Thermochemistry and Kinetics of Intermolecular Addition of Radicals

Subscriber access provided by LUNDS UNIV

A: Spectroscopy, Molecular Structure, and Quantum Chemistry

Thermochemistry and Kinetics of Intermolecular Addition of Radicals to Toluene and Alkylaromatics Lawrence Lai, and William H. Green J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.9b00817 • Publication Date (Web): 06 Mar 2019 Downloaded from http://pubs.acs.org on March 6, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Thermochemistry and Kinetics of Intermolecular Addition of Radicals to Toluene and Alkylaromatics Lawrence Lai1, William H. Green1* 1

– Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139

Abstract To better understand intermolecular radical additions to aromatic rings that take place in the pyrolysis of alkylaromatics at low to moderate temperatures (~450°C), the thermochemistry and kinetics of several reactions of this type are investigated using the CBS-QB3 level of theory. The calculated thermochemistry of the adduct radicals is significantly different from previous estimates; the average discrepancy in Gibbs free energy at 298 K is 5.3 kcal/mol. A group additivity value for aromatic pi radicals was developed to facilitate rapid accurate estimates for other molecules containing this functional group; average discrepancy in Gibbs free energy using the updated group additivity value is improved to 0.5 kcal/mol. Previous estimations of the rate coefficients of these reactions were found to be inaccurate due to the lack of important features such as loss of aromaticity in the compounds used as the training set for the estimation. Rate coefficients for addition of several different radicals to aromatic rings are reported. The reaction rates are comparable for ortho, meta, and para additions, slowest for addition to the substituted position, and insensitive to the length of alkyl chains attached to the aromatic reactant.

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 27

Introduction The pyrolysis of alkylaromatics is important in oil refining1-4, coal pyrolysis5, studies relevant to organic geochemistry6, and production of chemicals7,8. In the past, many reaction types relevant to alkylaromatic pyrolysis have been studied, such as unimolecular initiation and radical recombination6,9,10, disproportionation11-14, hydrogen abstraction6,15,16, beta scission2,10,17, and intramolecular addition6,10.

As part of the beta scission/intermolecular addition family, intermolecular addition to aromatic rings has not been studied extensively. Initiation, radical recombination6,18, and disproportionation14 reactions carry high relevance due to their production/elimination of two radicals, dramatically changing the reactivity of the pyrolysis system. Aliphatic beta scission directly reduces molecular weight and forms desired alkene products, while aromatic intermolecular addition (and its reverse, beta-scission of aromatic pi radicals) have more subtle effects. The chemistry of alkylaromatics is dominated by resonantly stabilized benzylic radicals2,19, whereas aromatic pi radicals are relatively difficult to form since they require loss of aromaticity20. Aromatic intramolecular addition reactions are more frequently studied than the intermolecular reactions, because they form fused ring aromatic species, associated with coke formation6,10.

Despite the relative lack of prior work in aromatic intermolecular addition, there have been some relevant studies. Previous moderate-temperature experimental work from this group2 and Lannuzel14 reports products such as 1-methyl-2-(phenylmethyl) benzene from alkylaromatic pyrolysis, which are likely products formed by the addition of benzylic radicals to aromatic rings. At very high temperature, addition of OH, H, and O to benzene rings is known to occur21-23. Shukla et al. observed addition of phenyl radical to benzene, and proposed this was important in polyaromatic hydrocarbon (PAH) formation24.


Page 3 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Computationally, our previous work2 generated a reaction mechanism for the pyrolysis of hexylbenzene using the Reaction Mechanism Generator (RMG)25. The resulting mechanism contains a substantial number of products formed through radicals adding to the aromatic bonds of hexylbenzene and other alkylaromatics, very few of which match our experiments, suggesting significant flaws in the model.

In the interest of gaining a better understanding of this class of reactions and to be able to generate more accurate chemical mechanisms, this work studies the addition of aliphatic and aromatic radicals to aromatic pi bonds. This reaction class has the potential to form multi ring aromatics connected by alkyl bridges, allowing for the possibility of further aromatic and coke formation, an active area of interest4,2628.

Aromatics connected by alkyl bridges are thought to be key structures in coal, petroleum, and heavy

oil fractionations29-31.

Methods In this work, quantum calculations are performed using Gaussian 0332, followed by the calculation of partition functions and fitting to NASA polynomials using the Cantherm software that is part of the RMG package25. Finally, group additivity estimates are derived through solving a least squares system of linear equations, based on previous group additivity values in the RMG database by Benson et al.33 and Lay et al.34

Gaussian 03 Calculations Thermochemical and kinetic data were calculated at the CBS-QB3 level of theory35 using the Gaussian 03 quantum chemistry package32. Geometry, energy, and frequency, calculations were done in the CBSQB3 level of theory; and hindered rotor calculations done in the B3LYP/CBSB7 level of theory. Hindered 3 ACS Paragon Plus Environment


Page 4 of 27

rotor calculations were made for each rotatable bond each treated as an independent rotor; rotatable bonds considered in this work are single bonds not bound to a ring structure, characterized by a rotation energy of < 10 kcal/mol. The dihedral angles were stepped in 10° increments followed by constrained optimizations for the full 360° rotation. These dihedral motions were projected out from the secondderivative matrix used to compute the harmonic frequencies of the other modes36. The energy eigenvalues for each rotor were computed in a basis set of sine and cosine functions, using the effective moment of inertia I(2,3) evaluated at the equilibrium geometry37.

