EPA/600/A-94/029
Development of transition metal oxide-zeolite catalysts to control
chlorinated VOC air emissions
C. A. Vogela and H. L. Greene^
•Air and Energy Engineering Research Laboratory, U.S. Environmental Protection Agency,
Research Triangle Park, NC 277 1 1 (US A)
bDepartment of Chemical Engineering, University of Akron, Akron, OH 44325-3906
(USA)
Transition metal oxide (TMO)-zeolite oxidation catalysts have been developed to
control chlorinated volatile organic compound (CVOC) emissions. Research has been
initiated to enhance the utility of these catalysts by the development of a sorption-catalyst
system. Zeolites with a high Al/Si ratio (e.g., Y- zeolite) provide active acid exchange sites.
The exchanged and calcined zeolite is impregnated with the same or different metal and
calcined to form the TMO film. These TMO-zeolites provide shape selectivity and three
types of active sites: unexchanged HY (Bronsted) acid active sites, highly active exchanged
sites, and impregnated TMO sites. They can achieve over 95% destruction efficiency at
relatively low temperatures (i.e., 300 to 350°C) of a humid, low concentration CVOC, and
are very resistive to poisoning. In the sorption-catalyst system, the zeolite first would
physically adsorb the CVOC at room temperature. The system would then be heated to
promote the TMO-zeolite catalytic reaction. The Y-zeolites can physically adsorb about 5%
VOC but are highly hydrophilic. Silicalite is a zeolite with very little Al, can adsorb up to
15% VOC, and is hydrophobic. Impregnated Silicalite is an effective adsorbent and fair
catalyst, -
1. INTRODUCTION
CVOCs from both air streams and remediation of contaminated waste waters can be
destroyed economically through catalytic incineration. The catalytic incineration of humid,
dilute CVOCs from stripping contaminated waste waters is cheaper and more effective than
carbon adsorption (either of water or stripping air) [1], While noble metals can very
effectively destroy VOCs, they are very easily poisoned by the products (specifically HC1)
and by-products (specifically Ch) of oxidizing CVOCs [2].
TMO-zeolite catalysts can destroy dilute or concentrated CVOC streams, which are
wet or dry, at relatively low temperatures without the formation of dangerous by-products,
such as dioxins. In the 300 to 350°C range with a space velocity of 3000 h-i, the reaction is
first order with respect to the CVOC and over 95% efficient.
Experimental work has focused upon the following CVOCs: methylene chloride
(MeCh), carbon tetrachloride (CCU), and trichloroethylene (TCE). Cobalt exchanged
zeolites (Co-Y) are very effective in destroying single carbon CVOCs. However, the use of
chromium, either in the exchanged or impregnated form, appears needed to destroy double
bonded CVOCs. The TMO-zeolite catalysts strongly promote the deep oxidation of the
chlorine atoms in the CVOC to HC1 with no other chloride by-product other than C\2 being
detected. However, CO rather than CO2 often is favored.
-------
2. EXPERIMENTAL
2.1 Materials
The zeolites tested were H-Y, a faujasite-type zeolite with a three dimensional
structure and high AhOj/SiC^ ratio; and Silicalite, a zeolite with essentially no A^Cb. The
H-Y was 1/16-inch (1.6 mm) LZ-Y64 pellets obtained from Union Carbide. The Silicalite
was made into pellets by bonding its powder with Silbond and then impregnated.
The catalysts tested were zeolites exchanged or impregnated with first row transition
metals (TM). In our nomenclature, TM-zeolite represents an exchanged TMO-zeolite. A
slash (/) following the zeolite shows the impregnated TM cation and its salt or acid that was
used to impregnate the zeolite. For example, Co-Y/CA is a cobalt exchanged Y-zeolite
which was impregnated with chromium using chromic acid.
2.2 Apparatus
The experimental setup is shown elsewhere [3]. A 25 mm I.D., 1 m long Pyrex
tube normally was packed with 5.4 cm of catalyst. For 500 cmVmin reactant flow, a space
time of 3000 hr-1 would result. The reaction temperature was controlled within 1°C of the
250 to 400°C setpoint by the two Lindberg tubular furnaces surrounding the Pyrex tube.
