вход по аккаунту


Electron-Transfer Transition States Bound or UnboundЧThat is the Question!.

код для вставкиСкачать
[30] T. J. Katr. N. Acton. J A m . Chem. Soc 1972, 94, 3281-3283.
[31] D. E Hunt, J. W. Russell, J Organomer. Clwm. 1972. 46. C22-C24.
[32] K. Jonas. B. Gabor, R. Mynott. K. Angermund. 0. Heinemann, C. Kruger.
A n g w . Cliem. 1997, t09, 1790; Angcw Chem. Inr Ed. Engl. 1997. 36, 1712.
[33] K. Jonas. P. Kolb, G. Kollhach, B. Gabor. R. Mynott. K. Angermund, 0
Heinemann, C . Krdger. Anjicw Chrm. 1997, 109, 1793; Angetr. Chem. Inr. E d
€fig/. 1997. 36. 1714.
[34] According to IUPAC the anion 1 2 - should he named dihydropentalendiide: it
is obtained by double deprotonation of a dihydropentalene The anion 1 2 - can
also be obtained from pentalene by double electron transfer and may therefore
be named as pcntiilene dianion. In the original publications the authors prefer
the graphic representation of 1*- as a dianion or as a complex ligand with a
formula with a closed cycle in either one of the five-membered rings. The
pentalene ligand IS a 10e ligand in the complexes 4-10, therefore 4 should be
considered as a V”’ complex.
[35] H. Butenschon, Dissertation. Universitst Hamburg. 1983.
[36] H. Butenschon, A. de Meijere, Tetrulzmdron 1986, 42. 1721 1729.
[37] T. Lendvai, T. Friedl, H. Butenschon, T. Clark. A de Meijeie. A n g w . Clirm.
1986, 98. 734-735; Angew. Chem. Inr. Ed. Engl. 1986. 25. 719-720.
[38] R. Haag. R. Fleischer. D. Stalke, A. de Meljere. Afipeii Cheni. 1995. 107.
1642-1644. Aiigeii. Cliem. I n l . Ed. Engl. 1995. 34, 1492- 1495
[39] R. Haag. Dissertation. Universitit Gottingen. 1995
Electron-Transfer Transition States: Bound or Unbound-That
is the Question!
Hendrik Zipse*
The breathtaking advances realized in the field of computational quantum chemistry within the past two decades have
made it possible to reliably predict transition state structures for
most reaction types in organic chemistry. The focus of controversy has moved from the characterization of typical organic
reaction mechanisms, such as pericyclic reactions,[’I to more
complex issues. such as reaction mechanisms in organometallic
or the detailed description of solvent effects. In
light of this development it is quite surprising that little structural detail is available for the most elementary reaction step in
chemistry, the exchange of electrons between two reaction partners. It was only recently realized that transition states for electron-transfer reactions might also have definitive structural requirement~.’~]
Traditionally, electron-transfer (ET) reactions
were qualitatively discussed in terms of outer-sphere and innersphere ET.I4] These terms were originally coined for classifying
ET reactions between transition metal complexes, which either
occur without breaking metal-ligand bonds (outer-sphere ET)
or with concomitant cleavage/formation of metal-ligand bonds
(inner-sphere ET).Is1 Implicit in this categorization is the assumption that little interaction exists between the electron
donor and acceptor in transition states of outer-sphere ET reactions, whereas the converse is true for inner-sphere ET reactions. Exactly this assumption was challenged by Eberson and
Shaik through application of the valence-bond curve-crossing
model (VBCM) to thermal ET reactions in open-shell systems.[31
According to this qualitative model transition states for dissociative electron transfer between radical anions and alkyl halides
is stabilized by maximum overlap between the donor SOMO
(singly occupied molecular orbital) and the alkyl halide LUMO
(lowest unoccupied molecular orbital), whereas the most dominant interaction for the competing S,2 transition state is that
between the donor n HOMO (highest occupied molecular orbital) and the alkyl halide LUMO. For the combination of ketyl
radical anions and alkyl halides, the corresponding orbitals are
[*] Dr. H . Zipse
Institut fur Organiche Chemie der Technischen Universitit
Strasse der 17 Juni 135. D-10623 Berlin (Germany)
Fax. Int. code +(30)3142-1102
e-mail: zips0531w
A n p m . Cltoiz. Inr Ed Ef7gI. 1997. 36, No. 16
shown in Scheme 1. The most important conseqiiences of this
are the following: a) Optimization of bimolecular orbital overlap is important for transition states of formal outer-sphere
electron transfer. If not prohibited by steric effects, transition
states for electron transfer should be bound more strongly than
either the reactant or the product complexes. b) Different selection rules apply for transition states of electron transfer and
nucleophilic substitution.
