вход по аккаунту


Kinetic Evidence for Reactive Intermediates.

код для вставкиСкачать
V O L U M E 9 N U M B E R YO
O C TO BER 1 9 7 0
P A G E S 751-818
Kinetic Evidence for Reactive Intermediates[**]
By Rolf Huisgen[*l
Spectroscopic methods have recently been developed that allow the direct detection of
reactive intermediates under favorable circumstances. The most general and reliable
method, however, is still the indirect one, which makes use of the principle of kinefic
competition; this method isrbased on the freedom of the intermediate to choose between
several reaction possibilities. The foIlowing discussion is addressed less to the specialist
in reaction mechanisms than to the outsider who wishes to obtain some idea of the value
and limitations of the kinetic method.
1. Fundamental Concepts
l t is not true that chemists became interested in reaction mechanisms only about 40 years ago. The reaction event fascinated even the chemist of earlier
periods, but the knowledge and methods required for
the experimental testing of views on reaction mechanisms were lacking. These workers, consequently,
often had to be satisfied with determining the stoichiometry of a reaction.
The methods that have been developed in the meantime have led to a rapid growth of our knowledge of
reaction mechanisms. The term reaction mechanism
leads us first to think of the structural changes in the
components during the reaction event. Changes in the
distance functions and bond lengths can be described,
possibly supplemented by electron density and charge
separation. The conceptual clarity of these pictures
and models was enhanced by the introduction of
energy relations. At the beginning of his course, the
chemistry student of today is taught to look on the
interplay of the molecules as a n “energy landscape”.
Depending on their internal energy content, the molecules or collision complexes move up and down in
their potential troughs. The reaction event is associated
with escape from such a trough, i.e. with passage over
its rim.
[ * ] Prof. R. Huisgen
Institut fur Organische Chemie der Universitat
8 Miinchen 2, Karlstrasse 23 (Germany)
[**I Based on a GDCh training course from October 6-9,
1969 in Munich.
Angew. Chem. internat. Edit.
Vol. 9 (1970)I No. I0
The structural and energetic approaches are combined
in the energy profile. The chemical potential of the
reacting system is plotted on the ordinate, preferably
as the free energy, and the abscissa is the reaction
coordinate, a geometrical parameter symbolizing the
interplay of successive breakage and formation of
bonds. The energy profile may consist of a single
barrier, the summit of which is referred to as the
transition state; the activation energy AG* must be
supplied in order to reach this peak. This is followed
by a descent into a trough with liberation of energy;
the amount by which the energy set free exceeds the
activation energy is known as the heat of reaction.
The molecules o r collision complexes usually leave
their potential trough at the lowest point, even if this
reaction path does not lead to the product with the
maximum bond energy. The chemical reaction is
controlled kinetically, i.e. by the height of the activation barrier, and not thermodynamically by the quality
of the product, unless a series of forward and reverse
reactions allows the thermodynamically favored
Many reactions that were described earlier by a simple
activation peak (Fig. la) have turned out to be more
complicated systems, frequently even involving several
intermediates (Fig. lb).
Transition states are not really “states” with finite
lifetimes, but snapshots. On the other hand the dips in
the energy profile, i.e. the intermediates, have definite
lifetimes. Energy profiles can resemble a cross section
through the Alps. A feature that they share with a
mountain panorama is that only the first peak can
usually be seen from the bottom, and one certainly
Reaction Coordinate +
Reaction Coordinate
Fig. 1. Energy profiles of reactions (a) without and (b) with intermediates.
-1 : transition states; : intermediates.
cannot see over the highest ridge. The rate-determining step of the system is the conquest of the
highest peak. This is the only peak about which information is provided by the overall kinetics with the
activation enthalpy and the activation entropy.
How does one obtain information about what happens
behind the first high ridge? In other words, how does
one detect intermediates? Only a few decades ago it
was considered irresponsible to postulate intermediates that could not be isolated and stored in flasks,
Nowadays scepticism is aroused by any claim to the
isolation of a reactive intermediate as such. The substance in question is all too often a compound from a
side-equilibrium of the kinetic system. The walls of the
energy trough in which the reactive intermediate lies
are often so low that the material cannot collect here,
but flows away rapidly with formation of the products.
To detect short-lived intermediates directly, it is
necessary either to produce a sufficiently high stationary concentration, as occurs e.g. in flash photolysis,
o r t o use an extremely sensitive method of analysis;
ESR spectroscopy, for example, responds even to very
small concentrations of free radicals. Though such
spectroscopic methods provide valuable information
on the nature of intermediates, they are not free from
the above shortcoming in the distinction of main path
and side-equilibria.
Elegant and reliable criteria for the detection of unisolable intermediates are based on kinetic comA+B-+C
Intermediate B
in main path
in side-equilibrium
petition. If there is only one energy peak between the
reactants and the product, the product cannot
contain anything that was not already present in the
starting system, i.e. that did not enter into the ratedetermining activation process; the energy profile is
unambiguous. A dip in the energy profile, on the other
hand, may be interpreted as a “breather” of the
reacting system. The system comes to rest here, and
can choose between several possible reactions. It can
enter bonding interactions with molecules that were
not involved in the first, rate-determining step, i.e. in
the formation of the intermediate.
A well-chosen competition system gives reliable
evidence for an intermediate, but not much more.
Information about its nature is limited; such information may be available from the type of the reactions
between which the intermediate chooses. As to the
structure of the intermediate, analogies or the introduction of further experimental criteria, such as the
steric course of the reaction, are necessary. Some role
must also be given to chemical intuition here, though
not to utter fantasy. Great caution should be exercised
in the postulation of new elementary reaction steps o r
in the assumption of more intermediates than are
verified experimentally and by convincing analogy.
