Catalytic aquathermolysis of Shengli heavy crude oil with an amphiphilic cobalt catalyst

2.1 Experimental material

In experiments, the heavy oil used was from the Shengli Oilfield, China, with a viscosity of 8960 mPa s at 80 °C and 167,400 mPa s at 50 °C.

Almost all of the chemicals used in the study were of analytical grade and purchased from the Sinopharm Chemical Reagent Co. Ltd. Dodecylbenzenesulfonic acid (DBSA) was provided by the Tokyo Chemical Industry Co. Ltd and its purity was above 90 %. All reagents were used as-received without further purification.

2.2 Synthesis of the active cobalt catalyst

About 120 mL of 0.1 mol/L cobalt nitrate (Ni(NO₃)₂) solution was accurately measured and introduced into a flask and incubated at 60 °C for 15 min in an oil bath. A 1 mol/L NaOH solution was slowly added drop-by-drop to keep the pH value of this solution at 11. After stirring for 20 min, the solution became opaque and then was vacuum-filtered with a Whatman membrane. The precipitate was washed three times with deionized water and then added to another flask, to which 10 g of DBSA was added. The flask was heated in an oil bath at 100 °C for 4 h, while the solution was stirred at 800 rpm. After that, the product was extracted with chloroform, and the by-products or unreacted solids were removed via centrifugation. The product was separated from chloroform by vacuum distillation and then dried under vacuum at 80 °C for 24 h. The collected product was a light blue powder, which was the desired catalyst.

2.3 Catalyst characterization

The elemental contents of C, H, S, and Co of the synthesized catalyst were determined with an elemental analyzer (Vario EL III, Elementar), combined with an atomic absorption spectrometer (ContrAA 700, Analytik Jena), and the content of O was calculated by subtracting the content of all other elements. The molecular structure was also characterized qualitatively with a Fourier transform infrared spectrometer (Nicolet 6700, Thermo) after being mixed with KBr.

The surface activity and interfacial activity between heavy oil and the catalyst solution were measured with a surface tensiometer (EasyDyne, Krüss) and a spinning drop interfacial tensiometer (TX 500C, CNG).

The catalyst stability was evaluated with a thermogravimetric analyzer (WCT-1D, Beijing Optical Instruments) coupled with a differential thermal analyzer (DTA) under N₂ atmosphere. The temperature was increased to 600 °C at 20  °C/min.

2.4 Catalytic aquathermolysis of Shengli heavy oil

A total of 140 g of heavy oil, 60 g of water, and 0.28 g of catalyst (0.2 wt%, mass ratio of the catalyst to heavy oil) was added to an autoclave. The temperature was raised to 250 °C in less than 1 h and was kept at 250 °C for 24 h. Then the autoclave was cooled down to room temperature with cooling water in less than 10 min. After that, the mixture was taken out, and dehydrated by distillation below 120 °C. Over 95 % water was removed. Then the viscosity of the heavy oil after aquathermolysis was measured with a Brookfield DV-III viscometer at both 80 and 50 °C. In order to obtain accurate data, the torque was kept at 50 % during measurements. The aquathermolysis experiments of heavy oil were repeated at least 3 times, and the data were averaged.

