NASA-CR-194354

FINAL REPORT

P

DESTRUCTION OF PROBLEMATIC AIRBORNE CONTAMINANTS

BY HYDROGEN REDUCTION USING A CATALYTICALLY ACTIVE,

REGENERABLE SORBENT (CARS)

July 1993

Prepared by:

John O. Thompson

James R. Akse

C_ontract Number: NAS2-13714

Ames Research Center

NatJonaI-Aeronaufdcs & Space Administration

',J ....... _Moffett Field, CA 94035-1000

UMPQUA RESEARCH COMPANY

AEROSPACE DIVISION

PO BOX 791- 125 VOLUNTEER WAY

MYRTLE CREEK, OR 97457

TELE: (503) 863-7770

FAX: (503) 863-7775

URC 80436

i: uiiiil ;k'i_

0

,,t-

I

O,Z

ult_

U¢

NU_U'_00f-4O

_ _KOZ +J O-

H_ ON

0 _ i-- .'='

- ,,t. o_ C_ U- LAI'_=,-, UJ _.k.

I I'-_ _''_

u.JC3 >.. Z :3_.. / .j>.,.j _j O"

Z. _ >-,<C)_

u_.4"

rr_

https://ntrs.nasa.gov/search.jsp?R=19940009630 2018-06-01T05:27:10+00:00Z

z

w

FINAL REPORT

w

_= -

DESTRUCTION OF PROBLEMATIC AIRBORNE CONTAMINANTS

BY HYDROGEN REDUCTION USING A CATALYTICALLY ACTIVE,

REGENERABLE SORBENT (CARS)

July 1993

_=tPrepared by:

=

w

J

John O. Thompson

James R. Akse

-L -

wContract Number: NAS2-13714

i

i

Ames Research Center

National Aeronautics & Space AdminislTaUon

Moffett Field, CA 94035-1000

i

• Z_=

w

w

UMPQUA RESEARCH COMPANYAEROSPACE DIVISION

PO BOX 791- 125 VOLUNTEER WAY

MYRTLE CREEK, OR 97457

TELE: (503) 863-7770FAX: (503) 863-7775

- d

w

fr.,_

URC 80436

= ,

_=

w

L_w

w

W

i

w

w

w

w

PROGRAM OVERVIEW

In future NASA missions, the TCCS will need to operate more efficiently over longer

time periods with the use of fewer expendables due to the increased demands of larger crewsand longer mission durations. The Space Station Freedom (SSF) baseline air regenerationsystem, a combination of carbon sorbents and a catalytic oxidation system, usesnon-regenerable sorbent beds and an oxidation catalyst which is run constantly and issusceptible to poisoning by halogen, sulfur, and silicon containing compounds. The resupplypenalties and energy demands of this system are high. Significant savings in both areas canbe realized by the use of regenerable sorbents which specifically eliminate catalyst poisons

allowing more efficient operation conditions to be used by the catalytic oxidizer.A Catalytically Active Regenerable Sorbent (CARS) system is an air purification

process which uses a sorbent bed composed of an activated carbon made catalytically activeby metals dispersed on its surface. The same bed can be used as the sorbent, and during

regeneration, as the catalytic reactor. A propedy designed regenerable bed's sorptioncapacity is also adjustable since the timing of the regeneration cycle in a parallel bed systemcan be continuously adjusted to always adsorb the contaminant load.

The feasibility of a Catalytically Active Regenerable Sorbent (CARS) system has beendemonstrated in this study. A complete adsorption and regeneration test performed on thebest CARS cOmbination cleady illustrat-es-the effecti_iless Of this technology for the treatment

of halogen and sulfur containing contaminants. The challenge mixture contained Freon-113,trichloroethylene, bromotrifluoromethane, and dichloromethane at 436%, 15000%, 11.2%, and95.6% of their Spacecraft Maximum Allowable Concentration (SMAC) respectively as well as

thiophene, a generalized organosulfur compound without a SMAC value. Approximately 100liters of this challenge stream was sorbed by a 7.8 cm z bed before breakthrough occurred.The bed was then purged, filled with hydrogen, and heated to 250 C for 500 minutes followedby 300 C for 1100 minutes. After the initial heating, none of the original sorbed contaminantswere detected demonstrating the high catalyiic hydi'-0-genation activity at these lowtemperatures. The multi-step hydrogenation of the remaining by-products required anadditional 1600 minutes completing the regeneration. The CARS bed was then serviceablefor a ne_v adsorption _n, The regeneration duration can be reduced considerably using

higher hydrogen pressures and reactor temperatures, or by improvements in the catalyst.Before the CARS approach to remediation of problematic trace airborne contaminants

can be used to replace expendable sorbent beds in the current TCCS configuration, additionalwork is needed in several areas. Future work will involve challenging a CARS bed with amixed contaminant stream representative of all classes of compounds in the TCCS model.

The use of layered thermally desorbable sorbents must be investigated to optimize thesorption capacity of the bed for a mixed contaminant stream, and the effect of additionalcontaminants on catalytic hydrogenation must be determined. Optimization of the catalyst's

activity and regeneration conditions will lead to a more efficient system, including the use ofmetal hydrides to initially provide and later scavenge unreacted hydrogen when a regeneration

is complete. Evaluation and optimization of the alkanes produced during regeneration willallow the catalytic oxidizer to operated at lower temperatures. The impact of a fully developed

CARS system on TCCS operation will include lower resupply penalties, greater flexibility inhandling contamination events, lower overall system weight, and possibly lower energy

consumption.

URC 80436 i

I

_= _.

TABLE OF CONTENTS

. =

= =

I

1.0

2.0

3.0

4.0

5.0

INTRODUCTION ................................................. 1

SELECTION OF MODEL CONTAMINANTS ............................. 2

ADSORPTION AND DESORPTION STUDIES ............................ 5

3.1 Single Component Adsorption Study ......................... 5

3.2 Desorptions .......................................... 6

3.3 Multiple Contaminant Adsorption Study ....................... 9

CATALYST ACTIVITY TESTING .................................... 11

4.1 Catalyst Test Apparatus ................................ 11

4.2 Hydrogen Reduction Reactions ........................... 13

4.3 Results of Catalyst Comparisons .......................... 14

THE FINAL TEST STAND CONFIGURATION AND RESULTS .............. 22

5.1

5.2

5.3

5.4

5.5

The Final Test Apparatus ............................... 22

The First Run ........................................ 23

The Second Run ...................................... 24

The Third and Fourth Runs .............................. 28

The Fifth Run: The Molybdenum Catalyst ................... 32

qr=e 6.0 TEMPERATURE, HYDROGEN PRESSURE, AND PRESSURE DEPENDENCE

FOR THE REACTION OF DICHLOROMETHANE OVER 5% RU, 20% PT

CATALYST .................................................... 39

7.0 CONCLUSION .................................................. 42

8.0 FUTURE WORK ................................................ 45

BIBLIOGRAPHY ...................................................... 46

REPORT DOCUMENTATION PAGE ....................................... 47

_=

L

_=._-i

URC 80436 ii

W

,win

E

w

D

= :

r=

_=

TABLES AND FIGURES

TABLE 1

TABLE 2

Single Contaminant Adsorption Results ........................... 3

Contaminant Generation Rates and SMAC Limits .................... 5

FIGURE NUMBER

1 Final Test Stand Configuration ................................. 4

2 Thermal Desorption of Freon-113 ................................ 7

3 Thermal Desorption of Dichloromethane ........................... 8

4 Multiple Contaminant Adsorption Study ........................... 10

5 Catalyst Test Apparatus ...................................... 12

6 Thiophene Reduction Reaction Steps ............................ 14

7 Baseline Catalyst v.s. 5% Nickel Catalyst for the Reduction ofFreon-113 at200 Degrees C ............................. 16

8 5%Pd,5%Pt v.s. 5%Ru,5%Pt Catalyst for the Reduction of

Freon-113 at i25 Degrees C. ,., ......................... 17

9 5%Pd,5%Pt v.s. 5%Ru,5%Pt Catalyst for the Reduction ofFreon-113 at 200 Degrees C ............................. 18

10 Baseline Catalyst Reduction of Bromotrifluoromethane ............... 20

11 Baseline/Molybdenum Catalyst Reduction ofBromotrifluoromethane ................................. 21

12,13,14 Final Test Stand's Second Regeneration Run .................... 25-27

15,16,17 Final Test Stand's Third Regeneration Run ...................... 29-31

18,19,20 Final Test Stand's Fourth Regeneration Run ..................... 33-35

21,22,23 Final Test Stand's Regeneration Run Using theBaseline/Molybdenum Catalyst .......................... 36-38

24 Temperature Dependence of Dichioromethane Hydrogenation ........... 40

25 First Order Relationship of Dichloromethane to Reaction Rate ........... 41

26 Hydrogen Pressure Dependence of Dichloromethane Reduction ......... 43

URC 80436 iii

_-= ]= =

I

w

z

=

=

W

r_

w

F_

w

l

i

1.0 INTRODUCTION

In future NASA missions, the TCCS will need to operate more efficiently over longer

time periods with the use of fewer expendables due to the increased demands of larger crews

and longer mission durations. The Space Station Freedom (SSF) baseline air regeneration

system, a combination of carbon sorbents and a catalytic oxidation system, uses non-

regenerable sorbent beds and an oxidation catalyst which is run constantly and is susceptible

to poisoning by halogen, sulfur, and silicon containing compoundsJ -4 By using a process in

which the sorbent beds are regenerated, while at the same time reducing the operating time

and exposure of the catalytic oxidizer to catalyst poisons, resupply penalties and energy

demand can be significantly reduced. A Catalytically Active Regenerable Sorbents (CARS)

system is an air purification process using a sorbent bed composed of an activated carbon

made catalytically active by metals dispersed on its surface. The same bed can be used as

the sorbent and the catalytic reactor. While a non-regenerable sorbent bed will be limited in

its sorption capacity to the least retained contaminant present in the air stream, a properly

designed regenerable bed's sorption capacity is unlimited in the sense that the regeneration

cycle for a parallel bed system can be continuously adjusted to always adsorb the

contaminant load.