A principle that applies in all natural sciences, and
which is known as Occam’s razor [I], enjoins the use
of the simplest possible description: Entia non sunt
multiplicanda praeter necessitatem.
What are reactive intermediates? It is not easy to add
chemically to the above definition as a “dip in the
energy profile”. One usually thinks first of short-lived
radicals, carbonium ions, carbanions, carbenes, azenes
(nitrenes), and the like; some compounds of these
classes can be intercepted and isolated. However, the
range extends far beyond the electron-deficient and
electron-rich compounds, as is shown by singlet oxygen [21, diimine [31, strained ring systems, unstable
oxidation states, o r energetically unfavored valence
tautomers. Photo-excited states all belong to this
group. Close relatives of classes of compounds that are
known to be stable sometimes also defy isolation, and
kinetic detection then becomes necessary, as illustrated
later for diphenylnitrilimine.
If many of the examples described below are taken
from our own work, this is not because we attach
particular importance to it, but is rather intended to
show that the kinetic detection of intermediates is a
routine method for research on reaction mechanisms.
2. Reaction Order as Evidence for Intermediates
If the reaction product contains a structural unit that
does not appear in the rate equation of the determining step, it is logical to conclude that an inter[l] William of Occam, born near London in 1285, died at
Munich 1349.
121 C. S. Foote, Accounts chem. Res. I, 104 (1968).
[31 For a review, see S. Hunig, H. R. Miiller, and W . Thier,
Angew. Chem. 77, 368 (1965); Angew. Chem. internat. Edit.
4, 271 (1965).
Angew. Chem. internat. Edit.
1 Vol. 9 (1970) No.
mediate is involved. The rate equation in this case
does not reflect the stoichiometry of the process as a
Aziridines of the type ( I ) can enter into cycloadditions
in which the CC bond is brokenL4.51. If we follow the
adduct formation with tetracyanoethylene a t 100 “C
dilatornetrically, we find that the rate equation does
not contain the concentration of tetracyanoethylene,
as long as it is present in an a t least equimolar amount [61. The equation is
d [adduct]/dt = k i [aziridine (I)]
Experiments with 10, 20, and 30 equivalents of tetracyanoethylene gave the same dilatometric first order
rate constant.
o-quinodimethane (5) as a n intermediate, since the
trans-diphenyl compound (4) gives a Diels-Alder
adduct (6) with cis-phenyl residuesal.
A classic example of the discrepancy between stoichiometry and the rate equation is found in the halogenation of acetone in aqueous acid. Lapworth[*] discovered that bromine participates with “zero order” and
that the rate is determined by the concentrations of
acetone and of acid:
d [acetonelid? = k2 [H,O+] [acetonel
CH302C p 0 2 C H 3
The chlorination proceeds with the same rate constant as bromination; in both cases an intermediate is
intercepted by the halogen, the formation of this intermediate determining the overall rate of the reaction.
A r = C6H40CH3-p
A step preceding the cycloaddition must therefore
alone be rate-determining. It is not the aziridine ( I ) ,
but the azomethine ylide (2), which is present in a low
equilibrium concentration and combines as a 1,3dipole with the olefinic dipolarophile. The reaction
with tetracyanoethylene is so fast that the reverse
reaction (2) + ( I ) is completely suppressed and the
rate of the adduct formation is determined only by the
ring cleavage ( I ) +2), i.e. by kl.
very fast
Regarding the nature of the intermediate, all we can
learn from the kinetic investigation is its high rate of
halogenation. Since the acid-catalyzed iodination and
The intermediate (2) can be detected stereochemically
racemization of (+)-methylethylacetophenone (8) also
as well as kinetically. The aziridine-trans-2,3-dicarboshow the same rate constant [91, it seems likely that the
xylic ester ( I ) gives a cycloadduct (3) with ester groups
enol form (7) of acetone is involved. Kinetic measurein the cis-2,5 positions. Such a steric course is inments give equal rate constants for the base-catalyzed
conceivable in a one-step process. A satisfactory
iodination and bromination of acetone, the reactions
explanation is provided by the conrotatory ring
being of zero order with respect to NaOCl and NaOBr,
cleavage ( I ) +(2) [51.
respectively[lol. The decisive role of the enolate ion
Interrelations of benzocyclobutenes with o-quinodi(9) was underlined by the observation that alkalimethanes have been postulated, but the valence
catalyzed racemization and H,D-exchange in the
tautomerism (4) +(5 ) was demonstrated only by
optically active ketone (8) proceed with equal rate
kinetic measurement of the adduct formation with
constants [11J.
tetracyanoethylene. As in the last example, the reaction
rate depends only on the concentration of the 7,s-diHsC -?! -C- C6H5
phenylbenzocyclobutene. The constant k2 is so large
that the reverse reaction (5)+(4) is of n o importance.
The steric course also confirms the occurrence of the
[4] H . W. Heine, R . Peavy, and A . J . Durbetaki, J. org. Chemistry 31, 3924 (1965).
[51 R . Huisgen, W. Scheer, C. Szeimies, and H . Huber, Tetrahedron Letters 1966, 397; R . Huisgen, W. Scheer, and H. Huber,
J. Amer. chem. SOC. 89, 1753 (1967).
[6] R . Huisgen, W. Scheer, and H . Mader, Angew. Chem. 81,
619 (1969); Angew. Chem. internat. Edit. 8 , 602 (1969).
Angew. Chem. internat. Edit. J Vol. 9 (1970) J No. I0
[7] R . Huisgen and H . Seidl, Tetrahedron Letters 1964, 3381.
[8] A . Lapworth, J. chem. SOC.(London) 85, 30 (1904).
[9] P . D . Bartlett and C. H . Stauffer, J. Amer. chem. SOC.57,
2580 (1935).
