You are on page 1of 41

1

3 Please cite this article:


4 Pahlavan, F., Hosseinnezhad, S., Samieadel, A., Hung, A., & Fini, E. (2019). Fused Aromatics To
5 Restore Molecular Packing of Aged Bituminous Materials. Industrial & Engineering Chemistry
6 Research, 58(27), 11939-11953.

1
8 Fused Aromatics to Restore Molecular Packing of Aged Bituminous Materials

9 Farideh Pahlavan,a Shahrzad Hosseinnezhad,a Alireza Samieadel, bAlbert Hung,a Elham Finic*
10 * Corresponding author. e-mail: elifini@gmail.com (E. H. Fini).

11 a. Division of Research and Economic Development, North Carolina A&T State University, 1601 E. Market St.,
12 Greensboro, NC 27411, USA

13 b. School of Sustainable Engineering and the Built Environment, Ira A. Fulton Schools of Engineering, Arizona
14 State University 660 S. College Avenue, Tempe, AZ 85287-3005

15 c. Arizona State University, 660 S. College Avenue, Tempe, AZ 85287-3005, Phone: 480-965-4273, USA

16 Abstract

17 This study incorporates computational and laboratory experiments to determine specific interaction

18 mechanisms between conjugated structures (interrupted and continuous conjugates) and selected

19 polyaromatics hydrocarbons (PAH) stacks such as those found in oxidized asphaltene molecules. The

20 theoretical results obtained by quantum-mechanical calculations and molecular dynamics simulations

21 show that both continuous and interrupted conjugates are effective to weaken the π-inter-sheet forces and

22 decrease the size of PAH nano-aggregates, with interrupted conjugates being more efficient than

23 continuous conjugates to exfoliate PAH stacks. Due to their limited partial entrance into the PAH stacks,

24 interrupted conjugates are not trapped inside the PAH interlayer, and can easily detach from the stacks.

25 This allows an interrupted conjugate not to be consumed in interaction with one PAH stack, making the

26 interrupted conjugate available to affect many other stacks in the matrix. The higher exfoliation capacity

27 of the interrupted aromatic conjugates is also evidenced in the results of our experiments based on UV-

28 Vis and rheometry analyses. This is reflected in a higher reduction in the polydispersity index and a

29 higher increase in both the crossover frequency and crossover modulus of aged bitumen (containing

30 stacks of large PAH referred to as asphaltene nano-aggregates) when it is doped with an interrupted

31 conjugate (fluoranthene) compared to when it is doped with a continuous conjugate (pyrene). The study

32 results contribute to the body of knowledge by providing an in-depth understanding of how variation in

2
33 the π-electron distribution of conjugated structures affects their efficacy to exfoliated self-assembled

34 stacked structures of polyaromatic hydrocarbons such as asphaltenes.

35 Keywords: Polyaromatics nano-aggregate, self-aggregation, exfoliation, aromatic conjugate, π-

36 conjugation, asphaltene

37 1. Introduction

38 Polycyclic aromatic hydrocarbons (PAHs) are a large category of organic compounds with fused benzene

39 rings arranged in various configurations. Carbon allotropes are the main members of this family,

40 including graphene,1 fullerene,2 and carbon nanotube (CNT).3 Graphene consists of fused sp2 hybridized

41 rings arranged in a two-dimensional layer, which is the basis of the other dimensions of the carbon

42 nanomaterial. Graphene sheets can be wrapped up to form zero-dimensional fullerene buckyballs, rolled

43 into one-dimensional carbon nanotubes, or stacked as three-dimensional graphite.4 In addition, we can put

44 molecules with graphene-like structures in the PAH class, such as the asphaltenes in bitumen.5

45 Electronically, the honeycomb structure of the PAH components as an extended π-electron

46 system can easily bind to the other π-systems through noncovalent π-stacking interactions.6-8 So,

47 polyaromatic sheets and tubes have a tendency to self-aggregate in the form of graphite-like stacking or

48 CNT bundles, respectively. This behavior reduces solvents' compatibility, preventing them from being

49 exploited for achieving certain electronic properties, which hinders their widespread application.9-10 Thus,

50 PAH aggregates need to be exfoliated directly via sonication in an aqueous solution of surfactants and

51 then stabilized through either noncovalent or covalent alteration in the carbonic layers using small

52 molecules.11-15 Despite the high dispersion stability obtained through covalent functionalization, this

53 approach has several disadvantages: structural damage, loss of inherent properties, and shortening of

54 tubes.16-19 Therefore, noncovalent functionalization with minor structural and electronic defects is a

55 superior approach to maintain the intrinsic properties of carbon nanoparticles.18, 20-21

3
56 Disturbing the π-π interfacial interactions between PAH fragments can be a key strategy to

57 achieve deagglomeration of self-aggregated polyaromatics, to produce high-throughput and high-quality

58 PAHs. Molecules containing small π-conjugated aromatic groups are capable of forming directional π-π

59 interactions with the graphenic surface of the PAHs, consequently interrupting the aggregation of the

60 PAHs.6, 22-25 Through this methodology, aromatic conjugate parts act as anchoring units for adsorption of

61 surfactants onto the hydrophobic surface of graphene and CNT to enhance their chemical compatibility

62 and dissolution properties.6, 26-27


Owing to the similarity in the basis of the atomic structure of

63 polyaromatics, this strategy is also applicable to other PAHs such as those found in bitumen, including

64 asphaltenes.

65 Asphaltene fragments have an architecture of 7-9 5, 28 condensed aromatic rings with heteroatoms,

66 which makes them highly polarizable with a strong tendency for self-assembly.29-30 Their structures

67 become even more polar with the chemical changes in the asphaltene structure that occur during oxidative

68 aging. This, in turn, increases the tendency for self-assembly and intermolecular associations in

69 asphaltenes, promoting the formation of nano-aggregates followed by agglomeration.31 The

70 agglomeration has considerable negative effects on the physical and rheological properties of bitumen and

71 consequently on the performance of asphalt pavement.32 To restore aged bitumen so it has the desirable

72 physicochemical and rheological properties of bitumen, there is a need to reduce the self-assembly and

73 size of asphaltene agglomerates in an attempt to enhance the dispersion of nano-aggregates in the asphalt

74 matrix. Environmental and economic concerns have focused increased attention on attempting to restore

75 oxidized asphaltene to have the original properties of asphaltene, using modifiers referred to as

76 rejuvenators.33-36

77 In the present study, we report a novel strategy to introduce the effective electronic and structural

78 factors of aromatic conjugates in the deagglomeration of polyaromatic assemblies. We hypothesize that

79 aromatic conjugates with different degrees of π-circulation, continuous and interrupted, have different

80 behaviors to intercalate dimerized asphaltenes and disturb their π-π stacking. To begin, using quantum

4
81 studies at the density functional theory (DFT) level, the intermolecular interactions between asphaltenes

82 are characterized before and after the addition of two types of conjugated exfoliants. We continue

83 theoretical studying using molecular dynamic (MD) simulations of the alteration of asphaltene stacking in

84 the presence of continuous and interrupted aromatic conjugates. We use experimental and rheological

85 measurements to support our computational findings about the correlation between the π-conjugation of

86 exfoliates and their ability to deaggregate PAH assemblies. The goal of this study is to contribute a basis

87 for investigators to select proper exfoliating agents for PAH assemblies, while reducing the need for

88 external mechanical forces and consequently decreasing the electronic and structural defects they cause.

89 2. Methods and Materials

90 Quantum-Mechanical Calculations. Quantum-mechanical calculations were performed at the framework

91 of density functional theory applying dispersion corrections (DFT-D), implemented in the DMol3

92 module37-38 of the Accelrys Materials Studio program package, version 6.0. The density functional used in

93 this study is the Perdew-Burke-Ernzerhof (PBE)39 formulation of the generalized gradient approximation

94 (GGA). Owing to the essence of our polyaromatic stacking systems, we included Grimme’s dispersion

95 correction40 in our calculations (PBE-D) to treat the long-range behavior of the functional.

96 DMol3 calculations were carried out at the “fine grid” level for the matrix numerical integrations.

97 The optimization was performed to reach the convergence criteria of 1.0 × 10−5 hartree for the energy

98 gradient, 2.0 × 10−3hartree Å-1 for the maximum force, and 5.0 × 10−3 Å for the maximum

99 displacement. A double numerical quality basis set with polarized function (DNP) was chosen for all

100 atoms.

101 The thermodynamic stability of the interacting complex was estimated through a binding energy

102 (Ebind) calculation. Ebind is the energy difference between the complex and its monomers in the lowest

103 energy state, 𝐸𝐶𝑜𝑚𝑝𝑙𝑒𝑥 and 𝐸𝐹𝑟𝑎𝑔𝑚𝑒𝑛𝑡 , respectively.

104 𝐸𝑏𝑖𝑛𝑑 = 𝐸𝐶𝑜𝑚𝑝𝑙𝑒𝑥 − (𝐸𝐹𝑟𝑎𝑔𝑚𝑒𝑛𝑡 1 + 𝐸𝐹𝑟𝑎𝑔𝑚𝑒𝑛𝑡 2 ) (1)

5
105 In this paper, we calculated binding energy for two series of systems; (1) between oxidized asphaltene

106 fragments in the dimer before and after addition of conjugated molecules as disrupting agents, Table 1,

107 (2) between each conjugated molecule and asphaltene monomer, Table 2. In the first category, 𝐸𝐹𝑟𝑎𝑔𝑚𝑒𝑛𝑡

108 is total energy of the oxidized asphaltene monomer and 𝐸𝐶𝑜𝑚𝑝𝑙𝑒𝑥 is total energy of the oxidized

109 asphaltene dimer isolated from the exfoliated complex (Asphaltene2/conjugated exfoliant).

110 The calculation of energy values took into account the corrections of basis set superposition error (BSSE)

111 evaluated by means of the counterpoise (CP) method41.

112 The harmonic oscillator model of aromaticity (HOMA) index is a widely used parameter of

113 aromaticity that is based on the molecular geometry (bond lengths). The HOMA index was originally

114 proposed by Kruszewski and Krygowski42 and then introduced as the generalized form:43

1 2
115 𝐻𝑂𝑀𝐴 = 1 − 𝑛 ∑𝑛𝑗=1 𝛼𝑖 (𝑅𝑜𝑝𝑡,𝑖 − 𝑅𝑗 ) (2)

116 where n is the number of bonds forming the ring, 𝛼𝑖 and 𝑅𝑜𝑝𝑡 are pre-calculated constants dependent on

117 the type of a given bond. Indeed, 𝛼𝑖 is a normalization constant and 𝑅𝑜𝑝𝑡 is the optimal bond length, for

118 which the compression to the double bond and extension to the single bond take the same energy. j refers

119 to the atom next to atom i. HOMA equals 1 for the fully aromatic molecules in which all bond lengths are

120 identical to the optimal value (𝑅𝑜𝑝𝑡 ). So, the HOMA value is close to 1 for aromatic systems, close to 0

121 for non-aromatic systems, and negative for anti-aromatic rings. The HOMA calculation only requires

122 molecular coordinates. So, we generated wavefunction (.wfn) files for all selected aromatic conjugates as

123 exfoliants at the B3LYP/6-31G* level of theory using the Gaussian09 program44 and employed them as

124 input files for the Multiwfn package45 to calculate HOMA values and contributions from each atom pair.