The catalytic regeneration is performed by purging the sorbent bed of oxygen with an

inert gas, then filling and maintaining pressure in the regeneration loop with hydrogen. The

catalyst bed is heated to operating temperature, 200 to 300 degrees C, while a pump

circulates hydrogen through the closed regeneration loop. A lithium hydroxide bed contained

in the regeneration loop scrubs the stream of volatile acid gases produced from halogen

reduction. When regeneration is complete, hydrogen and alkanes, the only compounds

remaining in the gas stream, may be....................................................purged from the sorbent bed and incinerated by the

catalytic oxidizer. Operated in conjunction with a CARS system, the catalytic oxidizer can be

operated more efficiently since its influent is more consistent and contains no catalyst poisons.

To determine the effectiveness of this air regeneration process, UMPQUA tested

catalysts prepared on activated carbon supports containing combinations of nickel,

molybdenum, palladium, ruthenium, and platinum to determine their relative effectiveness at

reducing halogenated organic compounds. In order to test the CARS concept a model

contaminant mixture containing dichloromethane, bromotrifluoromethane, trichloroethylene,

thiophene, and Freon-113 was developed after consultation with NASA AMES personnel. The

more active catalysts reduced these contaminants individually to alkanes at temperatures

ranging from 125 to 200 C.

URC 80436 1

I

w

w

w

.. =

!li

i

°__

= :

A comparison of the adsorption capacity of the catalyst to that of untreated activated carbon

revealed minor losses of sorption capacity due to the deposited metals ('Fable 1). This

retention of sorptive capacity allows the catalyst to also be used as a major component in the

adsorption bed. If needed, other thermally desorbable sorbents can be added to enhance the

sorbent bed's capacity.

Since the baseline catalyst was capable of reducing all of the contaminants chosen for

this project and also function as an effective sorbent, the test stand's final configuration was

developed (Figure 1). The test stand operates in two distinct modes: an adsorption cycle and

a catalytic regeneration cycle. In the adsorption cycle, contaminated air passes through a

sorbent bed composed of a mixture of 4 cc of the baseline catalyst and 4 cc of a Carbosieve,

a component necessary to retain the most volatile contaminant, bromotrifluoromethane. A

hydrocarbon analyzer monitors the effluent gas for contaminant breakthrough. In the catalytic

regeneration cycle, oxygen is purged from the regeneration loop by nitrogen, which is in tum

purged by hydrogen. The regeneration loop is then pressurized and maintained at 14 psig with

hydrogen. A pump recirculates hydrogen through the catalyst bed while it is heated to the

hydrogenation temperature. The catalytic hydrogenation of the halogenated contaminants

produces acid gases which are released into the recirculation loop where a room temperature

lithium hydroxide bed removes them.

The baseline catalyst; 5%Ru, 20%Pt on activated carbon; was run four times with

vadous contaminant concentrations and regeneration temperatures. Due to time constraints,

total regeneration of the bed by destruction of all contaminants was not performed.

The test stand was run with a catalyst containing 5%Ru, 20%Pt, and 10%Mo. The bed

was initially loaded with contaminants to saturation, as indicated by the breakthrough of

bromotrifiuoromethane. Total regeneration of this bed was performed in a timely manner

demonstrating the CARS concept. This catalyst showed a much greater activity for reduction

of the contaminants compared to the baseline catalyst.

2.0 SELECTION OF MODEL CONTAMINANTS

URC collaborated with NASA AMES personnel to select five contaminants which are

potential catalyst poisons for the TCCS catalytic oxidizer:. Freon-113 (CI2FC-CF2CI), the most

abundant contaminant in the TCCS model; dichloromethane (CH2CI2) and

bromotrifluoromethane (CBrF3), two other contaminants found in the TCCS model;

trichloroethylene (HCIC=CCI2), a common industrial solvent; and thiophene, a contaminant

URC 80436 2

w

= ,

L

w

-f_7

W

L--

W

SORBENT

MATERIAL

COLUMN

SIZE

Activated

Carbon

CONTAMINANT CONCENTRATION

mg/L

Activated 6cc Freon 113 8.15Carbon

5%Ru,20%Pt 6cc Freon 113 8.15

Catalyst

4cc Freon 113 0.475

5%Ru,20%Pt 4cc

Catalyst

4ccActivated

Carbon

5%Ru,20%Pt

Catalyst

5%Ru,20%Pt

Catalyst

4cc

4cc

Freon 113

5%Ru,20%Pt

Catalyst

CH2CI_

CH2C12

Thiophene

0.49

0.168

Carbosieve

S-III

0.198

0.43

SORPTION

mg/cc

3OO

235

122

107

11.6

13.7

122

FLOW

RATE

1370

1370

650

650

650

660

860

4cc TCE 1.78 274 770

5%Ru,20%Pt 4cc CBrF 3 0.35 1.9 710

Catalyst

Activated 4cc CBrF 3 0.35 1.9 710Carbon

ACF-1605 2cc CBrF 3 0.35 1.6 710

Cloth .45gm

4cc 13.6 680CBrF3 0.35

w

W

Table I. Single Contaminant Adsorption Results

Imm

W

w

URC 80436 3

CARS Adsorption/Recirculating Catalytic

Hydrogenation Test System

t:j

F_

"r--:

g

_J

E_

W

l

h

=--:

W

lIBe

HYDROCARBONANALYZER

o o

ADSORPTION,%

PATH .................::'

SORBENTREACTOR

.........?-I_--

SAMPLE

Z g

MAKEUPHYDROC,EN

REGULATOR _

HYDROGENSOURCE

TUBE FURNACE

PERISTALTIC PUMP

i

PRESSURIZEDCONTAMINATED

AIR

>.

HYDROGENRESERVOIR

LITHIUM HYDROXIDEBED

REGENERATION PATHFi gure 1. Final Test Stand Configuration

URC 80436 4

w

Irepresenting organosulfur sources particularly problematic for oxidation catalysts. The

following table is provided for comparison purposes 1. The third column contains a calculated

generation rate of contaminants for a 4 module space station with 31500 cubic feet of air

assuming that no air pudfication occurs over a period of 24 hours.

w

CONTAMINANT SMAC

mg/Liter

GENERATION RATESSF AIR VOLUME

mg/day Liter

Bromotrifluoromethane 0.6088 0.00052

Dichloromethane 0.0868 0.00193

Freon-113 0.383 0.0254

Thiophene ..........

Trichloroethylene 0.00054 0.000044

W

r_

w

U

I_:: $

L_

INO

w

Table 2. Contaminant Generation Rates and SMAC Limits

3.0 ADSORPTION AND DESORPTION STUDIES

3.1 Single Component Adsorption Study

To determine the system's final configuration, a comparison between sorption

capacities of the baseline catalyst and its untreated activated carbon substrate was

established (Table 1). If the carbon substrate's sorption capacity were significantly reduced by

the addition of catalytically active metals, another bed would be needed to trap and then

thermally desorb contaminants during regeneration into the catalytic reactor. If the sorption

capacity were not significantly reduced, the same bed could be used as both a sorbent and a

catalyst.

The sorption comparison between the baseline catalyst and the activated carbon was

established at two concentrations of Freon-113 and one concentration each of

dichloromethane and bromotrifluoromethane. At the high concentration of Freon-113 (8.15

mg/L = 8150 mg/m3), the catalyst adsorbed only 78% of the untreated carbon's capacity.

However, at a lower concentration of 0.48 mg/L, the catalyst adsorbed 88% of the untreated

carbon's capacity. For the concentrations of dichloromethane (0.168-0.198 rag/L) and

bromotrifluoromethane (0.35 mg/L), the catalyst adsorbed the same amount as the untreated

carbon. The reason for the catalyst's loss of sorptive capacity with respect to Freon-113 is

URC 80436 5

E

r =

W

L

L_._

uncertain, but the loss is limited at lower concentrations, which are closer to the TCCS model.

The limited loss of sorptive capacity due to the addition of catalytically active metal indicates

that the use of the same catalyst bed as both sorbent and catalytic reactor is feasible.

The baseline catalyst's sorption capacity varies greatly with the contaminant adsorbed.

Similar concentrations of Freon-113 and thiophene, 0.49 and 0.43 mg/L respectively,

adsorbed at over 100 mg/cc of catalyst; while 0.35 mg/L of bromotrifluoromethane was

adsorbed at only 1.9 mg/cc of catalyst. Dichloromethane, at an influent concentration of 0.198

mg/L, was adsorbed by the catalyst at 13.7 mg/cc. The adsorption data for CARS indicates

that there is an inverse relationship between a contaminants volatility and its adsorptivity.

Due to the catalyst's relatively low affinity for bromotrifluoromethane adsorption, two

other sorbents were tested. An activated carbon fiber derived from the carbonization of kynol

fibers (ACF-1605 distributed by American Kynol, Inc.) showed little capacity for

bromotrifluoromethane adsorption. More promising results were obtained by using Carbosieve

S-Ill (available through Supelco), a carbon based molecular sieve with a 15 to40 angstrom

internal pore matrix. This product was originally manufactured for the adsorption of low

molecular weight hydrocarbons in chromatography gas streams, making it suitable for the

adsorption of similarly sized halocarbons. The Carbosieve's capacity was established at 13.6

mg/cc with a stream concentration of 0.35 mg/L bromotrifluoromethane. Its capacity was 7.2

times greater than activated carbon's capacity. Previous work at UMPQUA has shown this

Carbosieve to be thermally regenerable. 5'6 The addition of carbosieve to the sorption/catalyst

bed will significantly increase the beds overall capacity when low molecular weight

halocarbons are present in the contaminated air stream.

