High Throughput Screening and Characterization Methods of Jordanian Oil Shale as a Case Study

Oil shale is an important possible solution to the problem of energy in Jordan. To explore the technical and the economic feasibility of oil shale deposits, numerous samples are analyzed using the standard Fischer Assay (FA) method. However, it would be useful to develop faster, cheaper, and reliable methods for determining the oil content of oil shale. Therefore, the aim of this work was to propose and investigate rapid analytical techniques for the screening of oil shale deposits and to correlate them with the FA method. The Omari deposit located east of Jordan was selected as a case study for analysis using thermogravimetric analysis (TGA) coupled with Fourier-transform infrared (FTIR), differential scanning calorimetry (DSC), elemental analysis, X-ray fluorescence (XRF), X-ray diffraction (XRD), scanning electron microscopy (SEM), and energy-dispersive X-ray (EDX) analysis. Results obtained from the TGA method were linearly correlated with FA with high regression factor (R2 = 0.99); a quadratic correlation (R2 = 0.98) was maintained between the FA and the elemental hydrogen mass content, and a quadratic correlation (R2 = 0.97) was found between the FA and the aliphatic hydrocarbons (FTIR peak at 2927 cm−1) produced in the pyrolysis zone. Although other techniques were less correlated, further investigation might lead to better results. Subsequently, these correlated techniques can be a practical alternative to the conventional FA method when, in particular, specific correlation is made for each deposit.


Introduction
Jordan is a country in the Middle East region that suffers from a lack of conventional fossil energy sources. In 2017, the country imported 94% of its energy needs, equivalent to 8.5% of its gross domestic product (GDP) [1]. To increase energy security, the kingdom has been encouraging investments in renewable energy-mainly solar and wind-and alternative energy that is principally oil shale. However, the economic feasibility of these projects is affected by crude oil prices, technical challenges, and environmental considerations.
Oil shale can be defined as sedimentary rocks including organic matter disseminated in a mineral matrix. The organic matter is combustible hydrocarbons, predominantly kerogen, which is insoluble in common organic solvents [2,3].
The kingdom is ranked the sixth country in the world in oil shale resources, with more than 70 billion tons of proven reserves distributed over 60% of Jordan's territory [4]. Recent scientific reports in power generation as well as in oil production [8].

Materials and Methods
Crushed and sieved oil shale samples were collected from the Omari deposit, which is located southeast of Amman and nearly reaches the Iraqi borders, as shown in Figure 1. Moreover, the oil yield was determined using the FA method according to standard test method ASTM D3094-90. In this method, 100 g of a dried sample of oil shale was ground into less than 2.38 mm particles and heated in a small aluminum oven (retort) to 500 • C at a rate of 12 • C/min and held at the final temperature for 40 min. The resulting vapors of oil, gas, and water were passed through a condenser and cooled with ice water into a graduated centrifuge tube to separate the oil from the water and measure the oil yield [27].
Thermal properties of the samples were evaluated using thermoanalyzer Jupiter STA 449 F5 (Netzsch, Germany) coupled with FTIR spectrometer (Vertex 70V Bruker Optik, Germany). The different samples obtained from the deposit were placed in vials and given a number from OS1 to OS6. Three samples from each vial were thermally analyzed. We assumed that these homogeneous and representative samples could be considered as an identical sample where the standard deviation of the thermal analysis is calculated. Thermogravimetric analysis (TGA), derivative thermogravimetry (DTG), and differential thermal analysis (DTA) were accomplished over a temperature range of 25-1100 • C under an ambient atmosphere of nitrogen with a heating rate of 20 • C/min. TGA and DTG were used for the thermal stability characterization of the investigated samples. DTA gives information about the changes in the energy of the materials (specific heat capacity and enthalpy). The gas products from TGA were simultaneously analyzed on the FTIR spectrometer with a resolution of 4 cm −1 and a number of 16 scans/0.3 • C by the implementation of Gram-Shmidt diagram.
Elemental analysis was done on Vario MACRO CHN ELEMENTAR Analysensysteme GmbH. Furthermore, the chemical composition of major oxides in the oil shale samples was analyzed by XRF using the sequential wavelength dispersive X-ray fluorescence spectrometer (Bruker S4 Pioneer, Germany). The identification of the structure of the crystalline solid shales and their mineral composition was examined by using the D4 ENDEAVOR X-ray Diffractometer from Bruker. The method of peak area against calibrated curves was used.
The Quantax 200 with XFlash 4010 detector (Bruker AXS machine, Czech Republic) was used for scanning electron microscopy (SEM) imaging of samples. SEM imaging was performed with a secondary electrons (SE) detector under high vacuum <6 × 10 −4 Pa. Before measurements, oil shale samples were sputtered with a nanolayer (10 nm) of gold for morphology determination. For EDX measurements, the microscope 1430 VP LEO Electron Microscopy Ltd (England) was used (non-sputter samples were tested).

