Use of Competition Kinetics with Fast Reactions of Grignard Reagents

Article abstract of DOI:10.1021/jo991777h

Competition kinetics are useful for estimation of the reactivities of Grignard reagents if the reaction rates do not differ widely and if exact rates are not needed. If the rate of mixing is slower than the rate of reaction, the ratios between the rates of fast and slow reagents are found to be too small. This is concluded from experiments in which results obtained by competition kinetics are compared with results obtained directly by flow stream procedures. A clearer picture of the reactivity ratios is obtained when the highly reactive reagent is highly diluted with its competitor. A fast reagent may account for almost all the product even when present as only 1 part in 100 parts of the competing agent. In this way allylmagnesium bromide is estimated to react with acetone, benzophenone, benzaldehyde, and diethylacetaldehyde ca. 1.5 × 10⁵ times faster than does butylmagnesium bromide. The rates found for the four substrates do not differ significantly, and it seems possible that there is a ceiling over the rate of reaction of this reagent, for example, caused by diffusion control. This may explain that competition kinetics using allylmagnesium bromide have failed to show kinetic isotope effects or effects of polar substituents with isotopically or otherwise substituted benzophenones. A recently reported α-deuterium secondary kinetic isotope effect for the reaction of benzaldehyde with allylmagnesium bromide was observed at -78°C, but was absent at room temperature. It is suggested that the reaction of benzophenone and benzaldehyde with allylmagnesium bromide has a radical-concerted mechanism since no radical-type products are produced and since no color from an intermediate ketyl is observed even at -78°C.

Full text of DOI:10.1021/jo991777h

1188
J . Org. Chem. 2000, 65, 1188-1192
Use of Com p etition Kin etics w ith F a st Rea ction s of Gr ign a r d
Rea gen ts
Torkil Holm
Department of Organic Chemistry, Technical University of Denmark, Building 201,
DK-2800 Lyngby, Denmark
Received November 16, 1999
Competition kinetics are useful for estimation of the reactivities of Grignard reagents if the reaction
rates do not differ widely and if exact rates are not needed. If the rate of mixing is slower than the
rate of reaction, the ratios between the rates of fast and slow reagents are found to be too small.
This is concluded from experiments in which results obtained by competition kinetics are compared
with results obtained directly by flow stream procedures. A clearer picture of the reactivity ratios
is obtained when the highly reactive reagent is highly diluted with its competitor. A fast reagent
may account for almost all the product even when present as only 1 part in 100 parts of the
competing agent. In this way allylmagnesium bromide is estimated to react with acetone,
benzophenone, benzaldehyde, and diethylacetaldehyde ca. 1.5 × 105 times faster than does
butylmagnesium bromide. The rates found for the four substrates do not differ significantly, and
it seems possible that there is a ceiling over the rate of reaction of this reagent, for example, caused
by diffusion control. This may explain that competition kinetics using allylmagnesium bromide
have failed to show kinetic isotope effects or effects of polar substituents with isotopically or
otherwise substituted benzophenones. A recently reported R-deuterium secondary kinetic isotope
effect for the reaction of benzaldehyde with allylmagnesium bromide was observed at -78 °C, but
was absent at room temperature. It is suggested that the reaction of benzophenone and benzaldehyde
with allylmagnesium bromide has a radical-concerted mechanism since no radical-type products
are produced and since no color from an intermediate ketyl is observed even at -78 °C.
In tr od u ction
the less reactive reagent the chance to get more than its
fair share of the substrate.5
Fast Grignard reactions usually have been measured
by flow methods of various types.1,2 This is convenient if
the half-life for the reaction is more than 5 ms. Reactions
10 or 20 times faster may still be measured, but flow
methods are then time-consuming, require delicate ap-
paratus, and are inaccurate. It would be convenient to
find reaction rates for fast reagents by comparison with
well-known rates of less reactive reagents. As shown by
Felkin and Frajerman,3 this is, however, not a straight-
forward procedure. When they allowed allylic Grignard
reagents to react with acetone in competition with
propylmagnesium bromide, the results were not consis-
tent. The reason was that reactions of the allylic reagents
were much faster than the mixing process, so that the
reactions were mixing controlled.
