close

Вход

Забыли?

вход по аккаунту

?

Experimental and Theoretical IR Spectra of the 2-Norbornyl Cation.

код для вставкиСкачать
[3] G C. Sinke, J. Phys. Chem. 71 (1967) 359.
[4] K . 0. Christe, W. W. Wilson, C. J. Schack, R. D. Wilson, Inorg. Synth. 24
(1986) 39.
[S] L H. Long. Q. Rev. Chem. SOC.7(1953) 155.
[6] J. Jander. J. Knackmuss, K. U. Thiedemann 2. Nalurforsch. B 30 (1975)
464.
[7] J. Jander, Adv. Inorg. Chem. Radiochem. 19 (1976) 1.
[XI P. Pulay, Mol. Phys. 17 (1969) 197.
[9] P. Pulay in H. F. Schaefer I11 (Ed.): Modern Theoretical Chemistry, Vol. 4,
Plenum Press, New York 1977.
[lo] W Sawodny, H. Hartner, R. Minkwitz, D. Bernstein, 1 Mol. Struct. 213
(1989) 145.
[ l l ] H. 1. Becher: 2. Phys. Chem. (Miinchen) N F 8 1 (1972) 225.
[12] K. 0. Christe, Spectrochim. Acia A 36 (1980) 921.
[13] H. Siebert. Anwendung der Schwingungsspektroskopie in der anorganischen
Chrmie, Springer, Berlin 1966.
catching is the region between 1440 and 1340 cm- Here the
2-chloronorbornane spectrum shows hardly any peak while
the IR spectrum of the cation exhibits bands of strong intensity.''] Subsequent quenching of the deposited material afforded exo-2-norborneol.
zoo0
3000
PI
fl
1600
lZ00
800
1
6
2
Experimental and Theoretical IR Spectra
of the 2-Norbornyl Cation**
By Wolfram Koch,* Bowen Liu, Doughs J. DeFrees,
Dionis E. Sunko,* and Hrvoj VanEik
Koch et al.['] were recently able to conduct a systematic,
high-level quantum-chemical investigation of the vibrational
spectrum of the 2-norbornyl cation, which unequivocally
showed that the 2-norbornyl cation assumes a nonclassical
geometry in the gas phase,['] in line with earlier theoretical
s t ~ d i e s . [ Not
~ * ~included
]
in that paper was the complete set
of harmonic frequencies and IR intensities which were obtained at high levels of ab initio theory, including electron
correlation effects. At the same time, Sunko and VanEikIS1
developed a new experimental technique, which allows the
study of short-lived reactive intermediates like carbocations
and carbocation-like species in cryogenic SbF, matrices by
IR spectroscopy.[61In this communication we wish to present a comparison of the experimental IR spectrum of the
matrix-generated 2-norbornyl cation with the theoretically
predicted spectrum.
2-Chloronorbornane and SbF, were codeposited on the
CsI window according to the previously described techniquec5] and the IR spectrum recorded (Figs. 1 a, 2a). On
allowing the temperature to rise to 150 K, the IR bands of
the chloride start to decrease with the appearance of new
signals corresponding to the 2-norbornyl cation. This spectrum remains unchanged up to 200 K. The final spectrum of
the matrix-generated carbocation (Figs. 1 b, 2 b, Table 1) is
significantly different from the spectrum of the precursor
chloride. The deformation region below 1500 cm clearly
shows bands not present in the chloride. Particularly eye~
[*] Dr. W. Koch
IBM Heidelberg Scientific Center,
Institute for Supercomputing and Applied Mathematics
D-6900 Heidelberg (FRG)
Prof. Dr. D. E. Sunko, Dr. H. VanCik
Department of Chemistry, Faculty of Science
University of Zagreb, 41000 Zagreb (Yugoslavia)
["I
Dr. B. Liu
IBM Almaden Research Center
San Jose, CA 95120-6099 (USA)
Dr. D. J. DeFrees
Molecular Research Institute
Palo Alto, CA 94304 (USA)
The work in Zagreb was supported by the Research Council ofCroatia and
the National Science Foundation (USA), grants JPF 545 and 841. We
thank D. Gerson of the IBM Dallas Engineering and Scientific National
Support Center for making available computing resource and facilitating
o u r calculations. W K . thanks M . D. Miller for helpful discussions.