2.5 Analysis of heavy oil samples

Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) is an effective technique for analyzing the molecular structure of complex hydrocarbon systems, such as heavy oil (Qian and Robbins 2001; Hughey et al. 2002). Combined with soft ionization methods, e.g., electrospray ionization (ESI), FT-ICR MS was applied to investigate catalytic aquathermolysis of heavy oil (Wu et al. 2005; Smith et al. 2008; Colati et al. 2013; Wang et al. 2013). 10 mg oil sample was firstly dissolved in 1 mL toluene, and then diluted with toluene/methanol (1:3, v/v) to yield a 0.2-mg/mL solution. After that, 15 μL ammonium hydroxide (28 %) solution was added to the 1-mL oil sample solution. The solution was vortex-mixed before analysis. The MS analysis was carried out on a Bruker Apex-ultra FT-ICR mass spectrometer with a 9.4 T superconducting magnet. The diluted sample solution was injected at 180 μL/h. The atmospheric pressure photo ionization source was purchased from Bruker Daltonics. Nitrogen was used as the drying and nebulizing gas. The nebulizing gas temperature was 200 °C. The nebulizing gas flow rate was 1.2 L/min, and the drying gas flow rate was 5.0 L/min. For ESI, the main operating parameters were emitter voltage, 4.0 kV; capillary column introduction voltage, 4.5 kV; and capillary column end voltage, −320 V. Ions were accumulated for 0.01 s in a hexapole with a 2.4-V direct current (DC) voltage. The optimized mass for Q1 was 250 Da. The extraction period for ions from the hexapole to the ICR cell was set to 1.2 ms. The RF excitation was attenuated at 12.5 dB and used to excite ions over the range of 115–800 Da. Spectra comprising 4 M data points were collected. The signal-to-noise ratio was improved by summing 128 scan FT-ICR datasets.

Mass spectra were internally calibrated using an extended homologous alkylation series (molecular ions of aromatic hydrocarbons and thiophenes) of high relative abundance in a mixed heavy oil within the mass range of 300–1000 Da. Data analysis was performed using custom software and the procedure has been described in detail elsewhere (Shi et al. 2009). The compounds were characterized by class (numbers of N, O, and S heteroatoms), type [rings plus double bond equivalence (DBE)], and carbon number. Species and their isotopes with different DBE values and carbon number were searched within a set ±0.001 Kendrick mass defect (KMD) tolerance (Hughey et al. 2001).

The saturate, aromatic, resin, and asphaltene (SARA) fractions of heavy oil samples before and after catalytic aquathermolysis were separated and determined using alumina column chromatography. The average molecular weights of oil samples, resins, and asphaltenes were determined by vapor pressure osmometry (VPO) in a toluene solution. The C, H, S, and N contents of resins and asphaltenes were measured with an elemental analyzer (Elementar Vario EL III). FT-IR spectra of the resins and asphaltenes were collected on a Nicolet 6700 Fourier transform infrared spectrometer. The samples were daubed on a KBr disk to obtain a thin film. ¹H NMR spectra of resins and asphaltenes were recorded on a Bruker ARX400 spectrometer. CDCl₃ was used as the solvent and TMS as the internal chemical shift standard.

2.6 Analysis of gas products

Gas products of heavy oil after catalytic aquathermolysis were analyzed by an Agilent 7890A gas chromatograph with an Abel column AB-5MS (60 m × 0.25 mm × 0.25 μm) and a flame ionization detector. High purity nitrogen was used as carrier gas.

3.1 Characterization of the synthesized catalyst

The FT-IR spectra of DBSA and the synthesized cobalt catalyst are shown in Fig. 1. The typical peaks of the benzenesulfonic acid group are the stretching vibration peak of O–H (907 cm⁻¹) and the stretching vibration peaks of sulfate S=O (1363, 1177, and 1129 cm⁻¹) in the DBSA spectrum. In the spectrum of the catalyst, the stretching vibration peak of O–H at 907 cm⁻¹ vanished, and the stretching vibration peaks of sulfate S=O also shifted to higher wave number due to the inductive effect of the cobalt atom. Therefore, the synthesized compound was cobalt dodecylbenzenesulfonate (CoSDB). The element analysis results proved the molecular structure and composition of the synthesized catalyst (Table 1). The elemental composition is consistent with the stoichiometric values. The slightly higher H and O contents of the experimental results compared to those expected theoretically might be due to water absorption during the sample preparation.