Scheme 1. Frontier molecular orbitals of ketyl radical anions and alkyl halides.
It already becomes apparent at this stage that the terms of
inner- and outer-sphere electron transfer alone are not well suited for discussing mechanistic details in electron-transfer processes, as they d o not describe the nature of the transition states
involved. In general electron-transfer transition states might be
characterized as “bound” or “unbound”; the latter describes a
situation in which only weak interactions exist between the reactants, whereas for the former the electron donor and acceptor
are in intimate contact in the transition state. Therefore, ET
reactions might be of the unbound, outer-sphere type involving
weakly bound transition states and avoiding formation of
bound intermediates o r products, o r they might be of the bound,
outer-sphere type in which formation of bound intermediates or
products is again avoided, but the more compact transition state
WILEY-VCH Verlag GrnbH. D-69451 Weinheim, 1997
0570-0833!97/3616-1697 S 17 50f.50 I1
has a defined structure. An upper limit of 1 kcalmol- of electronic interaction energy is commonly assumed for unbound,
outer-sphere ET transition states.131The distinction between unbound and bound transition states is of less importance for
inner-sphere ET reactions, as basically all reactions leading to
bond formation will have bound transition states. This leaves us
with three categories for the classification of ET transition
states: “unbound, outer-sphere ET”, “bound, outer-sphere
ET”, or “bound, inner-sphere ET”.
canthe VBCM predictions of Eberson and Shaik be verified?
One of the smallest model systems amenable to quantum mechanical studies is the reaction of the formaldehyde radical anion with methyl chloride (Scheme 2).16]The results obtained are
-. 0
Scheme 3. Competition between intramolecular nucleophilic substitution and intramolecular dissociative electron transfer in radical anions of substituted benzoylhaloalkanes. R = H, Ar; X = CI. Br. I ; n = 1. 2, 3.
the theoretical model study,[61the relative barriers for ET and
substitution reactions will strongly depend on the chain length.
This point was explored in more detail in another theoretical
0 study of smaller model systems (Scheme 3, R = H, X = C1j.r81
+ CI
For n = 1 the barrier for electron transfer is significantly lower
2.060 !lt 2.151
than for nucleophilic substitution at oxygen, simply due to the
fact that the ring strain is much higher in the transition state
of the latter process. For n = 2 the S,2 transition state can be
+ .CH3 + CI0 formed much more easily, whereas the ET transition state is
now burdened with excess ring strain. In agreement with experiment,
the reaction now proceeds mainly by the SN2process.
Scheme 2. Pathways for nucleophilic substitution and dissociative electron transfer
The most intriguing aspect of the experimental studies by
in the reaction of formaldehyde radical anion and methyl chloride [UMPZ(FC)/
6 - 31+G*].
Kimura et al. is, however, that the rate of disappearance of the
initially formed ketyl radical anions correlates well with the
in complete agreement with the qualitative predictions of the
reduction potential of the parent ketones as well as solvent
VBCM model. The structure of the SN2transition state leading
polarity indicators, irrespective of the actual ratio between the
to the 0-alkylation product is similar to that computed for other
substitution and the electron-transfer process. This observation
nucleophiles. The formaldehyde radical anion is oriented such
can only be explained by assuming a common mechanism for
that one of the carbonyl-group lone pairs points towards the
both processes in the sense that fission of the C-C1 bond as well
methyl halide carbon atom. The transition state for electron
as rearrangement of the charges advance to a similar degree for
transfer is distinctly different from that for the SN2alternative:
both pathways.
