Molecular structure and conformation of 1,3,5-tris(trifluoromethyl)-benzene as studied by gas-phase electron diffraction and quantum chemical calculations

Accepted Manuscript Molecular structure and conformation of 1,3,5-tris(trifluoromethyl)-benzene as studied by gas-phase electron diffraction and quantum chemical calculations Inna N. Kolesnikova, Olga V. Dorofeeva, Nikolay M. Karasev, Heinz Oberhammer, Igor F. Shishkov PII: DOI: Reference:

S0022-2860(14)00579-1 http://dx.doi.org/10.1016/j.molstruc.2014.05.065 MOLSTR 20664

To appear in:

Journal of Molecular Structure

Received Date: Revised Date: Accepted Date:

4 April 2014 24 May 2014 26 May 2014

Please cite this article as: I.N. Kolesnikova, O.V. Dorofeeva, N.M. Karasev, H. Oberhammer, I.F. Shishkov, Molecular structure and conformation of 1,3,5-tris(trifluoromethyl)-benzene as studied by gas-phase electron diffraction and quantum chemical calculations, Journal of Molecular Structure (2014), doi: http://dx.doi.org/ 10.1016/j.molstruc.2014.05.065

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Molecular structure and conformation of 1,3,5-tris(trifluoromethyl)-benzene as studied by gas-phase electron diffraction and quantum chemical calculations

Inna N. Kolesnikova 1*, Olga V. Dorofeeva 1, Nikolay M. Karasev 1, Heinz Oberhammer 2, Igor F. Shishkov 1 1

2

Department of Chemistry, Moscow State University, 119992 Moscow, Russia

Institut für Physikalische und Theoretische Chemie, Universität Tübingen, 72076 Tübingen, Germany

* Corresponding author E-mail address: [email protected] Abstract The molecular structure of 1,3,5-tris(trifluoromethyl)benzene (1,3,5-TTFB) was studied by gas-phase electron diffraction (GED) and quantum chemical calculations (B3LYP method with 6-31G(d,p) basis set and MP2 method with cc-pVTZ basis set). The best fit of the experimental scattering intensities (RG = 4.0%) was obtained for the structure of Cs symmetry. The differences between some geometric parameters were constrained at the values calculated at the MP2/cc-pVTZ level. The principal structural parameters rh1(∠ h1) determined by GED are (bond lengths in Angstroms and bond angles in degrees with 3σ in parentheses):

r(C C)=1.392(4),

r(C−C)=1.512(4),

r(C−F)av=1.346(2),

∠C–C(CF3)–

C–C(H)–C=119.1(2), ∠ (C–C–F)av=111.6(2). The structure of the carbon ring C=120.9(2), ∠ deviates from a regular hexagon due to the σ-electronegative effect of the CF3 groups. The geometric parameters of the trifluoromethyl groups deviate considerably from regular tetrahedron arrangement. The experimental structural parameters agree well with the results of B3LYP/6-31G(d,p) and MP2/cc-pVTZ calculations. The electron diffraction data are in agreement with nearly free rotation of the CF3 groups around Cmethyl-Cphenyl axis.

Keywords: 1,3,5-tris(trifluoromethyl)benzene, molecular structure, gas electron diffraction, quantum chemical calculations. 1

1. Intoduction Three decades ago structural studies of several methyl substituted benzenes [1-4] and benzotrifluoride [5] were performed by GED. In these studies the influence of the substituents on the valence angles and bonds lengths in the benzene ring was investigated. No assumptions about the overall symmetries of the molecules were made because of the almost free rotation of the substitutents around the C-C bonds. STO MO calculations resulted in six-fold barriers in some toluene derivatives [6], confirming almost free rotation of