RMG – Cantherm Partition functions were calculated using the rigid-rotor-harmonic-oscillator approximation with corrections for hindered rotors. Thermodynamic parameters were fitted to NASA polynomials using the Cantherm package of RMG37. A scaling factor of 0.99 was used for the frequency analysis with the recommended spin orbit corrections and bond additivity corrections35. Raw CBS-QB3 energy differences without bond additivity corrections were used to compute barrier heights in the forward direction of each reaction, to compute the high pressure limit rate coefficients using conventional transition state theory. For cases where there are optical isomers and/or internally symmetric hindered rotors, R*ln(Symmetry number) is subtracted from the entropy of the molecule, where the symmetry number is the product of external and internal symmetry divided by the number of chiral centers, and R is the gas constant. This addition assumes the species of interest is a homogenous mixture of all different optical isomers in all conformations.

RMG is an active open-source software project; as a result, the methods are constantly being improved. The version of Cantherm used in these calculations can be found using the GitHub version commit string


Page 5 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


in this work’s supporting information, allowing anyone to replicate the results. In the most recent version of RMG, Cantherm has been replaced by a related new software package, Arkane; but this work used the earlier RMG version with Cantherm.

Group Additivity Estimates A script was written using python package numpy to convert calculated thermochemistry to group additivity values. More information on the methodology of group additivity value development can be found in the supporting information.

Group additivity values do not intrinsically account for symmetry. As a result, the development of group additivity values for entropy uses thermochemistry calculations where entropy modifications due to symmetry are removed. When RMG is used to evaluate group additivity estimations, RMG’s algorithm will determine the appropriate symmetry number to account for external rotational symmetry, optical isomers, and internal symmetry due to hindered rotors; this calculation adjusts the entropy by subtraction of R*ln(Symmetry Number) of each of the species estimated. However, since this algorithm is still under development, it often misses optical isomers associated with ring structures. As RMG’s symmetry algorithm continues to improve in the future, estimations will improve.

Results and Discussion Aromatic Intermolecular Addition Reactions of Interest In this work, the aromatic intermolecular addition reactions between toluene/hexylbenzene and various radicals were studied. The radicals are the hydrogen atom, methyl radical, ethyl radical, benzyl radical,



Page 6 of 27

and 1-phenyl-1-ethyl radical; the selection of radicals is based on the relative radical concentrations in our system of interest found in previous work2. The products considered are shown in Table 1. Table 1 Aromatic intermolecular addition pairings considered for this work; possible resulting species are listed.

Aromatic Species

Radical Species

Resulting Species

H•

H•

•CH3

•C2H5


Page 7 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The purpose for the various types of pairings between aromatics and radicals is to gain an understanding of the effects of different types of aromatic side chains and radicals on the rate of aromatic intermolecular addition.

Thermochemistry of Intermolecular Addition Products The thermochemistry of all the resulting radicals shown in Table 1 (total of 24) were calculated at the CBS-QB3 rigid-rotor-harmonic-oscillator approximation + 1-D rotors level of theory. In addition, thermochemistry for recombination products with hydrogen for adducts of Toluene + H, CH3, and C2H5 were also calculated. This data is shown in Table 2; full NASA polynomials and heat capacity data can be found in the supporting information. This work finds limited matches between species reported here and species in the NIST Thermochemical Database38 or Burcat’s Third Millenium Database39; benchmarking was done for only for available matches in Table 3. In addition, we have previously reported benchmarks on other molecules computed using the same methods and software2,40.

Table 2 Thermochemical data of species calculated in this work. Heat capacities given in supporting information.

Species SMILES String Toluene + H• Ortho CC1=CC=C[CH]C1 Meta CC1[CH]CC=CC=1 Para CC1C=CC[CH]C=1 Substituted CC1[CH]C=CC=C1 Hexylbenzene + H• Ortho CCCCCCC1=C[CH]C=CC1 Meta CCCCCCC1[CH]C=CCC=1 Para CCCCCCC1C=CC[CH]C=1 Substituted CCCCCCC1C=C[CH]C=C1 Toluene + •CH3 Ortho CC1=CC=C[CH]C1C Meta CC1[CH]C(C)C=CC=1 Para CC1C=CC(C)[CH]C=1 Substituted CC1(C)[CH]C=CC=C1 Toluene + •C2H5

Enthalpy of formation (298 K) (kcal/mol)

Entropy (298 K) (cal/mol K)