3.2 Desorptions

The configuration of the final test stand depends on the temperature dependent

desorption of contaminants from the catalyst's surface. If desorption occurs at temperatures

below the catalytic hydrogenation temperature of 125 to 200 C, a single pass system will

bleed an excessive amount of contaminants before they are reduced. Desorptions were

performed on the activated carbon and baseline catalyst with Ioadings of Freon-113 and

dichloromethane (Figures 2 and 3). The temperature was raised from 23 C to 125 C at 1

C/minute with a continuous flow of nitrogen gas through the bed. The effluent gas was

analyzed'by shots on a gas chromatograph. These data indicate that: both the activated

carbon and the baseline catalyst desorbed significant concentrations of both contaminants

before an effective catalytic temperature was reached, and that the catalyst desorbed higher

concentrations of halocarbon sooner than the activated carbon. Since the contaminants

URC 80436 6

u

Thermal Desorption of Freon-113T=23to125 at 1C/min,21 cc/min Nitrogen

120 ......................................................................................................................................................................................................................................................10 ...................................................................................................................................................................................................................................................................................

w

W

O3E

V

cO

!

(.-O

LL.

Ea)

_1=LU

100-908070-60-50-40-

[]........................................................................ _h. ............................... A .........................................................................................................................................................................

[]

•ES]

30_ ........................................................._ ................................................................................................................................_'........._............................................................................204..................................................."_................................................................................................................................................................."........... .........., .................................

0 .....................................................................................................................................................................................................................................................................................

O_U_m_ - _ _ _ _ _ _0 50 100 150 200 250 300 350

Desorption Time (Minutes)

400

= =

w

w

[]Activated Carbon • Baseline Catalyst I

Figure 2. Thermal Desorption of Freon-113

URC 80436 7

Thermal Desorption of DichloromethaneT=23to125 at 1C/min,21 cc/min Nitrogen

3s]E 301 .........................................................................................................._ ..............................................................................................................................................................

W

z

i

C f T

0 5O 100 15O

"""A.,,, .,,..j..::.................................................................................

200 250 300

Desorption Time (Minutes)

- _ Activated Carbon ....Baseline Catalyst

= : Figure 3. Thermal Desorption of Dichloromethane

URC 80436

vr_

I

Zw

w

====

w

i :

desorb at a temperature below which the catalyst is effective, a single pass system in which

hydrogen passes through a heated bed loaded with contaminants will bleed an unacceptable

amount of contaminants into the hydrogen stream. This problem is compounded by the

formation of partially hydrogenated intermediates more volatile than the original contaminants.

There are two possibilities to circumvent this problem. Given enough gas volume, the

reactor can be pressurized with hydrogen and then heated until all halocarbons have been

reduced. Depending on the amount of halocarbons adsorbed on the bed, this method might

require high hydrogen pressures to supply sufficient reducing agent. This method allows

halogen acids to remain in contact with the catalyst blocking catalytically active sites and

requiring a long desorption time to clear the bed at the end of the regeneration cycle.

The second possibility involves recirculating hydrogen and partially reduced by-

products through the reactor and a lithium hydroxide bed to adsorb the acid gases. The acid

gases no longer compete for catalytic sites. Hydrogen consumed in the reactor can be

replaced by maintaining the system at a constant pressure using hydrogen as the makeup

gas. Recirculation is continued until all of the halocarbons are reduced. This is the process

configuration chosen as the final system configuration for this project.

3.3 Multiple Contaminant Adsorption Study

A 4cc bed of the baseline catalyst was challenged with a multiple contaminant stream

to determine how the component's adsorption behavior interact. The contaminant air stream

was assayed using the gas chromatograph and comparing the peak heights to those of

standard concentrations prepared in the lab. The air stream contained:

0.51 mg/L bromotrifluoromethane0.55 mg/L dichloromethane0.54 mg/L Freon-1130.47 mg/L thiophene0.38 mg/L trichloroethylene

The bed was challenged at a flow rate of 1000 cc/minute for 1270 minutes (Figure 4).

The sorbent retained an insignificant amount of bromotdfluoromethane, and

breakthrough occurred immediately. Dichloromethane, Freon-113, and thiophene exceeded

their influent concentrations after their complete breakthrough occurred. By performing a

manual integration of the adsorption curves, it was calculated that the sorbent initially retained

67 milligrams of dichloromethane but then released 46 milligrams before the end of the run,

thereby retaining only 21 milligrams. The sorbent also initially retained 150 milligrams of

Freon-113 but then released 83 milligrams, thereby retaining only 67 milligrams. Competition

from more strongly adsorbed species such as thi0phene and trichloroethylene displaces a

URC 80436 9

Lm

L

w

Z :

W

-- =

0.8_J_0.7

E 0.6

0.5.D

0.4

0.5

0.2

cO)0.1

0.01.0

_0.8E

co 0.6

o

0.4c

oco 0.2

(D

0.00.7

_J

_0.6

E 0.5c

.o 0.4

0.3(-Q)o 0.2c-Oo0.1

0.00.4

_.J

E O.3

c-O

+-.,o 0.2L-

C_D

©(D

0.0

URC 80436

I I I I I I |

_ D,CH:OROMETHAnE1

INFLUENT CONCENTRATION=0.55 mg/L

I 1 1 I

I I I I I I IFREON- 1 13

I 1 I I 1 l

I 1

THIOPHENE

I I I I

I I I

0

I I I I I I

INFLUENT CONCENTRATION=0.38 mg/L

- TRICHLOROETHYLENE ....f

, , ,200 400 600 800 1000 1200

THROUGHPUT (LITERS)

Figure 4. Multiple Contaminant Adsorption Study10

-- f

w

= =

=__m

= .

_=L___j

r===_

H

w

L_

L=

=_F.=

W

portion of the less strongly adsorbed species. A sorbent bed exposed to a multiple

contaminant stream will be limited in sorption time by the least retained, more easily displaced

contaminant.

The more volatile contaminants such as bromotdfluoromethane and dichloromethane

are not well retained by this type of activated carbon. The addition of Carbosieves which

have a greater affinity for the more volatile organics will greatly extend the effective adsorption

capacity of a sorbent bed.

4.0 CATALYST ACTIVITY TESTING

4.1 Catalyst Test Apparatus

Five catalysts were prepared and tested for halocarbon reduction. Not all catalysts

were tested on all of the contaminants, but comparisons were made among pairs of catalysts

to determine the varying activities of each. The apparatus used for catalyst testing, with the

exception of bromotrifluoromethane which is a gas at room temperature, is illustrated in

Figure 5.

A regulated flow of hydrogen is split with one stream passing through a gas bubbler to

mix with the contaminant vapor. The contaminant/_hydr0gen stream is then diluted to the

desired concentration with uncontaminated hydrogen from the second branch of the split

stream. Early attempts to mathematically determine the halocarbon content by indicated flow

rates of the contaminated and dilution stream proved untenable. Incomplete saturation of the

contaminated stream and the lack of calibration tables for the flow tubes in relation to a mixed

hydrogen/halocarbon stream combined with reading errors of the flow tubes and the effect of

slight temperature changes on the halocarbon part!aivapor pressure to produce large errors.

It was decided that the influent concentration could be more accurately determined by use of

a gas chromatograph and standards prepared by injecting a known quantity of halocarbon into

a flask containing a known volume of air. When bromotrifluoromethane was used as the

contaminant, its vapor was mixed directly with a hydrogen stream.

The mixed contaminant/hydrogen stream then passes through individual flow control

valves for each reactor. A backpressure regulator maintains a constant pressure drop across

the flow control valves for accurate control. The reactors are composed of 112 inch diameter

stainless steel tubing with a 0.035 inch wall and porous metal frits as end plugs to contain the

catalyst. A temperature controller using a low mass thermocouple to measure each reactor's

external surface temperature controls a heating tape wrapped around the reactor to maintain

URC 80436 11

E

W

rz_

W

]

| _"

°iQ,.

W

I

tr

.iii

>

cx3

L/1

I--

+-)(/1>.,%

I.L.

L

URC 80436 12

E °

= =

-]

w

i

Hir==_

L_

¢r,d

W

U

L:J

t

the reactors' temperatures. Insulation is placed around each reactor to minimize heat loss

and reduce the temperature gradients across the reactors.

Early in the catalyst testing, the method of effluent gas analysis was changed.

Originally, the effluent gases were to be bubbled through a sodium hydroxide solution and

analyzed for pH changes and halogen content to determine the catalysts' effectiveness at

halocarbon reduction. This method would not have indicated the types or quantities of

intermediate products. The effluent gases were instead analyzed using a gas chromatograph.

Information regarding the presence of unreacted halocarbons, partially hydrogenated

intermediates and the final products, alkanes, is more informative when comparing catalysts.

The large number of possible intermediate by-products precludes the identification and

quantitization of all peaks observed from the gas chromatograph. Catalyst comparisons were

based upon the production of alkanes and comparative heights of the other peaks.

4,2 Hydrogen Reduction Reactions

The reduction of most halogenated and sulfonated organic compounds is a series of

many individual reductions. For each halogen or sulfur to carbon bond, one hydrogen

reduction must occur. Dichloromethane, for example, requires two separate reductions

producing methane and two hydrogen chlorides:

CH2CI2+ 1H 2.... > CH3CI + HCI CH3CI+ 1H2-----> CH 4 + HCI

Results of catalyst runs in which dichloromethane was reduced show that chloromethane is

present in the effluent gas.

The hydrogenation process becomes more complicated when reducing Freon-113.

The complete reduction of Freon-113 to ethane requires six separate reductions' Each

chlorine and fluorine is individually removed, resulting in over thirty possible intermediate by-

products. Results of catalyst runs have also shown |hat the carbon-carbon bond in Freon-113

is susceptible to hydrogen reduction. Both methane and chloromethane have been detected

in its effluent gases.