125 As another test for the degree of π-conjugation, the quantum theory of atoms in molecules

126 (QTAIM) developed by Bader46 was utilized. The QTAIM analysis was carried out on the optimized

127 structures of interrupted aromatic conjugates to reveal the nature of carbonic linkages between the

128 benzenoids in these components. The eigenvalues of the Hessian matrix of electron density (𝜆𝑖 , 𝑖 =

6
129 1, 2, 3) at bond critical points (BCPs) were computed based on Bader’s QTAIM using the AIM2000

130 program.47

131 The main criterion to evaluate the effect of conjugated exfoliants on asphaltene aggregation is an

132 alteration in the π-π interactions between asphaltene monomers. Accordingly, noncovalent interactions

133 (NCI) between asphaltenes in their assemblies were characterized by visualizing reduced density gradient

134 (RDG) isosurfaces. Tracing the changes in the RDG in low-density regions is the basis of this method.48

135 The RDG (or s) quantities are calculated by the following equation:

1 |∇𝜌|
136 𝑠 = 2(3𝜋2 )1⁄3 𝜌4⁄3 (3)

137 where 𝜌 and ∇𝜌 are density and its first derivative, respectively. The NCI approach for the weak

138 interactions is insensitive to the computational method,48 so we used the Gaussian computations near the

139 level of DMol3 calculations performed in this study to generate electron densities. Accordingly, we used

140 the PBEPBE/6-31G* method in the Gaussian package to reproduce the electron density for the interacting

141 systems optimized by DMol3. To visualize the RDG surfaces, we used the Visual Molecular Dynamics

142 (VMD) program.49

143 Molecular Dynamics Simulation. The model was built in the MedeA® environment using the molecular

144 builder to perform a step-by-step construction of different structures with attached alkyl chains and

145 different functional groups.50 The PCFF+ forcefield was used in this study; this is an extension of the

146 PCFF forcefield and is an all-atom forcefield designed to provide excellent accuracy on hydrocarbon and

147 liquid modeling from ab initio simulations.51 This forcefield includes a Lenard-Jones 9-6 potential for

148 intermolecular and intramolecular interactions and specific stretching, bending, and torsion terms to

149 involve 1-2, 1-3, and 1-4 interactions. The initial ensemble was created using supercell features of the

150 MedeA® environment.

151 A cubic cell of 1728 heptane molecules was built by replicating the solvent molecule 12 times in

152 the x-, y-, and z-directions. Similarly, a slab of 18 oxidized asphaltene molecules was merged into the

153 solvent cell (9.5 wt% of solvent concentration) to form the initial simulation box with periodic boundary

7
154 condition applied to the simulation box. The simulation is composed of two LAMMPS stages, starting

155 with a minimization using the conjugate gradients method at a constant volume with a low average

156 density to avoid molecular overlaps. After the minimization phase, an NVT ensemble with a temperature

157 of 800 K and an annealing stage using an NPT ensemble with a decrease in pressure from 200 bar to 1 bar

158 at 350 K for 500 ps was performed. The high-temperature and high-pressure ensembles were used to

159 accelerate the convergence toward the liquid solvent density. The second stage of LAMMPS starts with

160 an NVT ensemble with a temperature of 350 K (target temperature) for a better equilibration with no

161 confining pressure for 2 ns, followed by an NPT for 20 ns. A Nose-Hoover thermostat and barostat52 were

162 utilized throughout all the simulation stages, and the time step was set to 1 fs. In all simulations, the short-

163 range interactions were calculated directly, while long-range interactions were measured with the particle-

164 particle-particle-mesh (PPPM) method. Non-bonded terms were calculated with a simple cutoff of 9.5 Å.

165 To investigate the aggregation behavior of oxidized asphaltenes in the presence of aromatic

166 conjugates and plausible exfoliation phenomena, pyrene and fluoranthene were doped to the system of

167 oxidized asphaltene at 10 wt% concentration. The simulation was continued for another 20 ns in the NPT

168 ensemble; the average temperature and pressure were monitored during the simulation to ensure the

169 system was in equilibrium. During the simulation, the coordinates for the center of mass of asphaltene

170 molecules were recorded to study aggregation, and the radial distribution function result was calculated

171 for the most-centered carbon atom of asphaltenes.

172 Materials and Experiment Methods. A Superpave PG 64-22 bitumen (Sharpe Brothers, Greensboro, NC)

173 was aged in the laboratory following the method documented by Bowers et al., to produce oxidized

174 asphaltene representing aged asphalt in highway pavements at the end of service life.53 In this method,

175 aging in a rolling thin-film oven (RTFO) was performed following ASTM D2872, in which samples were

176 placed in an RTFO at 163°C and 4 L/min airflow for 85 min to simulate conditions at an asphalt-mixing

177 plant. Afterward, aging was performed for two cycles in a pressure aging vessel (2PAV) according to

178 ASTM D6521, wherein approximately 50g of the bitumen was subjected to an air pressure of 2.10 MPa at

8
179 100°C for 40 h. The bitumen was then degassed in a vacuum oven at 170°C for 30 minutes at 15 kPa.

180 Asphaltenes were extracted from the aged bitumen by collecting the fraction insoluble in n-heptane.

181 Pyrene and fluoranthene were selected from the aromatic conjugates studied as exfoliants in this

182 paper; they were obtained from Sigma-Aldrich. To make solutions for UV-Vis absorption spectroscopy,

183 aged asphaltenes were mixed with 10% by weight of dopant (exfoliant molecules), and dissolved together

184 in toluene. The solution was then diluted to a concentration of 20 mg/L asphaltenes in an 80/20 v/v

185 hexane/toluene solvent mixture and left to flocculate for 24h. A portion of the flocculated solution was

186 measured by UV-Vis after a brief 5s ultrasonication to re-suspend (but not re-dissolve) the aggregates.

187 Another portion of the flocculated solution was centrifuged in an Eppendorf 5415C centrifuge at 10krpm

188 (8161g RCF) for 5 min in order to separate the soluble portion (supernatant) from the aggregates

189 (precipitate). Once separated, the precipitate was air dried and re-dissolved in toluene for UV-Vis

190 measurements. Absorption spectra from 290 – 800 nm were collected using a Cary 6000i UV-Vis-NIR

191 spectrophotometer (Varian) and a quartz cuvette with 1 cm pathlength (28B-Q-10, Starna Cells).

192 For dynamic shear rheometer analysis, a heat gun was used to warm up the spatula to recover 5gr

193 of aged bitumen and then transferred to 20ml Borosilicate glass vial; after that 10% of PAH (fluoranthene

194 or Pyrene) by the weight of bitumen was added to the vial and placed on the hotplate. The temperature

195 was measured with Infrared Thermometer and kept constant at 190C for 10 minutes while mixing was

196 performed. A Malvern Kinexus dynamic shear rheometer (DSR) was used to evaluate the effect of

197 conjugated aromatic exfoliants on the rheology of prepared samples. Parallel Plate geometry (4mm

198 diameter) was used to prepare bitumen specimen with 1.75mm thickness. In this study, a frequency sweep

199 test (0.0001 to 100 rad/second) at 1% strain was performed at two temperatures of 10C and 25C.

200 3. Results and Discussion

201 Molecular Models of PAH-Exfoliants. To date, a wide range of aromatic conjugates has been utilized as

202 intercalants and dispersants to noncovalently functionalize polyaromatic surfaces. The π-conjugated

203 molecules are inclined to be adsorbed over the polyaromatic surface by means of π-stacking interactions.

9
204 Therefore, the π-cyclic conjugated region in the exfoliating agents, in addition to its destructive effect on

205 the nano-aggregates, acts as an anchor for the other dangling segments that have dispersing effects and

206 prevent the PAHs from self-aggregation.

207 Specifically, in the case of bitumen, components with π-cyclic conjugation, including resins and

208 naphthene aromatics, have a predominant role in the stabilization of asphaltenes in the surrounding

209 medium. Several studies have probed the impact of resins and aromatics on asphaltene stacking behavior

210 and the resulting rheological and mechanical properties of the bitumen.54-58 However, the effect of π-

211 conjugation on asphaltene deagglomeration has not been explicitly reported. Therefore, in this study, we

212 evaluated the role of different aromatic conjugates in disturbing the π-π stacking of asphaltenes in the

213 aggregate.

214 Surfactants such as perylene, pyrene, and anthracene contain cyclic conjugates; these surfactants

215 have received particular attention in the exfoliation of graphitic compounds, since the surfactants'

216 hydrophobic surface is similar to the surface of graphene-like structures. The exfoliation of graphite or

217 CNT bundles is commonly accompanied by an external force, such as ultrasonication or shear mixing.

218 The external force can greatly help the molecules of solvent or surfactant intercalate into bulky PAH

219 layers/tubes, thereby increasing the interlayer space for the subsequent exfoliation to give mono- or few-

220 layered /bundled PAH molecules. The prolonged sonication used for the exfoliation of graphene in an

221 organic solvent consumes time and energy, and it has a destructive effect on the graphene flakes and

222 carbon nanotubes that affects their electronic properties.59-61

223 These challenges and issues can be overcome by using components that have a role in opening up

224 the PAH sheets in the aggregate to facilitate the interaction of surfactants with individual PAH sheets. A

225 promising solution could be the use of small π-conjugated aromatics that have a great desire to interact

226 with the outer edge of the polyaromatic surfaces without being drawn into the interlayer space in the

227 agglomerates. These cyclic conjugates can act as “key components” to open a sufficient space between

10
228 PAH layers in the aggregate for the penetration of surfactants and the subsequent exfoliation of the

229 aggregate.

230 The π-π interactions between aromatic conjugates and graphenic flakes determine the

231 effectiveness of the aromatic conjugate derivatives for PAH exfoliation, and post-exfoliation stabilization

232 depends on the electron density distribution on their aromatic surfaces.12 So, the number of fused rings is

233 important; a conjugate molecule with more fused rings has a greater influence on exfoliation due to

234 having more intense π-π interactions with graphene-like molecules.15 Therefore, it is worth exploring the

235 electronic characteristics of aromatic conjugates from a different attitude with a different function. The

236 degree of aromaticity is the intended molecular feature in this study to determine the capability of

237 aromatic conjugates to disturb the π-π interactions between oxidized asphaltenes in the agglomerates.

