EPA-650/2-74-006
December 1973
Environmental Protection Technology Series
-------
EPA-650/2-74-006
DEVICE FOR COLLECTION
AND
ASSAY OF AMBIENT GASES
by
Peter Tsang
Bendix Research Laboratories
Bendix Center
Southfield, Michigan 48076
Contract No. 68-02-0657
Project Element No. 1AA010
EPA Project Officer: Dr. Eugene Sawicki
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
December 1973
-------
This report has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does
mention of trade names or commercial products constitute endorsement
or recommendation for use.
11
-------
TABLE OF CONTENTS
SECTION 1 - INTRODUCTION 1
SECTION 2 - COLLECTOR SYSTEM DESIGN 5
2.1 Collector System 5
2.2 Collector Unit 5
2.3 Collector Cartridge 10
2.3.1 Cartridge Geometry 10
2.3.2 Cartridge Charging H
2.4 Preparation of Absorbent Packings 11
2.A.I Treatment of Celite 11
2.4.2 Coating of TEA-HC1 11
2.4.3 Cobalt Oxide (Co203) 12
2.5 Vacuum Pump (Cast Model 1531) 12
2.6 Power Supply 15
2.7 Timer 15
2.8 Operation/Maintenance 15
2.9 Interface Device 16
2.10 Parts List 16
SECTION 3 - DEVELOPMENT OF COLLECTION SYSTEM 21
3.1 Introduction 21
3.2 Identification and Characterization of Solid Chemical
Absorbents 21
3.2.1 Technical Approach 21
3.2.2 Evaluation Parameters 21
3.2.3 Test Setup 22
3.2.4 Evaluation Method 24
3.3 Evaluation of Candidates 28
3.3.1 Triethanolamine and Supports 28
3.3.2 Thermal Instability of TEA Packings 33
3.3.3 Amino-Alcohol Screening 33
3.3.4 Triethanolamine Hydrochloride (TEA-HC1) 34
3.3.5 Further Work with Celite 39
3.3.6 Recovery by Thermal Method (TEA.HCl/Celite) 39
3.3.7 The N02 Collector: TEA-HCl/Celite 44
3.3.8 NO Absorbent 48
3.3.9 Calibration of Peak Areas 57
3.4 Interference Studies 59
-------
3.5 Development of Collector System
3.5.1 First-Generation Collector Cartridge and
Field Collection System 64
3.5.2 Development of Collector Unit 69
3.6 Sampling Pump 75
3.7 Collection System (Air Sampler) 76
3.8 Interface Device 76
3.9 Preliminary Field Collection and Assay of N02 77
3.9.1 Experimental Procedure 77
3.9.2 Discussion 78
SECTION 4 - SUMMARY AND RECOMMENDATIONS 81
4.1 Summary 81
4.2 Recommendations 82
4.2.1 N02 Collector with TEA-HC1 82
4.2.2 Cobalt Oxide 83
4.2.3 TEA-HC1 Packing 84
4.2.4 Cobalt Oxide Packing 85
4.2.5 Collector Configuration 86
APPENDIX A - TYPICAL RECORDINGS OF COLLECTOR MATERIALS
EVALUATION A-l
APPENDIX B - DIRECTIONS FOR ASSEMBLING COLLECTOR FIELD
HOUSING . B-l
ii
-------
LIST OF ILLUSTRATIONS
Figure No. Title Page
1 Collector System 4
2 Collector System Housing 7
3 Collector Unit Parts 8
4 Specifics of Collector Unit 9
5 Pyrex Glass Connector Cartridge 10
6 Gas Pump and Collector Unit 13
7 Connection for Collector Unit/Filter to Pump/
Orifice 14
8 Details of Interface Device 17
9 Interface Device and Proportional Temperature
Controller 18
10 Schematic of Test Setup for the Evaluation of
Absorbents 23
11 TEA/Chromosorb AW DMCS 26
12 Celite (AW) 28
13 Pure TEA/Chromosorb G DMCS 30
14 TEA'HCl/Chromosorb G AW DMCS 37
15 Celite, Specially Treated (Acid-Washed, Neutralized
with Distilled H20, Dried and Fired at 1200°C
for 48 hr) 40
16 Blank, TEA-HCl/Glass Wool (By Heat) 43
17 TEA.HCl/Celite 45
18 Performance of N02 Collector TEA.HCl/Celite
(S02-Treated, 1000 ppm S02/Air, 1 £/min for
20 min) 47
19 Total NOx Collector (Aluminum Cartridge),
Co203/Chromosorb G 51
20 NOx Collector, Co203/Chromosorb G 80/100 Mesh 52
21 NOx Collector (Aluminum Cartridge, Co203/Chromo-
sorb G 45/60 Mesh), Carrier Air Contains 92%
Humidity 54
2.2 Estimation of Quantity of Sample Injected with
Calibration Sample Moisture 58
23 Effect of H2S (20.2 ppm) on TEA-HCl/Celite,
Carrier Air Contains 96% Humidity 60
24 Effect of S02 (1000 ppm/air) on TEA-HCl/Celite,
Carrier Air Contains 90% Humidity 62
25 Recovery from an N02 Collector, TEA'HCl/Celite
(S02-Treated), Which was Injected with
20 cc NOx (see Fi§- 1& f°r area count),
Stored in a Lab Drawer for 48 hr 64
26 First-Generation Collector Cartridge 66
27 First-Generation Dual Path Collector/Pump
Assembly 68
iii
-------
Figure Ho. Title
28 Second-Generation Aluminum Cartridge (00263) 70
29 Second-Generation Pyrex N02 Collector Cartridge
Shown with Aluminum Cartridge and Collector
Unit Subassembly 71
30 Collector Unit Assembly with Glass H02 Collector
and Aluminum Collector Cartridges 72
31 Details of the Collector Unit Tube 73
32 Luer-Coupler (for Interfacing the Second-Generation
Co203 Collector Cartridge to the Analyzer) 75
A-l General Material Evaluation, Showing the Inertness
of (1) Pyrex Tube, (2) Silane-Treated Glass Wool,
(3) Chromosorb T, and (4) Chromosorb W AW DMCS A-3
A-2 Chromosorb W AW DMCS A-4
A-3 TEA/Chromosorb W AW DMCS A-5
A-4 TEA/Regis GasPak FS A-6
A-5 TEA/Chromosorb T A-7
A-6 TEA/NEAT (Liquid) A-8
A-7 TEA Borate A-9
A-8 THEED A-10
A-9 Quadrol/Chromosorb W AW DMCS A-ll
A-10 NiS04 A-12
A-ll CoS04 A-13
A-12 CoO A-14
A-13 CuO A-15
A-14 Aluminum Cartridge (After Heating Cycle to
200°C and Cooled Down to Room Temperature) A-16
A-15 Teflon-Coated Aluminum Surface A-17
B-l Folded View of Collector System Housing B-2
B-2 Assembled Collector Unit B-2
Table No.
1 Parts List for the Collection System 19
2 Candidate Absorbents for NOX Collection 20
3 Materials Evaluated 35
4 Summary of Experimental Results 79
5 Comparison of Ambient N02~Concentration by
Technicon and Bendix Monitors (Date: 12-8-73) 80
iv
-------
SECTION 1
INTRODUCTION
Most current methods for collection and assay of ambient pollutant
gases involve liquid media in bubblers or impingers. These techniques
suffer from generally poor collection efficiency, limited sampling capa-
city, tedious or cumbersome operation, and special handling requirements.
They are also troublesome for storage and transportation of reagents,
equipment, and collected samples. These drawbacks have provided the im-
petus for developing better, simpler collection devices. Ambient gas
collectors based on solid absorbents offer the most attraction for de-
velopment, since they would possess the essential physical characteristics
to provide simplicity, ruggedness, large capacity for air sampling, sta-
bility, and wide selection of analytical methods.
Under EPA Contract No. 68-02-0657, Bendix Research Laboratories
developed such a solid-state collection device and transfer interface
device for nitrogen oxides. These devices have been evaluated under
laboratory conditions and are believed to be highly promising. However,
more rigorous field testing is necessary to validate their merits in
actual conditions. Presently only very limited field evaluation has
been conducted. The scope and main objectives of this development pro-
gram were as follows:
(1) Design, fabricate, and evaluate small, simple inexpensive
collection devices with solid chemical absorbents that
-------
quantitatively collect atmospheric gases under realistic
field conditions. The device should require a minimum of
non-technical training for proper use and should survive
mail shipment without damage or loss of sample. Devices for
collection and transportation of nitrogen dioxide (N0_) and
nitric oxide (NO) have first priority, with development of a
prototype device for these gases the primary goal. The phy-
sical design of the collection device should be such that it
can adapt readily to collection of lower priority gases such
as sulfur dioxide (SO ), carbon monoxide (CO), formaldehyde
(HCHO), and more complex organic compounds.
(2) Develop an interface device that quantitatively transfers the
gas sample, or its derivative, from the collector to an appro-
priate analytical instrument. The interface device should
operate as simply as possible consistent with quantitative
sample transfer to specific or multi-pollutant analyzers.
The analyzers could range from inexpensive manual intruments
to automated, multicomponent analyzers.
The extensive program effort resulted in the following:
• A long list of candidate solid chemical absorbents for NO,
NO , or both NO and NO (NO ) collection and their evaluation.
^ ^ X
• Selection of triethanolamine hydrochloride (TEA'HCl), m.p. 177-
179, for selective NO collection and cobalt oxide for NO and
NO collection.
-------
• A collector unit which is
• Capable of selective and quantitative collection of NCL and
NO
• Easy to assemble and disassemble
• Lightweight (<350g) and rugged
• Suitable for mailing
• Adaptable through an interface to a chemiluminescence NO
X
analyzer, a gas chromatograph, or a wet chemical analyzer.
o Easily installed in an air sampler, requiring no special
tools or skills.
• Reusable.
• An air sampling system capable of continuous, unattended opera-
tion using 12V DC or 110V AC power.
• A weatherproof shelter to house collector unit, sampling system,
timer, and batteries.
• An interface device for recovering pollutant samples from col-
lection cartridges for subsequent analysis.
The details of the different phases of the program are discussed
in the following sections. Section 2 contains the instrument description,
while Section 3 presents experimental techniques and data analysis. Sum-
mary and recommendations are given in Section A.
-------
Figure 1 - Collector System
-------
SECTION 2
COLLECTOR SYSTEM DESIGN
2.1 COLLECTOR SYSTEM
The collector system consists of the following major components:
» Collector units
o Vacuum sample pump
• Power supply
o Timer
o System housing
The overall configuration of the system is shown in Figures 1 and 2. The
collector pump and collector unit are housed on the top shelf; the second
shelf contains the timer and power supply and thus is the control shelf;
the bottom shelf holds the battery. A locked door is provided in the
front for access to the control switches and battery. The housing is made
from 16-gage aluminum with wooden shelves; it has a slanted roof to drain
rain or snow. Side holes provide ventilation and ports for a sampling
line. The entire housing is hinged so that the housing will fold toget-
her into a flat package for easy transporting when the collector system
components are removed. Instructions for assembly of the housing are
given in Appendix B.
