вход по аккаунту


On the BO Bond Length in Oxadiboriranes.

код для вставкиСкачать
[13] K. H. Dotz in Transition Metal Ccrrbene Complexes (Eds.: K. H. Dotz, H.
Fischer, P. Hofmann, F. R. Kreissl, U. Schubert, K. Weiss), Verlag
Chemie, Weinheim 1983,pp. 191-226.
[14] J. Wolf, R.Zolk, U . Schubert, H. Werner, J Organomet. Chem. 1988,340,
161 -178.
[15] A. Hohn, H. Werner, Chem. Ber. 1988, 121, 881-886.
[I61 H. Werner, U. Brekau, M. Dziallas, J. Organomel. Chem. 1991,406.237260.
[17] K. Weiss in Carbyne Complexes (Eds.: H. Fischer, P. Hofmann. F. R.
Kreissl, R. R.Schrock, U. Schubert, K. Weiss), VCH Verlagsgesellschaft,
Weinheim 1988,pp. 205-228.
[18] C. M. Stegmair, W. Ullrich. W. Schutt, P. Kiprof, F. R.Kreissl, Chem. Ber.
1992, 125, 1571-1573.
1191 E. 0. Fischer, T. L. Lindner, G. Huttner, P. Friedrich, F. R. Kreissl, J. 0.
Besenhard, Chem. Ber. 1977, 110, 3397-3404.
[20] H. Fischer, P. Hofmann, F. R. Kreissl, R. R.Schrock, U. Schubert, K.
Weiss, Carbyne Complexes, VCH Verlagsgesellschaft. Weinheim, 1988.
[21] C. M. Stegmair, W. Schutt, W. Ullrich, P. Kiprof, J. Ostermeier, F. R.
Kreissl, J. Organomer. Chem. 1993,447,251 -259.
[22] N . Ullrich, C. M. Stegmair, H. Keller, E. Herdtweck, F. R.Kreissl, Z.
Naiurjorsrh. B 1990,45, 921-925.
[23] W. Schutt. N. Ullrich, F. R. Kreissl, J. Organornet. Chem. 1991,408,C5CX.
[24] C. M. Stegmair, .I.
Ostermeier, W. Schutt, F. R. Kreissl, unpublished results.
1251 Crystal structure data for complex 2 a : C,,,H,,BF,N,O,S,W .0.25Et20.
0.125CH'CN. M , = 604.1 (without solvent), rhombohedra], R3, a =
6 = 3052.2(2), c =1402.1(l)pm, y =120", V =11312(2)x l o b pm', Z =
18, -80+3 "C, pcSIsd
= 1.596 (without solvent) g ~ m ' ~
, measured reflections (without R centering), of which 3683 were unique and 3467 observed with I > l.Oo(1) (w scan, 1.0" < 0 < 25.0', (0.75 + 0.25 . tanof"
+25%, + h , k, I, Mo,., 1 =71.073 pm, Enraf-Nonius CAD4).
R = 0.035, R, = 0.035, goodness of fit = 3.237, Ap,," (max./min) 0.611
-0.59 e k ' . There are large cavities, each of about 1000 A' around the
lattice positions (0,0,0), (1/3,2/3,2/3), and (2/3,1/3,1/3). In total. about
25% of the lattice is empty! Even with a complete occupation of the ether
positions, about 260 A 3 0 f unoccupied lattice positions still remain, almost
1 0 % of the cell volume. Small quantities of solvent incorporated in the
cavities are not observed by X-ray crystallography. The cations interact
with the anions through weak hydrogen bonds. Further details of the
crystal structure investigation may be obtained from the Fachinformationszentrum Karlsruhe, Gesellschaft fur wissenschaftlich-technische Information mbH, D-76344 Eggenstein-Leopoldshafen (FRG) on quoting
the depository number CSD-57414, the names of the authors, and the
journal citation.
[26] F. R. Kreissl, F. X. Miiller, D. L. Wilkinson, G. Miiller, Chem. Ber. 1989,
f22, 289-290.
[27] V. A. Semion, Y. T. Struchkov, Zh. Sfrukr.Khim. 1968, 9, 1046-1047.
[28] F. R.Kreissl, N. Ullrich, Chem. Ber. 1989, 122, 1487-1488.