Carbon-carbon double bonds are more susceptible to hydrogenation than carbon-

halogen bonds. When a bed of the baseline catalyst at room temperature was loaded with

trichloroethylene and purged with nitrogen and then purged with hydrogen, a cloud of

hydrogen chloride gas was expelled by the reactor. The reactor was warm to the touch

afterwards. The energy released by the hydrogenation of the double bonds also causes the

carbon-chlorine bonds to react. Trichloroethylene, when reduced at 125 degrees C, shows

methane in the effluent, indicating reduction of the carbon-carbon bond.

Most of the intermediate reduction by-products of thiophene have not been identified.

URC 80436 13

w

LJ

w It has been observed that methane, ethane, propane, and butane are all present in the

effluent when thiophene is catalytically reduced. Butane is the most abundant by-product,

indicating that the thiophene ring structure is most often cleaved at the sulfur-carbon bond

producing thiobutane. Thiobutane will reduce to butane and hydrogen sulfide. The presence

of the other alkanes implies that the carbon-carbon bonds in the ring structure are sometimes

cleaved. This may occur when the double bonds in the ring structure are reduced. The heat

of this reduction, compounded with the ring stresses, may produce a reduction between two

carbons producing methyl propyl sulfide or diethyl sulfide. The subsequent desulfonating of

these compounds would lead to methane, ethane, and propane production, as shown in

Figure 6.

=

s'--ctC_c/C +H2= Butane + H2ScJS_.c

+2H2 _ I I +H2 _ C_s/C_c/c +2H2=Methane + Propane + H2SC--C

--_ C_c'S_c/C +2H2=2 Ethane+ H2S

w

L.

w

L_

Figure 6. Thiopene Reduction Reaction Steps

4.3 Results of Catalyst Comparisons

Five different catalysts were prepared for this project using an activated carbon

support. The same type of carbon substrate and parameters for thermal processing were

used for all of the catalysts. The amount of metal loading is by percent of the original

substrates mass. The catalysts tested were:

5% Ruthenium, 20% Platinum (The baseline catalyst)5% Ruthenium, 20% Platinum, 10% Molybdenum5% Nickel5% Ruthenium, 5% Platinum

5% Palladium, 5% Platinum

These catalysts all showed varying but significant activity toward halocarbon reduction.

Results will be discussed in a series of comparisons between catalysts under similar

conditions of temperature, contaminant concentration, and reactor residence time.

URC 80436 14

w

=

4.3.1 5% Nickel v.s. Baseline Catalyst for Freon-113

The 5% nickel catalyst and the baseline catalyst were run under similar conditions to

determine their relative abilities to reduce Freon-113. The runs were performed at 200

degrees C and a reactor residence time of approximately 2 minutes with infiuent Freon-113

concentrations of 14.1 and 22 mg/L. The results of the effluent gas analysis are presented on

a bar graph in Figure 7. The retention times of the various by-products separated in a gas

chromatograph column are graphed against their peak heights. The identities of these by-

products and their actual concentrations were not established but are not necessary for a

specific comparison between the two catalysts. Generally, the larger, less hydrogenated by-

products will have a longer retention time. It should also be noted that a longer retention time

means that the gas chromatograph peak will be broader and represent a greater concentration

for the same peak height.

Under the previously mentioned reaction conditions, the nickel catalyst produced only

50% of the ethane (RT=35 seconds) and 36% of the methane (RT=19 seconds) that the

baseline catalyst produced. The nickel catalyst allowed more partially reduced by-products to

pass, indicating its lesser activity. Although a nickel catalyst could be effective where a longer

residence time/larger bed size would be feasible, the aerospace pdorities of low mass and

energy efficiency makes this catalyst less attracti_,e. No further catalytic runs were performed

on the nickel catalyst.

4.3.2 5%Ru, 5%Pt v.s. 5%Pd, 5%Pt

Catalysts composed of 5%Ru, 5%Pt and 5%Pd, 5%Pt were challenged with a series of

three contaminants: Freon-113, dichloromethane, and then bromotrifluoromethane. These

catalysts were tested simultaneously and received the same concentrations of contaminants.

The catalysts were challenged with Freon-113 at 125 degrees C (Figure 8) and then

200 degrees C (Figure 9). As in the previous comparison, the effluent gas is analyzed with a

gas chromatograph and the peak heights are graphed against their retention times (R'l'). As

expected, both catalysts showed greater activity at the higher temperature, even when

considering the increase in residence time. At both temperatures, the ruthenium catalyst

outperformed the palladium catalyst as indicated by its production of more methane and

ethane. The ruthenium catalyst increased its ethane production by a factor of nine due to the

rise in temperature, while the palladium catalyst increased its ethane production by a factor of

four. At 125 C, the palladium catalyst allowed less of the larger partially hydrogenated

products through (RT>70).

The palladium catalyst showed more promise when the columns were challenged with

URC 80436 15

w

Catal ,s[ ComparisonFreon-113 at 200 deg C

U

=

w

E ::

w

_ I

w

" 500

450

_o 400t-o_ 350

7- 300"" 250--

a_ 200-

150-

lOO-50-

019 35 54 67 78 146 194 278 420

GC Retention Times (Seconds)

I[_] RuPt1.9min,22mg/L _ Ni,2.3min,14.1mg/L I

Figure 7. Baseline Catalyst vs. 5% Nickel Catalyst forthe Reduction of Freon-ll3 at 200 Degrees C

URC 80436 16

= .

Catai ',, t Comparison125 C, 31.2mg/L Freon-113

800

700

054 69 80 148

GC Retention Time (seconds)

196 290

Pd,Pt,2.05min. _ Ru,Pt, 2.34 min.

Figure 8. 5%Pd, 5%Pt v.s. 5%Ru, 5%Pt Catalyst for theReduction of Freon-ll3 at 125 Degrees C

z

L

URC 80436 17

|7

U

Catalyst Comparison200 C, 35.6 mg/L Freon-113

" 2500-

. _=_

L

, =I

019 35 49 52 54 69 80 145 196 290

GC Retention Time (seconds)

Pd,Pt, 4.61 min. _ Ru,Pt, 4.24 min.

Figure 9. 5%Pd, 5%Pt, v.s. 5%Ru, 5%Pt Catalyst for theReduction of Freon-ll3 at 200 Degrees C

URC 80436 18

W

w

i

12 to 16 mg/L of dichloromethane for two weeks. Both reactors were at 125 C with contact

times of 1.8 to 2.4 minutes. The palladium catalyst allowed only half as much unreacted

dichloromethane through: 3 mg/L as compared to 6.3 mg/L for the ruthenium catalyst. The

palladium catalyst had almost three times the amount of intermediate product, chloromethane,

in its effluent but did produce slightly more methane.

Neither of these catalysts was able to match the effectiveness of the baseline catalyst

for dichloromethane reduction. When challenged with 49 mg/L of dichloromethane at a

contact time of 1.50 minutes and a temperature of 125 C, it allowed only 4.35 mg/L

breakthrough of dichloromethane. The extra 15% platinum content of the baseline catalyst is

the only difference between it and the 5%Ru, 5%Pt catalyst, and therefore is the reason for

the baseline catalyst's greater activity.

The catalysts were then challenged with 7.6 mg/L of bromotrifluoromethane at 125 C

and then 200 C. At 125 C, a reactor residence time of 6 minutes was established and the

effluent gas tested for methane and other by-products. The ruthenium catalyst produced three

times the amount of methane, indicating a greater activity. The ruthenium catalyst did allow a

significant amount of breakthrough, 6 mg/L of bromotrifluoromethane but the palladium

catalyst allowed 6.3 mg/L breakthrough. This greater performance was repeated at 200 C

where the ruthenium catalyst produced twice as much methane.

Overall, both catalysts produced results indicating the ability to reduce these three

halocarbons to alkanes. Although the palladium catalyst showed a greater activity toward

dichloromethane reduction, the ruthenium catalyst was more effective against Freon-113, the

major component in the TCCS model. This makes the ruthenium catalyst more useful for our

current purposes. The additional platinum in the baseline catalyst increases the activity

toward dichloromethane reduction.

4.3.3 The Addition of 10% Molybdenum to the BaselineCatalyst

The addition of molybdenum to the baseline catalyst was an attempt to enhance its

ability to react with sulfur containing compounds. Molybdenum by itself or in combination with

nickel or cobalt has demonstrated superior performance for the hydrodesulfudzation of sulfur

containing organic compounds. Since the system was already set up with the

bromotrifluoromethane infiuent system and the baseline catalyst had not been tested with this

influent, the baseline catalyst and the baseline catalyst plus 10% molybdenum were first

challenged with this contaminant. Figures 10 and t i track the reactors' influent

concentrations, residence times, and effluent gas components compared to the total amount

URC 80436 19

_-" a 5%Ru,20%Pt on Corbon

-- =

Z=-

L_

P !

40O

350

500

(-

._ 250(D(-

2000o

150(D

° 100

50

0

10

dt-O

o'3"9

c CY_

n 8

c-

7

.E 14

E 12

10

6

4o 2

Figure i0.

iI I I ; I I I

Ii

• RT=20 Methanb- [] RT=24 =

o RT=25- • • RT=70 CBrF_ ;i /o_

A o RT=210 J !/ \

i -_-e-

0 200 400 600 800 1000 1200 1400

Totol BromotrifluoromethQne Throu9hput (rag)

I I I I I I

i -I l L I 1 I _ _

I I 1 I I I

TEMPERATURE

CHANCE

125-150 C

Baseline Catalyst Reduction of Bromotrifluoromethane

--z

URC 80436 20

w

m

5%Au,20%Pt ,10,%oMo,/Ca rbo n

c-

.£i

q)Q_

C)(-9

400

350

300

250

200

150

100

50

0

= =w

|_10

dc0

o "_" 9

c- O4

t,___

c

7

._ 12

10

E 8iL _=(..)

i[i c 40

o 2

w

I I I I I I

_• • RT=20 Methane

\•"A__ 0 RT=25

4, RT=70 CBrF70 RT=210 "_

- \ •

I l I _1_ _--_lL ,_, ____

0 200 400 600 80C 1000 1200

Total Bromotrifluoromethane Throughput (mg)

I t I I i I I

I I I I

__+ O___ #,

I l l I _ Ii

TEMPERATURE

CHANGE

125-150 C

I I

#"v

1

Figure 11. Baseline/Molybdenum Catalyst Reductionof Bromotrifluoromethane

1400

URC 80436 21

w

w

jw

of bromotrifluoromethane throughput. The data was accumulated over a span of forty days

and included a change in temperature from 125 C to 150 C.

