United States
Environmental Protection
Agency
Atmospheric Research and
Exposure Assessment Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/S3-90/013 June 1990
&EPA Project Summary
Evaluation of Methods Used to
Desorb the Constituents
Adsorbed on the Charcoal
Contained in Automotive
Evaporative Canisters
David Dropkin
This summary represents a two-
part study which evaluated current
extraction methods for analyzing
charcoal canisters used to control
evaporative emissions in
automobiles. The second part of this
study investigated the use of solvent-
free extraction methods such as high
pressure CO2: soxhlet extraction and
vacuum transfer extraction. The
results of th@ solvent-free methods
were then compared to the CS2
soxhlet extraction method.
The results of this study showed that
the CSa method extracted up to 8% more
material (by weight) from the charcoal
than did the vacuum transfer method and
up to 15% more material (by weight) than
did the high pressure CO2 soxhlet extrac-
tion method. In addition, more total
hydrocarbons were measured with the
CSa method than were measured in either
the vacuum transfer or the high pressure
COa methods.
The vacuum transfer apparatus used
for this study was a modification of the
apparatus that had been used for a pre-
vious study. More of the higher molecular
weight hydrocarbons, mainly alkylated
aromatics greater than carbon number 9,
were measured in the extracts obtained
from the modified vacuum transfer ap-
paratus than were measured in the ex-
tracts from the previous method. Lower
quantities, however, of the lower
molecular weight hydrocarbons were
measured than before.
The high pressure COa soxhlet extrac-
tion method gave the lowest hydrocarbon
measurements of the three methods.
This Project Summary was developed
by EPA's Atmospheric Research and Ex-
posure Assessment Laboratory, Re-
search Triangle Park, N.C. to announce
key findings of the research project that
Is fully documented in a separate report
of the same title (see Project Report or-
dering Information at back).
Introduction
This laboratory conducted a two-part
study in order to determine the most ac-
curate, practicable and cost effective
method of evaluating charcoal canisters.
Charcoal canisters are used to collect
evaporative hydrocarbon emissions from
gasoline powered vehicles. When the
Manufacturers Operations Division (MOD)
at EPA began testing vehicles to determine
their compliance with the federal emission
standard, a significant number of the
vehicles failed to meet the evaporative
hydrocarbon emission standard. The
MOD suspected that these failures, were
caused by a decrease in the ability of the
charcoal canister to capture hydfocarbon
vapors. As an aid in their assessment of
canister performance, MOD requested
our laboratory to determine the nature
and quantity of the material adsorbed and
retained on the .charcoal contained on
these canisters. There was concern,
however, that using the present (CS2
soxhlet extraction) method in the
-------
analysis of the canisters might alter the
quantity and quality of hydrocarbons or
other materials desorbed from the canister's
carbon. In addition, it was also unknown
which of the current methods of desorbing
the hydrocarbon constituents from the carb-
on would be the most accurate, practicable
and cost effective.
The first phase of this study involved the
use of various solvents or a combination of
solvents to extract the charcoal in a soxhlet
extraction apparatus. We also ultrasoni-
cated the charcoal using CSa as the solvent.
Solvent extractions are most often the stand-
ard method to be compared with and usually
provide an indication of the maximum
amount of adsorbed material that can be
extracted from the charcoal. On the other
hand, solvent extraction introduces im-
purities, such as benzene in CSa, and the
potential for errors with many sample han-
dling steps. The results of the first phase of
this study show that CSa was the most prac-
ticable solvent for the extraction and analysis
of the charcoal. The total quantity of material
soxhlet extracted from the seven charcoal
samples analyzed was the same for the
various solvents tested, except for
melhanol. CSa was the preferred extracting
solvent because, in the GC/FID analysis of
the CSa extracts, the CSa solvent peak only
interfered with the measurement of the Cs
hydrocarbons. When using solvents such
as CHaCla or hexane, the solvent peak inter-
fered with the measurement of the Ce and C/
hydrocarbons.
The second phase of this study inves-
tigated solvent-free extractions of the char-
coal and included a modified vacuum
transfer method and a high pressure COa
soxhlet extraction method. We did not inves-
tigate the use of microwave desorption, ther-
mal desorption or super-critical fluid (using
COa under high pressure) extraction, to ex-
tract these charcoal samples because of
equipment unavailability and budget con-
straints.