+ +
On the BO Bond Length in Oxadiboriranes""
By Michael Biihl, Henry Frederick Schaefer HI,*
Paul von Rague Schleyer,* and Roland Boese*
sized and characterized by X-ray structure analysis. The accuracy was reduced due to disorder, but the refinement favored (model A) gave extremely long BO bond lengths
(1.545(5) and 1.510(6)A). These disagreed with the much
shorter BO distances (1.4O9l4I and 1.403
which were
calculated for the parent compound 1 a, both at the modest
(HF/3-21G)[41and at the higher (MP2/6-31G*)I31level. We
now report the results of a variety of quantum mechanical
calculations which reveal that a second refinement (model B), although more disordered, provides a better description of the molecular geometry than the previously published
crystal structure.
Conventional refinement of the X-ray diffraction data for
1 d (model A) resulted in an essentially linear CBBC arrangement; in accord with crystallographic centrosymmetry, the
oxygen atoms were found with 50 % probability on each side
of the B-B bond.13] The BBC angles of 178 and 182" exaggerated the trend towards linearity predicted computationalMore importantly, the
ly for the parent compound 1a.r3,41
Center for Computational Quantum Chemistry
University of Georgia
Athens, GA 30602 (USA)
Telefax: Int. code + (706)542-0406
Prof. Dr. P. von R.Schleyer
Institut fur Organische Chemie der Universitit Erlangen-Nurnberg
Henkestrasse 42, D-91054 Erlangen (FRG)
Telefax: Int. code + (9131)859-132
Priv. Doz. Dr. R. Boese
Institut fur Anorganische Chemie der Universitdt-Gesamthochschule
Universitdtsstrasse 5-7, D-45 117 Essen (FRG)
Telefax: Int. code + (201)183-2535
This work was supported by the Deutsche Forschungsgemeinschaft, the
Fonds der Chemischen Industrie, and the U S . National Science Foundation.
Verlagsgesellschaft mbH. 0-69451 Weinherm. 1993
Id, R = C(SiMe,),
3 0 bond lengths of this model (A) did not agree with the ab
initio calculated distances. An alternative, preliminary refinement with split positions, allowing for disorder of the
boron atoms as well as the oxygen atom (model B), afforded
BO separations of around 1.38 A which are in much better
agreement with the theoretical results. The R value of this
model B, however, was not significantiy better than that of
model A. Since the theoretical geometries of the parent system 1 a might be different from those of the highly substituted compound Id, it was decided to publish the resuIts for
model A, the conventional refinement.r3]
We now have investigated the substituted derivatives 1 b
and 1 c at the SCF/DZP level, as well as the parent oxadiborirane 1a at higher levels of theory including large basis
sets and more sophisticated treatments of electron correlat i ~ n . l ' *As
~ ]apparent from the results in Table 1, there is no
change in the theoretical conclusion. The computational evidence requires that 1a does not possess a BO bond longer
than about 1.39 A. The CCSD(T)/TZ2Pf value (1.388 A)
should be reliable. The geometry changes upon methyl (1 b)
Cmpd. R
[*] Prof. Dr. H. F. Schaefer 111, Dr. M. Biihl
l c , R = C(SiH,),
Table 1. Geometries of oxadiboriranes I a-1 d.
High-level ab initio calculations, an important source of
accurate structural data,"] can help decide among competing structural models obtained from experimental refinements.''] Recently the first oxadiborirane (1 d)[31was synthe-
la, R = H
1 b, R = CH,
I \
Cmpd. R
B-0 [A] B-B [A] B-H
SCF/3-21G [4]
MP2/6-31G* 131
exp. model A [3]
1.601(7) 1.544(4)
exp. model B
$ l0.00+ ,2510
B-B-H ["I
1.599(9) 1.510(7) 172.2(5)
1.607(7) 171.9(5)
Angew. Chem Inr. Ed.Engi. 1993,32,No.8
and trisilylmethyl (1 c) substitution are less than 0.01 A for
the BO separations (SCF/DZP level, lower part in Table 1).