[lo] P. D . Bartlett, J . Amer. chem. SOC.56, 967 (1934).
[lll S. K . Hsii, C. K . Ingold, and C . L . Wilson, J. chem. SOC.
(London) Z938, 78.
3. Competition of Forward and Reverse Reaction
In all the examples considered above, the rate-determining process was a tautomerization. In the absence
of an intercepting reagent, the reactive tautomers occur
in equilibrium concentrations so small that they
cannot be detected directly. This means that the formation of the reactive tautomer (rate k l ) is much
slower than its reverse reaction (k-1).A reaction order
of zero for the intercepting reagent indicates that the
combination of this reagent with the intermediate is
even much faster than the reverse reaction mentioned.
If k-l and k2 are of the same order of magnitude, this
does not affect the reliability of detection of the intermediate. The rate equation becomes somewhat more
complicated, and the intercepting reagent is now
involved as a function of its concentration. The DielsAlder reactions of cyclooctatetraene, which are accompanied by rearrangement, may be taken as a n
example. The addition of tetracyanoethylene is preceded by tautomerization to bicyclo[4.2.0]octatriene
The formation of (12) is followed dilatometrically,
evaluation as a first o r pseudo-first order reaction
being allowed by the use of a sufficiently large excess
of dienophile. Figure 2 shows dilatometric constants
( k d ) of this type as a function of the excess of dienophile and of the temperature. If the bimolecular
combination with the dienophile alone were ratedetermining, the results should be straight lines
passing through the origin; kd = k10 should be proportional to [D] (dienophile concentration).
The quantitative evaluation of the competition of
reverse and forward reactions makes use of the steady
state approximation (Bodenstein theorem) [13,141.
The change in the concentration of the intermediate
(11) is negligible compared with those of (10) and
(12), and is equated to zero.
[(II)] can be substituted in the rate equation for the
adduct formation:
Comparison with the overall equation for the dilatometric measurement
d [(12)l/dt
yields kd as a function of [D]:
(11); the equilibrium concentration of (11) found
indirectly from kinetic data is only 0.01 % in dioxane
at 100 “C (121.
Simple transformation gives the equation of a straight
h e for a plot of kd against kd/[D];kl is the intercept
on the ordinate and - k-l/k2 is the slope.
0 ZL
Tetracyanoethylene lrnolelll+
Fig. 2. Dilatometric rate constants kd for the reaction of cyclooctatetraene (0.02 mole/l) with excess tetracyanoethylene in dioxane at
various temperatures.
[12] R. Huisgen and F. Mietzsch, Angew. Chem. 76, 36 (1964);
Angew. Chem. internat. Edit. 3, 85 (1964).
7 54
In the extreme case k2/[D]
k-1, which was dealt
with in Section 2, this expression becomes kd = k l .
Since the kd values in Figure 2 give straight lines in
this new type of plot (Fig. 3), the kinetic system must
be correct. The competition of a first order reverse
reaction with a second order cycloaddition proves the
occurrence of an intermediate in this case. At an
infinitely high dienophile concentration, i.e. for
kd/[D]= 0,we obtain kd = kl. From the temperature
dependence of the extrapolated kl values, the following
values result for the activation process of the valence
tautomerization: AH* = 28.1 kcal/mole; A s * =
+1 e.u. 1151.
1131 M . Bodenstein, Z. physik. Chem. 85, 329 (1913).
1141 A . A . Frost and R. G . Pearson: Kinetics and Mechanism.
Wiley, New York 1961, 2nd Edit.; Kinetik und Mechanismus
homogener chemischer Reaktionen. Verlag Chemie, Weinheim 1964, p. 159.
[15] F. Mietzsch, Dissertation, Universitat Miinchen 1965.
Angew. Chem. internat. Edit. / Vol. 9 (1970) [ No. I0
Fig. 3.
II rnote-’s-’l
Dilatometric constants of Fig. 2 plotted against k d / l D l .
This tells us nothing about the nature of the intermediate. Bicyclooctatriene (11) has the quasi-planar
diene system required for the Diels-Alder reaction,
while (10) has not. In the equilibrium (13)+(14), in
which both tautomers can be isolated, only (14) again
adds dienophiles [161. The extrapolated k l values
should be independent of the nature of the dienophile,
as has in fact been shown for (10) and its phenyl
derivative 1131. Vogel, Kiefer, and Roth 1171 were able to
confirm the existence of (11) by its preparation at
-78 “C; (11) changes into (10) even below 0 “C.
The Diels-Alder adducts of cycloheptatriene (15) are
structurally derived from norcaradiene (16) [I*].
Kinetic measurements on the reaction with tetracyanoethylene gave k d values that were proportional
to the dienophile concentration over the entire range
measured (193. This means that the establishment
of the equilibrium (15)+(16) is fast in comparison
with the Diels-Alder reaction. The intermediate
cannot be detected kinetically in this case. If it is
assumed that the diene activity of (16) can be equated
[16] R . Huisgen, F. Mietzsch, G. Boche, and H . Seidl: Organic
Reaction Mechanisms. Chem. SOC.(London), Spec. Publ. 19, 3
[17] E. Vogel, H . Kiefer, and W . R . Roth, Angew. Chem. 76,
432 (1964); Angew. Chem. internat. Edit. 3, 442 (1964).
[I 81 For a review of the cycloheptatriene/norcaradieneproblem see G . Maier, Angew. Chem. 79, 446 (1967); Angew. Chem.
internat. Edit. 6, 402 (1967).
