Evaluating the Performance of an Electronic Nose for Dioxin Monitoring

The need to measure dioxin/furans in real or near real time in stack gases has been identified by the United States Department of Energy (DOE) in order to meet the Environmental Protection Agency’s (EPA) Maximum Achievable Control Technology (MACT) Standard.

A mixture of various dioxin/furans was sent to Electronic Sensor Technology for evaluation using a GC/SAW, as part of a preliminary evaluation of promising technologies such as fast chromatography. This article presents a description of the GC/SAW measurement system, calibration procedures, sample analysis procedures, values obtained, and minimum detection levels.

Evaluation Protocol

The evaluation sample contained a mixture of furans and dioxins at 1-10 nanogram per microliter levels in nonane.

Evaluation was carried out by injecting microliter quantities of undiluted sample, 1:50 diluted sample, and a 1:1000 diluted sample into an opentubular desorber fixed to the inlet of a GC/SAW vapor analyzer. For every single sample, the concentration of individual furans and dioxins were recorded and referenced to calibration standards of similar concentration.

Description of GC/SAW Technology

Electronic Sensor Technology manufactures fast chromatographs in two different models and both use surface acoustic wave (SAW) integrating detectors. The model 7100 features a handheld GC and sampling preconcentrator fixed to a support case with the help of a 6 foot umbilical cable. The second, a model 7100, is designed for portable or laboratory use and the chromatograph and vapor preconcentrator are incorporated into a benchtop case.

Both systems use an RS232 connection to interface with a Pentium laptop running proprietary control software. A complete range of post processing analysis and communications software is provided by links to features inherent in Microsoft Office and Windows 95.

These instruments can be configured to rapidly analyze a wide variety of semi-volatile and volatile compounds using the patented integrating SAW detector. The GC/SAW is can speciate and quantify furans and dioxins at the picogram level in less than 1 minute chromatogram using a temperature ramped DB-6 column.

For the assessment of the MSE samples, a 4100 system was used together with a model 3100 open-tubular desorber fixed to the inlet of the system. This accessory thermally vaporizes liquid injections, and the GC/SAW measurement system samples these vapors.

Handheld Model 4100 GC/SAW vapor Analyzer

Figure 1. Handheld Model 4100 GC/SAW vapor Analyzer.

Benchtop Model 7100 GC/SAW Vapor Analyzer

Figure 2. Benchtop Model 7100 GC/SAW Vapor Analyzer.

Block diagram of GC/SAW vapor measurement system

Figure 3. Block diagram of GC/SAW vapor measurement system.

Sample Preparation and Injection

Stock solution was obtained from MSE and two dilutions were performed in hexane. Injecting 20 µliters of stock solution into 1 milliliter of hexane resulted in a 50 to 1 dilution. A 1000 to 1 dilution was prepared by injecting 1 µliter of stock into 1 milliliter of hexane.

Attachment of Open-Tubular sample desorber attached in inlet of GC/SAW Vapor Analyzer

Figure 4. Attachment of Open-Tubular sample desorber attached in inlet of GC/SAW Vapor Analyzer.

A 10 µliter glass syringe was used to inject all samples and calibration standards. Sample injection and measurement was performed in two steps:

Step 1

1-10 µliters of sample is injected into middle of glass wool wick within a six inch long desorbtion tube attached to the inlet of the GC/SAW vapor analyzer.

Step 2

A desorbtion tube heater (280 °C) is placed over the glass desorbtion tube and vapor sampling (measurement cycle) by the GC/SAW is initiated by the operator.

The remainder of the measurement process was automatic and no further operator actions were required other than to annotate notes which identified the actions being taken or other applicable sample identification information.

Calibration Standards

Two calibration standards were procured from AccuStandard Inc. (25 Science Park, New Haven CT 06511). Each kit contained five furans (M8280B) and five dioxins (M8280A) as required by EPA 8280 Method.

The concentration of each analyte present in the mixture was 5.0 nanograms per µliter of toluene. A 10-to-1 dilution was used as a calibration level of 0.5 ng/µliter. The full analyte specifications as well as their TEQ rating are provided in the following table.

Analyte CAS No. TEQ*
2,3,7,8-tetrachlorodibenzo-p-dioxin 51207-31-9 1.00
1,2,3,7,8-Pentachlorodibenzo-p-dioxin 40321-76-4 0.50
1,2,3,4,7,8-Hexachlorodibenzo-p-dioxin 39227-28-6 0.10
1,2,3,4,6,7,8-Heptachlorodibenzo-p-dioxin 35822-46-9 0.01
Octachlorodibenzodioxin 3268-87-9 0.001
2,3,7,8-Tetrachlorodibenzofuran 1746-01-6 0.1
1,2,3,7,8-Pentachlorodibenzofuran 40321-76-4 0.05
1,2,3,4,7,8-Hexachlorodibenzofuran 55684-94-1 0.10
1,2,3,4,6,7,8-Heptachlorodibenzofuran 35822-46-9 0.01
Octachlorodibenzofuran 39001-01-0 0.001

Figure 5. Analyte Standards Used in Sample Evaluation.