the carbonyl carbon atom now aims towards the methyl group
One immediate question arising from the concept of bound
in CH,CI, and the distance between donor and acceptor is about
outer-sphere ET transition states is whether these transition
states can be bound as strongly as those of bond-forming reac0.5 8, larger. The barrier for the electron-transfer process
tions, or whether a limit exists beyond which transition state
amounts to + 5.3 kcalmol-l relative to a complex of the reacbinding inevitably leads to bond formation. This question was
tants at the QCISD(T)/6 - 31 +G(d)//UMP2/6 - 31 +G(d)
level of theory, and is 1.9 kcalmol-‘ less than that for the S,2
explored in reactions of ketyl radical anion donors with methyl
halide acceptors (Scheme 4).[91Variation of the donor strength
reaction. That both pathways involve significant binding between the two reactants is not only seen in the transition state
and leaving-group ability gives a series in which the reaction
barrier becomes progressively larger. A mechanistic changestructures but also in the activation entropies, which are negaover from C-oriented ET to C-oriented S,2 occurs at high bartive and of comparable magnitude in both cases. These theoretriers and short C-C distances, but all transition states can be
ical results nicely complement experimental studies by Kimura
linearly correlated with a. The parameter a is a relative measure
et al. on the competition of intramolecular substitution and
electron-transfer processes in radical anions of benzoylof transition state location that, according to the Marcus equation, can be expressed as a ratio of activation- and reaction-enhalo alkane^.[^] The product distribution obtained in these sysergy contributions.”” The clear separation between ET and SN2
tems (Scheme 3, R = Arj depends on the length of the chain
connecting the carbonyl group and the halide carrier, the halide
transition states illustrates that a “point of no return” exists at
around 2.4 8, for the bond distance between the central carbonyl
leaving group, and the solvent. For very short chains, such as
and alkyl carbon atoms. All transition states involving a C-C
n = 1, only electron transfer was observed, whereas an increase
distance that is smaller than this limiting value lead to SN2prodto n = 2 or 3 gives mainly nucleophilic substitution if chloride is
the leaving group. This can readily be explained as the conseucts, whereas transition states with longer C-C bond distances
lead to ET products. The common correlation observed for all
quence of differing structural requirements for the transition
states of the two competing processes. The connecting chain not
stationary points in this study can most reasonably be explained
only serves as a link between the two reactive centers, but also
by assuming that ET and SN2transition states belong to one
family of related reaction mechanisms.
moderates the conformational space available for the relative
The discussion of bound or unbound ET transition states is,
orientation of these two groups. If the orientational preferences
of course, not without controversy. A new theoretical study of
between the two groups are indeed as different as predicted by
0 WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1997
0570-083319713616-1698 $17.50+.50/0
Angew. Chem. Int. Ed Engl. 1997, 36, No. 16
o. ]
H3C -X
A third mechanism for electron transfer can be found in reactions of uncharged nucleophiles with radical cations of alkenes
and alkanes.“ 3 . 14] In the case of the former this involves initial
nucleophilic addition to the alkene radical cation and formation
of a distonic radical cation (Scheme 5, type A). Homolysis of
R = c H 3 , H, CN; X = CI, Br, I
ET- region
r(c-c)/A 2.5
aScheme 4 Correlation of the C-C bond distance in the transition state with the
Marcus parameter T .
the methyl chloride/formaldehyde radical anion system by
Bertran et al.“ ‘I reports an additional transition state for electron transfer that is close to that expected for an unbound,
outer-sphere ET transition state. The distance between the reactants is significantly larger in this structure, which is energetically much less favorable than the transition state for the O-alkylation pathway. A similar stationary point was found for the ET
reaction of methyl fluoride and formaldehyde radical anion.
Moreover, the bound ET transition state reported earlier16] was
also found in this study. However, this reaction path led to
substitution products, and this stationary point has correspondingly been designated a “C-S,2” transition state. Indeed, the
situation appears to be rather delicate in that the reaction pathway leads to different products, depending on the method chosen for following the reaction. This point was reinvestigated in
more detail by Shaik et al.;[”] the ET/S,2 reaction pathway is
sensitive to the coordinate system as well as the choice of wavefunction. This controversy will eventually be settled by generating the complete potential energy surface and employing more
sophisticated theoretical methods. But the results obtained so
far indicate a very flat potential energy surface, in which the
difference between going in one o r the other direction is rather
small. Certainly, the actual branching ratio between the two
pathways will not only depend on the reaction energetics, but on
the reaction dynamics as well.
A n g i , ~ Chrm In/ Ed Engl 1997, 36, N o 16
Nut + H3C-CH3
Scheme 5 . Mechanisms for inner-sphere electron transfer between uncharged nucleophiles and radical cations of alkenes (type A) and alkanes (types B.C)
the newly formed bond gives the donor radical cation and the
neutral alkene as products. This sequence is a classical example
for inner-sphere electron transfer between organic substrates. If
the radical cation of ethane, C,H;+, is used as a representative
model for hydrocarbon radical cations, electron transfer to uncharged nucleophiles proceeds either through bound, outersphere ET transition states (for PH, and SH,) or by a two-step
inner-sphere process that involves stepwise transfer of H +/H’
(Scheme 5, type B) or +CH,/’CH, (type C) redox pairs. The
latter two pathways are found for NH, and H,O as nucleophiles. Since the initial step of inner-sphere ET processes is
simply nucleophilic attack at the hydrocarbon radical cation, we
can reconfirm one major conclusion reached in the studies of
anionic systems: bound transition states for outer-sphere ET
and nucleophilic-substitution reactions are mutually exclusive
for the same system, but both belong to the same mechanistic
family in the sense that they share common structural and electronic features. Irrespective of the actual choice of ET through
bound transition states, an alternative pathway also exists
through unbound, outer-sphere ET transition states in these
systems. This option was explored for the reaction of NH, and
PH, with the ethane radical cation. In both cases the unbound
transition states for outer-sphere ET are energetically less favorable than the separated reactants, and show a much larger distance between the two reactants (>4 A). The pathways through
unbound, outer-sphere E T transition states are thus much less
attractive energetically than either of the two alternatives. In
matching theoretical with experimental results it is important to
consider also activation entropies, which are usually more favorable for pathways involving unbound, outer-sphere ET transition states.
Ab initio studies of small model systems predict that electron
transfer cannot only proceed through the traditionally anticipated pathways involving unbound, outer-sphere or bound, inner-sphere ET transition states, but that the bound, outersphere ET transition state is a serious contender in many cases.
,c WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1997
0570-0833/97/3616-1699 S 17 50+ 50 0
The latter shares some of the characteristics of transition states
for bond-forming reactions, such as clear stereochemical preferences and a certain proximity of the reactants. Given the dominance of ET pathways through bound transition states emanating from the theoretical studies, it appears desirable that the
interpretation of experimental results proceeds under consideration of all three reaction types. The combination of complementary experimental and theoretical methods seems to be the
most promising way to achieve deeper mechanistic insight into
electron-transfer processes. Promising links between theoretical
and experimental studies include relative activation energies
and entropies of related systems as well as kinetic-isotope effects
as one of the best observable descriptors of transition state
structure. Even though this highlight exclusively deals with ET
processes in open-shell systems, it should be noted that recent
experimental results for ET processes in closed-shell substrates" '] also suggest bound, outer-sphere ET transition states.
It therefore appears that the three types of ET processes described here can be used in a more general sense to categorize ET
German version. Angeii Chem 1997. fOY, 1773-I776
Keywords: a b initlo calculations
mechanisms . transition states
. reaction
[l] a) K. N. Houk. J. Gonzalez, Y Li. Acc. Chem. Res. 1995. 28. 81; b) K. N.
Houk, Y. Li, J. D. Evanseck, Angew. Chem 1992, 104,711, Angew. Chem. Int.
EN'. EngI. 1992. 31, 682.
[2] a ) U. Pidun. C. Boehme. G. Frenking, Angew. Chem. 1996. 108. 3008: Angar.
C h m . lnt. Ed. Engl. 1996. 35. 2817, b) S. Dapprich, G. Ujaque, F. Maseras,
A. Lledos. D. G. Musaev. K. Morokuma, J. Am Chem. Soc. 1996,lf8.11660.
[3] L. Eberson, S. S. Shaik. J. Am. Chem. Soc. 1990. 112. 4484.
141 a ) L Eberson, Adi. P h j x Org. Chem. 1982. 18. 79; b) Electron Transfer in
Orgunic Chmiistrj~,Springer, Heidelberg, 1987; c) J -M. Saveant, Adv P h j s .
Or8 Clieni. 1990,26, 1, d) B. Speiser, A n g w . Cheni. 1996. 108, 2623; Angeu,.
Cheni. I n t . Ed. Engl. 1996, 35. 2471
[S] a ) H. Taube, A n g m Cheni. 1984, 96. 315; Angeit Chenz. Inr Ed. Engl. 1984,
23. 329; b) Ehctron fiunsfer Reacrions of Compler Ions in Solution, Academic
Press, New York. 1970.
[6] G N Sastry. S. Shaik. J Am. Chem. Soc. 1995. 117. 3290.
[7] a ) N.Kimurd, S. Takamuku. Bull. Cheni. SOC.Jpn. 1991.64, 2433; b) J. Am.
Chcm. Sor. 1994. 116. 4087.
[XI G. N. Sastry. A. C. Reddy, S. Shdik, Angeu. Chrm. 1995, 107. 1619; Angeiv
Clretn. Ifit E d Engl. 1995, 34. 1495.
[9] G. N.Sastry. D. Danovich, S. Shaik, Angew. Chem. 1996. 108, 1208; Angew.
Chcvii. Int. Ed. En$. 1996, 35, 1098
[lo] S. S . Shaik. H. B. Schlegel. S. Wolfe. Theoreticol Asprcts of Phjsicul Orgunic
Clieniistrj-The S,2 Mechunlrm, Wdey. New York, 1992.
[ I l l J. Bertran. I. Gallardo. M. Moreno, JLM. Saveant, J Am. Chem. SOC.1996,
11K. 5737.
[12] G. N. Sastry, S . Shaik, J Phys. Chem. 1996. fO0, 12241.
[I 31 A. C. Reddy, G. N. Sastry, S. Shaik, J. Cheni. Soc. Perkin Truns. 2 1995, 1717
[14] A. C. Reddy. D. Danovich. A. Ioffe, S. Shaik, J. Chein. Sue. Perkin Tram 2
1995, 1525.
[15] L. M. Tolbert, J. Bedlek, M. Terapane. J. Kowalik. J. Am. Chem Sor 1997.1fY.
Amino Acid Derivatives by Multicomponent Reactions
Gerald Dyker"
Amino acids are one of the most important classes of naturally occurring substances and possess a variety of biological functions. The areas of application of the amino acid derivatives
range from sweeteners through pharmaceuticals to crop and
plant protection. Widely applicable methods for the synthesis of
a-amino acids".
are of great interest especially for the construction of compound libraries by combinatorial chemistry.
Multicomponent reactions as a special case of a domino processE3]are particularly fascinating since they facilitate rapid syntheses from simple building blocks.
A classic example of such a reaction is the Strecker synthesis,
which has been known for almost 150 years but is still as topical
as ever:[41 an aldehyde 1 (or a ketone) is condensed, using a n
acid catalyst, with an amine 2 and an alkali metal cyanide such
as 3 to give an a-aminonitrile 4, which can be hydrolyzed to the
amino acid 5 . In recent investigations on the Strecker synthesis
attempts have been made to optimize the reaction conditions[4b1
and to achieve stereoselective syntheses.r4c-h1 Usually chiral
amines such as 7 have been employed, which can, for example,
on reaction with 6 be converted to the thiophene-substituted
amino acid 8.r4c1
Prof. Dr. G Dyker
Fachbereich 6, Organische/Metallorganische Chemie
der Universitdt-Gesamthochschule
Lotharstrasse 1, D-47048 Duisburg (Germany)
Fax: Int. code +(203)379-4192
e-mail: dykerio
WILEY-VCH Verlag GmbH, D-69451 Weinheim 1997
In Ugi's four-component condensation, imine formation
from an aldehyde 1 and an amine 2 is likewise the initiating
a carboxylic acid 9 and an isonitrile 10 are the other
reaction components, which finally yield the bisamide 11. Both
for this reaction and the Strecker synthesis, the galactosylamine
12 is particularly suitable for carrying out a stereoselective reaction (synthesis of 13) .[4d.e, 5 f 1 With a n aminoglucopyranose as a
chiral auxiliary the stereoselectivity of the reaction can be further increased.[5b1Amino acids as condensation components
yield particularly impressive results. For instance, the iminodicarboxylic acid 17 can be synthesized from components 14-16
0570-083319713616-1700 S 17 SO+ 5010
A n p i t Chem Int Ed Engl 1997. 36, No 16
Без категории
Размер файла
449 Кб
unboundчthat, bound, transfer, electro, state, questions, transitional
Пожаловаться на содержимое документа