238 Accordingly, for this study, two kinds of aromatic conjugates were selected: those with

239 continuous π-conjugation and those with interrupted π-conjugation. The first choices to represent these

240 two kinds of conjugates are pyrene and perylene. Pyrene and perylene derivatives are the most popular

241 aromatics used as stabilizers in graphene exfoliation and carbon nanotube debundling.6, 12-14, 24, 62-68

242 Indeed, what distinguishes them from each other is their differing π-electron distribution, which plays a

243 decisive role in their respective interacting behaviors. To ensure the results of a comparison between the

244 two chosen categories of aromatic conjugates, we also selected a second representative from each

245 category. So there are four selected exfoliants: fluoranthene, perylene, pyrene, and benzo (a) pyrene.

246 Perylene and fluoranthene, in the category of interrupted aromatic conjugates, are obtained by

247 connecting two benzenoid aromatics through nearly single carbon-carbon bonds so as to form a new

248 ring.69-70 This makes the π-electron systems of two fused aromatics mutually independent, and their

249 interaction through the new joint ring is negligible.71-73 The π-cyclic conjugation in the central “empty”

250 ring is negligible; it would be non-aromatic or would have low intensity.74-78 Therefore, we are dealing

251 with the localization of π circulation in the interrupted aromatic conjugates, and subsequently, two

252 distinct aromatic moieties in their structures. In addition to studying interrupted conjugates, we also

11
253 examine the behavior of continuous aromatic conjugates; pyrene and benzo (a) pyrene are selected as the

254 rival molecules. Figure 1 portrays the molecular structures of the four aromatic conjugates used in this

255 research as exfoliating agents.

“Interrupted Aromatic Conjugates”

Perylene Fluoranthene
“Continuous Aromatic Conjugates”
Perylene

Pyrene Benzo (a) pyrene


256 Figure 1. Molecular structures of aromatic conjugate exfoliants.

257 As further evidence, we employ some indices to ensure the anti-aromatic or non-aromatic essence of the

258 central carbonic cycle in the selected interrupted conjugates.

259 Harmonic Oscillator Model of Aromaticity (HOMA) Index. Here, the local aromaticity for rings in our

260 considered aromatic exfoliants is quantified by means of the structure-based HOMA index to determine

261 continuity and discontinuity in the exfoliants' π-cyclic conjugation. A higher HOMA value corresponds to

262 a more aromatic nature of the ring in question, and hence, a more π-electron delocalization of the system.

263 A detailed inspection of the HOMA data given in Figure 2 reveals that in perylene and fluoranthene, the

264 central empty rings have significantly negative and low HOMA values, respectively. A negative value of

265 HOMA for the six-membered joint ring in perylene is a characteristic of anti-aromaticity, and a low value

266 of HOMA in the five-membered ring in fluoranthene indicates non-aromaticity. The HOMA values for

267 the rest of the rings in these molecules are close to 1, which means the rings are aromatic.

268

12
269

“HOMA values”
Pyrene

0.834 0.553
Benzo-pyrene

0.768 0.575 0.643 0.425 0.827


Fluoranthene

0.838 0.914 0.023


Perylene

0.756 -0.018
270 Figure 2. HOMA values for nonidentical rings in the selected aromatic exfoliants.

271 The second term in the HOMA equation (Equation 2) represents contributions from all bonds in the ring,

272 so the nature of the bonds determines the aromaticity character of the ring. The lengths of C-C bonds

273 between aromatics nuclei in the interrupted aromatic conjugates are about 1.48 Å; that matches well with

274 the C-C single bond length. Therefore, these carbonic linkages contribute negatively to the HOMA and

275 significantly break aromaticity. This observation is in line with the π current density maps of

276 perylene-type and fluoranthene-type compounds, which break up into two distinct benzenoid subunits,

277 and the π cyclic conjugation in the central six-ring or five-ring is negligible.73 As an example, the currents

278 on the C-C bonds connecting two naphthalenic moieties in perylene are weak.73, 79-80 So, the central ring

279 fails to show the benzenoid character, and naphthalene nuclei are predominant in the perylene structure,

280 based on the diamagnetic susceptibility measurements.81-82 This finding is consistent with nucleus-

281 independent chemical shifts (NICS) analysis predicting a very small intensity of π-electron circulation in

13
282 the central five- and six-membered rings.79, 83
In contrast, HOMA analysis for pyrene and its benzo-

283 derivative confirms their total aromatic structures.

284 Quantum Theory of Atoms in Molecules (QTAIM). To shed more light on the nature of carbon-carbon

285 linkages between benzenoids in fluoranthene and perylene, we employed QTAIM analysis. QTAIM

286 allows estimating the bond ellipticity (ε) at a bond critical point (BCP) along these C-C bond paths. The

287 bond ellipticity is a measure of the cylindrical charge distribution along the bond path or, in other words,

288 a measure of the π-character of the bonding. It is defined as in Equation 4.

𝜆
289 𝜀 = 𝜆1 − 1 𝑤ℎ𝑒𝑟𝑒 |𝜆1 | ≥ |𝜆2 | (4)
2

290 𝜆1 and 𝜆2 are eigenvalues of the Hessian matrix of electron density at the BCP perpendicular to the bond

291 path. In regard to the single and triple bonds, the bond is cylindrically symmetrical (𝜆1 = 𝜆2 ), then ε is

292 zero. Ellipticity increases from a single to a double bond and reaches the maximum value for a double

293 bond. The trend is reversed from a double to a triple bond, in which the cylindrical symmetry is regained.

294 The ε values for C-C bonds connecting two benzenoid aromatics in perylene and fluoranthene are 0.09

295 and 0.08, respectively, which are far from the ellipticity value for an aromatic bond, 0.23. Having

296 ε very close to zero for these bonds confirms their single bond nature.

297 Molecular Models of PAH Aggregate; Oxidized Asphaltene Dimer

298 Asphaltene is the most challenging fraction of bitumen because its structural features give it the

299 controlling role in the bitumen’s viscosity.84 They have more complex structure compared to the other

300 petroleum fragments in the asphalt.85 The continental or island model is the most predominant

301 conformation for asphaltene in the asphalt media,5, 28, 86 and it is the one considered in this paper. Based

302 on this model, asphaltene has a hand-like motif that consists of a condensed polyaromatic region

303 surrounded by alkyl chains. This component was introduced in the literature as the constituent of bitumen

304 with the most polarity and the highest molecular weight, due to the presence of heteroatoms (such as S, O,

305 and N) and a trace of some metals. The main distinction between asphaltenes and the other components of

306 asphalt is asphaltenes' tendency to self-aggregate and form stacks. Environmental factors can change the

14
307 asphaltene aggregation; this affects the rheology and morphological properties of the asphalt,

308 consequently altering the asphalt's performance.

309 Oxidation is one of the most important factors contributing to the hardening of asphalt and the

310 resulting pavement embrittlement.87-88 Based on the results of our DFT studies, asphaltenes have the

311 highest chemical reactivity toward oxidation compared to the other fractions of bitumen.89 These

312 chemical changes convert asphaltene to the more polar fractions of bitumen, which have a higher

313 tendency to agglomerate. For more detail about evolution of asphaltene molecules during oxidation,

314 please refer to Hung and Fini, 2019.90 Figure 3 shows the asphaltene model used in the present study; it is

315 an asphaltene-pyrrole structure well-known as the Yen-Mullins island model5, 28 modified by Greenfield-

316 Li91 and Fini et al.92 The asphaltene molecule subjected to oxidative aging is decorated with polar

317 carbonyl groups (Figure 3). The number of C=O groups and their positions have been described in our

318 previous work.93

319 The widely known model for an asphaltene agglomerate describes it as an assembly with an

320 aromatic core composed of stacked aromatic sheets surrounded by alkyl chains. Accordingly, the oxidized

321 asphaltene dimer employed in this study is a mimic of an aged asphaltene nano-aggregate, which is the

322 most stable conformation among our proposed orientations for the oxidized dimer93 (Figure 3).

Polyaromatic Aggregate Model

“Virgin Asphaltene” “Oxidized Asphaltene” “Oxidized Asphaltene Dimer”

323 Figure 3. The molecular structures of asphaltene, oxidized asphaltene, and oxidized asphaltene dimer.

324 Modeling of Asphaltene Aggregation Performance in the Presence of Aromatic Conjugate

325 Exfoliants

15
326 As far as we know, no effort has been made to use molecular modeling and electronic analysis to probe

327 the role of the aromaticity of conjugates on the deagglomeration and the size of asphaltene stacks.

328 Therefore, we will begin to unravel the mechanism of exfoliation through molecular modeling of oxidized

329 asphaltene aggregation and the effect of aromatics with different conjugation as exfoliants on the

330 deaggregation of asphaltene stacks. To do this, an interacting complex was designed to allow the most

331 stable oxidized asphaltene dimer to interact with our selected aromatic conjugates. Many researchers

332 believe that asphaltene agglomerates possess a fractal-like nature, probably owing to the presence of

333 interstitial pores within their assembly.94 Accordingly, to set up the initial structure for the “asphaltene

334 dimer/aromatic exfoliant” complex in the optimization process, the aromatic conjugate was located in the

335 less restricted position around the interlayer space of the dimer aggregate, to better feel the π-cloud of

336 asphaltenes.

337 Table 1 presents the exfoliating system of each aromatic conjugate separately with the asphaltene

338 dimer, to show how differences in the electronic structure of each aromatic conjugate cause different

339 interacting behavior. The intermolecular binding energy (Ebind) and the binding distance (dbind) provided

340 here are the two main descriptors of the interaction strength between polyaromatic planes (which are

341 asphaltenes in this study). As a side note, the binding distance reported here is the vertical distance

342 between asphaltene monomers, almost at the center of the asphaltene dimer. Different interaction patterns

343 were observed for two continuous conjugates and two interrupted conjugates. Aromatics with the

344 continuous circulation, including pyrene and benzo (a) pyrene, have a great tendency to go through the

345 interlayer region of the aggregate so that they are completely drawn into the space between asphaltenes in

346 their optimized interacting system. In contrast, the interrupted conjugates were stopped at the margin of

347 space between two asphaltene sheets by their central empty ring. Comparing the binding energy and

348 binding distance values before and after the addition of aromatic conjugates to the asphaltene dimer

349 shows that both types of aromatic components have reduced the π-stacking forces between polyaromatic

350 cores of the asphaltenes; this is demonstrated in Table 1 by a considerable decrease in the binding energy

351 (~32-37%) and a significant increase in the inter-sheet distance (~83-87%) between oxidized asphaltenes

16
352 in the dimer in the presence of aromatic exfoliants, in all complexes. Thus, DFT calculations show a

353 disrupting effect of aromatic conjugates on the intermolecular interactions between polyaromatic sheets in

354 the asphaltene aggregate. This is just for one conjugated exfoliant; in real systems with several exfoliants

355 for each aggregate, the changes in these values would be much more than what was reported here.

