a p p l i c at i o n N o t e
Gas Chromatography
Author
A. Tipler
PerkinElmer, Inc.
Shelton, CT 06484 USA
The Determination of Benzene
and Toluene in Finished Gasolines
Containing Ethanol Using the
PerkinElmer Clarus 680 GC
with Swafer Technology
Introduction
ASTM® Test Method, D-3606, is designed
to determine the benzene and toluene
content in gasolines using packed columns
in a 2-column backflush configuration.
This is an established method that was
originally developed to analyze gasoline
that did not contain ethanol. Ethanol is a
biofuel that is added to modern gasolines
to improve combustion efficiency. Significant quantities of this additive may be present – for instance
10% in the USA (E10) and 25% (E25) in Brazil. The presence of large amounts of ethanol in samples
caused problems with chromatographic co-elution with benzene when using the D-3606 method.
The method was revised (D-3606-07) to include an alternative column set designed to handle the
presence of the ethanol but there are still reports of problems with co-elution and further column
sets are under consideration.
This application note describes a method that is based on the original ASTM® D-3606 method with
the main difference being that capillary columns are used. This approach completely eliminated all
chromatographic interference from the ethanol (even solutions made up in pure ethanol could be
run), improved the quality of the chromatography in general and reduced the analysis time significantly
(by 50% or 75% depending on the column set).
Analytical approach
Initial Setup Procedure
The traditional D-3606 packed column method uses a 2-column
backflush configuration. The first column (precolumn)
was packed with 10% polydimethylsiloxane (PDMS) on a
Chromosorb®-W support. The second column (analytical
column) contained a packing of 20% 1,2,3-tris(cyanoethoxy)
propane (TCEP) on a Chromosorb®-P support.
The Swafer Utility Software was used to establish the carrier
gas pressures needed for this analysis. Figure 1 shows a
screen shot taken from this software showing the configuration
and suggested pressure settings used for the initial work.
In this application, these columns were replaced with narrowbore capillary columns containing the same stationary
phases as used in the packed columns. A split/splitless injector
was used to introduce the sample into the precolumn. A
Swafer™ system was used to manage the backflush process.
Although D-3606 specifies the use of a thermal conductivity
detector (TCD), in this work a flame ionization detector (FID)
was used as it was more suited to high-resolution capillary
chromatography and did not give baseline drift as column
pressures were changed during the backflushing process.
To begin, 0.3 µL of a solution of ethanol, benzene and
toluene in iso-octane was injected into the system using the
conditions shown in Figure 2. Figure 3 shows the resultant
chromatography from this injection. All components have
eluted from the precolumn (and into the TCEP column)
within 1.61 minutes and so this time was adopted as the
backflush point.
100mL/min
Injector at 200°C
S-Swafer
60cm x 0.100µm fused
silica restrictor
FID at 200°C
30m x 0.250mm x 0.25µm
Elite-1
Experimental conditions
30psig
hydrogen
20psig
hydrogen
50m x 0.250mm x 0.4µm
TCEP
GC oven = 100°C isothermal
Figure 2. System configured to monitor the precolumn chromatography to
establish the backflush point.
Benzene
Toluene
Note that hydrogen is used as the carrier gas. This enables
the run time to be reduced to increase sample throughput
and eliminates the need for increasingly expensive helium
as world stocks are depleting.
Iso- octane
Ethanol
A PerkinElmer® Clarus® 680 gas chromatograph with a
programmable split/splitless injector (PSS), flame ionization
detection (FID) and Swafer system was used to perform this
analysis. Further details on the Swafer technology may be
found on the PerkinElmer website (www.PerkinElmer.com/
Swafer.com). In this instance, the Swafer S6 configuration was
used with a fused silica restrictor tube on one of the outlet
ports to enable the chromatography on the precolumn to be
directly monitored by a detector. This enables the backflush
point to be easily and accurately established.
