatic amines. The equilibrium S02 pressure, however,
at 55°C was approximately 0.055 atm, whereas that for aniline was 0.83 atm.
Bright and Jasper^ ' prepared the 1:1 triethylamine-SO- complex and distilled
it at 93.5°C (751 mm). This indicates that the dissociation pressure of this
compound is less than 0.98 atm at 93.5°C. The inference from this is that
the more basic amines (trimethylamine, K, = 4.24; triethylamine, Kfa = 3.35;
vs. aniline, K, = 9.41) form more stable complexes with SO-- Steric factors
are also important, e.g., the SO- complexes of N, N-dimethyl-p-toluidine and
N,N-dimethyl-m-toluidine are much more stable than that of N,N-dimethyl-o-
toluidine. This is considered to be due to shielding of the nitrogen by the
(4)
ortho methyl group v ' . Steric effects will be discussed more fully later.
From the results of Hill and Fitzgerald, however, it also appears that
subsequent irreversible reaction of SO- with aromatic amines (primary,
secondary, or tertiary) and primary aliphatic amines does not occur rapidly
at temperatures in the 25-50°C range.
The reaction of ammonia and aliphatic amines with S00 under anhydrous
st\ 2
conditions has been studied spectrochemically^ . Ammonia reacts spontane-
ously with the evolution of heat to form thionylimide (HNSO), NH.+, HSO ",
S3 H J
S2°5 ' The reaction between methylamine and SO- produces methylthionyl-
amine (CH^NSO) and the pyrosulphite t(CH.jNH3)2S205] . The reaction between
dimethylamine and S02 was reported to give a material which could be puri-
fied by vacuum distillation without decomposition. 'This product was con-
sidered to be the molecular complex (CH3)2NH.S02 on the basis of spectral
-------
Sulfur dioxide and ammonia have been shown to undergo a redox reaction
which leads to ammonium sulfamate and free sulfur .
Bateman, Hughes, and Ingold*1 ' and Byrd^ ' have discussed the bonding in
amine-SC^ COItlP°un^s' It is agreed that the bonding is between sulfur and
nitrogen to form a charge-transfer complex:
Under severe conditions, S09 is a powerful oxidizing agent for many
(9) z
organic compoundsv ; for example, toluene is oxidized to benzoic acid, and
aniline decomposes explosively in the range of 200-400°C under 50-2500 atm
pressure. It is considered that under conditions suitable for sorption of
S0« from flue gases and regeneration of the sorbent, these reactions are
negligibly slow.
Other possible reactions which could lead to irreversible reactions of
organic amines with flue gas constituents include:
1. Oxidation of SO- to SO- by 0. - particularly in the
presence of water. It is known that in a solution of
sulfurous acid, in the absence of oxygen, dispropor-
tionation into sulfuric acid and free sulfur will occur
whereas in the presence of oxygen, formation of sulfuric
acid will occur^ .
Subsequent reaction with amino groups would form amine
sulfate salts which would not be thermally regenerable.
Redox processes between SO^ and oxides of nitrogen. It
is possible that oxidation-reduction reactions between NO, 0
and S09 could occur in a manner similar to the lead
-------
In addition, the complex equilibria between the oxides of
nitrogen, oxygen, and water can lead to the formation of
nitric acid which would compete with S02> probably irre-
versibly, for amine sites.
3. Reactions between oxides of nitrogen and amines. Dragq
and coworkers have investigated the interaction of nitric
oxide with amines^ , and reported that primary and sec-
ondary amines react readily with NO according to the follow-
ing scheme:
R2NHNO
The aliphatic tertiary amine, trimethylamine, forms an addition compound,
(CH«)~NN909, which is unstable. Weaker tertiary amines such as pyridine, a.nd
highly hindered primary amines such a t-butylamine, do not form addition
compounds.
(12)
Nitrogen dioxide has also been reported to react with amines to form
addition compounds which are stable only at low temperatures, and as with NO,
do not form with hindered amines, such as a, a'-lutidine. Probably more
important than competition with S0_ for sorption sites however, is the fact
that N02/N20^ reacts with a variety of organic materials; a few of these
reactions are listed below:
1. Nitration of paraffins, alcohols, ethers.
2. Formation of nitrous acid esters with alcohols.
3. Nitration of aromatic nuclei, even at 80°C.
4. Formation of nitrous acid and subsequent reaction with
amines:
HN02 + RNH2 - > ROH (deamination)
-------
Although tertiary amines are generally considered
unreactive toward nitrous acid, they sometimes react
with cleavage of an alkyl group:
R2N-CHRj, + HONO
Although many of these reactions do not occur at appreciable rates at
low temperatures (< 100°C) there may be sufficient reactivity over long
periods of use to severely damage a sorbent system.
Although these side reactions are all possibilitiess it is difficult to
estimate the relative importance they might have (depending on amine type,
temperature, humidity, concentration of various gases, etc.) or how they
might quantitatively affect regenerability.
Clearly, tertiary amines are to be preferred on the basis of greater
stability toward a variety of side reactions with S0_ and other agents, and
aliphatic amines are preferable because of their higher basicity and their
generally greater resistance to oxidation by most oxidizing agents, such as
atmospheric oxygen. This is not to imply, however, that tertiary aliphatic
amines are completely unreactive with respect to oxidation; nonperoxidic
agents can convert tertiary aliphatic amines to enamines, which are then
subject to further oxidation or hydrolysis:
[01 H2°
R2NCH2CHR' L J< > R2N-CH=CHR' — =L->> R'CH2CHO + R2NH
Peroxidic agents or ozone convert tertiary amines to N-oxides:
The substitution of alkyl groups on nitrogen generally results in a
maximum in base strength for the secondary amine and a decrease in basicity
for the tertiary amine. The order of base strength is:
R2NH > RNH2 > R3N > NRj
The basicity of an amine in aqueous solution is not necessarily directly
-------
amine acts as a proton acceptor.