Sorbent breakthrough data were obtained with sorbent challenged by reactant run at room
temperature. Breakthrough was reported when the outlet concentration exceeded 5% of the
inlet concentration.
2.3 Analysis
Reactant feed and product samples were collected with Hamilton CR 700-200
constant rate syringes. These samples were injected for analysis into an HP 5890 GC
followed by an HP 5970B mass selective detector (GC/MS). Oxygen and CVOC pickup
were determined at 300°C with a thermogravimetric analyzer (TGA). This TGA consists of
a DuPont Model 2100 thermal analysis system with a 2950 TGA.
Surface areas were determined with a Quantachrome Quantasorb Jr. BET surface
area analyzer. Catalyst and sorbent compositions were determined using a Philips PV9550
energy dispersive X-ray fluorescence (XRF) spectrometer. Zeolite acidities were obtained
using the TGA by temperature programmed desorption (TPD) of ammonia previously
adsorbed at 100°C.
All experimental measurements were made in accordance with EPA Air and Energy
Engineering Research Laboratory Quality Assurance Category 3 criteria. For example, three
to five CVOC GC/MS samples were taken for each run at inlet and outlet, and the outlet
concentration is accurate to ±10 ppmv.
2.4 Procedures
The procedures for preparing Co-Y/CA are described elsewhere [3]. The
Silicalite/CA powder was bonded with Silbond for pellet formation and then impregnated.
Additional information on preparing these pellets and powders washcoated onto a low-
surface-area inert cordierite core for support is provided in Reference 4.
3. RESULTS
3.1 Catalyst Characteristics
Table 1 gives BET, XRF, and acidity results for selected catalysts whose
conversions and selectivity are reported. The typical H-Y, from which most of the above
catalysts were made, had a surface area of 550 m2/g, a SiO^Al^Os ratio of 1.6, and 0.05 g
NHs/g [3]. Thus properly prepared, cation addition does not significantly reduce surface
area or SiO^AhOs ratio, and appears to slightly decrease the acidity.
2
-------
3.2 Conversion
Figures 1 and 2 show the conversion of MeCl2 with one CCLj run and of TCE,
respectively. As expected, conversion falls for decreasing temperature. These figures show
that the use of Co in these catalysts is very effective in destroying MeC^ (see Figure 1) but
not suitable for destroying TCE (see Figure 2).
Table 1
Catalyst Characteristics and Run IDs
Surface Area SiCV Catalyst Acidity (g Run E>b
Catalyst" (nfl/R) M£H Cation(%) NH^/g) TCE MeCh
Cr-Y 440 1.58 2.17 0.042
Co-Y/Co 524 1.56 4.6 • 0.039
Co-Y/CN 419 1.61 3.34/5.95
Sil/CA 370 high 10.98 0.005
Mixed bedc
Co-Y/CA 330 1.77 1.96/5.47 0.0
Co-Y/CR 521 1.56 3.47/4.95 0.045
Cr-Y/Co 388 1.59 1.71/5.47 0.048
Co-Y 560 1.56 1.1 0.046
1-4 5-8
9-11
12-14 15-16
17-20 »
26-29 21-22d
30-33 --
34-37 --
38-40 --
41-43 -
a All runs at approx. 2400 h-i and 13,000 ppm added water.
b See Figures 1 and 2, respectively, for MeC^ and TCE runs. Note all runs are in
descending temperature order.
c Mixed bed is a physical mixture 50% each by weight Cr-Y and Silicalite/CA.
d Runs 23, 24, and 25 were CCl4 whose conversion is in Figure 1.
120
100
0
'
§
O
60
40
20
..c°2/Co Co-Y/CN Mixed Bed Mixed Bed (CCI4)
240
260
260 300 320
Temperature (°C)
Figure 1. Conversion of single carbon CVOCs
3
340
360
-------
120
100
g
"w
I
O
20 -
' O
Or.Y
Co-Y/CN
A
Sil/CA
—O-
Mixed Bed
— .4- —•
Co-Y/CA
250
300 350
Temperature (°C)
400
Figure 2. TCE conversion
The extent to which the mass balances for carbon and chlorine arc closed (closure
%) is shown in Figure 3. For the 43 runs selected, closure % for carbon averaged 90.2%
and closure % for chlorine averaged 98.8% with standard deviations of 12.5 and 12.1,
respectively. These closure %'s are calculated based on the measured inlet CVOC and CC>2
and outlet product concentrations.