2.2 COLLECTOR UNIT
The collector unit consists of a cylindrical aluminum outer case
which houses two Pyrex glass cartridges. These cartridges, individually
protected by thin foam plastic covers, are joined tight with a Beckman
-------
Teflon connector (/M06). The assembled cartridges are secured into the
aluminum case by inserting the inlet and outlet cartridge ends into Beck-
man bulkhead Teflon fittings (/M27) attached to each threaded end-plate
of the case. The details of construction are illustrated in Figures 3
and 4.
For actual assembly, one of the collector cartridges is first in-
serted into the Teflon bulkhead fitting (#427) on the left end-plate
(Figure 3) and is secured by finger-tightening the nut on the fitting.
The second cartridge is joined to the first via the Beckman Teflon con-
nector (#406) with finger-tight sealing. Next, the bulkhead fitting
(#427) is removed from the right end-plate. The free end of the second
cartridge is inserted into the removed fitting and secured by tightening
the nut. This subassembly (cartridges and fittings) with foam plastic
covers is placed into the case, after which the right -end-plate and its
washer are tightened to the case. Finally, the bulkhead nut and the
screw nut are tightened onto the end of the fitting protruding from the
right end of the case. This completes the assembly of the collector unit.
As assembled, the unit weighs less than 350 g.
A Millipore Swinex 13 //I filter assembly made of plastic (catalog
no. SXG S0130S) is attached to each end of the collector unit. This
filter employs Gelman type-A glass fiber filters cut to size. The fil-
ter assembly has Luer-lock terminals and is attached to the bulkhead
Teflon connector with a nylon female Luer-lock adapter. This adapter
is made by cutting in half the Hamilton female/female Luer-lock connector
-------
SIDE PANELS
OPEN AS SHOWN
1/4-20 x 1/2
9 CM HASP
FRONT DOOR
OPEN AS SHOWN
t
I
3
r> f
30 <-
u
r-
1
i
DOOR HEIGHT
i
3
i*
«
I
' \
t
1
i :
i t
a
r
' 1
i
3
1
i
' 1
f
U
O>
«
E
«
c
«
i_
13 MM
SHELF
3 PLACES
t_.
7"
38 MM
ANGLE ALUM.
SHELF FRAMES
fr
3
ft-
D
»
1
50
•<- ALUM.
4COF
___^--
i
)
T!
1 u
) »
oi
M
I-
k
>
3
3
r
PIANO
HINGE
VENT HOLES
5 MM DIA.
FOLDED
Figure 2 - Collector System Housing
-------
00
Figure 3 - Collector Unit Parts
-------
SIDE PANELS
OPEN AS SHOWN
1/4-20 x 1/2
9 CM HASP
I
DOOR HEIGHT
s §
00 U
ID
r-
en £
in u
8
(«
13 MM
SHELF
3 PLACES
38 MM
ANGLE ALUM.
SHELF FRAMES
62 CM
FRONTDOOR
OPEN AS SHOWN
50 MM
ALUM. ANGLE
4 CORNERS
T
1
1/4-20
PIVOT
si
PIANO
HINGE
VENT HOLES
5 MM DIA.
FOLDED
Figure 2 - Collector System Housing
-------
OC
Figure 3 - Collector Unit Parts
-------
END PLATE
SCREW NUT
1
BULKHEAD NUT
STRAIGHT TEFLON
BULKHEAD FITTING
(BECKMAN #4271
TEA/CELITE IN
GLASS TUBE
BECKMAN TEFLON
CONNECTOR 406
(a) Collector Unit
Co2O3/CHROMOSORBG 45/60
AW DMCS
PYREX TUBE
FEMALE LUER
LOCK
0.396
1.905
- 0.432
(b) Optional Filter Adapter
Figure A - Specifics of Collector Unit
VO
-------
(Hamilton catalog no. 86505); it is inserted into the Teflon connector
and tightened with the screw nut. (The adapter can also be made by taper-
ing the inside of 1.9-cm long Teflon tubing from 0.396 cm ID to 0.432 cm
as shown in Figure 4.) The other end of the collector unit is also at-
tached to a Millipore filter through a similar adapter and a Hamilton
nylon male/male Luer-lock connector (Hamilton catalog no. 86506) which
inserts into the female terminal of the filter. This filter is connected
to the limiting critical orifice installed to the vacuum pump. This
filter assembly, however, needs to be evaluated for its inertness to NO
X
interconversion.
2.3 COLLECTOR CARTRIDGE
2.3.1 Cartridge Geometry
The collector cartridges for the two absorbents are iden-
tically constructed from Pyrex glass. Their dimensions are given in
Figure 5. The inlets and outlets are 0.635 cm OD Pyrex tubing which
readily inserts into the finger-tight Beckman Teflon connectors. In ana-
lysis, Beckman Teflon connectors are also used to connect the cartridge
to the analytical instrument.
0.635
1.575
d »>
/f- -^
^ •**
>x ^
^ ^
d 7 B7 to.
1.575
Figure 5 - Pyrex Glass Connector Cartridge
10
-------
2.3.2 Cartridge Charging
In charging the cartridge, the packing is first prepared
as directed. One end of the tube is packed with a silanized glass wool
plug (Applied Science #14502). The prepared packing is then transferred
into the cartridge with a clean spatula. The tube is constantly vibrated
by hand while it is being packed. When the tube is filled, the open end
is plugged with silanized glass wool. The difference in weights of car-
tridge before and after charging is the weight of packing material. It
is important that the cartridge be carefully packed to ensure consistency
in performance. A small difference in packing weight and pressure drop
of charged cartridge is an indication of consistent packing.
2.4 PREPARATION OF ABSORBENT PACKINGS
2.4.1 Treatment of Celite
The Celite 15/30 mesh was obtained from Johns-Manville as
a free sample. It is soaked in concentrated HC1 for over 72 hr, drained,
and washed to litmus paper neutral with distilled water. The washed Celite
is then dried at 50-80°C, and brought to 250°C for 2 hr prior to firing
at 1200°C for 12-48 hr. The Celite treated in this fashion is inert to
NO. It absorbs N0« at room temperature, but releases it at 170-200°C.
The cooled Celite is stored in a clean, capped jar placed in a dessicator.
2.4.2 Coating of TEA'HCl
About 20 g TEA-HC1 (Eastman 1916) is dissolved in 80 cc
distilled water. Then 40 g of treated Celite is added to the solution,
which is stirred and allowed to stand for 30 min. The solution is evapor-
ated to dryness at 80°C in a hood; during drying, the solution is occa-
sionally shaken. The dried TEA-HCl/Celite is then stored in an oven set
11
-------
at 110-120°C. Packing prepared in this manner is ready for charging into
the collector when cool. Each charged collector contains approximately
6g of packing. Before use, the charged collector needs to be conditioned
at 160-170°C with either N0 or air flow at 100-120 cc/min until no NO
L X
evolution is observed from the collector. The collector can then be
cooled in a clean atmosphere, capped, and stored for future use. CAUTION:
the conditioning temperature should not at any time exceed 170°C; beyond
this temperature, decomposition may occur, thereby impairing the perfor-
mance of the collector.
2.4.3 Cobalt Oxide
The cobaltic oxide (J. T. Baker 1688) is a fine grayish-
black powder (^200 mesh). It is first packed into a large Pyrex tube,
and heated at 500-550°C for 48 hr under nitrogen flow at 20-100 cc/min.
When cooled, the powder is physically mixed well with Chromosorb G AW
DMCS 45/60 mesh (Applied Science Lab) in approximately 1:2 by volume.
This mixture is then used to charge the NO absorbent cartridge. The
A
charging procedure is the same as that for the NO absorbent cartridge.
Prior to collection, the prepared cartridge must be conditioned at 420-
450°C under air flow at 20-120 cc/min either overnight or until it no
longer releases NO (as indicated by an NO analyzer). After cooling
A A
in a clean environment, the collector cartridge is ready for use, or
should be capped for storage.
2.5 VACUUM PUMP (Cast Model 1531)
This pump (obtained from Cast Manufacturing Corp., Benton Harbor,
MI 49022) is a 12V DC-operated Rotary Vane Vacuum/Compressor pump. It
12
-------
Figure 6 - Gas Pump and Collector Unit
-------
has a free air flow of 42 £/min. The pump draws 11.3 A current and
can be powered by a 12 V auto battery or by a 12 amp-rate battery charger
(110V AC). Figure 6 shows the pump and collector unit assembly.
The collector system flow rate is determined by the pressure drop
exerted by the collector unit and filters. In order to eliminate the
flow rate variations caused by uncertainty in the uniformity of each
packing, a critical limiting orifice is placed between the pump and the
collector outlet filter. The orifice provided with the unit gives a
flow rate of 4.86 &/min. Figure 7 shows the quick-connection of the col-
lector, pump, and orifice.
PIPE TOSWAGELOK
ADAPTOR
BECKMAN TEFLON
CONNECTOR (406)
\
SWAGELOK THREAD
TO PUMP
SWAGELOK
FITTING
BENDIX CRITICAL
ORIFICE (BENDIXESD)
CATALOG NO. 7011
FEMALE
LUER-LOCK
TO ACCEPT
CASSETTE/FILTER
TERMINAL
PIPE THREAD
SPECIALLY DRILLED ORIFICE P.87.8(
Figure 7 - Connection for Collector Unit/Filter to Pump/Orifice
14
-------
2.6 POWER SUPPLY
Either 110V AC or a 12V DC battery can be used as the power source.
With 110V AC the collector unit is powered by the output of a 12V, 12A,
battery charger (Model 126, R. N. Industries, Dearborn, MI 48126), which
is in turn supplied by the 110V AC power line. A conventional 12V car bat-
tery provides a maximum of 12 hr continuous operation; a heavy duty 280 A
truck battery can extend the operation time over 24 hr0
2.7 TIMER
The timer supplied with the collection system is a 110V AC-operated
Intermatic Model T103 (International Register Company, 2624 W. Washington
Blvd., Chicago 12, Illinois). This timer enables the collection unit to
sample the air for a preset collection time interval of 1-24 hr when
operated in the AC-mode.
2.8 OPERATION/MAINTENANCE
In actual operation, the operator carries the collector system to
the chosen site and connects the collection unit to the pump orifice by
finger-tightening the Beckman Teflon connector. (No special skills or
tools are necessary.) Then the operator merely turns on the switch (either
the battery or the timer). At the end of collection, the operator comes
back to the sampler, turns off the switch, removes the collector unit,
and caps the ends. (Either the Hamilton male Luer plug or a piece of
0.63 OD solid Teflon rod can serve as a cap for the collector unit. The
plug goes readily to the filter adapter, while the Teflon rod can cap
the unit terminals by replacing the filter adapter.) He can then carry
the unit to the laboratory or send it out by mail for analysis of the
collected oxides of nitrogen at a central laboratory.