Solvent-free extraction of the charcoal
would have been the preferred method be-
cause this method minimizes sample han-
dling and reduces the potential for heat
induced artifacts to be formed during the
extraction process. In addition, solvent free
extractions eliminate the solvent peak inter-
ference in the GC/FID measurement of the,
hydrocarbons in the extracts.
Procedure
Handling and Storage of
Charcoal Canisters
The seven canister samples analyzed for
this second part of the study were the same
as those analyzed in Part 1. Five of the
canister samples had been prepared and
analyzed for previous canister studies. The
other two canisters were taken from an in-
house evaporative emission project where
the loading and purging of the canisters had
been done under controlled conditions. One
canister was loaded with gasoline vapors
from a base stock gasoline and was labeled
gasoline-base. The other canister was
loaded with gasoline vapor containing 10%
oxinol (5% methanol + 5% t-butyl alcohol).
The preparation and handling of the samples
has already been described in Part 1. The
two unused (blank) charcoal samples that
were analyzed in Part 1 were analyzed in the
initial set up of the high pressure COa soxhlet
and vacuum transfer methods. No material
was extracted from the blanks (except water)
and, therefore, no hydrocarbons were
measured.
High Pressure CO-2 Soxhlet
Extraction
A high pressure soxhlet extractor (J&W
Scientific, Folsom, Calif.) was used to extract
the charcoal (Fig. 1). COa was used as the
Pressure Gauge
3000 ps! _
Needle Valve
extracting solvent. The high pressure CO2
extraction procedure used was the same
procedure as described in the operating and
supplementary instructions provided by
J&W Scientific.
The charcoal sample (approximately
1.00g) was weighed into a preconditioned
cellulose thimble and the thimble was then
placed into the soxhlet extractor. The soxhlet
extractor was placed in a pressure chamber
and the chamber was then sealed. The
chamber and filling lines were then
evacuated under moderate vacuum. After
the inlet and outlet valves to the chamber
were closed, the inlet lines were filled with
liquid COa and the lower part of the chamber
was placed into a preheated (60°C) water
bath. The chamber was immersed in the
water bath for 15 min. and then it was moved,
in a vertical position, to a ring stand. The
pressure chamber was then filled slowly with
liquid COa to its operating pressure of 820
PSI (ambient temperature of 21 °C). The
lower part of the loaded pressure chamber
was then placed back into the water bath and
All Metal Parts
304 Stainless Steel
To Vent
••*- Needle Valve
To COa Cylinder
Heated Water
Bath 40°C to 45°C
Aluminum Heat
Transfer Plate
All Glass Soxhlet
Apparatus
Cellulose Thimble
Magnetic Stirrer
Figure 1. High pressure COa soxhlet extractor.
-------
the flow of the 10°C water to the top of the-
extractor (cold finger condenser) was turned
on. All the extractions were done for 1 hour
after initially determining that the amount of
material extracted was the same if the extrac-
tion time was 1, 2 or 4 hours. The operating
pressure in the chamber was maintained at
its initial pressure during the entire 1 hour
extraction period. After the extraction was
completed, the chamber was removed from
the water bath, the chilled water for the con-
denser was turned off and the chamber was
then de-pressurized, slowly, to ambient. The
time required for de-pressurizing the cham-
ber was usually 20 minutes. The de-pres-
surized chamber was then opened quickly
and the extracted material was transferred
from the flask into a pre-weighed vial using
a disposable pipet. The vial was stored in a
-20°C freezer. The thimble containing the
extracted charcoal was placed in a desic-
cator for 1 hour before determining the
weight loss of the charcoal.
Vacuum Transfer
The vacuum transfer apparatus had been
modified (Fig. 2) in order to eliminate the
apparent limitations of the apparatus pre-
viously used (Fig. 3) to do the charcoal
extractions. The modifications were madeto
reduce the potential for vacuum leaks by
eliminating as many greased joints as pos-
sible and to eliminate cold spots in the
manifold so that the less volatile, relatively
higher molecular weight hydrocarbons
would be transferred into the receiving
flask.
This modified vacuum transfer apparatus
did not contain any joints that require
vacuum grease that would flow as soon as
heat was applied to the system and should,
therefore, be less likely to leak. All the con-
nections in the extraction region were made
using FETFE 0-rings. The top of the 250 ml.
distilling flask and one end of the manifold
were threaded so that a vacuum coupling
connector (Ace Glass, Vineland N.J.) could
be used instead of a greased joint. The
manifold was evacuated from the receiving
end of the apparatus instead of the center in
order to eliminate a cold spot at the vacuum
stopcock manifold connection. The joint at
the receiving end of the apparatus was made
by using a vacuum adapter with an inner ring
design (Ace Glass) to connect to a
centrifuge tube. The manifold vacuum
gauge was connected directly into the sys-
tem in order to eliminate another greased
connection and to make it possible to
monitor the vacuum conditions inside the
system during the extraction.