R N + H20 N R3NH+ OH"
The same would be true in reactions with Bronsted acids.
R3N -f 'HX
Reaction with electron acceptors larger than the proton (e.g. S02)
J f- \ J *•
involves close approach of the acceptor molecule to the nitrogen, and steric
effects (especially the so-called "F-strain") become the dominant factor in
determining the stability of the complex.
For example, the stability of amine-'trimethylboron adducts decreases with
increasing substitution of alkyl groups on the nitrogen much more rapidly
than does the base strength toward protons.
\ n D . xTn 1
R3B + R; N
\
Amine \ KJ4 J(CH,),B adduct]
EtNH2 5.6 x 10"4 0.0705
Et-NH 9.6 x 10"4 1.22
-4
Et3N 9.7 x 10 Too highly dissociated to
be measured.
Hence, the order of basicity toward the reference acid trimethylboron becomes:
If the reference acid is tri-tert-butylboron, steric factors become completely
dominant, and the order of stability is:
RNH2
A discussion of this behavior can be found in the work by Brown and
(13)
collaborators . Although tertiary amines are less basic than the corres-
ponding secondary amines, particularly toward bulky Lewis acids, the much
higher stability toward side reactions would make them the most promising
sorbents, providing the alkyl groups around the nitrogen are small.
-------
III. SCREENING OF BASIC SORBENTS
Our first task was to select basic nitrogen-containing polymers which
were known to be cospinnable with fiber-forming materials, and to obtain
information regarding the effects of temperature, S0_ concentration in the
gas phase, and moisture upon the capacity and rate of sorption. A styrene/
dimethylaminopropylmaleimide copolymer (SSQ) and a copolymer of 2-vinyl-
pyridine with 2-methyl-5-vinylpyridine (GIQ) were chosen as typical materials
covering a wide range in basicity from the strongly basic tertiary alkyl
amine in SSQ to the relatively weakly basic pyridine nitrogen in GIQ.
CH2 - CH - CH2 - CH
SSQ ^N\ GIQ
CH3 CH3
In addition to the above polymers, a number of other polymers were
evaluated and the results are included in this section.
A. STYRENE/DIMETHYLAMINOPROPYLMALEIMIDE COPOLYMER (SSQ) AND
2-VINYLPYRIDINE/2-METHYL-5-VINYLPYRIDINE COPOLYMER (GIQ)
In order to gain some information regarding the S09 sorption properties
of SSQ and GIQ, both as the pure materials and in fibers, samples were ex-
posed to pure SO at atmospheric pressure, and the weight gain was determined.
1. S09 Capacity of Pure Sorbents
One gram samples each of SSQ and GIQ were placed in a chamber containing
pure SOy gas at room temperature. The samples were weighed at two intervals
to determine SO- uptake.
-------
g/g
0.435
0.459
so2
Pure SSQ
mole/equiv
1.95
2.06
Uptake
g/g
0.461
0.976
Pure GIQ
mole/equiv
0.81
1.72
3 hours
24 hours
These data indicate that both of these materials form 2:1 complexes with
S0« under the pure gas. However, these secondary complexes appear to be weak
and do not form to any appreciable extent under conditions where the S02 con-
centration is lower (see Sec. VI). In addition, such 2:1 complexes have not
been reported for non-polymeric materials and are therefore of little import-
ance to the present work (see Sec. II).
2. SO Capacity of Sorbents in Polypropylene Fiber
Five-gram samples of fibers containing (1) 2 parts per hundred (phr) of
SSQ, (2) 2 phr GIQ, and (3) no S09 acceptor, were exposed to pure SO- in the
same fashion as above. At each time interval the weight gain of the blank
was subtracted, and the S0_ sorption of the acceptor in each fiber was calcu-
lated.
S02 Uptake
1.75 hours
18.75 hours
SSQ in
Polypropylene Fiber
g/g
J.410
).410
mole/equiv
1.84
1.84
GIQ in
Polypropylene Fiber
g/g
0.529
0.801
mole/equiv
0.93
1.41
These results show that the S02 sorption behavior of SSQ and GIQ under
pure S02 in polypropylene fiber is parallel to that of the pure materials.
The fibers used here were spun as part of another program, and contain much
lower levels of acceptor than used in later work.
3. SQ2 Capacity of SSQ in a Flow System
The effectiveness of dry SSQ in removing SO from a S02~N2 gas stream
containing approximately 11,000 ppm S02 was determined at several tempera-
-------
tures in the range 35-95°C. In each case a 1 g fixed bed of 60-100 mesh
SSQ was used.
At each temperature the S02 concentration in the effluent gas remained
below 50 ppm (our lower limit of detection) for a certain time, then rose
rapidly to the feed concentration (see for example, Figure 1).
The capacity of SSQ shows a strong temperature dependence. The break-
through point (where the S02 concentration in the effluent gas reaches that
of the feed gas, and no further sorption of S02 takes place) is reached
earlier at higher temperatures, and the total SO- capacity is consequently
smaller (see Experimental, Sec. X, for the relationship between breakthrough
time and capacity). For example, the breakthrough point at 35°C occurs at
approximately 0.190 g/g whereas at 95°C it occurs at 0.033 g/g. The data
(Table I) also show that at 35°C the SSQ bed is capable of reducing the S02
TABLE I
Sorption of S02 by SSQ (g/g)
Dry Presence of Water Vapor
1st Appearance Breakthrough 1st Appearance Breakthrough
T°C of S02 (>50 ppm) Point of S02 (>50 ppm) Point
95
85
75
55
35
0.0056
-
0.0636
0.116
0.162
0.0357
0.0602
0.110
0.139
0.187
0.0535
-
-
0.145
0.169
0.0781
-
-
0.185
0.196
Feed gas concentration approx. 11000 ppm
Flow rate 21.2 ml/min.