39.2 40.4 40.3 42.2

81.5 81.9 82.4 80.8

14.8 15.6 15.4 16.7

129.5 129.6 128.6 129.2

33.1 33.6 33.5 34.0

90.2 90.6 89.7 87.2



Ortho CCC1[CH]C=CC=C1C Meta CCC1[CH]C(C)=CC=C1 Para CCC1[CH]C=C(C)C=C1 Substituted CCC1(C)[CH]C=CC=C1 Toluene + Benzyl Radical Ortho CC1=CC=C[CH]C1Cc1ccccc1 Meta CC1[CH]C(C=CC=1)Cc1ccccc1 Para CC1C=CC([CH]C=1)Cc1ccccc1 Substituted CC1([CH]C=CC=C1)Cc1ccccc1 Toluene + 1-Phenyl-1-Ethyl Radical Ortho CC1=CC=C[CH]C1C(C)c1ccccc1 Meta CC1[CH]C(C=CC=1)C(C)c1ccccc1 Para CC1C=CC([CH]C=1)C(C)c1ccccc1 Substituted CC(c1ccccc1)C1(C)[CH]C=CC=C1 Toluene + C2H5 + H Stable Ortho CCC1CC=CC=C1C Ortho CCC1C=CC=CC1C Ortho CCC1C=CCC=C1C Meta CCC1CC(C)=CC=C1 Meta CCC1C=C(C)C=CC1 Meta CCC1C=C(C)CC=C1 Para CCC1CC=C(C)C=C1 Para CCC1C=CC(C)C=C1 Substituted CCC1(C)CC=CC=C1 Substituted CCC1(C)C=CCC=C1 Toluene + CH3 + H Stable Ortho CC1=CC=CCC1C Ortho CC1C=CC=CC1C Ortho CC1=CCC=CC1C Meta CC1=CC(C)CC=C1 Meta CC1CC(C)C=CC=1 Meta CC1=CC(C)C=CC1 Para CC1C=CC(C)CC=1 Para CC1C=CC(C)C=C1 Substituted CC1(C)CC=CC=C1 Substituted CC1(C)C=CCC=C1 Toluene + H + H Stable Ortho/Meta CC1=CC=CCC1 Ortho/Meta CC1=CCC=CC1 Ortho/Sub CC1C=CC=CC1 Meta/Para CC1C=CCCC=1 Para/Sub CC1C=CCC=C1

Page 8 of 27

27.4 28.1 28.3 28.3

98.0 99.7 98.2 95.4

58.9 59.6 59.3 60.1

115.5 118.8 117.0 113.3

53.2 53.4 53.0 53.8

121.5 130.3 126.5 119.1

4.9 5.6 5.0 5.5 7.2 5.5 5.6 7.4 6.3 5.8

96.5 97.2 98.5 97.4 97.5 97.4 98.1 96.7 96.1 94.7

10.5 9.7 10.3 9.8 12.8 11.1 10.4 12.8 11.0 11.3

88.7 88.8 89.4 88.3 89.7 89.4 89.0 87.6 87.0 86.6

17.6 17.0 17.3 19.3 19.7

80.5 80.3 81.1 79.6 80.4

Table 3 Table of thermochemistry values at 298.15K, in comparison with literature values.

Species

Enthalpy of formation (298.15K) Entropy of formation (298.15K) (kcal mol-1) (cal mol-1 K-1) This work Literature This work Literature 8 ACS Paragon Plus Environment

Page 9 of 27

11.6

12.041

76.5

76.942

50.5

49.743

75.2

76.143

Derived Thermochemistry Group Values for Intermolecular Addition Products The calculated thermochemistry values in Table 2 were compared to the previous RMG’s estimates in Figure 1 which uses a group additivity estimation method outlined by Benson33, and further refined by Lay et al.34 for radical species; some of these groups used include aromatic C-H, aromatic C-C, and primary C-C, etc. In this comparison, the Gibbs free energy deviation between RMG’s estimate and this work’s calculations differ by 5.3 kcal/mol on average, with a standard deviation of 4.5 kcal/mol at 298K. 16 14 12

Frequency

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


10 8 6 4 2 0 -5

-4

-3

-2

-1

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

G298K (RMG - CBS-QB3) Figure 1 Distribution of Gibbs free energy deviation (kcal/mol) for 24 calculated aromatic intermolecular addition products. Distribution for RMG estimates previously (white) and after adding the aromatic pi radical group (black) are shown.

Figure 1 suggests that there is a systematic error in RMG’s group additivity estimate, where RMG overestimates the single Gibbs free energy when compared to CBS-QB3 calculation, by a range of 1 to 14 kcal/mol, a significant difference that could yield dramatically different product distributions in kinetic models2. The cause of the large thermochemistry deviations are explained in the supporting information. 9 ACS Paragon Plus Environment


Page 10 of 27

In response to the large Gibbs free energy deviations, a new aromatic pi radical hydrogen bond increment (HBI) group34 was developed using the methodology outlined in the methods section, where 19 out of 24 thermochemistry calculations made in this work are used as training data. In this work, only one new radical group (named the aromatic pi radical in Table 4) was derived. The error of estimation for the test set (5 out of 24 species, mutually exclusive from the training set) was decreased from 6.9 kcal/mol using old estimates to 0.7 kcal/mol using the updated aromatic pi radical. The enthalpy and entropy correction of this group is compared to other radical groups that were previously used to describe these species in Table 4.