Backflush
Point
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
Figure 3. Precolumn chromatography of mixture of ethanol, benzene and
toluene in iso-octane.
To check the separation of the mixture on the complete
system, the restrictor is then disconnected from the FID
detector and the TCEP column connected to this detector,
as shown in Figure 4.
Figure 1. Screen shot taken from the Swafer Utility Software showing the S6
configuration used in this method and gas pressures used for the initial work.
2
100mL/min
Injector at 200°C
octane
- Iso
S-Swafer
60cm x 0.100µm fused
silica restrictor
Ethanol
30m x 0.250mm x 0.25µm
Elite-1
FID at 200°C
Figure 4. System configured to monitor the TCEP column chromatography to
check the final chromatographic separation.
Ethanol
Toluene
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
4.2
4.4
4.6
4.8
The effect of both lowering the inlet pressure and raising
the midpoint pressure to initiate backflushing provides a
cleaner separation between the ethanol and benzene peaks
on the TCEP column chromatography as shown in Figure 7.
This approach also shortens the chromatographic run time –
the full separation is complete within just 4 minutes!
Benzene
0.4
0.4
The next step was to check the chromatography of a gasoline
sample. For these samples, backflushing of the precolumn
was important to keep the bulk out of the gasoline material
off the sensitive TCEP column. To perform the backflush
technique, the inlet pressure was reduced and simultaneously,
the midpoint pressure at the S-Swafer was increased slightly.
Iso - octane
0.2
0.2
Figure 6. Combined precolumn and TCEP column chromatography of
mixture.
The resultant chromatogram is shown in Figure 5. Good
separation of all of the components is evident.
0.0
0.0
Toluene
50m x 0.250mm x 0.4µm
TCEP
GC oven = 100°C isothermal
Benzene
20psig
hydrogen
Precolumn chromatography
30psig
hydrogen
4.0
4.2
4.4
4.6
4.8
Toluene
Benzene
Because the precolumn chromatography is completed before
the TCEP column chromatography starts, the restrictor and
TCEP column can be connected simultaneously to the same
detector. This means that both chromatograms may be
monitored in the same run as shown in Figure 6. This
approach makes It very easy to establish and maintain
the backflush point and it provides a visible record of the
chromatography on the precolumn for QC purposes. This
information is not available from the original ASTM D-3606
method.
Ethanol
Figure 5. Chromatography of mixture eluting from TCEP column.
2.80 2.85 2.90 2.95 3.00 3.05 3.10 3.15 3.20 3.25 3.30 3.35 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 3.95
Figure 7. Chromatography eluting from the TCEP column for a typical
gasoline sample with a timed event to reduce the injector pressure and
increase the midpoint pressure at 1.61 minutes.
The Swafer Utility Software was further used to model
backflush conditions as shown in Figure 8.
3
ASTM® method D-3606 specifies the use of an internal
standard. In the original method this was 2-butanol and in
the updated method for gasoline samples that may contain
ethanol, methyl ethyl ketone (MEK) was specified.
On this system, it was found that MEK eluted after toluene
and so would extend the run time. 2-Butanol eluted between
benzene and toluene as shown in Figure 9 so this compound
was chosen as the internal standard to keep the chromatography to 4 minutes.
The final method for the determination of benzene and
toluene in finished gasoline containing ethanol, as used to
produce the chromatography in Figure 9, is listed in Table 1.
Under these conditions, the chromatography took just 4
minutes to complete. A short equilibration time is added to
stabilize the carrier gas pressures before the start of each
run. The total cycle time for each analysis is just 5.4 minutes
which enables 11 samples to be run each hour or 88 samples
over an 8-hour shift.
Figure 8. Screen shot taken from the Swafer Utility Software showing the
system in backflush mode (note that in practice, a single detector is used).
Ethanol
Toluene
2- Butanol (I.S.)