The molybdenum containing catalyst initially produced more methane than the baseline

catalyst, but its production decreased rapidly without a corresponding increase in the

production of other by-products. When comparing the total amounts of effluent by-products

from the two reactors, one notices that the molybdenum catalyst's effluent is lacking carbon

mass. The two reactors were fed the same contaminant concentrations, but the molybdenum

containing catalyst produced less of all of the by-products. This is only possible if the

molybdenum catalyst is producing and retaining tars or reducing the contaminants to

elemental carbon.

The catalysts were cleansed of by-products by a flow of hydrogen at operating

temperature. When an analysis of the effluent gas showed that no more by-products were

being desorbed by the catalysts, the reactors were cooled and the catalysts removed and

weighed. The molybdenum catalyst had gained 0.2 grams while the baseline catalyst had lost

a slight amount of weight, most likely due to the loss of ambient moisture previously adsorbed

before being placed in the reactor. Since the molybdenum catalyst did not produce any larger

compounds indicative of intermediates in tar production, it had most likely produced carbon.

This coking had not affected the catalyst's activity during this run; however, it might over a

longer span of time.

i u

I Ir_

i

-N

w

, w_,

i

5.0 THE FINAL TEST STAND CONFIGURATION AND RESULTS

5.1 The Final Test Apparatus

The final test system (Figure 1) is operated in two separate cycles: adsorption and

cataiytic regeneration. In the first cycle, a contaminated air stream flows through the

carbon/catalyst bed where the contaminants are transferred from the air to the sorbent

surface. The effluent from the sorption bed is continuously monitored for organic

contaminants by a hydrocarbon analyzer which uses a flame ionization detector and will

detect a breakthrough of less than 0.01 mglL of the components in the contaminant stream.

The sorbent bed is composed of 4cc of the baseline catalyst and 4cc of Carbosieve S-

Ill mixed together. The Carbosieve shows a much greater affinity for the smaller organics

such as bromotrifluoromethane. The incorporation of Carbosieve in the bed dramatically

increases the systems sorption capacity for the mixed contaminant load.

The second cycle is the catalytic reduction/sorbent regeneration cycle. During

regeneration the flow path is switched to include the CARS bed, a hydrogen source, a pump

URC 80436 22

! _--4

= _-----,

_- _-c-:

_- _- •

4 kz_E

Z

=

w

for gas recirculation, and a lithium hydroxide bed for acid gas removal. The system is first

purged with nitrogen to remove excess oxygen, which would react with hydrogen. The system

is then purged with hydrogen and pressurized to approximately 14 PSIG. Hydrogen is kept on

line to supply makeup hydrogen as it is used up in the reaction. The reactor is then heated at

a rate of 5 C/minute until operating temperature is reached.

As hydrogen is recirculated through the reactor, intermediate by-products and

nonreacted contaminants desorb into the hydrogen gas along with the acid gases and alkanes

produced by hydrogen reduction. The acid gases are removed when they react with the

lithium hydroxide bed to form nonvolatile lithium halides and water. The hydrogen sulfide

produced by thiophene reduction will form lithium sulfide, which is also nonvolatile. The

partially reduced organics, unreacted hydrogen, and alkane products are recirculated through

the catalytic reactor until an acceptably small amount of non-reduced organics are left in the

hydrogen stream. For the purposes of this project, the gases are then purged from the

reactor with a flow of nitrogen at operating temperature. The reactor is then cooled to prepare

for the next adsorption cycle.

In the TCCS model, Freon-113 is the major halogenated constituenL and thisis '

reflected in the model contaminant mixture chosen to test the recirculation reduction system.

The other components were added at the minimum measurable concentrations. In later runs,

the detection limit for bromotrifluoromethane was reduced, allowing for lower infiuent

concentrations more in proportion with its concentration in the TCCS model.

5.2 The First Run

The first contaminant stream tested Was composed of:

BromotrifluoromethaneDichloromethaneFreon- 113

ThiopheneTrichloroethylene

0.50 m_L0.078 mg/L0.96 mg/L0.086 mgiL0.092 mg/L

The adsorption was continued for 200 minutes at a flow rate of 500 cclmin, for a throughput of

100 liters. The only contaminant to break through was bromotrifluoromethane, which started

breaking through at 80 minutes and reached a concentration of 0.1 mg/L by the end of the

200 minutes.

The catalytic reduction/regeneration cycle for this adsorption was not maintained at any

one temperatureor hydrogenpressure. Parameterswere vaded to determinetheeffects on

the system. As the reactor bed was initially heated up, a variety of partially hydrogenated

contaminant by-products were detected in the recirculating hydrogen. None of the original

URC 80436 23

= =

=lee=

=

=

w

r=

w

w

Lm

L_

J

w

-=

contaminants were detected in the effluent, indicating that the contaminants were reduced at

least once before escaping the bed. The by-products of the initial reduction are either

readsorbed and reduced again or continue to recirculate. It was also noted that a removal of

some of the recirculating gas through the sample port caused a predictable reduction in the

contaminants' concentration due to dilution by makeup hydrogen. The larger, less volatile by-

products recovered partially from this reduction in concentration, whereas the more volatile

components did not. The reactor's carbon retains a portion of these larger contaminants and

re-equilibrates with the hydrogen gas when a sample is removed.

With over thirty possible intermediate by-products for Freon-113 alone, there are

difficulties in quantitatively describing the progression of a catalytic regeneration run. The

gas chromatograph has revealed fifteen major by-products whose peak heights have been

recorded throughout these regeneration runs. Methane, ethane, propane, and butane have

been identified by retention time. The catalytic reduction would be considered complete when

these are the only peaks remaining.

A desorption of the remaining by-products was performed at 200 degrees C and a

nitrogen flow rate of 10 cc/minute. It was found that some of the larger by-products were still

being desorbed 1200 minutes later. A higher desorption temperature will reduce the required

desorption time.

5.3 The Second Run

The next adsorption run was performed=a_ the same flow rate and duration as the first,

using the same influent. At the end of the run, a breakthrough of bromotdfluoromethane at a

concentration of 0.25 mg/L was detected. The previous desorption of contaminants was

incomplete, and these retained contaminants competed with the influent for adsorption sites.

During the regeneration cycle, the reactor was maintained at 200 degrees C for the

first 460 minutes and then at 250 degrees C for the rest of the run. Figures 12, 13, and 14

track the concentrations of fifteen of the by-products through the entire catalytic regeneration.

The regeneration continued for approximately eight hours at a time; the dotted vertical lines on

the graphs indicate the beginning of a day's run. Each night the system was cooled down to

room temperature and the makeup hydrogen was turned off. There is a noticeable drop in the

by-products' concentrations after each shutdown with the more volatile components being the

most affected. Small gas leaks in the system cause a loss of some non-adsorbed

contaminants.

There are trends in the contaminant concentrations relevant to all of the catalytic

regenerations. When hydrogen is initially recirculated through the system at room

URC 80436 24

w

v

|

-: -!

_='_

[,9--I--'

_.C

C_

T

(3(1)n

c-£)_ID

o

(3

Eo

_.C0

GO0

(.9

URC 80436

30000

25000

20000

15000

10000

5O0O

0550

3OO

25O

2OO

150

IO0

5O

0 -

I I

RT=22

0 200 400 600 800 10001200140016001800Recirculation Time (minutes)

Figure 12. Fina] Test Stand's Second Regeneration Run

25

m +

E :

L_

.,,..'

i

i

= =w

==

i

_ +

w

L.+

URC 80436

IOO0

F :

800 !_

_ :' RT=50! :

600 i :

i'400i

200 ] l -

0 ._ I ]: I I I 1 ..... Ii

I I f..... I I I i l

_ ,_"-w% =- i- !

_ ] I I I I I

, ,[i , _I ' ' ] '

_

120 i I I I

lOO - RT=82

4O - i_1

20 'i IkOll_ _ -

0 _ ! I _ I ! I -_'I'i_-0_I_II,00___ -

600

;i i

:+ i RT=IO0 -

,:Pr°pane'100

O I-- ---- I I I----_ t I

0 200 400 600 800 10001200140016001800Recirculotion Time (minutes)

Figure 13. Final Test Stand's Second Regeneration Run

6OO0

5000

4000

3000

2000

c-CD I000

T 0

__ 3500o

3000G_

._c 2500O_0 2O0OL

15000

0 1000

5000L

/z 00

00

500

400

300

200

26

E

W

w

J

w

w

E_

J

i

E

L_J

T

,,./O(t)

n

c-

O

O,.-I.-"

O

Eo

(-

6OO

O

URC 80436

1oo

80

6O

40

20

04OO

I 1 Li I I I I 1i:

- !! _,_. RT=150 -

i 1 -

I t ii I i l I I i

300 RT=200 -

200 -

100 _ -

, i_ _o,0

50I I il I_ I I I I I

4o RT=290 13O

2O

"350 I i-" I I i I I i

300 - i _

250 - 00_

200-150 - i " i Bu_(_ne

100 -

5o- _ i -t

0 I --_ --- , I I I L.... t Ii

40 I I i!q I I I I I

i RT=81 03o-

20 -

_lo ii

0 200 400 600 800 10001200140016001800

Recirculotion Time (minutes)

Final Test Stand's Second Regeneration RunFigure 14.

27

F

z

- =

i - il-

= =

===_.__...,..

temperature, the only detectable compounds are a trace of methane and sometimes ethane.