Angen. Chem. l n t . Ed. Engl. 29 (1990) N o 2
G
Fig. 1. IR spectra of (a) 2-chloronorbornane and (b) the cation generated from
2-chloronorhornane.
The structure of the nonclassical 2-norbornyl cation was
optimized in C, symmetry by using analytical gradient techniques. A standard 6-31G* basis set was used.[*]The effects
of dynamical electron correlation were included through second-order Moller-Plesset perturbation theory (MP2).'*] The
vibrational frequencies at this MP2/6-3 1G* level were then
obtained by numerical differentiation of analytical gradients, which consisted of more than 60 MP2/6-31G* analytical gradient computations, possibly one of the largest calculations ever done for a molecule of this size. All calculations
were performed using the vectorized IBM version of the
Gaussian 86 programt9]installed on an IBM 3090 with Vector Facility computer at the IBM Science and Engineering
National Support Center in Dallas, Texas. The results of the
calculations are presented in Table 1.
It is well documented in the recent literature that harmonic
frequencies computed at the MP2/6-31G* level of approximation in general tend to overestimate the experimental,
anharmonic frequencies by 4-7%.110-131This is due to the
combined effect of the neglect of anharmonicity and the
insufficiency of the theoretical method. Therefore, a uniform
scaling factor was applied to the theoretical frequencies to
achieve better agreement with the experimental spectrum
(Table 1). However, it should be noted that this uniform
scaling of all frequencies with only a single scaling factor
may even worsen the agreement with the experimental frequencies for some modes. Known examples are multiple
bond stretching and out-of-plane deformation modes for
which the computed, unscaled frequencies are sometimes
even smaller than the experimental frequencies.[", "I The
same pertains for high-frequency in-plane stretching modes,
which typically have a larger anharmonicity."'] The recent
review by Hess et al.[I4]describes many examples where the
comparison of theoretically determined harmonic frequencies with the experimental IR spectra has been successfully
applied to aid structure identification and to decide between
structural possibilities." The assignment of the experimental to the computed frequencies in Table 1 is based on the
best numerical match of the two sets of frequencies and
intensities. Figure 2 shows a graphical representation on the
Q VCH VerlagsgesellschafimbH, 0-6940 Weinheim. 1990
0570-0833/90j0202-0183$02.50/0
183
Table 1. Computed unscaled and scaled (by 0.96) and experimentally observed
frequencies and IR intensities for the 2-norbornyl cation.
No.
48
47
46
45
44
43
42
41
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Sym- Unscaled Scaled
metry frequency frequency
[cm-'1
[cm-'1
Calculated
intensity
[km mot-']
Exp.
frequency
[cm-'][a]
Exp. IR
intensity
[b]
A
7.0
6.2
0.0
0.0
0.9
1.2
18.0
0.6
20.9
0.1
3.7
6.9
0.2
18.3
19.1
27.1
10.1
10.9
11.9
0.1
3.5
4.2
0.4
1.o
1.8
0.5
6.5
0.0
0.2
3.5
1.6
2.8
0.3
0.1
0.3
2.8
4.1
5.0
18.0
0.7
0.0
16.0
0.5
8.4
3.1
0.0
24.4
9.1
3110
S
2990
vs
2980
vs
2875
m
1490
1445
1435
1380
1348
1305
S
A"
A'
A
A
A
A"
A
A
A
A
A
A
A'
A"
A'
A"
A
A'
A
A
A
A
A
A
A
A"
A'
A
A'
A'
A"
A
A"
A
A"
A
A
A
A
A
A"
A'
A'
A
A
A
A
3276
3262
3214
3214
3198
3198
3174
3150
3134
3116
3112
1569
1550
1528
1522
1488
1412
1368
1350
1334
1314
1305
1250
1226
1207
1204
1168
1141
1104
1087
1064
1017
1011
989
972
962
930
924
912
822
790
683
650
484
452
415
291
255
3145
3131
3086
3085
3070
3070
3047
3024
3009
2991
2988
1506
1488
1467
1461
1428
1355
1314
1296
1281
1262
1252
1200
1178
1159
1156
1122
1095
1060
1043
1022
976
970
950
933
923
893
887
875
789
757
656
624
464
434
398
279
244
1.