[19] R . Huisgen and W . D . Wirth, experiments performed in
Angew. Chem. internat. Edit. 1 Vol. 9 (1970) 1 NO. 10
to that of ( 1 4 ) (equal k 2 values for the addition of
tetracyanoethylene to the cyclohexadiene portion of
the two molecules), a 0.1 % equilibrium concentration
of (16) at 20 “C would lead to k d values of the order
The classification of nucleophilic aliphatic substitutions as 1S,
and S,2 by Hughes and Ingold is based on
the kinetic criterion (201. An ionization was postulated
as the rate determining step for the first order substitution. It is interesting to note that kinetic evidence for
the carbonium intermediate comes only from the
deviations from the first order.
The ionization of 4,4’-dimethylbenzhydryl chloride
(17) in aqueous acetone leads to a rather selective
carbonium ion, which can not only react with water
to give benzhydrol (19) but can also combine with
chloride ion formed to regenerate the starting material.
The kinetic situation, a combination of a reversible
and a consecutive system, is similar to that described
above, except that a second order reverse reaction (kz)
now competes with a pseudo-first order forward
reaction (k+) (since water is a component of the
solvent, its concentration remains practically constant) for the intermediate. As above, the Bodenstein
theorem is applied.
The ratio k~[Cle]/k+in the denominator is the competition constant of the reverse reaction and the carbinol formation. The hydrolysis of dimethylbenzhydryl chloride thus begins with the rate constant kl
as long as n o chloride ions are present. As the reaction
proceeds, the liberated chloride ions cause a decrease
in the overall hydrolysis constant. koveraiiis thus n o
longer really a “constant”, and the treatment as a first
order reaction is no longer a permissible approximation. In the special case, koverall falls to 72% of
the initial value after 90% conversion when a 0.04 M
solution of (17) in 85% aqueous acetone is usedr211.
If the chloride ion is present in the solution from the
outset, e.g. as 0.06 M LiCI, the hydrolysis begins
with only 54% of the koverall value found without the
addition of the salt. The hydrolysis constant of triphenylmethyl chloride in 85 % acetone is decreased by
a factor of 4 on addition of 0.02 M NaCl122J.
The influence of the chloride ion on the hydrolysis
constant is even greater than appears from the above
data. The kinetic scheme is further complicated by the
[20] For reviews, see C. K . Ingold Structure and Mechanism
in Organic Chemistry. G. Bell and Sons, London 1953, p. 306;
A . Streitwieser, Chem. Reviews 56, 571 (1956).
[21] L. C. Bateman, E. D . Hughes, and C. K . Ingold, J. chem.
SOC.(London) 1940, 974.
[22] C. G. Swain, C. B. Scott, and K . H . Lohmann, J. Amer
chem. SOC.75, 136 (1953).
7 55
fact that salts in general cause an increase in the ionization rate by increasing the ionic strength of the
solvent [231. This acceleration by the salt effect is overcompensated in the case of the chloride ion by the
regeneration of the organic chloride.
+ d IFlidt
+ d [Glldt
- d [D]/dt
- d [El/dt
k D [C] [D]
kE [C] [El
Division of the first differential equation by the second
I D 1 _ _ k_D_ _[Dl
-dd [El
4. Rate-Product Discrepancy
This term covers the evidence for intermediates on the
basis of incompatibility of the rate equation with the
stoichiometry (see Section 2). We shall use the term
rate-product discrepancy primarily to denote the
absence of quantitafive agreement.
If 0.05 M sodium azide is added to a solution
of 4,4'-dimethylbenzhydryl chloride ( I 7) in 85 %
acetone, the rate constant of the solvolysis increases
by 48%; the product contains 60% of the organic
azide (18) and 40% of the carbinol (19) [231. If the
increase in rate were due to a direct reaction of the
chloride (17) with the azide ion, only 32% of (18)
and furnishes, after transformation, the competition
d [El
d l n [El
The replacement of the differentials by finite differences
is allowed. If [Dlo and [El0 are the initial concentrations of the competing reactants and [D]e and [E]e
their concentrations at the end of the reaction, we
obtain an expression that contains only quantities that
are experimentally accessible.
The validity of this equation is unrestricted. If D and
E are present in a very large excess, there will be
practically no change in their concentrations during
the reaction. [D]o and [El0 are now constants, and
integration leads to the even simpler expression:
should be expected. The rate acceleration is in fact a
salt effect, which occurs to an equal degree on addition of 0.05 M tetramethylammonium nitrate. The
reaction with the azide ion thus takes place only at the
carbonium stage, where azide and water arecompeting.
The experimental procedure is to allow D and E to
compete for C in various ratios and to determine the
concentrations of the products F and G. The reliability
of the method is then shown by agreement in the
competition constants K.
This can be illustrated by the detection of benzyne as
an intermediate in the nucleophilic substitution of
halobenzenes by phenyllithium o r lithium piperidide.
The substitution of unactivated aryl halides with
strong nucleophiles proceeds with primary elimination,
as is revealed by a characteristic change in the position
of the substituent (cine substitution) 1241.
The liberation of benzyne (21) from fluorobenzene in
ether at 20 "C proceeds 25 times as rapidly with lithium
Reaction Coordinate-
Fig. 4. Competition of two reaction components D and E for an intermediate C.
How does one define and determine competition
constants? In Figure 4, the reactants D and E compete
for the intermediate C with formation of the products
F and G.
I231 L . C. Bateman, M . G . Church, E . D . Hughes, C. K . Ingold,
and N . A . Taher, J. chem. S O C . (London) 1940, 979.
Reaction Coordinate-
Fig. 5. Energy profiles for the nucleophilic substitution of halobenzene
with phenyllithium and with lithium piperidide
[24] For reviews, see R . Huisgen and J . Sauer, Angew. Chem.
72, 91 (1960); R . W . Hoffmnnn: Dehydrobenzene and Cyclo-
alkynes. Verlag Chemie, Weinheim 1967.