Open image in new window

Table 1

Elemental composition of the synthesized active cobalt catalyst

Elemental content	C,  %	H,  %	S,  %	O,  %	Co,  %
Measured	57.37 ± 0.59	8.94 ± 0.17	8.59 ± 0.15	16.67 ± 0.18	8.43 ± 0.12
Theoretical	60.94 ± 0.44	8.18 ± 0.14	9.03 ± 0.13	13.54 ± 0.21	8.31 ± 0.13

The surface tension of the CoDBS aqueous solution decreased to approximately 27 mN/m when the CoDBS concentration was 0.02 wt% (Fig. 2). The interfacial tension between the CoDBS aqueous solution and heavy oil was also measured by a spinning drop method. At a concentration of 0.2 wt%, CoDBS could reduce the interfacial tension from about 20 mM/m to 0.7 mM/m. The interfacial activity of the CoDBS catalyst will benefit the emulsification of water and heavy oil. The better mixing and increased interfacial area between water and heavy oil might improve the cracking of asphaltene and resin in heavy oil, as asphaltene is partially surface active and apt to adsorb at the interface during aquathermolysis.

Open image in new window

TGA-DTA (thermogravimetric analysis and differential thermal analysis) data of the synthesized catalyst are shown in Fig. 3. There is a significant weight loss and an exothermic peak at about 400 °C. This indicated that the catalyst does not decompose below 400 °C. That is to say, the CoDBS catalyst is thermally stable during the aquathermolysis process, usually below 300 °C.

Open image in new window

3.2 Viscosity reduction and properties of the heavy oil after aquathermolysis

High resolution MS was performed to study the changes of heteroatomic compounds in Shengli heavy oil before and after aquathermolysis. The results are shown in Figs. 4, 5, 6, 7, and 8 and Table 4. Because of the different ionizations of different compounds, the contents in Fig. 4 only show the relative contents. For example, the sulfur compounds have a low ionization ratio and they also show low contents in Fig. 4. In the negative-ion mode of ESI FT-ICR MS, heteroatomic compounds including O₂, N₁, N₁O₂, O₂S₁, O₃, O₁, N₁S₁, N₁O₁, N₂, N₂O₂S₂, S₁, and N₁O₂S₂ were detected in the heavy oil sample. After aquathermolysis, the contents of N₂O₂S₂, S₁, and N₁O₂S₂ decreased sharply, and were hardly detected. This indicates that the sulfur compounds are apt to crack during aquathermolysis because of the relatively small bond energy of C–S bonds, which also corresponds with the element analysis results. N₁-, O₁-, and O₃-type contents increased after aquathermolysis while O₂ type decreased.

Open image in new window

Open image in new window

Open image in new window

Open image in new window

Open image in new window

Table 4

Various fractions of the heteroatomic compound groups in heavy oil samples before and after aquathermolysis (negative ions)

Heavy crude oil		Heavy oil after aquathermolysis		Summary of changes
Group	DBE and carbon number distribution	Group	DBE and carbon number distribution
N1	9–15; C20–C35	N1	12–15;C20–C35	Some compounds decreased
N1O2	10–16; C20–C40	N1O2	10–15; C20–C40	Some compounds decreased
N1O1	13–18; C20–C35	N1O1	10–18; C20–C35	Parts of compounds decreased obviously
N1S1	11–17; C20–C35	N1S1	14–18; C20–C35	Some compounds decreased
N2	12–18; C20–C35	N2	12–18; C20–C35	Some compounds decreased
O2S1	3–9; C20–C35	O2S1	3–9; C20–C35	Parts of compounds decreased obviously
O1S1	6, 10; C24–C26	O1S1	6–8; C23–C27	Parts of compounds decreased obviously
O1	4–12; C18–C35	O1	4–11; C18–C35	Some compounds decreased
O2	1–8; C15–C35	O2	1–3; C15–C20	Parts of compounds decreased obviously
O3	4–11; C20–C35	O3	2–11; C18–C35	Parts of compounds decreased obviously
N1O3S2	2–4; C15–C30			Cracked totally
O3S2	2–6; C15–C30			Cracked totally
S1	6–15; C22–C50			Cracked totally
N1O2S1	11–13; C25–C30			Cracked totally
N1O2S2	4–8; C20–C35			Cracked totally

The type and content of N₁ compounds are shown in Fig. 5. The N₁ compounds in the negative-ion mode are nonbasic nitrogen compounds including pyrrole, benzopyrrole, carbazole, benzocarbazole, and dibenzocarbazole. Their DBE ranges from 9 to 16, and the carbon number is from 20 to 40 (C₂₀ to C₄₀). After aquathermolysis, the DBE of the compounds ranged from 12 to 15, and the carbon number was from 20 to 40. Benzocarbazole-type compounds (DBE 12) and dibenzocarbazole (DBE 15) had the highest content. The N₁ compounds with DBE of 9–11 decreased. Almost no DBE 6 (benzopyrrole type) and DBE 3 (pyrrole type) compounds were detected.

According to the results of O₁ and O₂ compounds, the oxygen compounds in heavy crude oil are mainly phenolics and carboxylic acids. Because the carboxylic acid compounds are thermally instable, they tend to crack into hydrocarbons and carbon dioxide during aquathermolysis. The phenolic compounds mainly include alkyl phenolic and alkyl benzophenolic (from benzophenolics to quadribenzophenolics) compounds. The O₁ class of the negative-ion mode of ESI FT-ICR MS results is shown in Fig. 6. For heavy oil samples before and after aquathermolysis, the distributions of DBE and carbon number were continuous. DBE was from 4 to 9, while the carbon number was from 18 to 35 (C₁₈ to C₃₅). Since only the hydroxyl compounds can be ionized under the negative ESI mode, and the minimal DBE was 4, it would be deduced that the O₁ class compounds were mainly alkyl phenolics and their benzo-derivatives. The O₁ class compounds are relatively stable, and their content did not change obviously after aquathermolysis.

As shown in Fig. 7, the DBE of O₂ class compounds in the heavy oil sample was from 1 to 7, and their carbon numbers were from 16 to 35. After aquathermolysis, the DBE was from 1 to 3, and the carbon numbers were from 16 to 20. Furthermore, the contents also decreased obviously. The O₂ class compounds of DBE 1 should be carboxylic acid compounds, and the DBE 2 should be naphthenic acids. The DBE 3-7 should be polyring naphthenic acids. Both carboxylic acids and naphthenic acids are active and easily cracked. As a result, the contents of O₂ class compounds decreased significantly after catalytic reaction. The results of O₂S₁ class compounds are shown in Fig. 8. Like O₂ class compounds, their contents decreased noticeably. Their DBE was from 4 to 8, and the molecular structures should be thiophene rings fused with naphthenic acids up to benzothiophene fused with dual ring naphthenic acids.

3.3 Changes of the molecular structures of asphaltenes and resins

The above results show that the synthesized catalyst effectively reduced the viscosity of heavy oils during catalytic aquathermolysis, and the SARA analysis also shows decreases in the asphaltene and resin contents. Since the asphaltene and resin are the high-molecular weight components, they might be the direct reason for the high viscosity of the heavy oil. It is necessary to study the change of the molecular structures of asphaltene and resin during the aquathermolysis process. FT-IR was used to characterize the ratio of groups to that of methyl groups, \(n_{{{\text{CH}}_{ 2} }} /n_{{{\text{CH}}_{ 3} }}\) of resin and asphaltene, which means the length of the aliphatic side chains, through Eq. (1) (Wang et al. 2009). According to the FT-IR spectra in Fig. 9, the calculated results are listed in Table 6. The \(n_{{{\text{CH}}_{ 2} }} /n_{{{\text{CH}}_{ 3} }}\) of asphaltene and resin fractions after aquathermolysis were smaller than those of the original ones. This indicates that parts of the side chains were cracked during aquathermolysis. The side chains of resin were more apt to be cracked than those of asphaltene.

Open image in new window

Table 6

\(n_{{{\text{CH}}_{ 2} }} /n_{{{\text{CH}}_{ 3} }}\) of asphaltenes and resins

Sample	\(n_{{{\text{CH}}_{ 2} }} /n_{{{\text{CH}}_{ 3} }}\)
Heavy crude oil
Asphaltene	0.212
Resin	0.644
Heavy oil after aquathermolysis
Asphaltene	0.177
Resin	0.171

$$n_{{{\text{CH}}_{ 2} }} /n_{{{\text{CH}}_{ 3} }} = 2.93\,A_{1460} /A_{1380} - 3.70$$

(1)

where \(n_{{{\text{CH}}_{ 2} }}\) and \(n_{{{\text{CH}}_{ 3} }}\) are the number of methylene and methyl groups; A ₁₄₆₀ and A ₁₃₈₀ are the IR absorption intensity at 1460 and 1380 cm⁻¹, respectively.

An improved Brown–Ladner method was adopted to calculate the molecular structures of resins and asphaltenes with the data of elemental analysis, average molecular weight, and ¹H-NMR (Suzuki et al. 1982; Ali et al. 1990; Strausz et al. 1992). The calculated results are listed in Table 7. The ¹H-NMR spectra of resins and asphaltenes are shown in Figs. 10 and 11. The aromaticity (f _A) and aromatic condensation index (H _AU/C _A) of asphaltene and resin were calculated from Eqs. (2) and (3) (Liang 2000). The parameter H _AU/C _A establishes the relation between the average number of aromatic carbons and the number of carbon atoms probably incorporated into polycondensed structures of the asphaltenes and resins.

Table 7

Average structure parameters of resin and asphaltene fractions extracted from heavy oil before and after aquathermolysis

Sample	Asphaltene		Resin
	Heavy crude oil	Heavy oil after aquathermolysis	Heavy crude oil	Heavy oil after aquathermolysis
MW, g/mol	10,031	9078	1454	1403
C content, wt%	81.1	80.2	85.2	85.1
H content, wt%	8.5	8.4	10.0	10.1
C/H	0.796	0.791	0.707	0.699
Relative content, %
H A *	10.54	6.49	7.25	6.88
H α *	12.72	11.43	15.96	15.48
H β *	58.01	58.24	60.40	60.70
H γ *	18.73	23.84	16.39	16.95
H AU/C A	0.48	0.37	0.61	0.61
R	106.6	101.8	13.8	13.9
R A	74.6	62.3	8.5	8.5
R N	32.0	39.6	5.3	5.3
C	678.2	606.4	103.2	106.6
C A	300.2	251.0	36.1	36.2
C N	128.1	158.3	21.1	21.4
C P	249.9	197.1	46.1	49.1
f A	0.44	0.41	0.35	0.34
f N	0.19	0.26	0.20	0.20
f P	0.37	0.33	0.45	0.46

Open image in new window

Open image in new window

$$f_{\text{A}} = \frac{{{{C/H}} - \left( {{{H}}_{\alpha }^{ *} + {{H}}_{\beta }^{ *} + {{H}}_{\gamma }^{ *} } \right)/2{{H}}_{\text{T}}^{ *} }}{{C/H}}$$

(2)

$$\frac{{{{H}}_{\text{AU}} }}{{{{C}}_{\text{AU}} }} = \frac{{{{H}}_{\text{A}}^{ *} / {{H}}_{\text{T}}^{ *} + {{H}}_{\alpha }^{ *} /2{{H}}_{\text{T}}^{ *} }}{{{{C/H}} - ({{H}}_{\alpha }^{ *} + {{H}}_{\beta }^{ *} + {{H}}_{\gamma }^{ *} )/2{{H}}_{\text{T}}^{ *} }}$$

(3)

In Eqs. (2), (3), and Table 7, C/H is the atomic ratio of carbon to hydrogen, which is calculated from the elemental analysis data. H _A ^* refers to the relative content of aromatic hydrogens; H _α ^* , H _β ^* , and H _γ ^* refer to relative content of hydrogens which are in α, β, and γ positions relative to aromatic rings: \({{H}}_{{\text{T}}} ^{*} = {{H}}{}_{{\text{A}}}^{*} + {{H}}_{\alpha } ^{*} + {{H}}_{\beta } ^{*} + {{H}}_{\gamma } ^{*}\). C, C _A, C _N, and C _P are the total carbon number, aromatic carbon number, naphthenic carbon number, and paraffinic carbon number, respectively. R is the total number of rings; R _A is the number of aromatic rings; and R _N is the number of naphthenic rings. f _A is the ratio of the number of aromatic carbons to the total number of carbons; f _N is the ratio of the number of naphthenic carbons to the total number of carbons; and f _P is the ratio of the number of paraffinic carbons to the total number of carbons. The higher the H _AU/C _A ratio, the lower the aromatic condensation degree.

For asphaltene, both f _A and H _AU / C _A decreased after catalytic aquathermolysis. The reduction in aromaticity could be attributed to the rupture of polyring structures after the removal of heteroatoms or the hydrogenation of unsaturated groups, while the increase in the aromatic condensation degree might originate from the aggregation of aromatic units after the loss of side chains. For resin, the H/C, f _A and H _AU / C _A changed slightly after catalytic aquathermolysis, almost the same as those of the crude resin. This is consistent with the previous results that the content of the resin changed much more but its molecular weight remained the same. It seems that during the aquathermolysis, the side chains of asphaltene might be more easily cracked than the side chains of resin, and parts of asphaltene might be transferred to resin after cracking.

The detailed structures of asphaltene and resin are also shown in Table 7. The average molecular structure of asphaltene changed significantly after aquathermolysis, while the resin remained ostensibly unchanged. Almost all the parameters of asphaltene, such as R _T, R _A, C _T, C _A, and C _P, decreased after aquathermolysis. The smaller R _T and C _T of asphaltene proved that the cracking and ring-opening reactions occurred during aquathermolysis. The increased R _N and C _N also indicated that the aromatic rings were opened by hydrogenation. The decreased H _AU/C _A and increased f _N also indicated the lower aromatic condensation degree of the asphaltene after aquathermolysis. The changes of C _P and f _P further illustrated that the alkyl side chains were cracked off relatively easily over the catalyst. All these changes in the molecular structures of asphaltene benefited viscosity reduction and partially improved the quality of heavy oil. The changes of resin were more complicated than asphaltene. After aquathermolysis, part of the original resin was cracked into relatively small molecules, leading to a decrease in the average molecular weight, aromatic, and naphthenic ring numbers of the resin. Meanwhile, part of the cracked asphaltenes were of lower molecular weight and hence classified into resin fraction. After catalystic aquathermolysis, the resins might be of higher molecular weight, aromatic, and naphthenic ring numbers than the original resin. Therefore, the newly formed resin fraction had almost the same average molecular structure parameters compared with that of the original resin fraction.

In general, the catalytic aquathermolysis had more influence on the structure of the asphaltene than that of the resin. This is because the heteroatomic compounds, especially the sulfur and oxygen heteroatomic compounds, are usually found in the asphaltene fraction, where the bond energy of the C–S and C–O is lower than that of the C–C in every fractions. Therefore, the weak bonds are easier to crack by the catalyst during catalytic aquathermolysis and hence the high macromolecule content in asphaltene and resin fractions reduced. Moreover, asphaltenes are interfacially active compounds, and tend to adsorb at the water/oil interface. The addition of catalyst could reduce the interfacial tension force and improve the mixing of water and heavy oil, and meanwhile, the active metals, cobalt ions, were also concentrated at the interface. More importantly, H₂O would act as an H-donor in the catalytic reaction, where its role was same as that in heavy oil hydrotreatment. Therefore, the asphaltene could be cracked more easily.