Results and Discussion
The absence of an ideal method to analyze oil shale has led to the usage of a variety of methods to overcome various limitations of the popular method(s) of analysis. The results of these analyses are discussed here.

Fischer Assay
Six oil shale samples labeled from OS1 to OS6 were obtained from the Omari deposit covering a wide range of oil content in Jordanian oil shale, which has an average content of about 10 wt.% [7]. The oil yield was determined using the standard FA method. Table 1 shows that the oil content of the tested samples varied between 2.9 to 12.7 wt.%, and the oil density varied between 0.951 to 0.967 g/mL. Furthermore, it was noticeable that the shale oil content was inversely proportional to its density, which could be related to the lower density of oil compared with the inorganic matter.

Thermogravimetric Analysis
TGA is a rapid technique widely used to record weight loss of oil shale as a function of temperature. DTA is a complementary technique usually used in the same TGA instrument to represent the temperature difference developed between the oil shale sample and a reference. Both techniques can be used to provide important information on the behavior of heated oil shale samples and assign the maximum temperature where the basic thermodecomposition of kerogen occurs. Moreover, thermogravimetric analysis coupled with FTIR (TGA-FTIR) as a detector was used for online identification of evolved gases of the TGA in the infrared spectrometer gas cell. This coupling can be useful in studying the mechanism of kerogen pyrolysis, which gives a real-time identification of TGA products at different temperatures. Furthermore, gases evolved at the temperature observed for oil weight loss (TGA/DTG) were coupled with the thermal transitions gained by coupling to the DSC. The simultaneous DSC analysis was also associated with endothermic transitions around the region of gas evolving from oil pyrolysis. Additionally, the variation of the amount of evolved gases during pyrolysis and thermal degradation of the oil shale sample could be detected by Gram-Schmidt thermogram where the FTIR absorption intensity of evolved gases was plotted as a function of temperature.

TGA and DTG
All six oil shale samples collected from Omari deposit were subjected to the above-mentioned techniques under a nitrogen environment over a temperature range of 25 to 1100 • C and at a heating rate of 20 • C/min. In general, all samples showed similar trends with slight discrepancies in the pyrolysis region. Figure 2A shows representative TGA and DTG curves of one of the oil shales sample (OS6, FA 12.70 wt.%). The total weight loss was 52.0 wt.%, which passed through three distinct stages: (1) drying, (2) pyrolysis, and (3) mineral decomposition. In the drying stage (up to 370 • C), less than 1.5 wt.% was lost due to moisture loss, which included surface moisture, shirking core evaporation of moisture, and the moisture in small capillaries of the oil shale particles [2]. In the pyrolysis stage (370-560 • C), a weight loss of 14.0 wt.% was recorded. Here, kerogen was thermally decomposed to form shale oil, shale gas, and char. Furthermore, the mineral matter may have decomposed to produce carbon dioxide and release combined water. Finally, in the mineral decomposition stage (more than 600 • C), about 31.0 wt.% loss was observed. In this stage, the different minerals, e.g., calcite, kaolinite, and dolomite, decomposed to form mainly carbon dioxide and the metal oxide. Moreover, employing the first derivative for the TGA to generate the corresponding DTG showed an average maximum temperature of pyrolysis (T max ) of 472.4 ± 9.1 • C.