356 Table 1. The exfoliation systems of continuous and interrupted aromatics. Ebind is binding energy between
357 asphaltenes (kcal/mol), dbind is the interlayer distance between asphaltenes (Å).

“Oxidized Asphaltene Dimer (side view)”

Ebind= -48.8

dbind= 3.48

“Oxidized Asphaltene Dimer+Aromaric Complex (Top view)”


Benzo (a) Pyrene Exfoliation Complex
Pyrene Exfoliation Complex

Ebind= -31.6 Ebind= -30.9


dbind= 6.52 dbind= 6.45

17
Fluoranthene Exfoliation Complex
Perylene Exfoliation Complex

Ebind= -33.2 Ebind=-33.9


dbind= 6.36 dbind= 6.41

358 Table 1 shows that the disrupting effect of continuous conjugates is more intense compared to the effect

359 of interrupted aromatics. For the continuous conjugates, the variation of the binding distance between

360 asphaltenes (∆𝑑𝑏𝑖𝑛𝑑 ) from the oxidized asphaltene dimer to its exfoliated system reached more than 3.0

361 Å. This result is related to the structure of continuous conjugates being nearly identical to the aromatic

362 core in the asphaltene; this allows for a stronger π-π interaction compared to the interrupted conjugates

363 (Table 2). However, this interacting characteristic causes the continuous conjugates to insert completely

364 between asphaltenes and occupy the interlayer space. So, they leave less room for other intercalating

365 molecules and consequently decrease the extent of deagglomeration.

366 Table 2. The interacting systems of continuous and interrupted conjugates with an asphaltene monomer.
367 Ebind is the binding energy between asphaltene and aromatic (kcal/mol).

“Oxidized Asphaltene Monomer+Aromaric Complex (Top view)”

18
Benzo (a) Pyrene Interacting Complex
Pyrene Interacting Complex

Ebind= -17.8 Ebind= -18.1

Fluoranthene Interacting Complex


Perylene Interacting Complex

Ebind= -16.4 Ebind=-16.8

368 In conclusion, our molecular modeling results show that the small aromatic conjugates have a wedge role

369 in the deaggregation of polyaromatic assemblies. Wedging into the interlayer can be crucial in the

370 exfoliation process; opening up the PAH layers helps the other intercalants and stabilizers to insert into

371 the space between molecular sheets. The aromatics selected based on our theory increase the interlayer

372 spacing between PAH layers in the asphaltene stacking properly. However, the interrupted aromatic

373 conjugates are more practical compared to the continuous aromatic conjugates, due to the way the

374 interrupted conjugates interact; they stay in the margin of aggregates and make more interlayer space

375 between PAH sheets in their complexes. Therefore, they can play the “key” role to prepare the

19
376 polyaromatic aggregate to disperse into individual PAH fractions, reducing or eliminating the need for

377 external mechanical forces such as sonication.

378 Proposed mechanism. The precipitation and exfoliation mechanisms in bitumen are controlled by

379 dispersion interactions between asphaltene fragments and exfoliants. Aromatic conjugate exfoliants

380 reduce the size of the aggregations by serving as solvating agents to disrupt the π-π interactions that drive

381 self-association in the precipitate asphaltene fractions. The schematic depiction of the proposed

382 mechanism is shown in Scheme 1. The combination of π orbitals of one of the asphaltene fragments with

383 the π system of the aromatic exfoliant disturbs the π-π stacking interactions between asphaltenes in the

384 oxidized dimer, resulting in a greater separation distance between two asphaltenes in the dimer. As the

385 van der Waals force is proportional to 1⁄𝑟 6 , increasing the π-π stacking distance (r) above 5 Å can

386 significantly weaken the strong intersheet interactions of the graphene-type layers.95 This structural

387 deformation can cause the intercalation step, in which dispersant and stabilizer molecules penetrate the

388 empty space created between two asphaltene planes, imposing an extra spatial hindrance between

389 polyaromatic sheets and facilitating their separation. This can lead to the break-up of oxidized asphaltene

390 agglomerates, yielding small assemblies. Splitting in the asphaltene stacking through their interactions

391 with aromatic conjugates can be confirmed through visualization of the electronic perturbations imposed

392 by an aromatic intercalant.

393

20
Scheme 1. The mechanism of interrupted conjugate intercalation in asphaltene aggregation and the disturbance of π-π
stacking.

394 Noncovalent Interaction Analysis Using RDG Isosurfaces. We use reduced density gradient (RDG)

395 plots to visualize the extension of electrostatic and van der Waals interactions between oxidized

396 asphaltenes in their dimer before and after the introduction of the aromatic conjugate. RDG isosurfaces

397 qualitatively justify how the aromatic conjugate exfoliants decrease the stability of aggregations and

398 intermolecular interactions between asphaltenes. Accordingly, 3D-plots of RDG for an asphaltene dimer

399 before and after exfoliation are presented in Figures 4-a and 4-b. Discrete asphaltenes after exfoliation are

400 isolated from their complex with fluoranthene. The gradient isosurfaces presented in the following

401 correspond to 𝑠 = 0.5 𝑎𝑢 and are colored based on a blue-green-red (BGR) color scale.

402

21
“Oxidized Asphaltene Dimer” “Disturbed Asphaltene Dimer”
Top view
a. b.

403 Figure 4. Noncovalent interaction analysis of the intermolecular interaction between asphaltene
404 monomers in their (a) oxidized; and (b) exfoliated dimers (top-views).

405 As shown in Figure 4-a, the predominant interactions in the oxidized asphaltene dimer are van der Waals

406 forces between fragments, which reveal themselves as the extended greenish isosurfaces. The absence of

407 red and blue surfaces in this color scheme indicates that there are no significant steric repulsive forces and

408 no significant hydrogen bonding attractive forces in the asphaltene dimer. The absence of hydrogen

409 bonding in the asphaltene stacks was also reported in the previous studies. 96 In Figure 4-b, the noticeable

410 point in the RDG isosurfaces for the asphaltene dimer after exfoliation is the disappearance of van der

411 Waals and electrostatic interactions (green RDGs) between asphaltenes.

412 Using Molecular Dynamics Simulations to Evaluate the Deagglomeration Effect of Aromatic

413 Conjugates

414 To investigate the effect of exfoliants with different π-conjugation on the aggregation of oxidized

415 asphaltenes after the nano-aggregates formed in an equilibrium state, pyrene and fluoranthene were added

416 separately to the system in the amount of 10 wt% of the oxidized asphaltene molecules. All simulations

417 were performed at a constant temperature of 350 K and pressure of 1 bar for 20 ns. The original distance

418 between the aromatic molecules was more than 1 nm. Eight molecules of pyrene and eight molecules of

419 fluoranthene (10 wt% of the asphaltene fraction) were added to investigate their effect on the oxidized

420 asphaltene aggregation.

22
421 Average Aggregation Number (gz). The distance between the centers of mass (COM) in each pair of the

422 oxidized asphaltenes was calculated every 500 ps during the MD simulations to provide a measure of

423 stacking. A distance of less than 0.85 nm was accepted as a stacking distance, and the aggregation number

424 was calculated based on this criterion.97-98 The z-average aggregation number, gz, was calculated using the

425 following equation, where ni is the number of aggregations containing gi monomers. 99-100

∑ 𝑛𝑖 𝑔𝑖3
426 𝑔𝑧 = ∑𝑖 2 (5)
𝑖 𝑛𝑖 𝑔𝑖

427 The average aggregation numbers for oxidized asphaltene, oxidized asphaltene with pyrene, and oxidized

428 asphaltene with fluoranthene were calculated for 20 ns and for each 500 ps. Asphaltenes in solvent tend to

429 form nano-aggregates following self-interacting. With the extent of the simulation, nano-aggregates of
97
430 asphaltene start forming clusters as presented in previous studies . The results shown in Figure 5

431 illustrate that after 25 ns, the average aggregation number starts to increase when there are no dopant

432 molecules present. By adding pyrene and fluoranthene exfoliants to the ensemble containing already

433 formed nano-aggregates of oxidized asphaltenes, the increase in average aggregation number is prevented

434 and at some points reduced more than an equilibrated ensemble of oxidized asphaltene molecules. There

435 is a greater reduction in aggregation number from adding fluoranthene molecules than from adding

436 pyrenes.

6
Average aggregation number

Dopant
5

3 Oxidized asphaltene
Oxidized asphaltene + 10% Pyrene
2
Oxidized + 10% fluoranthene
1

0
0 5 10 15 20 25 30 35 40
Time (ns)
437

23
438 Figure 5. Average aggregation number for the oxidized asphaltene molecules in heptane with and without
439 the presence of continuous and interrupted aromatic conjugates (pyrene and fluoranthene, respectively).

440 Radial Distribution Function (RDF). In addition to the average aggregation number, the radial

441 distribution function (g(r)) was calculated for the subset consisting of the most-centered carbon atom of

442 each oxidized asphaltene molecule. RDF results help to understand the most probable stacking distance of

443 asphaltenes and also the change in the form of interaction of oxidized asphaltene molecules due to the

444 addition of aromatic conjugates. The RDF is shown in Figure 6; it is calculated from binning of pairwise

445 distance to the maximum force cutoff of 9.5 Å using the “compute rdf” command of LAMMPS source

446 code.

447 In Figure 6, the first peak of the radial distribution function before the addition of either pyrene or

448 fluoranthene shows that for parallel stacking of asphaltenes, the stacking distance with the highest

449 probability is 4.51 Å. After the addition of aromatic exfoliants, this peak probability occurred at a greater

450 distance between two asphaltenes in the stack (a change from 4.51 Å for pure oxidized asphaltene to 4.98 Å

451 for oxidized asphaltene with pyrene, and to 4.61 Å for oxidized asphaltene with fluoranthene). This can be

452 due to the weaker interactions of oxidized asphaltenes with each other after adding aromatic conjugates,

453 which is in agreement with our DFT results (Table 1). This phenomenon also happened for unparalleled

454 stacking (known as T-shaped stacking) that occurred at distances higher than 6 Å.

24
90
80 4.6075Å
70
60 4.9875Å
Oxidized asphaltene
50
g(r)

40 4.5125Å Oxidzed asphaltene +


10% Pyrene
30
Oxidized asphaltene +
20 10% Fluoranthene
10
0
0 2 4 6 8 10
455 Distance (Å)

456 Figure 6. Radial distribution function, g(r), for the most-centered carbon atom of oxidized asphaltene
457 molecules in heptane with and without the presence of aromatic compounds pyrene and fluoranthene.