100
I
S
vs
m
S
vs
1288
1280
S
1160
W
1120
m
1056
1040
970
m
m
928
900
890
878
m
S
1200
W
m
vs
1
1200
'.
VCH L4rlagsgesellschaff mbH, 0-6940 Weinhelm, 1990
100
Fig. 2. Bar representation of (a) the IR spectrum of 2-chloronorbornane,
(b) the experimentally observed IR spectrum of the 2-norbornyl cation, and
(c) the calculated IR spectrum of the 2-norbornyl cation, computed at the
MP2/6-31G* level for the nonclassical, bridged structure (scaled by a factor of
0.96).
same scale of the experimental spectrum and the scaled, computed spectrum to allow an easy comparison of the two
spectra. The overall agreement of the two spectra (Fig. 2 b, c)
is reasonably good if one bears in mind that experimentally
we are not dealing with an isolated cation in the gas phase
but with ion pairs which obviously may have a slightly different spectrum.
In the CH stretching region we experimentally observe
four bands at 31 lO(s), 2990(vs), 2980(vs), and 2875(m)
cm- The corresponding computed bands, after scaling, are
at 3145 and 3131(s), 3047(vs), 3009(vs), and 2988(m) cm-',
respectively. Before scaling, the computed frequencies in this
region differ from the experimental ones by 5.2-8.2 YO(average 6.2%), which is very similar to the 6.5 YOaverage difference reported for CH and NH stretches in heterocyclic
arenes by Simundirus et al. using a comparable level of approximation.[121However, more significant than the CH
G
BOO
W
[a] The assignment of the experimental frequencies to the computed ones is
based purely on the best numerical match. [b] w, weak; m, medium; s, strong;
vs, very strong.
184
I
stretching region, which can be expected to show similar
frequencies for most carbocations, is the lower-frequency
region of the spectrum. Here, below 1600 cm-', the agreement is even more satisfactory, in particular when bands
with strong and very strong intensity are considered. The
very strong bands in the experimental spectrum occur at
1435, 1305, and 878 cm-', while their computed counterparts (after scaling) are at 1428, 1296, and 875 cm-', respectively.
The agreement of the computed and experimental spectra
supports all previously reported evidence that antimony pentahalide complexes of the 2-norbornyl cation have a bridged,
nonclassical structure. This was inferred by Yunnoni et a1.['61
from the 3C cross-polarization magic-angle-spinning (CPMAS) spectra of a fluoroantimonate of the 2-norbonyl
cation as well as by Luube["] and by Montgomery et al.['*]
from the X-ray structure of a fluoroantimonate of the
1,2,4,7-tetramethylnorborn-2-ylcation and of 2-methoxy1,7,7-trimethylbicyclo[2.2.l]hept-2-yliumfluoroborate, respectively. Similar conclusions were reached by Sounders
OS70-0833/90/0202-0lX4$02.50/0
Angew. Chem. In1 Ed. Engl. 29 (1990) No. 2
and Kates119]based on the deuterium isotopic perturbation
effect. An additional factor that supports this conclusion can
be found by analyzing further the CH stretching region. In
the hypothetical classical 2-norbornyl cation, the hyperconjugative interaction of the C ( 3 j H bonds with the formally
empty p orbital at C(2) should exert a significant stabilizing
effect. This would lead to a weakening of the C(3)-H bonds,
which should be reflected in a sizeable decrease in the frequencies of the corresponding CH stretching modes. Such a
shift was recently demonstrated for the I-methylcyclopentyl
cation,[51where strong hyperconjugative interaction of the
C(2)-H bonds with the cationic center at C(l) caused a decrease in the frequency of the C(2)H stretch of about
220 cm- * to 2775 cm-’, in perfect agreement with accurate
calculations for the parent cyclopentyl cation.[20]In a delocalized, bridged structure ascribed to the 2-norbornyl cation,
such an interaction should be small and the only bonds that
could contribute to a hyperconjugative stabilization are
C(3)-H and C(7j H . The normal mode analysis of the computed spectrum shows that the modes corresponding to the
two lowest CH stretching modes, computed at 2991 and
2987 cm-’, indeed involve hydrogens at C(3) and C(7). The
absence of a significant hyperconjugative decrease in the
stretching frequencies is additional, albeit negative, evidence
for the bridged structure of the cation. The same conclusion
was reached some twenty years ago[’’] on the basis of secondary kinetic isotope effect measurements in solvolysis of
2-norbornyl brosylates.[221
Received: September 20, 1989 [Z 3556 IE]
German version: Angew. Chem. 102 (1990) 198
CAS Registry number:
2-Norbornyl cation, 24321-81-1.
