Qualitative and Quantitative Analysis of Heavy Crude Oil Samples and Their SARA Fractions with 13 C Nuclear Magnetic Resonance

: Nuclear magnetic resonance (NMR) approaches have unique advantages in the analysis of crude oil because they are non-destructive and provide information on chemical functional groups. Nevertheless, the correctness and e ﬀ ectiveness of NMR techniques for determining saturates, aromatics, resins, and asphaltenes (SARA analysis) without oil fractioning are still not clear. In this work we compared the measurements and analysis of high-resolution 13 C NMR spectra in B 0 ≈ 16.5 T (NMR frequency of 175 MHz) with the results of SARA fractioning for four various heavy oil samples with viscosities ranging from 100 to 50,000 mPa · s. The presence of all major hydrocarbon components both in crude oil and in each of its fractions was established quantitatively using NMR spectroscopy. Contribution of SARA fractions in the aliphatic (10–60 ppm) and aromatic (110–160 ppm) areas of the 13 C NMR spectra were identiﬁed. Quantitative fractions of aromatic molecules and oil functional groups were determined. Aromaticity factor and the mean length of the hydrocarbon chain were estimated. The obtained results show the feasibility of 13 C NMR spectroscopy for the express analysis of oil from physical properties to the composition of functional groups to follow oil treatment processes.


Introduction
Knowledge of the chemical composition of crude oil is necessary both for fundamental research and technological processes [1]. Despite the high efficiency of the available methods of oil separation and concentration of individual classes of compounds, there is a need for instrumental approaches to obtain additional detailed information about the structure and properties of the obtained fractions, in addition to characterizing complex oil systems in situ [2,3]. Unless modern tools of in-depth analysis are developed and exploited, information about the nature of oils, and products of their fractioning and processing, remain incomplete.
Use of radiospectroscopic techniques such as nuclear magnetic resonance (NMR) and electron paramagnetic resonance (EPR) [4], and their double (NMR and EPR) resonance combinations [5,6] in high magnetic fields (B 0 > 1.5 T) have greatly expanded the possibilities for the selective detection of oil structural fragments and functional groups mainly due to the higher spectral resolution. applications for biological investigations and material science [39], one should expect the rapid development of 13 C DNP for oil studies.
Although attempts to find a correlation between the SARA analyses and NMR measurements have been undertaken for some time, in a well-known paper on this topic [19] it is emphasized that "the results cannot be generalized for other samples. To do that, it is mandatory to incorporate a wider range of experimental data of crude oils characterization and their respective" fractions.
Our current work presents the new results of a comprehensive investigation of samples of four crude oils and their sixteen SARA fractions to obtain data on structural group composition using quantitative 13 C NMR spectroscopy in the magnetic field of B 0 ≈ 16.5 T (f RF = 175 MHz). Information on the content of general functional groups obtained by 13 C NMR spectroscopy can be useful for fast prediction of oil product properties that change under different types of treatment, in addition to the development of a fingerprinting approach.

Materials and Methods
For the current investigation we took four heavy (viscous) oils of various origin and their four SARA fractions (a total of 20 samples). A list of the studied samples and the viscosities of the initial oil is presented in Table 1.

SARA Fractionation
According to the ASTM 2007 standard, oil samples were divided into their fractions: saturates, aromatics, resins, and asphaltenes (SARA). Asphaltenes were separated from oils by precipitation with heptane (40:1) for 24 h and subsequent purification from maltenes (saturate, aromatic, and resin) in a Soxhlet extractor [40,41]. The results of SARA analysis of all studied oil samples are reported in Table 2. The analytical parameters, tools, and separation conditions for conducting SARA analysis are shown in Table 3.

13 C NMR Spectroscopy
NMR experiments on the initial oil samples (1)(2)(3)(4) and their fractions (1s-1as, 2s-2as, 3s-3as, 4s-4as) were performed on a Bruker Avance III HD 700 MHz spectrometer equipped with a quadruple resonance ( 1 H, 13 C, 15 N, 31 P) QCI CryoProbe. All samples of crude oils and their fractions were diluted in deutered chloroform (CDCl 3 ) solvent. Field lock and shimming were achieved using the deuterium signal of CDCl 3 . 13 C NMR spectra were recorded using 90 • pulses with 11.7 µs pulse length and with inverse gated broadband proton decoupling (zgig pulse program); relaxation delay between pulses was 9 s and acquisition time was 3.5 s; spectrum width was set to 220.0 ppm; the number of scans was 3200. An exponential digital filter with the line broadening factor of 10 Hz was applied to process 13 C NMR spectra prior to Fourier transformation. Measurements were conducted at the temperature of 30 • C. All NMR spectra were integrated after baseline correction, and at least three integration values were taken for each calculation. The relative standard deviation of the results of manual integration did not exceed 3%. The integration of the resonance lines in the 13 C NMR spectra was carried out with respect to the 13 C signal of the CDCl 3 solvent, for which the value was taken as 1. Estimation of molar fractions of primary (C p ), secondary and quaternary (C sq ), tertiary (C t ), aromatic (C ar ) carbons, aromaticity factor (F CA ), and mean chain length (MCL) of aliphatic hydrocarbons was carried out in a way similar to our previous work (see References [40,42] and data presented in Supplementary Materials).