Felkin improved his measurements by adding the
acetone as vapor, highly diluted with nitrogen. While the
relative reactivities for allyl- and propylmagnesium
bromide were found by simple mixing (dropwise, mag-
nestic stirring) to be 7:1, this ratio changed to 700:1 by
using the slow addition. Felkin also compared the reac-
tivities of allyl-, crotyl-, and 2-pentenylmagnesium bro-
mide toward both acetone and methyloxirane and since
he did not obtain results which were mutually consistent,
he knew that the results were improved, but not correct.
Mixing control corrupts the results of competition
kinetics when a very fast reaction is compared with a
slow reaction. In this case the rate ratio found is much
lower than the true ratio. Dilution of the substrate, as
in Felkin’s case, improves the results and indicates
higher rate ratios, but the extreme reactivity of allyl-
magnesium bromide is seen only when it competes with
other fast Grignard reagents. Diluted 1:140 with benzyl-
magnesium bromide, allylmagnesium bromide still com-
sumes 97% of the test substrate benzophenone (see
below).
When competition kinetics were first introduced in
1926, Francis4 explained that the method is unreliable
when very fast reagents are used. The ratio between the
products from the two competing reactions tends to be
statistically controlled, that is, they are determined by
the concentrations of the competing reagents, rather than
kinetically controlled by the reactivity of the competing
reagents. The reason for this is that when the two
solutions obtain contact in the initial phase of mixing,
the highly reactive reagent is depleted locally. This gives
The purpose of the present work has been to check the
reliability of competition kinetics with Grignard reagents
by comparing rate measurements made by competition
methods with reaction rates already measured by flow
stream procedures. The investigation was thought to be
important since competition kinetics has been widely
(1) Holm, T. Acta Chem Scand. 1967, 19, 2753.
(2) Holm, A.; Holm, T.; Huge-J ensen, E. Acta Chem Scand. 1974,
19, 781.
(3) Felkin, H.; Frajerman, C. Tetrahedron Lett. 1970, 13, 1045.
(4) Francis, A. W. J . Am. Chem. Soc. 1926, 48, 655.
(5) Tolgyesi, W. S. Can. J . Chem. 1965, 43, 3, 343.
10.1021/jo991777h CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/27/2000

Use of Competition Kinetics with Grignard Reagents
J . Org. Chem., Vol. 65, No. 4, 2000 1189
used for mechanistic purposes, for example, for obtaining
kinetic isotope effects.6-8
substrates. In a few cases rates have been measured for
reactions with half-lives as short as 40 µs.11 In such
experiments the reacting solutions were forced into the
mixing “T” under high pressure (5-7 bar), producing a
liquid speed in the 0.22 mm reaction tube of 14 000 mm
s-1. Mixing took place in two phases, the first of which
was the production of an emulsion and the second of
which was the homogenization of the emulsion by diffu-
sion. If the liquids had different refractive indexes, the
emulsion phase could be observed in the glass mixing
chamber as an opalescent cone which in the described
experiment had a length of 0.5 mm corresponding to a
mixing time of 35 µs. The very fast mixing is possible
because of the low viscosity of ether; for aqueus solutions
1000 µs is considered very fast mixing.12
Mixing in the flow stream mixing chamber under these
conditions is very much faster than, for example, mixing
by pouring one solution into another. Since the separa-
ration of the mixing phase and the reaction phase is
required if the true reaction rates are to be obtained, the
high-speed experiments are a valuable “answer book”
when the results from competition kinetics are to be
evaluated. This is true even if the high-speed results are
very crude, being based on a single reading of the
temperature increase after a reaction time of 80 µs.
Discu ssion
Na tu r e of th e Gr ign a r d Rea gen t. Grignard reagents
are complex mixtures of the Schlenk components dialkyl-
magnesium, alkylmagnesium halide, and magnesium
halide, which are solvated with ether and are in a
fluxional state in which the magnesium ligands, alkyl
groups, halide ions, and solvent molecules are exchanged
in very fast equilibria.9,10 It is obvious that the simplest
competition experiment would be to add a very low
concentration of a substrate to a large excess of two
competing Grignard reagents. This procedure may be
called an SD (substrate deficiency) experiment. Using SD,
the constitution of the Grignard reagents will be very
little influenced by the presence of the substrate and the
reaction products. It is unavoidable, however, that the
two Grignard reagents exchange ligands and, for ex-
ample, form the mixed dialkylmagnesium. For this
reason, exact results should not be expected and reaction
rates determined by competition are bound to differ from
rates determined directly.
