The Molecular Structure of Methylfluoroisocyanato Silane - American

Dec 31, 2014 - Brooks H. Pate,. § ... Department of Chemistry, University of Virginia, McCormick Road, P.O. Box 400319, Charlottesville, Virginia 229...
1 downloads 0 Views 560KB Size
Subscriber access provided by ERCIYES UNIV

Article

The Molecular Structure of Methylfluoroisocyanato Silane: A Combined Microwave Spectral and Theoretical Study Gamil A. Guirgis, Jason S. Overby, Timothy J. Barker, Michael Henry Palmer, Brooks H. Pate, and Nathan Andrew Seifert J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/jp511354j • Publication Date (Web): 31 Dec 2014 Downloaded from http://pubs.acs.org on January 6, 2015

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 free 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 accessible to all readers and 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.

The Journal of Physical Chemistry A 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 24

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

The Molecular Structure of Methylfluoroisocyanato Silane: a Combined Microwave Spectral and Theoretical Study Gamil A. Guirgisa, Jason S Overbya, Timothy J. Barkera Michael H. Palmerb, Brooks H. Patec, Nathan A. Seifertc a

Chemistry & Biochemistry, College of Charleston, Charleston, SC 29424 USA. b School of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 3JJ, UK. c Department of Chemistry, University of Virginia, McCormick Road, P.O. Box 400319, Charlottesville, VA22904-4319, USA. Abstract The structure of methylfluoroisocyanato silane (Me-SiHF-NCO) has been deduced by a combination of microwave (MW) spectra including data from 12,13

C, 14,15N, and 28,29,30Si isotopomers, and ab initio calculations. The rotational

constants (RC) for the most abundant isotopes are A = 6301.415(45), B = 1535.078(39) and C = 1310.485(39) MHz. The symmetric quartic centrifugal distortion constants have been identified, using the Ir representation for C1 symmetry, which includes the 3-fold rotor. The spectra of the isotopomer combinations gave a partial substitution structure where the C2Si3, Si3N4 and N4C9 bond lengths are 1.8427(70), 1.7086(77) and 1.2120(90)Å; although the C2Si3N4 angle is close to tetrahedral (109.71º (52)), the Si3N4C9 angle is wide (157.69º (18)). The rotational constants are only consistent with a transorientation for each of the dihedral angles (HC2Si3N4, C2Si3N4C9 and Si3N4C9O10). The structural analysis was completed by calculations of the equilibrium structure, using MP3 in conjunction with an aug-cc-pVTZ basis set (434 Cartesian basis functions). This gave A = 6240.324, B = 1518.489 and C = 1297.819 MHz. The equilibrium structure bond lengths for C2Si3, Si3N4 and N4C9 were 1.8485, 1.7147 and 1.1947 Å, with the C2Si3N4 and Si3N4C9 angles 109.55

and 156.67º respectively. Although the SiNC polynomial bending

ACS Paragon Plus Environment

1

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 24

surface is complex, the data points can be fit to the simple form V(x) = 50.36(91) x4 - 7.53(44) x5, with comparatively little loss of accuracy. The Arotational constant is strongly influenced by the Si3N4C9 angle, and smaller bases lead to this angle being nearly linear. The theoretical results suggest a very heavily polarised molecule, which is supported by the positions of the local electron density minima within the bonds and electron density calculations. 1. Introduction Recently, we reported combined microwave spectral (MW) and theoretical studies of two new isocyanates, namely difluoroisocyanato-silane (1, HF2SiNCO)1 and difluoroisocyanatomethyl-silane (2, MeSiF2NCO),2 both of which showed wide SiNC angles of 154.6 and 157.5º respectively. These molecules (1 and 2) are examples of quasi-linear or ‘floppy’ molecules, which have large amplitude vibrational modes, and often have vibrationally excited states close to the ground state.3,4,5 Spectral analyses under these circumstances often show extreme centrifugal distortion which complicates both MW and infrared (IR) spectral interpretation,6 and the necessity to use much higher terms such as quartic and sextic terms in the potential energy surfaces. The calculated equilibrium structures for both 1 and 2, determined coupled cluster (CCSD(T)), configuration interaction (MP2)

using

and density

functional (DFT, B3LYP) methods, gave satisfactory agreement with experimental rotational constants. We have now synthesised the closely related compound methyl (fluorosilyl) isocyanate, MeSiHFNCO (3) (Figure 1A,B) as a chiral mixture, and

ACS Paragon Plus Environment

2

Page 3 of 24

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

investigated its molecular structure by similar combined MW and theoretical methods. The present theoretical study for compound 3, includes extensive MW experimental study alongside Moller-Plesset and density functional (DFT, especially B3LYP) calculations of the equilibrium structure.