458 Molecular Exfoliation Mechanisms. Figure 7 shows a snapshot sequence for the trajectory of the system

459 of oxidized asphaltene with pyrene. The molecular mechanism for the exfoliation of asphaltene

460 aggregates occurs in three steps. At the first step (Figure 7a,), continuous aromatic conjugates (in white)

461 interact with asphaltene aggregates (in green) and intercalate into the intersheet gap between asphaltenes,

462 with a consequent increase in the intermolecular space within the asphaltene aggregate. In the next step

463 (Figures 7b and 7c), aromatic conjugates disturb the π-π forces between PAH faces of asphaltenes and

464 deagglomerate their assemblies. Finally (Figure 7d and 7e), aromatic conjugates act as terminators to

465 prevent asphaltene aggregates from re-stacking through the conjugates' π-stacking.

466 The Figure 8 snapshot sequence shows a slightly different exfoliating behavior for the

467 fluoranthene exfoliants. Due to their interrupted π-conjugation, fluoranthene molecules have a weaker π-π

468 interaction with the aromatic zone of the asphaltenes, so they mostly detach from the deagglomerated

469 asphaltene stacks after exerting their disrupting effect on the π-stacking of asphaltenes. This phenomenon

470 causes fluoranthenes to be less effective than pyrenes in preventing the re-stacking of asphaltenes.

471 Fluoranthenes act weakly as terminators, so this job is relegated to other fragments in the real bitumen;

25
472 the fluoranthenes are not consumed, and they are free to attack the remaining asphaltene aggregates and

473 initiate more deagglomeration. The diffusion coefficient (reported in supporting information) is higher for

474 fluoranthene than for pyrene, which can be due to fluoranthenes' relatively looser interactions with

475 oxidized asphaltene molecules that result in more freedom of movement.

476 In the RDF plot reported in the previous section, the increase in the intensity of the first peak for the

477 fluoranthene-containing system is more than that of the pyrene-containing system. This observation can be

478 justified through mechanisms defined for both types of conjugate exfoliants. The radial distribution function

479 for asphaltene studies gives an understanding of the "degree of separation" of asphaltene molecules. Therefore,

480 asphaltenes in the case of pyrene are more dispersed in the ensemble, while for fluoranthene, they are more in

481 touch with each other. Looking at the average aggregation number and RDF results simultaneously show us

482 that pyrene may disperse the asphaltene aggregates better and keep them separated, but fluoranthene can better

483 reduce the size of nanoaggregates and prevent them from forming large aggregates.

“Pyrene Exfoliation Mechanism”


a. b. c.

d. e.

484 Figure 7. Snapshot sequence for exfoliation of the oxidized asphaltene aggregates by 10 wt% pyrene.

26
“Fluoranthene Exfoliation Mechanism”
a. b. c.

d. e.

485 Figure 8. Snapshot sequence for exfoliation of the oxidized asphaltene aggregates by 10 wt% fluoranthene.

486

487 UV-Vis Analysis of Aged Asphaltenes and Their Complex with Aromatic Intercalants.

488 Figure 9-a shows the UV-Vis absorption spectra of aged asphaltenes and asphaltenes mixed with either

489 pyrene or fluoranthene (10 wt% solids). All solutions were partially flocculated in 80% hexane solvent.

490 When the flocculated suspension was separated into soluble supernatant and insoluble precipitate

491 fractions by centrifugation, almost all of the added dopant (~98%) in both cases remained in the

492 supernatant (Figure 9-b). No significant difference in asphaltene solubility was observed between doped

493 and undoped mixtures. Compared to solutions in pure toluene, the pyrene and fluoranthene peaks are

494 blue-shifted 2-3 nm in 80% hexane.

27
495 a. b.

496 Figure 9. (a) UV-Vis absorption spectra of flocculated solutions of aged asphaltenes with and without
497 added pyrene or fluoranthene. (b) Absorption spectra of the flocculated mixtures separated into their
498 soluble fraction (supernatant in 80% hexane) and insoluble fraction (precipitate re-dissolved in toluene).

499 Figure 10 shows the absorption spectra of the pyrene-asphaltene mixtures with the asphaltene absorbance

500 subtracted out to more clearly show the pyrene spectra. The pyrene absorbance in both the flocculated

501 suspension and the separated supernatant are similar, but subtraction of one spectrum from the other

502 reveals slight differences. First, there appears to be a very weak, broad absorption peak in the flocculated

503 suspension around 358 nm, just ahead of the primary absorption peaks. This absorbance could be

504 indicative of pyrene associating with the asphaltene agglomerates in solution, as aggregation of

505 polyaromatics tends to cause a red-shift in absorbance and fluorescence.101 A similar comparison of the

506 fluoranthene-doped mixtures could not detect any shifts in the fluoranthene spectrum (data not shown),

507 which may indicate a weaker interaction with asphaltenes under these conditions compared to pyrene.

508 These observations confirm the quantum mechanical prediction that continuous conjugates have a

509 stronger interaction with PAH aggregate (Tables 1 and 2). On a related note, fluoranthene is known to be

510 more soluble than pyrene in both hexane and toluene.102 However, this conclusion is tentative because the

511 effect is very small, and an alternative explanation is that the weaker absorbance of fluoranthene in this

512 wavelength range (more than five times lower absorption coefficient for the leading absorption peak

28
513 compared to pyrene) makes detecting any interactions much more difficult. For future studies,

514 fluorescence spectroscopy may be more sensitive for detecting any such interactions.

515 In addition, the pyrene peaks in the supernatant are slightly red-shifted compared to the non-

516 separated, flocculated suspension. The reason for this shift is unknown, although one possibility is that

517 asphaltene aggregates in suspension are solvated or swollen with excess toluene. Centrifugation of the

518 aggregates into a compact precipitate mass could release some of this solvent, which would appear as an

519 increase in toluene concentration to the dissolved molecules in the supernatant and cause the

520 aforementioned spectral shift.

521

522 Figure 10. UV-Vis absorption spectra of the asphaltene-pyrene mixture comparing the whole
523 (flocculated) suspension to the separated supernatant.

524 The asphaltene absorbance was subtracted out in order to focus on the pyrene spectra. The difference

525 (flocculated minus supernatant) shows a slight red-shift of the pyrene peaks in the supernatant, while the

526 flocculated sample shows a weak absorption peak around 358 nm (blue arrow) that may be indicative of

527 dimers or aggregates.

528 Rheological Analysis for Evaluating the Effect of Small π-conjugated Aromatic Molecules on the

529 Aged Bitumen

530 The effect of π-conjugated aromatics molecules on the aged bitumen was studied using rheometry

531 analysis utilizing a dynamic shear rheometer. To do so a frequency sweep test was applied to assess the

29
532 crossover points, which are the corresponding values at which shear storage modulus (G´) and shear loss

533 modulus (G˝) have the same magnitude. The crossover point is defined by two symbols of crossover

534 frequency, c, and crossover modulus, Gc. In this point, a transition from liquid-like to solid -like

535 materials happens. Gc has been correlated to polydispersity by several researchers mainly in the polymer

536 industry and few in the asphalt industry.103

537 PDI is related to the ratio between the weight average molecular weight (Mw) and the number

538 average molecular weight (Mn). The PDI in rheology is typically defined as 100,000/crossover modulus,

539 where the crossover modulus has units of Pascals. Application of Gc for calculating PDI was investigated

540 by Zeichner and Patel who presented a relationship between the crossover modulus of polypropylene

541 melts and molecular weight distribution.104 Even though asphalt molecular structure differs from that of

542 polymers, its asphaltene portion (heptane insoluble) which are the focus of this paper is composed of self-

543 assembled nano-aggregates of few nanometers in size with a molecular weight of approximately 750

544 g/mol.105-106 PDI has been also correlated to the point where a sol-gel shift occurs in asphaltene-rich

545 bitumen. 107-108

546 Farrar et al. used Christensen and Anderson model for bitumen, referred to as the dispersed polar

547 fluid model for utilizing the crossover modulus as a parameter related to asphalt structure and its

548 evolution with aging.103, 109 They found an approximately linear relationship between the oxygen uptake

549 and the inverse of the log of the crossover modulus, 1/log(|G*c |).103

550 Liu et al. investigated the evolution of bitumen during oxidation aging and used the change in

551 crossover values as indicators of aging progression. They showed G∗c continuously decreased for both

552 neat and modified bitumen as aging time increased.110 As aging progresses the amount of polar functional

553 groups in bitumen increases giving rise to intermolecular interactions causing asphaltene aggregation.

554 This, in turn, promotes the formation of nano-aggregates of different sizes due to molecules affinity to

555 cover hydrophilic side chains of oxidized asphaltenes formed during aging. Such different nano-

556 aggregates will increase the polydispersity of bitumen matrix.


30
557 As shown in Figure 11-a, crossover values of aged bitumen doped with either of additive were

558 higher than un-doped bitumen. In addition, comparing the effect of the two aromatics exfoliants additives,

559 it can be seen the crossover values for fluoranthene-modified bitumen are higher than that of the pyrene-

560 doped bitumen indicating fluoranthene is more efficient in changing crossover values. It should be noted

561 that the observed changes in crossover values are from a direct measurement without any transfer

562 function or model. To further interpret observed changes, here we use prior arts relating crossover values

563 to extent of asphaltene aggregation as a plausible justification for fluoranthene’s having higher capacity to

564 break self-assembled asphaltene nano-aggregates than pyrene. The latter conclusion was independently

565 made from our DFT and MD analysis in this paper; therefore, regardless of whether increase of crossover

566 values stems from reduction of asphaltene nano-aggregates and decrease of polydispersity in bitumen, we

567 have shown the two exfoliants have capacity to break nano-aggregates, and that fluoranthene is more

568 efficient than pyrene to do so. Accordingly, addition of two aromatic conjugates is postulated to exfoliate

569 the asphaltene nano-aggregates and reduce the aggregation size. This is further observed in calculated

570 polydispersity index (PDI), where fluoranthene’s PDI (0.015) is nearly 60% less than that of pyrene

571 (0.024); a lower PDI for fluoranthene is in-line with the result of molecular dynamics simulations

572 indicating breakage of nano-aggregates to smaller and more frequent similar-size particles due to

573 introduction of fluoranthene to aged bitumen.

574

575

576

577

578

579

580

31
581 a.

100
Gc = 4.08 MPa
c = 0.19 (rad/sec)
10
Modulus G', G" (kPa)

1 Gc = 6.74 MPa

c = 2.51 (rad/sec)
0.1

0.01

0.001
0.001 0.01 0.1 1 10 100
Frequency (rad/sec)
Fluoranthene + Aged bitumen-G'
Fluoranthene + Aged bitumen-G"
Pyrene + Aged bitumen-G'
Pyrene + Aged bitumen-G"

582
583 b.

1.00E+09
Complex viiscosity (Pa.s)

1.00E+08

1.00E+07

1.00E+06

1.00E+05
0.0001 0.001 0.01 0.1 1 10 100
Frequancy (rad/sec)
Fluoranthene + Aged bitumen Pyrene + Aged bitumen
584
585 Figure 11. a) Storage modulus (G’) and loss modulus (G”) of rejuvenated bitumen measured at 10C,
586 b) complex viscosity of two modified `.