When the reactivities of two substrates toward the
same Grignard reagent are to be compared, a small
concentration of the Grignard reagent is added to an
excess of the competing substrates, a GD (Grignard
deficiency) experiment. The constitution of the Grignard
reagent is, however, changed when it is added to an
excess of a typical substrate (ketone, ester, etc.), because
of the fast coordination equilibria. Alkoxides formed in
the reaction likewise tie up active alkyl to give complexes
of low reactivity. The results therefore do not pertain to
the reactivity of the Grignard reagents as such, and there
may even be a change of mechanism when going from
SD to GD conditions.
An alternative way to obtain competition between two
substrates is to react a highly diluted solution containing
both the competing substrates with a normal concentra-
tion of the Grignard reagent for a short time, for example,
by quenching the reaction mixture, a TD (time deficiency)
experiment.The carbonyl carbon kinetic isotope effect in
benzophenone reacting with tert-butylmagnesium chlo-
ride has been determined by both methods, and while
no effect was observed6 when using GD, a significant
effect was found7 using TD.
Competition kinetics with highly reactive reagents as
mentioned above require the use of very low concentra-
tions of the substrate. A limit to the degree of dilution
is, however, set by the possibility of measuring the
product distribution. The reaction mixture has to be
concentrated and the various and inevitable impurities
in the Grignard reagent tend to dominate in the GC
analysis.
Resu lts
SD Exp er im en ts. It has been found in the experi-
ments reported here that if moderately reactive reactants
and dilute substrate solutions are used, reaction rates
found by competition kinetics correspond fairly well with
rates obtained by flow stream methods, and the results
are to some extent mutually consistent. With higly
reactive Grignard reagents such as allylmagnesium
bromide, or with highly reactive substrates such as R,â-
unsaturated ketones, the rate ratios between the fast and
the slow reagents found in this way are much smaller
than the true values. High dilution of the test reagent
improves the results, but the rate ratios measured are
still much too low.
In Table 1 are shown rate data for benzophenone
reacting with a series of Grignard reagents. The relative
rate for each reagent was found in competition with
benzylmagnesium bromide, but the table, for conven-
ience, gives the rates relative to the rate for methylmag-
nesium bromide.
In Table 2 are shown the ratios between the reaction
rates of methylmagnesium bromide and some alkylmag-
nesium bromides as calculated from the product ratios
when the two Grignard reagents in equal concentrations
are competing for benzalacetophenone. It is obvious that
the reactivities of the fast reagents have had an “unfair”
competition and show too low reactivity. All though the
order of increasing reactivity is the same in the two
columns, experiments with other combinations of the
reagents gave results which were not consistent. The rate
ratio in the direct competition of isopropylmagnesium
bromide and ethylmagnesium bromide was found, for
example, to be 3.90, while the values in Table 2 would
indicate a ratio of 470/240 ) 1.96. Direct competition
between phenyl- and ethylmagnesium bromide gave a
ratio of ethyl to phenyl of 6.9 against 7.3 from the values
in Table 2.
Mixin g P r ocess. Flow stream reactivity measure-
ments have been performed for the reaction of many
Grignard reagents in the reaction with a variety of
(6) Yamataka, H.; Matsuyama, T.; Hanafusa, T. J . Am. Chem. Soc.
1989, 111, 4912 and references cited herein.
(7) Holm, T. J . Am. Chem. Soc. 1993, 115, 916.
(8) Yamataka, H.; Matsuyama, T.; Hanafusa, T. Chem. Lett. 1987,
647.
(9) Handbook of Grignard Reagents; Silverman, G. S., Rakita, P.
E., Eds.; Marcel Decker Inc.: New York, 1996.
(10) Grignard Reagents, New Developments; Richey, H., Ed.;
Wiley: New York, 1999.
(11) Holm, T. Acta Chem. Scand. 1991, 45, 925.
(12) Caldin, E. F. Fast Reactions in Solution. Blackwell Scientific
Publications: Oxford 1964. Page 37.