No other standards were available for comparison with the MSE samples, and hence in this study, any quantification of isomers lower than tetra is based not upon a calibration standard but upon an estimated response factor.

Selection of GC Method

The GC/SAW vapor analyzer can perform dioxin analysis and quantification in less than 1 minute chromatogram as well as at slower speeds such as 20, 50, or more seconds. There is a trade-off in resolving power with better resolution being obtained at longer and slower chromatograms, as shown in Figure 5.

Resolution vs speed displayed for 18 °C /sec, 7 °C /sec, and 3 °C/sec column ramping rate

Figure 6. Resolution vs speed displayed for 18 °C /sec, 7 °C /sec, and 3 °C/sec column ramping rates.

A 20 second chromatogram was achieved with a linear increase of column temperature from 60 °C to 200 °C within 20 seconds for quantification of the MSE sample. The complete GC method was constructed using a graphical method as shown in Figure 6.

The GC method steps are developed by dragging placeholders from the vertical toolbar into a horizontal line at the bottom of the dialog screen of Figure 6. Each placeholder corresponds to an action or step with parameters set by the operator. This method commences with a 30 second sample (preconcentrate) time, move valve to inject position, inject sample, ramping of the column temperature, and taking of data for 20 seconds after the injection.

At the end of the method is a 15 second bake cycle to ‘clean’ the crystal detector that is activated and the column temperature returns to 60 °C.

GC Method dialog screen showing method used to evaluate MSE samples

Figure 7. GC Method dialog screen showing method used to evaluate MSE samples.

Analysis Time Requirements

In automatic mode, it is essential for each analysis to contain the following basic steps with their minimum values. The values used for the MSE samples are shown for comparison in the following table.

Minimum (Sec) MSE Sample (Sec)
Inject Sample into Desorber 2 5
Preconcentrate Vapor Sample 15 30
GC Analyze 10 20
Recovery of Column & Detector 15 30
Total Cycle Time 42 85

Calibration Procedures

Chromatogram of furan standards after entry of proper response factors and retention times into peak identification file

Figure 8. Chromatogram of furan standards after entry of proper response factors and retention times into peak identification file.

Instrument calibration involved injection of standards of known concentration. A response factor specific to each analyte was produced by division of SAW detector ‘counts’ by the concentration.

The response factor (Hz/pg), retention time, peak name and percentage variation allowed in retention time (Percent spread) were entered into a calibration table and this completed calibration (Single point). Other available features within the software include multiple point calibration and interpolation. Checking of proper calibration was done on a regular basis by injecting furan or dioxin mixtures of known concentration.

Operator entry of retention time windows, peak labels, and response factors completes system software calibration

Figure 9. Operator entry of retention time windows, peak labels, and response factors completes system software calibration.

Minimum Detection Limit

For instruments’ detection limits, the GC/SAW is determined by signal to noise, and the noise or detected peak amplitudes acquired with a blank injection of pure hexane into the GC/SAW are specified to be less than 1 picogram.

A 3 picogram minimum detection level is obtained when the system is operated at a signal to noise ratio of 3, and a minimum detection level of 10 picograms would be obtained when the system is operated at a higher signal-to-noise ratio of 10.

Blank injection chromatogram of 10 uliters of hexane compared with 2 uliter dioxin standard (0.5 ng/uliter)

Figure 10. Blank injection chromatogram of 10 uliters of hexane compared with 2 uliter dioxin standard (0.5 ng/uliter).

Evaluation of method detection limits was done by multiplying the standard deviation of seven replicate measurements by 3.14. Method detection limits differed between 10 and 30 picograms using this method with 10 picogram injections. RSD values for manual injections were typically 20% or less.

Quality Control/Assurance Procedures

Electronic Sensor Technology utilizes ISO9000 procedures throughout the manufacture and testing of all GC/SAW instruments. The company also maintains an on-site calibration laboratory, where EPA quality assurance and quality control for all performance tests are implemented.

The laboratory director logged and maintained samples obtained from MSE. Additionally, the laboratory operators controlled the quality of calibration standards throughout the testing of the MSE samples. All GC data taken on the MSE samples was logged and archived on the company server, and each data record was labeled and time-stamped based on the laboratory’s Quality Assurance procedures.

Evaluation Results – Undiluted MSE Sample

An average of three 1 µliter injection-measurements was used to arrive at the following analyte concentrations:

Chromatogram from 1 uliter injection of undiluted MSE sample

Figure 11. Chromatogram from 1 uliter injection of undiluted MSE sample.