concentration in a gas stream containing approximately 11,000 ppm at a flow
rate of 21.2 ml/min. to less than 50 ppm until a capacity of 0.162 g/g is
reached. At 95°C, however, SSQ can reduce the SO- concentration in a similar
gas stream to less than 50 ppm until only 0.0056 g/g is sorbed. In addition,
-------
11000
10000
8000
6000
4000
2000
>50
Feed gas level
E
CL
CL
eg
O
UJ
10
JL
80 100 120 140 180 200 220 240 260 280
Time in Minutes
-------
if a sample loaded with S02 at a lower temperature is raised to a higher
temperature, SI
later section.
temperature, S09 is driven off. Regeneration behavior is described in a
4. S(>2 Capacity of SSQ in the Presence of Water Vapor
The reaction of amines with SO- in the presence of water should lead
to a more favorable equilibrium for the reaction, since S09 would be more
tightly bound as the bisulfite or sulfite salt than as the association
! (1)
complex .
Samples of SSQ were equilibrated with water vapor at room temperature,
then treated with a water-saturated S09-N9 gas mixture and the SQ_ content
^. £, £.
of the effluent gas stream determined as a function of time. The S0_ capa-
city was significantly increased, particularly at higher temperatures. At
lower temperatures, where the capacity of the dry polymer is already high,
the improvement was much smaller, as is to be expected (Table I).
B. POLYETHYLENEIMINE (PEI)
Montrek 12 (Dow Chemical Co.) polyethyleneimine, a liquid, was placed
on Chromosorb W (a calcined silica) as a support material at a level of 1 g
Montrek on 5 g of Chromosorb. At 35°C the breakthrough point occurred at
16.5% of the 1:1 stoichiometric value (Table II). However, since poly-
ethyleneimine has a very low equivalent weight (43), this is a relatively
TABLE II
Sorption of SC>2 by Various Materials (g/g)
PEI on Chromosorb
Benzoylated PEI
GIQ
Aminopolyurea
IRA-400
VTQ
SOo-N2 mixture 10,000 ppm S02
Flow rate 21.2 ml/min.
15
35°
150°
35°
35°
35°
31°
95°
75°
Dry Water Vapor
0.246 0.990
0.91
0.100
0.0099
0.074
0.08
0.138
0.188
Stoichiometric
1:1 Capacity
1.47
0.260
0.568
0.164
-------
high capacity on a weight basis (0.246 g/g). At 150°C. considerable sorption
of SO. occurred as determined by integration of the time vs. effluent S02
curve (0.91 g/g, 60.9% of stoichiometric), but poor correlation with the
weight gain (0.221 g/g, 14.8% of stoichiometric) was obtained. Examination
of the sample after the run showed that extensive darkening and degradation
had taken place, and loss of volatile decomposition products probably caused
the discrepancy.
Polyethyleneimine has been spun into polypropylene fiber at high temper-
atures (about 300°C) without significant decomposition. The decomposition
observed in the present work is probably due to reaction with S02 at high
temperatures. The higher capacity for S02 at 150° compared to 35° is
probably due to increased rate of side reactions as discussed in Section II.
Although in its present form PEI is not useful it seems that polyethylene-
imine merits further consideration since: (1) polyethyleneimine and its de-
rivatives made by reaction with alkyl halides, anhydrides, etc. are cospinn-
able with polypropylene; (2) its low equivalent weight gives high potential
S0_ capacity.
The presence of moisture has a strong effect upon the SO. uptake of
unsubs-tituted PEI. At 35°C, dry PEI .on Chromosorb reaches the breakthrough
point at a capacity of 0.246 g/g, whereas in the presence of moisture, the
breakthrough occurs at a capacity of about one gram of SO- per gram of PEI
(Table II). A blank run with Chromosorb W support material alone under the
same conditions showed only an insignificant SO- absorption.
We prepared a solid derivative of PEI by reaction with benzoyl chloride.
The softening point of this product is 90°C, and the equivalent weight,
determined by titration of the amino groups, is 246.3 (the equivalent weight
of the unsubstituted material is 43). Thus, somewhat more than half of the
amine nitrogen was benzoylated, and is inactive. This material was evaluated
in the presence of water vapor, and breakthrough was reached at 38.6% of stoi-
chiometric capacity, a loading of 0.100 g SO- per gram of sorbent. Thus it
appears that benzoylation of polyethyleneimine to this degree not only raises
the equivalent weight, thereby reducing the potential capacity, but the remain-
ing free amino groups are less reactive toward SO-, probably because of steric
effects.
-------
Polyethyleneimine is 9 highly branched polymer containing primary,
secondary, and tertiary amino groups. Since primary and secondary amino
groups lead to undesirable side reactions, and may be responsible for the
extensive decomposition observed at elevated temperatures, the most prom"
ising course may be to react the polymer with a lower-alkyl alkylating
agent to convert all the amino groups to tertiary amino groups. This
would, in addition, lead to only a relatively small increase in equiva-
lent weight. This is planned for fixture work.
C. POLYVINYLPYRIDINE
Polyvinylpyridine (GIQ) was exposed to a 11,000 ppm SO--N- gas stream
at 35° in the same manner as SSQ. The breakthrough occurred at only 0.0099
g/g, less than 27. of the stoichiometric 0.568 g/g (Table II), This is
probably due to the much weaker basicity of the pyridine nitrpgen as com-
pared to the tertiary aliphatic nitrogen of SSQ, and a consequently higher
SOj partial pressure of the complex.