the

CF3

groups

in

benzotrifluoride,

2,6-difluorobenzotrifluoride

and

2,6-

dichlorobenzotrifluoride. In order to investigate OH…F hydrogen bonding and its geometrical consequences, compared with the ‘parent’ compounds, computational studies (MP2/631G(d,p)) were performed for benzotrifluoride [7], 1,3-bis(trifluoromethyl)benzene [8] and their OH derivatives, 2-trifluoromethylphenol [7], 2-trifluoromethylresorcinol [8] and 2,6bis(trifluoromethyl)phenol [8]. In order to obtain the information on the additivity of structural substituent effects in benzene derivatives, quantum chemical calculations at MP2/6-31G(d,p) level of theory were performed for 1,4-bis(trifluoromethyl)benzene [9] and 1,3,5-tris(trifluoromethyl)benzene [9]. Although numerous theoretical studies on trifluoromethyl substituted benzenes have been reported up to now, to our knowledge, no experimental data are available for such compounds except for benzotrifluoride [5]. In the present study we investigate the gasphase structure of 1,3,5-tris(trifluoromethyl)benzene (1,3,5-TTFB) by means of GED and quantum-chemical calculations to investigate the effect of three trifluoromethyl groups on the ring geometry.

2. Quantum chemical calculations Quantum chemical calculations were performed with the Gaussian03 program package [10] using the B3LYP [11, 12] and MP2 [13] levels of theory. The B3LYP/631G(d,p) calculations, with different starting orientations of the three CF3 groups, resulted in two minima. In both structures one fluorine atom of each CF3 group lies in the plane perpendicular to the ring (“orthogonal” CF3 conformation). The global minimum corresponds to the structure of CS symmetry, in which perpendicular fluorine atoms are disposed on the opposite side of the ring (Figure 1). The second minimum corresponds to the structure of C3v symmetry with the perpendicular fluorine atoms located on the same

2

side of the ring (Figure 1). The geometries of these two conformers were fully optimized with B3LYP/6-31G(d,p) and MP2/cc-pVTZ methods. The energy difference between CS and C3v conformers, as calculated at the MP2/cc-pVTZ level, is less than 0.2 kJ/mol. The geometrical parameters in both conformers are also very similar; differences in the bond angles are within 0.3˚ and in bond lengths within 0.002 Å. These differences are even smaller for the ring skeleton, less than 0.05˚ and 0.0002 Å, respectively. In both conformers the carbon skeleton deviates slightly from planarity with all C С being equal. Similarly, all three C–С distances possess equal length. Furthermore, all three CF3 groups possess a local C3v symmetry and equal geometric parameters except for the torsional angles. The calculated harmonic vibrational frequencies of both conformers are also very similar with the lowest frequencies for CF3 torsional vibrations of 9 cm-1 and 8 cm-1, respectively. The calculated barrier height for CF3 group rotation around Cmethyl-Cphenyl axis is 0.5 kJ/mol. These data indicate nearly free rotation of the CF3 groups around the C−C bonds. Since GED intensities are weakly sensitive to the orientation of the CF3 groups, the scattering intensities and radial distribution curves for CS and C3v conformers are very similar. Furthermore, effects of free or hindered internal rotation on the scattering intensities of 1,3,5-TTFB are rather small.

3. Electron diffraction analysis 3.1. Experimental A commercial sample of 1,3,5-TTFB with purity of 99% was obtained from Aldrich Chemical Co. and used without further purification. Electron diffraction intensities were recorded using the electron diffraction apparatus at Lomonosov Moscow State University. Information about the experimental conditions for all datasets used in the present investigation is given in Table 1. The electron wavelength was calibrated with CCl4 scattering patterns. The optical densities were measured using an Epson Perfection 4870 photo scanner. The data were processed with the program UNEX [14] using standard routines. The final modified intensity curves are shown in Figure 2. The experimental intensity curves with backgrounds and their numerical values are available as Supporting Information (Table S1).

