CHEMICAL & ENERGY analysis

Speciation of Chlorinated Hydrocarbons

in Reformate using the Agilent
7200 GC/Q-TOF

Solutions for Your Analytical Business

Markets and Applications Programs

ABSTRACT
Authors
The presence of chlorine in petroleum products can be harmful to equipment
Frank David, Bart Tienpont
and refining processes. In addition to total chlorine determination, hyphenated
and Pat Sandra
techniques are needed to identify and quantitate the different chlorinated
Research Institute for Chromatography, species. Speciation of chlorinated hydrocarbons in reformate was performed
Kortrijk, Belgium using the Agilent 7200 GC/Q-TOF high resolution mass spectrometer.
Fabien Chainet, Michel Thomas Using extracted ion chromatograms at accurate mass, excellent selectivity
and Eric Lemaire and sensitivity were achieved, enabling the detection and quantitation
of chlorinated hydrocarbons in the very complex reformate matrix. These
IFP Energies Nouvelles, Solaize, France
compounds were determined at concentrations below 2 pg on-column,
Sofia Nieto corresponding to 0.5 ppm (mg/kg) in the reformate matrix.
Agilent Technologies, Santa Clara, USA INTRODUCTION
The presence of chlorine in the petroleum industry is of considerable
importance because of the potential formation of hydrogen chloride in liquid
or gas streams [1]. This chloride-containing gas can deactivate downstream
catalysts and cause undesired reactions. Chlorine is a poison recovered on
Pd/Al2O3 catalysts during selective hydrogenation [2]. The presence of HCl
in a hydrocarbon matrix leads to the formation of chlorinated hydrocarbons
and can also promote the polymerization of olefins to produce “green” oils,
mainly containing C6-C18 hydrocarbons [3]. Even if the concentration of HCl is
relatively low, it can still interfere with operating processes that use hydrogen
and can also cause corrosion problems in equipment such as pipes, valves
and compressors.
In addition, HCl is considered a hazardous material and must be eliminated before release to the environment [1].
Recently, a review on total chlorine determination was published by Doyle et al. [4]. Chlorine can be measured using
X-ray fluorescence (XRF) with a quantitation limit (LOQ) of 4 mg/kg in petroleum products and derivatives [5], by
ICP-OES [4] and also by ICP/MS [6]. Micro-coulometric methods are currently the most widely used techniques for the
determination of total chlorine, using standard test methods ASTM D4929 [7] or NF EN 14077, with an LOQ of around
1 mg/kg (ppm).
However, these methods determine the total chlorine concentration rather than identifying specific organochlorine
compounds. In the petroleum industry, speciation of chlorine is necessary as these compounds need to be identified
and quantitated, in order to develop systems for trapping these molecules. To our knowledge, there are no speciation
methods for chlorine in petroleum products available in the literature.

Typically, chlorinated organic compounds are analyzed by gas chromatography (GC) and liquid chromatography (LC)
coupled to mass spectrometry or specific detectors such as electron capture detector (ECD), atomic emission detector
(AED) and inductively coupled plasma mass spectrometry (ICP/MS). GC-ECD lacks specificity and the ECD detector
response is strongly dependent on the organic molecule [8, 9]. Chlorobenzene, for instance, gives a very low response in
ECD. GC-AED can also be used, but chlorine detection is not very sensitive and the co-elution of hydrocarbons at very high
levels (3-4 orders of magnitude higher than the chlorine compounds) exceeds the selectivity of the detector [10]. Despite
spectral interferences at m/z 35 and low ionization potential for chlorine, GC-ICP/MS can be performed, as shown by
Peters et al. [11], for the detection of three compounds (dichloromethane, trichloroethane and trichloroethylene) in
pentane with LODs around 2 mg/L. However, GC-ICP/MS configurations are not widely available.

Due to the complexity of reformate samples, containing about 250 hydrocarbon species, and the chlorine concentration at
low (or sub-) ppm level, selective and sensitive techniques must be used. Compound specific analysis can be performed
using two-dimensional GC (heart-cutting) [12] or GC-MS/MS in multi-reaction monitoring (MRM) mode [13]. However,
these techniques can only be applied to selected target compounds that are available as standards and for which
retention time and mass spectra (MRM transitions) can be determined. In organochlorine speciation, the availability
of a screening (non-target) method that allows for the detection and quantitation of chlorinated hydrocarbons (alkanes
and aromatics) in gasoline, naphtha or reformate samples at low ppm levels would offer interesting possibilities for the
petrochemical industry.