1190 J . Org. Chem., Vol. 65, No. 4, 2000
Holm
Ta ble 1. Rea ction Ra tes for Rea ction of RMgBr w ith Ben zop h en on e Rela tive to th e Ra te of Rea ction of
Meth ylm a gn esiu m Br om id ea
kRMgBr/kMeMgBr
kRMgBr/kMeMgBr
competition flow stream
competition
flow stream
kinetics
RMgBr
kinetics
RMgBr
kinetics
kinetics
CH3MgBr
C6H5MgBr
C3H7MgBr
C2H5MgBr
1
2
6.1
8.5
1
1
11
24
i-C3H7MgBr
t-C4H9MgBr
C6H5CH2MgBr
58
100
98
70
89
303
a
Based on either competition experiments between 0.5 M RMgBr and 0.5 M C6H5CH2MgBr reacting with benzophenone (10-3 M) or
reported flow stream rate measurements.11 The pseudo-first-order rate constants for the reaction of benzophenone with 0.5 M
methylmagnesium bromide and 0.5 M benzylmagnesium bromide in ether are 0.3 and 91 s-1, respectively.
Ta ble 2. Rea ction Ra tes for Rea ction of RMgBr (0.25 M) w ith Ben za la cetop h en on e (2 × 10-4 M) Rela tive to th e Ra te of
Rea ction of Meth ylm a gn esiu m Br om id e (0.25 M)a
kRMgBr/kMeMgBr
kRMgBr/kMeMgBr
competition
kinetics
flow stream
kinetics
competition
kinetics
flow stream
kinetics
RMgBr
RMgBr
CH3MgBr
C6H5MgBr
t-C4H9MgBr
1
33
34
1
55
89
C3H7MgBr
C2H5MgBr
i-C3H7MgBr
175
240
470
550
1050
1400
a
Based on either (1) competition experiments between RMgBr and CH3MgBr (1:1) or (2) reported11 flow stream rate measurements.
The pseudo-first-order reaction constant for benzalacetophenone reacting with methylmagnesium bromide is 20 s-1
.
Ta ble 3. Ra tios betw een Ra tes of Rea ction of 0.02 M ter t-bu tylm a gn esiu m Ch lor id e w ith a solu tion of ben zop h en on es
0.2 M in ea ch a
run
ketone 1
Ph2CO
Ph2CO
Ph2CO
ketone 2
3-fluoro-Ph2CO
4,4′-dimethyl-Ph2CO
3-methoxy-4′-chloro-Ph2CO
3-methoxy-4′-chloro-Ph2CO
competition
flow stream
1
2
3
4
1:4
2.7.1
1:4
1:9
8.0:1
1:3.8
1:31
4,4′-dimethyl-Ph2CO
1:6
a
Method: GC measurements of the 1,2-addition products.
Ta ble 4. P r od u ct Distr ibu tion in by th e Rea ction of a n Alk ylm a gn esiu m Br om id e 10-3 M w ith 2 Com p etin g Su bstr a tes
in Eth er ea l Solu tion s, 0.5 M Ea ch
CH3MgBr
C2H5MgBr
C6H5MgBr
CH2dCHCH2Br
PhCHO/PhCOPh
> 3000:1
1.9:1
1.95:1
2:1
Et2CHCHO/PhCOPh
CH3CO CH3/PhCOPh
PhCHO/PhCHdCHCOPh
4:1
1.5:1
More consistent results seem to be obtainable by taking
the Grignard reagents in uneven concentrations. Using
a 1:100 mixture of ethyl/methylmagnesium bromide, the
ethyl adduct still accounts for 10 times as much of the
product as the methyl adduct. Assuming first-order
kinetics with respect to the Grignard reagents, the
relative reactivity is therefore 1000:1, which is close to
the flow stream results.
GD Exp er im en ts. GD may give misleading results
even with rather slow reactions. When a millimolar
solution of phenylmagnesium bromide reacted with a
solution 1 M in acetone and 1 M in benzophenone, the
product composition indicated a ratio between the rate
of the two addition reaction of 4:1, while the rate
constants for the two substrates reacting with phenyl-
magnesium bromide are 43 and 0.3 s-1, a ratio of 143:1.