Table 4 Comparison of enthalpy and entropy for the hydrogen bond increments for the aromatic pi radical (this work), secondary radical between two pi bonds, secondary allyl radical, and tertiary alkyl radical34

Aromatic Pi Radical Secondary radical between two pi bonds Secondary Allyl radical Tertiary alkyl radical

Enthalpy 298K (kcal/mol) 75.3 76.0 85.6 96.5

Entropy 298K (cal/mol K) 1.6 -4.1 -3.8 5.2

In addition to the numerical group value, the substructure corresponding to this group must be defined. We require the hydrocarbon parent of aromatic pi radicals to contain a 1,3-cyclohexadiene or 1,4cyclohexadiene structure, both of which will receive an identical hydrogen bond increment correction. The H removed to form the radical must be on a tetrahedral carbon adjacent to at least one double bonded carbon, and the carbon must be member of the cyclohexadiene ring. Additionally tertiary aromatic pi radicals are treated identically to secondary aromatic pi radicals.


Page 11 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The performance of the updated group values can be observed in Figure 1, where the Gibbs free energy deviation is centered much closer to zero, with an average of 0.5 kcal/mol and a standard deviation of 0.6 kcal/mol.

Future work for thermochemistry estimations in intermolecular addition While the developed HBI values in this work are developed with sufficient depth for RMG to estimate the thermochemistry of species that may arise in alkylaromatic pyrolysis systems, the degree to which these group values can be extrapolated to other unknown aromatic pi radicals is also not well understood, and other methods of thermochemistry estimation should be considered.

One of the other methods developed in this group by Li et al.44 employs the use of machine learning to allow for the machine to identify characteristics that users might not be able to easily identify. One of the drawbacks of this machine learning method cited by the authors is the lack of fundamental understanding of the criteria used to estimate the various aspects of thermochemistry in their molecules of interest. Contrary to this machine learning approach, group additivity estimations allow for the user to clearly view the contributions of each of the different groups that led to the thermochemistry estimation of a molecule. This work indicates that users of RMG and other chemical models should be wary of the shortcomings of both estimation methods when estimating the thermochemistry of unknown compounds, and consider other necessary alternatives based on these shortcomings.

Since the heat of formation of aromatic intermolecular additions are only slightly negative, the reverse of these reactions are fast, and aromatic pi radicals have a short lifetime as a result. These reactions are important if the radical has other fast decomposition pathways; for example, H + Toluene  Aromatic Pi



Page 12 of 27

radical  CH3 + Benzene. Note that if there is no other fast decomposition channel, the aromatic intermolecular addition reaction will rapidly equilibrate. In this case, the addition reactions are important if the concentration of aromatics are high enough, at sufficiently low temperatures that the equilibrium constant is large, such that the product Kc*[aromatics] > 1, since in that case a large fraction of the radicals in the system will be aromatic pi radicals.

Kinetics of Intermolecular Addition The rate coefficients for 21 of the 24 reactions outlined in Table 1 were calculated at the CBS-QB3 level of theory (Table 5). Transition state geometries can be found in the supporting information. Table 5 Modified Arrhenius rate parameters and rate at 723K for aromatic intermolecular addition reactions in CBS-QB3 level of theory.

Species Toluene + H• Ortho Meta Para Substituted Hexylbenzene + H• Substituted

A (cm3/mole-s)

N

Ea (kcal/mol)

Rate at 723K (cm3/mol-s)

1.0x109 1.1x109 1.1x109 2.2x108

1.4 1.4 1.4 1.6

4.5 5.4 5.4 6.4

1.3x1013 1.3x1013 1.3x1013 6.7x1012

8.8x107

1.7

6.1

6.7x1012

Toluene + •CH3 Ortho Meta Para Substituted

6.8x103 6.7x103 4.0x103 1.4x102

2.3 2.3 2.3 2.6

9.0 9.8 9.9 10.4

2.7x1010 3.0x1010 1.6x1010 5.3x109

Toluene + •C2H5 Ortho Meta Para Substituted

1.0x103 1.8x103 5.7x102 2.2x101

2.6 2.6 2.7 3.0

7.7 8.3 9.2 9.0

3.5x1010 4.4x1010 2.1x1010 7.7x109

2.9 2.8 2.8 3.2

11.3 12.5 12.4 11.9

1.2x1010 3.6x1010 1.8x1010 5.1x109

3.1

11.4

5.2x109

Toluene + Benzyl Radical Ortho 7.5x101 Meta 3.3x102 Para 2.0x102 Substituted 3.0 x100 Toluene + 1-Phenyl-1-ethyl Radical Ortho 1.0x101


Page 13 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Meta Para Substituted


6.4x100 3.8x100 2.9x10-2

3.2 3.2 3.7

11.3 11.4 12.1

9.7x109 4.8x109 9.2x108

RMG’s training reaction algorithm allows extrapolation of known reaction rate coefficients to analogous reaction types; the objective of this work is to therefore appropriately populate RMG’s training reaction database to allow for the accurate extrapolation of reaction rates to the many possibilities of aromatic intermolecular addition in our systems of interest; more details on how RMG utilizes training reactions can be found in Gao et al.25. This section outlines the trends that were observed across our calculations to assess if the training reactions are appropriately populated to improve future estimates.