In this application note, the use of the Agilent 7200 GC/Q-TOF high resolution mass spectrometer for the analysis
of chlorinated hydrocarbons in reformate is described. Ion extraction at accurate masses is used for detection and
quantitation of the target compounds.
EXPERIMENTAL

Samples: A reformate sample from IFP Energies Nouvelles (Solaize, France) was used as the test matrix. Basic properties
of this reformate are listed in Table 1. A stock solution of 10 chlorinated hydrocarbons was prepared in heptane at a
concentration of 100 mg/kg (ppm). The reference compounds are listed in Table 2. A diluted stock solution was also
prepared in heptane at 10 ppm. Spiked matrix solutions at concentrations between 0.5 and 20 ppm (per chlorinated
compound) were prepared by spiking the reformate test matrix with the stock solution.

GC/Q-TOF instrument:

Analyses were performed on an Agilent 7890B GC equipped with a split/splitless inlet. Injection was performed using
an Agilent 7693 ALS. The GC was coupled to an Agilent 7200 Q-TOF mass spectrometer.

Analytical conditions:

0.2 µL of sample was injected in split mode (1/50) at 250°C. Separation was performed on a 100 m x 0.25 mm ID x 0.5
µm HP-PONA column (19091Z-530). The carrier gas was helium at 1.5 mL/min constant flow. The oven was programmed
from 35°C (5 min) at 4°C/min to 250°C (10 min). The transfer line was set at 280°C.

The 7200 Q-TOF was operated in MS mode using electron impact (EI) ionization. The source temperature was 230°C. The
scan rate was 5 Hz in HR (high resolution) mode and the mass range was 50 to 300 amu. No internal mass referencing
was used, but the instrument was mass calibrated before each run using the keyword command (MassCal) in the
sequence.

2
Results and discussion

The total ion chromatogram obtained for the analysis of the diluted calibration stock (10 ppm) is shown in Figure 1.
Taking into account the injection volume (0.2 µL) and the split ratio (1/50), the injected amount corresponds to 40 pg
per compound on-column. Seven compounds were easily detected. However, compounds 2, 3 and 4 were masked by
the heptane solvent peak. For the detected compounds, accurate mass spectra were obtained and searched against the
standard NIST library. The chloroalkanes from chloropentane to chlorononane, all have the ion at m/z 91.031 in common.

Fig 1: Base Peak Chromatogram of the analysis of a test solution containing chlorinated
hydrocarbons in heptane (10 ng/µL concentration, 40 pg on-column).

Using the formula generator option, this fragment ion was correctly identified as [C4H8Cl]+, as illustrated by the high
resolution mass spectrum of chloroheptane (compound 7) in Figure 2. At m/z 93.028 the chlorine isotope peak is
detected. For these ions, the mass error was found to be lower than 4 ppm.

Fig. 2: High resolution accurate mass spectrum of peak 7 (1-chloroheptane)

Using extracted ion chromatograms at accurate mass, very high selectivity can be obtained. This is illustrated in Figure
3. In Figure 3A, the total ion chromatograms are compared for a reformate sample (top) and the same reformate sample
spiked at 1 ppm level with the chloroalkanes (bottom). These profiles show the high complexity of the reformate
sample, mainly containing aromatic hydrocarbons (toluene, xylenes, C3-, C4-, C5- benzenes) and paraffins. If the ion
chromatograms are extracted using a window of +/- 0.5 amu, as is done on low resolution instruments such as a single
quadrupole or a triple quadrupole instruments, the chloroalkane compounds cannot be selectively detected in the spiked
sample.
3
Fig. 3A: Comparison of TIC obtained for reformate sample (top) and reformate sample spiked at
1 ppm level with chlorinated hydrocarbons (bottom).