170
(1) drying, (2) pyrolysis, and (3) mineral decomposition. In the drying stage (up to 370 °C), less than 171 1.5 wt.% was lost due to moisture loss, which included surface moisture, shirking core evaporation 172 of moisture, and the moisture in small capillaries of the oil shale particles [2]. In the pyrolysis stage   When correlating the TGA mass loss in the pyrolysis region (370-560 • C) with the FA oil content, an outstanding linear correlation (R 2 = 0.99) between the two analysis methods was established ( Figure 2B). Each point in Figure 2B represents an average of three trials, and the corresponding error bars demonstrate the precision of the experiments. Therefore, the TGA method can be used as a reliable alternative for the FA method. Because kerogen type and mineral matrix might vary according to the deposit type [9], such correlation should be utilized based on a specific calibration. Moreover, the produced correlation in Figure 2B can be also described as in Equation (1).
where FA denotes the oil weight percent in the oil shale determined by the FA method, and TGA represents the percent of weight loss at the pyrolysis range. Such correlation is in accordance with the literature [28].

TGA-FTIR
Thermogravimetric analysis coupled with FTIR as a detector was used for online identification of evolved gases of TGA for all samples (OS1 to OS6). The analysis was repeated three times in the infrared spectrometer gas cell, and the average result for each sample is presented in Figure 3. The spectra showed two distinct regions: pyrolysis and mineral decomposition. During the pyrolysis region (370-560 • C), the online FTIR spectrum showed broad bands between 2793-3060 cm −1 assigned to aliphatic (C-H stretching) with a peak area that was directly proportional to oil content. It is noteworthy to mention that CO 2 also evolved at this stage (pyrolysis), which was confirmed by the broad bands between 2225-2410 cm −1 assigned to CO 2 stretching with no obvious correlation related to the oil content of the samples. Furthermore, the dactyloscopic region (500-1500 cm −1 ) showed the characteristic peaks related to the inorganic and the organic matrix. In the second region (thermal decomposition), weak bands with respect to C-H aliphatic stretching (2793-3060 cm −1 ) were shown, which could have been related to the low oil content remaining after pyrolysis. On the other hand, stronger bands were noticed for CO 2 (2225-2410 cm −1 ) due to the thermal decomposition of minerals, which produced CO 2 .  217 Figure 4A shows the 3D FTIR spectra between wavenumbers (2800-300 cm −1 ) for the six oil shale 218 samples at the pyrolysis temperature range (370-560 °C). Quantification was done based on the 219 integration of peak areas at 2927 cm −1 and 2342 cm −1 for the aliphatic hydrocarbons and CO2, 220 respectively. Moreover, a correlation was made between the FA and the areas of the two sets of peaks, 221 as shown in Figure 4B. It was obvious that a good quadratic correlation (R 2 = 0.97) was maintained 222 for the aliphatic hydrocarbons, as shown in Equation (2).
Where Y is the area of aliphatic hydrocarbons (peak at 2927 cm −1 ), and X is the oil content of FA

225
On the other hand, there was no clear correlation between the FA and the peak area of CO2 (2342 226 cm −1 ). However, two regions representing two mechanisms could be envisaged where there was a 227 certain production of CO2 up to FA 7.0 wt.% and then it dropped.  Figure 4A shows the 3D FTIR spectra between wavenumbers (2800-300 cm −1 ) for the six oil shale samples at the pyrolysis temperature range (370-560 • C). Quantification was done based on the integration of peak areas at 2927 cm −1 and 2342 cm −1 for the aliphatic hydrocarbons and CO 2 , respectively. Moreover, a correlation was made between the FA and the areas of the two sets of peaks, as shown in Figure 4B. It was obvious that a good quadratic correlation (R 2 = 0.97) was maintained for the aliphatic hydrocarbons, as shown in Equation (2).
where Y is the area of aliphatic hydrocarbons (peak at 2927 cm −1 ), and X is the oil content of FA (wt.%).
On the other hand, there was no clear correlation between the FA and the peak area of CO 2 (2342 cm −1 ). However, two regions representing two mechanisms could be envisaged where there was a certain production of CO 2 up to FA 7.0 wt.% and then it dropped.  However, for the same type, the specific heat increased with higher oil yield [2]. The same trend was 242 found for all oil shale samples (OS1 to OS6), as exemplified in Figure 6, where specific heat capacity 243 ( Figure 6A) and enthalpy ( Figure 6B) were directly proportional to the oil content.