15
-------
Practically no maintenance of the collector system is required,
other than changing the filter paper periodically. The critical orifice
may be air-jet cleaned to remove any trapped particulates if it is ac-
cidentally contaminated. The pump requires no lubrication.
2.9 INTERFACE DEVICE
The interface device accepts one collection cartridge and is capable
of providing sufficient heat to bring the collector to the temperature
necessary to release absorbed nitrogen oxides for analysis. The device
itself is a Kaowool-insulated aluminum block with holes drilled for the
collector cartridge, a 250 W Hotwatt cartridge heater, a 0-500°C platinum
temperature sensor (RFL 24928, 100 H), and a thermocouple. The specif-
ications of the interface are given in Figure 8. The sensor and heater
are operated by an RFL Model 71 Proportional Temperature Controller (RFL
Industries, Inc., Boonton, N. J. 07005), which is powered by a 110V AC
line. Figure 9 is a photograph of the interface device and the tempera-
ture controller. A 16-cm long rod is provided on the device for conven-
ience in setting it up on an iron stand.
In operation, the interface device is set at the desired temperature
(160-170°C for the N02 collector; 402°C for the NO collector); then the
proper collector cartridge is inserted for analysis. The collector car-
tridge, of course, should be connected to an appropriate analyzer prior to
its insertion into the interface device.
2.10 PARTS LIST
A parts list for the collection system is given in Table 1.
16
-------
19.304 MM REAM THRU
SILVER
SOLDER
vv
-16 CM-
3.17 MM
50 MM DEEP
(FOR THERMOCOUPLE)
.9.525 MM
DRILL THRU
(FOR CARTRIDGE HEATER)
ALUMINUM
4.76 MM
50 MM DEEP
(FOR PT-SENSOR)
KAOWOOL INSULATION
i
BRASS
TUBE
/
| flit»-
T
I
1
4WV
•T
I
I
i.
L
L
-9CM-
u
i-.
IV."
Figure 8 - Details of Interface Device
-------
00
Figure 9 - Interface Device and Proportional Temperature Controller
-------
Table 1 - Parts List for the Collection System
Item Description, Model No.
1 Collection System Housing Prototype
2 Collector Unit (Aluminum case)
3 Collector Cartridges (Pyrex tubing)
4 Plastic Filter, Millipore Cat. No.
SXG S0130S, Swinex 13, No. 1 size
5 Beckman Bulkhead Teflon-Connectors
Cat. No. 427
6 Beckman Teflon Connectors Cat.
No. 406
7 Hamilton female/female Luer-lock
connector No. 86505
8 Hamilton male/male Luer-lock
connector No. 86511
9 Teflon adaptor for filter with
Luer-lock terminals
10 Bendix Critical Limiting Orifice
(for flow) Cat. No. 7011
11 Cast Vacuum Pump, Model 1531,
12V DC Motor
12 Timer, Intermatic, Model T 103
13 R. H. Battery Charger
Model 126
14 Proportional Temperature
Controller and Platinum Sensor,
RFL Model 71
15 Cartridge Heater, Hotwatt, 250W
16 Interface Device
Manufacturer
Bendix Research Laboratories
Southfield, MI 48076
Bendix Research Laboratories
Bendix Research Laboratories
Millipore Filter Corp.
Bedford, MA
Beckman, Inc.
Southfield, MI 43075
Beckman, Inc.
Hamilton, Box 7500
Reno, Nevada
Hamilton
Bendix Research Laboratories
Bendix Environmental Science
Division, Baltimore, MD 21204
Cast Manufacturing Corp.
Benton Harbor, MI 49022
International Register Co.
Chicago 12, Illinois
R. N. Industries
Dearborn, MI
RFL Industries, Inc.
Boonton, NJ 07005
Bendix Process Instruments Div.
Ronceverte, W. Va.
Bendix Research Laboratories
19
-------
N>
O
Table 2 - Candidate Absorbents for NO Collection
GC phases/sorbent
Inorganic Bases
Alcohols, amines, amlno-alcohols
Inorganic Salts
DuPonc FCX-330
Halocarbon 11-14
Halocarbon 13-21
Fluorolube MD-10
Aroclor 1232
Chlorinated Biphenyl-32
Silicone Oil GE, SF96
Armeen SD
Hi-EFF-lOB
Molecular Sieve
5A, 13A, 13X
Activated Carbon
Alumina
NaO-CaO
NaOH
LiOH-H20
Triethanolamine
Triethanolamine hydrochloride
2-(2-butoxyethoxy)ethanol
1,1',l"-nitrilo-tri-2-propanol
Triethyl N-tricarboxylate
Triethanolamine Borate
N ,N'-(2-hydroxyethyl)-Piperazine
N,N-Bis(2-hydroxyethyl)-o-toluene-sulfonamide
2,2-Bis(hydroxymethyl)-2,2',2"-nitrilotriethanol(Bis-tris)
2,2',2", 2"'-(ethylenedinitrilotetraethanol) (THEED)
2,2',2", 2"'-(ethylenedinitrilotetra-2-propanol) (Quadrol)
Polyoxyethylene(8) Ethylenediamine
Tr is-(hydroxymethyl)aminomethane
Tr is-(hydroxye thy Daminome thane
1,1,1-Tris-(hydroxymethy1)ethane
6-amino-l-hexanol
1.5-pentanediol
NaClO. /Alumina
(Cr03+H3P04)/Quartz
(Na2Cr20? + HjSO^) /Glass Fiber
CuO
CoO
C°2°3
Fe2°3
FeSO
PbO
Pb02
CuSO,
-------
SECTION 3
DEVELOPMENT OF COLLECTION SYSTEM
3.1 INTRODUCTION
The total development program included five major efforts: (1)
identification and characterization of solid chemical absorbents and
supports, (2) development of collector cartridge and collector unit, (3)
fabrication of interface device, (4) survey and selection of air sampling
pump, and (5) assembly of collection system. Each of these efforts is
discussed in the following subsections.
3.2 IDENTIFICATION AND CHARACTERIZATION OF SOLID CHEMICAL ABSORBENTS
3.2.1 Technical Approach
The effort of this phase centered around the rapid screening
of candidate materials to be used as absorbents for the oxides of nitrogen.
Those absorbents could be either of two forms: liquids that could be
conveniently coated onto an inert solid support, or solids that could
be either used by themselves or physically mixed with some suitable inert
support. Candidates were chosen from literature references or from ex-
amination of chemical structures. They included: (a) inorganic bases,
(b) gas chromatography stationary phases or solid sorbents, (c) organic
bases, and (d) inorganic salts. Table 2 summarizes the candidates chosen
for evaluation.
3.2.2 Evaluation Parameters
Candidate absorbents were screened on the basis of
» Ability to completely absorb either NO, N0_, or both (NO )
£ X
21
-------
• Inertness to NO , so that the absorbent surface does not
x*
catalyze NO+NO interconversion.
If an absorbent met these requirements, it was evaluated for its ability
to quantitatively release collected NO upon heating.
X
3.2.3 Test Setup
The absorbents were evaluated using the simple test fixture
shown in Figure 10. This fixture has a one-liter mixing bulb for sample
preparation which accepts the gas flows from the NOX tank and from the
diluent nitrogen tank. Appropriate flow controllers and needle valves
are provided. The pressure of the bulb can be read from the pressure
gage connected to it. The outlet of the bulb goes to a 50 cc buffering
volume through a valve. This volume is open to a vacuum vent line or
to the sample inlet of the absorbent to be evaluated.
The sample inlet branches, going to the NOX analyzer
(Bendix Chemiluminescence NOV Analyzer) directly, or through the collector
A
first and then to the analyzer and pump. A Pyrex cold-trap (ice water)
is placed between the analyzer and the collector.
A septum port is installed in the line between the mixing
bulb and the buffering volume for withdrawing prepared sample. Another
septum port placed at the ambient inlet line is for sample injection.
Of course, when the valves before and after the buffering volume are
both open, the prepared sample is continuously drawn, either directly or
via the collector path, into the NO analyzer. The nitrogen line and the
A
vacuum vent allow the gas preparation system to be purged. All the gas
22
-------
N)
OJ
AMBIENT AIR
V
SCRUBBER
INJECTION
SEPTUM
FOR SAMPLE
WITHDRAWAL
SEPTUM
NEEDLE
VALVE
50 CC BUFFER
VOLUME
VACUUM
VENT LINE
DIRECT PATH (BYPASS)
CONNECTION FOR
COLLECTOR TO BE
EVALUATED
FLOW
CONTROL
ICE
COLD
TRAP
Figure 10 - Schematic of Test Setup for the Evaluation of Absorbents
TO NOX-
ANALYZER
AND PUMP
-------
lines are Teflon, and connectors are either Beckman Teflon-type or stain-
less steel. NO is supplied from a tank of calibrated mixture prepared
A
by Cryogenic Sales, Inc. The sample prepared by this system has a con-
centration range of 1 ppm in N0_ and 0.1 ppm in NO.
For evaluation, absorbent material was packed into a 2.5-
cm OD Pyrex tube with 0.64-cm OD inlet and outlet. Silanized glass wool
was used for end plugs. The tube had a ground-glass joint to allow easy
packing.
3.2.4 Evaluation Method
The Bendix Chemiluminescence NOV Analyzer gives readouts
A.
of [NO ] and [NO] in ppm. When a steady concentration of the oxide is
X
present in the sample gas, the analyzer readout is a straight line (con-
tinuous mode). If the sample concentration has a short lifetime or is
pulsed, a peak readout appears. The latter occurs when a sample is in-
jected into the ambient air stream.
In evaluating a candidate as a potential absorbent, a
steady concentration was first established in the gas system through
the "direct path". The readouts were recorded as [NOX]D and [NOjp.
Then the sample stream was routed to the "collector path" with a can-
didate absorbent in an inert cartridge. The readouts were recorded as
[NO..]., and [NO] . The difference between [NOV] and [NO] represents the
ALL A
concentration of [NO-]. That is,
[N02]D - [N°X]D - [N°]D
[N°2]C = [N°X]C ' [N°]C
-------
When [NO ] = 0, the candidate material totally absorbs or removes the
A L»
nitrogen oxides. If [NO ] = [NO] , the material absorbs or removes
AC C
N0_. [NO] = [NO] means that the candidate absorbent causes no change
in the concentration of NO, thus implying the absence of N0? -»• NO conver-
sion. More concisely,
[NOV]., = 0 Total removal of NOV
A C X
[NOVK = [NO] Total removal of N0_
A C C ^
[NO] = [NO] No conversion of NO, to NO.
When these are not the case, the absorbent either removes NO, incompletely
and/or catalyzes its conversion to NO. The percentages of NO, removal and
conversion can be calculated as follows:
[N°2]D = [N°X]D - [N°]D
[N°2]C = [N°X]C - [N°]C
' [N°2]C
A [N00] .
i -, 2 removal ......
- removal = , , x 100
% N02 absorption = (% N02 removal) - (% NO conversion)
The evaluation of triethanolamine (TEA) coated onto Chromosorb W DMCS,
as shown in Figure 11, serves as a sample calculation.