D. AMINOPOLYUREA
A polyurea containing tertiary amino groups was prepared in the labor-
atory by the reaction of 2,4-tolylenediisocyanate w;Lth N-methyl-bis(amino«-
propyl) amine:
The SO sorption characteristics at 35° are shown in Figure 2. Comparison
with Figure 1 (SSQ at 35°) shows that this material does not give as sharp
a rise to the breakthrough point as does SSQ. This indicates that the
rate of sorption, at 35° at least, is slower. In addition, only 45%
(.074 g/g) of stoichiometric capacity was reached (Table II) whereas SSQ
reached almost 85% (0.187 g/g) of stoichiometric under the same conditipns.
-------
CO
12000
11000
10000
8000
6000
4000
2000
<50
Feed gas level
120 .140 160
Time in Minutes
180 200 220 240
-------
E. ANION EXCHANGE RESIN
Dry anion exchange resins have been studied as potential SCL sorbents
Of the materials studied, the best performance was obtained with Rohm and
Haas IRA-400, a commercial quaternary ammonium chloride anion exchange resin,
but the S02 concentration range in this study was much higher (.018 - 1 atm)
than that used in our work. We considered it of interest to evaluate this
resin by our method to see how effective ion exchange resins are in compari-
son with the free-base polymers we have been evaluating. Results with poly-
ethyleneimine and SSQ run under similar conditions show that the anion
exchange resin is definitely inferior to either SSQ or PEI (Table II). The
quaternary ammonium halide is a salt, and reaction with S0_ would require
displacement of the halide:
Iffi (C) o x Iff) Q
-iTciu + so =^ - N^ksor + uci
An anion exchange resin in the hydroxyl form, however, should react readily
with S02:
Resins in the hydroxyl form have low thermal stability, however.
Free base polymers can form salts directly with S02 in the presence of
water:
—NH HS03
Ion exchange resins, therefore, must depend upon secondary forces to
bind S0_, whereas with free base resins true complex formation takes place.
F. STYRENE-METHYLVINYLETHER-DIMETHYLAMINOPROPYLMALEIMIDE
TERPOLYMER (VTQ)
VTQ, a terpolymer of styrene, methylvinylether, and dimethylamino-
propylmaleimide (mole ratios 1:1:2), has been evaluated in the absence of
water vapor at 95°C and 75°C. This material has a 1:1 stoichiometric S0_
-------
capacity in the same range as SSQ (0.249 g/g for VTQ, 0.223 g/g for SSQ).
Its breakthrough point occurred at 0.138 g/g at 95°C and at 0.188 g/g at
75°C (Table II). This is substantially better performance than that of
SSQ. Further work with VTQ is in progress.
-------
IV. FIBERS
A. POLYPROPYLENE-POLYETHYLENEIMINE FIBERS
1. Spinning
Spinning conditions used were those designed for low levels of polymeric
additives. Compatibility with the polypropylene (Hercules Pro-fax 6501 +
heat stabilizer) was satisfactory, and good fibers of about 15 denier were
obtained. These fibers contained 9-11% of PEI by weight. Some exploratory
work will be required to determine the optimum conditions and techniques
of S02 acceptor loading and fiber spinning for obtaining fibers having the
maximum amounts of S0» acceptor consistent with good fiber properties.
2. Effect of Temperature
S0? sorption experiments were run on an undrawn 9.17, PEI fiber in the
presence of moisture at temperatures of 35°, 123°, and 150°C. The respective
sorptions are shown in Table III. At the two higher temperatures a darken-
ing of the fiber occurred, indicating decomposition,
TABLE III
SO- Sorption of Polypropylene Fiber Containing 9.1% PEI
In Presence of Water Vapor
Temp.
123
150
35
35
35
g/g on fiber
0.0127
0.0189
0.0167
0.00435
0.0253
g/g on sorbent
0.141
0.208
0.184
0.0478
0.228
(degradation apparent)
(degradation apparent)
(dry)
(11.1% PEI in fiber)
10,000 ppm SO- in S02-N2 mixture
21.2 ml/minute flow rate
-------
3. Effect of Moisture
S02 sorption of an undrawn 9.1% PEI fiber at 35°G, both dry and in the
presence of moisture, was determined. The respective sorptions were 0.00435
g/g and 0.0167 g/g (Table III). This increased sorption of a PEI fiber in
the presence of moisture is in agreement with the results obtained from
samples of pure PEI and SSQ equilibrated with water vapor at room temperature.
High hydrophilicity of fibers may be advantageous in improving their per-
formance as S0_ sorbents.
4. Drawn vs. Undrawn Fibers
Drawing of fibers increases the orientation of the polymer chains.
Although the changes in diffusion and sorption properties are complex, and
depend upon many variables, the general effect of orientation is to decrease
the permeability of the polymer matrix to substances such as water '
However, this decrease in permeability is greater in the direction parallel
to the direction of orientation than normal to it. To gain some information
on what the magnitude of this effect might be, the SO- capacities of drawn
(3:1) and undrawn samples of a 9.1% PEI fiber in the presence of moisture at
35°C were compared. The respective S0« sorptions were 0.0168 g/g and 0.0147
g/g. The slightly higher S0_ sorption capacity and faster rate (Figure 3) of
the drawn fiber may be attributable to its greater (1.73-fold) surface area.
It would appear that this effect was sufficiently great to more than offset
any decrease in sorption which might be expected to occur due to an increase
in the orientation of the fiber. Further work will be needed to establish
the generality and magnitude of the effect of orientation on rate and capacity,
5. Effect of Polyethyleneimine Level
In a comparison of the S02 sorption capacity of two undrawn polypropyl-
ene-PEI fibers containing different amounts (1.17. and 11.1%) of PEI, at 35 °C
in the presence of moisture, a positive correlation between the PEI level and
the S02 sorption capacity was observed. The S02 sorptions were 0.0167 g/g and
0.0253 g/g (Table III).
-------
N>
10000
8000
6000 h
4000
2000
0
A PEI on CHROMOSORB 21.2 ml/mm.