140
40
Figure 3. % Closure on Mass Balances for Carbon and Chlorine
4
-------
3.3 Selectivity
Note that all the above runs were made with addition of about 13,000 ppm of water
vapor. Figure 4 gives the production ratios for the CO/CO2 and C12/HC1. Use of the Run
IDs as defined in Table 1 will give the catalysts/CVOC combinations, and the temperatures
are in Figures 1 and 2 for single carbon CVOCs and TCE, respectively. For example, run
ID 30 which has the highest Ch production is a TCE run at 325 °C using Co-Y/CA (see
Figure 2). The C12/HC1 ratios for TCE arc much higher than those for MeCh- However,
this Ch production varies for different catalysts. For example, CrY produces significant Cl2
during TCE oxidation [runs 1-4] while CoY/CN [runs 12-14] shows no detectable Ch
formation. Similarly, CO production with TCE is about five times greater with CrY than
with Sil/CA catalyst. Only COj is formed in the oxidation of a humid CCU stream with a
total absence of CO as exemplified by run IDs 23, 24, and 25.
CO/C02 100XCI2/HCI
A
40
Figure 4. Chlorine and carbon products
3.4 Deactivation
Table 2 shows some selected deactivation results. During these tests, the catalyst
was continuously challenged for 12 days with wet (approx. 13,000 ppm water) CVOC feed
at 600°C except for short periods when the temperature was reduced to 275°C to measure
conversion. The catalyst cation and surface areas dropped only slightly. However,
significant losses in conversion, oxygen and TCE pickup, and acidity were observed.
Figure 5 shows the falloff for several of these catalysts at various times during these
deactivation studies. For these TMO-zeolite catalysts, the chromium exchanged catalysts
appear much more stable than the chromium impregnated ones.
-------
Table 2
Characteristics of Fresh and Aged Catalysts
Catalyst
Age Surface
(days) Area
(m2/g)
A1203
Catalyst Oxygen TCE Conv. Acidity
Cation Pickup Pick- (%) (g
(oxides)(%) (%) up(%)
Cr-Y
Co-Y/CN
Sil/CA
Co-Y/CA
Co-Y/CR
Cr-Y/Co
sJO
^
c
0
'c/5.
>
C
0
O
0 440
12 412
0 419
12 339
0 370
12 331
0 330
12 250
0 521
12 440
0 388
12 348
1 \AJ
J
BO
70
60
50
40
30
oo
h — ~__
ft — ^~~— -
,\ "A"' -
-'. \
°
4 \
* %
' xv
\ \
x V '• "v
" x t\
V*\\
* * Vo.
" " '*'::X
B\\
1.58
1.59
1.61
1.46
high
—
1.77
1.73
1.56
1.56
1.59
1.57
2.17
1.54
3.34/5
3.22/5
10.98
9.98
1.96/5
1.92/5
.95
.37
.47
.03
3.47/4.95
3.32/3
5.47/1
5.40/1
.84
.71
.48
3.53
1.53
3.30
2.33
--
—
3.20
1.87
3.61
1.76
3.51
2.60
3.31
1.39
2.75
0.96
--
—
1.73
1.18
3.75
1.12
4.30
0.91
92 0.042
75 0.031
92 0.041
43 0.032
72 0.005
33
95 0.043
50 0.034
90 0.045
33 0.037
90 0.048
76 0.036
Cr-Y
B
o---^
A
o
^x
''-> X
r ,
0 50
\\
100
Ct-Y/Co
A
D
-.
• ^
0
N A
v^.
i
150
Co-Y/CA
O
D
^~_-,
^~ —
A...
o
i
200
Co-Y/CR
~- ^_
^^— - .