As the system is heated at a rate of 5C/minute, all detectable compounds start desorbing from

the carbon surface and increase in concentration in the hydrogen gas. The more volatile by-

products of hydrogenation increase in concentration in the gas stream faster since they are

not well retained in the carbon bed. Some intermediate by-products, RT=22 and RT=50, are

quickly reduced to alkanes; while others such as RT=810 require more time to be reduced.

During about the first half of the run, intermediate products like RT=25, RT=48, RT=55, and

RT=70 are being produced by the hydrogenation of larger by-products and are also

undergoing reduction.

The peak at RT=810 is a by-product of thiophene reduction. It is most likely

thiobutane, the sole source for the butane and propane by-products. This conclusion is

supported by the fact that the production of propane and butane stops when RT=810 is

depletedl

When the reactor temperature was raised from 200 to 250 C, indicated by the vertical

dashed line, the concentration of all by-products in the hydrogen stream increased due to

desorption from the carbon catalyst. There is also a noticeable increase in the rate of

contaminant reduction. An increase in reactor temperature increases the amount of thermal

energy available to activate the hydrogenation reactions. The higher temperature also

increases the mobility of adsorbed species on the carbon surface and in the gas phase, and

decreases the concentration of adsorbed species. The kinetics are most probably dominated

by the thermal energy term.

5.4 The Third and Fourth Runs

The third and fourth runs were both performed with a new contaminant stream

consisting of:

BromotdfluoromethaneDichloromethaneFreon- 113

ThiopheneTrichloroethylene

0.068 mg/L0.083 mg/L1.67 mg/L0.075 mg/L0.081 mg/L

This contaminant stream more closely reflects the ratios of contaminants in the TCCS model.

Both of these adsorptions were performed at a rate of 500cc/minute for 200 minutes.

Negligible breakthrough was detected in both runs.

The catalytic regeneration for the third run is shown in figures 15, 16, and 17. The

regeneration temperature was ramped from 23 to 250 C at 5C/minute. Since this regeneration

run was ramped to a higher temperature, the reducible components disappear more rapidly.

URC 80436 28

w

w

i | Li i.--.,

7 -

_J

U

; .'=

I L

I .F.,

03

C-O_

-r-

O(1)

n

J::cLo

ET_O

(D

EO

F-(D

co(D

(._9

URC 80436

55000 I I r "]

t3oooo METHANEi

20000

15000

10000

5OO0

O, , I J I

550 lip I , I I I t

300I-I ! i -H RT=22

25017 i i -200 I-I

i i150 Pl i , -la , i i

,oo1t I ' I -5oI_ i '_ -o_: i, , ' , ,'

35001 ' l ' ' '3000 b .. ,'_ --J I RT= 25 -

_oo1__-,,7_ooorl _.._.i1500 FI '---- q_ll !H ' 1

10001- / I1'_1'4__

_ooH - i"-"n ,,I,_ l i _ _1 i I ! _1_

_oooooI- ,,.,.;' i ' i ' ' ' -| _/_'_! I i ETHANE

8oooo__ !-."..%. ! _ -_ooooI-, I _r_,,,_ i _

iz ! ' i"_-,-,,.I

20000h, t io_ ', !, _, ,

_°°°/._t_i i' ' '

600 _-i400!

200 I !

1 'o.: , , _, I. ,0 500 1000 1500 2000

Recirculation Time (minutes)

Figure 15. Final Test Stand's Third Regeneration Run

29

w

z

LJ=

w

i

6O

(-C_

Q_I

_3C(3(I)

El_

C-ELE)

ED0

O

F0L.

(-C_)

03(3

(_9

URC 80436

J00 I I I 1

150 t

I00 1

5_q I ] I I I -

12ooo_ , ' i-' '10000 -_

8000 -% RT=55

6000 -k

4°°° i ".% ,

2°°°/- _'__n,_- J l -'ltlt:l_.lll_ _i_L ....... I --

_1_ == i v_ _- vvvvv vv v

8000 l if,, • i' I v , --_r---

ooo11 ! , -It I -_

2000H i i i _ '

F,.__,300 1-_ i i i RT=82 i -

I]1 1 ii i r ]

l l,_ ! i I !11% i i i

_ooM _ i i i ! -II "__ I ! _ '

_,L ....¢t 4\#-"'r'__

100

5O

Oq

0

Figure

r I

500 1000

16.

IRT= I00!

Propane

I L

1500 2000Recirculotion Time (minutes)

Final Test Stand's Third Regeneration Run30

t

_=!

w

L

_4

iii '

1200 tT_

800 II _ I

•_' 0,11 "i,,_ii.,.._I

Q_

t-

%,_

o

oEo

c-O

o0

120

URC 80436

100

80

60

4-0

20

oq

550

500

250

200

150

100

50

Oq

200

150

lO0

50

o, i_

0

i

i

l

1

I I

tl 1

I

RT= 150

I

RT=200

I

RT=290

I I

I

[

I I

5OO 000

--.....%RT=420Butane

i

I i

' IRT=81 0

I

5OORecirculotion Time (minutes)

I I

I

I ,

Figure 17.

! I_

2000

Final Test Stand's Third Regeneration Run

31

w

=

w

L

W

m

iw

_=

m

u

h_w

w

Some contaminants are very resistant to hydrogen reduction as exemplified by peaks RT=25,

RT=48, and RT=70. Even after 2000 minutes, they still remain at significant concentrations in

the gas stream. One possible delay is the poor exposure of contaminants to the catalyst. In

an attempt to rectify this problem, the hydrogen reservoir was reduced in size for the fourth

run. This reduced the gas volume outside the reactor from 59 cc to 18 cc. The results of the

fourth regeneration are shown on figures 18, 19, and 20.

A comparison of the first 2000 minutes of the two runs indicates that no significant

improvement in halogen reduction rate occurred due to the removal of the hydrogen reservoir.

At 1937 minutes into the fourth run, the temperature was raised to 300 C to increase the rate

of hydrogenation. All of the compounds present in the recirculation loop increased in

concentration due to thermal desorption. The contaminants at RT=48 and RT=70 were

reduced at a faster rate due to the 50 degree temperature increase. The by-product at RT=25

effectively mimics the methane concentration. Its concentration only decreases in between

run times when leakage occurs. Its total inability to be reduced when the regeneration system

is operating implies that it is probably a cyclic alkane or an inorganic by-product.=

5.5 The Fifth Run: The Molybdenum Catalyst

The fifth run was performed using a mixed bed composed of 4 cc of the 5%Ru, 20%Pt,

10%Mo catalyst and 3.8 cc of Carbosieve recovered from the previous sorbent bed. The

molybdenum containing catalyst was previously shown to be more active than the baseline

catalyst in reducing bromotrifluoromethane. The same contaminant mixture and adsorption

procedure used in runs 3 and 4 were used for this run. A little more breakthrough occurred in

this adsorption run, but the amount was less than 0.01 mg/L. The extra breakthrough was

probably due to the lower amount of Carbosieve in the sorbent bed.

The reactor temperature was ramped from 23 to 250 C at 5C/min for the first day of

this regeneration (Figures 21, 22, and 23). After reaching temperature, none of the odginal

challenge contaminants was identifiable in the recirculation gas mixture demonstrating that the

first hydrogenation step occurs very rapidly. All of the compounds previously shown to be

prone to hydrogen reduction were reduced significantly faster. The peak at RT=25 still

showed no reduction. The peak at RT=290 appears to be the result of two different

compounds with the same retention times. One of the compounds is the same reducible one

found in the other regeneration runs. The other is produced while thiobutane, RT=810, is

depleted and then remains at a constant concentration. Its resistance to hydrogen reduction,

along with its retention time being equal to that of a component found in butane lighter gas,

implies that it is isobutane. The molybdenum catalyst may cause a rearrangement of a methyl

URC 80436 32

w

E_

W

- __- -

--7

i i._4

iL

O3

C'-

I

E)(D

Q_

c-O_E)L.

E_o

-I--'

(3

EO

c-

O

O3O

C.D

60000

50000

40000

30000

20000

10000

o:350

3o0

25o

2oo

15o

too

J il i_l: I l r--, ii i

I i i ' METHAN[Ei i, i -r

' 00 i i

iii I i t _1

_T=22

50 -

0i __ _L ........ i__J I ,I3500 I _ I

oooI ' ''i'''i2500_iI1_ i RT=2_ -

_ooo__ i l ',-,

_oo_ , _ __.___ooo_ ' _ A isoo ,'_.i ---o_* ....... ' ! 'l i, I ,

140000 _ _ , __.._,., _THAN_ =

._oooof _.,..._ i _r'_ I,ooooo, ¢- ,_-_._, _-.-.__ 48000o ] i i -"'"_L,,_J

, i ,t"- I60000 ] ] li I I ! q

, 1 i _ I i4oooo- , , ,, , I . i 4i ] f: I_oooo- i : :i , 1 _ 4

_ooo_ __.__''___ ._i

1400_ ! _T=48

1ooo_ __800 H _"-- T__,,__ I I600 H _ I ", : ! "@_,@_1._I ,400 , ' i " ' i I

0 500 1000 1500 2000 2500 3000 3500Recirculotion Time (minutes)

Figure 18. Fina] Test Stand's Fourth Regeneration Run

URC 80436 33

6OO

5OO

w 400

_. 300

i 200

IO0

URC 80436

:

0

Figure 19.

|

500 1000 1500 2000Recircutation Time (minutes)

Final Test Stand's Fourth Regeneration Run

34

I i|' _..

I_ropan.ei I1 II i

2500 3000 3500

= ,

w

r_

J

== =

L

i

__!=

t_

L_

F_

600 1

5OO

400 i

300 t I

200

1oo_ • i01 ' ------_.A¢--- -@_

2000

1600 _I

1200 _q

o _°°___ 400"_ o--_ -:::-- |_ _o4_

] :I

i:

I

' i'RT= 150

II

I_ I'"_'E_ 120 I-_q3n_ I?[ i ,

_c 8o Ili !II i

4oll f ic- 800 !