DSC
The amount of heat needed for retorting oil shale depends predominantly on its specific heat capacity. Therefore, DSC analysis was performed for all samples in triplicate. Figure 5 shows, using OS6 (FA 12.7%) as an example, the DSC curve in addition to its TGA and DTG profiles. Results showed a similar DSC profile as a function of temperature for all samples. Furthermore, the peaks denoted an endothermic behavior in the pyrolysis zone (370-560 • C) with a heat capacity in the range of (0.133-1.228 J/(g·K)) due to the thermal decomposition of kerogen. Moreover, the specific heat of oil shale depended on its type, which differed in the composition of mineral and kerogen content. However, for the same type, the specific heat increased with higher oil yield [2]. The same trend was found for all oil shale samples (OS1 to OS6), as exemplified in Figure 6, where specific heat capacity ( Figure 6A) and enthalpy ( Figure 6B) were directly proportional to the oil content.

234
The amount of heat needed for retorting oil shale depends predominantly on its specific heat 235 capacity. Therefore, DSC analysis was performed for all samples in triplicate. Figure 5 shows, using 236 OS6 (FA 12.7%) as an example, the DSC curve in addition to its TGA and DTG profiles. Results

237
showed a similar DSC profile as a function of temperature for all samples. Furthermore, the peaks 238 denoted an endothermic behavior in the pyrolysis zone (370-560 °C) with a heat capacity in the range 239 of (0.133-1.228 J/(g·K)) due to the thermal decomposition of kerogen. Moreover, the specific heat of 240 oil shale depended on its type, which differed in the composition of mineral and kerogen content.

241
However, for the same type, the specific heat increased with higher oil yield [2]. The same trend was 242 found for all oil shale samples (OS1 to OS6), as exemplified in Figure 6, where specific heat capacity 243 ( Figure 6A) and enthalpy ( Figure 6B) were directly proportional to the oil content.

Gram-Schmidt
The amount of evolved gases resulted from the TGA could be detected and evaluated using the Gram-Schmidt thermogram [29]. For OS1 to OS6, Gram-Schmidt showed increase in the amount of evolved gases as a function of oil content, where the maximum gas productions were around 450 • C and 760 • C for the kerogen pyrolysis and the mineral decomposition, respectively (Figure 7, OS6 at a heating rate of 20 • C/min as an example). This behavior mirrored the peaks found in the DTG shown in Figure 2A. Figure 8 shows two sets of oil shales. The first group represents the low oil content (OS1-OS3), and the second is the high oil content (OS4-OS6). Although the first group contained lower oil content, it showed a higher amount of evolved gases. This might have been related to their lower heat capacity required by the oil for pyrolysis (as shown in Figure 6). Conversely, the second group produced fewer gases due to their higher heat capacity.

Elemental and Mineral Analysis
Elemental analysis is an accurate standard method used for determining the carbon-hydrogen-nitrogen (CHN) content of kerogens by combustion. However, it could be used for determining the origin and the evolution of oil shale with limited accuracy [2]. XRF was conducted to determine the major oxides present in the oil shale samples, whereas XRD was used as a qualitative analytical method for the identification of the crystal phase of minerals in the oil shale samples. SEM revealed information about the solid shale including outer morphology, chemical composition, crystalline structure, and orientation of materials making up the sample. EDX was one of the analytical techniques used for the elemental analysis or the chemical characterization of a sample. The interaction of a source of X-ray excitation and a sample gave a unique atomic structure for each element, which allowed a unique set of peaks on the electromagnetic emission spectrum.

Elemental analysis
CHN analysis was conducted for all oil shale samples (OS1 to OS6). Figure 9A shows a correlation between the FA (wt.%) and the mass content of each element, where a directly proportional relationship between the two was observed. Out of the three possible correlations, elemental hydrogen mass content provided the best correlation to the FA with a good quadratic correlation (R 2 = 0.98), as shown in Equation (3). Correlating the hydrogen content to the FA method is preferred because it is predominantly found in the organic matter (kerogen), while carbon is found in both inorganic and organic fractions.
where H represents the weight percent of hydrogen content in the oil shale sample, and FA denotes the oil weight percent in the oil shale determined by the FA method. Furthermore, the atomic ratios H/C and N/C were calculated and evaluated against the oil content. There was a relatively fixed N/C atomic ratio (0.01-0.02), whereas a wider range of H/C atomic ratio (0.69-1.12) was found. Values of the H/C ratio could be used as a preliminary indicator of the aromatic nature of kerogen. However, the atomic ratios of H/C and O/C of the kerogen content (not the oil shale) could be assigned on the van Krevelen diagram to determine the type of kerogen [10]. It is reported that the kerogen of Jordanian oil shale has an average content of 10 wt.% and atomic H/C and O/C ratios of 1.3 and 0.101, respectively [2].