The structural differences between 1 c and 1 d should be even
Energy considerations also argue against model A: a partially optimized structure 1 a', based on the published X-ray
structure (i.e. 1 a with a fixed three-membered BBO ring but
with freely optimized (B)-H positions), was 12.4 kcalmolless stable (CISD/DZP level) than the fully optimized geometry (1 a, Czv)."ilFinally, the averaged IGLO "B NMR
chemical shiftF8]calculated by using this hypothetical structure (la'), 6 = 81.1, does not agree with the value observed
experimentally for 1 d (6 = 65.7).[31In contrast, the computed structures for 1 a give very good agreement: 6 = 67.0 (II'//
MP2/6-31G*) and 66.3 (II'//CISD/DZP).r91
Prompted by these apparent deficiencies of model A, the
second, split model B was refined
The final
molecular structure (Fig. 1) can be viewed as a superposition
perpendicular to the BBO ring plane is expected to shorten
the observed BO distance artificially. Also, the SiMe, positions should be affected by the disorder of the BBO ring.
Some evidence for this can be detected by investigation of the
ADPs of the respective atoms, but it was not possible to
resolve them in split positions. Based on the ab initio data in
Table 1, we recommend 1.39 8, as reference value for the
equilibrium BO bond length in oxadiboriranes. While crystallographic criteria (i.e. R values) could not decide between
the two alternative structural models A and B for Id, our
theoretical results (geometries, energies, and NMR chemical
shifts) unambiguously demonstrate that model B (Fig. 1)
affords the more realistic experimental geometrical parameters.
This again demonstrates a general admonition: when their
results do not agree with high level ab initio computations,
experimentalists are encouraged to consider other possible
interpretations of their data if this leads to better accord
between theory and experiment.
Received: February 9, 1993 [Z 5857 IE]
German version: Angew. Chem. 1993. /Of, 1265
Fig. 1 . Molecular structure of 1d in the solid state (model B). The ellipsoids
correspond to 50% probability ofelectron occupation, hatched and unhatched
atoms are equivalent by crystallographic centrosymmetry, the disordered oxygen and boron atoms are drawn with dashed outlines, and the hydrogen atoms
are omitted for clarity. Selected distances [A] and angles ["I (standard deviation
in parentheses): B-Bl' 1.599(9), B-0 1.347(7), B1'-0 1.365(7), B-C 1.607(7),
B1'-C' 1.510(7). 0-B-BI' 54.4(3), B-0-B1' 72.3(4), C-B-Bl' 172.2(5), C'-BI'-B
of two independent molecules disordered about the crystallographic inversion center. The "averaging" of the B and B1
as well as BI' and B positions, leads to the apparently elongated BO separations of model A. The BO distances deduced
from model B f1.347(7) and 1.365(7) A) are now in much
better agreement with the results of the ab initio calculations
(1.378 A for l c , Table 1). In contrast to model A, these BO
bond lengths are also within the usual range observed for
trigonal-coordinated boron (1.28 - 1.43 A, mean value
1.365 A),["] and are close to the BO separations in Me,B-0BMe, (1.359(4) A[',]), an acyclic reference compound.
Likewise, the BBC bond angles from model B (172.2(5)
and 171.9(5)") are in better accord-with the ab initio values
for lc(172.8",Table 1)thanthoseofmodel A(177.7(3)"and
182.3(3)o[3,'31).Finally, the computed energy of 1 a", employing the BBO ring dimensions derived from model B, is
only 1.5 kcalmol-' higher than that of the fully optimized
structure of 1 a (CISD/DZP level).
In summary, the refinement employing model B results in
much better agreement of theory and experiment and represents the molecular geometry of 1 d more reliably than that
derived from model A.[31However, correlations between the
refined parameters (ADPs) increase the error margins of the
experimental data beyond the commonly used 3 o criterion.
In addition, the large vibrational motion of the oxygen atom
Angew. Chem. Inl. Ed. Engl. 1993. 32, No. 8
[I] See for example A. Domenico, I. Hargittai, Accurate Molecular Sfrurtures,
Oxford University Press, Oxford, 1992.
[2] Experimental and theoretical borane and carborane geometries have been
assessed by means of computed energies and chemical shifts: M. Buhl, P.
von R. Schleyer, J. Am. Chem. Soc. 1992, 114,477.
[3] P. Paetzold, L. Geret-Baumgarten, R. Boese, Angew. Chem. 1992. 104,
1071; Angew. Chem. Int. Ed. Engl. 1992,31, 1040; see also "highlight": H.
Grutzmacher. ibid. 1992, 104, 1358 and 1992,31, 1329.
[4] C. Liang, L. C. Allen, J. Am. Chem. Soc. 1991. 113, 1878.