Rate ratios for the GD reactions of tert-butylmagne-
sium chloride with pairs of substituted benzophenones
are given in Table 3 together with the ratios obtained
from SD flow stream rate measurements. The Hammett
F for the reaction was 3.0 using the flow stream values,
but of the order of 1.5 using the competition experiments.
The much lower Hammett F either may indicate a change
in mechanism when going from SD to GD or may more
likely be due to the leveling effect on the results caused
by mixing control of these rather fast reactions. The
pseudo-first-order rate constant for 0.02 M tert-butyl-
magnesium chloride reacting with 0.25 M benzophenone
in ether is 400 s-1
.
Rea ction s of Allylm a gn esiu m Br om id e. The reac-
tivity of allylmagnesium bromide is extremely high, and
the rates of its reactions with carbonylic substrates may
not be measured by flow stream procedures. Felkin used
competition kinetics and found the relative reactivity of
allylmagnesium bromide and propylmagnesium bromide
toward acetone to be 700:1. In the present work a series
of competition experiments was performed using allyl-
magnesium bromide and benzylmagnesium bromide
competing for very dilute solutions of benzophenone or
acetone. The Grignard reagents were used in ratios of
allyl:benzyl ) 1:1.3, 1:11.6, and 1:128. In all experiments
the allylic addition product dominated and only trace
amounts of the benzylic adduct were found as shown in
Table 5.
It is surprising that even in a dilution of 1:128 the
allyllic product accounts for 97% of the product in both
reactions. The reactivity toward acetone of allylmagne-
sium bromide relative to benzylmagnesium bromide is
in this experiment 0.90/0.07 × 26 ) 3343. The ratio
between allylmagnesium bromide and the 68 times less
reactive propylmagnesium bromide would then be 227 000,
which is 325 times the ratio found by Felkin.3 Toward

Use of Competition Kinetics with Grignard Reagents
J . Org. Chem., Vol. 65, No. 4, 2000 1191
Ta ble 5. Rela tive Rea ctivity of Allylm a gn esiu m Br om id e
a n d Ben zylm a gn esiu m Br om id e tow a r d Ben zop h en on e
a n d Aceton e a t 20 °C a t Va r iou s Con cen tr a tion s
Ta ble 6. Rela tive Ra tes for th e Rea ction s of
Allylm a gn esiu m Br om id e (0.05 M) w ith Ben za ld eh yd es
Deter m in ed by GC on th e Rea ction P r od u ctsa
[C3H5MgBr] [PhCH2MgBr] [Ph2CO] [(CH3)2CO] kallyl/kbenzyl
kPhCHO/kd
kPhCDO/kd -PhCHO
-78 °C
-78 °C
-78 °C
25 °C
25 °C
25 °C
0.992 ( 0.005
0.948 ( 0.005
1.045 ( 0.007
1.004 ( 0.005
1.004 ( 0.005
1.000 ( 0.007
5-PhCHO
5
0.35
0.07
0.007
0.35
0.07
0.007
0.0035
0.088
0.45
0.81
0.90
0.45
0.81
0.90
0.45
0.113
2.5 × 10-4
2.5 × 10-4
2.5 × 10-4
221:1
190:1
30:1
103:1
240:1
26:1
k
k
k
k
PhCHO/kPhCDO
PhCHO/kd -PhCHO
PhCDO/kd -PhCHO
PhCHO/kPhCDO
5
6.9 × 10-4
6.9 × 10-4
6.9 × 10-4
6.9 × 10-4
6.9 × 10-4
5
a
The results are the average of two experiments.
35:1
290:1
tion, reaches a limit for the rate of its reaction. The
limitation could be the rate of diffusion.
benzophenone the ratio kallyl/kbenzyl is 0.90/0.007 × 26x )
3342, and kallyl/kpropyl ) 3342 × 28 ) 9.4 × 104.
Ben za ld eh yd e a n d Allym a gn esiu m Br om id e. A
recent report describes the determination of the second-
ary deuterium kinetic isotope effect (SDKIE) of allylmag-
nesium bromide reacting with R-deuteriobenzaldehyde
using substrate deficiency competition kinetics.14 The
reactions were carried out at -78 °C, and a normal KIE
(kH/kD ) 1.04) was taken as proof of a single electron
transfer (SET) as the rate-determining step. The authors
suggest that the same mechanism is operating at room
temperature and also that this SET mechanism is
operating with benzophenone.