Figure 1(A,B). The molecular skeleton of 3, which is a staggered silaethane. The C1 symmetry leads to non-equivalence during the methyl group internal rotation, best seen from the Newman projection in Figure 1B, where all three Hatoms are non-equivalent.

2. Synthetic procedure for methylfluoroisocyanato silane. Slow addition of methyldichlorosilane to silver cyanate under vacuum gave crude methyldichlorosilyl isocyanate, after cooling to -100 oC and the volatile products removed by degassing for few minutes. The reaction flask is warmed to room temperature and stirred for 24 hrs. Bulb-to-bulb separation of the product using 3 U-tubes, cooled to -40, -80 and −196 °C, in sequence under vacuum, collected the isocyano product, methyldichlorosilane and volatile products respectively. The isocyano product was then fluorinated by antimony trifluoride, without a solvent. The final product 3 was further purified by trapto-trap distillation, and identified by infrared, nuclear magnetic resonance (1H

ACS Paragon Plus Environment

3

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

and

19

F) and microwave spectroscopy. This compound (3) is not indexed in

Scifinder Scholar, and no other reports of its synthesis are known. The reaction scheme is shown in Figure 2. Figure 2. The synthetic route

The NMR chemical shifts for CH3-SiHF-N=C=O (3, CDCl3 solution, measured from Me4Si) for each nucleus were: 1H (a) CH3 δ 0.35 ppm (a doublet of doublets, 3JHH 6.5 Hz); (b) H(Si) δ 4.77 ppm (a doublet of quartets, 2JHF of 68.9 Hz); 19F δ -137.7 ppm (a doublet of quartets, 2JFH of 69.0 Hz. The MW absorption discussed below also confirms the identity of the compound.

3. Microwave spectral methods The rotational spectrum of CH3SiHFNCO was studied using a chirped-pulse Fourier transform microwave (CP-FTMW) spectrometer at the University of Virginia, operating in the 6.5 to 18 GHz range. The chirped pulse methodology used in this study has been described in detail previously,1,2 so only brief details of the experiment will be reported here. The sample was prepared by mixing CH3SiHFNCO (3) vapor with Ne gas (GTS Welco) at 3.4 atm, giving a total sample concentration of ~0.2%. Approximately 12 000 valve injection cycles, using five pulsed nozzles, were completed at 4

ACS Paragon Plus Environment

4

Page 4 of 24

Page 5 of 24

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

Hz to generate a time-averaged spectrum of 120 000 molecular free induction decays (FIDs). Each FID was recorded for 20 µs, generating a Dopplerbroadened linewidth of approximately 120 kHz at full-width/half-maximum. The resultant spectrum has a dynamic range of 3600:1 with approximately 6500 observable lines with a signal/noise ratio greater than 3:1; the rotational temperature measured for the detected species was approximately 1 K. Due to the C3-symmetric geometry of the off-axis methyl rotor, each observed transition was doubled due to high-barrier internal rotation effects. The doubling corresponds to pure rotational transitions in states corresponding to the two irreducible representations (A/E) of the C3 point group. The transitions corresponding to the E symmetry state cannot be fit by the standard rigid rotor Hamiltonian; this arises from the presence of a non-zero, second-order perturbative correction to the rigid rotor rotational constants in the overall torsion-rotation Hamiltonian.7 This second-order effect causes the effective rotational constants of the E state to be functions of the angular momentum of the rotor, causing a breakdown of the rigid rotor approximation for this state. Therefore, the spectrum was first assigned by fitting the A symmetry state (where the perturbative coefficients are zero) to a rigid rotor Hamiltonian using the CALPGM program suite.8 This fitted A state was then used for constant scaling and prediction of isotopologic spectra. The coupled A/E spectrum was then refit using XIAM,9 for a fit value for V3 of 480(19) cm-1; this value is close to the MP2/6-311++g(d,p) value of 436 cm-1 giving confidence in the resulting ACS Paragon Plus Environment

5

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 6 of 24

fit. A summary of the present microwave results is shown in Table 1. A full listing of the fit constants for the parent and other isotopic species can be found in the Supplementary Material (SM, Tables S1 to S11). As well as the high abundance isotopic data (1H, 12C 14N, 16O), the high dynamic range afforded by the experiment was sufficient to detect spectra from both 13C centres, as well as singly-substituted

