Rapid and Automated Identification of Components in Petroleum Based Samples

Scott J Campbell1

1SpectralWorks Ltd, The Heath Business & Technical Park, Runcorn, UK

First Published: Gulf Coast Conference 2005


Analytical science has changed over recent years, as instrument technology has improved it has become possible to detect lower and lower levels of compounds and separate more and more complex mixtures. This has lead to a shift away from the analysis of a few target analytes in a sample to the desire to identify all the analytes in a sample. This approach however moves the bottle neck from the acquisition to data processing. In this poster we show the analysis of a petroleum based sample using a new easy to use and intuitive vendor independent mass spectrometry data processing application to enable the rapid and automated detection of components within our sample.


Headspace analysis was carried out on a sample of water that had been contaminated with gasoline. A 1ml sample of the contaminated water was placed in a 10ml headspace vial. The vial was placed on a LEAP Combi PAL and a headspace analysis was carried out. The sample was injected into a Thermo Trace DSQ – GC-MS system under the following conditions:

Oven Method: 40oC (5.00 min), Ramp 20.0 (deg/min), 320oC (5.00 min), Split 60 (ml/min), Injector Temperature 150oC, Constant Flow 1.00 (ml/min), using a 25m 0.25mm id RTX 5 Column DSQ: 8 Scans per second over the mass range 50 to 650 in EI mode.

Typical TIC of Gasoline Sample

Figure 1. Typical TIC of Gasoline Sample

Found Components from AnalyzerPro

Figure 2. Found Components from AnalyzerPro

Subsection of TIC Showing Found Components

Figure 3. Subsection of TIC Showing Found Components

Table 1. Textual Report from AnalyzerPro

Textual Report from AnalyzerPro
Spectrum from Component 5

Figure 4a. Spectrum from Component

Spectrum from Component 6

Figure 4b. Spectrum from Component 6

Spectrum from Component 7

Figure 4c. Spectrum from Component 7

Processing Method Editor

Figure 5. Processing Method Editor

Results and Discussion

The results from the analysis yielded a complex chromatogram as shown in Figure 1. Typically, existing software packages either fail to find components or report too many false positive with lead to complications if the results are to be further processed using statistical techniques such as principle component analysis (PCA). Manual analysis isn’t an alternative as it is tedious and extremely time consuming to fully extract all the information within the sample. Figure 2 shows the components that were found in the Total Ion Chromatogram (TIC) using AnalyzerPro. The found components are shown with blue markers and the total processing time to generate the results took less than a minute. If we examine the subsection of the TIC as shown in Figure 3, we can see the presence of co-eluting peaks which could easily be missed during a manual examination of the data. The analysis using the novel spectral refinement algorithms used in AnalyzerPro to easily differentiate the three component in the cluster. The spectra for the three components found in the subsection of the TIC are shown in Figures 4a-c. The processing parameters are set in a single, easy to use method editor and is shown in Figure 5 with the method used to process our sample. Additionally, our found components could have been searched against the NIST Library or a user Target Library for identification. A subsection of the textural report that is generate by AnalyzerPro is shown in Table 1.


AnalyzerPro provides a rapid and intuitive method for analysis of complex samples.