The molecular structure of trichloroethenylgermane, CH2=CH-GeCl3, has been determined by electron diffraction and supported by quantum chemical calculations on CH2=CH-MX3 (M = C, Si, Ge, Sn and X = H, Cl). An equilibrium syn conformation with C, symmetry is obtained both experimentally and theoretically where one of the Ge-Cl bonds eclipses the C=C bond. The barrier of internal rotation about the C Ge bond is determined to be V-3 = 5.3(7) kJ mol(-1) using a dynamic model to simulate the internal motion. The most important structure parameters (estimated r(e)/angstrom and angle/degree) are: r(C-Ge) = 1.911(5). r(C=C) = 1.345(5). r(Ge-C17) = 2.122(2), <C=C-Ge = 120.4(5). <C-Ge-C17 = 111.0(2), and <C17-Ge-C18) = 109.4(5) where the C17 atom is in the C=C-Ge plane and the C18 atom is out of the C=C-Ge plane. Uncertainties are estimated total standard deviations (sigma(tot)) given as: sigma(tot) = [sigma(2)(scale) + (2 sigma(1sq))(2)](1/2) for bond lengths where sigma(scale) = 0.001r and sigma(1sq) is the least-squares standard deviation using a diagonal weight matrix and sigma(tot) = 2 sigma(1sq) for the other parameters. The assignment of some of the fundamental frequencies has been discussed. The internal agreement between the calculated molecular geometry obtained from B3LYP and MP2(F) using the cc-pVQZ basis set and to the experimental geometry is not good, whereas the molecular geometry obtained from a CCSD/cc-pVTZ calculation is in very good agreement with experimental geometry. This investigation shows that the gas electron-diffraction method is capable of determining the rotational barrier for small suitable molecules with acceptable accuracy.

Quantum chemical calculations using levels up to MP2(Full)/aug-cc-pVTZ have been applied. B3LYP calculations using the 6-31G* basis set reveal that there are four conformations of bis(chloroimino)butanedinitrile. The planar anti-ZZ conformer with C2h symmetry is the most stable conformer. The non-planar EE conformer with C2 symmetry, the non-planar EZ conformer with C1 symmetry and the non-planar ZZ conformer with C2 symmetry are 16.8, 22.7, 27.2kJ/mol, respectively, less stable than the planar anti-ZZ conformer according toB3LYP/6-31G* calculations. Calculated frequencies for the planar anti-ZZ conformer have been compared with observed frequencies, and some reassignments have been proposed.

B3LYP and MP2(F) quantum chemical calculations have been performed for 2-propen-1-ol using the cc-pVTZ basis set. Five stable conformations are detected and related to the two conformations observed experimentally. The calculated normal frequencies are scaled for two conformers and their assignments are compared. Similar quantum chemical calculations predict three stable conformers for cis-3-chloro-2-propen-1-ol. Quantum chemical force field calculations reveal large amplitude motions about the CC bond for all conformers. Least-squares refinements of the gas electron diffraction data using a mixture of two static conformers verify the presence of these two conformers. The agreement with the experimental data for this type of refinements is not satisfactory which indicates that large amplitude models should be used. Describing the torsions about the CC bond as frame work vibrations in harmonic large amplitude angular motion of Gaussian distributions for two conformations improved the agreement with the experimental data considerably. The root-mean-square angular amplitude of the torsions for the two conformations refined to 28.6(12)° and 21.4(46)°, with dihedral angles CCCO equal to −125.5(17)° and 180(fixed)°, respectively. Some ra distances (Å) and rα angles (degrees) obtained are: r(CC)=1.5044(20), r(CC)=1.3435(16), r(CO)=1.4299(14) and r(CCl)=1.7366(19), <(CCC)=125.60(12), <(CCO)=111.0(2), <(CCCl)=123.61(10). Parenthesized values are [σ2lsq +(0.001×r)2]1/2 for bond distances, where σlsq is the least-squares standard deviation. For the angles σlsq is given. There is good agreement between the experimental results and the quantum chemical calculations.