29

Si,

30

Si and

natural abundance; these concentrations are:

15

N isotopologues, each in

29

Si (4.67%),

30

Si (3.1%),

13

C

(1.1%), 15N (0.4%). Except the 15N species, the others were fit using XIAM to account for the internal rotation, assuming internal rotation parameters fixed to the fit values determined from the parent species. The transitions for the

15

N

isotopologue, were sufficiently weak to prevent assignment of the A/E-coupled fit, and only an A-state fit was determined by using CALPGM. The rotational constants afforded by the isotopologues enable direct determination of the experimental structure of the planar CSiNC backbone of CH3SiHFNCO via application of Kraitchman’s equations.10

These equations

return the Cartesian coordinates of all detected isotopically-substituted atoms in the molecule, and can be directly compared to the ab initio structure. However, atoms close to a principal axis (here principally the O atom) can subsequently generate some error in the rs coordinates. The calculated inertial axis structure (Figure 3A,B), which is discussed below, demonstrates this aspect.

ACS Paragon Plus Environment

6

Page 7 of 24

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

4. Theoretical methods. The electronic structure, including electric field gradients (EFG) and derived 14

N nuclear quadrupole coupling constants (NQCC), for the title compound (3),

were determined using the GAUSSIAN-0911 suites of programmes. The final (MP3) wave-function was analysed for local electron density minima (Del(ρ) = 0.0) in the bond skeleton using the Atoms in Molecules (AIMPAC)12,13 procedure. Initial studies using the largest Pople-style basis (6-311++G (3df, 3pd)),14,15 which was successful with molecules 1 and 2 above failed to give an acceptable value for the A-rotational constant. We found that the MP2 method with

the

aug-cc-pVTZ

basis

set

(H[4s3p2d],

C,N,O,F

[5s4p3d2f],

Si[6s5p3d2f]).16,17 gave an acceptable A-rotational constant. When this was refined by MP3 calculation, the agreement with experiment was good, as shown in Table 2. 5. Results. All molecular conformations for 3 have C1 symmetry; cis/trans conformers at HCSiN, CSiNC and SiNCO conformers were studied, but the values of the observed rotational constants (RC) show that only the isomer with all-trans dihedral angles gave values close to the observed RC. The MP3 equilibrium structure (and dipole moment components) lying in the inertial axis frame are shown in Figure 3a,b. The MP3 atomic coordinates of the equilibrium structure are shown in the SM(S5).

ACS Paragon Plus Environment

7

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

Figure 3 a,b. The a,b-(3a) and a,c (3b)-inertial axis frames. Table 1. A summary of the principal microwave spectroscopic results. for the parent isotopic species; using overall A/E symmetry states of the ground state torsion-rotation spectrum. Normal /rs Rotational constants / MHz A B C F (CH3 local B-rotational constant)) Θ (3-fold rotor a-axis projection) Quartic centrifugal distortion / kHz ∆J ∆JK ∆K δJ δK 14 N Nuclear quadrupole coupling / MHz 1.5χaa 0.25(χbb-χcc) Bonds (Ẵ) and Angles (º) C2Si3 Si3N4

MW (XIAM) 6301.415(45) 1535.078(39) 1310.485(39) 155.7(66) 58.6(10) 0.742(23) 41.50(14) -25.39(10) 0.067(13) 25.6(19) 2.655(11) 0.023(18) 1.8427(70) 1.7086(77)

ACS Paragon Plus Environment

8

Page 8 of 24

Page 9 of 24

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

N4C9 C2Si3N4 Si3N4C9 C2Si3N4C9

1.2120(90) 109.71(52) 157.69(18) 158(2) a,b

Table 2. The principal theoretical results Property

aug-ccpVTZ MP3 Total energy -597.16267 / a.u. Cartesian 434 basis functions Rotational constants / MHz A 6240.3242 B 1518.4894 C 1297.8185 b 163.4 GHz F c 62.8 Θ