303
The XRF analysis of the mineral matter resulted from tested oil shale samples, as shown in Table   304 2, indicated that this oil shale contained mainly calcite (CaCO3) and quartz (SiO2). Furthermore, the 305 data showed no trend in the mineral composition that could be correlated with the oil content. The

XRF
The XRF analysis of the mineral matter resulted from tested oil shale samples, as shown in Table 2, indicated that this oil shale contained mainly calcite (CaCO 3 ) and quartz (SiO 2 ). Furthermore, the data showed no trend in the mineral composition that could be correlated with the oil content. The mineral content of the tested samples agreed with the reported average mass concentrations of major inorganic content of oil shale deposits in the central part of Jordan [7].      Figure 11 shows scanning electron micrographs for selected oil shale samples that covered the whole range of oil content given the symbols A (OS1 2.9 wt.%), B (OS3 7.0 wt.%), and C (OS6 12.7 wt.%) with variable magnification from 1 kx (magnification-1000 times) to 35 kx (magnification-35,000 times). The three samples showed amorphous and fragmented surfaces with no distinct clusters of organic materials. EDX analysis (Figure 12) showed the same chemical components detected by XRF with no obvious trend that could be correlated with oil content, which might have been related to the anisotropy of the surface. Figure 13 shows the micrograph with colored mineral distribution mapped on the surface of oil shale and quantified at a rate of 10.67 Kcps in a pulsed mode and multi-line scanning. It could be observed that the following elements, C, O, Mg, Al, Si, S, K, Ca, Fe, Cl, and Zn, were uniformly distributed across the whole surface. Conversely, phosphorus (P) was located in certain spots with less distribution because it originated from the apatite, as found in the XRD analysis. Calcium then silicon and sulfur were found at the highest level, which is in accordance with the literature values of Jordanian oil shale deposits (30-35% CaCO 3 , 10-26.3% SiO 2 , 4.3-4.8% SO 3 ) [7]. On the other hand, a more precise line scan analysis was conducted and is presented in Figure 14. The line analysis presented the same mineral types found by XRD and XRF analyses and showed no localization of the organic matter. Moreover, it demonstrated the diversity of mineral content and that there was no apparent relationship with the oil content.   348 Figure 14. Line analyses of SEM with energy-dispersive X-ray (EDX) detector for A (12.5 wt%), B (7.0 wt%), and C (2.9 wt%) samples.

Conclusions
In this work, several high throughput screening techniques for determining shale oil content were applied as an alternative for the conventional Fischer Assay (FA) method. The Omari oil shale deposit, which is located east of Jordan, was characterized and adopted as a case study. Under the test conditions, the following conclusions were drawn: 1.
The TGA method could be used as an alternative method to the conventional FA method due to the high linear regression factor (R 2 = 0.99) obtained between the weight loss of the organic content in the pyrolysis range (370-560 • C) obtained by the TGA and the oil yield (wt.%) obtained by the FA method.

2.
Elemental analysis could be used for replacing the conventional FA method. A good quadratic correlation (R 2 = 0.98) was maintained between the FA and the hydrogen mass content. 3.
The TGA-FTIR method could be used as another quick method for replacing the conventional FA method. A good quadratic correlation (R 2 = 0.97) was maintained between the FA and the aliphatic hydrocarbons (FTIR peak at 2927 cm −1 ) produced in the pyrolysis zone. 4.
XRF analysis of the mineral matter revealed that this oil shale contains mainly calcite and quartz with low concentrations of other metal oxides (Al 2 O 3 , P 2 O 5 , Fe 2 O 3 , MgO, K 2 O, and TiO 2 ). 5.
XRD analysis revealed the presence of the following minerals: quartz, ankerite, calcite, apatite, and clay, which matches oil shale composition. 6.
SEM analysis revealed an amorphous oil shale surface with no distinct clusters of organic material, which was proven by area mapping and line analysis.