3

3.2. Structural refinements The analysis of GED data was carried out by applying the least squares method to the molecular intensities using UNEX program. In this program, the molecular geometry is specified in a format of a Z-matrix. From the scaled force field obtained by MP2/cc-pVTZ method, the root-mean-square amplitudes of vibrations and perpendicular amplitude corrections for 1,3,5-TTFB conformers were calculated using the SHRINK program [15]. Both conformers are characterized by three rotors around the C(sp2)–C(sp3) bonds. However, the use of a dynamic model with three rotating CF3 groups poses a challenging problem. Therefore, two static models with CS and C3v symmetry corresponding to the energy minima were used in the GED analyses. To reduce the correlation between very similar parameters, their values were refined in groups (see Table 2). The differences between parameters within each group were constrained to calculated values. In addition to the geometrical parameters, the vibrational amplitudes of non-hydrogen distances were refined. Most of the amplitudes were refined together in groups depending on the interatomic distances. These analysis demonstrated that neither of the two conformers with fixed torsion angles nor mixtures of both conformers fitted the GED intensities very well. For a mixture of СS : С3v = 3 : 2 the RG value of 6.8% was obtained. The best fit of the GED intensities (RG = 4.0%) was obtained with a rigid model possessing CS molecular symmetry, but allowing the torsional angles of both symmetrical CF3 groups, C2-C3-C8-F14 and C6-C5C9-F17, to deviate from the exact perpendicular orientation. A partially dynamic model based on CS geometry with two hindered rotated CF3 groups and free rotation around the C1-C7 bond did not improve the fit of the experimental intensities. The results of this GED experiment together with geometry optimization at B3LYP/6-31G(d,p) and MP2/cc-pVTZ level are listed in Table 2. The mean observed amplitudes for bonded non-hydrogen distances in 1,3,5-TTFB are listed in Table 3. The amplitudes for all distances of 1,3,5TTFB are given in Table S2. The error limits were estimated as three times standard deviations (3σ) in the least-squares calculations. On the whole, the agreement between the observed and calculated amplitudes is satisfactory. The radial distribution curve is shown in Figure 3. 3.3 Results and discussion In previous studies of benzene derivatives with trifluoromethyl substituents the CF3 group was predicted to be in “orthogonal” orientation with one of the CCF planes being 4

perpendicular to the plane of the ring [7-9]. Our theoretical investigations of 1,3,5-TTFB confirmed these results and indicated the existence of two possible conformers of C3v and СS symmetry. To study internal rotation, different models were considered in this work. The single conformer models of C3v and СS symmetries with fixed torsional angles of rotation and their mixture were eliminated due to rather high R-factors. The CS model, allowing an effective average torsional angles of both symmetrical CF3 groups resulted in the best fit to the experimental GED intensities with RG = 4.0%. The refinements have been carried out using various initial values for the rotational angle φ(C2 C3–C8–F14). However, the results were independent of initial values. The refined φ value (Table 2) deviates by about 20° from the equilibrium value and may be regarded as an effective value which simulates the practically free rotation of the CF3 group around the C−C bond. An interesting structural feature of substituted benzene compounds is the ring deformation. The C C C bond angle adjacent to the substitutent (αipso) is the most sensitive geometrical parameter characterizing the ring deformation. This angle may change by several degrees depending on the nature of the substitutent. Very electronegative substitutents strongly increase this bond angle, e. g. αipso = 121.7° in chlorobenzene [17] and 123.4° in fluorobenzene [18]. The electronegativity of the CF3 group (χ = 3.3) is estimated to be between those of fluorine (χ = 4.0) and chlorine (χ = 3.0). This suggests a considerable ring deformation in 1,3,5-TTFB with a considerable increase of αipso. On the other hand, empirical correlations were observed between αipso and the substitutent electronegativity, indicating a systematic difference between the changes generated by second row and third row elements [19]. This result was later confirmed by MO calculations [20]. According to these correlations, a very small ring deformation is expected for CF3 substituted benzene molecules. An experimental GED study of benzotrifluoride results in a slightly distorted ring with a small decrease of the αipso angle (119.7(2)°) and an increase of the C C C angle in ortho position (αortho = 120.4(2)°) [5]. However, later theoretical investigations of benzotrifluoride [7, 9] at the MP2/6-31G(d,p) level predicted an increase of αipso (120.8º) and a decrease of αortho (119.4°) [7, 9]. These contradicting results between experiment and theory for benzotrifluoride make the results of the present study concerning ring deformation in 1,3,5-TTFB especially interesting. Previous theoretical investigations of 1,3,5-TTFB using the MP2/631G(d,p) method result in values for αipso and αortho of 121.0º and 119.0º, respectively [9]. In the present work the MP2/cc-pVTZ calculations predict very similar values of αipso = 5