I il i© ! iI ! I

o oooF i !

4°°I.,._I !_oo_--_.,-_-._II i I

oqlt_ !J i' i!

!

i

ii

[__1

o

m1

300 ,

250 f

200

150

100 _1)

I I :1I

_T=29_) -

Butane ]

' ti i

I_T=81E_

]

i!

! i I t

ii '

I I

3000 3500irculation Time (minutes)

_!

i i:

i' !i '

i

500 1 30 1500 2000 2500Re,

URC 80436

Figure 20. Final Test Stand's Fourth Regeneration Run

35

140000

120000

100000

8OO00

60000

4OO00

20000

O_

w

150000

I

I I

w

= m

w

i

36URC 80436

100000- [ I

_oooo_ ii ii0_--" I _ i ..... I I II I

II2000 .....

Jl.i _i I Ir ...... I

,ooo_.t i % _/ I i *.4

_oo_._ _

0 200 400 600 800 1000 1200 1400 1600Recirculation Time (minutes)

Figure 21. Fina] Test, Stand's Regenerat,ion Run Using theBase] ine/Mo]ybdenum Cat, a]yst

w

= T

w

W

i

= =

wml

IL!

W

m

68.+_,,

._c

©T

¢>(2_

._c(22)_

(D7O

O

EO$._

c-O

cooO

URC 80436

lO I

8 -

6 -

4 -

2 -

IO( m

2000 t I

1500

1000

500ti

2000i I

, ootl,°°°ItI

20 I

lS

10-

,)1' )

'ii

i

, i

i

1] _I :

I

'i

J i

i

!

I I I

I I i I

I L__

I I

l I

I I

I

RT=50

RT=55

I ....

RT=70

i il i

i I

RT=82

5

i I0(m l [ i i ) i i m J

, i !4000 L I I i : I lal I I

i|. ! .--._.._'-'_ __._ooob 12, _ -,_-_-

/ IT i RT=IO0_ooor y , _ro_one,ooo

O_L_ I I i .j I I d I

0 200 400 600 800 1000 1200 1400 1600Recirculation Time (minutes)

Figure 22. Final Test Stand's Regeneration Run Using theBasel i ne/Molybdenum Catalyst

37

w

# !

L

_=

iii-,

--.=

L=J

iiii

i

_g

U3

r-

T

O

cl

c-c3_OL

O

O

EOL

t--(D

(_9

URC 80436

300

250

200

150

1O0

5O

080O

6OO

4OO

2OO

0

150

12O

90

60

30

0,"

1600

1400

1200

1000

8OO

600

400

200

O;

1000 _

800 -

600 -

400 -

,ii

"_1 I i I ,I I

RT=I 50

--: -----:"---- --'-:m W_H_,0_Hm-Loo I

1 I ]--I I I I

! RT=200

!

t

I I I : I I I II _ I

i ' i RT=290

_ , !_.,_,_,¢.,,_,_I ! "-_ i

- i i _

- i: i

i= I ,F_,; • Ii

_ I I

_ ,,_..ow"_ i

-4

I I

•"_"'%-i".........'....%.,_,,m -RT=420 -

Butane i

I 1 I

I I

RT=810 -

0 200 400 600 800 1000 1200 1400 1600Recirculotion Time (minutes)

Figure 23. Final Test Stand's Regeneration Run Using theBasel i ne/Molybdenum Catalyst

38

w group to the center of the carbon chain during reduction.

At 500 minutes, the temperature was raised to 300 C. As previously observed, all

remaining contaminants increased in concentration in the gas stream. By 1600 minutes

almost all of the reducible components were destroyed. A mechanical failure precluded the

total reduction of RT=48.

|

z

ZZZ

==w

E :

_4

=

6.0 TEMPERATURE, HYDROGEN PRESSURE, AND PRESSURE DEPENDENCEFOR THE REACTION OF DICHLOROMETHANE OVER 5% RU, 20% PTCATALYST

The kinetic factors which determine the reaction rates during regeneration of a CARS

bed include temperature, the hydrogen partial pressure, and the concentration of the

contaminant. For contaminants such as Freon-113 which are hydrogenated in multiple steps,

a kinetic analysis requires a detailed consideratioh of the sequential reactions which are;L IiZZ _Z£

involved in the complete hydrogenation reaction. Such an analysis is complex and must

consider not only the catalytic surface reactions but also the competition for surface catalytic

sites from the various Species present in the gas phase. Such an investigation is beyond the

scope of the present study. In order to gather preliminary kinetic information on a relatively

simple hydrogenation reaction, dichloromethane's reaction rate as a function of concentration,

hydrogen pressure, and temperature was determined.

Using the fixed bed reactor system previously shown in Figure 5, the temperature

dependence of rate of dichloromethane disappearance was determined. The reactor was

challenged at three different temperatures (125 C, 150 C,and 200 C) with increasing

concentrations of dichloromethane until breakthroug h occurred. The breakthrough

concentration is dependent on the reaction rite and the catalyst bed size which is constant,

and if all catalytic sites are assumed to be saturated at breakthrough, represents the reaction

rate. A plot of the logarithm of the breakthr0ughconcentration versus reciprocal temperature

is shown in Figure 24. The activation energy calculated for this Arrhenius plot is 51.5 KJ/mole

(12.3 Kcal/mole). Using this activation energy increasing the temperature from 125 C to 250

C will increase the reaction rate -41 times.

The determination of the dependence of the reaction rate on the dichioromethane

concentration also assumes that all catalytic sites in the CARS bed are saturated. The

reaction conditions including temperature, hydrogen pressure, and flow rates were held

constant. A plot of the dichloromethane concentration versus reaction rate is shown in Figure

25. All data for this plot was recorded well after breakthrough of dichloromethane had

URC 80436 39

=

10

Temperature Dependence of

Dichloromethane Hydrogenation

I I

U

w

L

U

c

8

cp

E

@

0Ct2

c-O

00@

0.1

0.0020

5t.5 kd/mole

1 "_"_ 200 C 2.5 kcal/mole

I I

0.0022 0.0024 0.0026

1/degrees K

Figure 24. Temperature Dependence of Dichloromethane Hydrogenation

URC 80436 40

r

r

z z= =

w

c

oh

E

q)

orY

co

oE)¢)

rY

12

First Order Relationship of Dichloromethane

to the Reaction Rate at 125 Degrees C anda 1.5 Minute Contact Time

2

0

I I I I I

I I I I I

25 50 75 1O0 125

Dichloromethane Influent Concentration (mg//L)

0

Figure 25. First Order Relationship of Dichloromethaneto Reaction Rate

5O

w

m_U

URC 80436 41

-=. :

m

w

J

J

occurred. The slope indicates a linear dependence on the dichloromethane concentration. In

general during regeneration, this information means that a CARS bed will initially see high

reaction rates due to the high concentrations of desorbed contaminants. As the regeneration

proceeds, the reaction rate will decline, and at some point a decision is made to stop based

upon the time required for further regeneration relative to the current regenerated sorption

capacity.

The hydrogen pressure was varied between 0.08 and 1.00 atmospheres with a

constant influent level of dichloromethane. The system was maintained at 1 atmosphere with

nitrogen used as the dilution gas. A plot of the reaction rate versus the hydrogen

concentration is shown in Figure 26. At higher concentrations of hydrogen, the curve gives a

nearly linear relation between hydrogen pressure and reaction rate; however, as the hydrogen

pressure is lowered the curve bends over giving lower reaction rates than expected. The

curvature of this plot is most likely due to competition for adsorption site between nitrogen and

hydrogen which depresses the reaction rate at lower hydrogen pressures. The reaction rates

for regeneration can be increased by raising the hydrogen pressure within those limits

imposed by competitive adsorption from the contaminant species. Since hydrogen adsorbs

ve_ poorly, its pressure can likely be raised several orders of magnitude before it can

effectively compete for adsorption sites with a species such as trichloroethylene.

7.0 CONCLUSION

The feasibility of a Catalytically Active Regenerable Sorbent (CARS_) with non-catalytic

sorbent additions that adsorbs problematic airborne contaminants containing hydrogen and

sulfur, and then chemically converts them to more readily treatable by-products during

regeneration has been demonstrated in this study. This assertion is based on analysis of

adsorption, thermal desorption, catalytic hydrogenation, and regeneration data. The

contaminants tested include Freon-113, trichloroethylene, bromotrifluoromethane,

dichloromethane, and thiophene. Individual adsorption data for these compounds indicated

that the more volatile constituents of the mixture, dichloromethane and bromotrifluoromethane,

sorbed poorly on the baseline CARS. This problem was overcome by the addition of a non-

catalytic sorbent, Carbosieve, which is tailored to sorb such low molecular weight highly

volatile organics and which can be thermally desorbed. Thermal desorption data show that

desorption of contaminants precedes the onset of high catalytic activity, and consequently, the

regeneration (catalytic hydrogenation) of a CARS bed must proceed in a recirculating system

where desorbed species have additional contact with the catalyst. Gas analysis by GC during

URC 80436 42

F

L _

U

-- 4W

i

LI

c

E

Oh

E

q3

0rY

C0

L)0(t)

Od

1.0

0.9

0.8

0.7-

0.6-

0.5-

0.4--

0.3-

0.2-

0.1 -

0.00.0

Hydrogen Pressure Dependence ofDichloromethone Reduction

I I I I I I I I I II

Influent= 97 mg/L Dichloromethane qTemperature= 125 C /Contact time= __

I I I I I I I I I

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Hydrogen Pressure (Atmospheres)

L_

L--

I

Figure 26. Hydrogen Pressure Dependence ofDichloromethane Reduction

__:=

L_

U

URC 80436 43

m

r 7

i

U

w

I

regeneration of a CARS bed in a recirculation system proved that the hydrogenation by-

products are simple alkanes. The complementary acid gases (HCI, HF, H2S, and HBr) are

removed in a room temperature LiOH H20 bed within the loop.