Benzene
3.25
3.30
3.35
3.40
3.45
3.50
3.55
3.60
3.65
3.70
Figure 9. Chromatography of a typical gasoline sample with 2-butanol added as an internal standard.
4
3.75
3.80
3.85
3.90
3.95
Tolerance to ethanol
Table 1. Full Experimental Conditions.
Gas Chromatograph: PerkinElmer Clarus 680
Oven:
100 °C isothermal for 4 min. 0.5-min
equilibration time
Injector:
Programmable Split/Splitless Injector (PSS)
100 mL/min. split at 200 °C
Quartz Liner Part No. N6121001
A Split/Splitless (capillary) injector may
also be used.
Detector:
Flame Ionization at 200 °C
Air 450 mL/min. hydrogen 45 mL/min.
range x1, attenuation x8
Backflush Device:
S-Swafer in S6 configuration
S-Swafer Kit part no. N6520272
Swafer Kit part no. N6520270 and N9306262
for existing Clarus GCs with PPC
Precolumn:
30 m x 0.25 mm x 0.25 µm Elite-1
Part No. N9316010
Since there were concerns with the original ASTM® D-3606
method when used with gasoline samples containing significant quantities of ethanol, a solution containing a low
concentration of benzene in ethanol was injected. The result
is shown in Figure 10. This clearly shows that this method
has good tolerance to ethanol even when present at ~100%
levels when determining low levels of benzene.
Linearity
ASTM® Method, D-3606 specifies that the system should be
calibrated using 7 standard solutions as listed in Table 2.
Table 2. Calibration Solutions for ASTM® D-3606.
Concentration (%v/v)*
Solution
Benzene
Toluene
1
5
20
Analytical Column: 50 m x 0.25 mm x 0.4 µm TCEP
Part No. NR210048
2
2.5
15
Midpoint Restrictor: Fused silica, 60 cm x 0.100 mm
Part No. N9316601
3
1.25
10
4
0.67
5
5
0.33
2.5
6
0.12
1
7
0.06
0.5
Carrier Gas:
Hydrogen
Carrier Gas Pressure 30 psig for 1.62 min., then 5 psig
Inlet:
programming by timed event until end of run
Midpoint: 20 psig for 1.61 min. then 25 psig by timed
event until end of run
Injection:
0.3 µL by autosampler in fast mode
4 solvent washes following injection
Sample Preparation: 1 mL 2-butanol (I.S.) added to 25-mL
volumetric flask and then gasoline sample
added to bring total volume to 25 mL.
*Prior to addition of internal standard
A set of standard solutions was used to check linearity and
calibrate the system being used for this application work.
Some examples of the chromatography are given in Figure 11.
20
16
Toluene
12
8
3.25
3.30
3.35
3.40
3.45
3.50
3.55
3.60
Toluene = 0.5% v/v
2 - Butanol (I.S.)
Ethanol
24
4
2-Butanol (I.S.)
Benzene = 0.06% v/v
Standard 7
28
3.65
3.70
3.75
3.80
3.85
3.90
3.95
550
500
Toluene = 5% v/v
Standard 4
450
400
300
250
200
150
100
50
2-Butanol (I.S.)
Benzene = 0.67% v/v
350
0
3.25
Benzene
3.35
3.40
3.45
3.50
3.40
3.45
3.50
3.55
3.60
3.65
3.70
3.75
3.80
3.85
3.90
3.95
600
3.55
3.60
3.65
3.70
3.75
3.80
400
300
Butanol (I.S.)
-
500
Figure 10. Chromatogram of standard mixture containing 0.05% v/v benzene
and 0.5% v/v toluene in pure ethanol with added internal standard.
Standard 1
Benzene = 5% v/v
3.30
3.35
3.85
3.90
3.95
Toluene = 20% v/v
800
700
3.25
3.30
900
Figure 11. Example calibration solution chromatograms.