120.9 and αortho = 119.1º. The DFT/B3LYP/6-31G(d,p) method predicts a slightly smaller ring distortion with of αipso = 120.5 and αortho = 119.5º. The values for αipso and αortho obtained from the GED investigation (αipso = 120.9(2)° and αortho = 119.1(2)º) are in agreement with the MP2/cc-pVTZ calculations (Table 2) and correspond to a considerable ring deformation. The results for these angles in 1,3,5-TTFB agree with the general observation that substituents with negative σ-inductive effect increase αipso and decrease αortho in the benzene ring. The structural parameters of the trifluoromethyl group with F–C–F 106.9(2) differ considerably from the regular tetrahedral arrangement, in close agreement with the values predicted by B3LYP (107.2°) and MP2/cc-pVTZ methods (107.4°, see Table 2). The value for the F–C–F angles given in Reference 9 (111.6°) seems to be incorrect, since the value of the C–C–F angles is also larger than tetrahedral. The average C–F bond length (rh1 = 1.346(4) Å) is much larger than that in CF4 (rg 1.317(2) Å)[21]. The MP2 method predicts slightly shorter and the B3LYP slightly longer C–F bonds. Table 4 compares (C–F)av and C(sp2)–C(sp3) single bonds in methyl and trifluoromethyl substituted benzene compounds. The (C–F)av bonds are equal in benzotrifluoride and 1,3,5-TTFB. Within their experimental uncertainties the C–C single bonds are equal in all compounds, except benzotrifluoride, where this bond is shorter. 4. Concluding remarks In the present study the molecular geometry of 1,3,5-TTFB in the gas phase was determined experimentally by GED and was investigated theoretically by means of B3LYP/6-31G(d,p) and MP2/cc-pVTZ calculations. The calculations predict the presence of two conformers with C3v and CS symmetry which are very close in energy. In these equilibrium structures one CCF planes of each CF3 group is perpendicular to the benzene ring and the CF3 groups are almost freely rotating around Cmethyl–Cphenyl axes. Since the use of a dynamic model with three rotors is rather unrealistic, various rigid models were applied. The best fit (RG = 4.0%) was obtained for a model of CS symmetry with two symmetrical CF3 groups rotated away from the perpendicular equilibrium orientation by about 20°. The results for 1,3,5-TTFB reflect the trends observed for substituents with a negative σ-inductive effect, i.e. increase of αipso and decrease of αortho, confirming a wellknown fact that σ-interactions between the substituent and the ring are mainly localized in the immediate vicinity of the substitution. The experimental structural parameters obtained

6

in the GED study are in good agreement with those predicted by B3LYP/6-31G(d,p) and MP2/cc-pVTZ calculations.

Acknowledgments This research was supported by the Russian Foundation for Basic Research under Grants the 11-03-00716-a and No. 12-03-91330-NNIO-a and by the Deutsche Forschungsgemeinschaft, Grant OB28/22-1.

Appendix A. Supplementary material Supplementary material associated with this article can be found in the online version.