Figure 2 compares the Evans-Polanyi relationship (activation energy vs enthalpy of reaction) of the aromatic intermolecular addition reactions in Table 5 to aliphatic intermolecular additions, the latter of which had kinetics and thermochemistry approximated using RMG for additions to propylene, by the radicals H, CH3, C2H5, Benzyl, and 1-phenyl-1-ethyl. It could be seen in this figure that the Evans-Polanyi relationship between the two classes of radical addition are extremely different. The slope of the EvansPolanyi plot for aromatic intermolecular addition reactions is much steeper than usual, making us suspect it is not reliable, e.g. it is possible that aromatic + H reactions are so different from the other reactions they should be isolated on a different curve. The differences of aromatic intermolecular addition reactions from its aliphatic counterpart exemplifies the need for the calculation of training reactions to improve the accuracy of estimates in this reaction class’ kinetics.



14.0

Benzylic Activation Energy (kcal/mol)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 27

yli

12.0

z

l

y k l A

n Be c

10.0

Alkyl

8.0 6.0

H

4.0

H

2.0 0.0

-40

-30 -20 -10 Enthalpy of Reaction (kcal/mol)

0

Figure 2 Evans-Polanyi relationship for aromatic intermolecular addition reactions calculated by CBS-QB3 (blue circles) and aliphatic intermolecular addition reactions approximated by RMG using propylene + radical (red squares). It can be observed that the two classes of reactions exhibit very different Evans Polanyi relationships.

Comparison with Previous RMG estimates Figure 3 shows the RMG estimated rate coefficients and CBS-QB3 calculated rate coefficients for toluene + five different types of radicals (averaged among four addition positions ortho/meta/para/substituted). At our temperature of interest (723K), RMG overestimates the rate coefficient by 2-4 orders of magnitude, depending on the type of radical added.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


log[Rate Coefficient (cm^3/mol s)]

Page 15 of 27

14 12 10 8 6 4 2 0 -2

1

1.5

2

2.5

3

3.5

4

4.5

5

-4 1000/T Figure 3 RMG estimated rate coefficients (dashed) and CBS-QB3 calculated rate coefficients (solid) for 1. Toluene + H (Blue), 2. Toluene + CH3 (Red), 3. Toluene + C2H5 (Green), 4. Toluene + Benzyl (Purple), and 5. Toluene + 1-phenyl-1ethyl (Orange). Rate coefficients for each class of reaction are averaged between four different addition sites. It can be observed that RMG’s rate estimates are much faster than CBS-QB3 calculations because of RMG’s questionable source.

The original source of these RMG’s estimates come from CBS-QB3 calculations using 1,3,5-hexatriene or 3-methyl-1,3,5-hexatriene as model compounds. Due to the lack of an aromatic structure of the model compound, the effects of losing aromaticity are not properly captured. The previous RMG estimates have an activation energy between 0.4 – 8.5 kcal/mol based on the site of addition and radical species used, whereas the CBS-QB3 calculations have an activation energy between 6.4 – 16.1 kcal/mol, highlighting the importance of capturing loss of aromaticity in aromatic intermolecular addition.



Effects of alkyl chain on aromatic species Figure 4 shows the rate coefficients of toluene + H and hexylbenzene + H in the substituted position calculated using CBS-QB3. The calculated rates for the two reactions differ by no more than factor of 1.5 across all temperatures shown. Log [Rate Coefficient (cm^3/mol s)]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 27

12 11 10 9 8 7 6 5 4 1

2

3 1000/T

4

5

Figure 4 CBS-QB3 calculated rate coefficients for toluene + H (Blue Solid) and hexylbenzene + H (Black dashed) in the substituted position. Rate coefficients for two reactions differ by no more than factor of 1.5 across all temperatures shown.

This suggests that the length of the side chain does not affect the rate coefficient significantly. In the interest of conserving computational resources, training reactions for the aromatic intermolecular addition of other radicals to longer chain alkylaromatics (such as hexylbenzene) were not constructed.


Page 17 of 27

Dependence of rate to the radical structure The effects of radical added is displayed in Figure 5; rates are averaged among the four different sites of addition. This figure outlines the general trend that larger more stable radical groups tend to result in slower rate coefficients.

log[Rate Coefficient (cm^3/mol s)]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


14 12

H

10 8 6

Alkyl

4 2 0 -2

1

Benzyl ic 2

3

4

5

-4 1000/T Figure 5 Rate coefficients for 1. Toluene + H (Blue Solid), 2. Hexylbenzene + H (Black Dashed), 3. Toluene + CH3 (Red Dotted), 4. Toluene + C2H5 (Green Dash Dotted), 5. Toluene + Benzyl (Purple Dash Double Dotted), and 6. Toluene + 1phenyl-1-ethyl radical (Orange Solid). Rate coefficients for each class of reaction are averaged between four different addition sites. The calculated rate coefficient is slower for the addition of larger radical species.

The activation energy increases with the size of radical added from ~7 kcal/mol (+H) to ~15 kcal/mol (+Benzyl/+1-phenyl-1-ethyl). This is likely due to the relative stabilities of these radicals; due to high stability of benzyl radicals (through resonance stabilization), the energy difference between the reactants and the transition state of aromatic intermolecular addition reactions are much higher, and thus, create a larger activation barrier for radicals to be added to the aromatic ring.