aug-ccpVTZ MP2 -597.15747

aug-ccpVTZ HF/ B3LYP -598.21831

6-311**G [3df,3pd] MP2 -597.13347

6-311**G [3df,3pd] HF/ B3LYP -598.21349

434

434

314

314

6205.9862 1513.5552 1291.8729

6104.3822 1500.8527 1284.6281

6192.3760 1511.8617 1293.0796

6116.7144 1506.0636 1290.6131

2.231 -1.541 -0.677 2.795

2.064 1.557 0.552 2.644

Dipole moments µAA 1.840 2.076 2.084 µBB -0.909 -1.553 -1.698 µCC -0.931 -0.550 -0.516 µTotal 2.686 2.650 2.738 Quartic centrifugal distortion / kHz ∆J 0.57 0.47 ∆JK 29.73 28.43 ∆K -19.04 -3.81 δJ 0.05 -0.02 δK 14.15 13.78 14 N Nuclear quadrupole coupling terms / MHz 1.5χaa 2.760 2.713 2.677 0.25(χbb-χcc) 0.006 0.029 0.0004 Bonds (Ẵ) and Angles (º) H1 C 2 1.0896 1.0894 1.0907 C2Si3 1.8486 1.8468 1.8527 Si3N4 1.7147 1.7216 1.7152 N4 C 9 1.1947 1.2083 1.1953 C9O10 1.1606 1.1757 1.1677 H7Si3 1.4717 1.4716 1.4752 F8Si3 1.5973 1.6085 1.6123 H1C2Si3 110.6 110.7 110.5 C2Si3N4 109.6 108.9 110.2

0.49 27.10 -5.98 0.01 12.95

0.31 29.1 0.37 0.01 -0.80

1.0889 1.8396 1.7086 1.2051 1.1724 1.4661 1.6038 110.8 109.2

1.0907 1.8471 1.7065 1.1948 1.1661 1.4716 1.6062 110.5 110.4

ACS Paragon Plus Environment

9

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

C2Si3H7 113.9 C2Si3F8 109.4 Si3N4C9 156.7 N4C9O10 177.4 N4Si3F8 107.4 H7Si3F8 107.4 H1C2Si3N4 178.1 H1C2Si3H7 -59.7 H1C2Si3F8 60.5 C2Si3N4C9 177.0 F8Si3N4C9 -64.2 Si3N4C9O10 -177.9 Footnotes to Table 2

114.2 109.6 152.4 176.7 107.7 107.3 178.2 -59.9 60.6 177.8 -63.4 -179.0

114.0 109.4 163.2 178.0 107.5 107.0 178.4 -59.3 60.5 -173.8 -54.7 -178.2

113.9 109.5 158.4 177.3 108.0 107.2 178.7 -59.5 60.6 -176.3 -57.2 179.0

Page 10 of 24

113.8 109.3 165.6 178.3 107.7 107.0 178.7 -59.1 60.4 -168.5 -49.3 -177.9

a. The number of rows in the Table 2 has been reduced by removal of most bond angles. The coordinates for the MP3 equilibrium structure (in Supplementary Material as S11) enable all these to be regenerated. b. F is the reduced rotational constant of the CH3 rotor c. 3-fold rotor a-axis projection (degrees)

5.1 Comparison of the MW and theoretical data. (a) Rotational constants and centrifugal distortion constants. Comparison of the experimental and MP3 calculated RC show absolute errors in A, B and C: 61, 17 and 13 MHz, where the calculated are low by about 1%. Since the two sets of results reflect different structures, substitution and equilibrium respectively, there is no necessity for the two sets to be identical. However, the close values are indicative of a close relationship. Attempts to evaluate the centrifugal distortion constants by the MP3 method proved prohibitive in CPU cost. Again, comparison of data from Tables 1 and 2 shows that all the theoretical results are numerically smaller than experiment. The MP2 results show better agreement with experiment than the corresponding

ACS Paragon Plus Environment

10

Page 11 of 24

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

B3LYP values, while the aug-cc-pVTZ results are slightly better than the 6311G**[3df,3pd] set. (b) 14N nuclear quadrupole coupling (NQC) terms The correlation between the observed and calculated values are particularly good; using the LaPlace relationship, we obtain χAA 1.781, χBB -0.843 and χCC 0.938 MHz; these are to be compared with the MP3 (and B3LYP in parentheses) values χAA 1.840 (1.809), χBB -0.908(-0.846) and χCC -0.932 (0.963) MHz respectively. All of the off-diagonal NQC terms are also small, with the MP3 values: χAB

0.243, χAC 0.179 and χBC -0.037 MHz. Compared

with many NQC terms at 14N, these values are numerically small, and the small difference between χBB and χCC is a further indicator of the close to cylindrical electric field gradients at the N-atom. (c) Comparison of the equilibrium and partial substitution structures. Comparison of the partial substitution structure and various calculated equilibrium structure bond lengths and angles from Tables 1 and 2 shows generally close agreement between the limited sets of data. The MW (rs) and calculated (re) bond length comparisons show the calculated are consistently slightly longer. The calculated values for the two angles C2Si3N4 and Si3N4C9 are slightly smaller, both by 3.8%. In contrast, the values for the dihedral angle C2Si3N4C9 are significantly different (158 and 177º for rS and re respectively); there is no explanation for that difference. Although no MW data is available for the C9O10 double bond, in all calculated values, the N4C9 is slightly larger;

