Solvent effects in the Grignard reaction with alkynes

Article abstract of DOI:10.1002/poc.534

Kinetic studies were carded out on the reaction of phenylmagnesium bromide with hex-1-yne in diethyl ether, and in binary mixtures of diethyl ether with toluene, chlorobenzene and dichloromethane. The reaction was accelerated by addition of non-donating solvents. The replacement of a coordinated solvent molecule by the alkyne is necessary for the reaction to proceed, according to density functional theory (DFT) calculations with B3LYP/6-31 + G* method. The non-donating solvents accelerate the reaction by shifting the replacement equilibrium in favour of the complex formation. An analysis in terms of the Koppel-Palm equation revealed a rate decrease with increase in solvent polarity and polarizability. Copyright

Full text of DOI:10.1002/poc.534

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2002; 15: 701–705
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.534
Solvent effects in the Grignard reaction with alkynes
A. Tuulmets,1* V. PaÈllin,1 J. Tammiku-Taul,1 P. Burk2 and K. Raie1
1Institute of Organic and Bioorganic Chemistry, University of Tartu, Tartu 51014, Estonia
2Institute of Chemical Physics, University of Tartu, Tartu 51014, Estonia
Received 5 February 2002; revised 31 March 2002; accepted 23 April 2002
ABSTRACT: Kinetic studies were carried out on the reaction of phenylmagnesium bromide with hex-1-yne in diethyl
ether, and in binary mixtures of diethyl ether with toluene, chlorobenzene and dichloromethane. The reaction was
accelerated by addition of non-donating solvents. The replacement of a coordinated solvent molecule by the alkyne is
necessary for the reaction to proceed, according to density functional theory (DFT) calculations with B3LYP/6–
31 ‡ G* method. The non-donating solvents accelerate the reaction by shifting the replacement equilibrium in favour
of the complex formation. An analysis in terms of the Koppel–Palm equation revealed a rate decrease with increase in
solvent polarity and polarizability. Copyright  2002 John Wiley & Sons, Ltd.
KEYWORDS: Grignard compounds; Grignard reaction; alkynes; kinetics; solvent effects; DFT calculations
INTRODUCTION
reaction of hex-1-yne with phenylmagnesium bromide
was selected as a model process.
Recently, we carried out a kinetic investigation of the
Grignard reaction with acetylene.1,2 In a solution to new
details of the reaction mechanism, we observed large
solvation effects which depend strongly on the steric
requirements of both the Grignard reagent and the
donors. However, the non-specific solvation effects were
not investigated in that work.
In addition to a general interest, the effects of non-
specific solvation can be informative concerning the
reaction mechanism. In an earlier series of studies we
were able to show quantitatively that in the Grignard
addition reaction to ketones the transition states are of
lower polarity than the reagents and, therefore probably
have a cyclic structure.3–6 A change in the polarity of the
medium was achieved by addition of n-heptane or
dichloromethane to the Grignard reagent prepared in a
donor solvent. Linear plots of log k vs Kirkwood function
were obtained for the reactions of pinacolone with
ethylmagnesium and n-propylmagnesium bromides,4
dipropylmagnesium3,5 and diphenylmagnesium.6 The
susceptibility of the reactions to the polarity of the
medium was remarkably dependent on the donating
ability of donors (ethers, amines).6,7
RESULTS AND DISCUSSION
The kinetics of the slow but sufficiently exothermic
reaction were followed thermographically by means of
the initial rates method in the same way as in our previous
work8 (for details, see Experimental). Phenylmagnesium
bromide was prepared in diethyl ether and the required
reagent solutions were obtained by addition of appro-
priate amounts of toluene, chlorobenzene or dichloro-
methane to the initial reagent. The previously used
n-heptane3–6 was not suitable since it caused precipitation
of the reagents. Reaction media of different polarity/
polarizability were generated in this way, while the
specific solvation of the reagents presumably remained
unchanged. The addition of toluene decreased the
polarity and increased the polarizability of the solution,
as can be seen in Table 1, whereas admixtures of
chlorobenzene and dichloromethane caused increases in
both the polarity and polarizability.
The results of the kinetic measurements are presented
in Table 2. The rate constants are mean values for 3–4
parallel runs. The reproducibility of the data is within
In this work, we attempted to apply the same method to
the Grignard reaction of alkynes with the goal of
obtaining an insight into the solvation phenomena. The
Æ5%. The rate constant for the reaction in pure diethyl
À1
ether, 5 Â 10À5
M
sÀ1 at 20°C, is in reasonable
agreement with kII = 2 Â 10À4
M
sÀ1 at 31.5°C deter-
À1
*Correspondence to: A. Tuulmets, Institute of Organic and Bioorganic
Chemistry, University of Tartu, 2 Jakobi Str., Tartu 51014, Estonia.
E-mail: tuulmets@chem.ut.ee
Contract/grant sponsor: Estonian Science Foundation; Contract/grant
number: 4630.
mined by Dessy and Salinger.10 It can be seen from Table
2 that all the added non-donating solvents increase the
reaction rate regardless of their possible polarity/polariz-
ability contribution.
Copyright  2002 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2002; 15: 701–705