The C-C bond distance in molecular benzene is used by several electron diffraction groups to calibrate the electron wavelength in a gas electron diffraction experiment. It is therefore important to compare the applied r(a) value against the currently best available ab initio r(e) result. A high level CCSD(T = Full)/cc-pVTZ calculation, which has proven to give r(e) distances close to the experimental values, gave r(e)(C-C) = 1.392 angstrom. When our wavelength calibration distance r(a)(C-C) = 1.3975 angstrom is corrected for atomic displacements in the curvilinear approach and for anharmonic vibrations, it matches exactly the calculated r(e)(C-C) value. From a B3LYP/cc-pVTZ molecular force field the distance correction terms (d(hn)) and the root-mean-square vibration amplitudes were computed both in the linear and the curvilinear approximations.Gas electron diffraction intensities for benzene were registered on Fuji imaging plates. The data are very reproducible and revealed that systematic discrepancies might be present. A method to investigate systematic errors in electron diffraction is proposed. A multiplicative correction likely due to very small errors in the applied sector correction could be estimated. Applying this new correction the agreement is improved due to the high reproducibility of the imaging plates, however, hardly all systematic errors are removed. The agreement is very good and the errors left are unlikely to be caused by the Fuji imaging plate system. (c) 2006 Elsevier B.V. All rights reserved.

The molecular geometry of meta- and para-fluoronitrobenzene has been determined by gas electron diffraction and quantum chemical calculations using HF, MP2 and DFT, and 6-311G**, 6-311++G** and cc-pVTZ basis sets. The obtained geometrical parameters (rg, in Å, ∠α in degrees) for meta-fluoronitrobenzene are: r(C–C)av=1.397(4), r(N–O)av=1.227(3), r(C–N)=1.484(3), r(C–F)=1.333(8), r(C–H)av=1.107(16), ∠C1C2N=118.3(6), ∠C5C6F=120.6(24), ∠ONO=125.3(37), ∠C1C2C3=122.9(24), ∠C3C4C5=119.2(18) and ϕ(C–N)=0 (fixed) and for para-fluoronitrobenzene: r(C–C)av=1.393(2), r(N–O)av=1.232(3), r(C–N)=1.479 (dependent), r(C–F)=1.338(12), r(C–H)av=1.124(27), ∠C6C4F=118.7 (fixed), ∠ONO=124.2(23), ∠C4C5C6=123.8(17), ∠C3C2N=119.1 (fixed) and ϕ(C–N)=0 (fixed).Estimated error limits are three standard deviations from least-squares refinement using a diagonal weight matrix. B3LYP/6-311G** has been used to calculate the molecular structure for all sixteen different molecules XC6H5 (X=F, Cl, Br, I) and ortho-, meta- and para-XC6H4NO2 (X=F, Cl, Br, I). These general trends are predicted by the calculations: the C–X bond in XC6H5 is longer than the C–X bond in ortho-XC6H4NO2, the C–X bond in ortho-XC6H4NO2, is shorter than the C–X bond in meta-XC6H4NO2, the C–X bond in meta-XC6H4NO2 is slightly longer than the C–X bond in para-XC6H4NO2, the variation in the C–X bond length is approximately the same for X=F and Cl, smaller for X=Br and smallest for X=I. These trends are essentially also found experimentally. The C–X bond lengths are all calculated too long compared to experimental values.

The molecular structure of ortho-fluoronitrobenzene (o-FNB) has been investigated by gas-phase electron diffraction and ab initio MO calculations. The geometrical parameters and force fields of o-FNB were calculated by ab initio and DFT methods. The obtained force fields were used to calculate vibrational amplitudes required as input parameters in an electron diffraction analysis. Within the experimental error limits, the geometrical parameters obtained from the gas-phase electron diffraction analysis are mostly in agreement with the results obtained from the ab initio calculations. The main results are: the molecular geometry of o-FNB is nonplanar with a dihedral angle about C–N of 38(3)°. The r g (C–F) bond is shortened to 1.307(13) Å in comparison with r g (C–F) = 1.356(4) Å in C6H5F.

