An Alternative Gas Chromatography Setting for Geochemical Analysis

Gas chromatography is one of the most important techniques used to characterize gas composition of geochemical samples. This work presents a new gas chromatography setting to characterize gaseous composition using three detectors: a Flame Ionization Detector (FID), a Thermal Conductivity Detector (TCD) and a Flame Photometric Detector (FPD). The new gas chromatography setting developed is able to quantify hydrocarbons gases (C 1 -C 5 ) and non-hydrocarbons (H 2 S, CO, CO 2 , N 2 , O 2 and H 2 ) in only one injection and in the same analysis. This alternative gas chromatography is able to reduce time of analysis and save sample that can be used for other kind of analyses.


Introduction
Gas Chromatography (GC) technique is widely used in geochemical studies [1][2][3]. One of its extensive applications is the determination of gaseous hydrocarbons (C 1 -C 5 ) and non-hydrocarbons gases (N 2 , CO 2 , H 2 , O 2 ) in marine sediments. The knowledge of composition and concentration of these gaseous is used to constrain the origin of gases [4].
Gas chromatography is applied to conventional petroleum exploration to determine the quantity of light hydrocarbon gases (methane, ethane, ethylene, propylene, propane, butane, iso-butane, butylene, and pentane) in marine sediments which are proxies for deeper petroleum reservoirs. Differences in concentration of the gases above may reflect origin of hydrocarbons and their upward migration to seabed [5]. The gas composition in combination with stable isotopic composition (δ 13 C-CH 4 ) can be used to distinguish thermogenic from biogenic gases [6].
Other GC application is for the exploration of unconventional gas reservoir such as gas shale as gas hydrates [7]. Thermal maturity provides an indication of the maximum palaeo temperature reached by a source rock, which determines the type and quantity of the hydrocarbon product from kerogen. Therefore, gas shale is characterized by widespread hydrocarbons gases purity content. Furthermore, it is possible to distingue thermogenic, biogenic or combined biogenic/ thermogenic origin using gas composition and additional geochemical parameters [8,9].
Composition of trapped gas in gas hydrate can also be determined using gas chromatography techniques [10]. Methane is the most common gas in gas hydrate, but others gases like CO 2 and H 2 S also occur [11,12]. The gas composition of hydrates will determine its crystalline structure and stability [13].
This work presents an alternative gas chromatography setting to characterize gaseous composition using an innovative device containing three detectors: (1) Flame Ionization Detector (FID), (2) Thermal Conductivity Detector (TCD) and (3) Flame Photometric Detector (FPD). In this paper we will present the results of standard gases used to methodology development of the new gas chromatography setting which is able to quantify hydrocarbons gases (C 1 -C 5 ) and nonhydrocarbons (H 2 S, CO, CO 2 , N 2 , O 2 and H 2 ) in only one injection and in the same analysis.

Equipment descriptions
The analytical methodology was developed using a Shimadzu Gas Chromatograph Model GC-2014. Helium was used as carrier gas at a 30 ml/min flow rate. This device is equipped with FID detector use to quantify principally light hydrocarbons (C 1 -C 5 ), CO and CO 2 . Hydrogen and air, both with 30 ml/min flow rate, are used as combustion gas in the FID detector at 200°C. The methanation reaction of CO and CO 2 occur using methanator (MTN-1) with excess of hydrogen. The methane produced is detected in the FID. For CO, CH 4 , CO 2 , N 2 , O 2 and H 2 detection it is used a TCD detector at 200°C. The reference gas and the capillary makeup flow rates are 30 ml/min. A FPD detector at 200°C is used to detect H 2 S. Hydrogen and air were used as detector gases at 30 ml/min flow rate.
The gas chromatograph is equipped with a Carboxen capillary column (30 m length, 0.53 mm ID), two Hayesep Q columns (2 m length, 3.0 mm ID) and a Molecular Sieve column (3 m length, 3.0 mm ID). The GC conditions are as follows: starting temperature of the oven of 80°C (hold for 3.5 min) ramp of 8°C min -1 to 170°C (hold 10.2 min).
The injection system is equipped with three different loops. In a single injection, the gas sample is distributed in the loops and to the specific detectors. The FID detector has a 100 µL loop, while the TCD and FPD detectors have loop volumes of 500 and 250 µL, respectively. The properties of the columns and volumes of loops used in the development of the methodology can be seen in table 1.
The gases required for operation of the equipment are helium (99.999% purity), hydrogen (99.999% purity) and nitrogen (99.999% purity). Mixture of helium and methane (C 1 ), ethane, ethylene, propane, propylene, butane, butylenes and pentane (C 2 -C 5 ) is used as standards. The methane concentration is 1,000 ppm (v/v) while C 2 - Other standard gases used are a mixture of helium and CO, CO 2 , O 2 , N 2 and H 2 with 2,500 ppm (v/v) of concentration. Finally, 10 ppm of H 2 S in helium also was used in this work. All gas standards were provided by White Martins Company.