702
A. TUULMETS ET AL.
Table 1. Polarity and polarizability parameters of the solvents
diethyl ether:
investigated9
log kII ˆ
Solvent
Dielectric constant, e Refractive index, n
À5:65ꢀÆ0:36† À ꢀ1:02 Æ 0:61†Y ‡ ꢀ9:4 Æ 1:4†P ꢀ4†
R ˆ 0:925; SE ˆ 0:068; n ˆ 12
Diethyl ether
Toluene
Chlorobenzene
Dichloromethane
4.267
2.38
5.69
8.93
1.353
1.496
1.524
1.424
Although the overall correlation is satisfactory, the
polarity and polarizability terms are statistically signifi-
cant at the 95% confidence level. The absolute value of
the susceptibility factor in the polarity term is consider-
ably smaller than those for the reactions with ketones.4–6
However, the negative sign of the constant indicates a
smaller polarity of the transition state in comparison with
the reactants. This seems to be consistent with the four-
centre mechanism postulated by Dessy et al.13 for this
reaction and subsequently supported by evidence of
nucleophilic assistance in the reaction (see, e.g., Refs 1,
14 and 15). In contrast, the preponderant solvent
polarizability term, suggesting a charge separation during
the activation process, is less readily understood in this
context. Therefore, a significant contribution of solvation
equilibria shifts in the last term of Eqn. (4) as an artifact
cannot be excluded.
The solvent polarizability effect was excluded for the
Grignard reaction with ketones on the basis that the
experimental points for both n-heptane and dichloro-
methane additions fit the same log k vs Kirkwood
function correlation.4 For the reaction with alkynes, both
medium effects can be involved. The rate constants were
accordingly analyzed with the Koppel–Palm equa-
tion11,12 [Eqn. (1)] where the terms for constant specific
solvation were omitted:
log k ˆ log k0 ‡ yY ‡ pP
ꢀ1†
According to Ref. 11, the polarity, Y, and polarizability,
P, were calculated as
To obtain a more reliable understanding of the data,
DFT calculations were performed. The main question
was whether an alkyne molecule can directly attack the
organomagnesium compound via an SEi type of reaction
or whether a preliminary replacement of a solvent
molecule is necessary. The species involved in a
Grignard reagent and their coordination complexes with
an alkyne were studied in the gas phase using the DFT
B3LYP/6–31 ‡ G* method. Solvation was modelled
using the supermolecule approach (for details, see
Calculation Methods). We present here only the most
relevant data since an extensive computational investiga-
tion of the reaction is in progress, the results which will
be published elsewhere.
Y ˆ ꢀ" À 1†=ꢀ2" ‡ 1†
P ˆ ꢀn2 À 1†=ꢀn2 ‡ 1†
ꢀ2†
ꢀ3†
The dielectric constants (e) and refractive indexes (n) for
the solvent mixtures were calculated assuming additivity
of the properties (by volume fractions). The Grignard
reagents were considered as solvated with two ether
molecules on average, so in the calculations 2 mol of
ether per mole of reagent were not included in the solvent
composition.
From the data in Table 2, we obtained the following
equation for phenylmagnesium bromide solvated with
In order to save computing time, Grignard reagents
were modeled by MeMgBr, dimethyl ether was taken for
Table 2. Kinetic data for the reaction in various media at 20°C
Concentration (mol lÀ1
PhMgBr Hex-1-yne
)
Molar fraction
of the ether
Initial rate (dC/dt) Â 105 kII Â 105
Solvent
(mol lÀ1 sÀ1
)
(l molÀ1 sÀ1) kII[E] Â 105
Y
P
Et2O
1.0
0.503
0.503
0.503
0.503
0.487
0.473
0.446
0.503
0.487
0.473
0.446
0.503
0.528
0.528
0.528
0.528
0.512
0.496
0.469
0.528
0.512
0.496
0.469
0.528
1.25
1.24
1.42
4.13
1.19
1.24
1.34
2.85
1.20
1.19
1.19
1.85
4.72
4.68
5.38
15.6
4.80
5.28
6.31
10.7
4.82
5.11
5.71
6.46
36.53
33.39
25.42
26.36
35.71
37.70
41.20
18.08
35.86
36.48
37.28
10.92
0.343 0.178
0.338 0.182
0.315 0.198
0.274 0.216
0.345 0.181
0.346 0.183
0.350 0.188
0.372 0.223
0.352 0.180
0.359 0.181
0.372 0.184
0.414 0.200
Et2O ‡ PhMe
Et2O ‡ PhMe
Et2O ‡ PhMe
Et2O ‡ PhCl
Et2O ‡ PhCl
Et2O ‡ PhCl
Et2O ‡ PhCl
Et2O ‡ CH2Cl2
Et2O ‡ CH2Cl2
Et2O ‡ CH2Cl2
Et2O ‡ CH2Cl2
0.92
0.62
0.23
0.96
0.92
0.84
0.22
0.94
0.88
0.77
0.15
Copyright  2002 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2002; 15: 701–705