25
-------
INCOMPLETE I\I02 - REMOVAL
CONVERTS N02-" NO
(NO] c = 0.090
Since,
Figure 11 - TEA/Chromosorb AW DMCS
[N0]n = 0.025 ppm, [N0y]_ = 0.50 ppm,
JJ A U
[N0]c =0.09 ppm,
=0.12
ppm
26
-------
[NOJ_ = [NOV]_ - [NOl = 0.50 - 0.025 = 0.475 ppm
£• U A JJ U
= [NOX]C - [NOJC = 0.12 - 0.09 = 0.03 ppm
Therefore
A N02 = [N02]D - [N02]c = 0.475 - 0.03 = 0.445 ppm
A NO = [NO] - [NO] = 0.090 - 0.025 = 0.065 ppm
C D
Thus
% [N02+N0] = (A NO/[N021D) 100 = (0.065/0.475) 100 = 11.6%
% [N00] , . = 82.1%
2 'absorption
For the injection mode of operation, the same basic calculations are used,
except that the areas of the generated peaks are used. Each area is ob-
tained by triangulation of a given peak.
In the quantitative recovery study, the elution of the oxides
of nitrogen from the collector is also in the form of a peak. For the per-
cent recovery estimate, the peak area of the eluant from the collector is
counted as A . The average area of the peaks obtained by injecting
the same sample through the direct path is calculated as A.. . The
direct
ratio A /Adi t Denotes c^e effectiveness in recovery. When this
ratio is equal to 1, a 100% recovery is achieved. The analytical accuracy
of the approach was somewhat limited by the triangulation method; however,
27
-------
it should suffice for the present application. An electronic integrator
would definitely increase the accuracy of analysis.
3.3 EVALUATION OF CANDIDATES
3.3.1 Triethanolamine and Supports
The recent report by Levaggi et al. on the superior per-
formance of triethanolamine (TEA) on Celite-22 fire brick prompted us
to attempt to reproduce their work. Unfortunately, we observed persistent
N09 to NO conversion. Since Celite washed with concentrated nitric acid
exhibited extensive conversion of N02 to NO (Figure 12), we decided to
Figure 12 - Celite (AW)
28
-------
study other support materials: Corning special textured glass beads;
Chromosorbs G and W, both of which are AW-DMCS (acid washed and silane
treated); Chromosorb T (Teflon); and Regis Gas Pak FS (Teflon-coated
diatomaceous earth). Absorbent packing made by coating TEA on these
supports activated the conversion of N02 to NO, despite the fact that
the supports themselves were found to be inert. These puzzling results
are illustrated in Figures A-l through A-5 of Appendix A.
We speculated that evaporating the methanol solution of TEA
during the coating of the support might have introduced impurities or
damaged the inert surface of the support. Finally, after repeated failure
to obtain an inert packing on any of these supports (including an attempt
to coat TEA onto the support without solvent), it occurred to us that TEA
itself might be impure. This postulation was later confirmed by the ob-
servation that TEA itself, without a support, gave a 10% conversion of
N02 to NO (Figure A-6).
A pure sample of TEA (clear liquid) was then obtained and
stored in the dark. A packing was made by dissolving 5 g of this sample
in methanol and evaporating the solution onto 20 g Chromosorb G AW DMCS
80/100 mesh. After conditioning at 80°C for 4 hr with 40 cc/min helium
flow, this packing was then evaluated in a polypropylene tube. As illus-
trated in Figure 13(a) and (b), this packing completely and selectively
absorbed N02 and was inert to NO.
In this experiment, the original sample concentration was:
[NOX1D = 1.2 ppm
29
-------
u>
o
INDIRECT x '
A - 408
INOI0- [N0,lc- INOIC
CONVERSION ABSENT
10O* NO2 ABSORPTION
'"""COLLECTOR" '
A-38.5
Figure 13(a) - Pure TEA/Chromosorb G DMCS - Injection Mode
-------
UJ
a.
IN°! DIRECT
0.145
[NOX] DIRECT
1.2 PPM
[IMO]C=[NOX]C=[NO]D
.. 100% N02- REMOVAL
, 5MIN.
I « >•
NO,
•NO CONVERSION
ABSENT
[NO]COLLECTOR
0.14 PPM
Figure 13(b) - Pure TEA/Chromosorb G DMCS - Continuous mode
-------
[NO]D = 0.14 - 0.145
After passing the collector,
[NOY]_ = 0.14 - 0.145
A L*
[N0]c = 0.14 - 0.145
That is,
[NOX]C = [N0]c = [NO]D, or
[N0?] = 0 after passing the collector.
These values indicate 100% N0? removal and no NO-^NO conversion.
Similarly, for the injection mode,
• 109.8;
[NOY]
D
A[NO]D
l[NOY]
= 39.6;
c
That is,
= [N0]c = [NO]D
which represents 100% NO- removal and no NO-^NO conversion.
32
-------
3.3.2 Thermal Instability of TEA Packings
The recovery of the absorbed N02 from the TEA/Chromosorb G
collector by elution into the NO analyzer was attempted using thermal
A
methods. Evaporation of TEA occurred above 80°C, and the color of the
packing changed to orange above 150°C, thus indicating possible decom-
position of the packing. An NO elution peak was obtained, however, at
A
this temperature. This heat-treated packing subsequently converted NO-
to NO and had lost its capability for NO- removal. A new collector was
made by packing the TEA/Chromosorb G into a stainless tubing which was
then subjected to gas chromatographic evaluation for bleeding. Persistent
bleeding occurred at and above 75-80°C, despite extensive conditioning
at this temperature.
These results clearly indicated that materials having higher
boiling points or melting points than TEA should be sought to improve
thermal stability. Solid candidates were particularly attractive. Most
of the candidates shown in the amines, alcohols, and amino-alcohol column
of Table 2 were chosen on the basis that their chemical structure resembled
that of triethanolamine and their melting points were higher. Triethano-
lamine hydrochloride (m.p. 177-179°C), and Bis-Tris (m.p. 103-104°C) are
typical examples of this kind.
3.3.3 Amino-Alcohol Screening
Quite extensive but rapid screening of the available amino-
alcohols resulted in the identification of triethanolamine hydrochloride
(TEA-HC1) as a potentially reusable NO- collector material. All the
33
-------
other amino-alcohols either converted NO- to NO or failed to absorb N02
or NO. Figure A-7 to A-9 in Appendix A are recordings from tests of some
of the amino-alcohols. Because of the time constraint, each material
was rapidly screened for its affinity for N02 and/or NO, and its surface
activity toward NO, to NO conversion. Although not every recording trace
was presentable, the necessary data were recorded for calculation. The
results of the evaluation for all materials are summarized in Table 3.
3.3.4 Triethanolamine Hydrochloride (TEA'HCl)
The TEA'HCl was a fine white powder, fluffy in texture.
It was physically mixed with Chromosorb G AW-DMCS 80/100 mesh in a 1:1
ratio by volume, and then packed into a Pyrex tube for evaluation. The
performance of TEA'HCl /Chromosorb G AW-DMCS, illustrated in Figure 14 (a)
and (b), showed 100% N02 removal and the absence of N02 to NO conversion
[NOX]C = [N0]c = [HO]D = 0.14 - 0.15;
[N°]
The preparation of TEA'HCl/Chromosorb G packing was somewhat
difficult due to the fluffy nature of TEA'HCl. Moreover, although TEA-
HC1 is very soluble in water, Chromosorb G, W, T, etc. are all hydrophobic.
Therefore, some other support should be sought. TEA'HCl can be transfor-
r n
med into larger particles. Cold-pressing at 21 x 10 kg/m for 20 min
produces a cake that will break into chips. These chips, however, did
not have adequate mechanical stability, since they gradually broke down
into a powder.
34
-------
Table 3 - Materials Evaluated (1 of 2)
Material
Results and Remarks
Support
1. Celite 22
2. Corning textured glass bead
3. Chromosorb G AWDMCS
4. Chromosorb T (Teflon)
5. Chromosorb W AWDMCS
6. Regis Gas Pack FS (Teflon coated
diatomaceous earth)
7. Silane treated glass wool
Specially acid washed and fired sample absorbs
N02t is inert to NO at room temp., gives off
absorbed NC>2 above 150°-200CC, chosen for TEA-
HC1 support.
Inert to NO, N02, has low surface area and capa-
city for coating material.
Inert to N02 and- NO, hydrophobic; Chrom. G 45/60
mesh can be best mixed with the Co203 powder to
afford a decent gas flow, chosen for
support.
Inert to N02, NO
Container
1. Pyrex glass
2. Aluminum
3. Teflon coated aluminum
Inert to NO, N02, chosen for both lab test and
collector cartridge construction.
After going through heating cycle to 200°C,
converts
Inert to N02, NO
Absorbent Candidates
(Inorganic)
1. Co203/Chromosorb G
2. NiS04
3. CuO (wire)
4. Cu20/Chromosorb G
5. CoS04
6. CoO
Absorbs NOX completely up to 350°C. Releases
the absorbed quantitatively above 390°C.
Does not absorb N02, converts N02-|'NO
Slightly absorbs N02. Slightly converts N02-N0
Converts
Converts
100% NOX removal
Does not absorb N02-
u>
-------
Table 3 - Materials Evaluated (2 of 2)
u>
Absorbent Candidates
(Organic)
Material
1. 1,5-pentanediol/Celite
2. THEED/Chromosorb W
3. Quadrol/Chromosorb W
4. Polyoxyethylene(8)ethylene
diamine/Chromosorb T
5. TEA-Borate
6. 2-(2-butoxyethoxy)ethanol/
Celite
7. 1,1',l"-nitrilo-tri-2-
propanol/Celite
8. Impure TEA coated on
a. Chromosorb G
b. Chromosorb W
c. Chromosorb T
d. Regis Gas Pak FS
e. Without support
9. Pure TEA/Chromosorb G
10. Triethanolamine hydrochloride
(TEA-HC1)
11. Bis-Tris
12. N,N'-Bis-(2-hydroxyethyl)-
piperazine
13. N,N'-Bis-(2-hydroxyethyl)-p-
toluenesulfonamide
14. Phenyldiethanolamine succinate
(Hi-EFF 10B)
15. Tris-(hydroxymethyl)-
aminomethane
16. 1,1-tris-(hydroxymethyl)-ethane
Results and Remarks
61% N02 removal, 11%. NO removal
<10% HO-, removal, 50% N02->-NO conversion
Does not remove N02 to any significant extent
Significantly converts N02-»-NO (70%) .
Does not remove N02
80% N02 removal, converts 13% N02-*NO
78% N02 removal, slightly converts N02-»-NO.
100% N02 removal Converts N02-"NO(^50%)
4% N02-»-NO conversion
16%
17%
40%
10%
0% conversion, 100% N02 removal Poor in
thermal stability
0% conversion, 100% N02 removal, chosen N02
absorbent
Inert at room temp.