Journal of Chemical Engineering & Process Technology
The gas chromatograph configuration was designed to detect C 1 -C 5 , CO, CO 2 , O 2 , N 2 and H 2 in only one injection. Injection of gases is made using syringes of BD PrecisionGlide ™ coupled of three-way tap rotation and BD PrecisionGlide ™ hypodermic needle. Figure 1shows four rectangles with different colors. Each color represents a possible pathway which the gas sample will be directed until its detection. The load position is shown in figure 1, where the gases are filling the sampling loops.
After injection (A), the gas sample is divided in three loops (B, C and D) with different volumes. The gas sample will fill the loop B with 250 µL, loop C with 500 µL and loop D with 100 µL and will be carried to the FPD, TCD and FID detectors, respectively, by helium. The gases pathway after injection is described below:

Way 1 -Red rectangle
After starting time, the valve 93 allows 250 µL of the sample to be carried until the Haysep Q (column 1). After gases separation the H 2 S is detected in the FPD (Figure 1).

Way 2 -Green rectangle
The valve 93 also allows the sample to the loop with 500 µL and, therefore, the gas helium carries the sample to Q Hayesep column (column 2) and after that to the molecular sieve packed column (column 3). After gas detection in TCD the valve 92 allows a back flush flux to discard the undesirable compounds. These CO, CO 2 , O 2 , N 2 and H 2 will not be detected in TCD (Figure 1).

Way 3 -Blue rectangle
After starting time, the valve 91 allows 100 µL of the sample to be carried to the Carboxen 1006 (column 4). There are two main directions controlled by valve 94: to the FID detector (where C 1 -C 5 will be detected by FID detector) or to the methanator (dashed blue rectangle) (where after methanation, CO and CO 2 will be detected by FID detector).

Results and Discussions
Detection of C 1 -C 5 , CO 2

and CO by FID detector
The oven's temperature was maintained constant during the other optimization. Table 2 shows the gas chromatography temperature program during this work.
Initially, it was injected the standard containing a mixture of hydrocarbon (C 1 -C 5 ) to adjust their signal behavior and retention time. The separation and identification of hydrocarbons is made with the capillary column (Carboxen 1006) and the FID detector. For this purpose, the valve 91 is allows the injection of 100 µL in the column, and the valve 94 avoids deviation of the sample to the methanator. In the beginning, it was chosen an isotherm temperature (100°C) but in   do not have great affinity and cross the column earlier than C 2 -C 5 . The purpose of this column is only to separate the hydrocarbons C 2 -C 5 to the LMW. After the last LMW compound detection in TCD, the valve 92 is closed. This valve will act as a back flush flux in order to discard the retained fraction (C 2 -C 5 ) in column 2 and therefore, only CO, CO 2 , N 2 , O 2 and H 2 will reach the TCD detector. Figure 2C shows the results of LMW gases detection in the TCD: CO 2 (3.0 min) is the first compound that is detected in TCD followed by H 2 (4.2 min), O 2 (5.4 min), N 2 (6.7 min) and CO (9.9 min). Although is possible to detect C 2 -C 5 in TCD it was chosen to discard these compounds because their detection take more than 50. Furthermore, C 2 -C 5 are in low concentration, therefore, it is necessary to use one specific detector for this purpose.

Detection of H 2 S by FPD detector
When the valve 93 is open, 250 µL of the sample is also carried to the Haysep Q (column 1). A standard containing 10 ppm (v/v) of H 2 S is injected. In our first trial, no signal after followed injections was observed. How the H 2 S is a reactive gas, probably it reacted with the tubing material (stainless steel). Therefore, it was changed the tubing and loop to sulfinert material to avoid some possible reaction. After that, it was possible to detect H 2 S how can be viewed on the retention time 17.05 min ( Figure 2D).