D
"A""
o
,
250
Co-Y/CN
"TT~~~— —
y\
••j^....Q
O
1
300
Sil/CA
"— — -__^
_^
350
Time (hr)
Figure 5. Catalyst deactivation w'rth time
3.5 Sorbents
Table 3 shows the capacity and breakthrough data for some of the sorbents tested to
date. As expected, the zeolites tested to date do not have as high an adsorption capacity as
carbon for the CVOCs. However, the zeolites do not lose their capacity as rapidly at lower
CVOC concentrations and do not appear to be adversely affected by relatively high levels of
water vapor.
6
-------
Table 3
Collected Sorption Results
n>
i
2
3
4
5
6
7
8
9
10
Catalyst
Carbon
Carbon
Carbon
Silicalite
Cr-Y
-SiVCA
Mixed Bedc
Mixed Bed
Mixed Bed
Mixed Bed
Adsorbate
TCE
TCE
MeCl2
TCE
TCE
TCE
TCE
TCE
MeG2
CCLj
Adsorb
Cone
(ppmv)
1181
240
250
1050
1125
1049
1100
240
1050
1055
Water Cone
(ppmv)
12,114
13,918
14,265
14,632
none
11,115
10,902
14,000
12,151
12,051
Break-
through
Capacity^
36.0
20.0
3.6
10.6
2.2
8.6
5.8
6.1
3.3
2.0
Saturation
Capacity13
40.0
27.0
5.0
16.9
5.9
14.7
10.4
9.4
5.5
3.8
BA11 runs were carried out at 23°C and at a space velocity of 2360 hr-i.
b Reported capacities are weight percent adsorbate adsorbed per adsorbent.
cMixed bed consisted of a 50-50 weight percent (wt %) physical mixture of Silicalite/CA
and Cr-Y.
4. DISCUSSION
4.1 Activity
Properly prepared TMO-Y catalysts using several of the first row transition metals
give excellent activity in the oxidative destruction of single carbon HVOCs. However, no
data gave satisfactory TCE destruction using TMO-Y catalysts without the use of chromium.
The modified Co-Y catalysts used in Reference 5 were modified to Co-Y/CA catalysts by
impregnating them with chromic acid Much data exist (e.g., Figures 2 and 3 and Reference
5) that show greater than 95% CVOC destruction in the 300 to 350°C temperature range and
space velocities of about 3000 hr1. Reference 6 shows the expected falloff of conversion
with increasing space velocity. Temperatures of less than 275°C, with initially satisfactory
conversion, are often unsatisfactory due to activity loss caused by coking [5].
However, the acidity data in Table 1 are largely incomplete. The strength of acid
sites is better reflected by the peak ammonium desorption temperature [3]. This peak
temperature acidity is more indicative of catalyst activity than is total ammonium
adsorption/desorption.
4.2 Mechanisms and Selectivity
The CO/CO2 ratios do not appear to be affected by the addition of water. However,
the introduction of water can have a significant effect on Ch production. For example, at
300°C the percent C\2 increased from the 0 to 2% range with water to about 20% for dry
TCE feed [6].
For the TMO-zeolite oxidation of single carbon CVOCs, it is proposed [3] that the
HVOC is adsorbed on a Bronsted acid site. It then combines with a proton from this site to
form a carbonium ion. Oxygen is adsorbed on a cationic site as a disassociative species
(i.e., O) which oxidizes the TM (e.g., from Co-1"2 to Co*3) and forms O-. The carbonium
ion reacts with O- to form a surface intermediate and chlorine product.
For CCU, the surface intermediate is phosgene (COCli) or COChH+j and the
chlorine product is C\2- Since HC1 is the highly predominant product, this Ch either
-------
remains adsorbed or readsorbs and then desorbs as HC1 if sufficient water is present. With
a humid feed, COC12 reacts with water on the surface or in the vapor phase as:
COC12 + H2O --> CO2 + 2HC1
ForMeCl2, the surface intermediate is COHC1 which decomposes directly to CO
and HC1; and the chlorine product is HC1.