Angew. Chem. internat. Edit. Vol. 9 (1970)
1 No. 10
piperidide as with phenyllithium [251. The energy
profile (Fig. 5 ) contains an additional dip; the first
intermediate is the o-lithiohalobenzene (201, which
can also be intercepted. The slow, rate-determining
formation of benzyne is followed by fast additions of
nucleophilic agents to the short-lived intermediate.
Competition tests with lithium piperidide and phenyllithium in various ratios give x = 0.23. Though slower
by a factor of 25 in the liberation of benzyne, phenyllithium is 4.4 times as fast as lithium piperidide in the
addition to (21) 1261. This discrepancy between the
overall kinetics and the product composition makes
the assumption of an intermediate (20) unavoidable.
particular benzenediazonium o-carboxylate (24) 1271
and benzo[d]thiadiazole 1 ,I-dioxide (25) [281. The
competing reactants for the detection of an identical
intermediate in all these systems must be compounds
that are neither electrophilic nor strongly nucleophilic,
since reactions with the precursors of the benzyne
must be avoided.
The cycloadditions to furan and cyclohexadiene to
form (22) and (23) were found to be suitable. Under
identical conditions, the benzyne from various sources
adds to the two dienes with the same competition
constant (Table 1) of 21.5 (average value) 1293. In view
of the difference in the starting systems, it is difficult
to imagine an intermediate other than C6H4 (21) as a
“common denominator”; the only other possibility
would be a complex of (21) with the solvent.
Table I . Competitive addition of furan ( k F ) and cyclohevadiene ( k c )
to benzyne (21) from various precursors in tetrahydrofuran a t 50°C[291.
[A782 6;
Lithium piperidide
Fig. 6. Competition of phenyllithium and lithium piperidide in the
addition t o benzyne. Experiments with Auorobenzene (0)
and chlorobenzene (0)
in boiling ether 1261.
Figure 6 shows a plot of the concentration ratio of the
competing reactants against that of the products. The
experimental data give a straight line passing through
the origin. The agreement of the x values for experiments with fluorobenzene and with chlorobenzene
indicates that the intermediate responsible for the
product formation is free from halogen. Can any other
information be obtained?
5. Common Intermediate from Different
If one and the same intermediate is formed from different starting systems, the nature of the generating
systems often provides evidence on the structure of the
intermediate. The occurrence of a common intermediate should not be concluded merely from qualitative preparative findings. Rigid proof can be provided
only by agreement of the competition constants in a
quantitative competition test.
A pertinent example is again offered by the structure
of benzyne; C6H4 was at first regarded with much
suspicion. In addition to the above path via metalated
halobenzenes, other precursors became known, in
[25j R . Huisgen and J . Sauer, Chem. Ber. 92, 192 (1959).
[26] R . Huisgen, W. Mack, and L . Mobius, Tetrahedron 9 , 29
(1960). Confirmation of the competition constant: Th. Kauffmann, H . Fischer, R . Nurnberg, M . Vestweber, and R . Wirthwein, Tetrahedron Letters 1967, 2911.
Angew. Chem. internat. Edit.
1 Vol. 9
/ No.
kF/kc 20.8
A recent study made use of the competitive addition
of benzyne to the 1,4- and 9,lO-positions of 1,4-dimethoxyanthracene [301. The competing reactants thus
belong to one and the same molecule. The identity of
the intermediate was confirmed and extended to two
other starting systems. However, the competition
constant was only 2.3.
One critical point is that the conclusion of a common
intermediate is permissible only if the competition
constant is sufficiently different from 1. Very “hot”
intermediates react with the first partner encountered,
i.e. randomly. Competition constants of 0.5 to 2 do
not rule out the possibility of substantially diffitsioncontrolled processes. In such cases, independent experiments with different ratios of the competing reactants
are particularly important.
Phenylations of aromatic compounds have been
known for more than 70 years. They were ascribed to
the free phenyl radical by Wielandr311 and by Hey[321.
Such phenylations have been observed e.g. in the
homolysis of nitrosoacetanilide (26), phenylazotri[27] M . Stiles and R . G. Miller, J. Amer. chem. SOC.82, 3802
[28] G. Wittig and R . W. Hoffmann, Chem. Ber. 95,2718 (1962).
1291 R . Huisgen and R . Knorr, Tetrahedron Letters 1963, 1017.
[30] B. H . Klandermann and T. R . Criswell, J. Amer. chem.
SOC. 91, 510 (1969).
[ 3 1 ] H. Wieland and K . Heymann, Liebigs Ann. Chem. 514,
145, 154 (1934).
I321 W . S . M . Grieve and D.H . Hey, J. chem. SOC. (London)
1934, 1197; D . H . Hey, ibid. 1934, 1966.
phenylmethane (27), and dibenzoyl peroxide (28) in
aromatic solvents C ~ H S S(s = substituent).
Competitive phenylations again confirm the identity
of the intermediate responsible for the aromatic substitution in the decomposition of (26)-(28) (Table 2).
The 0-, m-, and p-positions of monosubstituted
benzenes can compete intramolecularly for the phenyl
radical, and this leads to a definite isomer ratio of the
monosubstituted biphenyls. Mixtures of aromatic
compounds allow the determination of intermolecular
competition constants, as is shown in Table 2 for the
example naphthaiene/benzene.
Table 2. Isomer ratios in the free-radical substitution (phenylation) of
aromatic compounds using nitrosoacetanilide (26). phenylazotriphenylmethane (27). or dibenzoyl peroxide (28) as phenyl generator.
Aromatic substrate
intermediate is demonstrated conclusively by the
numerical determination of competition constants.
Hardly any intermediate other than (32) could conceivably be common to the above four systems [361.