® 9.1% PEI 21.2 ml/min. J-/ draw ratio
A 11.1% PEI 2l.2ml/min. undrawn
o 9.1 % PEI 21.2 ml/mm. undrawn
40
80
100 120 140
Time in Minutes
160 180 200 220
Figure 3. Effect of flow rate, PEI level, and draw ratio upon rate of S02 sorptioa
-------
6. Polypropylene Fiber
As a control, a similar fiber without PEI was spun, and its sorption of
SO- determined at 35°C in the presence of moisture. Sorption was negligible,
and the weight gain of the sample was only 0.00093 g/g.
7- Sorption Rate
The sharpness of the breakthrough curve, which represents the sharpness
of the advancing front of S0? through the sorbent bed, is dependent upon the
rate of sorption of S0? by the sorbent (assuming the chemical equilibrium
between the gas and sorbent phases ;Ls favorable).
Rate-controlling factors can be: 1), the rate of reaction of S02 with
the acceptor polymer; 2), the rate of diffusion of S0? through the sorbent
matrix; 3), the surface area of the sorbent; 4), flow rate of gas through the
bed. For example, if any of the above factors were highly unfavorable, a
curve such as (1) would be expected, whereas if equilibration between gas and
sorbent were extremely rapid a curve such as (2) would be expected. Of course,
the more closely actual curves approach (2), the less SO™ "leakage" and the
more effective the system.
Effluent
so2
Concentration
time
Dry SSQ and PEI on Chromosorb, for example, give curves which approach
type (2) (see Figures 1, 3).
-------
The effect of flow rate, PEI level, and draw ratio upon the shape of the
breakthrough curves was briefly investigated (Figure 3). Increased draw ratio
and increased PEI level both gave more favorable curves. In the case of the
drawn fiber, the greater surface area and shorter diffusion path through the
fiber are probably responsible, and in the case of the higher PEI level, the
diffusivity of SO^ in the fiber may be higher since the mixture contains more
PEI and less polypropylene, although increased hydrophilicity may also be a
factor.
B. POLYPROPYLENE-SSQ FIBER
In one investigation of the SO- sorption capacity of an undrawn poly-
propylene fiber containing 8.337. SSQ in the presence of moisture; at 35°C, a
high percentage of the 1:1 stoichiometric capacity was attained. Although the
sorption did not proceed with great rapidity, it steadily increased, and at
termination of the experiment a loading of 0.0136 g S0»/g fiber (0.163 g/g on
SSQ) had been reached (see Figure 4). Although the rate was relatively slow
compared to that of SSQ powder (cf Figure 1) the capacity compares favorably.
Considerations concerning the rate of SO- sorption were discussed in the pre-
vious section, and it is probable that the rate 6f sorption by polypropylene-
SSQ fibers can be increased substantially by the use of the same techniques.
-------
tsJ
12000
11000
10000
8000
6000
4000
2000
>50
Feed gas level
20 40 60 80 100 130
Minutes
160
190
220 250
Figure 4. SG2 sorption of undrawn polypropylene-SSQ fiber
-------
V. REGENERATION
A. POLYPROPYLENE-POLYETHYLENEIMINE FIBER
The performance of an undrawn polypropylene-PEI fiber (11.1% PEI) in the
presence of moisture has been determined during an SO, sorption-regeneration-
sorption operation (Table IV). In the initial sorption, breakthrough occurred
at 0.0252 g/g on fiber. During the regeneration only 0.0097 g/g of the sorbed
S02 was released, probably indicating irreversible reaction between S0? and
PEI. A second sorption under the same conditions gave a S00 uptake of only
TABLE IV
Regeneration of PolypromOpJie-PEI Fiber
(1)
(2)
(3)
(4)
Sorption
Loss on regeneration
Residual S02
Second sorption
Total (3 + 4)
g/g on Fiber
0.0252
0.00971
0.0155
0.00874
0.0242
g/g on PEI
0.228;
0.0873
0.141
0.0787
0.220
Sorption - 35°C, 10,000 ppm SO_-N_ mixture, 21.2 ml/min, water
vapor.
Regeneration - 95°C, N-, 21.2 ml/min, dry.
0,00874 g/g. Possible reasons for this behavior have been discussed earlier
in this report.
B. MULTIPLE SORPTION-REGENERATION WITH SSQ
In accordance with plans to study the efficiency of S02 acceptors in
multiple sorptions and regenerations, the behavior of 60-100 mesh SSQ in the
presence of moisture was investigated in this respect. Data were obtained
on four S02 sorptions at 55°C and three intervening regenerations at 95°C.
-------
Feed gas concentrations of 10-12,000 ppm SO,, (saturated with water vapor)
were used for the sorption cycles at 55'C at a flow rate of 21.2 ml/min. Re-
generation was accomplished by raising the sample temperature to 95°C.and
flushing with dry nitrogen at the same flow rate.
There is a loss in SO,, capacity with succeeding cycles. A capacity of
0.181 g/g on the first sorption was reduced to 0.112 g/g on the fourth sorp-
tion. However, the removal of S02 on each regeneration appears to be fairly
complete (Figure 5). This would suggest a deactivation of the sorbent by the
process, rather than a cumulative irreversible sorption of S02(as occurs with
PEI). A possible explanation for this is hydrolysis, leading to subsequent
internal salt formation:
N
Evidence that water is important in the deactivation process is given by
work now in progress where it appears that dry SSQ can be carried through
several cycles without significant loss in capacity.
The data are somewhat erratic, especially in the second regeneration and
the third sorption. This may be due to changes in the moisture content of
the sample because of changes in structure which are occurring.
Regeneration of SSQ
Residual S02 After
Cycle No. Sorption (g S02/g SSQ) Regeneration (g/g)
1 0.181 0.0056
2 0.158 -0.0201
3 0.089 0.0167
4 0.112
Sorption - 55°C, 10-12,000 ppm S02-N2 mixture, 21.2 ml/min,
water vapor.