For TCE and Cr-Y, the proposed mechanism [7] will have active cationic sites.
Initially, oxygen adsorbs as O-, as discussed above, followed by the adsorption of an
association with TCE at the same site. Then a cyclic intermediary is formed in which the
double Cr=O and C=C bonds are saturated due to the formation of Cr-C and O-C bonds in
this cyclic intermediate. Two paths are proposed: one leads to the formation of two COHC1
intermediates and the other to one COHC1 and one phosgene. Thus the direct formation of
HC1, CO, and some CO2 is proposed. While some by-product chlorine formation may
occur in a parallel reaction, most of the little C12 formed is likely from the Deacon reaction:
2{HC1} + [O] < > [H2O] + {C12}
(Note: [ ] and { } represent different sites.)
This reaction appears is to be promoted for dry TCE feeds and higher catalyst temperatures.
Especially note that the above proposed TCE mechanism sites are Cr cationic rather than the
acid sites on which the C12 was adsorbed in the CCU mechanism.
4.3 Sorbent/Catalyst
Y-zeolite has a very high alumina/silicate ratio which provides acid sites, gives it
many exchange sites, and makes it an excellent catalyst. However, these same properties
make it hydrophilic and a poor adsorbent. Silicalite, a zeolite with very little alumina, has
low acidity and very few exchange sites, limiting it to an impregnation catalyst. Cr-Y is a
very good catalyst but is a poor adsorbent. Silicalite/CA is a good adsorbent but is only a
fair catalyst. The 50-50 wt % mixture of Cr-Y and Silicalite/CA serves both as a good
catalyst and good adsorbent. Carbon, an excellent adsorbent, cannot be used because of its
inability to withstand the reaction temperature.
Ongoing work includes investigating heat effects on existing catalysts and the use of
other zeolites. The 600° C deactivation work, herein reported, suggests that there should be
little drop in catalyst activity and likely little drop in physical adsorption capacity due to
poisoning or temperature effect. However, possible coking may dictate using temperatures
higher than needed for reaction.
4.4 Deactivation and Chromium Issue
The accelerated aging studies were carried out at 600°C, a very severe condition for a
high alumina zeolite (Y-zeolite). At these conditions, chromium was lost whether it was
exchanged or impregnated, but the exchanged form appears more stable. As previously
reported [3], at increased temperature and Cr2C»3 content, the Deacon reaction increases the
Cl^Cl ratio for both TCE and MeCl2 oxidation with the Cr-Y/CA catalyst. Also, Manning
[8] found that during the oxidation of perchloroethylene C12 reacted with chromium in the
catalyst to form chromium oxychloride (CrO2Cl2). However, in our deactivation studies,
dark green residue (Cr2C»3) rather than red-orange (CrO2Cl2) was observed. Experimental
determinations of these valence states, involving Auger Electron Spectroscopy (AES) or X-
ray photoelectron spectroscopy (XPS), would involve a significant change in research
approach. Avoiding or explaining possible artifacts, such as might be caused by the high
vacuum associated with running XPS, could in itself be a major effort.
8
-------
Metal sintering also has been observed on metal loaded zeolites. This sintering
causes a loss of catalytic surface area due to crystal growth in the catalyst phase. High
temperatures promote, and water vapor accelerates this sintering. Two mechanisms [9] for
metal crystal growth are (1) crystal migration (movement of metal crystals, collision, and
agglomeration); and (2) atomic migration (detachment of metal atoms from crystallites,
migration over the surface, and capture by larger crystallites). The loss of BET surface area
for both O2 and TCE as shown in Table 4 reflects the loss of crystallinity and resulting
catalyst activity loss.
Several of these deactivation mechanisms suggest that high temperatures, especially
at high chromium loadings, promotes much higher Cl2 levels than would occur otherwise.
Current plans are to run long term deactivation studies at a lower temperature (e.g., 450°C).
EPA's current research plan is to find a substitute for chromium in the catalysts used
to control TCE and other CVOCs. Efforts to minimize these effects through lower
chromium levels that have higher low temperature activity, higher stability, and even higher
selectivity (especially those effectively eliminating Cl2 formation) are underway.