Table 3. Cycloadditions of diphenylnitrilimine (32) to a.@-unsaturated
carboxylic esters to form the pyrazolines (33) and (34); intramolecular
Chlorobenzene 1331
% 0% m% PPyridine [341
% 2% 3-
Yield of
(33) i 1341
Generator of (32)
(33) :(34)
( %)
% 4Naphthalene [331
% 1% 2(Naphthalene/benzene) 1351
a) Methyl cinnamate (R = C6HS)
N(CzH& at 80 "C in benzene
in acetonitrile
N(C4H9)3 in anisole at 1 6 0 T
(31) in anisole at 160°C
67 : 33
63 : 37
57 : 43
59 :41
72 :28
16 : 24
b) Methyl crotonate (R = CHI)
Benzonitrile oxide is an isolable compound, but dimerizes rapidly in solution to give diphenylfuroxane.
All efforts to isolate the related diphenylnitrilimine
(32) have been unsuccessful because of its fast further
reactions. However, cycloadditions can be readily
carried out with (32) in situ, i.e. on liberation in the
presence of suitable dipolarophiles, and this reaction
has provided preparative routes to five-membered
Systems as different as the base-induced elimination
of HX from (a-chlorobenzy1idene)phenylhydrazine
(29) and (a-nitrobenzy1idene)phenylhydrazine (30)
and photolysis o r 160 "C-thermolysis of 2,5-diphenyltetrazole (31) produce the same cycloadducts with
multiple bond systems. The occurrence of a common
[33] R. Huisgen and R . Grashey, Liebigs Ann. Chem. 607, 46
[34] D. H. Hey, C. J . M. Stirling, and G. H. Williams, J . chern.
SOC.(London) 1955, 3963.
1351 R . Huisgen, F. Jakob, and R . Grashey, Chern. Ber. 92,
2206 (1959).
N(CzH& in benzene at 20 "C
Photolysis of (31) in benzene at 20 "C
In the 1,3-dipolar cycloaddition of diphenylnitrilimine
(32) to methyl cinnamate, the two addition paths
leading to (33) and (34), R = C6H5, compete with
each other. The systems yielding (32) are compared in
pairs in Table 3; care must be taken to ensure comparability of the reaction conditions with regard to
temperature and solvent. The competition constant is
independent of the temperature only if the competing
reactions happen to have the same activation enthalpy.
In the case of methyl crotonate, the photolysis of diphenyltetrazole (31) was also examined. Though the
ratios (33) : (34) agree satisfactorily within the series
of experiments a or b, their values are only 1.3 :1
and 3 : 1 respectively.
The intermolecular competition of two olefinic dipoIarophiIes for diphenylnitrilimine was therefore also
[36] J . S. Clovis, A . EckeN, R . Huisgen, and R . Sustmann,
Chern. Ber. 100,60 (1967).
Angew. Chem. internat. Edit.
Vol. 9 (1970)
1 No. I0
measured (Table 4). The thermolysis of diphenyltetrazole (31) gives constants that agreed with that of
the reaction (29) + tributylamine when tributylamine
and tributylammonium chloride were added, i.e. when
the medium is made identical. The magnitude of the
competition constant, x = 7.5, makes the common
intermediate convincing.
Table 4. Competition of ethyl cinnamate and indene for diphenylnitrilimine (32) in anisole at 160°C [ 3 6 ] .
Generator of (32)
Cinnamate adduct
Substituents may stabilize or destabilize intermediates.
A measure of this is provided by the selectivities as
mPasured by the competition constants in a standard
system D + E.
In the S N ~hydrolysis of alkyl halides, the recombination of the ions competes with the reaction of the
carboniumjon with water (Section 3). The quantitative
evaluation gives x = k c i e / k H , o . Though the data in
Table 5 sre only approximate values, the decreasing
seIectivities clearly reflect the diminishing resonance
stabilization of the carbonium intermediate.
Indene adduct
(31) f N(C4Hv)s f HN(C4Hg)tCIS
7.57, 7.36
7.67. 7.36
( 3 1 1 alone
9.50, 9.22
Table 5. Competirion consants for the reactions of carbonium ions
with pairs of nucleophiles. CI@/HzO (hydrolysis of RHal), and
N p I H 2 0 121-23. 371.
Triphenylmethyl chloride
4,4’-Dimethylbenzhydryl chloride
4-Methylbenzhydryl chloride
Benzhydryl chloride
rwl-Butyl chloride
6. Selectivity of the Intermediate
If t h e intermediate lies in a relatively deep energy
trough, it has sufficient time to choose carefully between the components D and E (Fig. 7). The size of
the competition constant y. provides a measure of the
se2ectivity of an intermediate. Highly selective intermediates a r e commonly referred to as “cold”. These
are contrasted with the “hot” intermediates, which
react with the components D and E over very low
activation barriers. The extreme case of a hot intermediate is one that reacts on the first collision with D
o r E, i.e. whose reaction is diffusion-controlled. The
terms “hot” and “cold” should preferably be used in
comparisons between several structurally related intermediates rather than in an absolute sense.
3 100
The selectivity of an intermediate may be influenced
not only by substituents but also by the medium. The
photochlorination of saturated hydrocarbons proceeds
by a free-radical chain; the position of the substitution
is determined by the abstraction of hydrogen from the
alkane in the reaction with the chlorine atom. In the
chlorination of 2,3-dimethylbutane (35) the hydrogen
may be removed from a tertiary o r from a primary
carbon. Since the tertiary alkyl radical (37) is energetically favored over the primary one (36), the tertiary
hydrogen reacts preferentially. The competition con-
Reaction Coordinate
+ C1’
Table 6 . Photochlorination of 2.3-dimethylbutane (35) at 55 “C;
solvent dependence of the competition constant for the substitution o n
the tertiary and primary carbon atoms 1381.