The measurements at -78 °C were reproduced in this
laboratory (kH/kD ) 1.045), but no KIE was observed
when the reaction was performed at 25 °C (kH/kD ) 1.00
Table 6). The last fact may be due either to the apparent
ceiling over the rate of reaction of allylmagnesium
bromide (diffusion control?) or to a change of mechanism
at room temperature. If a polar concerted mechanism,
has a higher energy of activation than an SET mecha-
nism an intersection of the Arrhenius plots is possible
so that SET prevails at - 78 °C and the polar mechanism
at room temperature.
On the other hand, the similarity of the rates of
reaction of allylmagnesium bromide and several rather
unrelated substrates is obvious, Table 4. Measurement
of the relative reactivity of benzaldehyde and 3-fluo-
robenzaldehyde indicated a near-zero Hammett F for the
reaction. Yamataka likewise found a Hammett F near
zero for the reaction with benzophenone with allylmag-
nesium bromide.15 It seems strange that no effect of polar
substituents should be found in an electron-transfer
reaction.
The relative reactivity of allylmagnesium bromide
increases when it is competing in a high dilution as
shown in Table 5. Formally the reaction rates are then
not first-order with respect to the reagent, especially at
concentrations above 0.1 M, when the reaction rates
reach a maximum and no longer increase with the
concentration.
Attempts were made to use a GD approach for compar-
ing a number of pairs of carbonylic compounds (Table 4).
While benzaldehyde was found in competition experi-
ments many thousand times more reactive than ben-
zophenone toward the slowly reacting methylmagnesium
bromide, it appears only slightly more competitive in the
reaction with allylmagnesium bromide. Ethylmagnesium
bromide is of moderate reactivity with ketones, but the
reactions with aldehydes and with R,â-unsaturated ke-
tones are exceedingly fast. The competition between
benzaldehyde and benzalacetophenone for ethylmagne-
sium bromide indicated a reactivity ratio of 1.5:1.
Although allylmagnesium bromide appears to add
somewhat faster to aldehydes and acetone than to
benzophenone in the GD experiment (Table 4), it is
surprising that benzaldehyde, diethylacetaldehyde, and
acetone seem to vary so little in reactivity.
The only Grignard reagent which reacts with benzal-
dehyde with a rate which is slow enough to be measured
by the flow stream procedure is methylmagnesium
bromide. The pseudo-first-order rate constant for 0.02 M
benzaldehyde reacting with 0.5 M methylmagnesium
bromide was found to be about 1000 s-1. Competition
between a mixture of 0.97 M methylmagnesium bromide
Allylmagnesium bromide is known16 to react with
inversion of the triad, and a six-center cyclic concerted
mechanism seems a possibility. If the carbon-magne-
sium bond is homolyzed by magnesium transfer, a tightly
coupled radical pair may form. If the radical pair should
is looked upon as a transition state, the reaction has a
radical concerted mechanism, Figure 1.
The radical concerted mechanism, which is virtually
one step, agrees with the fact that no ketyl coloring is
observed in the reaction of benzophenone and benzalde-
hyde with allylmagnesium chloride. Likewise there are
no radical byproducts. Stepwise ET reactions of tert-
butylmagnesium chloride and other Grignard reagents
with benzophenone or benzaldehyde produce a strong
ketyl color and many byproducts resulting from radical
reactions.
and 0.007 M allylmagnesium bromide for a 5 × 10-4
M
benzaldehyde solution produced the addition products in
the ratio 1:3.9, which indicates a rate for the addition of
allylmagnesium bromide of 1000 × 0.97/0.007 × 3.9 )
ca. 5.4 × 105 s-1. In competition with benzylmagnesium
bromide the rate of addition of benzophenone to allyl-
magnesium bromide was, depending on the relative
concentrations, found to be up to 4.1 × 105 s-1, Table 5.