5
Retention time precision
2.5
Retention time precision is a good indicator of a robust
method. Table 4 shows excellent relative standard deviations
obtained for the three components from the 100 gasoline
sample runs.
Response Ra�o
2.0
y = 0.4096x
R² = 0.9995
1.5
Table 4. Retention Time Precision Obtained from 100
Injections of a Single Gasoline Sample.
1.0
0.5
Retention Time
Component
0.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
Concentra�on (% v/v)
Mean (min)
RSD (%)
Benzene
3.478
0.020
2-Butanol (I.S.)
3.563
0.021
Toluene
3.887
0.019
Figure 12. Calibration plot for benzene.
Analyte carry-over
7.0
6.0
y = 0.2935x
R² = 0.9992
Response Ra�o
5.0
4.0
3.0
2.0
1.0
0.0
0.0
5.0
10.0
15.0
20.0
25.0
This method is designed to monitor levels of benzene and
toluene over a wide range of concentrations and so it is
important that cross-contamination between successive
sample injections is minimized. To check the levels of carryover from one injection to the next, Calibration Solution 1
(see Table 2) was run and then immediately followed by
an injection of iso-octane solvent. Figure 14 shows the
chromatograms. The carry-over value of 0.013% from an
injection of toluene at a 20% v/v concentration represents
a potential interference of 0.0026% v/v in the determined
concentration which is far below the 0.5% v/v minimum
calibration level.
Concentra�on (% v/v)
Figure 13. Calibration plot for toluene.
900
Standard 1
800
700
To check the quantitative precision, a gasoline sample was
run 100 times and the standard deviations were calculated
for the quantitative results. These data are shown in Table 3.
For each analyte, the quantitative repeatability requirements
of D-3606 are easily met.
500
Benzene
400
300
200
100
0
3.25
6.70
3.30
3.35
3.40
3.45
3.40
3.45
3.50
6
Std Dev D-3606-07 Requirement
Benzene
0.723
0.003
0.032
Toluene
5.713
0.092
0.191
3.65
3.70
3.75
3.80
3.85
3.60
3.65
3.70
3.75
3.80
3.85
3.90
3.95
6.30
6.20
6.10
6.00
Carry-Over = 0.013%
Component Mean
6.40
Non Detected
Concentration (% v/v)
3.60
Carry- Over = 0.016%
6.50
3.55
Iso-Octane
6.60
Table 3. Quantitative Precision Obtained from 100
Injections of a Single Gasoline Sample.
2-Butanol (I.S.)
600
Toluene
Quantitative precision
5.90
5.80
3.25
3.30
3.35
3.50
3.55
3.90
3.95
Figure 14. Carry-over study showing chromatogram of high concentration
standard mixture followed by injection of solvent.
Conclusions
• The combination of modern capillary columns with the
Swafer technology has taken a mature method and
improved the quality of the data and reduced the run time.
• Baseline separation of ethanol, benzene, toluene and the
2-butanol internal standard has been demonstrated.
• Even though the method relies on significant carrier gas
pressure and flow rate changes, the quantitative and
peak retention time precisions are excellent.
• The chromatographic run time has been reduced from
about 8 minutes for the original method (or 16 minutes
for some revised column sets) to just 4 minutes. The total
cycle time of the chromatographic analysis (including
pressure equilibration) is 5.4 minutes enabling 88 samples
to be analyzed during an 8-hour working shift.
• The method is able to analyze samples with low levels of
benzene in the presence of high levels of ethanol.
PerkinElmer, Inc.
940 Winter Street
Waltham, MA 02451 USA
P: (800) 762-4000 or
(+1) 203-925-4602
www.perkinelmer.com
For a complete listing of our global offices, visit www.perkinelmer.com/ContactUs
Copyright ©2010, PerkinElmer, Inc. All rights reserved. PerkinElmer® is a registered trademark of PerkinElmer, Inc. All other trademarks are the property of their respective owners.
009420_01