Regeneration - 95°C, N-, 21.2 ml/min, dry.
-------
Sorption
Regeneration
Flow Rate
55°, IO-/2000 ppm SOS, Water vapor
95°, N2, dry
21.2 ml/mm.
o
CM
o>
.20
,18
.16
.14
.12
.10
.08
06
.04
.02
0
',02
— •
— i
••• i
i
- i
" i
"~ A
-f
• j
i
" J
l
i
"i
_'
'
_'
•
1 ;
1
T
f
<
••
> ->
A
ii
'
1
1
!
f /
/
i V
Ja
/
/
/
/
/
/
/
/
/
ww»
- J
r
A
^•r
<
'
'
•1
P /
, S02
Sorbed
*
/
/
, /
/ f ^
/ '
/
/
'
J
f
/
jr
/
/
/
( 1
/
/
r ^/
S02 Remaining
^ after Regeneration
i i I
4 CYCLE NO.
Figure 5. Sorption-regeneration of SSQ
-------
VI. SORPTION ISOTHERMS
The equilibrium constant for gas-solid reactions of the type shown below
is equal to the partial pressure of the gaseous component
=^ Ca°(s)
This is in agreement with the phase rule, C - P + 2 = degrees of freedom. In
the above example there are two solid phases and one gas phase, and there are
two components, hence there is one degree of freedom, i.e., pressure or temper-
ature. As a result, if the partial pressure of CO- in the system is greater
than the dissociation pressure of CaCO- at a given temperature, the reaction
will proceed to completion.
In the case of reaction of SO- with a polymeric amine
S°
2(g)
so2
this treatment does not apply, since there is only one gas phase and one solid
phase present at any time. A polymeric amine chain reacted to any degree of
completion would be physically inseparable from other chains where the reaction
has proceeded to either a greater or lesser degree, i.e., a solid solution is
formed. This system has two degrees of freedom, and at constant temperature
the amount of gas sorbed will vary with the partial pressure.
The basic difference between an adsorption process and our SO- sorption
system is that in the Langmuir equation the maximum capacity corresponds to a
monomolecular layer of adsorbed molecules on the surface of the adsorbent,
whereas in our system the maximum capacity is related to the stoichiometric
amount of SO- which can react with the amine sites in the acceptor polymer.
The sorption of SO by dry SS.Q was determined at temperatures of 55, 75,
and 95°C under varying SO- partial pressures. Results are shown in Table V.
The data were obtained by treatment of 1 g samples of dry SSQ with S02~N2 gas
streams containing various levels of SO- at a flow rate of 64.0 ml/min. When
the SO- concentration in the effluent gas was the same as in the feed gas,
-------
TABLE V
S02 Capacity of Dry SSQ
95°C
p S02 (atm) Weight Gain (Y) (g/g)
0.00327
0.00526
0.00752
0.0108
0.0170
0.0113
0.0196
0.0297
0.0360
0.0585
£/Y
0.290
0.268
0.244
0.300
0.291
0.000435
0.00382
0.00526
0.00900
0.0171
0.0219
75°C
0.0081
0.0501
0.0644
0.0974
0.119
0.134
0.054
0.0762
0.0817
0.0924
0.144
0.163
0.000435
0.00526
0.0117
0.0150
0.0195
55°C
0.0029
0.129
0.149
0.161
0.163
0.015
0.0407
0.0787
0.0933
0,120
equilibrium was attained. The S0_ uptake of each sample was determined by
weight gain. Plots of g S02/g SSQ vs. equilibrium SO- pressure are shown
in Figure 6. Sorption at 55° and 75°C shows typical type I (Langmuir)
curves in the S02 pressure range studied, and a plot of p/Y vs. p gives a
straight line (Figure 7). Capacities at.various pressures were calculated
from the 55°C plot and are shown in Figure 6.
-------
U)
0.16
0.14
0.12
0.10
.08
.06
.04
.02
0
.002
006
Equilibrium S02 Pressure (atm)
,010
.014
.018
Figure 6, Dissociation pressure of SSQ-S02 reaction product.
-------
.16
.14
.12
.10
.08
.06
.04
.02
CL
75<
P S09 (atm)
1
1
I
1
I
1
1
I
-------
At 95°C, however, the capacity vs. pressure curve is nearly linear, and
the points on the p/Y vs. p plot are badly scattered. It may be that the
curvature of the capacity vs. pressure curve is so slight that it is not dis-
tinguishable with our experimental data, or Langmuir behavior may not be followed
at this temperature,
The Langmuir adsorption isotherm can be expressed as
Y bp
Y = 1 + bp Y = g S02 sorbed/g SSQ
Y = maximum g S02/g SSQ
it follows that b = constant
p = partial pressure SO-
The slope of the p/Y vs. p plot equals 1/Y and the intercept equals 1/Y b.
75° 55°
Slope (l/Ym)* 5.07 5.49
Y 0.197 g/g 0.182 g/g
m !
Intercept ( b )* 0.0533 0.0126
b m 95.3 436.7
* Determined by the method of least squares.
The maximum capacity of 0.182 g/g at 55°C is 81.6% of the stoichiometric
capacity of SSQ (0.223 g/g) and, of course, has meaning only if the Langmuir
isotherm is followed at much higher S02 pressures.
A series of log p vs - plots at various loadings (isosteres) were
made from the data, and AH of sorption calculated from the slope of the lines
for several loadings of SO, on SSQ:
-------
Isosteric Heat of Sorption of S02 °n Pyy SSQ
Loading, g/g AH. Real/mole
0.02" - 18.3
0.06 - 17.9
0.12 - 14.7
These figures are in the appropriate range for this type of process,
and show, as expected, that the heat; of sorption is lower at higher
loadings.