5. CONCLUSIONS
Zeolite catalysts, exchanged or impregnated with transition metal cations, have been
specifically compounded to deliver high oxidation activity in destruction of a variety of
CVOCs.
By judicious choice of exchanged or impregnated cations zeolite catalyst selectivity
may be tailored to produce more CO2 than CO and to produce HC1 with essentially no Cl2.
Low temperature activity of these catalysts is limited below 275°C by their tendency
to reversibly deactivate through a coking process.
Because zeolites can be configured to be both effective catalysts and sorbents, they
are potentially useful in a dual role of first physisorbing and then (upon heating) acting as
effective catalysts in CVOC control.
REFERENCES
1. Hylton, T., "Performance Evaluation of the TCE Catalytic Oxidation Unit at
Wurtsmith AFB", AFESC Report ESL-TR-91-43, April 1992
2. Foger, K. and H. Jaeger, "The Effect of Chlorine Treatment on the Dispersion of PT
Metal Panicles Supported on Silica and Alumina", J. Catal, 22, 64 (1985)
3. Chaterjee, S. and H.L. Greene, Applied Catalysis A: General, 9_8_, 139 (1993)
4. Chaterjee, S. et ah, J.Catal., 138. 179 (1992)
5. Chaterjee, S. et al., Catalysis Today, JJ, (1992) 569-596
6. Ramachandran, B., M.S. thesis, University of Akron (expected 1993)
7. Chaterjee, S., Ph.D. dissertation, University of Akron (1993)
8. Manning, M.P., Hazard. Waste, 1 (1),41 (1984)
9. Rachapudi, R., ongoing work toward M.S. thesis at the University of Akron
DISCLAIMER
This paper has been reviewed in accordance with the U.S. Environmental Protection
Agency's (EPA) peer and administrative review policies and approved for presentation and
publication.
-------
-------
AEERL-P-1122
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before comp
1. REPORT NO.
EPA/600/A-94/029
2.
4. TITLE AND SUBTITLE
5. REPORT DATE
Development of Transition Metal Oxide-Zeolite
Catalysts "to Control Chlorinated VOC Air Emissions
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
C. A. Vogel (EPA) and H. L.Greene (Univ. of Akron)
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
University of Akron
Department of Chemical Engineering
Akron, Ohio 44325-3906
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
CR 819695-01
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Published paper;
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES AEERL project officer is Chester A. Vogel, Mail Drop 61, 919 /
541-2827. Presented at 2nd International Symposium: Characterization and Control
oL.Odours and VOC in the Process Industries. Louvain-la-Neuve. Belgium, 11/3-5/93
paper discusses the development of transition metal oxide (TMO)-zeo-
lite oxidation catalysts to control chlorinated volatile organic compound (CVOC) air
emissions. Research has been initiated to enhance the utility of these catalysts by
the development of a sorption-catalyst system. Zeolites with a high.Al/Si ratio
(e.g., Y-zeolite) provide active acid exchange sites. The exchanged and calcined
zeolite is impregnated with the same or different metal and calcined to form the
TMO film. These TMO-zeolites provide shape selectivity and three types of active
sites: unexchanged HY (Bronsted) acid active sites, highly active exchanged sites,
and impregnated TMO sites. They can achieve over 95% destruction efficiency at
relatively low temperatures (i. e. , 300 to 350^C) of a humid, low concentration
CVOC, and are very resistive to poisoning. , In the sorption-catalyst system, the
zeolite first would physically adsorb the CVOC at room temperature.-^The system
would then be heated to promote the TMO-zeolite catalytic reaction. The Y-zeolites
can physically adsorb about 5% VOC but are highly hydrophylic. Silicalite is a zeo-
lite with very little Al, can adsorb up to 15% VOC, and is hydrophobic. Impregnated
Silicalite is an effective adsorbent and fair catalyst.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Sorption
Organic Compounds
Volatility Oxidation
Chlorination
Ion Exchange Resins
Catalysis
Pollution Control
Stationary Sources
Zeolites
Silicalite
13B
07C
20M
07B
07D
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)'
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
------- |