N02 removal up to 62%, converts N02->NO slightly
Does not absorb N02
Inert at room temp.
<10% N02 removal, 2% conversion N02~NO.
<10% N02 removal, 2% conversion N02"NO.
-------
INOIC INO,IC - IMO|Q
.. 10O* NO2 - REMOVAL
THERE IS NO
CONVERSION
-\ h
Figure 14(a) - TEA-HCl/Chromosorb G AW DMCS - Continuous Mode
00
-------
J°-COLllCIOB *
|NOIO INOIC INO.IC
10(7^ NO7 ABSORPTION
INO,ICOLL£CIOR x
S, • 72 6
t ***** H
Figure 14(b) - TEA-HCl/Chromosorb G AW DMCS - Injection Mode
-------
3.3.5 Further Work with Celite
Since Celite was the only available support that was not
hydrophobia, we decided to re-examine it as a possible support. The
previous acid treatment used concentrated HNO., and a short soaking time.
In the retest, Celite 22 firebrick (from Johns Mansville) was soaked in
concentrated HC1 for over 72 hr, washed with distilled water until neutral
to litmus paper, dried at 80°C, heated to 250°C for 2 hr, and fired at
1200°C for 48 hr. Figure 15 shows that Celite treated in this manner
was inert to NO and absorbed N0_ at room temperature up to 45%. The
absorbed N02 was released at 150-200°C.
3.3.6 Recovery by Thermal Method (TEA-HCl/Celite)
Blank
TEA'HCl was tested to determine whether it would decompose
to give off NOV. For this test, a tube was packed with TEA'HCl and silane-
A.
treated glass wool. Figure 16 indicates that the TEA'HCl did give off
some NOY at a temperature of 160-170°C. The evolution was most likely
A
the N0_ absorbed by the TEA-HC1, since another heating cycle up to 200°C
did not cause further evolution.
TEA-HCl/Celite
TEA'HCl was weighed and dissolved in distilled water. Pre-
weighed, acid-treated Celite was added and the solution was allowed to
stand an hour. The water was then evaporated at 80°C with occasional
stirring; the dried packing was then transferred into the Pyrex cartridge.
Prior to evaluation, this collector cartridge was conditioned at 100-120°C
for 4 hr with 120 cc/min air flow, then heated to 170°C to check for
39
-------
CELITE ABSORBS
CELITE DOES NOT
CONVERT NO2-NO
INDIRECT 'N°ICELITE
0.31 0-32
-4-
-4-
•4-
-4-
-4-
INOI DIRECT
0.30 • 0.31 PPM
1 1 1
Figure 15(a) - Celite, Specially Treated (Acid-Washed, Neutralized with Distilled H20,
Dried and Fired at 1200°C for 48 hr): Inert to NO, absorbs N02 at
Room Temperature
TO DIRECT
PATH
-------
0% ABSORPTION
AT 200° C
[NOX]CEL|TE 200° C
1.02-1.03
IN°X) DIRECT X 2
1.02- 1.03 PPM
[-•-5MIN.-*-]
TOCELITE
AT 200° C
+*»J
Figure 15(b) -
Celite, specially treated (Acid-washed, neutralized with
distilled H20, Dried and Fired at 1200°C for 48 hr): at
200°C, does not absorb N0?
-------
•C-
fo
CELITE CAN BE CLEANED
ABSORBED NO2 CAN BE
RELEASED BELOW 200 C
5MIN.
|NOXI X 1
Figure 15(c) - Celite, Specially Treated (Acid-Washed, Neutralized with Distilled
Dried and Fired at 1200°C for 48 hr): Releases Absorbed
NOX When Heated
-------
THETEA-HCI
WAS CLEANED
BY HEAT CYCLE
O
>
u
I I
Figure 16 - Blank, TEA-HCl/Glass Wool (by Heat)
-------
release of NO . Its performance at room temperature is given in Fig-
A.
ure 17(a). TEA-HC1 absorbed no NOY at 165-170°C as indicated in Fig-
A
ure 17(b).
3.3.7 The NO- Collector: TEA-HCl/Celite
That TEA'HCl/Celite was a true NO- collector was further
confirmed by the test results given in Figure 18. This figure shows
clearly that the packing not only removed NO- from the sample air without
disturbing NO at room temperature, but also completely released the NO-
at an elevated temperature. This experiment utilized the injection mode;
the volume used for each injection was 5 cc. The collection efficiency
was estimated:
Therefore
= 323
-------
o
o
I
tr
o
t-
u
8
o
I
Figure 17(a) - TEA-HCl/Celite at room temperature, inert to NO, absorbs
N02 completely
-------
Q.
ID
O
Figure 17(b) - TEA-HCl/Celite at 165-170°C, N02 completely passes through
46
-------
|NOIC • INO.]C - INOID
lOOfc ABSORPTION OF NO2
(H. CONVERSION
'in,.'"RECOVERY
10OXRECOVERY
NO. X I
AHfCOVERY 12"01 a "N021
Figure 18 - Performance of N02 Collector TEA-HCl/Celite (SOo-Treated, 1000 ppm SOo/Air
1 4/min for 20 min) .
-------
Thus,
A[NQ j ^ 0 and [N0]c ^ [NO^ ^ [NO]D
A total of 20 cc was introduced into the collector in 4
sequential injections. The recovery peak between 120-170°C had a cal-
culated area of 1240 counts, that is, about 4 times the area for each 5
cc injection of NCL. In other words, TEA'HCl/Celite released its absorbed
N09 completely:
A = 1240 = 4 ArMA , = 4 x 323
recovery [NO-]-
The above collector material went through many temperature cycles before
these measurements [Figure 17(b) for example] and it also underwent an
SCL treatment in which 1000 ppm S09 in air was passed through the collector
for 20 min at 1 £/min flow rate. Furthermore, the ambient air carrier
used in these experiments contained up to 95% humidity. Nonetheless, the
material retained its performance as a good N0~ collector.
3.3.8 NO Absorbent
An effort was made to identify an NO absorbent or a total
NO absorbent to complement the N00 absorbent TEA-HCl/Celite.
X 2,
Since it was known that inorganic oxides can interact with
NOY (for example, Pb00 forms a nitrate with nitrogen oxides), the follow-
A Z
ing inorganic compounds were chosen for further study: CuO, Cu90, CoSO,,
Co90_, CoO, NiSO,. The cobaltous and cobaltic oxides proved to be ex-
cellent total NOV absorbents, whereas the other salts evaluated showed
A
48
-------
negative results. Figures A-10 to A-13 show the results obtained from
some of these oxides.
The capability of cobaltic oxide to absorb NOX is clearly
depicted in Figures 19 and 20. In these experiments the sample NO was
quantitatively injected into the collector where it was completely ab-
sorbed and later released without loss at 395-420°C. The packing was
prepared by first heating Co-O. (grayish-black powder) in a Pyrex tube
at 500-550°C with air flow for 48 hr. When cool, the oxide was physically
mixed with Chromosorb G AW-DMCS 80/100 mesh (or 45/60 mesh) in 1:2 ratio
by volume, then packed into a Pyrex tube or an aluminum collector cart-
ridge. Prior to evaluation, the collector was conditioned at 420°C over-
night or until it was free from any pre-absorbed NOV. During the eval-
A
uation of the collector with an NO analyzer, the ambient air carrier gas
A.
contained 92% humidity.
As shown in Figure 20(a), the cobalt oxides completely ab-
sorbed the NO at 250°C. The area count for each injection was. established
by injecting samples into the collection system with the absorbent at
390-400°C at which temperature no absorption occurred.
A[NO]. .
injected
A1[NO] „
recovered
A2[NO] „ = 104
recovered
These values indicate 100% recovery.
49
-------
Figure 20(b) depicts the quantitative recovery of NO as
A
follows:
An = Arwn 1 = 1224 (for each inJection at 405-410°C)
U INUXJinjected
A- (recovery from 1 injection) = 1360 ^ A
A,, (recovery from 2 injections) = 2640 ^ 2A
A (recovery from 3 injections) = 4468 ^ 3A
A, (recovery from 4 injections) = 5472 ^ 4A
and,
A2/A1 = 1.94, A3/A1 = 3.28, ^/^ = 4.02 .
Results represented in Figure 21 were obtained on the col-
lector Co-O./Chromosorb G AW-DMCS 45/60 contained in an aluminum cartridge.
This figure also demonstrates that the collector absorbed and released
NOV quantitatively as shown below:
A
(a) For each 1 cc injection of NO sample, the average area
A
count was A.. = 545 (1 cc through a direct path).
(b) Injection of a 1 cc sample into a hot collector at 430°C
afforded an area count of A_ = 560
50
-------
REPEAT
IN°X) DIRECT = °-70
2.5 MIN.
x COLLECTOR =
.'. 100% ABSORPTION
'N°x' DIRECT X 1 = 0.65 PPM
r
"^DIRECT*1
0.13 PPM
Figure 19 - Total NOX Collector (Aluminum cartridge), Co.O^/Chromosorb G
-------
Ul
to
IN°)COLLECTOR AT
390 - 400° C
A = 97 (~ 0% ABSORPTION)
Figure 20(a) - NO Collector, Co203/Chromosorb G 80/100 mesh - Quantitative
Collection and Recovery of NO
-------
5 i
Figure 20(b) - NOX Collector, C0203/Chromosorb G 80/100 Mesh - Quantitative Collection
and Recovery of NOX
01
UJ
-------
A3cc= 1627.5
|N°«JDIRECTX6
A,, = 1008
. = 54.5
Figure 21(a) - NOX Collector (Aluminum Cartridge, Co203/Chromosorb G 45/60
raesh) - Area Estimate of Sample Injected, Carrier Air
Contains 92% Humidity
54
-------
Figure 21 (b) - NOX Collector (Aluminum Cartridge, Co203/Chromosorb G 45/60
mesh) - Quantitative Recovery of 2 cc Sample Injected
Ln
-------
AABSORBED ; 2^560 (B)
ARECOVERY = 102°
.. - 100% RECOVERY
ABSORPTION = 4 X 560 (See B)
= 2240
ARELEASED =2569.6
. 100% RECOVERY
Figure 21(c) - NOx Collector (Aluminum Cartridge, Co203/Chromosorb G 45/60
mesh) - Quantitative Recovery of 2 cc and 4 cc Sample Injected
-------
(c) A 2 cc sample injected into the collector at ^300°C was
totally absorbed; when the collector temperature was raised
to 430°C, a peak was obtained with an area count of
A, = A . _ 1100 = 2 A", = 2 A.
3 recovery from 2cc 1 2
= 2.0; A/A = 1.96
(d) A. = A 2569.6 •*. 4 x A.
4 recovery from Ace 2
A, = A o 1020 = 4 x A..
5 recovery from 2cc 1
It was therefore clear that an NO absorbent had been identified. More-
A.
over, this cobalt oxide could serve as an NO absorbent with a precolumn
such as TEA or TEA'HCl to first remove N02. This tandem arrangement was
therefore chosen for the design of the NO , NO collection unit.