[l] W. Koch, D. J. De Frees, B. Liu, J. Am. Chem. Soc. 111 (1989) 1527.
[2] Recent reviews on the 2-norbornyl cation: a) C. A. Grob, Acc. Chem. Res.
16 (1983) 426; b) H. C. Brown, ibid. 16 (1983) 432; c) G. A. Olah, G. K. S .
Prakash, M. Saunders, ibid. 16 (1983) 440; d) C. Walling, ibid. 16 (1983)
448; e) H. C. Brown (with comments by P. v. R. Schleyer): The Non-Classrcul Ion Problem Plenum, New York 1077; 0 P. A. Vogel: Carbocalion
Chemistry Elsevier, New York 1985, Chapter 7.8.6 (p. 281).
[3] K. Raghavachari, R. C. Haddon, P. von R. Schleyer, H. F. Schaefer, 111,
J. Am. Chem. So<. 105 (1983) 5915.
[4] M. Yoshimine, A. D. McLean, B. Liu, D. J. DeFrees, J. S. Binkley, J. Am.
Chem. Soc. 10s (1983) 6185.
[5] H. VanEik, D. E. Sunko, J. Am. Chem. SOC.111 (1989) 3742.
161 M . Saunders et al. recently reported a similar matrix isolation technique.
They used IR spectroscopy to monitor the formation of acyl cations from
the corresponding acyl chlorides: R. M. Jarret, Ny Sin, T. Ramsey, M.
Saunders, J. Phys. Org. Chem. 2 (1989) 51.
[7] Further evidence that the 2-norbornyl cation has indeed been generated
stems from the fact that the cation generated from 7-chloronorbornane is
the same species (identical spectra) as the one formed from the 2-chloronorbornane. It is known that solutions of 7-chloronorbornane in SbFJ
SO, mixtures rearrange rapidly, yielding, after quenching, 94 % of 2-exonorborneol: see P. v. R. Schleyer, W. E. Watts, R. C. Fort, Jr, M. B.
Comisarow. G. A. Olah, J. Am. Chem. Soc. 86(1964) 5679. Details on this
work will be published in a forthcoming paper dealing with various Wagner-Meerwein rearrangements and hydride shifts in a cryogenic matrix.
[8] For a description of the basis set and Moller-Plesset perturbation theory,
see W. Hehre, L. Radom, P. von R. Schleyer, J. A. Pople: Ab inifio Molecular Orbirul Theory, Wiley Interscience, New York 1986, chapter 4.
[9] M. J. Frisch, J. S . Binkley, H. B. Schlegel, K. Raghavachari, C. F. Melius,
R. L. Martin, J. J. P. Stewart, F. W. Bobrowicz, C. M. Rohlfing, L. R.
Kahn. D. J. DeFrees, R. Seeger, R. A. Whiteside, D. J. Fox, E. M. Fluder,
J. A. Pople, Gaussian 86, Carnegie-Mellon Quantum Chemistry Publishing Unit, Pittsburgh, PA, USA, 1984.
[lo] R. F. Hout, B. A. Levy, W. J. Hehre, J. Compur. Chem. 3 (1982) 234.