7

References [1] R. Seip, G. Schultz, I. Hargittai, Z. Naturforsch., 32A (1977) 1178-1183 [2] F. Pang, J. Boggs, P. Pulay, G. Fogarasi., J. Mol. Struct., 66 (1980) 281-287. [3] A. Domenicano, G. Schultz, M. Kolonits, I. Hargittai., J. Mol. Struct., 53 (1979) 197209. [4] Almenningen A., Hargittai, I., Samdal, S., Brunvoll, G., Domenicano, A., Lowrey, A., J. Mol. Struct., 96 (1983) 373-377. [5] G. Schultz, I. Hargittai, R. Seip., Z. Naturforsch., 36a (1981) 669-673. [6] T. Schaefer, G. H. Penner, J. Mol. Struct. (Theochem), 138 (1986) 305-310. [7] A. Kovács, I. Kolossváry, G. I.; Csonka, I. Hargittai., J. Comput. Chem, 17 (1996) 1804-1819. [8] A. Kovács, I. Hargittai, J. Mol. Struct. (Theochem), 455 (1998) 229 –238. [9] A. Kovács, I. Hargittai, Struct. Chem., 11 (2000) 193 - 200. [10] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A. Montgomery, T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, and J.A. Pople, Gaussian, Inc., Wallingford CT, 2004. [11] A. D. Becke, Phys. Rev. A, 38 (1988) 3098. [12] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B, 37 (1988) 785. [13] C. Møller, M.S. Plesset, Phys. Rev., 46 (1934) 618. [14] Yu.V. Vishnevskiy, 2009. . [15] V.A. Sipachev, J. Mol. Struct. THEOCHEM, 22 (1985) 143. [16] S. L. Masters, S. J. Atkinson, M. Hölbling, K. Hassler, Struct. Chem., 24 (2013) 1201– 1206. [17] G. Schultz, I. Hargittai, A. Domenicano, J. Mol. Struct., 68 (1980) 281. 8

[18] L. Nygaard, I. Bojesen, T. Pedersen, and J. Rastrup-Andersen, J. Mol. Struct., 2 (1968) 209. [19] A. Domenicano, A. Vaciago, and C. A. Coulson, Acta Cryst., 31 (1975) 1630. [20] A. R. Campanelli, A. Domenicano, F. Ramondo, I. Hargittai, J. Phys. Chem. A 108 (2004) 4940-4948 [21] C.W.W. Hoffman, R.L. Livengston, J. Chem. Phys., 21 (1953) 565.

9

Table 1. Experimental conditions of gas-phase electron diffraction experiment. Long camera

Short camera

Camera distance (mm)

362.16

193.82

Nozzle temperature (K)

298

298

Accelerating voltage (kV)

60

60

Electron wavelength (Å)

0.049260

0.049196

Number of films used

3

3

Range of s value (Å-1) a

3.8–17.8

7.0–32.6

Scale factor (conformers’ mixture )

1.116 (17)

0.987 (24)

Scale factor

1.109(16)

0.970(22)

a

s =4πλ-1sinθ/2, where θ is the scattering angle and λ is the electron wavelength.

10

Table 2. Molecular structure of СS conformer of TTFB obtained by gas-phase electron diffraction and quantum chemical calculations Parameter a

GED

rh1(∠ h1)

B3LYP/

MP2/

MP2/

6-31G(d,p)

cc-pVTZ

6-31G(d,p)b

re(∠ e)

re(∠ e)

re(∠ e)

Independent parameters r(C С)

1.392(4)

1.395

1.391

1.395

r(C–С)

1.512(4)

1.508

1.503

1.501

c

r(C7–F10) r(C7–F11)

1.348(2) 1.345(2)c

1.351 1.349

1.341 1.338

1.335 1.335

r(C–H)

1.079f

1.083

1.079

1.081

∠C–C(CF3)–C ∠C–C(H)–C ∠C2–C3–C8

d

120.9(2) 119.1(2)d 119.6(2)

120.5 119.5 119.5

120.9 119.1 119.6

121.0 119.0 119.5

∠C1–C7–F10 ∠C1–C7–F11 ∠F10–C7–F11

111.1(2)e 111.9(2)e 106.9(2)