3.3.9 Calibration of Peak Areas
A calibrated sample of 1.1 ppm NO (Matheson Gas Products,
Inc.) was used for the area/mass equivalent calculation. Various volumes
of this sample were injected to establish an average area count.
Each cubic centimeter of calibrated sample containing 1.4
ng produced an area count of 4, (Figure 22) making each area count equiv-
lent to 0.35 ng NO. Based on the calculation, each cubic centimeter of
sample prepared by the laboratory test setup contained on the order of
10 g. This sample size was injected into collectors being evaluated
to test their absorption efficiency.
57
-------
Ui
oo
LAB SAMPLE
(X 05)
0 tec
> 0 35 r»9'AREA COUNT
> FOR AN INJECTION OF
LAB SAMPLE. EACH cc
CONTAINS - 2 X ID-7 CM
CALIBRATION SAMPLE 11.1 PPM NO)
I1.4 rt
NT IN AREA
INO.I x 5
LABORATORY PREPARED
SAMPLE
INJECTION VOLUME
Ice A-634
Figure 22 - Estimation of Quantity of Sample Injected with Calibration Sample Mixture
-------
3.4 INTERFERENCE STUDIES
The effects of moisture, hydrogen sulfide (H2S), and sulfur dioxide
(S0_) on the performance of the collectors were investigated. Moisture
alone up to 92% did not affect the performance of the cobalt oxide and
TEA'HCl/Celite. Likewise the combination of H2S (20.2 ppra in N2 from a
permeation tube) and up to 96% humidity had no synergistic effect on
TEA'HCl/Celite (see Figure 23). Passing 1000 ppm S02 in air at a flow
rate of 1 «,/min for 20 min through the TEA-HCl/Celite had no effect.
The support retained its superior performance. The presence of S02 and
moisture did not change the concentration of [NO] or [NO.J in the ab-
sence of TEA'HCl/Celite. However, an effect was noted when [NO ] and
A L»
[N0]r were compared with or without the presence of both moisture and
L*
SO . A 6 to 30% reduction of NO passing through the collector was noted
in the presence of both 1000 ppm SO and moisture (see Figure 24). Thus
NO might have been converted to N0» which was in turn absorbed. However,
this effect was produced by SO™ in a concentration 100 times that of the
nitrogen oxide in the sample (1 cc NO sample with an area count about
A
_Q
90 was equivalent to about 3 x 10 g, while 1 cc of 1000 ppra SO- con-
tained about 3 x 10~ g). This high a concentration is unlikely to be
encountered in actual field conditions.
Figure 25 shows the recovery from the SO -treated TEA'HCl/Celite
collector which had been injected with 20 cc NOV (see Figure 18 for area
A
count), then stored in the lab drawer for 48 hr prior to thermal release.
The area count of the recovery peaks obtained from this collector was
more than equivalent to the 20 cc injection. (A = 1596. A0_ =
^ recovery /U cc
4 x 323 = 1292 per Figure 18), indicating a 100% release from the
collector.
59
-------
§
I*
s
< j;
tu ID
INO,ID
WITHOUT H2S
A • 70.5
INO,I0
• 1cc
H2S
A =70
|NO,]0
WITHOUT H2S
A = 7«
Ibl
INOXIDWITH
1 ccHjS
A. = 73.6
So" 7 S
5£ K"
t
Figure 23(a) - Effect of H2S (20.2 ppm) on TEA-HCl/Celite, Carrier Air Contains
96% Humidity - Direct Path (Bypassing Collector), Injected
Volume: 1 cc
-------
I N°l DIRECT PATH
WITHOUT H2S
A = 89
.'. |NOID- |NOIC-90
.'. THERE IS NO CONVERSION
.. |NO.Ic=(NOIci91.8
.. 100% NO2 - ABSORPTION
100% NO • PASSING THROUGH
INOl COLLECTOR
» I cc HjS
A =90
A/A,- 1.02
.. NO EFFECT ON
NO2- COLLECTION
INO,I COLLECTOR
WITH 1ccH2S(20.2PPM)
A = 91.8
INO.I COLLECTOR
WITHOUT HjS
A = 93.8
Figure 23(b) - Effect of H2S (20.2 ppm) on TEA-HCl/Celite, Carrier Air Contains
96% Humidity - l^S Does Not Affect the Performance of the Collector
-------
ro
WITHOUT SOZ
A«5I9
SO2 HAS NO EFFECT ON
THE CONCENTRATION
OF MO, IN THE ABSENCE
OF COLLECTOR SURFACE
WITHOUT SO2
[NO 1 X 5
"DIRECT
A = 519
WITH SO2
WITHOUT SO2
INOI DIRECT XZ
WITH 1 cc SO2 11000 PPM}
INDIRECT X2
A -90
Figure 24 (a) - Effect of S02 (1000 ppm/air) on TEA-HCl/Celite, Carrier Air Contains
90% Humidity - Direct Path
-------
^WITHOUT SO2 " 8Si8
SWITH S02 ° 80'8 (1ccl
A'WITH S02 ' 61 (2 ccl
.'. S02 LOWERS THE INOJ
78
WITH 2 cc
SO2 WITH 1 cc SO2
J I
WITHOUT SO
INOXICOLLECTOR X 2
WITHOUT SO2
Figure 24(b) - Effect of S02 (1000 ppm/air) on TEA-HCl/Celite, Carrier Air Contains
90% Humidity - [802] Lowers [NOjJ Passing Through the Collector
-------
AREA COUNT FOR 20 cc = 1292
AREA COUNT OF RECOVERY = 1596
100% RECOVERY (THE EXCESS MIGHT
BE DUE TO THE "PICK-UP" FROM THE
LAB-AIR)
(N°xl RECOVERY X
COLLECTOR INSERTED
INTO INTERFACE
HEATER AT 170° C
' I0*»
Figure 25 - Recovery from N02 Collector, TEA-HCl/Celite (S02~treated),
which was injected with 20 cc NO^ (see Figure 18 for area
count), stored in a lab drawer for 48 hr.
3.5 DEVELOPMENT OF COLLECTOR SYSTEM
3.5.1 First-Generation Collector Cartridge and
Field Collection System
The collector cartridge is required to be: (a) rugged,
(b) lightweight, (c) good in heat transfer, (d) easily packed and un-
packed, and (e) able to interface with a suitable analytical instrument
for analysis.
The first three requirements can be easily satisfied by
choosing an aluminum shell for the cartridge. The last two are met by
64
-------
having a Luer-Lock terminal on the cartridge for syringe needle adaption
(to a gas chromatograph, for example), and a stem that can be easily
opened at one end for charging and discharging the cartridge. Such a
design is shown in Figure 26. In this first-generation cartridge, the
barrel can be opened near one end. A V-clamp holds the barrel tight.
This clamp is tightened by an Allen-head screw. Both terminals are
tapered to allow Luer-Lock connection. As shown in Figure 26(b), one
of the Luer terminals is made by inserting a specially made Luer-end
screw into the barrel.
The inside surface of the cartridge was coated with Teflon
to ensure its inertness, because inertness was essential for selective
NO- collection. The surface was evaluated for its catalytic activity
in converting NO- to NO before and after the Teflon coating. The results
showed that bare aluminum catalyzed conversion, whereas the Teflon-coated
surface was inert (Figures A-14 and A-15 in Appendix A). For NO col-
lection, however, the conversion is not a problem. Therefore, two identical
aluminum collector cartridges were made, one of which was internally
Teflon-coated for the NO- collection absorbent, and the other uncoated
for the total NOV absorbent..
A
Parallel or dual-path collections could be employed, with
one collector for NO- and the other for NOV. The difference between
2. A
the two would represent the concentration of NO. This concept is reflected
in Figure 27 which illustrates the dual-path collector/pump assembly. The
collector cartridge was supported by an aluminum tube lined with soft pad-
ding. A filter was located in front of the cartridge and another behind
65
-------
Figure 26 (a) - First-Generation Collector Cartridge - Aluminum Cartridge
-------
10 MM
6.350x907 THRO
BOTH ENDS
1 s — ""
in ^ 1.17MM »-
J i
, f
6. 5 MM 30°
127 MM
X1
V.
\
fc.
^
\
«/*
\
"*
^
^
Y
1/10 MM
6.3
S
o
i-
I
1
5 MM
1
I
256 SCREW
(ALLEN HEAD)
2.80 MM DRILL
6.60 MM
P 8 7.803-3
Figure 26 (b) - First-Generation Collector Cartridge - Details of the
Cartridge
-------
ffl
oc
Figure 27 - First-Generation Dual Path Collector/Pump Assembly
-------
it. The filters were actually Beckman Teflon Connectors (#504) containing
a piece of Gelman Type-A glass filter paper cut to size. A terminal
cut from a Hamilton female/female Luer connector was inserted into the
Teflon connector to enable the filter to be attached onto the cartridge.
A limiting flow critical orifice was located between the pump and the
filter.
The drawback to this approach is that the NO concentration
is found indirectly by subtraction, and two collector units are required
each time. A more direct method is desirable.
3.5.2 Development of Collector Unit
The development of a collector unit was based on a tandem
arrangement of an NO- collector and an NO collector, with the former
L. X
serving as the precolumn. When air is being sampled, the first collector
removes N00, leaving the NO to be taken up by the NO collector. This
2. A
arrangement thus afforded a direct method of sampling NO- and NO. Only
one collection unit was necessary when these two collectors were pack-
aged in tandem.
This collector unit should house the two collector cartridges
sturdily, should be easily assembled or disassembled, and should be easily
attached to the pump for field collection. Figures 28 to 31 represent
a design possessing such features. In this design, the NO- collector
cartridge is a disposable Pyrex tube (for the non-reusable TEA), which
has an 0.63 OD inlet on one end, and a Luer groundglass taper (Kontes
Glass Co, Vineland, N. J. //K-663500-0244) on the other. The glass taper
readily fits into the aluminum cartridge (containing Co-O-) through the
inlet hole in its end.
69
-------
0.432
ALL DIMENSIONS
IN CENTIMETERS
1.27 TPF TYPICAL
- 1.905 -
0.940
(a) CARTRIDGE BODY AND END PLUGS
(0.635 OD 28/2.54 cm THREAD)
(1/428THRD)
-—^^ — - 1.80
1 1
1 1
r- '
i i.
t
- 1 80
0 94Q — ^
-- *—
.4
0.279 DIA.
/
0.432 -»•
p-i
—
0.058
i^ v n mr
(SLOT
*- 0.792 -»
1.575 »•
* -•"•'
DEEP
1 /
t \
0.396
(b) LUER - END SCREW
Figure 28 - Second-Generation Aluminum Cartridge
n
o
-------
Figure 29 - Second-Generation Pyrex NC>2 Collector Cartridge shown with
aluminum cartridge and collector unit subassembly
-------
ro
TEA TUBE
FOAM PLASTIC
SHIELD
OXIDE TUBE
ALUMINUM HOUSING
STRAIGHT TEFLON
BULKHEAD FITTING
TEA/CELITE
IN GLASS TUBE
CO203/CHROMASORB G
IN ALUMINUM TUBE
SPONGE RUBBER
PRESSURE PAD
Figure 30 - Collector Unit Assembly with Glass N02 collector and aluminum
collector cartridges.