Results and Discussion
The 13 C NMR spectra of crude oils 1-4 dissolved in CDCl 3 are shown in Figure 1.

Results and Discussion
The 13 C NMR spectra of crude oils 1-4 dissolved in CDCl3 are shown in Figure 1. The broadening of resonance lines in the aromatic region of the spectrum in the transition from the light oil sample (1) to the heavier oil samples (2-4) is observed. Table 4 shows the results of estimating the molar content of various carbon groups of samples 1-4 made by integration of the corresponding areas of 13 C NMR spectra. Table 4. Molar fractions (%) of primary (Cp), secondary and quaternary (Csq), tertiary (Ct), aromatic The broadening of resonance lines in the aromatic region of the spectrum in the transition from the light oil sample (1) to the heavier oil samples (2)(3)(4) is observed. Table 4 shows the results of estimating the molar content of various carbon groups of samples 1-4 made by integration of the corresponding areas of 13 C NMR spectra. Table 4. Molar fractions (%) of primary (C p ), secondary and quaternary (C sq ), tertiary (C t ), aromatic (C ar ) groups, aromaticity factor (F CA ), and mean chain length (MCL) of aliphatic hydrocarbons based on 13 C NMR spectra of samples 1-4. A quantitative analysis of the composition of the oil samples studied by NMR showed that as the oil viscosity increases, a decrease of C p and C sq parameters and increase in C t and C ar parameters are observed. Moreover, the concentration of aromatic groups (C ar ) in heavy oil sample 4 increases almost two times compared to the less viscous oil sample 1. In addition, with increasing viscosity, the aromaticity factor (F CA ) shows an obvious tendency to increase.