111.1 111.6 107.2

110.9 111.6 107.4

111.1

φ(C2–C3–C4–C5) φ(C1–C2–C3–C8) φ(C2–C3–C8–F14)

-0.4f 177.7f -67.4(8)

0.1 177.5 -88.8

-0.4 177.7 -88.7

Dependent parameters φ(C1–C2–C3–C4) φ(C5–C6–C1–C7)

0.2f -177.4f

-0.1 -176.9

0.2 -177.4

φ(C6–C1–C7–F10)

88.7g

88.5

88.7

RG

4.0

111.6

a

Bond lengths in Å and bond angles in degree with three standard deviations in parentheses. For atom numbering see Figure 1. RG is the goodness of fit factor [16] in %. b

Ref. 9

c,d,e

Refined in one group. Differences between parameters in the group were assumed at the values from MP2/cc-pVTZ calculation

f

Assumed at the values of MP2/cc-pVTZ calculation

g

This dihedral angle corresponds to an exact perpendicular orientation of the C7-F10 bond. The deviation from 90° is due to slight non-planarity of the carbon skeleton.

11

Table 3. Average values of calculated and observed vibrational amplitudes for bonded nonhydrogen distances in 1,3,5-TTFB.a GED

Calcb

GEDc

ra

u

u

(C–F)av

1.346

0.045

0.050(2)

C C

1.392

0.044

0.050(2)

C–C

1.512

0.048

0.054(2)

Distance

a

Values of distances (ra), and amplitudes (u) are in Å.

b

MP2/cc-pVTZ calculation

c

These amplitudes were refined in one group; differences between their values were assumed from theoretical calculation.

12

Table 4. Comparison of the average C-F and C-C bond lengths in 1,3,5-TTFB and related compounds determined from GED studies a Parameter 1,3,5-TTFBb benzotrifluoridec 1,3,5-trimethylbenzened (C–F)av

1.346(4)

1.345(3)

C–C

1.512(4)

1.504(4)

1.509(2)

trimetylbenzenee

1.511(8)

a

Bond lengths are in Å, rh1 parameters are given for 1,3,5-TTFB, rg for all others. bThis work. c Ref. 5. d Ref. 4. eRef. 1.

13

F(10)

F(10) F(11)

F(11)

F(12)

F(12) C(7)

C(7) H(21)

C(1)

C(1)

H(21)

H(19)

C(6) C(2)

C(2) C(5)

C(5) C(9)

F(16)

F(13)

C(8) C(4)

F(13)

F(16)

C(9)

C(8) C(3)

C(3) F(14)

F(17) F(18)

H(19)

C(6)

H(20)

F(15)

F(17)

C(4)

F(18)

F(14) F(15)

H(20)

Figure 1.

14

sM(s)

ΔsM(S)

0

5

10

15

s, Å -1

20

25

30

35

Figure 2.

15

f(r)

CF

F... F

C ...C

F... F

C C

C... F

C ...C

C ...C

C ...C C ...C

Δf(r) 0

C... F

C... F C ...C

C C

1

2

3

r, Å

4

C... F

C ...C

5

C... F

6

7

Figure 3.

16

Figure captions Figure 1. Molecular models of 1,3,5-TTFB (CS and C3v symmetry) with atom numbering. Figure 2. Experimental (open circles) and theoretical (solid line) molecular intensities sM(s) and the difference curve ΔsM(s) for CS conformer model. Figure 3. Experimental (open circles) and theoretical (solid line) radial distribution curves f(r) with difference curve Δf(r) for 1,3,5-TTFB. The distribution of non-hydrogen distances is indicated by vertical bars.

17

   

The molecular structure of 1,3,5-tris(trifluoromethyl)benzene was investigated. Gas-phase electron diffraction and quantum chemical calculations were applied. The best fit was obtained for the structure of Cs symmetry. The carbon ring was found to deviate from a regular hexagon.