The numbers indicate that benzophenone and benzalde-
hyde are of similar reactivity toward allylmagnesium
bromide. Competition experiments between allyl- and
benzylmagnesium bromide for highly diluted acetone
indicated a pseudo-first-order reaction constant k1 ) 5
× 105 for the reaction of acetone with 0.25 M allylmag-
nesium bromide.
The extreme reactivity of allylmagnesium bromide and
the apparent similarity of the reaction rates in the
reactions with widely different substrates and the ap-
parent absence of kinetic isotope effects and effects of
polar substituents give the impression that the reagent,
at room temperature and having a sufficient concentra-
(13) Holm, T. Acta Chem. Scand. 1969, 23, 579.
(14) Gajewski, J . J .; Bocian, W.; Harris, N. J . Olson, L. P.; Gajewski,
J . P. J . Am. Chem. Soc. 1999, 121, 326.
(15) Yamataka, H. Chem. Lett. 1988, 1367.
(16) Benkeser, R. A.; Young, W. G.; Broxtyerman, D. A.; J ones, A.
J r.; Peaseczynsky, S. J . J . Am. Chem. Soc. 1969, 91, 132.

1192 J . Org. Chem., Vol. 65, No. 4, 2000
Holm
substrate in dry ether was prepared in a 5 mL disposable
syringe. The syringes were connected by a polyethylene tube,
and the substrate solution was pressed into the Grignard
reagents. The reaction mixture was poured into an ammonium
chloride solution, and the ether layer was separated, washed
with water, and dried over magnesium sulfate. A 10 µL sample
was injected into the gas chromatograph. The exact composi-
tion of the Grignard reagent mixture was found by reacting a
small amount of the mixture with an excess of the substrate
and analyzing by GC after workup.
F igu r e 1.
Gr ign a r d Deficien cy. A mixture of 2 mL of solutions 0.5
M in each of the competing substrates was prepared in a 10
mL disposable syringe. A 5 mL sample of a 10-3 M Grignard
reagent was prepared in a 5 mL disposable syringe. The
syringes were connected, and the Grignard reagent was
pressed into the substrate solution. Workup was done as in
the SD experiment. The exact composition of the substrate
mixture was found by reacting a small amount with an excess
of Grignard reagent and analyzing by GC.
Con clu sion
Competition kinetics may be useful as a semiquanti-
tative method for comparing the relative reactivity of two
Grignard reagents toward a substrate. The rate ratios
found between a fast and a slow reagent tend to appear
to be too low. The reactivities of highly reactive reagents
and substrates may only be demonstrated when the
competition is performed with the reagents in uneven
concentrations. Allylmagnesium bromide at room tem-
perature seems to reach a maximum reaction rate, which
cannot be exceeded and which is almost the same with
many different carbonyl compounds.
KIE for [r-D]Ben za ld eh yd e. Competition was measured
between solutions of d5-benzaldehyde and (1) normal and (2)
[R-D]benzaldehyde, 0.5 M in each, reacting with 0.05 M
solutions of allylmagnesium bromide at (1) -78 °C and (2) 25
°C. Since the normal and the pentadeuterio reaction products
were completely separated in GC, the KIEs could be directly
calculated from the peak areas.
Exp er im en ta l Section
Ga s Ch r om a togr a p h y. A HP 5890 Series II instrument
was used equipped with a standard injector, an FI detector,
and a 100 m × 0.2 mm × 0.33 µm HP-5 capillary column. The
retention times for the various reaction products were observed
using an optimized temperature program, which allowed the
best possible separation of the product peaks from the numer-
ous peaks of byproducts. A HP 3354 integrator yielded the
peak areas, which were normalized according to the effective
carbon numbers (ECN) in the various products.
Ma ter ia ls. Benzyl- and allylmagnesium bromide were
prepared under argon in ether (distilled from benzophenone
ketyl) and sublimed magnesium by the slow (6 h) addition of
distilled benzyl and allyl bromide using agitation. d5-Benzal-
dehyde was obtained from Cambridge Isotope Laboratories.
[R-D]benzaldehyde was prepared according to ref 14.
Com p etition Kin etics. Su bstr a te Deficien cy. A mixture
of 2 mL of solutions 0.5 M in each of the two competing
Grignard reagents was prepared in
a 10 mL disposable
syringe. A 5 mL sample of a 2 × 10-4 M solution of the test
J O991777H