.
Tri-chloro furan/dioxin 308 pg/µliter Uncalibrated estimate
2378 TetraFuran 0 Unresolved from group
2378 TetraDioxin 478.7 pg/µliter At least one other Tetra present
12378 PentaFuran 478.7 pg/µliter Good retention time match
12378 PentaDioxin 598.7 pg/µliter Poor retention time match
123478 HexaFuran 549.3 pg/µliter Good retention time match
123478 HexaDioxin 429.0 pg/µliter Good retention time match
1234678 HeptaFuran 681.7 pg/µliter Good retention time match
1234678 HeptaDioxin 1133.3 pg/µliter Good retention time match
OctaFuran/Dioxin 2200 pg/µliter Octa Furan & dioxin not resolved

Evaluation of 50 to 1 MSE Sample

An average of three 5 µliter injection-measurements was employed to arrive at the following analyte concentrations:

Injection of 10 µliter of MSE sample diluted 50 to 1. Inset table shows measured amounts for each analyte

Figure 12. Injection of 10 µliter of MSE sample diluted 50 to 1. Inset table shows measured amounts for each analyte.

.
Tri-chloro furan/dioxin 9.44 pg/µliter Uncalibrated estimate
2378 TetraFuran 0 Unresolved from group
2378 TetraDioxin 22.8 pg/µliter At least one other Tetra present
12378 PentaFuran 17.8 pg/µliter Good retention time match
12378 PentaDioxin 29.8 pg/µliter Poor retention time match
123478 HexaFuran 22.7 pg/µliter Good retention time match
123478 HexaDioxin 19 pg/µliter Good retention time match
1234678 HeptaFuran 24.3 pg/µliter Good retention time match
1234678 HeptaDioxin 33.5 pg/µliter Good retention time match
OctaFuran/Dioxin 54.9 pg/µliter Octa Furan & dioxin not resolved

Evaluation of 1000 to 1 MSE Sample

An average of three 10 µliter injection-measurements was used to arrive at the following analyte concentrations:

Injection of 10 µliter of MSE sample diluted 1000 to 1. Inset table records measured amounts for each chromatogram

Figure 13. Injection of 10 µliter of MSE sample diluted 1000 to 1. Inset table records measured amounts for each chromatogram.

.
Tri-chloro furan/dioxin 0.732 pg/µliter Uncalibrated estimate
2378 TetraFuran 0 Unresolved from group
2378 TetraDioxin 4.31 pg/µliter At least one other Tetra present
12378 PentaFuran 2.56 pg/µliter Good retention time match
12378 PentaDioxin 4.37 pg/µliter Poor retention time match
123478 HexaFuran 3.17 pg/µliter Good retention time match
123478 HexaDioxin 2.93 pg/µliter Good retention time match
1234678 HeptaFuran 5.67 pg/µliter Good retention time match
1234678 HeptaDioxin 4.48 pg/µliter Good retention time match
OctaFuran/Dioxin 6.78 pg/µliter Octa Furan & dioxin not resolved

Peak Identification

It is possible to identify individual furan and dioxin isomers by retention time matching with known standards. An MSE sample chromatogram with overlays (in red) of chromatograms for 5 furan and 5 dioxin calibration standards are shown below.

This figure shows expanded portions of a 50 second duration chromatogram. It is also possible to see good retention time identifications for 12378 Penta Furan, 123478 Hexa Furan, and 1234678 Hepta Furan. Similarly good identifications for 1234678 Hepta Dioxin and 123478 Hexa Dioxin are also obvious.

The peak containing tetra-isomers comprises of two major peaks, which are not well resolved. The leading edge shoulder is considered to be a good match with the retention time of 2378 furan, while the peak detected by the software peak detector is not a good match for 2378 dioxin. Based upon this longer chromatogram, it is estimated that there is about 250 pg/µliter of 2368 Furan present in the earlier identified 478.7 pg/µliter 2378 dioxin peak (undiluted sample). The remainder of this peak is not 2378 dioxin but more likely contains other isomers such as 1368 or 1234 tetra dioxin.

Additionally, the software reported 12378 Penta dioxin (598.7 pg/µliter) in the undiluted sample and this is a misidentification since the retention time does not match the standard. The actual isomer could be 12478 or 12347 Penta dioxin.

Comparison of 50 second chromatogram results with furan and dioxin standard

Figure 14. Comparison of 50 second chromatogram results with furan and dioxin standards

Summary of Results

To conclude, 2378 dioxin is 0 although it may be of a much lower concentration than other tetra dioxins that were detected. In addition, 12378 Penta-dioxin is also listed as zero because the Penta-dioxin peak detected failed to match the retention time of standards.

Summary of Results

.
Tri-chloro furan/dioxin 308 pg/µliter Uncalibrated estimate
2,3,7,8-tetrachlorodibenzo-p-dioxin 200 pg/µliter Uncalibrated estimate
2378 TetraFuran 0 Unresolved estimate
2378 TetraDioxin 478.7 pg/µliter Other Tetra dioxins present
12378 PentaFuran 0 Good retention time match
12378 PentaDioxin 549.3 pg/µliter Other Penta dioxins present
123478 HexaFuran 429.0 pg/µliter Good retention time match
123478 HexaDioxin 681.7 pg/µliter Good retention time match
1234678 HeptaFuran 1133.3 pg/µliter Good retention time match
1234678 HeptaDioxin 2200 pg/µliter Octa Furan & dioxin not Resolved

This information has been sourced, reviewed and adapted from materials provided by Electronic Sensor Technology.

For more information on this source, please visit Electronic Sensor Technology.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit