1

Molecular dynamics analysis on hydrogenation process of asphaltene aggregate for 1

medium and low temperature coal tar 2

Xiaokang Han1,2, Wenzhou Yan1, Xu Liu3,4, Dong Li*3,4, Yonghong Zhu3,4 3

4

1School of Management, Xi’an University of Architecture and Technology, Xi’an 710055, People’s 5

Republic of China 6

2Project Control Department, Hualu Engineering & Technology Co., Ltd., Xi’an 710065, People’s 7

Republic of China 8

3School of Chemical Engineering, Northwest University, Xi’an 710069, People’s Republic of China 9

4The Research Center of Chemical Engineering Applying Technology for Resource of Shaanxi, Xi’an 10

710069, People’s Republic of China 11

12

Corresponding author: Dong Li 13

ORCID: 0000-0002-4578-0595 14

Telephones: +86-029-88373425 15

Fax: +86-029-88305825 16

Email: lidong@nwu.edu.cn 17

18

Abstract 19

In this paper, the average molecular structure models of middle-low temperature coal tar (MLCT) 20

asphaltene before and after hydrogenation were obtained by 1H-NMR characterization. Then the 21

aggregation structure models of MLCT asphaltenes before and after hydrogenation was constructed in 22

Materials Studio 8.0 software. The comparative analysis of structural characteristics between original 23

and hydrogenated asphaltenes were carried out. The final configuration, molecular rotation radius and 24

2

interaction energy of asphaltene aggerates before and after hydrogenation were analyzed by molecular 25

dynamic simulation in different solvents. 26

The conclusions are as follows: (1) The MLCT asphaltenes exist in the form of asphaltene aggregates 27

before and after hydrogenation. After structural optimization, the original asphaltene molecules become 28

2-3 asphaltene molecules with approximately parallel, orderly arrangement and polycyclic aromatic 29

hydrocarbons (PAH) stacking aggregate structure. The molecular accumulation number of its aggregate 30

is less than that of petroleum asphaltene, and more complex nano-aggregates are formed by self-assembly 31

among multiple aggregates. The accumulation number of asphaltene molecules after MLCT 32

hydrogenation are less than that of original asphaltene. (2) The total interaction between MLCT primary 33

asphaltene and hydrogenated asphaltene mainly comes from the intermolecular Van der Waals force. 34

Compared with original asphaltene, the van der Waals energy and electrostatic energy of hydrogenated 35

asphaltene are smaller and more affected by temperature. This shows that the intermolecular interaction 36

of hydrogenated asphaltene is weak and the structure of the aggregate is unstable. (3) Aromatic 37

compounds, strong basic compounds and strong polar compounds have good dissociation effect on 38

MLCT asphaltenes and can effectively inhibit the aggregation and asphaltenes. 39

Keywords: Meddle-low temperature coal tar, Asphaltene aggregate, Hydrogenation, Molecular 40

dynamics, Materials Studio 41

1 Introduction 42

Medium and low temperature coal tar is a by-product in the process of coal gasification, production of 43

semi-coke and upgrading of low-rank coal (Sun et al. 2015; Zhu et al. 2015). Asphaltene of middle-low 44

temperature coal tar refers to the substances that are insoluble in n-alkanes with low carbon number but 45

soluble in aromatics. Its structural unit takes the condensed aromatic ring system as the core, with a 46

number of naphthenic rings and alkyl side chains, as well as sulfur, nitrogen, oxygen and various metal 47

heteroatoms. It’s similar to crude oil that asphaltene is the most complex and difficult removal component 48

in coal tar, which usually accounts for 15-30wt% of coal tar. Coal tar asphaltene has the following 49

3

characteristics compared with crude oil asphaltene (Zhu et al. 2016; Berrgmaan et al. 2003; Sabbah et al. 50

2011): 51

I: The aromatic carbon ratio (fa) of coal tar asphaltene is about 0.7, the aromatic ring condensation 52

parameter (HAu/CA) is about 0.5, which are higher than that of all kinds of crude oil asphaltenes; 53

II: The heteroatom (O+N+S) content of coal tar asphaltene is about 10wt%, which is higher than 54

that of all kinds of crude oil asphaltenes; 55

III: The hydrogen-carbon atomic ratio of coal tar asphaltene is less than 1 and the ratio of aromatic 56

core carbon number to aromatic ring peripheral carbon number is less than 1.5; 57

IV: The molecular weight of coal tar asphaltene is about 500 to 600 which is lower than that of 58

ordinary crude oil asphaltene; 59

V: The alkyl side chain of condensed aromatic ring of coal tar asphaltene is mainly methyl and 60

ethyl, which is shorter than that of general petroleum asphaltene. 61

For a long time, many scholars have studied the structural unit model of asphaltene, but because of the 62

complexity of the molecular structure of asphaltene, the molecular structure model of asphaltene is still 63

controversial (Karimi et al. 2011; Wargadalam et al. 2002). At present, it is considered that the molecular 64

structure unit models of asphaltenes are mainly "group island type", "continental type" and "mixed type" 65

(Fernando et al. 2013, Gray et al. 2011). Schuler et al. (2015, 2017, 2020) demonstrated the first direct 66

observation of wide diversity of asphaltene molecular motifs with peri- and cata-condensed polycyclic 67

aromatic hydrocarbons (PAHs). Various types of coal-derived asphaltene structures were obtained by the 68

atomic force microscopy. A large amount of resin and asphaltene in heavy oil make it form a relatively 69

stable colloidal dispersion system. The stability of colloidal dispersion system is closely related to the 70

aromatic ring, aliphatic chain, heteroatomic structure, temperature, pressure, solvent and other external 71

conditions of asphaltene (Xu et al. 2016). The change of the above influencing factors will destroy the 72

stability of the colloidal system, resulting in the existence of asphaltene in the form of aggregation and 73

precipitation (Jian et al. 2014). As asphaltene aggregation and sedimentation will cause many problems, 74

such as tar fluidity reduction, pipeline blockage, catalyst deactivation and so on, many scholars have 75

used different methods to study the behavior of asphaltene aggregation. 76

Dickie et al. (1967) proposed the “Yen” model of asphaltene colloid aggregation for the first time. The 77

4

“Yen” model explains the aggregation mechanism of asphaltenes from molecular state to cluster state, 78

that is, a single aromatic layer can be stacked to form a PAH complex, which is further associated into a 79

micelle, and when the concentration of asphaltene is high enough, the micelles associate with each other 80

to form aggregates. The Yen model provides an important theoretical basis for the measurement and 81

analysis of phase separation asphaltenes. With the deepening of asphaltene research and the development 82

of analytical technology, Mullins et al. (2010) proposed an improved “Yen-Mullins” model. The basic 83

characteristics of the “Yen-Mullins” model are as follows: the asphaltene is mainly "continental" 84

molecular structure, and the asphaltene is distributed in the form of monomer molecules, nano-aggregates, 85

clusters and flocculationn in the solution. "Yen-Mullins" model proposed that the π-π force between 86

aromatic rings is the dominant force for asphaltene stacking to form molecular clusters. However, it's 87

important to note that Redelius et al. (2000, 2007) and Acevedo et al. (2010) investigated the physical 88

chemistry properties of asphaltenes or bitumen using solubility parameters of asphaltenes and their 89

fractions. Redelius et al. (2000) proposed a new bitumen model showing that the stability of bitumen 90

depends on the mutual solubility of all components. This statement is contradictory to the “old” ideas of 91

bitumen consisting of a dispersion of asphaltene micelles. The results indicated that the polar interactions 92

and hydrogen bonding are extremely important for the solubility properties, stability of bitumen and 93

mechanical properties of bitumen. In addition, Bake et al. studied three different source asphaltene 94

including immature source roc, petroleum, and coal to obtain their chemical compositions, molecular 95

architectures and structure-solubility relationships. It showed that π-π stacking is strongest in planar 96

asphaltenes with high aromatic content, whereas steric disruption is larger in less planarasphaltenes with 97

higher aliphatic content. Furthermore, Eyssautier et al. (2011 and 2012) studied the asphaltene 98

nanoaggregate with extensive experiment using the small angle neutron and X-ray scattering (SANS, 99

SAXS) measurements. One of the important finding is that the asphaltene nanoaggregate can be 100

represented by a dense aromatic core (PAH stacking in the middle) and a shell highly concentrated in 101

aliphatic carbon, which were best described by a disk of total radius 32 Å with 30% polydispersity and 102

a height of 6.7 Å. A hierarchical aggregation model, from free monomers to nanoaggregates and then 103

assemble into high aggregation number fractal clusters, was identified as a multi-scale organization by 104

Eyssautier et al. (2011). 105

5

The principle of the molecular dynamics (MD) method is to use the finite difference method to accurately 106

solve the Newtonian equations of motion of the simulated system in order to calculate the conformational 107

and wave changes of all particles in the system at the same time. MD simulation can not only predict and 108

determine the properties and structural characteristics of various systems, such as intermolecular 109

interactions, coordination characteristics of molecules in solution and so on. In recent years, MD 110

simulation has become an important tool for the study of asphaltene aggregation, which has been widely 111

used to study the structure of asphaltene association of heavy oil and made some progress. 112

At present, the research on asphaltene aggregate by MD is mainly in the following three aspects: (1) The 113

effect of alkyl side chains on aggregation. In the MD simulation of the asphaltene structure model, 114

Pacheco-Sanchez et. al (2004) observed that the asphaltene aggregation structure is consistent with the 115

“Yen” model, that is, there are three geometric configurations of Face-to-Face, offset π-stacked and T-116

shaped between the asphaltene aggregate. The alkyl side chain plays a role in reducing the size of 117

asphaltene aggregates. Jian et al. (2014) found that the longer alkyl side chain is not conducive to the 118

accumulation of aromatic core of asphaltenes, but is conducive to the accumulation of asphaltenes. (2) 119

The effect of temperature and solvent on the asphaltene aggregates. Zhang et al. (2007) found that the 120

distance of asphaltene dimer decreased with the increase of temperature. The simulation of asphaltene 121

binder in different solvents by Carauta et al. (2005) showed that the aromatic nuclear packing distance 122

of asphaltene association in poor solvents such as heptane is smaller than that in good solvents such as 123

toluene. (3) The interaction between asphaltene and other components, such as resin, water, inhibitor and 124

so on. Alvarez et al. (2006) obtained the interaction potential curves of asphaltene-asphaltene, asphaltene-125

resin and resin-resin systems by simulation. At present, it is considered that the aggregate structure of 126

asphaltene is electrostatic interaction, hydrogen bond, van der Waals interaction and other forces (Coata 127

et al. 2012; Lima et al. 2017). 128

In this paper, the average molecular structure model of LCT asphaltene before and after hydrogenation 129

was obtained by MLCT asphaltene characterization. Then the aggregate structure models of asphaltene 130

before and after hydrogenation were constructed in Materials Studio 8.0 software. The optimal 131

conformation was obtained by the kinetic optimization. Also, the structural characteristics of original and 132

hydrogenated asphaltene were compared. Then the effects of different solvents on the structural 133

6

characteristics, aggregation state and packing characteristics of the asphaltene aggregate were explored 134

by MD simulation of the asphaltene aggregate before and after hydrogenation in different solvents. This 135

helped to analyze the final configuration and molecular rotation radius of the asphaltene aggregate from 136

the point of view of structure. Finally, the interaction energy of asphaltene aggregate was analyzed from 137

the point of view of energy to explore the dominant force leading to asphaltene aggregate and the effect 138

of MLCT hydrogenation on asphaltene aggregate structure. 139

2 Average molecular structure model of coal tar asphaltene 140

The 1H-NMR spectrum of MLCT asphaltene before and after hydrogenation were shown in Fig. 1. The 141

detailed hydrogenation process was consistent with previous studies published by our research group (Li 142

et al. 2017). According to the 1H-NMR spectrum, hydrogen was divided into different types according 143

to chemical shift and integrated (Lalvani et al. 1991). The results are normalized and shown in Table 1. 144

12 10 8 6 4 2 0 -2

Original asphaltene

Hydrogenated asphaltene

Chemical shift/ppm 145

Fig. 1 The 1H-NMR spectrum of LCT original and hydrogenated asphaltenes 146

Table1 Hydrogen atoms content of original and hydrogenated coal tar asphaltenes 147

Chemical shift /10-6 Symbol Attribution

Asphaltene

Original Hydrogenated

0.5–1.0 Hγ

γ position of aromatic side chain

and further hydrogen

0.0793 0.0318

7

1.0–2.0 Hβ β hydrogen of aromatic side chain 0.2879 0.1405

2.0–4.0 Hα α hydrogen of aromatic side chain 0.3045 0.5167

6.0–9.0 HA Aromatic hydrogen 0.3283 0.3110

Based on the data in Table 1, the average asphaltene structural parameters before and after LCT 148

hydrogenation were calculated by Brown-Ladner formula (Gu et al. 2006). The results are shown in 149

Table 2. 150

Table2 The calculation results of average structural parameters of original 151

and hydrogenated coal tar asphaltenes 152

Parameters Symbol

Original

asphaltene

hydrogenated

asphaltene

Aromaticity fA 0.737 0.727

Hydrogen substitution rate around aromatic ring σ 0.454 0.317

Condensation degree parameters of aromatic ring HAU/CA 0.531 0.629

Total atomic number CT 35.56 29.66

Number of aromatic carbon atoms CA 26.21 21.57

Total number of hydrogen atoms HT 29.81 25.12

Number of saturated carbon atoms Cs 9.35 8.09

Carbon atom numbers in aromatic rings Ci 12.28 7.27

Carbon atoms numbers in periphery

of the aromatic ring

Cap 13.93 14.30

α carbon atom numbers in aromatic ring system Cα 4.41 6.49

Aromatic ring number RA 7.14 5.64

Total number of rings Rt 8.96 8.32

Naphthenic ring number Rn 1.82 2.68

Carbon atom numbers in alkyl side chain n 2.73 2.18

Based on the calculation and analysis of the above data, the average molecular structure of MLCT 153

original and hydrogenated asphaltene were calculated and speculated as shown in Fig. 2. This original 154

8

asphaltene molecular structure is similar to the results of Schuler et al. (2015) obtained by atomic force 155

microscopy, which further explains the rationality of this asphaltene structures. 156

157

Fig. 2 The speculated average structural of original and hydrogenated coal tar asphaltenes 158

3 Simulation flow and parameter setting 159

The dimer composition model of MLCT asphaltene before and after hydrogenation and the composition 160

model of benzene, nitrobenzene, pyridine, quinoline and n-heptane were constructed in the 3D Atomistic 161

Document module in Materials Studio software, which were optimized by the molecular mechanics and 162

molecular dynamics. Then the simulated cell with periodic boundary conditions was constructed in the 163

Amorphous cell molecular mechanics module. The dynamics of the cell box was optimized under the 164

isotherm and isobaric (NPT) system. Finally, the MD simulation lasting 1500 ps was carried out on the 165

asphaltene aggregate before and after hydrogenation in different solvents at 298 K and 598 K, 166

respectively. The pre-1000 ps was used as the equilibrium process and the data of the last 500ps was 167

extracted for analysis and calculation. The temperature and pressure of the simulation process were set 168

to 298.15 K and 0.1 MPa. Compass force field was worked as the simulation force field, the 169

BERENDSEN method controls the simulation pressure, the ANDERSEN method controls the simulation 170

temperature, the GROPUBASED method controls the non-bond energy interaction in the simulation 171

process, the integration time is set to 1500 ps and the integration step is 1fs. 172

4 Results and discussion 173

4.1 Model of asphaltene aggregate 174

Based on the average molecular structure model of MLCT asphaltene before and after hydrogenation 175

obtained in section 2, the composition models of original and hydrogenated asphaltenes were constructed. 176

9

The kinetic optimization process is shown in Fig. 3. The most stable conformations of original and 177

hydrogenated asphaltenes were obtained after MD optimization shown in Fig. 4. 178

0

30

60

90

0 5 10 15 20 25 30 35 40 4580

120

160

200

240

Original Asphaltene

En

erg

y/k

cal·

mo

-1

Hydrogenated Asphaltene

Optimization step/ps

179

Fig. 3 Dynamic optimization state of LCT original and hydrogenated asphaltenes 180

Eight asphaltene molecules were randomly placed in Materials Studio and optimized by dynamics 181

program in Forcite module to obtain aggregation structure of asphaltene molecules, as shown in Fig. 5. 182

183

Fig. 4 The most stable conformations of original and hydrogenated coal tar asphaltenes 184

185

Fig. 5 Aggregation structural diagram of original and hydrogenation coal tar asphaltenes 186

10

As can be seen from Fig. 4, after structural optimization, the eight original asphaltene molecules 187

randomly placed before the simulation are transformed into a PAH stacking aggregate structure with 188

approximately parallel and orderly arrangement of 2-3 asphaltene molecules, indicating that MLCT 189

asphaltene molecules don’t exist in the form of a single molecule in the heavy oil system, and 2-3 190

asphaltene molecules accumulate in one direction to form small nano-aggregates. Multiple aggregates 191

self-assemble to form more complex nano-aggregates, which extend in different directions. Therefore, 192

in the MLCT colloid, the asphaltene molecules that make up the colloidal core are the aggregate of 193

several asphaltene molecules. The existence form of asphaltene molecules is the aggregation state of 194

microscopic partial order and macroscopic disorder. Therefore, the dimer of MLCT asphaltene was 195

selected as the research object in the follow-up study. 196

Hydrogenated asphaltene molecules shown in Fig. 5. an also form a PAH stacking aggregate structure, 197

which exists in the form of aggregates, but the number of molecules is relatively small. This may be due 198

to the reactions, such as naphthenic ring opening, aromatic ring saturation, alkyl side chain breaking and 199

removal of impurity atoms in the process of hydrogenation, resulting in the weakening of the interaction 200

between asphaltene molecules. In addition, according to the study of Ren et al. (2014), it was found that 201

the aggregate numbers of petroleum asphaltene were 4 to 6, which were slightly higher than that of 202

MLCT asphaltene. This probably because the small molecular weight, the small number of aromatic 203

rings and the small core of condensed aromatic rings of MLCT asphaltene, also the small interaction 204

between aromatic rings and the weak association between molecules. 205

4.2 MD simulation of asphaltene aggregate 206

The dimer composition models of MLCT asphaltene before and after hydrogenation and the composition 207

model of benzene, nitrobenzene, pyridine, quinoline and n-heptane were optimized by molecular 208

mechanics and molecular dynamics. The simulated unit cells with optimized periodic boundary 209

conditions were constructed. The structural models of original and hydrogenated asphaltenes dimer were 210

shown in Fig. 6. The original cell structure diagrams of solvent molecule and asphaltene molecule were 211

shown in Fig. 7 (taking original asphaltene as an example). 212

11

213

Fig. 6 Optimized configuration of original and hydrogenated coal tar asphaltene dimers 214

215

Fig. 7 The periodic protocell structures of asphaltene dimer in benzene, nitrobenzene, pyridine, 216

quinoline and n-heptane 217

As shown in Fig. 6, the optimized asphaltene has a layered structure, which verifies the classical theory 218

obtained from X-ray diffraction experimental data that the asphaltene aromatic sheets were stacked to 219

form a PAH stacking complex, which was further aggregated into micelles. The micelles were further 220

aggregated with each other to form aggregates. In addition, the interlayer spacing of the optimized 221

asphaltene aggregate measured by the measure/charge tool in Materials Studio were 1.99, 1.90 Å (before 222

hydrogenation) and 2.41, 2,32 Å (after hydrogenation), respectively. It can be seen that the relative error 223

between the asphaltene layer spacing measured by measure/charge and the experimental data were less 224

than ±4.8%. The layer spacing of hydrogenated asphaltene is larger than that of original asphaltene. 225

4.2.1 MD Analysis from structural angle of asphaltene aggregate 226

(1) Configuration of asphaltene aggregate 227

In different solvents, the kinetically optimized final configurations of asphaltene aggregates extracted 228

from periodic cells were shown in Fig. 8. It can be seen that the face-to-face layered structures of both 229

12

original and hydrogenated asphaltenes have not been completely destroyed in all kinds of solvents at 298 230

K. The structure of a single asphaltene molecule has little change indicating that the structure of MLCT 231

asphaltene is stable at normal temperature. The disbanding effect of various solvents on asphaltene is 232

poor. At high temperature of 598 K, the configurations of asphaltene in different solvents changed 233

obviously. The structure of single asphaltene molecule also changed indicating that the structure of 234

MLCT asphaltene was unstable at 598 K. Some solvents can destroy the intermolecular force so that the 235

asphaltene aggregate can be disbanded. The disbanding effect of coal tar asphaltene aggregate in different 236

solvents was shown in Table 3. 237

238

Fig. 8 The final configurations of asphaltene dimers in benzene, nitrobenzene, solvent-free state, 239

pyridine, quinoline and n-heptane 240

Table3 Aggregate state of coal tar asphaltenes dimers in different solvents 241

Solvent

Original asphaltene Hydrogenation asphaltene

Damage to the asphaltene

layer structure

Molecular

deformation

Damage to the

structure

Molecular

deformation

Benzene Destroyed Bending

Completely

destroyed

Small

Nitrobenzene Completely destroyed Bending

Completely

destroyed

Small

13

Solvent-free Intact None Intact None

Pyridine

Increased spacing,

slightly destroyed

Bending Intact Small

Quinoline

Increased spacing,

Destroyed,

Small Intact Small

n-heptane Intact None Intact None

(2) Molecular radius of gyration of asphaltene aggregate 242

The radius of gyration (Rog) represents the extension of the asphaltene molecule in the solvent. The 243

higher the Rog value is, the larger the molecular size is, the more the molecule stretches in the solvent. 244

Because the structural unit of asphaltene is centered on the condensed aromatic ring system with some 245

naphthene and alkyl side chains, the molecular Rog data are calculated based on the condensed aromatic 246

core of asphaltene molecules. The specific calculation formula is as follows (Chen et al. 2007): 247

𝑅𝑜𝑔 = √{(𝑥𝑎

2+𝑦𝑏2+𝑧𝑐

2)

𝑁} (1) 248

In the formula: xa、yb、zc represent the spatial coordinates of the atom, where the center of mass of the 249

asphaltene molecule as the origin. N represents the total number of atoms in the system. The Rogs of 250

asphaltenes before and after LCT hydrogenation in different solvents were shown in Fig. 9. 251

1.5

1.6

1.7

1.8

1.9

2.0

2.1

1.0

1.2

1.4

1.6

1.8

2.0

0 300 600 900 1200 15001.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

0 300 600 900 1200 15000.9

1.2

1.5

1.8

2.1

2.4

2.7

3.0

Solvent-free Qiunoline Pyridine n-heptane Benzene Nitrobenzene

Original 298 K

Hydrogenated 298 K

Original 598 K Ro

g/n

m

Time/ps

Hydrogenated 598 K

252

Fig. 9 Comparison of Rogs of asphaltene dimers in different solvents 253

14

As shown in Fig. 9, the Rogs of MLCT asphaltene were affected little by temperature in n-heptane solvent, 254

while in other solvents, the Rogs of MLCT asphaltene at 598 K was obviously larger than that at 298 K, 255

showing that temperature has almost no effect on the size of MLCT asphaltene in n-heptane solvent, 256

while in other solvents, the size of MLCT asphaltene is larger at high temperature. This is due to the fact 257

that the solvent dissolves asphaltene molecules more strongly and the alkyl side chain of asphaltene 258

molecules is more extended at high temperature, while n-heptane has almost no dissolution effect on 259

asphaltenes at any temperature. This law is not only consistent with the objective law, which confirms 260

the reliability of the simulation, but also proves that high temperature is beneficial to the disbanding of 261

MLCT asphaltenes from the view of molecular dynamics. 262

It can also be seen from Fig. 9 that the change of Rog of hydrogenation asphaltene was more obvious 263

than that of original asphaltene, which is due to the fact that the condensed aromatic core of 264

hydrogenation asphaltene was smaller than that of original asphaltene and the condensation degree of 265

aromatic ring is lower. The average number of aromatic rings of original asphaltene is 8.533 and the 266

condensation degree parameter of aromatic ring is 0.432, while the average number of aromatic rings of 267

hydrogenation asphaltene is 7.057, the condensation degree parameter of aromatic ring is 0.492. 268

Therefore, the morphological change of the side chain of hydrogenated asphaltene had a greater influence 269

on itself and the change of molecular Rog for hydrogenated asphaltene was more significant. 270

At 298 K, the solubility of asphaltene aggregate was poor and the change of molecular Rog was not 271

obvious. Thus, the molecular Rog of asphaltenes at 598 K were analyzed. The Rogs of original and 272

hydrogenated asphaltene in solvent at 598 K were as follows: Nitrobenzene > Quinoline > Benzene > 273

Pyridine > Solvent-free > n-heptane (Original asphaltene); Benzene > Nitrobenzene > Solvent-free> 274

Quinoline > Pyridine > n-heptane (Hydrogenation asphaltene). 275

4.2.2 MD Analysis from energy angle of asphaltene aggregate 276

In the COMPASS force field, the configuration energy (EC) of the aggregate system consists of two parts: 277

the no-bond energy (Eno-bond) and the bond energy (Einternal). Among them, Einternal is caused by molecular 278

deformation, including bond expansion energy (EB), bond angular energy (EA), bond torsion energy (ET) 279

and bond inversion energy (EI), Eno-bond caused by internal interactions, including van der Waals 280

15

interaction energy (Evdw) and electrostatic interaction energy (Ee). The calculation formulas are as follows 281

(Yue et al. 2016): 282

𝐸𝐶 = 𝐸𝑛𝑜−𝑏𝑜𝑛𝑑 + 𝐸𝑖𝑛𝑡𝑒𝑟𝑛𝑎𝑙 283

(2) 284

𝐸𝑖𝑛𝑡𝑒𝑟𝑛𝑎𝑙 = 𝐸𝐵 + 𝐸𝐴 + 𝐸𝑇 + 𝐸𝐼 (3) 285

𝐸𝑛𝑜−𝑏𝑜𝑛𝑑 = 𝐸𝑣𝑑𝑤 + 𝐸𝑒 286

(4) 287

𝐸𝑣𝑑𝑤 = ∑ 𝜀𝑖𝑗[2𝑖,𝑗 (𝑟𝑖𝑗

0

𝑟𝑖𝑗)9 − 3(

𝑟𝑖𝑗0

𝑟𝑖𝑗)6] (5) 288

𝐸𝑒 =∑ 𝑞𝑖𝑞𝑗𝑖,𝑗

𝜀𝑖𝑗𝑟𝑖𝑗 (6) 289

In the formula, εij is the energy parameter, rij0 represents the stable distance between atoms, qi and qj 290

represent the atomic charge. Based on the MD simulation of the asphaltene aggregate before and after 291

hydrogenation, the final configuration of asphaltene aggregate was statistically analyzed by using the 292

analysis tool of the Forcite module. The Eno-bond and Einternal of the asphaltene aggregates in different 293

solvents before and after hydrogenation were analyzed at 298 K and 598 K, respectively. 294

(1) Interaction energy 295

The comparisons of Eno-bond of asphaltene aggregates in different solvents were shown in Fig. 10. It can 296

be seen from Fig. 10 that the Evdw and Eno-bond of original and hydrogenated asphaltenes in various solvents 297

had the same trend at 298 K and 598 K, indicating that the total interaction between coal tar asphaltene 298

molecules mainly comes from Evdw. The driving force of asphaltene aggregates before and after 299

hydrogenation was Evdw, which may be due to the ring opening and heteroatom removing reactions during 300

asphaltene hydrogenation process leading to the decrease of asphaltene molecular weight, aromaticity, 301

degree of aromatic ring condensation and the content of heteroatom. However, the structural change of 302

asphaltene was not enough to have an important effect on the aggregate driving force considering of the 303

lack of hydrogenation depth. Li et al. (2007) found that the main driving force of petroleum asphaltene 304

aggregate came from electrostatic interaction, which is due to the large interlayer spacing of petroleum 305

asphaltene nano-aggregates. Thus, the long-range Ee was dominant. 306

It can be seen from Fig. 10 that the electrostatic energy of asphaltene before and after hydrogenation was 307

16

less than 0 under the solvent-free simulation condition. But the absolute values of Evdw and Ee of 308

hydrogenated asphaltenes were smaller than that of original asphaltene, also more affected by 309

temperature. This shows that the intermolecular interaction of hydrogenated asphaltene is weak and the 310

structure of the aggregate is unstable. The Ee of hydrogenated asphaltene was lower than that of original 311

asphaltene. The content of heteroatoms is obviously lower than that of original asphaltene. It is speculated 312

that the existence of heteroatoms may be one of the reasons for the aggregation of asphaltene. 313

There was no significant difference in the interaction energy of asphaltene in various solvents at 298 K, 314

which indicates that the asphaltene aggregate is stable at 298 K and the aggregate structure is not easy to 315

be destroyed. When the temperature increases to 598 K, the absolute values of EC, Evdw and Ee of 316

asphaltene decreased, proofing that the asphaltene aggregate is unstable and the intermolecular 317

interaction force is weakened at high temperature, which is beneficial to molecular disbanding. Different 318

solvents have different effects on the disassembly of asphaltene aggregate. The interaction energy of 319

original asphaltene decreased significantly in the solvents of nitrobenzene, quinoline, benzene and 320

pyridine. The interaction energy of hydrogenated asphaltene decreased significantly in the solvent of 321

nitrobenzene indicating that the aggregate structure of asphaltene was destroyed. This is consistent with 322

the previous configuration analysis of asphaltene in various solvents. Therefore, the disbanding effects 323

of various solvents on asphaltenes were as follows: Nitrobenzene > Quinoline > Benzene > Pyridine > 324

Solvent-free > n-heptane (Original asphaltene); Benzene > Nitrobenzene > Solvent-free> Quinoline > 325

Pyridine > n-heptane (Hydrogenation asphaltene). 326

17

-40

-30

-20

-10

0

10

-40

-30

-20

-10

0

10

Eno-bond

Evdw

Ee

n-hep

tane

Quin

olin

e

Byr

idin

e

Solve

nt-fr

ee

Nitr

oben

zene

En

erg

y/k

acl

·mo

l-1

Original 298 K

Ben

zene

n-hep

tane

Quin

olin

e

Byr

idin

e

Solve

nt-fr

ee

Nitr

oben

zene

Ben

zene

Hydrogenated 298 K

Original 598 K

Hydrogenated 598 K

327

Fig. 10 Comparison of non-bond energy of asphaltene aggregates in different solvents 328

Benzene and nitrobenzene solvents can seriously destroy the aggregate structure of original asphaltene 329

and completely break the binding of asphaltene after hydrogenation. This shows that both solvent of 330

benzene and asphaltene form a strong Evdw before and after hydrogenation, so the Evdw within asphaltene 331

molecules is destroyed, indicating that the Evdw is an important driving force for the aggregate and 332

precipitation of asphaltenes. 333

Due to the large electronegativity of -NO3 group in nitrobenzene, the polarity of nitrobenzene is larger. 334

The asphaltene is the most polar component in coal tar system (Sun et al. 2015). According to the similar 335

miscibility principle, dipoles and electrostatic interactions are formed between asphaltene and 336

nitrobenzene molecules. Therefore, the Ee within asphaltene molecules is weakened and the asphaltene 337

aggregate structure is destroyed. 338

Benzene and nitrobenzene solvents can completely dissociate the hydrogenation asphaltene and seriously 339

destroy the aggregate structure of the original asphaltene. However, the damage of nitrobenzene to the 340

aggregate structure of the original asphaltene is more significant than that in the benzene solvent. This is 341

because nitrobenzene has strong polarity and aromaticity. It can not only interact with asphaltenes by 342

Evdw, but also interact with asphaltenes by Ee, which further affects the structure of asphaltene aggregate. 343

This shows that the Ee is one of the reasons leading to the association and settlement process of 344

asphaltenes. 345

18

Quinoline and pyridine solvents can destroy the aggregate structure of original asphaltene, but have little 346

effect on asphaltene aggregate after hydrogenation. It can be explained that the pyridine and quinoline 347

are strongly alkaline solvents (Lu et al. 2008). The content of heteroatoms in original asphaltene is high 348

producing hydrogen-bonded acid, which forms the hydrogen bond with pyridine or quinoline molecules, 349

furtherly destroying the aggregate structure of asphaltene. However, the number of heteroatoms in 350

hydrogenation asphaltene decreased. So, the disbanding effect on hydrogenation asphaltene is not 351

obvious. The n-heptane solvent has no effect on original and hydrogenated asphaltenes because of non-352

polarity characteristic. 353

To sum up, aromatic compounds, strong basic compounds and strong polar compounds have good 354

dissociation effect on MLCT asphaltenes and can effectively inhibit the aggregation and sedimentation 355

of asphaltenes. Therefore, when designing or synthesizing inhibitors and scavengers for asphaltene 356

sedimentation, we should comprehensively consider the aromaticity, alkalinity and polarity of substances. 357

（2）Molecular deformation energy 358

Einternal reflects the deformation of molecular structure to a certain extent. The Einternal of asphaltenes in 359

solvents before and after MLCT hydrogenation were shown in Figure 9 at 298 K and 598 K. As Fig. 11 360

showed that the Einternal of the original and hydrogenation asphaltene aggregates was greatly affected by 361

temperature. The Einternal of hydrogenation asphaltene was smaller and more affected by temperature. 362

When the temperature increased from 298 K to 598 K, the Einternal of the original asphaltene aggregate 363

increased from 1300 to 1400 kcal·mol-1 and that of the hydrogenated asphaltene aggregate increased 364

from 680 to kcal·mol-1, indicating that temperature has a significant effect on the stability of the 365

asphaltene aggregate, especially has a greater effect on the hydrogenation asphaltene. This is due to the 366

small number of aromatic rings of asphaltene after hydrogenation. The layered structure between 367

asphaltene molecules is unstable and easy to be destroyed. 368

When the temperature is 298 K, the Einternal of asphaltene aggregate before and after hydrogenation 369

changed little in different solvents showing that the structure of asphaltene aggregate is stable and keep 370

small solubility in all kinds of solvents is normal temperature. When the temperature was 598 K, the 371

Einternal and ET of original asphaltene aggregate in nitrobenzene, benzene and pyridine solvents increased 372

obviously demonstrating that the C-C bonds in original asphaltene molecules in nitrobenzene, benzene 373

19

and pyridine solvents are twisted leading to the deformation of asphaltene aggregate. Although the Einternal 374

of hydrogenated asphaltene aggregate were different but with little difference in various solvents showing 375

that the deformation of hydrogenated asphaltene aggregate in solvents is very small. 376

0

300

600

900

1200

1500

0

200

400

600

800

0

300

600

900

1200

1500

1800

0

300

600

900

En

erg

y/k

acl·

mo

l-1

Einternal

EB E

A E

T E

I

Original 298 K

Hydrogenated 298 K

n-hep

tane

Quin

olin

e

Byr

idin

e

Solve

nt-fr

ee

Nitr

oben

zene

Ben

zene

n-hep

tane

Quin

olin

e

Byr

idin

e

Solve

nt-fr

ee

Nitr

oben

zene

Ben

zene

Original 598 K

Hydrogenated 598 K

377

Fig. 11 Comparison of bond energy of asphaltene aggregate in different solvents 378

5 Conclusion 379

The average molecular structure models of MLCT asphaltene before and after hydrogenation were 380

obtained by the characterization of the asphaltenes. Also, the aggregation structure models of MLCT 381

asphaltenes before and after hydrogenation were constructed in Materials Studio 8.0 software. The 382

optimal conformation was obtained through the kinetic optimization. The comparative analysis of 383

structural characteristics between original and hydrogenated asphaltenes were carried out. In order to get 384

the effects of different solvents on the structural characteristics, aggregation state, accumulation 385

characteristics and dominant force of asphaltene aggregates, the final configuration, molecular rotation 386

radius and interaction energy of asphaltene aggerates before and after hydrogenation were analyzed by 387

molecular dynamic simulation in different solvents. The conclusions are as follows: 388

(1) The unit structure and aggregate structure models of MLCT asphaltene before and after hydrogenation 389

were constructed and the kinetics optimization and MD simulation were carried out. After structural 390

optimization, the original asphaltene molecules become 2-3 asphaltene molecules with approximately 391

20

parallel, orderly arrangement and PAH stacking aggregate structure, while multiple aggregates self-392

assemble to form more complex nano-aggregates, extending in different directions. The MLCT 393

asphaltene molecules after hydrogenation also form a PAH stacking association structure, which exists 394

in the form of aggregates, but the stacking number of asphaltene molecules is less than that of original 395

asphaltene. (2) The total interaction between MLCT original and hydrogenated asphaltenes mainly come 396

from the intermolecular van der Waals interaction, while the main driving force of petroleum asphaltene 397

aggregate comes from electrostatic interaction. Compared with original asphaltene, the absolute values 398

of van der Waals energy and electrostatic energy of hydrogenated asphaltene are smaller and more 399

affected by temperature. The intermolecular interaction of hydrogenation asphaltene is weaker and the 400

structure of hydrogenation asphaltene is unstable. (3) The disbanding effect of solvents on asphaltene is 401

as follows: Nitrobenzene > Quinoline > Benzene > Pyridine > Solvent-free > n-heptane (Original 402

asphaltene); Benzene > Nitrobenzene > Solvent-free> Quinoline > Pyridine > n-heptane (Hydrogenation 403

asphaltene). (4) The aromatic compounds, strong basic compounds and strong polar compounds have 404

good dissociation effect on MLCT asphaltenes and can effectively inhibit the aggregation and 405

sedimentation of asphaltenes. Therefore, when designing or synthesizing inhibitors and scavengers for 406

asphaltene sedimentation, it needs comprehensively consider the aromaticity, alkalinity and polarity of 407

substances. 408

The research results can not only provide a theoretical basis for the fluctuation control of product quality, 409

blockage of pipeline heat exchangers and oil storage in the current coal tar processing process, but also 410

provide technical support for continuous and stable operation and macro-control of industrial units. 411

Acknowledgments 412

Declarations 413

Funding: The financial support of this work was provided by the National Natural Science Foundation 414

of China (21978237) 415

Conflicts of interest/Competing interests: N/A 416

Availability of data and material: N/A 417

Code availability: N/A 418

21

Authors' contributions: 419

Xiaokang Han: Writing - Review & Editing, Conceptualization, Methodology, Writing - original draft. 420

Wenzho Yan: Visualization, Investigation, Writing - original draft. 421

Lu Xu: Software, Investigation, Writing - original draft. 422

Dong Li: Validation, Formal analysis. 423

Yonghong Zhu: Resources, Data curation. 424

References 425

1. Sun JM, Li D, Yao RQ (2015) Modeling the hydrotreatment of full range medium temperature 426

coal tar by using a lumping kinetic approach. Reaction Kinetics Mechanisms and Catalysis 114: 427

451-471. 428

2. Zhu YH, Zhang YH, Dan Y (2015) Optimization of reaction variables and macrokinetics for the 429

hydrodeoxygenation of full range low temperature coal tar. Reaction Kinetics Mechanisms and 430

Catalysis 116: 433-450. 431

3. Zhu YH, Huang JL, Dan Y (2016) Analysis and characterization of medium and low temperature 432

coal tar asphaltene, Acta Petrolei Sinica (Petroleum Processing Section) 32: 334-342. 433

4. Berrgmann U, Groenzin H, Mullins O C (2003) Carbon K-edge X-ray raman spectroscopy 434

supports simple, yet powerful description of aromatic hydrocarbons and asphaltenes. Chemical 435

Physics Letters 369: 184-191. 436

5. Sabbah H, Morrow A L, Pomerantz A E (2011) Evidence for island structures as the dominant 437

architecture of asphaltenes. Energy Fuels 25: 1597-1604. 438

6. Karimi A, Qian K, Olmstead W N (2011) Quantitative evidence for bridged structures in 439

asphaltenes by thin film pyrolysis. Energy Fuels, 25: 3581-3589. 440

7. Wargadalam V J, Norinaga K, Iino M (2002) Size and shape of a coal asphaltene studied by 441

viscosity and diffusion coefficient measurements 81: 1403-1407. 442

8. Alvarez-Ramírez, Fernando, Ruiz-Morales Y (2013) Island versus archipelago architecture for 443

asphaltenes: Polycyclic Aromatic Hydrocarbon Dimer Theoretical Studies. Energy Fuels, 27: 444

1791-1808. 445

9. Schuler B, Meyer G, Pena D (2015) Unraveling the molecular structures of asphaltenes by atomic 446

22

force microscopy. Journal of the American Chemical Society, 137: 9870-9876. 447

10. Schuler B, Zhang Y, Collazos S, Fatayer S, Meyer G, Perez D, Guitian E, Harper M R, Kashmiris 448

J D, Pena D (2017) Characterizing aliphatic moieties in hydrocarbons with atomic force microscopy. 449

Chemical science 8: 2315−2320. 450

11. Schuler B, Zhang Y, liu Fang, Pomerantz A E, Andrews A B, Gross L, Puachard V, Banerjee, S, 451

Mullins O C (2020) Overview of Asphaltene Nanostructures and Thermodynamic Applications. 452

Energy & Fuels 34: 15082-15105. 453

12. Gray M R, Tykwinski R R, Stryker J M (2011) Supramolecular assembly model for aggregation 454

of petroleum asphaltenes. Energy Fuels, 25: 3125-3134. 455

13. Xu C, Shi Q (2016) Structure and modeling of complex petroleum mixture. United Kingdom, 456

Springer International Publishing 9-16. 457

14. Jian C, Tang T, Bhattacharjee S (2014) Molecular dynamics investigation on the aggregation of 458

Violanthrone-based model asphaltenes in toluene. Energy Fuels, 28: 3604-3613. 459

15. Dickie J P, Yen T F (1967) Macrostructures of the asphaltic fractions by various instrumental 460

methods. Analytical Chemistry 39: 1847-1852. 461

16. Mullins O C (2010) The modified Yen model. Energy Fuels 24: 2179-2207. 462

17. Redelius P G (2000) Solubility parameters and bitumen. Fuel 79: 27−35. 463

18. Per Redelius (2007) Hansen Solubility Parameters of Asphalt, Bitumen and Crude Oils in the 464

Hansen solubility Handbook, CRC Press, Boca Raton, FL:151-176. 465

19. Acevedo S C, Castro A, Vasquez E, Marcano F, Ranaudo M (2010) Investigation of Physical 466

Chemistry Properties of Asphaltenes Using Solubility Parameters of Asphaltenes and Their Fractions 467

A1 and A2. Energy Fuels 24: 5921−5933. 468

20. Eyssautier J, Levitz P, Espinat D, Jestin J, Gummel J & Grillo I (2011) Insight into asphaltene 469

nanoaggregate structure inferred by small angle neutron and x-ray scattering. Journal of Physical 470

Chemistry B 115: 6827. 471

21. Eyssautier J, Henaut I, Levitz P, Espinat D, Barre L (2012) Organization of Asphaltenes in a Vacuum 472

Residue: A Small-Angle X-ray Scattering (SAXS)−Viscosity Approach at High Temperatures. Energy 473

Fuels 26: 2696−2704. 474

23

22. Eyssautier J, Espinat D, Gummel J, Levitz P, Becerra M, Shaw J, Barre L (2012) Mesoscale 475

Organization in a Physically Separated Vacuum Residue: Comparison to Asphaltenes in a Simple Solvent. 476

Energy Fuels 26: 2680−2687. 477

23. Gray M R, Tykwinski R R, Stryker J M (2011) Supramolecular assembly model for aggregation 478

of petroleum asphaltenes. Energy Fuels, 25: 3125-3134. 479

24. Pacheco-Sanchez J H, Alvarez-Ra mirez F, Martínez-Magadan J M (2004) Morphology of 480

aggregated asphaltene structural models. Energy Fuels 18: 1676-1686. 481

25. Jian C, Tang T, Bhattacharjee S (2014) Molecular dynamics investigation on the aggregation of 482

violanthrone-based model asphaltenes in toluene. Energy Fuels 28: 3604-3613. 483

26. Zhang L, Greenfield M (2007) Molecular orientation in model asphalts using molecular 484

simulation. Energy Fuels 21: 1102-1111. 485

27. Carauta A N M, Seidl P R, Chrisman E C A N (2005) Modeling solvent effects on asphaltene 486

dimers. Energy Fuels 19: 1245-1251. 487

28. Alvarezramirez F, Edgar Ramirezjaramillo A, Ruizmorales Y (2006) Calculation of the 488

interaction potential curve between asphaltene-asphaltene, asphaltene-resin, and resin-489

resinsystems using density functional theory. Academic Radiology 2: S432-S433. 490

29. Costa L M, Stoyanov S R, Gusarov S (2012) Density functional theory investigation of the 491

contributions of π–π stacking and hydrogen-bonding interactions to the aggregation of model 492

asphaltene compounds. Energy Fuels 26: 2727-2735. 493

30. Lima F C D A, Alvim R D S, Miranda C R (2017) From single asphaltenes and resins to 494

nanoaggregates: a computational study. Energy Fuels 31: 11743-11754. 495

31. Li D, Cui W G, Zhang X, Meng Q, Zhou Q C, Ma B Q, Li W H. Production of clean fuels by catalytic 496

hydrotreating a low temperature coal tar distillate in a pilot-scale reactor. Energy Fuels 2017, 3110: 497

11495-11508. 498

32. Lalvani S B, Muchmore C B, Koropchak J (1991) Lignin-augmented coal depolymerization 499

under mild reaction conditions. Energy Fuels 5: 347-352. 500

33. Gu XH, Shi SD, Zhou M (2006) Study on the molecular structure of asphaltene in the residue 501

of direct liquefaction of Shenhua coal. Journal of Coal Industry 31: 785-789. 502

24

34. Ren Q, Zhou H, Dai ZY (2014) Multi-scale study of the aggregation structure of heavy oil 503

asphaltene. International Conference on Molecular Simulation and Information Technology 504

applications[C] 389-399. 505

35. Chen ZL, Xu WR, Tang LD (2007) Theory and practice of molecular simulation. Chemical 506

Industry Press 123-131. 507

36. Li YF, Lu GW, Sun W (2007) Molecular dynamics study on the combination of petroleum 508

bitumen, Acta Petrolei Sinica (Petroleum Processing Section), 23: 28-34. 509

37. Sun Z H, Li D, Ma H X (2015) Characterization of asphaltene isolated from low-temperature 510

coal tar. Fuel Processing Technology 138: 413-418. 511

38. Lu GW, Li YF, Song H (2008) Microscopic mechanism of petroleum asphaltene aggregation 512

and sedimentation. Petroleum exploration and development 35: 73-78. 513