ACS Paragon Plus Environment

11

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

this implies slightly more double bond character for C9O10, and we have noted this previously.6,18 The theoretical values for the Si3N4C9 angle show more variation, and range from 156.7º (MP3), 152.4º (MP2) and 163.2º (B3LYP); the experimental value is 157.69(18)º. As mentioned above, smaller basis sets used gave nearly linear values for this angle. A significant difference between the observed MW value (158(2)º), and the theoretical values is the dihedral angle C2Si3N4C9 where the theoretical values are close to the trans-value of 180º. This would lead directly to a difference in the B,C-rotational constant balance. The final MP3 structure, showing the heavy atom bond lengths and angles is shown in Figure 4; analysis of the wavefunction to determine Mulliken gross atomic charges is discussed below.

Figure 4. The MP3 equilibrium structure. 5.2 The methyl group rotational spectrum Due to line intensity limitations, we were only able to fit the internal rotation in the normal species and the silicon isotopologues. All species have standard rigid

ACS Paragon Plus Environment

12

Page 12 of 24

Page 13 of 24

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

rotor fits, and these species have SPFIT A-state fits and XIAM9 internal rotation fits. All of the transitions in the main MW rotational spectrum were doubled owing to internal rotation splitting from the methyl group. For many transitions this A and E state rotational splitting was similar in order of magnitude to the nuclear quadrupole hyperfine structure arising from the 14N nucleus, and hence required special attention. The final analysis of the spectrum was carried out using the XIAM program to perform a combined fit for internal rotation and quadrupole hyperfine structure. A set of the observed transitions and their assignment is shown in the Supplementary Information (Table S1). Only a small subset of observed transitions could be precisely fitted using this approach, since many transitions showed unresolved A-E and/or quadrupole hyperfine splitting. The MW experimental 3-fold rotation barrier (V3) was determined as 480(19) cm-1. The MP2 method, used to determine the theoretical 3-fold rotation barrier (Figure 5), gives a result close to experiment, with V3 464.3 cm1

.

ACS Paragon Plus Environment

13

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

Figure 5. Me-SiHF-NCO MP2 energy surface for HCSi angle bending; the B3LYP surface is very similar.

5.3 The energy surface for SiNC bending. For a variety of HSiNC, CSiNC and SiNCO dihedral angles with reasonable bond and angle parameters for the MeSiHFNCO molecule, test computations showed that the all-trans conformer was the only one having rotational constants in the correct proportions in comparison with the MW results. As in structures 1 and 2,1,2 the most important single parameter is the angle SiNC. In the skeletal bending study, the SiNC angle was held at fixed values between 80º and 270º, whilst the remainder of the molecule was allowed to relax; the resulting energy surface is very similar for both the MP2 and B3LYP

ACS Paragon Plus Environment

14

Page 14 of 24

Page 15 of 24

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

methodologies, and the former is shown in Figure 6. These surfaces clearly are very different from parabolic; the shallow global minimum shows

the

equilibrium SiNC angles are: 163.0º (B3LYP) or 152.4º (MP2); these are almost averaged in the MP3 result shown earlier in Figure 4 (156.7º). Differences for this SiNC parameter between the methodologies used here are not unusual, and have been discussed previously;1,2 the calculated SiNC angles for HF2SiNCO (1) are: 171.2 (B3LYP), 167.7 (MP2), 154.9 (MP4) and 154.6° (CCSD(T))1,2 exemplify this. The numerical fitting of the potential energy surface (V) is normally given in terms of energy (V, cm-1) versus angle (x-x0, radian) where x0 is the minimum value.1,2 If the potential arises from a symmetrical environment, then only even terms in the power series expansion occur (x2N where N = 1,2,3….), leading to harmonic, quartic etc. Since the Me-SiHF group is an unsymmetrical environment, the potential energy surface for bending of the SiNC angle also allows odd terms (xN, where N = 1,2,3….). Successive fitting of the data points with a general series of polynomials of form: V = Σn (an*xn) necessary to give a uniform tight fit, showed that the closest polynomial which is effectively an exact fit (Fit1) is: V = -30.1x3 + 62.4x4 + 13.9x5 -22.2x6 -10.39x7 +10.2x8 + 1.9x9 -1.4x10, where all terms were larger than their asymptotic standard errors (SE).