-------
VII. DISTRIBUTION COEFFICIENTS
A rapid method for obtaining thermodynamic distribution coefficients
between SO^ in the gas phase and S0~ in solid sorbents has been investigated.
This method involves the use of the candidate sorbent as a substrate in a gas
chromatography column^ ' .
A sample of S09 is injected into a helium carrier gas stream and the
le of gi
determined.
volume of gas required to carry the sample through the column (V ) is
If it is assumed the pressure drop through the sample is negligible
V (S0_ in solid ph^se)
V ~ (S09 in gas phase)
S £•
t = retention time
r
F = carrier gas flow rate, ml/min.
V = retention volume
r
V = volume of solid absorbent
s
C = distribution coefficient
Determinations were carried out at three temperatures using a 1-1/2" x
1/4" O.D. column packed with 0.175 g 60-100 mesh SSQ.
C
4800
1330
740
A plot of log C vs. 1/T yields a straight line (Figure 8). Work is in
progress to determine the usefulness of these data in evaluation of sorbents.
-------
8 r-
7 -
6 -
5
4 h
o
o»
o
1000
9
8
7
6
5
4
i/T°Kxi03
I !
2.5
2.6
2.7
2.8
2.9
3.0
Figure 8. Log distribution coefficient vs. 1/T for SSQ
-------
VIII. EFFECT OF FLOW RATE
One-gram beds of 60-100 mesh SSQ were contacted with a 5200 ppm SO.-N,
gas stream at several flow rates at 55°C. The loading at breakthrough was
determined.
Relative Space
Velocity
Flow Rate
ml/min.
21.2
43.0
87,5
177.0
271.0
g S02/g SSQ
0.139
0.129
0.142
0.128
0.129
1
2.06
4.12
8.33
12.78
The capacities at each flow rate agree within + 5 percent, indicating
that equilibrium is attained in each case. If the density of SSQ is con-
sidered to be about 1, the absolute space velocity at a flow rate of 271.0
ml /rain, is 16,260 ml/hr/ml. Further experiments are being carried out at
higher flow rates to determine where non-equilibrium conditions begin.
-------
IX. EFFECT OF NO-
Although N02 is a minor constituent of flue gases, higher levels than
those originally present in the gas (~. 00025%) might occur because of further
oxidation of NO in the sorbent bed. In view of the potential reactivity of
N02 with amine sorbents a brief investigation of the gross effect of NO.
upon SSQ was undertaken.
A sample of dry SSQ was treated with N02 until 68.4% (0.1601 g/g) of
the stoichiometric amount was taken up. This sample was then exposed to a
5260 ppm S02-N2 gas mixture at a flow rate of 42 ml/min at 55°C. Break-
through was reached at 0.026 g/g. Another sample, handled in the same way
but without the N02 treatment broke through at 0.129 g/g. Although the NO.
treatment reduced the SO. capacity of the sample, the total number of amine
groups reacting (with N02 and SO,,) was greater (80.1% vs. 58%). This may be
due to either partial displacement of the NO. by SO. or the affinity of the
remaining free amine sites for SO. is not reduced by the presence of NO. on
other sites as much as if they were occupied by SO..
Although this experiment indicates that the presence of NO may have
serious effects upon the SO. capacity of materials such as SSQ, it would be
much more informative to conduct further experiments in the presence of
relatively low levels of NO., NO, and other gases and SO. simultaneously;
this would present a more competitive situation and would approximate the
conditions of service more closely. In addition, kinetic effects, such as
slow displacement of NO. by SO. (which would lead to an apparent equilibrium
favoring the NO. complex over the S02 complex) would be minimized. Modifica-
tion of our present apparatus will be required to carry out this work.
Experiments to determine the effect of other flue gas components are
under way.
-------
X. EXPERIMENTAL
The apparatus (Figure 9) for the determination of SO. concentration con-
sists of a F&M Model 720 dual-column, programmed-temperature chromatograph
with a Leeds and Northrup Speedomax G recorder. Helium carrier gas is used
at a flow rate of 40 ml/min. A 25 ml. gas sampling loop is used with a Perkin-
Elmer 154-0068 Precision Gas Sampling System. The chromatographic column is a
1-foot by 1/4-inch stainless steel tubing packed with W. R. Grace grade 12,
60-100 mesh silica gel. At 135°C the retention times for nitrogen and SO,
are 48 sec. and 209 sec. respectively. Gas mixtures of S02 and nitrogen
containing 5260 ppm, 435 ppm, and 41 ppm SO- were obtained from the Matheson
Company. Other gas mixtures are prepared in the laboratory by blending
Linde LC-3 99.997% nitrogen gas and Matheson anhydrous grade 99.98% S02- Flow-
meters are used to set the approximate relative amounts of SO- and N-. The
actual concentration of SO- is determined by the area of the SO- peak in the
gas chromatogram. The relationship of peak area to concentration was estab-
lished over a wide range of SO- concentration by simultaneous determination of
the SO- content of a known volume of gas by reaction with standardized iodine
solution.
The apparatus for exposing the sample to the gas stream consists of a
fan-circulating air oven containing a 6-1/2-foot by 1/4-inch coil of stainless
steel tubing to bring the gas mixture to thermal equilibrium. This leads to a
14 cm by 10 mm cylindrical stainless steel chamber. A weighed amount of the
candidate SO -acceptor is placed in this chamber and sandwiched between two
wads of glass wool. The gas passes through this chamber, then out of the oven
to the gas sampling loop of the G-C.
The gas containing 41 ppm SO- gives a barely discernible peak on the G.C.;
this represents our lower limit of detection.
In each run the SO- concentration in the feed gas was set at the desired
concentration. The exact concentration was determined by the gas chromato-
graph. In most cases the flow rate of gas through the samples was 21.2 ml/min.
All SO acceptors were screened to 60-100 mesh particle size.