Fig. 3B: Comparison of EIC obtained for reformate sample (top) and reformate sample spiked
at 1 ppm level with chlorinated hydrocarbons (bottom). EIC performed at unit mass resolution
(91.0309 +/- 0.5 amu)

Fig. 3C: Comparison of EIC obtained for reformate sample (top) and reformate sample spiked
at 1 ppm level with chlorinated hydrocarbons (bottom). EIC performed at high mass resolution
(91.0309 +/- 20 ppm)
4
This is illustrated in Figure 3B, showing the EIC at m/z 91.0309 +/- 0.5 amu. No difference can be observed between the
sample (top) and the spiked sample (bottom). The main peaks detected in these EICs are the aromatic hydrocarbons that
give the well know tropylium ion at m/z 91 in their mass spectra. This fragment corresponds to [C7H7]+ and the exact
mass is 91.0542. The first 4 peaks correspond to toluene (19.5 min), ethylbenzene, m- and p- xylene and o-xylene (24 - 26
min). These compounds all have the tropylium ion as the most abundant ion in their mass spectra and low resolution MS
is unable to differentiate the tropylium ion from the [C4H8Cl]+ ion. If ions are extracted at a mass accuracy of +/- 20 ppm,
the target compounds can be selectively detected, as illustrated in Figure 3C. In the unspiked sample, no chloroalkanes
were detected (top chromatogram), while compounds 6, 7, 9 and 10 were detected in the spiked sample.

The EIC signals at accurate mass for a standard (10 ppm of compounds in heptane, corresponding to 40 pg on-column)
and for a spiked matrix sample (10 ppm compounds in matrix) are compared in Figure 4. The peak areas are quite similar,
indicating that no quenching or ion suppression occurs in the matrix. Only slight differences are noted in retention time
and peak width, caused by matrix effects on the chromatography (solvent effects).

Using this method, all chlorinated compounds were detected at low ppm or even sub-ppm levels in the matrix (low
ppb levels on column). To take maximum advantage of the selectivity, the ions chosen for EIC extraction all contained
chlorine (see Table 2). Even though in some cases chlorine-containing ions were of relatively low abundance in a
compound spectrum, it allowed efficient detection of the compounds in the very complex matrix. This is illustrated in
Figure 5 for chlorobenzene using the ion at 112.007 and in Figure 6 for benzylchloride using the ion at m/z 126.023. Ion
chromatograms were extracted at +/- 20 ppm m/z window. The high resolution mass spectra are shown in these figures
and the mass accuracy was calculated using the formula generation option. Chlorobenzene showed a mass error of -3.03
ppm and benzylchloride a mass error of -3.17 ppm. It should be emphasized that these mass errors were obtained in the
complex reformate matrix.

Fig. 4: Comparison of EIC obtained for a 10 ppm standard solution (top) and reformate sample
spiked at 10 ppm level with chlorinated hydrocarbons (bottom). EIC performed at high mass
resolution (91.0309 +/- 20 ppm)

5
Fig. 5: Detection of chlorobenzene in reformate (spiked at 10 ppm level) - EIC @112.0074 +/- 20 ppm
Mass spectrum is given in insert

Fig. 6: Detection of benzylchloride in reformate (spiked at 10 ppm level) - EIC @126.0231 +/- 20 ppm
Mass spectrum is given in insert

6
 Properties Methods Values
Density (g/cm3) NF EN ISO 12185 0.8316
Boiling range (◦C) ASTM D2887 82-197
RON 101
Paraffins (%wt) 5.0
Isoparaffins (%wt) 14.7
Naphthenes (%wt) IFPEN 9302 1.1
Aromatics (%wt) 78.4
Olefins (%wt) 0.8
Chlorine (mg/kg) NF EN 14077 <2
Table 1

peak tR (min) solute formula MW exact mass fragment

1 9.949 2-chloro-2-methylpropane C 4H9Cl 92 77.0158 C 3H6Cl
2 16.463 1-chloro-2-methylbutane C 5H11Cl 106 62.9996 C 2H4Cl
3 16.747 2-chloro-pentane C 5H11Cl 106 62.9996 C 2H4Cl
4 18.784 1-chloropentane C 5H11Cl 106 62.9996 C 2H4Cl
5 23.390 chlorobenzene C6H5Cl 112 112.0074 C6H5Cl
6 23.955 1-chlorohexane C6H13Cl 120 91.0309 C 4H8Cl
7 28.834 1-chloroheptane C7H15Cl 134 91.0309 C 4H8Cl
8 30.921 benzylchloride C7H7Cl 126 126.0231 C7H7Cl
9 33.334 1-chlorooctane C 8H17Cl 148 91.0309 C 4H8Cl
10 37.484 1-chlorononane C 9H19Cl 162 91.0309 C 4H8Cl
Table 2

7
CONCLUSION

Sub-ppm chlorinated hydrocarbons can be analyzed in reformate samples using GC/Q-TOF high resolution MS analysis.
The high selectivity offered by accurate mass ion extraction allows the target compounds to be detected and quantified
in this extremely complex matrix. Sensitivity is better than 2 pg on column, corresponding to 0.5 ppm (mg/kg) in the
reformate matrix.