32
587 Figure 12 compares the rheological changes measured at 25C for aged bitumen before and after

588 the intrusion of pyrene and fluoranthene. As shown in Figure 12-a, the crossover modulus of the aged

589 bitumen is 2.67MPa at a frequency of 0.631; it is increased by introducing each type of conjugate

590 intercalant. The value of the crossover modulus for pyrene at a temperature of 25C was measured to be

591 approximately 6MPa, and occurred at around 80Hz. However, it was observed that the crossover value for

592 the fluoranthene scenario at the same temperature was much higher and occurred at a significantly higher

593 frequency that was beyond the detectable frequency range. In addition, our results demonstrate that the

594 complex viscosity for the aged bitumen is more reduced by the inclusion of fluoranthene compared to the

595 pyrene intercalant (Figure 12-b). This reduction is more obvious at lower frequencies, which are

596 analogous to higher temperatures.

597 Rheological studies showed that both aromatic conjugates (continuous and interrupted) were

598 effective on decreasing the aggregate size and reducing the complex viscosity, while the aged bitumen

599 modified with interrupted conjugates had more uniform and smaller aggregate size than the continuous

600 one. Based on the exfoliation mechanism introduced in the MD section, interrupted conjugates such as

601 fluoranthene provide more space for the functionality of the other molecules in bitumen to interact with

602 deagglomerated asphaltene stacks and keep them dispersed in the medium. So aromatics with interrupted

603 conjugation act as exfoliating agents to deagglomerate more of the asphaltene assemblies that intensified

604 during oxidative aging.

605 a.

33
100
Gc = 2.67 MPa
c = 0.63 (rad/sec)
10
Modulus G',G" (MPa)

1
Gc = 6.72 MPa
0.1 c = 100 (rad/sec)

0.01

0.001
0.001 0.01 0.1 1 10 100
Frequancy (rad/sec)
Aged bitumen-G' Aged bitumen-G"
Pyrene + Aged bitumen-G' Pyrene + Aged bitumen-G"
Fluoranthene + Aged bitumen-G' Fluoranthene + Aged bitumen-G"
606
607 b.

1.00E+08
Complex Viscosity (Pa.s)

1.00E+07

1.00E+06

1.00E+05

1.00E+04
0.001 0.01 0.1 1 10 100

Frequancy (rad/sec)
Aged bitumen Pyrene + Aged bitumen
608 Fluoranthene + Aged bitumen

609 Figure 12. a) Storage modulus (G’) and loss modulus (G”) of aged bitumen and
610 rejuvenated bitumen measured at 25C, b) Complex viscosity of two modified bitumen.

34
611 4. Conclusion

612 This paper studies the role of aromatic conjugates as exfoliants to disrupt a PAH nano-aggregate model

613 such as those found in asphaltene stacks. The results of our DFT analysis showed that both types of

614 conjugated exfoliants can increase the intermolecular distance between asphaltenes in a dimer, and

615 consequently disfavors agglomeration. While continuous conjugates show higher binding energy with

616 asphaltene than do interrupted ones, the continuous conjugates fully enter the inter-sheet spacing and get

617 consumed in one PAH stack. The interrupted conjugates are more efficient for exfoliation, since even

618 though they increase the inter-sheet spacing, they only partially enter the PAH stack leaving room for

619 surfactants and other intercalants to enter the interlayer space and facilitate exfoliation. In addition, due to

620 their limited partial entrance into the PAH stacks, interrupted conjugates are not trapped inside the PAH

621 interlayer, and can easily detach from the stacks. This allows an interrupted conjugate not to be consumed

622 in interaction with one PAH stack, making the interrupted conjugate available to affect many other stacks

623 in the matrix. This was further corroborated by Molecular dynamics simulation showing interrupted

624 conjugates mostly detach from the deagglomerated asphaltene stacks after disrupting the π-stacked

625 structure of asphaltenes. This can be attributed to their weaker π-π interactions with asphaltenes and their

626 partial entrance into the interlayer spacing.

627 UV-Vis confirmed our theoretical findings showing both types of conjugated exfoliants were

628 effective at interacting with asphaltene, with the continuous conjugate having a more noticeable

629 interaction. This can be attributed to continuous conjugates fully entering the inter-sheet spacing and

630 having stronger π-π interactions with PAH stack.

631 Based on rheology analysis, both types of exfoliants reduced the polydispersity index (PI) of aged

632 bitumen indicating both exfoliants decreased the agglomeration of asphaltene molecules. The PI of self-

633 aggregated asphaltene was reduced by 60% in the presence of pyrene; this reduction was even higher in

634 the presence of fluoranthene indicating the interrupted conjugates are more effective exfoliants than

635 continuous conjugates. The latter observation was further evidenced by a significant increase in both

35
636 crossover frequency and crossover modulus in the fluoranthene-exfoliating complex. The significant

637 enhancement in the crossover point is attributed to a decrease in nano-aggregate size. The addition of

638 pyrene to the aggregated system increased the crossover modulus to 4.08 MPa; the crossover modulus for

639 the fluoranthene increased to 6.74MPa. This effect was further corroborated by comparing the complex

640 viscosity of self-aggregated asphaltene in the presence of pyrene or fluoranthene.

641 Based on the results of this study, conjugated aromatics (especially interrupted ones) can be

642 promising exfoliants for the deagglomeration of polyaromatic aggregates such as self-assembled stacks of

643 oxidized asphaltene.

644 Acknowledgment

645 This research is sponsored by the National Science Foundation (Award No: 1546921) and the US

646 Department of Transportation via the Center for Highway Pavement Preservation. The content of this

647 paper reflects the view of the authors, who are responsible for the facts and the accuracy of the data

648 presented.

649 Supporting Information

650 The Supporting Information is available free of charge on the ACS Publications website.

651 • Diffusion Coefficient Analysis

652 References

653 1. Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.;
654 Grigorieva, I. V.; Firsov, A. A., Electric field effect in atomically thin carbon films. science 2004, 306
655 (5696), 666-669.
656 2. Krätschmer, W.; Lamb, L. D.; Fostiropoulos, K.; Huffman, D. R., Solid C60: a new form of
657 carbon. Nature 1990, 347, 354.
658 3. Iijima, S., Helical microtubules of graphitic carbon. Nature 1991, 354 (6348), 56.
659 4. Geim, A. K.; Novoselov, K. S., The rise of graphene. Nature materials 2007, 6 (3), 183-191.
660 5. Mullins, O. C., The modified Yen model. Energy & Fuels 2010, 24 (4), 2179-2207.
661 6. Ehli, C.; Rahman, G. A.; Jux, N.; Balbinot, D.; Guldi, D. M.; Paolucci, F.; Marcaccio, M.;
662 Paolucci, D.; Melle-Franco, M.; Zerbetto, F., Interactions in single wall carbon
663 nanotubes/pyrene/porphyrin nanohybrids. Journal of the American Chemical Society 2006, 128 (34),
664 11222-11231.
665 7. Lu, J.; Nagase, S.; Zhang, X.; Wang, D.; Ni, M.; Maeda, Y.; Wakahara, T.; Nakahodo, T.;
666 Tsuchiya, T.; Akasaka, T., Selective interaction of large or charge-transfer aromatic molecules with
667 metallic single-wall carbon nanotubes: critical role of the molecular size and orientation. Journal of the
668 American Chemical Society 2006, 128 (15), 5114-5118.

36
669 8. Su, Q.; Pang, S.; Alijani, V.; Li, C.; Feng, X.; Müllen, K., Composites of graphene with large
670 aromatic molecules. Advanced materials 2009, 21 (31), 3191-3195.
671 9. Dyke, C. A.; Tour, J. M., Covalent functionalization of single-walled carbon nanotubes for
672 materials applications. The Journal of Physical Chemistry A 2004, 108 (51), 11151-11159.
673 10. Li, Q.; Kinloch, I. A.; Windle, A. H., Discrete dispersion of single-walled carbon nanotubes.
674 Chemical Communications 2005, (26), 3283-3285.
675 11. Guardia, L.; Fernández-Merino, M.; Paredes, J.; Solis-Fernandez, P.; Villar-Rodil, S.; Martinez-
676 Alonso, A.; Tascón, J., High-throughput production of pristine graphene in an aqueous dispersion assisted
677 by non-ionic surfactants. Carbon 2011, 49 (5), 1653-1662.
678 12. Parviz, D.; Das, S.; Ahmed, H. T.; Irin, F.; Bhattacharia, S.; Green, M. J., Dispersions of non-
679 covalently functionalized graphene with minimal stabilizer. Acs Nano 2012, 6 (10), 8857-8867.
680 13. An, X.; Simmons, T.; Shah, R.; Wolfe, C.; Lewis, K. M.; Washington, M.; Nayak, S. K.;
681 Talapatra, S.; Kar, S., Stable aqueous dispersions of noncovalently functionalized graphene from graphite
682 and their multifunctional high-performance applications. Nano letters 2010, 10 (11), 4295-4301.
683 14. Lee, D.-W.; Kim, T.; Lee, M., An amphiphilic pyrene sheet for selective functionalization of
684 graphene. Chemical Communications 2011, 47 (29), 8259-8261.
685 15. Sampath, S.; Basuray, A. N.; Hartlieb, K. J.; Aytun, T.; Stupp, S. I.; Stoddart, J. F., Direct
686 exfoliation of graphite to graphene in aqueous media with diazaperopyrenium dications. Advanced
687 Materials 2013, 25 (19), 2740-2745.
688 16. Moniruzzaman, M.; Winey, K. I., Polymer nanocomposites containing carbon nanotubes.
689 Macromolecules 2006, 39 (16), 5194-5205.
690 17. Coleman, J. N.; Khan, U.; Blau, W. J.; Gun’ko, Y. K., Small but strong: a review of the
691 mechanical properties of carbon nanotube–polymer composites. Carbon 2006, 44 (9), 1624-1652.
692 18. Tasis, D.; Tagmatarchis, N.; Bianco, A.; Prato, M., Chemistry of carbon nanotubes. Chemical
693 reviews 2006, 106 (3), 1105-1136.
694 19. Skakalova, V.; Kaiser, A.; Dettlaff-Weglikowska, U.; Hrncarikova, K.; Roth, S., Effect of
695 chemical treatment on electrical conductivity, infrared absorption, and Raman spectra of single-walled
696 carbon nanotubes. The Journal of Physical Chemistry B 2005, 109 (15), 7174-7181.
697 20. Banerjee, S.; Hemraj‐Benny, T.; Wong, S. S., Covalent surface chemistry of single‐walled carbon
698 nanotubes. Advanced Materials 2005, 17 (1), 17-29.
699 21. Sun, Y.-P.; Fu, K.; Lin, Y.; Huang, W., Functionalized carbon nanotubes: properties and
700 applications. Accounts of chemical research 2002, 35 (12), 1096-1104.
701 22. Hecht, D. S.; Ramirez, R. J.; Briman, M.; Artukovic, E.; Chichak, K. S.; Stoddart, J. F.; Grüner,
702 G., Bioinspired detection of light using a porphyrin-sensitized single-wall nanotube field effect transistor.
703 Nano letters 2006, 6 (9), 2031-2036.
704 23. Hu, L.; Zhao, Y. L.; Ryu, K.; Zhou, C.; Stoddart, J. F.; Grüner, G., Light‐Induced Charge
705 Transfer in Pyrene/CdSe‐SWNT Hybrids. Advanced Materials 2008, 20 (5), 939-946.
706 24. Zhao, Y.-L.; Stoddart, J. F., Noncovalent functionalization of single-walled carbon nanotubes.
707 Accounts of chemical research 2009, 42 (8), 1161-1171.
708 25. Fujigaya, T.; Nakashima, N., Non-covalent polymer wrapping of carbon nanotubes and the role
709 of wrapped polymers as functional dispersants. Science and technology of advanced materials 2015, 16
710 (2), 024802.
711 26. Lu, J.; Yang, J.-x.; Wang, J.; Lim, A.; Wang, S.; Loh, K. P., One-pot synthesis of fluorescent
712 carbon nanoribbons, nanoparticles, and graphene by the exfoliation of graphite in ionic liquids. ACS nano
713 2009, 3 (8), 2367-2375.
714 27. Liang, Y. T.; Hersam, M. C., Highly concentrated graphene solutions via polymer enhanced
715 solvent exfoliation and iterative solvent exchange. Journal of the American Chemical Society 2010, 132
716 (50), 17661-17663.
717 28. Pomerantz, A. E.; Wu, Q.; Mullins, O. C.; Zare, R. N., Laser-based mass spectrometric
718 assessment of asphaltene molecular weight, molecular architecture, and nanoaggregate number. Energy &
719 Fuels 2015, 29 (5), 2833-2842.