-------
Flowmeter
Sample
Chamber
Mixing
Chamber
H2SO4
'Glass Woo I
Spray Trap
Waste
Flowmeter
H2O
Vaporizer
Gas
Chromatograph
-------
Plots of SO- concentration in the effluent gas vs. time were made and the
SO- taken up by the sample at various time intervals was determined by graphical
integration of the area between the curve and the line drawn at the feed gas
concentration. In most cases good correlation between the weight gain of the
sample at the breakthrough point and the graphically determined SO- sorption
was obtained.
In cases where water vapor was used, the gas stream was bubbled through a
gas washing bottle containing water at 25°C (puori = 0.032 atm) until equilibrium
H/U
between the SO- in the feed gas and the SO- in the water was attained. Fiber
samples were immersed in water for 5 minutes, then blotted dry. Powder samples
were equilibrated with water vapor by placing them in a desiccator containing
water at room temperature. In all cases, excess water on the samples was re-
moved by the gas stream in the first few minutes of flow.
-------
XI. FUTURE RESEARCH PROGRAM
A. SCREENING OF BASIC ABSORBENTS
Evaluation of absorbents will be continued with particular emphases on
(1) selectivity toward S02 in the presence of other acidic flue gas constit-
uents, (2) stability in operating environment (operating environment shall
consist of flue gas having the following compositions:
Volume Percent
Component Composition I Composition II
N2 74.9 81.7
C02 14.7
H20 7.25 2.6
02 2.8 4.2
S02 0.3 11.5
NO 0.05
X
Fly Ash 0.2 (wt) 3. (wt)
and temperatures in the range of 250-300 °F for sorption. The upper tempera-
ture limit shall be based on thermal regeneration requirements), (3) high
temperature capacity, and (4) reversibility.
B. FIBER OPTIMIZATION
The effect of such variables as fiber denier, cross section, loading of
absorbent, crystallinity and orientation, hydrophilicity, stability, sample
packing and configuration, etc. on rate and capacity will be determined.
C. KINETICS
The effect of SO- concentration, flow rate, temperature, and factors
related to fiber optimization upon rate of S02 absorption will be studied.
The data derived in this area will be obtained in such a menner that it can
be presented in terms of space velocities and partial pressures of S02>
-------
D. REGENERATION
The more promising fibers will be carried through multiple cycles of
absorption and regeneration to determine what changes in rate or capacity
occur, and to optimize regeneration conditions. Evidence of irreversible
reactions occurring will be sought. To the extent that such reactions occur,
either during sorption or regeneration, chemical regeneration techniques will
be investigated in addition to thermal regeneration methods.
E. SCALE-UP
Promising fibers will be spun in sufficient quantity to permit fabrica-
tion into filter packs for testing with gas mixtures simulating typical S02~
containing waste gases.
F. PLANT CONCEPTUALIZATION
In the latter stages of this work, a chemical engineer will be assigned
to this project to develop and outline plant designs for the practical appli-
cation of this method of SO. removal. Conceptual designs for a source yield-
6 ;
ing 2 x 10 SCFM flue gas of composition I (previous table) and a source
yielding 1 x 10 SCFM of composition II will be considered, if applicable.
Such conceptualization, will include heat and material balances, engineering
flow diagram, and sketches of novel and innovative equipment. The Project
Officer will be called upon to supply guidance and information in this work.
In addition, the above engineer and the Project Officer will be consulted
throughout the work to insure that the type of information and data being
accumulated is in a form suitable for engineering calculations, and further,
that no important points of information are being overlooked.
-------
XII. REFERENCES
(1). A. E. Hill, J. Am. Chem. Soc., 53 2598 (1931).
(la). See also ref. (4).
(2). Houben-Weyl, Methoden der OrganjLschen Chemie. 4th ed. Vol. XI/2
Georg. Thieme Verlag, Stuttgart, 1958, Chap. V, p 732.
(3). A. E. Hill and T. B. Fitzgerald, J. Am. Chem. Soc. .57 250 (1935).
(4). W. E. Byrd, Inorganic Chem. 1 762 (1962).
(5). A. B. Burg, J. Am. Chem. Soc. 65 1629 (1943).
(5a). J. R. Bright and J. J. Jasper, J. Am. Chem. Soc. j>5 1262 (1943).
(6). T. Hata and S. Kinumaki, Nature, 203 1378 (1964).
(7). J. L. Boivin, Can. J. Chem. _34 827 (1956).
(8). L. C. Bateman, E. D. Hughes, C. K. Ingold, J. Chem. Soc. 243 (1944).
<9). Selective Oxidation Processes, A.C.S., Washington, D.C., 1965, p. 52 ff.
(10). N. V. Sidgwick, The Chemical Elements and Their Compounds, Volume II
Oxford University Press, London, 1950, p. 908.
(11). R. Longhi, R. 0. Ragsdale, and R. S. Drago, Inorganic Chem. I 768 (1962)
(12). P. Gray and A. D. Yoffee, Chem. Revs. 5_5 1069 (1955)
(13). H. C. Brown and R. B. Johannesen, J. Am. Chem. Soc. 75 16 (1953)
and earlier papers in this series.
(14). R. Cole and H. L. Shulman, Ind. Eng. Chem. 52 859 (1960).
(15a). S. W. Laskoski and W. H. Cobbs, J. Poly. Sci ^6 21 (1959).
(15b). A. Peterlin and H. G. Olf, J. Poly. Sci. Part A-2 4 587 (1966).
(16); S. Glasstone, Textbook of Physical Chemistry, second edition
D. Van Nostrand Company, Inc., Princeton, N.J. 1946 pp 779, 845.
(17). Dal Nogare and Juvet, Gas-Liquid Chromatography, Interscience
Publishers, New York, 1962, pp 9, 364.
(18). M. A. Muhs and F. T. Weiss, J. Am, Chem. Soc. 84 4697 (1962).
------- |