ACS Paragon Plus Environment

15

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

Figure 6. Me-SiHF-NCO B3LYP energy surface for SiNC angle bending; the MP2 surface is very similar. Most experimental ‘fits’ do not go above quadratic and quartic functions, and in order to allow comparisons with that procedure, we sought the simplest power series fit, which leads to acceptably small differences between calculated and fit values close to the minimum, and contains only cubic + quartic and quartic + quintic terms; this is the quartic equivalent to the more usual harmonic and cubic potential energy terms (binary fits): V(x) = ax2 – bx3 (in this instance, for a physically meaningful potential, the terms have opposite signs). The results (Table 3), show a comparison with the binary fits for HF2SiNCO(1)1 and MeSiHFNCO(2)2 under the same fitting procedure. These fits show the same features for all three molecules, where a predominance of quartic terms occurs.

ACS Paragon Plus Environment

16

Page 16 of 24

Page 17 of 24

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

Table 3. Comparable fits for two term polynomials (Fit2) with standard errors (SE) in parentheses. The dominant contributions to the potential energy are quartic (b), but the anharmonicity was introduced by either cubic (a) or quintic (c) terms. MeSiHFNCO (3)

MeSiF2NCO(2)

HF2SiNCO(1)

bx4

cx5

bx4

cx5

bx4

50.4(9)

- 7.5 (4)

53.3 (17)

– 14.8 (11) 59.7 (6)

– 20.7(30)

ax3

bx4

ax3

bx4

ax3

bx4

-28.7(7)

48.1(4)

-47.4(4)

14.7(9)

-5.9(25)

44.2(9)

cx5

Generally, the terms in Fit 2 have small errors, but the data for 2 is less similar than for 1 or 3. A simple measure of the quartic component in the three molecules, from the quartic and quintic surfaces when normalised (b2/(b2+c2)) is: 99(3), 94(2) and 89(1)% respectively. Given the inexact ‘fits’ to these 2term equations, this probably indicates a relatively uniform high quartic component over the three molecules. 5.4 The electron distribution. The present Paper is largely devoted to determination of the molecular structure of compound (3); the ab initio methods generate a full structure, similar to the partial substitution structure obtained experimentally. The calculated structure also generates the electron distribution, which is briefly described here. Two major methods for determination of polarity in chemical bonds use either Mulliken analysis,20 where the density of basis functions (AOs) on individual

ACS Paragon Plus Environment

17

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

centres are summed This method has the disadvantage of ignoring the contributions of density to adjacent centres when diffuse functions are present. Alternatively we have used the Atoms in Molecules (AIMPAC) approach,12,13 which involves direct integration of the electron density (ED) about the atomic centres, The ‘atomic basins’ where the ED is summed for a particular atom are determined by the ‘critical points’ (CP). These are shown in red in Figure 7, and represent the effective position where change of atomic density from one atom in the bond to the other atom occurs. A previous study has drawn attention to the similarities and differences between these two (and other) methods.21 For some silane derivatives, the total integrated ED is not easily interpreted by standard electronegativity considerations; this has been attributed to how the core electrons are treated.20,21 The positions of the CP in comparison to the respective covalent radii (CR) is appropriate. Although several series of CR are used, some widely used values22 for H, C,N,O, F and Si atoms are 25, 67, 56, 48, 42 and 110 pm. The CP results shown indicate that the Si atom in particular appears to be particularly small in relation to its neighbours C, N and F. Similarly, the CP of the C-atom in the N=C=O unit lies close to C and distant from the N and O-atoms, implying a low C-atom density. This is consistent with the Mulliken charges shown (Figure 7). The AIMPAC integrated ED values (electrons) for the individual groups of atoms, are: CH3 (-0.71e), SiHF(+1.55e), NCO(-0.84e), which show a much larger charge separation than the Mulliken analyses. However, the two methods are making quite different summations;

ACS Paragon Plus Environment

18

Page 18 of 24

Page 19 of 24

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

the summation over the ‘atomic basins’ includes density from the ‘tail’ of that on the neighbouring atom(s). Total density results for the main atoms are Si(+3.16), N(-1.78), C(+2.12), O(-1.17), F(-0.89), HSi(-0.72e) respectively. In contrast, the CP positions and Mulliken overall charges shown seem in qualitatively related over the two methods. The Mulliken charges expressed as local bond dipoles,24 by arrows in Figure 7, also conform to the more usual notions of electronegativity between bonded elements.