The procedure used to extract the char-
coal was the same as had been used for the
previous extraction apparatus. Smaller
quantities of the charcoal (5.00 to 20.00g.)
To Cold Trap
and Vacuum
Pump
Vacuum Gauge
1 - Liquid Nitrogen Dewar
2 - Heating Mantle to Heat to
400°C
Figure 2. Previous vacuum transfer apparatus.
Needle Valve
0-8/77/77
"O*-Rings
\
Liquid Nitrogen Dewar
To Cold Trap
and Vacuum
Pump
Ace Glass
Vacuum Seal
Flask Connector
1 - Liquid Nitrogen Dewar
2 - Heating Mantle to Heat to
400°C
Figure 3. Present vacuum transfer apparatus.
To Vent
Vacuum Gauge
0-8/77/77 Hi Vacuum
"O"-Rings
•O'-Ring
f— (Ace Glass "O"-Ring
Seal)
10mL Graduated Tube
(Pyrex)
I "*— Liquid Nitrogen Dewar
-------
were extracted instead of the 40.00g used
previously in order to conserve the samples.
Gas Chromatographic
Equipment and Procedures
All the extracts were analyzed using a
Psrkin Elmer Model 3920 gas
chromatograph equipped with a flame
ionization detector. An'aliquot of the sample
was vaporized into a tedlar bag and then
introduced into the GO system using a 10 mL
sample loop. This method of sample intro-
duction had been used in the previous
canister study and the procedure already
described In Part I. A 60m x 0.32 mm i.d.
DB-1 (J &W Scientific) capillary column with
a 1.0 u film thickness was used to separate
the C4to C13 hydrocarbons. Upon injection
of the sample to the capillary column, the
sample is cryogenically trapped in the
column which was initially cooled to -95°C.
After the one-minute injection period, the
temperature to the capillary column is
ramped to -45° C and then slowly increased
at the rate of 8°C/min up to a final tempera-
ture of 235°C. The Helium carrier gas flow
was maintained at 1.5 mL/min using a Math-
eson Model 8240 mass flow controller
(Matheson Corp., East Rutherford N.J.). The
peak area and peak retention time data was
recorded and plotted on a Perkin Elmer
Sigma 15 Chromatographic data station. In
addition, the peak area and peak retention
time data was stored on cassette tape for
later transfer to a computer using a standard
RS 232 serial port.
Results And Discussion
The CSa extraction method extracted up
to 8% more material (by weight) from the
charcoal than was extracted by using the
modified vacuum transfer procedure, for six
of the seven samples tested. In addition, the
CSa extraction method extracted up to 15%
more material (by weight) from the charcoal
samples than was extracted by using the
high pressure COa soxhlet extraction proce-
dure (Table 1 and Fig. 4).
Table 1. Comparison of Charcoal Extraction Methods - Percent (by Weight) Material Extracted from Charcoal
Methods
Sample Number
A152I0237
A151/0420
A15U0320X
A151/0282
A1S2/OS99
Gasoline-base
Gasoline-blend
CSa
Soxhlet
Extraction
31.10
44.59
36.60
35.20
25.60
22.40
20.85
Modified
Vacuum
Transfer
23.30
47.36
33.98
30.40
19.40
21.40
18.20
High Pressure
C02Soxhlet
Extraction
16.60
36.80
24.00
28.20
17.80
14.30
19.70
50
40
Weight 30
Loss
(Percent)
of
Charcoal
20
10
| CS2Soxhlet
Ijj Vacuum Transfer
|3] HPCOzSoxhlet
0237 0420 0320X 0282 0599 Base Blend
Sample Number
Figure 4. Comparison of charcoal extraction methods - weight loss (percent) of charcoal.
-------
Up to 4% more total hydrocarbons on
carbon was measured by the GC/FID
analysis of the extracts from the €82 soxhlet
extraction method in comparison to the
hydrocarbons measured in the extracts from
the vacuum transfer method. Moreover, up
to 8% more total hydrocarbons on carbon
was measured in the CSa extracts than were
measured in the extracts from the high pres-
sure COa soxhlet extraction procedure
(Table 2 and Fig. 5).