Limits of Detections (LODs)
Limits of Detection (LODs) of each gas are shown in Table 3. Methane, carbon dioxide and carbon monoxide can be analyzed using TCD and FID detectors. This is advantage because if the samples have low concentration, it will be necessary to use FID detector due its lower LODs and higher sensitivity. Carbon dioxide and carbon monoxide are determined by FID only due the methanator actuation. Even TCD is known as universal detector, in this gas chromatography there are no this temperature, butane, butylene and pentane had their retention time above 50 min therefore, the time of analysis would stay infeasible. It was chosen a temperature program with two ramps to reduce the retention time for these compounds. The initial temperature was fixed at 35°C for three minutes and after methane (0.9 min), ethylene (2.2 min) and ethane (2.8 min) separation, the temperature is increased to 100°C using a heating rate of 50°C min -1 . The temperature must remain at 100°C for 10 min and during this period, propane and propylene were separated in 7.4 and 7.9 min, respectively. Finally, using a heating rate of 25°C min -1 , the temperature is increased until 250°C, hold for 12 min to separate butylene (12.7 min), butane (13.3 min) and pentane (18.4 min). The chromatogram for C 1 -C 5 is shown in Figure 2A.
After the C 1 -C 5 analysis, CO and CO 2 are analysed with the injection of 100 µL of standard containing 25,000 ppm (v/v) of CO and CO 2 . The separation and identification of these gases is made with a capillary column (Carboxen 1006) and the FID detector, respectively. Carbon dioxide and carbon monoxide CO and CO 2 must be reduced to methane in the methanator before their detection by FID detector. This procedure is achieved by closing the valve 94, allowing the CO and CO 2 passing through the methanator.
The main purpose of this optimization is to analyse hydrocarbon (C 1 -C 5 ), CO 2 and CO in the same run. Methane will be first detect in the FID detector and after its signal integration, CO is the next compound (retention time in 1.7 min). Because ethylene has the retention time (2.2 min) similar to CO (1.7 min), the valve 94 must be kept closed. This valve will be kept closed until CO 2 is detected by the FID detector. This procedure will allow ethylene and ethane with 2.2 min and 2.8 min respectively, get in the methanator because CO 2 has a retention time of 5.3 min. After CO 2 signal integration, the valve 94 will actuate again and C 3 -C 5 will go direct to the FID detector. Figure 2B shows the chromatogram of CO (1.7 min) and CO 2 (6.4 min).

Detection of CO, CO 2 , N 2 , O 2 and H 2 by TCD detector
For detection of CO, CO 2 , N 2 , O 2 and H 2 in TCD detector it was used a standard containing 25,000 ppm (v/v) of CO, CO 2 , N 2 , O 2 and values of LODs for C 2 -C 5 in TCD detector because the configuration was performed to detect these gases only in FID detector. Sulfur dioxide is only detected in the selective FPD detector and the LOD was 0.001 ppm v/v . FID detector is nonspecific for H 2 S compound and there is no value of LOD for H 2 S in TCD detector because our standard gas contained 10 ppm v/v which is insufficient concentration to get some signal in this detector.

Conclusion
The alternative gas chromatograph developed and presented here was able to quantify hydrocarbons (C 1 -C 5 ) and non-hydrocarbons (H 2 S, CO, CO 2 , N 2 , O 2 and H 2 ) gases in only one injection in the same equipment. Our methodology is very useful in situations where gas samples are scarce for multiple injections such as in head space mud samples from marine sediments.
Our work demonstrated that it is possible to detect low gas content (ppm v/v ) and high concentration (% v/v ) using FID detector for light hydrocarbons (C 1 -C 5 ). Samples contain high concentration (% v/v ) of CO 2 and CO can be detected by the TCD detector, whereas samples containing low concentrations (ppm v/v ), need to be directed to the methanator redirected to FID detector.
Other advantage of this device is the H 2 S detection by FPD at the same time with other gases, which normally has to be done in separate equipment. Finally, our new equipment setting will extend the gases characterization reducing the time analysis and saving samples that can be used in other analyses such as stable isotopes. Using this alternative gas chromatography, many types of geochemical samples can be characterized with this device.