Figure 7. The critical points (CP, red) showing the integrated electron density minima along the bonds. The closer these are to a neighbour atom, the lower the electron density on that atom, while the converse is true. The atomic Mulliken charges are shown (blue), and the corresponding bond dipoles (black)

ACS Paragon Plus Environment

19

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

6. Discussion The molecule 3 shows similar structural results to 11 and 22, where in all cases, the complex MW spectrum, including the methyl group internal rotor, accompanies the quasi-linear structural unit and asymmetric Si centre. The substitution pattern about Si can only lead to a staggered or eclipsed sila-ethane structure of C1 symmetry; the rotational constants make only the staggered choice acceptable. The only significant difference between the structural unit from interpretation of the MW spectrum and the theoretical study is with the C2Si5N6C7 dihedral angle. The MW analysis for this last angle leads to 158(2)° while the MP3 value is 177.9º, more close to a trans staggered conformer. The dihedral angles H1C2Si5N6 and Si5C6N7O8 are trans. There is a nearly linear NCO structure. The wide SiNC angle clearly shows the level of interaction at the N-atom with both Si and C neighbours. The very low value observed in the MW spectrum for (χbb - χcc) implies a nearly linear environment. This is in good agreement with the theoretical data, which shows the absence of a CP, normally associated with a classical lone pair, on the N-atom. The energy surfaces of the MeSiF2NCO (2) and MeSiHFNCO(3) molecules at the MP2 level show similar SiNC bending profiles; however, the difluoro compound shows a significant bulge on the high SiNC angle (dissociative energy path), which is absent in the present case (3). This appears to preclude a second conformer lying close to the lowest energy surface of 3; we have obtained best fits to a set of polynomial functions to the MeSiHFNCO surface,

ACS Paragon Plus Environment

20

Page 20 of 24

Page 21 of 24

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

but no second root to the equations which lies under the SiNC envelope has been found. In all cases, the dominant terms in the potential are the quartic and quintet terms, and the simplest potential for 3, namely V(x) = 50.36(91) x4 7.53(44) x5 seems a satisfactory result. The current calculations indicate a very polar molecule (the MP3 calculated dipole moment is 2.83D). The Mulliken based bond dipole moments (Figure 7)23 reflect the usual concepts of electronegativity F > O > N > C, but show that although the methyl group is nearly neutral, charge is passed to the Si atom, which donates strongly to the F and N atoms, as well as the H(Si) atom. Integration of the electron density using AIMPAC, along the bond axis to obtain the position of zero electron density gradient (CP) (Figure 7), also suggests that the molecule is very polar. This difference from the Mulliken data may arise from the ‘atomic basin’ concept. For a bond A-B the density relating to the adjacent atom (B), beyond its CP is incorporated into the density of atom A. The effect is loss of density from B and increase in that for A. In contrast, the Mulliken process attributes all density to the centre (A or B) which carries the basis function. The atomic polarities of the Mulliken method (for the same wave-function) are thus lower than those from the ‘atomic basin’ approaches. Associated Content Supporting Information Rotational parameters for parent isotopic species; silicon, carbon, and nitrogen isotopologues, Kraitchman coordinates, assigned transitions, A-symmetry state transitions can be found in supplementary material at http://pubs.acs.org . Author Information

ACS Paragon Plus Environment

21

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

Corresponding Author *E-mail:[email protected] Tel.: 864-953-5943 *E-mail: m.h.palmer@ ed.ac.uk Tel.: (44) 131 650 4765 Notes The authors declare no competing financial interest. Acknowledgments GAG gratefully acknowledge the partial financial support of this study by the Camille and Henry Dreyfus Foundation by grant no. SI-14-007. MHP thanks the NSCCS super-computing service for continuing facilities. BHP acknowledges funding support from the National Science Foundation (NSF) Major Research Instrumentation program (award CHE-0960074) and NSF Division of Chemistry (award CHE-1213200).