[5] Geometries have been fully optimized in C,, symmetry for 1 a and in C,
symmetry for 1 band Ic, at the SCFjCISD [6a], and CCSD(T) levels [6 b],
employing the following contracted Huzinaga/Dunning basis sets [6c]:
DZP (9s5pld/4s2pld) for B (a, = 0.7). C (a, = 0.75), and 0 (a, = 0.85).
(11s7pld/6s4pld) for Si (ad= 0.51, (4slp/2slp) for H (a,= 0.75); TZP
(lOs6pld/Ss/3pld) for B and 0, (5slp/3slp) for H ; TZ2Pf (10s6p2dlf/
5s2dlf) for B (a, =1.4, 0.35, I( = 0.5) and 0 (a, =1.7, 0.425, af =1.4),
TZP for H.
[6] a) P. Saxe, D. J. Fox, H. F. Schaefer 111, N. C. Handy, J. Chem. Phys. 1982,
77,5584; J. E. Rice, R. D. Amos, N. C. Handy, T. J. Lee, H. F, Schaefer 111.
ibid 1986, 85, 963; b) A. C. Scheiner, G. E. Scuseria, J. E. Rice, H. F.
Schaefer 111, ibid. 1987, 87, 5461; G. E. Scuseria, C. L. Janssen, H. F.
Schaefer 111, ibid. 1988,89,7382; c) S. Huzinaga, ibid. 1965.42.1293; T. H.
Dunning, Jr., ibid. 1970, 53, 2823; ibid. 1971, 55, 716.
[7] l c was optimized by employing the Turbomole program: a) M. Haser, R.
Ahlrichs, J. Comput. Chem. 1989, 10, 104. b) R. Ahlrichs, M. Bar, M.
Haser, H. Horn, C. Kolmel, Chem. Phys. Leri. 1989, 162, 165.
[8] Chemical shifts have been evaluated by using the IGLO (Individual Gauge
for Localized Orbitals) method (W. Kutzelnigg, Isr. J. Chem. 1980, 19. 193;
M. Schindler, W. Kutze1nigg.J. Chem. Phys. 1982,76,1919),employing the
following contracted Huzinaga basisset (S. Huzinaga, Approximate Atomic Wave Functions, University of Alberta, Edmonton, 1971): 11' (9s5pld,
5s4pld) for B (a, = 0.5) and 0 (a,,= l.O), (3s.2~)for H. See also reference
[7] and W. Kutzelnigg, M. Schindler, U. Fleischer, in N M R , Basic Princ.
Prog. 1990, 165.
[9] Similar values were obtained with P. Pulay's GIAO program by employing
a TZP basis set; H. Sulzbach, P. von R. Schleyer, unpublished results.
[lo] Crystal structure of I d : Cell data identical to those reported in reference
[3]; refinement with disordered boron and carbon atoms with SOF = 0.5;
2098 independent reflections, 1799 of which observed with Fo 2 4cr(F);
anisotropic temperature factors for non-hydrogen atoms; 172 parameters
refined; R = 0.0420, R , = 0.0451. (Model A: 163 parameters, R = 0.0424,
R, = 0.0520) Further details of the crystal structure investigation may be
obtained from the Fachinformationszentrum Karlsruhe, Gesellschaft fur
wissenschaftlich-technische Information mbH, D-76344 Eggenstein-Leopoldshafen (FRG) on quoting the depository number the names of the
authors CSD-400 204 and the journal citation.
I111 A. F. Wells, Slrurturai Inorganic Chemistry, 5th ed., Ciarendon Press. Oxford, 1984.
[12] G. Gundersen, H. Vahrenkamp, J. Mol. Struct. 1976, 33, 97.
[13] ELF calculations (ELF = electron localization function) (A. Savin. A. D.
Becke, J. Flad, R. Nesper, H. Preuss, H. G. von Schnering. Angew. Chem.
1991,30,421; Angew. Chem. Int. Ed. Eng., 1991,409) indicate that the BB
bond i n l a is bent outward, but the BO bonds are strongly polarized
toward oxygen and can hardly be regarded as having bent character. We
thank M. Kaupp, Stuttgart, for this analysis.
Verlagsgesellschafi mbH, 0.69451 Weinherm, 1993
0570-0833/93/0808-/155 $ 10.00+.25/0
Без категории
Размер файла
292 Кб
bond, length, oxadiborirane
Пожаловаться на содержимое документа