The results of the comparison between the
CSa extraction method and the vacuum
transfer method were the same as had been
reported in a previous canister study. More
material (by weight) had been extracted from
the charcoal with the CSz soxhlet extraction
method than had been extracted by the
vacuum transfer method. In addition, the
total hydrocarbons on carbon that were
measured in the extracts with the CSa
soxhlet extraction method were greater than
the total hydrocarbons measured in the ex-
tracts from the vacuum transfer method.
The total hydrocarbons on carbon
measured in the CSz extracts were greater
than the total hydrocarbons measured in the
vacuum transfer extracts because 082 ex-
tracts more of the high molecular weight,
alkyl aromatics than the other method. In
our previous canister studies we had deter-
mined that used "aged" canisters contained
significant quantities of alkyl aromatics.
Table 2. Comparison of Charcoal Extraction Methods - Percent (by Weight) Hydrocarbon on Carbon from GC/FID
Methods
Sample Number
A152/0237
A151/0420
A151/0320X
A151/0282
A1S2/0599
Gasoline-base
Gasoline-blend
CSz
Soxhlet
Extraction
21.54
33.13
14.96
22.87
15.39
15.01
11.90
Modified
Vacuum
Transfer
17.34
32.13
13.43
21.15
15.63
13.82
10.90
High Pressure
C02 Soxhlet
Extraction
14.54 ,
25.53
10.97
18.19
10.31
Not Measured
6.69
Percent 20
(by Weight)
Hydrocarbon
on Carbon
CSz Soxhlet
Vacuum Transfer
_1 HP COz Soxhlet
0237 0420 0320X 0282 0599 Base Blend
Sample Number
Figure 5.. Comparison ofcharcoaf extraction methods - percent (by weight) hydrocarbon on carbon from GC/FID.
-------
Conclusions
The results from the second phase of this
study show that, at present, the CS2 soxhlet
extraction method is the most practicable
and efficient method for desorbing and
analyzing the Ce+ constituents adsorbed on
charcoal. This method, however, can only
aid in determining the composition of the
relatively higher molecular weight com-
pounds on the charcoal, since the more
volatile components are lost through the
condenser system.
High pressure COa soxhlet extraction is
not a practicable method for extracting these
charcoal samples because this method
does not completely extract the material ad-
sorbed on the charcoal. Moreover, the
results showed that the lower molecular
weight, more volatile, hydrocarbons were
not retained in the extracts.
The modification of the vacuum transfer
apparatus did not improve the method suffi-
ciently to be able to recommend this method
for extracting charcoal rather than the CSa
soxhlet extraction method. More of the alky-
lated aromatic hydrocarbons were
measured in the extracts from the modified
vacuum method than were measured using
the previous apparatus. The quantities of the
alkyl aromatics measured using this
modified vacuum method were, however,
lower than measured by using the CSa ex-
traction method. Moreover, lower quantities
of the lower molecular weight hydrocarbons
were measured in the extracts using this
modified apparatus than measured using
the previous apparatus.
Recommendations
A mass balance involving the amount of
material extracted and the quantity of
hydrocarbons measured in the extracts can-
not be attained using the CSa soxhlet extrac-
tion method. A second extraction of the
charcoal, using the vacuum transfer
method, must be done in order to determine
the quantity of the hydrocarbons less than
Ce adsorbed on the carbon. The vacuum
transfer method is time consuming, how-
ever, since it requires one day to extract and
analyze one sample. Further modification of
the apparatus is required to reduce or
eliminate vacuum leaks in order to retain the
lower molecular weight hydrocarbons. Two
other charcoal extraction methods, thermal
desorption and High Pressure COa super-
critical fluid extraction, should be inves-
tigated since these methods are
solvent-free, require minimum sample han-
dling, eliminate exposure to toxic CS2
vapors, and might be more cost effective
than soxhlet extraction with CSa.
David Dropkln is the EPA Project Officer (see below)
The complete report, entitled "Evaluation of Methods Used to Desorb the Constituents
Adsorbed on the Charcoal Contained in Automotive Evaporative Canisters - Part
I," (Order No. PB 90-188 830/AS; Cost $17.00, subject to change) and "Evalua-
tion of Methods Used to Desorb the Constituents Adsorbed on the Charcoal
Contained in Automotive Evaporative Canisters - Part II," (Order No. PB 90-188
848/AS; Cost $17.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
------- |