References

[1] O.K. Karan, M.A. Ay, K. Karhaman, A. Selmen, “Combating [9] S. Pedersen-Bjergaard, S.I. Semb, J. Vedde, E.M.
green oil formation in a CCR reformer”, PTQ 2013, Q3, 29. Brevik, T. Greibrokk, “Comparison of GC-ECD, GC-MS
and GC-AED for the determination of polychlorinated
[2] B. Didillon, J. Cosyns, C. Cameron, D. Uzio, P. Sarrazin, J.P.
biphenyls in highly contaminated marine sediments”,
Boitiaux, “Industrial evaluation of selective hydrogenation
Chromatographia, 1996, 43, 44.
catalyst poisoning”, Stud. Surf. Sci. Catal., 1997, 111, 447.
[10] C. Brede, S. Pedersen-Bjergaard, “State-of-the art of
[3] J. Cosyns, O. Ducreux, Q. Debuisschert, and F. Lepeltier,
selective detection and identification of I-, Br-, Chlorine-,
Removal of chlorinated compounds in hydrocarbon cuts, (6-
and F-containing compounds in gas chromatography and
6-2012)
liquid chromatography”, J. Chromatogr. A, 2004, 1050, 45.
[4] A. Doyle, A. Saavedra, M.L.B. Tristão, L.A.N. Mendes,
[11] G.R. Peters, D. Beauchemin, “Versatile interface for
R.Q. Aucélio, “Spectrometric methods for the determination
gas chromatographic detection or solution nebulization
of chlorine in crude oil and petroleum derivatives — A
analysis by inductively coupled plasma mass spectrometry:
review”, Spectrochim. Acta B, 2013, 86, 102.
preliminary results”, J. Anal. At. Spectrom., 1992, 7, 965.
[5] ASTM D6443, Standard Test Method for Determination
[12] F. Chainet, C.P. Lienemann, J. Ponthus, M.
of Calcium, Chlorine, Copper, Magnesium, Phosphorus,
Courtiade, O.F.X. Donard, “Development of heart-cutting
Sulfur, and Zinc in Unused Lubricating Oils and Additives
multidimensional gas chromatography coupled to time of
by Wavelength Dispersive X-ray Fluorescence Spectrometry
flight mass spectrometry for silicon speciation at trace
(Mathematical Correction Procedure) (2004).
levels in gasoline samples”, J. Chromatogr. A, 2012, 1264,
[6] F.G. Antes, M.F.P. dos Santos, R.C.L. Guimarães, J.N.G. 80.
Paniz, E.M.M. Flores, V.L. Dressler, “Heavy crude oil
[13] N.S. Chary, S. Herrera, M.J. Gómez, A.R. Fernández-
sample preparation by pyrohydrolysis for further chlorine
Alba, “Parts per trillion level determination of endocrine-
determination”, Anal. Methods, 2011, 3, 288.
disrupting chlorinated compounds in river water and
[7] ASTM D4929, Standard Test Methods for Determination wastewater effluent by stir-bar-sorptive extraction
of Organic Chloride Content in Crude Oil (2007). followed by gas chromatography–triple quadrupole mass
spectrometry”, Anal. Bioanal. Chem., 2012, 404, 1993.
[8] A. Schwarz, K.G. Heumann, “Two-dimensional on-line
detection of brominated and iodinated volatile organic
compounds by ECD and ICP-MS after GC separation”, Anal.
Bioanal. Chem., 2002, 374, 212.

Agilent Products are for Research Use Only.

Not for use in diagnostic procedures.
Information, descriptions and specifications in this
publication are subject to change without notice.

© Agilent Technologies, Inc. 2015

Solutions for your analytical business Published in USA, March 26, 2015
Markets & Applications Programs 5991-5684EN

www.solutions-to-win.com