2-Chloro-2,2-difluoroacetamide has been studied by electron diffraction (ED) and ab initio Hartree–Fock (HF) calculations with 6-31G* and 6-311++G** basis sets and DFT calculation with a 6-311++G** basis set. The ab initio calculations predict one conformation with the C–Cl bond approximately orthogonal to the CCON skeleton and a slightly nonplanar NH2 group. Two different methods for calculation of vibrational corrections have been used, and their influence on the determination of the structure parameters has been discussed. A molecular force field has been determined, and the calculated fundamental frequencies have been tentatively assigned. The structural parameters were refined using ab initio results as constraints in the analysis. The structural parameters obtained using the curvilinear approach are: rg(N–H4)=1.048(4), rg(CO)=1.214(1), rg(C–N)=1.366(4), rg(C–C)=1.568(2), rg(C–Cl7)=1.771(2), rg(C–F8)=1.350(2), ∠OCN=125.8(2), ∠CCN=114.5(3), ∠CCF8=110.7(1), ∠Cl7CF8=109.0(1), ∠CNH4=114.1(9). Bond distances are in Å and bond angles in degrees. Uncertainties are one standard deviation from least-squares refinement using a diagonal weight matrix and inclusion of the uncertainty in the electron wave length. The barrier heights when the C–Cl is syn (α=0) and anti (α=180) to the C–N bond are of the order of 8.9(9) and 5.0(5) kJ/mol respectively, with a minimum at 96(6)°. The theoretical calculations are in good agreement with the experimental results.

The molecular structure of dimethyl-N-nitramine was reinvestigated with gas-phase electron diffraction (ED) and ab initio calculations. Ab initio calculations using different basis sets and HF, MP2 and DFT all predict a molecule with C(s) symmetry and a pyramidal amine N bond configuration. The vibrational spectra were interpreted from the scaling of the harmonic force field, and vibrational amplitudes required for the ED analysis were calculated from this scaled force field. The following values (r(g) bond lengths in Angstrom and angle(alpha) angles in degrees with errors equal to three standard deviations) were found for the main parameters: r(N-O) = 1.232(3), r(N-N) = 1.387(3), r(N-C) = 1.466(3), r(C-H)(ave) = 1.114(9), angle CCN = 116.1(6), angle CNC = 122.4(27), angle ONO = 127.6(12), angle NCH(ave) = 109.9(18). The sum of the bond angles around the amine N atom is 354.6(28)degrees. The geometrical parameters obtained from the ED analysis are in agreement with the ab initio calculations except that a more pyramidal amine N bond configuration is predicted by ab initio. (C) 1999 Elsevier Science Limited. All rights reserved.

An improved version of the previously proposed modified autocorrelation power spectrum is described. The enhanced resolution of this spectrum in the frequency domain is demonstrated for lead tetrachloride and benzene data. A background which may simultaneously correct gas electron diffraction data for both multiplicative and additive long periodic (low frequency) errors is suggested. Application of this background revealed systematic non-constant multiplicative errors in our data. The multiplicative corrections previously done to the calculated intensities of lead tetrachloride by modifications of the Pb scattering factor could now be substituted by a multiplicative background correction to the experimental intensities. For weakly exposed benzene data, the applied ‘blackness correction’ seemed to introduce a linear multiplicative correction of negative slope. The need for this correction disappeared for benzene data when the exposure of the photographic plates was increased.

The molecular structure of gaseous 4-chlorobenzaldehyde has been determined by a joint analysis of gas electron diffraction data, rotational constants from microwave spectroscopy, and constrained by results from ab initio calculations. The ab initio calculations have been performed at the HF/6-311G** level of theory. The planar C-s symmetry structure was found to be the only stable conformation. The torsion of the formyl group has been treated as a large amplitude motion. The most important structure parameters (r(g),) from the joint analysis with estimated total errors (in parentheses) are:(C-C)(mean) = 1.398(1)Angstrom, C-Cl = 1.734(3) Angstrom, C-C(= O)= 1.482(10) Angstrom, C = O = 1.216(5) Angstrom, (CCClC = 121.0(5)degrees, and (CCCHOC = 120.2(8)degrees. A scaled molecular force field has been determined. The ground state rotational constants have been determined from microwave data. (C) 1998 Elsevier Science B.V.