[ l l ] D. J. DeFrees, A. D. McLean, J. Chem. Phys. 82 (1985) 333.
[12] E. D. Simandirds, N. C. Handy, R. A. Amos, J. Phys. Chem. 92 (1988)
1739.
[13] See also the detailed discussion in chapter 6.3 (p. 226) of [8].
[14] B. A. Hess, L. J. Schaad, P. Carsky, R. Zahradnik, Chem. Rev. 86 (1986)
709.
Angen Chem. Inl. Ed. Engl. 29 (19901 N o . 2
G
[15] For a recent application, see G. Maier, H. P. Reisenauer. W. Schwab, P.
Carsky, B. A. Hess, L. J. Schaad, J. Am. Chem. Soc. 109 (1987) 5183.
[16] C. S . Yannoni, V. Macho, P. C. Myhre, J. Am. Chem. Soc. 104 (1982) 907,
7380.
1171 T. Laube, Angew. Chem. 99 (1987) 580; Angew. Chem. Inr. Ed. Engl. 26
(1987) 560.
[18] L. K. Montgomery, M. P. Grendze, J. C. Huffman, J. Am. Chem. Soc. 109
(1987) 4749.
[19] M. Saunders, M. R. Kates, J. Am. Chem. SOC.105 (1983) 3571.
[20] P. von R. Schleyer, J. W. de M. Carneiro, W. Koch, K. Raghavachari, J.
Am. Chem. Soc. 111 (1989) 5475.
[21] J. M. Jerkunica, S . Borcic, D. E. Sunko, Chem. Commun. 1967. 1302.
[22] For a summary of subsequent uses of isotope effects in this system, see
p. 203 in [2e].
Peptides as Conformational Switch:
Medium-Induced Conformational Transitions
of Designed Peptides **
By Manfred Mutter * and Re& Hersperger
The design of peptides adopting predetermined secondary
and tertiary conformations is of utmost interest in structurefunction studies as well as in protein de novo design.[‘*21
Empirical predictive schemes[31and experimental studies on
synthetic model peptides I4] have allowed the delineation of
design principles for the construction of peptides which exhibi t tailor-made conformational proper tie^.'^] More recently, the amphiphilicity of secondary structures has been used
as a powerful design principle for the stabilization of helical
and f5-sheet conformations in solution and for the design of
self-associating peptides in the de novo design of tertiary
structures.f2.6,1’
We have now used these design principles to construct
peptides which exhibit solvent-induced conformational transitions. Most notably, this type of peptide may serve as a
“conformational switch”, the predetermined change in the
configurational properties being intimately related to
changes in the physicochemical and biological properties. To
this end, three different primary sequences of the same overall composition (4 A, 4 L, 3 E, 3 K) were designed (Fig. 1).
14 - mer : 4 A , 4 L , 3E,3K
t
I
a1
:
II :
I II :
2
3
4
5
I
7
a
9 I D 1 1 1 2 1 3
143s
AC - E L A L K A K A E L E L K A
G - NH2
K L G - NH2
AC - E L L A K K A L E A E A L K G NH2
AC - E A L
EKA LKEA LA
-
Fig. 1. Design of three tetradecapeptides (with an additional C-terminal glycine). a) Primary sequences of peptides 1-111 (A = alanine, L = leucine,
E = glutamic acid, K = lysine). In the “helical wheel” [9] and P-sheet representation, the hydrophilic residues (E, K) are indicated by the sequence number
and a black square.
[*I Prof. Dr. M. Mutter, Dr. R. Hersperger
Sktion de chimie de I’Universite Lausanne
Rue de la Barre 2, CH-1005 Lausanne (Switzerland)
[**I This work was supported by the Swiss National Science Foundation.
VCH Verlugsgesellschafr mbH, 0-6940 Weinheim, 1990
0570-0833/90/0202-0i85 $02.SOjO
185
Документ
Категория
Без категории
Просмотров
0
Размер файла
348 Кб
Теги
experimentov, theoretical, norbornyl, cation, spectral
1/--страниц
Пожаловаться на содержимое документа