References 1 Guirgis, G. A.; Wang, Z.; Lirjoni, J.; Palmer, M. H.; Obenchain, D. A.; Peebles, R. A.; Peebles, S. A.. The molecular structure of difluoroisocyanato silane: A combined microwave spectral and theoretical study. J. Molec.Struct, 2010, 983, 5-11. 2 Guirgis, G. A.; Overby, J. S.; Palmer, M. H., Peebles, R. A.; Peebles, S. A.;Elmuti, L. F.; Obenchain, D. A.; Pate, B. H.; Seifert, N. A.. Molecular Structure of Methyldifluoroisocyanato Silane: A Combined Microwave Spectral and Theoretical Stud J. Physical Chem. A 2012, 116, 7822-7829. 3 Neely, G. O.. Interpretation of extreme centrifugal distortion in isocyanic acid, isothiocyanic acid, and their deuterium derivative. J. Mol. Spectrosc. 1968, 27, 177-196. 4 Winnewisser, M.. Dynamics of small, floppy molecules studied by highresolution spectroscopic techniques. J. Molec. Struct. 1985, 126, 41-66. 5 Koput, J. Characteristic patterns in microwave spectra of quasi-symmetric top molecules of the WH3XYZ type. J. Molec. Spectrosc. 1986, 118, 448-458. 6 Palmer, M. H.; Nelson, A.D.. The structures of the azido-, isocyanato- and isothiocyanato- derivatives of methane and silane and their derivatives. A comparison of ab initio with experimental results. J. Molecular Struct. 2004, 689, 161-173. 7 Lin, C. C.; Swalen, J. D.. Internal rotation and microwave spectroscopy. Rev. Mod. Phys., 1959, 31, 841-892. 8 Pickett, H. M.. The fitting and prediction of vibration-rotation spectra with spin interactions. J. Mol. Spectrosc. 1991, 148, 371-377. 9 Hartwig, H.; Dreizler, H.. The microwave spectrum of trans-2,3dimethyloxirane in torsional excited states. Z Naturforsch 1996, 51, 923-932.

ACS Paragon Plus Environment

22

Page 22 of 24

Page 23 of 24

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

10 Kraitchman, J. Determination of molecular structure from microwave spectroscopic data. American J. Phys. 1953, 21, 17-24. 11 Gaussian 09, Revision D.01; Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R., Scalmani, G., Barone, V; Mennucci, B; Petersson,. G. A. et al, Gaussian, Inc., Wallingford, CT, 2009. 12 Bader, R.F.W. Atoms in Molecules: A Quantum Theory; Clarendon Press, Oxford, 1990. 13 Biegler-König, F.; Nguyen-Dang, T. T.; Tal, Y.; Bader, R. F. W. Calculation of the average properties of atoms in molecules/ J. Phys. B 1981, 14, 2739-2751. 14 Krishnan R.; Binkley J. S.; Seeger, R.; Pople, J.A., Self-consistent molecular orbital methods. XX. A basis set for correlated wave functions. J. Chem. Phys. 1980, 72, 650-654. 15 McLean, A.D.; Chandler, G. S. Contracted Gaussian basis sets for molecular calculations. I. Second row atoms, Z = 11-18. J. Chem. Phys. 1980, 72, 56395648. 16 Woon, D. E.; Dunning, T. H., Gaussian basis sets for use in correlated molecular calculations. III. The atoms aluminum through argon. J. Chem. Phys. 1993, 98, 1358-1371. 17 Kendall, R.A.; Dunning, T.H.; Harrison, R. J., Electron affinities of the firstrow atoms revisited. Systematic basis sets and wave functions, J. Chem. Phys. 1992, 96, 6796-6806. 18 Palmer, M.H., Deviations from idealized geometry, a comparison of structural data from experimental and ab initio studies. Part II. The pseudohalogen acids HNXY. J. Mol. Struct. 1991, 246, 321-328. 19 Mulliken, R. S., Electronic population analysis on LCAO-MO linear combination of atomic orbital-molecular orbital molecular wave functions. J. Chem. Phys. 1955, 23, 1833–1840 20 Palmer, M. H.,On the charge distribution in ethanes and disilanes and correlations with equilibrium bond lengths; an ab initio study. J. Mol. Struct. (Theochem), 2000, 500, 225–243 21 Fodi, B.; Palmer, M. H.; McKean, D. C., The effects of fluorine and chlorine substitution on bond lengths in ethanes and disilanes: comparisons of ab initio and experimental information. J. Molec. Struct.2000, 500, 195-223. 22 Cotton, F. A.; Wilkinson, G. 1988. Advanced Inorganic Chemistry (5th ed.). Wiley. p. 1385. 23 Palmer, M. H.; Findlay, R.H.; Gaskell, A.J., Electronic charge distribution and moments of five- and six-membered heterocycles. J.Chem.Soc. Perkin Trans 2,1974, 420-428. Table of Contents Graphic (The MP3 equilibrium structure)

ACS Paragon Plus Environment

23

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 ACS Paragon Plus Environment

24

Page 24 of 24