The electron wavelength of the short camera distance electron diffraction diagrams of benzene was calibrated against the CC distance of the molecule. This wavelength, when applied to the long camera distance data, repeatedly gives a CC distance about 0.3% longer than the applied calibration distance. The effect is demonstrated by the analysis of benzene data from eight plates recorded at each of the two applied camera distances. Our presently applied photometer procedures and numerical data reduction, which includes digital Fourier filtering of the photometer data, are described. A new variant of the autocorrelation power spectrum is illustrated for the benzene data.

The gas-phase electron diffraction data for [Me2Cl2Ga2(μ-Cl)2] are consistent with a trans model of C2h symmetry and bond distances (ra) Ga-C = 194.9(7) pm, Ga-Clt = 212.9(3) pm and Ga-Clb = 233.9(3) pm.

The gas-phase electron diffraction data (GED) for 1-aza-closo-dodecaborane, 1-NB11H12, can be fit by four models with C5v symmetry almost equally well. The MP2/6-31G* single point calculations, as well as IGLO (individual gauge for localized orbitals) computed chemical shifts were able to decide among these possibilities. The final experimental geometry was selected on the basis of the agreement between the IGLO B-11 chemical shifts calculated for the various GED models and the experimental values. This optimal choice was supported by the energetic criteria. The shorter N-B separation [r(g) = 1.716(9) angstrom] and the expansion of the pentagonal belt adjacent to nitrogen [r(g) = 1.825(6) angstrom] are the main distortions, compared with the icosahedral ''parent'' B12H122- [r = 1.77 angstrom]. The mean length of the B-B bonds in the hemisphere opposite to nitrogen is 1.791(5) angstrom. The IGLO delta B-11's computed for the best model agree with the experimental values. In particular, the chemical shift of the antipodal B-12 is reproduced reasonably well (IGLO delta = 5.5 ppm; experimental delta = 2.8 ppm).

Dirhenium heptoxide, Re2O7 (1), presents new structural features in both the crystalline state and the gas phase. When the solvated complex O3Re-O-ReO3(THF)2 (2) is crystallized from tetrahydrofuran (THF). it has a verv unsymmetrical bent oxygen-bridge structure (149.8(3)-degrees; 174 vs 209 pm). A reinvestigation of the gas-phase structure of neat Re2O7 at 230-degrees-C shows a bent molecule O3Re-O-ReO3. This bending causes the bridging Re-O distances to be longer (1.89 angstrom) than those observed for the same compound in a previous gas-phase electron diffraction study at a higher temperature (500-degrees-C). These differences are accounted for by the flatness of the bending potential of the Re-O-Re unit within this temperature range. Thermogravimetry of Re2O7(THF)2 (2) and Re2O7(dme) (3) show clean decomposition at 108 and 137-degrees-C, resp. In the latter, crystalline ReO3 Of high phase purity is formed. The solvate complex of Re2O7(dioxane)2 (4) does not decompose within a narrow temperature range (65-150-degrees-C).

A recent electron diffraction investigation found a trigonal prismatic coordination for gaseous hexamethyltungsten. Removing one ligand from a D3h symmetric prism, distorted prismatic geometries of C(s) symmetry seem possible for a pentasubstituted compound. Only five of the seven geometrical parameters of a C(s) configuration for molybdenum pentachloride could be determined. Two different distortions of the prismatic configuration both fitted the data equally well and as well as possible. Common to these configurations was the same rectangular and close-to-square arrangement of four chlorine atoms. The results may be described either as distorted prismatic configurations or as distorted C4v geometries; the distortion of the prism is actually toward the C4v square pyramid.

Benzene data recorded on Fuji imaging plates had previously revealed systematic errors most likely due to errors in the sector correction. An improved sector correction obtained from the data then gave better u-values and lower R-factors for the benzene structure. Accordingly a new direct measurement of the sector seemed to be of interest The new measurements of the sector agrees favorably with the improved sector correction previously estimated from the benzene scattering pattern. Optimal start values for s for the new sector correction were 3.25 and 6.5 Å -1 for the long and the medium camera distance data respectively, these s-values corresponds to a distance of about 15 mm from the center of the sector.