All the reactions listed in Figure 5 were added as training reactions in the RMG database, since the addition of different radicals can yield a 6 order of magnitude difference in rate coefficient (between H radical and 1-phenyl-1-ethyl radical).

Effects of addition position Figure 6 shows the effect of the different positions of addition in toluene + H. Addition to the ortho, meta and para positions have comparable rate coefficients, and addition to the substituted position has the slowest rate, which is an order of magnitude slower than addition to the other positions.

13 log [Rate Coefficient (cm^3/mol s)]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 27

12 11 10 9 8 7 6 5 1

2

3 1000/T

4

5

Figure 6. Rate coefficients for 1. Toluene + H (Ortho) (Blue Solid), 2. Toluene + H (Meta) (Red Dashed), 3. Toluene + H (Para) (Purple dotted), and 4. Toluene + H (Substituted) (Black Dash Dotted). Differences in the rate coefficients up to an order of magnitude can be observed.


Page 19 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


This work finds that these trends also apply to the addition of larger radical groups. We surmise that the steric effects from adding to the substituted position contribute towards the energy of the transition state, making addition to the substituted position more difficult by 1.1-2.0 kcal/mol. In the interest of appropriately populating RMG’s training reaction database, reaction groups were further designed for different addition positions due to the potential for difference in kinetics.

Summary to rate coefficients of Aromatic Intermolecular Addition This work finds that the previous RMG estimates for aromatic intermolecular addition rate coefficients to be highly overestimated compared to CBS-QB3 calculations done in this work, reason being the inappropriate use of 1,3,5-hexatriene and 3-methyl-1,3,5-hexatriene as a model compound to estimate these reactions, which does not account for effects resulting from breaking the aromatic structure. Other patterns observed in the kinetic rates of aromatic intermolecular addition are: (1) length of the alkyl side chain does not significantly affect rate, (2) increase in stability of radical reactant leads to a slower rate, and (3) addition to the substituted position is slower than addition to the three other positions. RMG’s training reaction database was appropriately populated based on these trends to ensure maximum accuracy in rate estimates in the future for aromatic intermolecular addition reactions.

Conclusions This work investigates the thermochemistry and kinetics of intermolecular addition, grounded by CBSQB3 calculations. Aromatic intermolecular addition (and its reverse) is a class of reactions that could carry significant relevance in the reactivity of aromatics at moderate temperatures (~450°C)2. Upon investigating the thermochemistry of the products of this class of reactions, it can be found that Benson et al.’s group additivity and Lay et al.’s hydrogen bond increment corrections implemented in the RMG 19 ACS Paragon Plus Environment


Page 20 of 27

software package are insufficient for estimating the thermochemistry of this class of species. As a result, a new radical correction is derived through CBS-QB3 calculations, and the resulting group value will be instrumental to estimating the thermochemistry of aromatic pi radicals. Previous rate estimates for these reactions made using RMG are far from the true values, due to the software making analogies with wrong model species. Detailed rate calculations are used in this work to better understand the characteristics of this class of reactions. The newly calculated thermochemistry and rate coefficients added to RMG’s database are expected to significantly improve the accuracy of future rate estimations.

Associated Content Supporting Information Github version commit strings, heat capacity data and NASA Polynomials for all calculated species, source of thermochemistry error estimates, transition state geometries, and least squares methodology.

Author Information Corresponding Author William H. Green Phone: 617-253-4580 Fax: 617-324-3127 E-mail: [email protected]


Page 21 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Acknowledgements The authors gratefully acknowledge the support of Saudi Aramco for funding this study under Contract Number 6600023444, and thank Dr. Ki-Hyouk Choi, Mohnnad Alabsi, and Massad S. Alanzi for their input.

The authors thank Mark Payne, Sarah Khanniche, Mengjie Max Liu, Matthew Johnson, and Mark Goldman, for their contributions to this work through expertise in RMG, intuition towards the chemistry of alkylaromatics, and detailed understanding of molecular symmetry. We thank all RMG developers, including but not limited to Ryan Gillis, Alon Grinberg Dana, Colin Grambow, and Richard West for the development and maintenance of the software that enabled this investigation.

References (1) Carr A. G.; Class C. A.; Lai L.; Kida Y.; Monrose T.; Green W. H. Supercritical Water Treatment of Crude Oil and Hexylbenzene: An Experimental and Mechanistic Study on Alkylbenzene Decomposition. Energy Fuels 2015, 29, 5290-5302. (2) Lai L.; Gudiyella S.; Liu M.; Green W. H. Chemistry of Alkylaromatics Reconsidered. Energy Fuels 2018, 32, 5489-5500. (3) Freund H.; Olmstead W. N. Detailed Chemical Kinetic Modeling of Butylbenzene Pyrolysis. Int. J. Chem. Kinet. 1989, 21, 561-574. (4) Gudiyella S.; Lai L.; Borne I. H.; Tompsett G. A.; Timko M. T.; Choi K.-H.; Alabsi M. H. ; Green W. H. An Experimental and Modeling Study of Vacuum Residue Upgrading in Supercritical Water. AIChE J., 2018, 64, 1732-1743. (5) Xie K.-C. Structure and Reactivity of Coal; Springer, New York, 2015; pp. 119-179.