37
720 29. Jiménez-Mateos, J. M.; Quintero, L. C.; Rial, C., Characterization of petroleum bitumens and
721 their fractions by thermogravimetric analysis and differential scanning calorimetry. Fuel 1996, 75 (15),
722 1691-1700.
723 30. Lesueur, D., The colloidal structure of bitumen: Consequences on the rheology and on the
724 mechanisms of bitumen modification. Advances in colloid and interface science 2009, 145 (1-2), 42-82.
725 31. Petersen, J., Transportation Research Circular E-C140: A Review of the Fundamentals of Asphalt
726 Oxidation: Chemical, Physicochemical, Physical Property, and Durability Relationships. Transportation
727 Research Board of the National Academies, Washington, DC 2009.
728 32. Le Guern, M.; Chailleux, E.; Farcas, F.; Dreessen, S.; Mabille, I., Physico-chemical analysis of
729 five hard bitumens: Identification of chemical species and molecular organization before and after
730 artificial aging. Fuel 2010, 89 (11), 3330-3339.
731 33. Fini, E. H.; Kalberer, E. W.; Shahbazi, A.; Basti, M.; You, Z.; Ozer, H.; Aurangzeb, Q., Chemical
732 characterization of biobinder from swine manure: Sustainable modifier for asphalt binder. Journal of
733 Materials in Civil Engineering 2011, 23 (11), 1506-1513.
734 34. Yang, X.; You, Z.-P.; Dai, Q.-L., Performance evaluation of asphalt binder modified by bio-oil
735 generated from waste wood resources. International Journal of Pavement Research and Technology
736 2013, 6 (4), 431-439.
737 35. Yang, X.; You, Z.; Dai, Q.; Mills-Beale, J., Mechanical performance of asphalt mixtures
738 modified by bio-oils derived from waste wood resources. Construction and Building Materials 2014, 51,
739 424-431.
740 36. Sun, Z.; Yi, J.; Huang, Y.; Feng, D.; Guo, C., Properties of asphalt binder modified by bio-oil
741 derived from waste cooking oil. Construction and Building Materials 2016, 102, 496-504.
742 37. Delley, B., An all‐electron numerical method for solving the local density functional for
743 polyatomic molecules. The Journal of chemical physics 1990, 92 (1), 508-517.
744 38. Delley, B., From molecules to solids with the DMol 3 approach. The Journal of chemical physics
745 2000, 113 (18), 7756-7764.
746 39. Perdew, J. P.; Burke, K.; Ernzerhof, M., Generalized gradient approximation made simple.
747 Physical review letters 1996, 77 (18), 3865.
748 40. Grimme, S., Density functional theory with London dispersion corrections. Wiley
749 Interdisciplinary Reviews: Computational Molecular Science 2011, 1 (2), 211-228.
750 41. Schwenke, D. W.; Truhlar, D. G., Systematic study of basis set superposition errors in the
751 calculated interaction energy of two HF molecules. The Journal of chemical physics 1985, 82 (5), 2418-
752 2426.
753 42. Kruszewski, J.; Krygowski, T., Definition of aromaticity basing on the harmonic oscillator model.
754 Tetrahedron Letters 1972, 13 (36), 3839-3842.
755 43. Krygowski, T. M., Crystallographic studies of inter-and intramolecular interactions reflected in
756 aromatic character of. pi.-electron systems. Journal of chemical information and computer sciences 1993,
757 33 (1), 70-78.
758 44. M. Frisch, G. T., H. B. Schlegel, G. Scuseria, M. Robb,; J. Cheeseman, G. S., V. Barone, B.
759 Mennucci and; G. Petersson, I., Wallingford, CT 2009, vol. 200.
760 45. Lu, T.; Chen, F., Multiwfn: a multifunctional wavefunction analyzer. Journal of computational
761 chemistry 2012, 33 (5), 580-592.
762 46. Bader, R., Atoms in Molecules: A Quantum Theory. Oxford, England: 1990; pp 188-250.
763 47. Biegler‐König, F.; Schönbohm, J., Update of the AIM2000‐Program for atoms in molecules.
764 Journal of computational chemistry 2002, 23 (15), 1489-1494.
765 48. Johnson, E. R.; Keinan, S.; Mori-Sanchez, P.; Contreras-Garcia, J.; Cohen, A. J.; Yang, W.,
766 Revealing noncovalent interactions. Journal of the American Chemical Society 2010, 132 (18), 6498-
767 6506.
768 49. Humphrey, W.; Dalke, A.; Schulten, K., VMD: visual molecular dynamics. Journal of molecular
769 graphics 1996, 14 (1), 33-38.

38
770 50. Sun, H.; Mumby, S. J.; Maple, J. R.; Hagler, A. T., An ab initio CFF93 all-atom force field for
771 polycarbonates. Journal of the American Chemical Society 1994, 116 (7), 2978-2987.
772 51. Waldman, M.; Hagler, A. T., New combining rules for rare gas van der Waals parameters.
773 Journal of computational chemistry 1993, 14 (9), 1077-1084.
774 52. Plimpton, S., Fast parallel algorithms for short-range molecular dynamics. Journal of
775 computational physics 1995, 117 (1), 1-19.
776 53. Bowers, B. F.; Huang, B.; Shu, X., Refining laboratory procedure for artificial RAP: A
777 comparative study. Construction and Building Materials 2014, 52, 385-390.
778 54. Nellensteyn, F., The constitution of asphalt. J. Inst. Pet 1924, 10, 311-325.
779 55. Pfeiffer, J. P.; Saal, R., Asphaltic bitumen as colloid system. The Journal of Physical Chemistry
780 1940, 44 (2), 139-149.
781 56. Andersen, S. I.; Keul, A.; Stenby, E., Variation in composition of subfractions of petroleum
782 asphaltenes. Petroleum Science and Technology 1997, 15 (7-8), 611-645.
783 57. Spiecker, P. M.; Gawrys, K. L.; Kilpatrick, P. K., Aggregation and solubility behavior of
784 asphaltenes and their subfractions. Journal of colloid and interface science 2003, 267 (1), 178-193.
785 58. Spiecker, P. M.; Gawrys, K. L.; Trail, C. B.; Kilpatrick, P. K., Effects of petroleum resins on
786 asphaltene aggregation and water-in-oil emulsion formation. Colloids and Surfaces A: Physicochemical
787 and engineering aspects 2003, 220 (1-3), 9-27.
788 59. Khan, U.; O'Neill, A.; Lotya, M.; De, S.; Coleman, J. N., High‐Concentration Solvent Exfoliation
789 of Graphene. Small 2010, 6 (7), 864-871.
790 60. Hennrich, F.; Krupke, R.; Arnold, K.; Rojas Stütz, J. A.; Lebedkin, S.; Koch, T.; Schimmel, T.;
791 Kappes, M. M., The mechanism of cavitation-induced scission of single-walled carbon nanotubes. The
792 Journal of Physical Chemistry B 2007, 111 (8), 1932-1937.
793 61. Skaltsas, T.; Ke, X.; Bittencourt, C.; Tagmatarchis, N., Ultrasonication induces oxygenated
794 species and defects onto exfoliated graphene. The Journal of Physical Chemistry C 2013, 117 (44),
795 23272-23278.
796 62. Dong, X.; Shi, Y.; Zhao, Y.; Chen, D.; Ye, J.; Yao, Y.; Gao, F.; Ni, Z.; Yu, T.; Shen, Z.,
797 Symmetry breaking of graphene monolayers by molecular decoration. Physical review letters 2009, 102
798 (13), 135501.
799 63. Zhang, M.; Parajuli, R. R.; Mastrogiovanni, D.; Dai, B.; Lo, P.; Cheung, W.; Brukh, R.; Chiu, P.
800 L.; Zhou, T.; Liu, Z., Production of graphene sheets by direct dispersion with aromatic healing agents.
801 Small 2010, 6 (10), 1100-1107.
802 64. Jang, J.-H.; Rangappa, D.; Kwon, Y.-U.; Honma, I., Direct preparation of 1-PSA modified
803 graphene nanosheets by supercritical fluidic exfoliation and its electrochemical properties. Journal of
804 Materials Chemistry 2011, 21 (10), 3462-3466.
805 65. Englert, J. M.; Röhrl, J.; Schmidt, C. D.; Graupner, R.; Hundhausen, M.; Hauke, F.; Hirsch, A.,
806 Soluble Graphene: Generation of Aqueous Graphene Solutions Aided by a Perylenebisimide‐Based
807 Bolaamphiphile. Advanced Materials 2009, 21 (42), 4265-4269.
808 66. Simmons, T. J.; Bult, J.; Hashim, D. P.; Linhardt, R. J.; Ajayan, P. M., Noncovalent
809 functionalization as an alternative to oxidative acid treatment of single wall carbon nanotubes with
810 applications for polymer composites. ACS nano 2009, 3 (4), 865-870.
811 67. Ehli, C.; Guldi, D. M.; Herranz, M. A.; Martín, N.; Campidelli, S.; Prato, M., Pyrene-
812 tetrathiafulvalene supramolecular assembly with different types of carbon nanotubes. Journal of
813 Materials Chemistry 2008, 18 (13), 1498-1503.
814 68. Fujigaya, T.; Nakashima, N., Methodology for homogeneous dispersion of single-walled carbon
815 nanotubes by physical modification. Polymer journal 2008, 40 (7), 577.
816 69. Gutman, I.; Durdevic, J., Fluoranthene and its congeners-A graph theoretical study. MATCH-
817 COMMUNICATIONS IN MATHEMATICAL AND IN COMPUTER CHEMISTRY 2008, 60 (2), 659-670.
818 70. Donaldson, D.; White, J., The crystal and molecular structure of perylene. Proc. R. Soc. Lond. A
819 1953, 220 (1142), 311-321.