-------
2.54 OD, 20 THREAD/2.54 CM
(1-20 THRD)
ALL DIMENSIONS
IN CENTIMETERS
\
')
—
1.27
„
26.67
1.27
f
.54
1
L.
(a) ALUMINUM TUBE
2.54 OD, 20 THREADS/2.54 CM
(1-20 THRD)
\
fr
Xs
0.305 +•
~\
1^
r /
F=-~-~
1
•4
W
1.27
r— 2.997
1
P-87-803-3
A =
0.953 OD, 24/2.54
CM THRD
1.27 DRILL
(b) END-PLATE
Figure 31 - Details of the collector unit tube
73
-------
The aluminum cartridge with Co_0_ packing is assembled
by pressing the end plugs into the tube. These end plugs are removable.
The outlet of the aluminum cartridge has a Luer-Lock terminal provided
by the Luer-end screw (Figure 28). Though not shown in the sketch, the
Luer-end screw has a Luer-Lock ring attached onto the slot cut for it.
In assembly, the Pyrex tube is first inserted into the Beckiaan bulkhead
Teflon connector on the end-plate and tightened; this then attaches to the
aluminum cartridge via the Luer ground taper. The resulting subassembly
is then carefully placed into the collection tube and the end-plate is
tightened. The assembled unit has the Luer-Lock end of the NOX collector
as one of the terminals.
For analysis, the unit is disassembled, and the cartridges
are inserted separately into the interface device for recovery of the
collected nitrogen oxides. A special Luer coupler shown in Figure 32
is designed for the aluminum cartridge. This coupler readily goes into
the hole at one end of the cartridge and serves as the terminal for
analysis.
Although this design appears to be very attractive, the glass
Luer taper unfortunately restricts the sample flow and introduces mechanical
weakness because of its thin, narrow, fragile neck. To alleviate these
problems, a design with two identical Pyrex tubes joined by Beckman Teflon
connectors fitting into a unit housing is recommended. This design has
been described in detail in Section 2. One unit of such design was
mailed from Bendix to EPA and arrived at the destination safely.
74
-------
ALL DIMENSIONS
IN CENTIMETERS
0.305 DRILL ,
THROUGH
Figure 32 - Luer-coupler (for interfacing the second-generation
collector cartridge to the analyzer)
3.6 SAMPLING PUMP
Since the sampling pump is one of the key components of the sampl-
ing system, currently available pumps were surveyed. Desirable features
for this pump are: (1) low power DC- or 110V AC-operated; (2) capability
for continuous 24 hr unattended operation; (3) capability for providing
high air flow rates, greater than 10 £/min if possible.
Literature was requested from over thirty manufacturers of pump and
air samplers. The majority do not make DC motors for the pump. Most of
the DC-operated vacuum pumps on the market are low in air flow rate (less
than 10 £/min) and are not designed for continuous operation with any
restrictions other than particle filter paper (<100 mm Hg) . The most
promising pumps that can most closely meet the requirements are Model
107C DC20 from Thomas Industries Inc. (Sheboygan, Wise. 53081) and Model
1531 from Cast Manufacturing Corp. (Benton Harbor, MI 49022). Both
pumps deliver 35-42 Jl/min free air flow, and remain operative against
500-600 mm Hg vacuum; they both can be powered by 12V DC source.
75
-------
The Cast pump was chosen for the present program since its manu-
facturer had a better delivery time than Thomas Industries. This Cast
Rotary Vane Pump weighs 8 lb., has a 1/10 HP power requirement, and
draws 11.3 A for operation. A 12V battery or 12V battery charger (12
amp-rate) can serve as power supply for this pump.
3.7 COLLECTION SYSTEM (AIR SAMPLER)
The system is required to collect samples continuously up to 24 hr
in the field with either DC or AC power. This capability is provided
by the installation of an interval timer, and 12V battery charger to serve
as power supply and rectifier, or a 280 amp hr truck battery.
The development effort was concentrated on the design of a portable,
weatherproof and tamper-proof housing for the collection unit, vacuum pump,
interval timer, power supply (battery charger) and battery. This has been
accomplished with a housing made of aluminum; it contains three shelves
for respectively the collector unit and pump; timer and battery charger
controls; and battery. Locks are provided for the housing. For dis-
assembly and transportation, the housing can be easily detached and folded.
The details have been given in Section 2.
3.8 INTERFACE DEVICE
When thermal release was established as a satisfactory means of
recovering the trapped oxides from the collectors for analysis, the inter-
face design became a relatively simple matter. Two approaches were tried.
The first approach utilized an aluminum tube to serve as the oven
body which readily accepted the collector cartridge. Thermal power was
supplied from a thin integral heater/sensor foil (Thermofoil heater) manu-
factured by Minco Products, Inc. The Thermofoil heater was easily wrapped
76
-------
around the aluminum heating tube with pressure-sensitive adhesive. The
heater took either DC or AC power source; its heating power was limited
by the resistance of the heater foil, the power supplied, and the mass
and geometry of the heat sink. Repeated trials, however, failed to
produce an oven capable of bringing the collector to the desired temper-
atures. Custom manufacturing would be necessary to accomplish the task
and was too expensive to pursue.
The second approach resulted in the interface design depicted and
described in Section 2. This device had a variable temperature setting
ranging from ambient to 500°C; this range was made possible by a propor-
tional temperature controller and platinum sensor.
3.9 PRELIMINARY FIELD COLLECTION AND ASSAY OF N02 with TEA-HCl
This was only a preliminary and limited experiment. It was designed
to check the performance of the collection material under a more realistic
field condition. A continuous chemiluminescence NO -monitor, and a con-
A
tinuous Saltzman NO -monitor (Technicon Autoanalyzer) were used for
comparative study.
3.9.1 Experimental Procedure
This experiment was conducted at Wayne County Department
of Health, Air Pollution Control, Detroit. Air sample was taken from a
sampling manifold leading into the laboratory from the rooftop. One mani-
fold line was led to a Bendix chemiluminescence NO monitor, and another
A
to a Technicon Wet Chemistry Analyzer for comparative, continuous NO,
N0? monitoring. From the same manifold, air samples were drawn into
77
-------
the collector by a small diaphragm pump for 24 hours continuous NO--
collection. The collectors were tubes packed with Triethanolamine hydro-
chloride mixed with Chromosorb G AWDMCS (45/60 mesh) in 1.5/1 ratio.
Approximately 6 grams of packing was used for each collector. For assay,
the same elution/area count technique as previously described was used
as in the screening tests with a Bendix NO -analyzer. The experimental
A
results are summarized in Table 4. Table 5 gives the hourly [NO ] average
read outs by Technicon and chemiluminescence analyzers.
3.9.2 Discussion
The experimental results indicate that the present ambient
and assay method with solid absorber is promising when the flow rate
is kept around 100 cc/min. The comparison data were obtained by a con-
tinuous Saltzman technique using a Technicon Autoanalyzer, and by a
chemiluminescent analyzer running at the same time on the same sample.
As shown by the attached Table 5, the hourly averages of the two tech-
niques do not agree satisfactorily and neither do the 24 hr averages.
The present collection and assay technique show consistently higher
results than both other methods. While it is possible that these pre-
liminary data contain systematic errors due to variations in flow, the
lack of agreement of the two other techniques suggests the presence of
inadequacies in the methods employed. The possibility definitely exists
that the present collection and assay technique gives correct results,
in which case the other two techniques would give results seriously in
error on the low side. It would be worthwhile to remove these uncer-
tainties and obtain consistent results both in controlled laboratory
conditions and in the field.
78
-------
Table 4 - Summary of Experimental Results
Collector
Number
2
3
16
Flow
Rate
1 lit/min.
1 lit/min.
105 cc/min.
Collection
Time
24 hrs.
24 hrs.
24.2 hrs.
Daily Average [N02]
(By Technicon and
Chemiluminescence)
0.035 ppm
0.025 ppm
0.032 ppm
Measured
Collected
Quantity
11.99 Mg
4 Mg
(a) by area count
quantity 10
Collected
(Calculated
Daily Ave
90.7 Mg
64.8 Mg
8.8 Mg
17
80 cc/min. for 4 hrs.
75 cc/min. for 20.2 hrs.
24.2 hrs.
0.032 ppm
12 ug
(b) by weighing
13.85 pg
(a) by area count
7.6 ug
(b) by weighing
7.4 Mg
6.29 Mg
-------
Table 5 - Comparison of Ambient NC^-Concentration by
Technicon and Bendix Monitor* (Date: 12-8-73)
.me (hr.)
0-1
1-2
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23-24
Technician
ppm
0.050
0.050
0.040
0.040
0.030
0.030
0.030
0.040
0.040
0.030
0.030
0.030
0.020
0.020
0.020
0.020
0.020
0.030
0.030
0.030
0.030
0.030
0.030
0.030
Bendix
ppm
0.035
0.035
0.030
0.030
0.020
0.020
0.025
0.030
0.030
0.025
0.025
0.020
0.015
0.020
0.015
0.015
0.020
0.025
0.025
0.020
0.025
0.020
0.020
0.020
Tech: Bendix
Ratio
1.43
1.43
1.33
1.33
1.50
1.50
1.20
1.33
1.33
1.20
1.20
1.50
1.33
1.00
1.33
1.33
1.00
1.20
1.20
1.50
1.20
1.50
1.50
1.50
80
-------
SECTION 4
SUMMARY AND RECOMMENDATIONS
4.1 SUMMARY
Simple collection and assay of NO and N02 using solid chemical ab-
sorbents was proven feasible. The selective collection of these oxides
is possible with a tandem arrangement of two absorbent cartridges: The
first, containing triethanolamine hydrochloride packing, completely absorbs
NO- from the air sample without affecting the NO concentration. The sec-
ond, filled with cobalt oxide packing, quantitatively removes NO. These
collected oxides are later eluted and analyzed separately.
The interface device developed to transfer the sample for analysis
accepts one collector cartridge and heats it to a temperature which re-
leases the nitrogen oxide. At 160-170°C, the absorbed N02 is quantitively
released from the TEA-HC1 collector as NO; cobalt oxide gives up its cap-
tured NO completely at 400-420°C. When the collector is coupled via this
interface to a suitable instrument, such as a chemiluminescence NO ana-
' x
lyzer, the NO and N0_ are easily measured. When cooled, the collector
cartridges are ready for reuse in sample collection.
For convenient handling, a special collector unit was developed.
This unit is easy to assemble and to install in a collector system; more-
over, it is easy to disassemble. Its rugged, lightweight construction
enables its safe and inexpensive transportation within the postal system.