Page 22 of 27

(6) Leguizamon Guerra N. C.; Lizardo Huerta J. C.; Lorgeoux C.; Michels R.; Fournet R.; Sirjean , B.; Randi A.; Bounaceur R.; Burkle-Vitzthum V. Thermal Cracking of n-Butylbenzene at High Pressure: Experimental Study and Kinetic Modelling. J. Anal. Appl. Pyrolysis, 2018, 133, 234-245. (7) Chen Q.; Froment G. F.; Thermal Cracking of Substituted Aromatic Hydrocarbons I. Kinetic Study of the Thermal Cracking of i-Propylbenzene. J. Anal. Appl. Pyrolysis, 1992, 21, 27-50. (8) Chen Q.; Froment G. F.; Thermal Cracking of Substituted Aromatic Hydrocarbons II. Kinetic Study of the Thermal Cracking of n-Propylbenzene and Ethylbenzene. J. Anal. Appl. Pyrolysis, 1991, 21, 51-77. (9) Klippenstein S. J.; Harding L. B.; Georgievskii Y. On the Formation and Decomposition of C7H8. Proc. Combust. Inst., 2007, 31, 221-229. (10) Kislov V. V.; Mebel A. M.; Aguilera-Iparraguirre J.; Green W. H. Reaction of Phenyl Radical with Propylene as a Possible Source of Indene and Other Polycyclic Aromatic Hydrocarbons: An Ab Initio/RRKM-ME Study. J. Phys. Chem. A, 2012, 116, 4176-4191. (11) Terry J. O.; Futrell J. H. Disproportionation and Combination Reactions of Simple Alkyl Radicals: Methyl, Ethyl, n-Propyl, and Isopropyl. Can. J. Chem., 1967, 45, 2327-2333. (12) Zhu R. S.; Xu Z. F.; Lin M. C. Combination and Disproportionation Reactions of Alkyl Radicals: An Ab Initio Kinetics Study For CH3 + C2H5. Prepr. Pap.-Am. Chem. Soc., 2004, 49, 447-448. (13) Mousavipour S. H.; Homayoon Z. A Theoretical Study on the Kinetics of Disproportionation Versus Association Reaction of CH3 + C2H5. J. Phys. Chem. A., 2003, 107, 8566-8574. (14) Lannuzel F.; Bounaceur R.; Michels R.; Scacchi G.; Marquaire P.-M. An Extended Mechanism Including High Pressure Conditions (700 bar) for Toluene Pyrolysis. J Anal. Appl Pyrolysis, 2010, 87, 236-247.


Page 23 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(15) Semenikhin A. S.; Savchenkova A. S.; Chechet I. V.; Matveev S. G.; Liu Z.; Frenklach M.; Mebel A. M. Rate Constants for H Abstraction from Benzo(a)pyrene and Chrysene: A Theoretical Study. Phys. Chem. Chem. Phys., 2017, 19, 25401-25413. (16) Mai T. V.-T.; Ratkiewicz A.; Duong M. V.; Huynh L. K. Direct Ab Initio Study of the C6H6 + CH3/C2H5 = C6H5 + CH4/C2H6 Reactions. Chem. Phys. Lett., 2016, 16, 102-109. (17) Ismail H.; Park J.; Wong B. M.; Green W. H.; Lin M. C. A Theoretical and Experimental Kinetic Study of Phenyl Radical Addition to Butadiene. Proc. Combust. Inst., 2005, 30, 1049-1056. (18) Burklé-Vitzthum V.; Michels R. Mechanistic Modeling of the Thermal Cracking of Decylbenzene. Application to the Prediction of Its Thermal Stability at Geological Temperatures. Ind. Eng. Chem. Res. 2003, 42, 5791-5808. (19) Sabbe M. K.; Reyniers M.-F.; Van Speybroeck V.; Waroquier M.; Marin G. B. Carbon‐Centered Radical Addition and β‐Scission Reactions: Modeling of Activation Energies and Pre‐exponential Factors. Chem. Phys. Chem., 2008, 9, 124-140. (20) Mebel A. M.; Lin M. C.; Morokuma K. Theoretical Study of Potential Energy Surface and Thermal Rate Constants for the C6H5 + H2 and C6H6 + H Reactions. J. Phys. Chem. A., 1997, 101, 3189-3196. (21) Lorenz K.; Zellner R. Kinetics of the Reactions of OH-Radicals with Benzene, Benzene-d6 and Naphthalene. Ber. Bunsenges. Phys. Chem., 1983, 87, 629-636. (22) Yang K. Free Radical Reactions Initiated by Ionizing Radiation. II. Rate Constants for Hydrogen Atom Addition Reactions with Mono-olefins, Butadiene and Benzene. J. Am. Chem. Soc., 1962, 84, 37953799. (23) Da Costa I.; Fournet R.; Billaud F.; Battin-Leclerc F. Experimental and Modeling Study of the Oxidation of Benzene. Int. J. Chem. Kinet., 2003, 35, 503-524.