Solvent (concentration in
Carbon tetrachloride (4.0)
Nitromethane (4.0)
Propionitrile (4.0)
Dioxane (4.0)
Chlorobenzene (4.0)
Benzene (2.0,4.0. 8.0)
p-Xylene (4.0)
Mesi t ylene
8.0,15. 32
Reaction Coordinate
Fig. 7. Energy profiles (a) with a “cold” and (b) with a “hot” intermediate.
Angew. Chem. internat. Edit. Vol. 9 (1970)
No. 10
[37] A . G . Ogston, E . R. Holiday, J . S. L. Philpot, and L. A.
Stocken, Trans. Faraday S O C . 44, 45 (1948).
[381 G. A . Russell, J. Amer. chem. S O C . 80,4987 (1958).
stant can be determined from the ratlo of the two substitution products (Table 6), taking into account the
fact that (35) contains 1 2 primary and 2 tertiary
The preference for the tertiary H is relatively small,
the competition constant being 3.7; the chlorine atom
is a “hot” species. RussellC3sl observed a striking increase in selectivity if the chlorination is carried o u t in
aromatic solvents (Table 5). It must therefore be
concluded that a stabilizing interaction 3ccurs with
the aromatic compound; the chlorine atom becomes
“colder”. A x-complex of the chlo.ine atom with the
aromatic system is postulated as an intermediate.
The measurement of competition constants also allows
the determination of the relative rates of many substrates with a given intermediate. This leads e.g. to the
substitution rules for free-radical phenylation when
many experiments of the type shown in Table 2 are
carried out. After the deterfiination of the o : m : p
ratio in the reaction of the phenyl radical with monosubstituted benzenes, it was necessary to determine
the intermolecular competition constant of benzene
with its monosubstituted derivatives. The combination
of these data gives the “partial rate factors” of Table 7;
these are relative values based on a value of 1.0 for H
in benzene. The rate constants of the electrophilic
bromination of monosubstituted benzenes extend
over 24 powers of ten. In the free-radical substitution,
the substituent effect is compressed to less than one
power of ten; thus the phenyl radical is a less selective
Table 7. Partial ratc factors of the free-radical phenylation of aromatic
compounds [39].
&:: 0:::
Table 8. Competition of dipolarophile pairs for diphenylnitrilimine
(32) in benzene a t 80°C; experimental x values in italics, the first
column giving the relatwe addition constants.
dimethyl fumarate
9 2.5, 2.6
11.41, 1.41
dimethyl acetylenedicarboxylate
ethyl acrylate
methyl methacrylate
methyl propiolate
norbornene t
5 ;;;;
ethyl cyanoformate
+ ethyl
E 1.00
+ styrene
I .62
7. Kinetic Differentiation of Simultaneous
A multiplicity of products may be formed via a common intermediate o r via energy profiles that are
separated from the outset. It is not too rare to find that
one and the same product is formed from the same
precursor by different reaction paths.
Thus it is now known that in the chlorination of olefins
in the liquid phase, both the addition to the double
bond and the allylic substitution can proceed by a
free-radical or by a cationic mechanism [40al. The
free-radical chain can be suppressed by operation
under oxygen.
When the pyrimidyl azide (38) is heated to 150-160 “C
with diethyl fumarate, 1,2-dihydronaphthalene, 1,3dimethoxybenzene, o r aniline, the substitution products
The detection of diphenylnitrilimine (32) as a common intermediate was described in Section 5. The
same method was used for the stepwise construction
of the characteristic scale of dipolarophile activities by
competition of various pairs of dipolarophiles for the
1,3-dipole. Part of a larger volume of data is shown in
Table 8c401. Each double arrow denotes a pair competing for diphenylnitrilimine. To keep the experimental errors within limits, frequent additional
pairings are recommended. The arbitrary assignment
of a value of 1.0 to the addition constant k2 (ethyl
crotonate) leads to the relative rate constants (on left
in Table 8).
[39] Data from G. H. Williams: Homolytic Aromatic Substitution. Pergamon Press, Oxford 1960, p. 72.
1401 A. Eckeil, R . Huisgen, R . Sustmann, G . Wallbillich, D .
Grashey, and E. Spindler, Chem. Ber. 100, 2192 (1967).
(38) = R-N,
f 42)
5 1.86, 1.89, 1.81
[40a] M . L . Poutsma, J. Amer. chem. SOC.87,2161,2172,4285
Angew. Chem. internat. Edit. 1 Vol. 9 (1970) 1 No. 10
(39), o r (40) plus (41), o r (42) plus (43), or (44) are
formed with elimination of nitrogen 141,421. Do these
formally similar reactions proceed by the same
mechanism? The rate of nitrogen evolution from (38)
was measured in chloronaphthalene with and without
additives (Table 9). This experiment discloses a dichotomy of the reaction paths.
‘Table 9. Kinetics of nitrogen evolution from 2-arido-4,6-dimethyIpyrimidine /381 (69 mmoles/l) in chloronaphthalene a t 180 “C 1411.
Additive (mmoles/l)
Diethyl fumarate (948)
Diethyl fumarate (357)
Diethyl Eurnarate (231)
1,2-Dihydronaphtha!ene (948)
trans-Stilbene (91 I)
I,3-Dirnethoxybenzene (948)
Aniline (948)
Benzonitrile (948)
3 .O
Fumaric ester and olefins “induce” the liberation of
nitrogen from the azide (38); they enter into the rate
equation. Division of the pseudo-first order constants,
measured with excess fumaric ester, by the fumaric
ester concentration gives constant k2 values. This
contrasts with the aromatic compounds, the addition
of which does not increase the rate of the self-decomposition of (38). kl(se1f-decomposition) also varies
only slightly in a large number of solvents. Despite the
formal similarity of the products, therefore, two
fundamentally different reaction paths are used. The
CC double bond combines with the azide belore the
release of the nitrogen to form the A2-triazoline (45),
which very rapidly loses nitrogen at the high reaction
temperature. Enamines add azides particularly rapidly
and allow the isolation of corresponding cycloadducts
with (38) even at low temperatures f431.