Sample
Figures 2-5 show 13 C NMR spectra of saturated, aromatic, resin, and asphaltene fractions of the studied oil samples. There is also a trend towards broadening of resonance lines in the aromatic area of the spectrum in the transition from the light saturates fraction to the heavier fractions. Quantitative data on the proportions of primary, secondary, tertiary, quaternary, and aromatic carbon atoms of the studied oil fraction samples were also obtained and are presented in Tables 5-8.  Quantitative analysis of the composition of saturated compounds (1s), aromatics (1ar), resins (1r), and asphaltene fraction (1as) of oil sample 1 studied by NMR showed that as the fraction gravity increases, a decrease in parameters C p , C sq and increase in C t , C ar , F CA are observed. In addition, with increasing fraction gravity, there is a tendency for the MCL parameter to increase up to 13.3.  Quantitative analysis of the composition of saturated compounds (2s), aromatics (2ar), resins  Table 6. Molar fractions (%) of primary (C p ), secondary and quaternary (C sq ), tertiary (C t ), aromatic (C ar ) groups, aromaticity factor (F CA ), and mean chain length (MCL) of aliphatic hydrocarbons based on 13 C NMR spectra of samples 2, 2s-2as. Quantitative analysis of the composition of saturated compounds (2s), aromatics (2ar), resins (2r), and asphaltene fraction (2as) of oil sample 2 studied by NMR showed that as the fraction gravity increases, C p and C sq parameters decrease. However, the concentration of tertiary carbons (C t ) in this case also decreases on average, unlike in the case of sample 1. The concentration of aromatic carbons (C ar ) takes the lowest value (4.3%) for the sample of saturates fraction 2s; then, as fraction gravity increases, this parameter evenly increases. In addition, with increasing fraction gravity, there is a tendency for the F CA and MCL parameters to increase.  Quantitative analysis of the composition of saturated compounds (3s), aromatics (3ar), resins (3r), and asphaltene fraction (3as) of oil sample 3 studied by NMR showed that increasing the fraction gravity is accompanied by a decrease in the Cp, Csq, and Ct parameters and an increase in the parameter Car. In addition, with increasing fraction gravity, there is a tendency for the FCA and MCL parameters to increase.  Table 7. Molar fractions (%) of primary (C p ), secondary and quaternary (C sq ), tertiary (C t ), aromatic (C ar ) groups, aromaticity factor (F CA ), and mean chain length (MCL) of aliphatic hydrocarbons based on 13 C NMR spectra of samples 3, 3s-3as. Quantitative analysis of the composition of saturated compounds (3s), aromatics (3ar), resins (3r), and asphaltene fraction (3as) of oil sample 3 studied by NMR showed that increasing the fraction gravity is accompanied by a decrease in the C p , C sq , and C t parameters and an increase in the parameter C ar . In addition, with increasing fraction gravity, there is a tendency for the F CA and MCL parameters to increase.  The behavior of the studied parameters for the fraction samples of the heaviest oil 4 is the same as for the fraction samples of the lightest oil 1. A quantitative analysis of the composition of saturated compounds (4s), aromatics (4ar), resins (4r), and asphaltene fraction (4as) extracted from oil sample 4 studied by NMR showed that an increase in the fraction gravity correlates with a decrease in parameters Cp and Csq, and an increase in Ct, Car, FCA, and MCL. Thus, the maximum value of the mean chain length (MCL) equal to 15.8 is observed for the asphaltene fraction of oil sample 4. Figure 6 shows a summary diagram of changes of the obtained structural group parameters Cp, Csq, Ct, Car, FCA, and MCL for the studied oil samples. This diagram allows us to analyze how the obtained parameters change according to two criteria: the gravity of the fraction (growing from saturates to asphaltenes) and viscosity of the sample (an increase within the same fraction from sample 1 to sample 4). Thus, for the Cp parameter there is a general decrease of approximately twofold in values as the gravity of the fraction increases. However, analysis of the dependence of the parameter Cp on the sample viscosity among individual fractions shows an ambiguous picture: for saturates and asphaltenes there is a slight decrease in the values of Cp, while for the aromatic and  Table 8. Molar fractions (%) of primary (C p ), secondary and quaternary (C sq ), tertiary (C t ), aromatic (C ar ) groups, aromaticity factor (F CA ), and mean chain length (MCL) of aliphatic hydrocarbons based on 13 C NMR spectra of samples 4, 4s-4as. The behavior of the studied parameters for the fraction samples of the heaviest oil 4 is the same as for the fraction samples of the lightest oil 1. A quantitative analysis of the composition of saturated compounds (4s), aromatics (4ar), resins (4r), and asphaltene fraction (4as) extracted from oil sample 4 studied by NMR showed that an increase in the fraction gravity correlates with a decrease in parameters C p and C sq , and an increase in C t , C ar , F CA , and MCL. Thus, the maximum value of the mean chain length (MCL) equal to 15.8 is observed for the asphaltene fraction of oil sample 4. Figure 6 shows a summary diagram of changes of the obtained structural group parameters C p , C sq , C t , C ar , F CA , and MCL for the studied oil samples. This diagram allows us to analyze how the obtained parameters change according to two criteria: the gravity of the fraction (growing from saturates to asphaltenes) and viscosity of the sample (an increase within the same fraction from sample 1 to sample 4). Thus, for the C p parameter there is a general decrease of approximately twofold in values as the gravity of the fraction increases. However, analysis of the dependence of the parameter C p on the sample viscosity among individual fractions shows an ambiguous picture: for saturates and asphaltenes there is a slight decrease in the values of C p , while for the aromatic and resin fractions the values of C p slightly increase. A similar trend in changes in individual fractions is observed for the parameter C sq . However, it was noticed that the reduction in C sq for saturates and asphaltenes as the viscosity of the sample increases occurs more sharply than for C p . Analysis of the parameter C t showed that for each fraction its value increases strongly as the viscosity of the samples increases. Furthermore, by the gravity of the fraction increases, there is an increase in C t values for samples 1 and 4 and a decrease for samples 2 and 3. The behavior of the C ar parameter as the gravity of the fraction increases is exactly opposite to the case of C p : its values increase sharply. However, its values within individual fractions as the sample viscosity increases have a general tendency to decrease. As an exception to this rule, however, there is a sharp increase in the values of the C ar parameter of sample 3 for saturates and asphaltene fractions. In addition, there is a very sharp drop of the C ar parameter value (down to 4.3%) of sample 2 for the aromatics fraction. The remaining parameters-aromaticity factor (F CA ) and mean chain length (MCL) of aliphatic hydrocarbons-are indirect. For example, the values of parameter F CA strongly depend on the values of parameter C ar . Therefore, the behavior of its value both with increasing the gravity of the fraction and increasing the sample viscosity within a separate fraction is similar to the behavior of parameter C ar , however, its changes occur more smoothly. The values of the mean chain length (MCL) of aliphatic hydrocarbons, as expected, grow both as the gravity of the fraction and the sample viscosity within a separate fraction increase. 13 C NMR spectroscopy can significantly expand the repertoire of research methods to study structural-group composition of oil and oil fractions, and adequately describe qualitatively and quantitatively not only the elemental composition but also the molecular structures of natural organic material, its fractions, and intermediate and target products. These characteristics determine both the properties of objects and the strategy of various technological schemes for oil upgrading or for extraction of different fractions and groups.

Conclusions
Regarding specific results, we determined the contribution of fractions of oil in the aliphatic (10-60 ppm) and aromatic (110-160 ppm) areas of the 13 C NMR spectra. We demonstrated that NMR spectroscopy methods make it easy to establish the quantitative presence of the main hydrocarbon components in any fraction of oil. The quantitative fractions of the main functional groups constituting oil hydrocarbons and hydrocarbons of the main oil fractions (saturates, aromatics, resins, and asphaltenes) in several samples were determined and their variations were shown.