A complete adsorption and regeneration test performed on the best CARS combination

clearly illustrates the effectiveness of this technology for the treatment of halogen and sulfur

containing contaminants. The challenge mixture used contained Freon-113, trichloroethylene,

bromotrifluoromethane, and dichloromethane at 436%, 15000%, 11.2%, and 95.6% of their

Spacecraft Maximum Allowable Concentration (SMAC) respectively as well as thiophene, a

generalized organosuifur compound without a SMAC value. _ Approximately 100 liters of this

challenge stream was run through a 7.8 cm3 bed before breakthrough occurred. The bed was

then purged, filled with H2, and heated to 250 C for 500 minutes followed by 300 C for 1100

minutes. After the initial heating during the regeneration cycle, none of the sorbed

contaminants were detected demonstrating the high catalytic hydrogenation activity at these

low temperatures. Subsequent GC analysis reveals the multistep hydrogenation mechanism

by which more complex organics are hydrogenated. The fact that Freon-113 requires six

steps for complete hydrodehalogenation slows the overall rate at which regeneration occurs.

Nevertheless, after 1600 minutes the regeneration was complete and the CARS bed

serviceable for a new adsorption run. This performance can be improved by increasing the

hydrogen pressure and reactor temperature, and by improvements in the catalyst.

The CARS approach to remediation of problematic trace airborne contaminants can be

used to replace expendable sorbent beds in the current TCCS configuration. The impact of

the use of CARS in the TCCS will be lower resupply penalties, greater flexibility in handling

contamination events, lower overall system weight, and possibly lower energy consumption.

This last assertion is based on a projected lower catalytic oxidizer operation temperature due

to an influent which is well defined consisting of known concentrations of alkanes without any

catalyst poisons, and possibility of operating the catalytic oxidizer only during a regeneration.

There are other potential applications of the CARS concept. The Environmental Protection

Agency policies on the release of chlorofluorocarbons (CFCs) will require industry to capture,

recycle, replace, or destroy their CFCs. The current list of regulated CFCs can only be

expected to expand, and the need for new technologies to deal with industrial use of these

chemicals will grow. The CARS technology can be used to capture relatively low

concentrations of CFCs and then to destroy their harmful characteristics.

w

URC 80436 44

w

=w

L_

w

iU

8.0 FUTURE WORK

The practicality of replacing expendable sorbent beds in the TCCS with a CARS

system depends on its ability to load all contaminants to practical levels, and then to

catalytically hydrogenate oxidation catalyst poisons without undue influence from other

adsorbed species. This initial NRA project has investigated only a few of the haiogenated

organics composing the TCCS model. The adsorption characteristics of the other TCCS

components, including those which are not oxidation catalyst poisons, must be determined in

order to design a bed which will be effective against a realistic mixed stream. The effect that

the presence of the other contaminants will have on the halogen reduction and separation

during the regeneration phase must also be investigated.

It was shown in this investigation that the addition of molybdenum to the baseline

catalyst significantly changes its catalytic characteristics. The molybdenum increased the

catalysts effectiveness but caused it to produce elemental carbon. The effects from adding

other base metals to the baseline catalyst should be investigated.

At the end of a regeneration cycle, unreacted hydrogen will be present in the

regeneration loop. Methods using a regenerable metal hydride bed should be investigated

which will release hydrogen into the regeneration loop when needed and will readsorb the

excess hydrogen at the end of a regeneration run.

A method for detecting the breakthrough of halogenated organics as a class of

compounds must be developed to determine when a regeneration is necessary. The most

promising method would be that of using an infrared adsorption system tuned to those

frequencies indicative of a carbon chlorine or _rbc>n'_u0rine bond. A simgar detector could

be used to determine when catalytic regeneration is complete.

=

r_w

w

w

=

URC 80436 45

FE 4

m

=

m_

W

W

w

W

r_

.

,

.

.

.

o

BIBLIOGRAPHY

Lamparter, R. A., "Analysis of Trace Contaminant Control and Process Material

Management Systems for Space Station Freedom, Lockheed Missile & SpaceCompany", Contract NAS8-36406, May 1989, George C. Marshall Space Flight Center.

Link, D. E. and Angeli, J.W., "A Gaseous Trace Contaminant Control System for the

Space Station Freedom Environmental Control and Life Support System", presented atthe 21st International Conference on Environmental Systems, San Francisco,California, July 15-18, 1991.

Wydeven, T., "A Survey of Some Regenerative Physico-Chemical Life Support

Technology", NASA Technical Memorandum 101004, 1988

Olcott, T.M.,"Design, Fabrication, and Evaluation of a Trace Contaminant ControlSystem", Paper No. 75-ENAs-23, presented at the 5th Intersociety Conference onEnvironmental Systems (ICES), San Francisco, 1975

Atwater, J.E. and Holtsnider, J.,"Thermally Desorbable Toxin and Odor Control

Cartridge", Contract NAS9-18337 Final Report, July 1990, Lyndon B. Johnson SpaceCenter.

Atwater, J.E. and Holtsnider, J., "Airbome Trace Organic Contaminant Removal Using

Thermally Regenerabie Multi-Media Layered Sorbents, July 1991, SAE TechnicalPaper 911540, In Press, SAE Transactions, 1991.

W

w

URC 80436 46

E

W

REPORT DOCUMENTATION PAGE fo,r. _pp,o_,_oMBNo,ozo4-orBe

! . I . ..I i i i==. ,= , ,

PuI_]{ rtaorlm9 Im_defl for Ihlt ¢Ol_lctl,_rt Of IrlformltlOn fl MllMihrd |O I,_tfiqe I I'toul' Nr ¢¢lgOeltl. m¢ivdtt_j tht, tlmf/of reviewing If_lrud=0nt le==*¢hlng i=lttm 9 dito source|,

gi{l_rm 9 _ mifmtalmrdithe #lti mN_14, und :=ml_etin9 aM rev,twmo the co IR¢Io¢101informwtlgn, lem_ (Ommtntt reaerdln9 tillsbutddln_t_'lm=t¢ o_I,¥ other itl:,e¢1t_f this¢ol|¢tctlOfl O| tnfofm=llo¢_, |n¢l.udir_g SUggetflOnl for reclu¢ln_ ihlt I_wdlm, (g Wishl/_l|On Hltdqu|rtert Servlce_, £Hrl'=,'loretw _O/Infof_lt|ot_ (3t'_rltIOnt a_ I_elmOf|l_ IZll ;elferlon

Daws H_ghwiy, tu_ll IZO4, Af._|to_, VA -sll0_4.10_, =ha Io the Olfl¢_, pf Management aM llUdgltt. PiIDlr_ot_. _tdu=;Uon PfI_i_t (0704.0 IIU_ Wxd_JngtOet. 0C _0_03.

1. AGENCY USE ONLY (Leave blankJ 2. REPORT gAtE .I. REPORT IYPE AND DATES COVIR|D

4, TITLE AND SLIB1;I-TLE' " ......... $i _F'-UNDING NUMBER_ l

Destruction of Problematic Airborne Contaminants ByHydrogen Reduction Using A Catalytically Active,Regenerable Sorbent (CARS) ..

|, AUTHOR(S) ..........

John O. Thompson and James R. Akse

_7, P|RFORMING O.RGANIZATION NAIVE(S) A_IO AUDRE_;_=(I_St ....

Umpqua Research CompanyP.O. Box 791 - 125 Volunteer WayMyrtle Creek, OR 97457

g, SPON(ORING/MONITORING AGENI_Y'_IAM||$) AND AI_DRESS(E$) "

Ames Research Center

National Aeronautics & Space AdministrationMoffett Field, CA 94035-1000

' . k' = I I I I

1 t, SUN_LEMENTARY NOTES

NRA 2-34944 (LML)

NONE

B. PERFORMINO ORGANIZAT-iON

REPORT NUMIIR

URC 80436

• JR _. I

10. SPONSORING/ MONITORIN_AGENCYREPORTNUMBER

NRA 2-34944 (LML)

1_, DISTRIBUTION/AVAILAliiLil"Y STATEMENT

NASA

t.t. ABSTRACT (Maximum ;/00 word;)

12"_'.DISTRIBUTIONCODE

NASA

Thermally regenerable sorbent beds have been demonstrated to be a highly efficient means for removal

of toxic airborne trace organic contaminants aboard spacecraft. The utilization of the intrinsic weightsavings available through this technology has not been realized since many of the contaminants

desorbed during thermal regeneration are poisons to the catalytic oxidizer or form highly toxic oxidation

by-products in the Trace Contaminant Control System (TCCS). Included in this class of compounds are

nitrogen, sulfur, silicon, and halogen containing organics. The catalytic reduction of these problematic

contaminants using hydrogen at low temperatures (200-300 C) offers an attractive route for their

destruction since the by-products of such reactions, hydrocarbons and inorganic gases, are easily

removed by existing technology. In addition, the catalytic oxidizer can be operated more efficiently dueto the absence of potential poisons, and any posttreatment beds can be reduced in size. The

incorporation of the catalyst within the sorbent bed further improves the system's efficiency. Thedemonstration of this technology provides the basis for an efficient regenerable TCCS for future NASA

missions and can be used in more conventional settings to efficiently remove environmental pollutants.

.s ..... _ .._. SUBJECT_¢RMSAirborne Contaminant Control, RegenerableHydrogenation, Holocarbon Destruction

Sorbents, Catalytic

; .i- ____ rrrr i

17. SECURITY CLASSIFICATION t11, SECURITY CLA$SlflCA?JC)N lB.OF REPORT OF THiS PAGE

Unclassified Unclassified

...... g0'45NSN7S4Q.0t-_80.$$00 URC 6 47

SECURITY CLASSIFICATIONOF ABSTRACt

Unclassified

15, NUMBER OF PAGES

46t=, P_i_ CODE

20, LIMif'_ION OFAIISTRACTi

UL

' _.,,,.d,_Fo:_._?_.(R..2_)