4 -, 4
( 4 2 ) - (44j
The last example will illustrate the simultaneous
occurrence of two reaction paths in one and the same
system producing the same product. The formation of
cyclobutanones from ketenes and olefins has recently
been the subject of mechanistic studies, in which most
criteria point to a one-step multicenter addition, i.r.
a simple activation peak (Fig. la), between the starting
components and the product ‘441. The second possible
reaction path, which passes through a zwitterionic
intermediate, occurs only in the addition of ketenes to
enamines; in this case optimum stabilization of the
intermediate is possible in accordance with (49).
Depending on whether 1-pyrrolidinoisobutene (48)
o r dimethylketene (47) is chosen in excess, the reaction
yields more of the cyclobutanone (50) or of the 1 : 2
adduct (51) ‘451. This points to a common intermediate
(49), which either undergoes ring closure to form (SO)
or enters into cycloaddition with a second molecule of
dimethylketene to form (51).
k, +(CHJ,C=C=O
Kinetic measurements show that the rate is first order
with respect to (47) and (48). The dependence of the
k value on the polarity of the solvent would be rather
small for the formation of the zwitterion (49) in the
rate-determining step: k(acetonitrile)/k(cyclohexane) = 78.
It was only the quantitative competition experiment
that showed the simultaneous occurrence of the concerted formation ( k c ) of the cycloadduct (50) and of
the path involving the zwitterion (kI)1461. The enamine
(48) was allowed to react with six equivalents of (47)
in increasing dilution with the solvent. If the reaction
proceeded entirely via the intermediate (49), the
yields of (50) and ( 5 1 ) should be related as follows:
On the other hand, the elimination of nitrogen from
(38) to form the pyrimidylazene (46) precedes the
interaction with phenol ethers o r with aniline. It is
therefore possible to distinguish between the azide
cycloaddition and the azene reaction by a simple
kinetic measurement. (45) and (46) are reactive intermediates; the olefinic component is involved only in
the formation of (45).
This simple mass law relation was not obeyed. Only
after the subtraction of a quantity of the cyclobutanone (50) formed by the path with k c which is
[411 R . Huisgen, K . v. Fraunberg, and H . .
Letters 1969, 2589.
[421 R . Huisgen and K . v. Fraunberg, Tetrahedron Letters
1969, 2595.
I431 R . Fusco, S. Rossi, and S. Muioranu, Tetrahedron Letters
1965, 1965.
1441 R . Huisgen, L . A . Feiler, and G . Binsch, Chern. Ber. 102,
3460 (1969).
[451 P. O t t o , L . A . Friler, and R . Huisgen, Angew. Chem. 80,
759 (1968); Angew. Chem. internat. Edit. 9, 737 (1968).
[461 R . Huisgen and P . Otto, J. Amer. chem. SOC.91, 5922
Angew. Chem. internat. Edit.
/ Vol. 9 (1970) 1 No. I 0
76 1
characteristic for each solvent [43 % of the (50) formed
in acetonitrile,63 %in chloroform, 92 %in cyclohexane]
was the above relation valid for the remainder of the
material. The difference in the solvent dependence of
the two branches of the overall reaction is typical:
Overall reaction
Concerted addition ( k c )
Path via the zwitterion ( k r )
The formation of the zwitterion (49) is associated
with strong charge separation, corresponding to an
increase in solvation during the activation process,
i.e. to strong promotion by the polarity of the solvent.
8. Closing Remarks
It is meaningless to a r . which metho is the most
significant in the elucidation of reaction mechanisms
and in the detection of reactive intermediates. Without
chemical kinetics c471, our knowledge would be much
poorer and more fragmentary. The kinetic method
can often positively rule out postulated reaction paths;
it can show with certainty the occurrence of an intermediate. Information on the nature of the intermediate
is not provided with the same certainty, since several
mechanisms sometimes follow the same kineticscheme.
However, the number of possible intermediates can be
cut down by the methods described in Section 5, and
it may also be possible to establish the structure of the
intermediate p e r exclusionem.
Our knowledge of reaction mechanisms has swelled
over the last 40 years into a broad stream fed from
many tributaries. The work of Hughes and JngoldI201,
who first made systematic use of the kinetic method in
the elucidation of nucleophilic aliphatic substitution,
played a n important part in the early phase of this
Received: November 10, 1969
I A 782 IEl
German version: Angew. Chem. 82, 783 (1970)
Translated by Express Translation Service, London
[47] The following are recommended as introductions to
chemical kinetics of increasing depth and detail: R. Huisgen in
Houben-Weyl-Miiller: Methoden der organischen Chemie.
4th Edit., Vol. III/l, p. 101, G. Thieme Verlag, Stuttgart 1955;
A. A . Frost and R . G. Pearson: Kinetics and Mechanism. Wiley,
New York 1961 ; Kinetik und Mechanismen homogener chemischer Reaktionen. Verlag Chemie, Weinheim 1964; A . Weissberger: Technique of Organic Chemistry. Vol. VIII ( 2 Volumes),
2nd Edit., Interscience Publ., New York 1961.
Angew. Chem. internat. Edit. ] Vol. 9 (1970)
No. 10
Без категории
Размер файла
1 035 Кб
evidence, intermediate, kinetics, reactive
Пожаловаться на содержимое документа