Page 24 of 27

(24) Shukla B.; Susa A.; Miyoshi A.; Koshi M. Role of Phenyl Radicals in the Growth of Polycyclic Aromatic Hydrocarbons. J. Phys. Chem. A., 2008, 112, 2363-2369. (25) Gao C. W.; Allen J. W.; Green W. H.; West R. H. Reaction Mechanism Generator: Automatic Construction of Chemical Kinetic Mechanisms. Comput. Phys. Commun., 2016, 203, 212-225. (26) Choi K.-H.; Aljishi M. F. United States Patent US 8,535,518 B2, 2013. (27) Morimoto M.; Sato S.; Takanohashi T. Effect of Water Properties on the Degradative Extraction of Asphaltene Using Supercritical Water, J. Supercrit. Fluids, 2012, 68, 113-116. (28) Zaker A.; Guerra P.; Wang Y.; Tompsett G.; Huang X.; Bond J.; Timko M. Evidence of Heterogeneous Catalytic Activity of ZSM-5 in Supercritical Water for Dodecane Cracking. Catal Today, 2018, 101, 3189-3196. (29) Haenel M. W. Recent Progress in Coal Structure Research. Fuel, 1992, 71, 1211-1223. (30) Artok L.; Su Y.; Hirose Y.; Hosokawa M.; Murata S.; Nomura M. Structure and Reactviity of Petroleum-Derived Asphaltene. Energy Fuels, 1999, 13, 287-296. (31) Mojelsky T. W.; Ignasiak T. M.; Frakman Z.; McIntyre D. D.; Lown E. M.; Montgomery D. S.; Strausz O. P. Structural Features of Alberta Oil Sand Bitumen and Heavy Oil Asphaltenes. Energy Fuels, 1992, 6, 83-96. (32) Frisch M. J.; Trucks G. W.; Schlegel H. B.; Scuseria G. E.; Robb M. A.; Cheeseman J. R.; Montgomery J. A. Jr.; Treven T.; Kudin K. N.; Burant J. C, et al. Gaussian 03, Revision C.02; Gaussian, Inc.: Wallingford, CT, 2003. (33) Benson S. W.; Cruickshank F. R.; Golden D. M.; Haugen G. R.; O'Neal H. E.; Rodgers A. S.; Shaw R.; Walsh R. Additivity Rules for the Estimation of Thermochemical Properties. J. Chem. Phys., 1958, 29, 279-324.


Page 25 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(34) Lay T. H.; Bozelli J. W.; Dean A. M.; Ritter E. R. Hydrogen Atom Bond Increments for Calculation of Thermodynamic Properties of Hydrocarbon Radical Species. J. Phys. Chem., 1995, 99, 14514-14527. (35) Petersson G. A.; Malick D. K.; Wilson W. G. Calibration and comparison of the Gaussian-2, Complete Basis Set, and Density Functional Methods for Computational Thermochemistry. J. Chem. Phys., 1998, 109, 10570-10579. (36) Sharma S.; Raman S.; Green W. H. Intramolecular Hydrogen Migration in Alkylperoxy and Hydroperoxyalkylperoxy Radicals: Accurate Treatment of Hindered Rotors. J. Phys. Chem. A., 2010, 114, 5689-5701. (37) Allen J. W.; Green W. H. CanTherm User's Guide, Reaction Mechanism Generator. http://reactionmechanismgenerator.github.io/RMG-Py/users/cantherm/index.html/ (Accessed July, 11, 2018). (38) Chase M.; NIST-JANAF thermochemical tables 4th ed., New York: American Institute of Physics for the National Institute of Standards and Technology, 1998. (39) Burcat A.; Ruscic B.. Third Millennium Ideal Gas and Condensed Phase Thermochemical Database for Combustion with Updates from Active Thermochemical Tables. Argonne National Laboratory, 2005. (40) Lai L.; Khanniche S.; Green W. H. Thermochemistry and Group Additivity Values for Fused Two Ring Aromatic Species and Radicals. Submitted to J. Phys. Chem. A., 2019. (41) Roux M. V., Temprado M., Chickos J. S., Nagano Y. Critically Evaluated Thermochemical Properties of Polycyclic Aromatic Hydrocarbons. J. Phys. Chem. Ref. Data, 2008, 4, 1855-1996. (42) Narayanaswamy K.; Blanquart G.; Pitsch H.; A Consistent Chemical Mechanism for Oxidation of Substituted Aromatic Species. Combust. Flame, 2010, 10, 1879-1898. (43) Ruscic B.; Boggs J. E.; Burcat A.; Csaszar A. G.; Demaison J.; Janoschek R.; Martin J. M.; Morton M. L.; Rossi M. J.; Stanton J. F.; Szalay P. G.; Westmoreland P. R.; Zabel F.; Berces T. IUPAC Critical



Page 26 of 27

Evaluation of Thermochemical Properties of Selected Radicals Part I. Phys. Chem. Ref. Data., 2005, 34, 573-656. (44) Li Y.-P.; Han K.; Grambow C. A.; Green W. H. Self-Evolving Machine: A Continuously Improving Model

for

Molecular

Thermochemistry.

J.

Phys.

Chem.

https://pubs.acs.org/doi/abs/10.1021/acs.jpca.8b10789


A.

2019,

accepted,

in

press.

Page 27 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC Graphic


Thermochemistry and Kinetics of Intermolecular Addition of Radicals

Recommend Documents