39
820 71. Haigh, C.; Mallion, R., Rationalisation of Relative» Ring-Current «Sizes in Polycyclic,
821 Conjugated Hydrocarbons. Croatica Chemica Acta 1989, 62 (1), 1-26.
822 72. Randić, M., Aromaticity of polycyclic conjugated hydrocarbons. Chemical Reviews 2003, 103
823 (9), 3449-3606.
824 73. Monaco, G.; Zanasi, R., On the additivity of current density in polycyclic aromatic hydrocarbons.
825 The Journal of chemical physics 2009, 131 (4), 044126.
826 74. Randić, M.; Balaban, A. T., Partitioning of π-electrons in rings of polycyclic conjugated
827 hydrocarbons. Part 1: Catacondensed benzenoids. Polycyclic Aromatic Compounds 2004, 24 (3), 173-
828 193.
829 75. Balaban, A. T.; Randić, M., Partitioning of π-electrons in rings of polycyclic conjugated
830 hydrocarbons. Part 3. Perifusenes. New Journal of Chemistry 2004, 28 (7), 800-806.
831 76. Vukicevic, D.; Randic, M.; Balaban, A. T., Partitioning of π-electrons in rings of polycyclic
832 conjugated hydrocarbons. Part 4. Benzenoids with more than one geometric Kekulé structure
833 corresponding to the same algebraic Kekulé structure. Journal of mathematical chemistry 2004, 36 (3),
834 271-279.
835 77. Balaban, A. T.; Randić, M., Partitioning of π-electrons in rings of polycyclic conjugated
836 hydrocarbons. 5. Nonalternant compounds. Journal of chemical information and computer sciences 2004,
837 44 (5), 1701-1707.
838 78. Balaban, A. T.; Randić, M., Partitioning of π-electrons in rings of polycyclic conjugated
839 hydrocarbons: Part 6. Comparison with other methods for estimating the local aromaticity of rings in
840 polycyclic benzenoids. Journal of mathematical chemistry 2005, 37 (4), 443-453.
841 79. Fias, S.; Fowler, P. W.; Delgado, J. L.; Hahn, U.; Bultinck, P., Correlation of Delocalization
842 Indices and Current‐Density Maps in Polycyclic Aromatic Hydrocarbons. Chemistry-A European Journal
843 2008, 14 (10), 3093-3099.
844 80. Radenković, S.; Bultinck, P.; Gutman, I.; Ðurđević, J., On induced current density in the
845 perylene/bisanthrene homologous series. Chemical Physics Letters 2012, 552, 151-155.
846 81. Shiba, H.; Hazato, G., On the Diamagnetism of Perylene. I. Bulletin of the Chemical Society of
847 Japan 1949, 22 (3), 92-96.
848 82. Hazato, G., On the Diamagnetism of Perylene. II. Bulletin of the Chemical Society of Japan 1949,
849 22 (4), 151-157.
850 83. Radenkovic, S.; Tosovic, J.; Đurđević Nikolić, J., Local aromaticity in naphtho-annelated
851 fluoranthenes: can the five-membered rings be more aromatic than the six-membered rings? The Journal
852 of Physical Chemistry A 2015, 119 (20), 4972-4982.
853 84. Branthaver, J. F.; Petersen, J.; Robertson, R.; Duvall, J.; Kim, S.; Harnsberger, P.; Mill, T.;
854 Ensley, E.; Barbour, F.; Scharbron, J. Binder characterization and evaluation. Volume 2: Chemistry;
855 1993.
856 85. Schuler, B.; Meyer, G.; Peña, D.; Mullins, O. C.; Gross, L., Unraveling the molecular structures
857 of asphaltenes by atomic force microscopy. Journal of the American Chemical Society 2015, 137 (31),
858 9870-9876.
859 86. Schuler, B.; Fatayer, S.; Meyer, G.; Rogel, E.; Moir, M.; Zhang, Y.; Harper, M. R.; Pomerantz,
860 A. E.; Bake, K. D.; Witt, M., Heavy oil based mixtures of different origins and treatments studied by
861 atomic force microscopy. Energy & Fuels 2017, 31 (7), 6856-6861.
862 87. Lu, X.; Isacsson, U., Effect of ageing on bitumen chemistry and rheology. Construction and
863 Building materials 2002, 16 (1), 15-22.
864 88. Qin, Q.; Schabron, J. F.; Boysen, R. B.; Farrar, M. J., Field aging effect on chemistry and
865 rheology of asphalt binders and rheological predictions for field aging. Fuel 2014, 121, 86-94.
866 89. Mousavi, M.; Pahlavan, F.; Oldham, D.; Hosseinnezhad, S.; Fini, E. H., Multiscale Investigation
867 of Oxidative Aging in Biomodified Asphalt Binder. The Journal of Physical Chemistry C 2016, 120 (31),
868 17224-17233.
869 90. Hung, A. M.; Fini, E. H., Absorption spectroscopy to determine the extent and mechanisms of
870 aging in bitumen and asphaltenes. Fuel 2019, 242, 408-415.

40
871 91. Li, D. D.; Greenfield, M. L., Chemical compositions of improved model asphalt systems for
872 molecular simulations. Fuel 2014, 115, 347-356.
873 92. Martín-Martínez, F. J.; Fini, E. H.; Buehler, M. J., Molecular asphaltene models based on Clar
874 sextet theory. RSC Advances 2015, 5 (1), 753-759.
875 93. Pahlavan, F.; Mousavi, M.; Hung, A. M.; Fini, E. H., Characterization of oxidized asphaltenes
876 and the restorative effect of a bio-modifier. Fuel 2018, 212, 593-604.
877 94. Gawrys, K. L.; Matthew Spiecker, P.; Kilpatrick, P. K., The role of asphaltene solubility and
878 chemical composition on asphaltene aggregation. Petroleum science and technology 2003, 21 (3-4), 461-
879 489.
880 95. Spanu, L.; Sorella, S.; Galli, G., Nature and strength of interlayer binding in graphite. Physical
881 review letters 2009, 103 (19), 196401.
882 96. Hansen, C. M., HANSEN SOLUBILITY PARAMETERS: A User’s Handbook, Per Redelius ,
883 Chapter9: Hansen Solubility Parameters of Asphalt, Bitumen, and Crude Oils, . 2007, 151-175.
884 97. Goual, L.; Sedghi, M.; Wang, X.; Zhu, Z., Asphaltene aggregation and impact of alkylphenols.
885 Langmuir 2014, 30 (19), 5394-5403.
886 98. Samieadel, A.; Oldham, D.; Fini, E. H., Investigating molecular conformation and packing of
887 oxidized asphaltene molecules in presence of paraffin wax. Fuel 2018, 220, 503-512.
888 99. Sengers, J. V.; Kayser, R.; Peters, C.; White, H., Equations of state for fluids and fluid mixtures.
889 Elsevier: 2000; Vol. 5.
890 100. Sedghi, M.; Goual, L.; Welch, W.; Kubelka, J., Effect of asphaltene structure on association and
891 aggregation using molecular dynamics. The Journal of Physical Chemistry B 2013, 117 (18), 5765-5776.
892 101. Oliveira, D.; Baba, K.; Teizer, W.; Kasai, H.; Oikawa, H.; Nakanishi, H., Stopped-Flow Studies
893 of the Formation of Organic Nanocrystals in the Reprecipitation Method, Y. Masuda (Ed.), 167, DOI:
894 10.5772/17303. In Nanocrystal, InTech: 2011.
895 102. Acree Jr, W. E., IUPAC-NIST Solubility Data Series. 98. Solubility of Polycyclic Aromatic
896 Hydrocarbons in Pure and Organic Solvent Mixtures—Revised and Updated. Part 3. Neat Organic
897 Solvents. Journal of Physical and Chemical Reference Data 2013, 42 (1), 013105.
898 103. Farrar, M. J.; Turner, T. F.; Planche, J.-P.; Schabron, J. F.; Harnsberger, P. M., Evolution of the
899 crossover modulus with oxidative aging: Method to estimate change in viscoelastic properties of asphalt
900 binder with time and depth on the road. Transportation Research Record 2013, 2370 (1), 76-83.
901 104. Zeichner, G.; Patel, P. In A comprehensive evaluation of polypropylene melt rheology,
902 Proceedings 2nd world congress of chem eng, Montreal, 1981.
903 105. Mullins, O. C.; Sabbah, H.; Eyssautier, J.; Pomerantz, A. E.; Barré, L.; Andrews, A. B.; Ruiz-
904 Morales, Y.; Mostowfi, F.; McFarlane, R.; Goual, L., Advances in asphaltene science and the Yen–
905 Mullins model. Energy & Fuels 2012, 26 (7), 3986-4003.
906 106. Zadshir, M.; Oldham, D. J.; Hosseinnezhad, S.; Fini, E. H., Investigating bio-rejuvenation
907 mechanisms in asphalt binder via laboratory experiments and molecular dynamics simulation.
908 Construction and Building Materials 2018, 190, 392-402.
909 107. Mousavi, M.; Hosseinnezhad, S.; Hung, A. M.; Fini, E. H., Preferential adsorption of nickel
910 porphyrin to resin to increase asphaltene precipitation. Fuel 2019, 236, 468-479.
911 108. Laukkanen, O.-V.; Pellinen, T.; Makowska, M., Exploring the observed rheological behaviour of
912 in-situ aged and fresh bitumen employing the colloidal model proposed for bitumen. In Multi-Scale
913 Modeling and Characterization of Infrastructure Materials, Springer: 2013; pp 185-197.
914 109. Christensen, D.; Anderson, D., Chemical-Physical Property Relationships for Asphalt Cements
915 and the Dispersed Polar Fluid Model. Washington, DC 1992, 37 (3), 1279-1291.
916 110. Liu, F.; Zhou, Z.; Zhang, X.; Wang, Y., On the linking of the rheological properties of asphalt
917 binders exposed to oven aging and PAV aging. International Journal of Pavement Engineering 2019, 1-
918 10.
919

41

View publication stats

You might also like