81
-------
A simple weatherproof collection system was built for field collec-
tion, based on the concept of a collector unit containing the solid chem-
ical absorbents. Its Cast vacuum pump and interval timer enable the sam-
pler to operate up to 24 hr continuously without attention. This operation
can be powered either by a 12V DC high capacity battery or by a 110V AC
power supply with a 12V DC output.
Although the collectors developed under this contract met the per-
formance requirements in the laboratory, extensive field testing may lead
to possible design improvements. The major effort of the program was in
the screening and evaluation of chemical absorbents, so the collector unit
represents an experimental prototype. Additional work will be required
to rigorously test and further improve the designs. Recommendations for
further work are detailed in the following subsections.
4.2 RECOMMENDATIONS
4.2.1 NO- Collector With TEA-HC1
TEA-HC1 (m.p. 177-179) selectively collects N02 and is inert
to NO at room temperature. At higher temperatures, its affinity for NO-
decreases and it catalyzes the conversion of N02 to NO. Its desirable
property as an NO absorbent, however, returns when the temperature is
decreased. In use, the TEA-HC1 cartridge is heated to 160-170° to recover
the absorbed N0«, thereby regenerating the absorber. If the temperature
exceeds 170°C, however, some discoloration of the absorbent occurs, after
which N0« to NO conversion is noted with the cartridge at room temperature.
Limiting the temperature to 170°C, therefore, is a necessary precaution
with this cartridge.
82
-------
The ultimate lifetime of the TEA'HCl absorbent should be
determined. Laboratory tests confirmed its performance up to ten heating
cycles, but more extensive cycling should be evaluated. If the absorbent
has a long lifetime, it can be encased in a more permanent material, such
as Teflon-coated aluminum instead of the present disposable Pyrex. The
metal case has the obvious advantage of nonbreakage.
Absorbent capacity also should be further evaluated. The
present collector cartridge was designed in a size convenient to pack
and handle without regard to the ultimate absorbent capacity. When the
absorbent's maximum capacity has been determined, the optimum size for
the cartridge can be established.
4.2.2 Cobalt Oxide
Since both cobaltous and cobaltic oxides absorb NO com-
X
pletely, conversion of NO- to NO is not a problem. The total absorption
capacity of each oxide should be measured, however, to design an optimum
collection cartridge.
Preliminary tests indicate that cobaltic oxide quantitatively
absorbs and releases NO independently from the N0_ absorption. If this
is true, one could use cobaltic oxide alone for the collection and assay
of the nitrogen oxides. The analyzer, such as a gas chromatograph, would
have to differentiate the NO- from NO, since they would evolve simultan-
eously from the collector. Further investigation of this phenomenon
could lead to a single absorbent collector.
83
-------
The extensive and combined effects of possible interfering
species, such as moisture, SO-, peroxide, and C02, on collector perfor-
mance could be examined. It has been established that moisture present
in air up to 96% humidity does not affect the performance of the collectors.
The TEA-HC1 absorbent was subjected to 1000 ppm S02 at a flow rate of U/
min for 20 min and still retained its performance. Nonetheless, evalua-
tion of collection efficiency in the presence of both moisture and S02
in more common concentrations should be systematically investigated.
The preliminary data indicated that the presence of highly concentrated
S02 and moisture (96%) did lower the concentration of NO passing through
the TEA'HCl. This may be caused by the conversion of NO to N02, the
latter then being absorbed. The conversion may be attributed to the
combined effect of SO-, moisture, and the surface of the absorbent. In
the absence of the absorbent, no change in concentration was observed.
This indicates the need for a systematic study of the effect of interfer-
ences on both absorbents under realistic field conditions.
The nature of bonding of the nitrogen oxides to the absor-
bent surface was not studied for either absorbent. This phenomenon should
be investigated using IR and DTA techniques to determine what inherent
properties of the absorbents might degrade their performance. Such an
investigation might also reveal what other types of compounds could serve
as NO and N0_ absorbers.
A.2.3 TEA'HCl Packing
This absorbent is a fine, white powder which, when used by
itself as a collector packing, introduces great flow restriction and re-
sults in a poor flow rate. Physical mixing of such a fine powder with an
84
-------
inert support such as Chromosorb G AW DMCS 45/60 does not produce an
especially good packing. The best packing is produced by solution-coating
the TEA'HCl (in distilled water) onto a support such as Celite. This
technique gives a high-capacity packing with desirable flow characteristics,
Two points with respect to this collector packing are worth
further consideration. First, the Celite surface should always be com-
pletely covered by the absorbent. Celite absorbes N0_ at room temperature
and, although it gives up the captured N0_ at 170-200°C, the efficiency
of the absorbent is better preserved and controlled when no extra foreign
surface is present. Second, the collection efficiency of this absorbent
at flow rates higher than 120 cc/min has not been evaluated. Since field
collection may involve flow rates > l£/min, the absorbent efficiency under
such conditions should be determined.
4.2.4 Cobalt Oxide Packing
Commercial cobalt oxide is a grayish-black fine powder
(^200 mesh). This powder must also be mixed with some inert, more coarse
support (Chromosorb G, 45/60) to afford a reasonable flow rate. Other
means of making packings of lower flow restriction that should be further
investigated are
• Hot-pressing or cold-pressing the oxide into a cake, firing the
cake at high temperature, then crushing the cake into chips for
packing.
o Solution-coating Co(NO_) onto a support such as Celite, drying,
then coverting to Co20 at 200-250°C in air.
85
-------
With the first method, the packing's capability of retaining its shape
without breaking down to a powder is the major question. Moreover, the
firing temperature chemically converts the Co20_ into CoO and Co^,
which may affect the desirable properties as an absorber. The second
method gives good flow characteristics as evidenced by its use with the
TEA-HCl/Celite packing; however, removing the residual nitrate from the
packing is quite difficult. In fact, conditioning of packing prepared
by this method for over a week at 450°C in air did not result in a clean
absorbent. A systematic evaluation of these packings should be undertaken.
4.2.4 Collector Configuration
The present design of the collector unit contains two Pyrex
cartridges. Although these cartridges are ruggedly constructed, they can
be broken during assembly or disassembly. When the necessary tests are
performed and the reusability and lifetime of the absorbents are fully
established, a nonbreakable cartridge design can be considered. Experi-
mental work showed that aluminum cartridges coated inside with Teflon
did not affect N0_ collection. The Teflon coating is required for the
TEA'HCl/Celite cartridge, since any NO™ to NO conversion results in ana-
lytical error.
The final cartridge design will still require a precaution
with respect to overheating, since, extremes of temperature will induce
decomposition of the packing and the coating. A longer range program
may prove worthwhile, therefore, to develop a collection material of
higher thermal stability.
86
-------
APPENDIX A
TYPICAL RECORDINGS OF
COLLECTOR MATERIALS EVALUATION
A-l
-------
APPENDIX A
TYPICAL RECORDINGS OF COLLECTOR MATERIALS EVALUATION
This appendix contains a collection of recordings representing the
evaluation of various absorbents and container materials. These record-
ings are given in Figures A-l through A-15.
A-2
-------
Figure A-l - General material evaluation, showing the inertness of (1)
Pyrex tube, (2) silane-treated glass wool, (3) Chromosorb
T, and (4) Chromosorb W AW DMCS
-------
P>J
o
CM
X
[NO] 2 = [N0]c
[NOX]2=[NOX]C
THE SUPPORT IS
INERT
P-8 7-803-3
1
0
n
0
o
s
00
cc
8
o
0
cc
X
o
0
z
^
o
8
6
X
t—
^
HI
cc
o
o
2
I
^- ^^_
Figure A-2 - Chromosorb W AW DMCS
-------
.. INCOMPLETE NO2 - REMOVAL
CONVERTSNO2- NO
IN°) DIRECT
0.020
IN°I
COLLECTOR
u
X! COLLECTOR
'N°lCOLLECTOR
0.08
1 1 1 1 h
Figure A-3 - TEA/Chromosorb W AW DMCS
-------
THERE'S NO APPARENT
ABSORPTION OF NO2;
SIGNIFICANTLY CONVERTS NO2
INOICOLLECTOR X0'5
0.315
Q
i UJ
I &
o
INOI DIRECT 0-MPPM
Figure A-4 - TEA/Regis GasPak FS
-------
. INCOMPLETE
N02 - ABSORPTION
EXCESSIVE N02 - NO
CONVERSION
Figure A-5 - TEA/Chromosorb T
A-7
-------
[NOx]TEA(,jq)
0.22
0.12
^DIRECT
0.020
Figure A-6 - TEA/NEAT (liquid)
A-8
-------
DIRECTX
2.3 PPM
1 O
CM
X
cc
O
O CM
O
o
[NOX] DmECT X5
2.24
DIRECT* 2
0.08
IN°X]COLLECTOR X 5
0.28
Figure A-7 - TEA Borate
-------
f *
Figure A-8 - THEED
-------
I [N0]c
0.30
9*
INSIGNIFICANT
N02-REMOVAL
i
Figure A-9 - Quadrol/Chromosorb W AW DMCS
A-11
-------
to
DOES NOT ABSORB NO2. BUT
CONVERTS NO2-NO
IN°) COLLECTOR X °'5
0.02
[NO
IN°) DIRECT X0'5
H 1-
Figure A-10 - NiSO,
-------
DOES NOT ABSORB
NOo
[N°1 COLLECTOR X0-5
0.09 PPM
X1
[NOX] COLLECTOR
0.22 - 0.24
n
o
oo
Figure A-ll - CoSO.
A-13
-------
IN°] DIRECT X1
0.22 PPM
Figure A-12 - CoO
A-14
-------
IN°.1CU20
s
0.69
0.32
H H
H 1 1-
INDIRECT
0.30 PPM
-\ 1-
Figure A-13 - CuO
-------
INOX]
0.59 ALUMINUM
[NOX1 DIRECT
0.60 PPM
(4-5MIIM.-*)
IN01ALUMINUM
INCREASES WITH
TIME INDICATING
CONVERSION
• CONVERTS NO2
TO NO
• DOES NOT ABSORB
NO2
\
INO'DIRECT
0.38 PPM
t
TO ALUMINUM
Figure A-14 - Aluminum cartridge (after heating cycle to 200°C and
cooled down to room temperature)
A-16
-------
-------
APPENDIX B
DIRECTIONS FOR ASSEMBLING
COLLECTOR FIELD HOUSING
(1) Place folded housing in flat position with back side down (large
door with hasp up).
(2) Open door and remove legs.
(3) Close door and turn unit over (front down).
(A) Remove wing nut and screw holding cover in folded position.
(5) Unfold housing by pulling upper section so it swings upward and
toward bottom.
(6) Stand housing upright with cover in full open position.
(7) Remove five remaining wing nuts holding sides in folded position
(8) Swing sides into position running studs into appropriate holes
and fastening with wing nuts.
(9) Lay the housing on one side and install legs on the other side
with 1/2" long 1/4-20 bolts; match one leg with clearance hole
to stud head.
(10) Turn housing onto other side and install remaining two legs.
B-l
-------
Figure B-l - Folded View of Collector
System Housing
Figure B-2 - Assembled Collector Unit
------- |