SOLVENT EFFECTS IN GRIGNARD ALKYNE REACTION
703
Table 3. Complexation energies (DE), enthalpies (DH) and
Gibbs free energies (DG) for the reaction Z ‡ HCꢂ
CCH3 → ZÁHCꢂCCH3 calculated at the B3LYP/À61±31 ‡ G*
level of theory [all values are given in kcal mol and are
corrected for basis set superposition error (BSSE)]
Table 4. Estimated Gibbs free energies and equilibrium
constants at 20°C for the complex formation with propyne in
dimethyl ether solution according to Eqns (7a) and (7b)
DG
Complex
(kcal molÀ1
)
K
Z
DE
DH
DG
MgBr2ÁMe2OÁHCꢂCCH3
MeMgBrÁMe2OÁHCꢂCCH3
Me2MgÁMe2OÁHCꢂCCH3
8.68
7.14
4.22
3.3 Â 10À7
4.7 Â 10À6
7.1 Â 10À4
MgBr2
À8.35
À4.88
À5.37
À2.84
À3.92
À1.58
À8.31
À4.38
À5.06
À2.46
À3.66
À1.08
À0.25
4.50
2.81
6.82
4.45
7.07
MgBr2ÁMe2O
MeMgBr
MeMgBrÁMe2O
Me2Mg
Me2MgÁMe2O
equilibria
Z Á 2E „ ZE ‡ E
A ‡ ZE „ ZEA
ꢀ7a†
ꢀ7b†
the ether and propyne for alkynes. The gas-phase data are
reasonably applicable to poorly polar ether solutions.
As shown in Table 3, the solvation of the magnesium
compounds causes a substantial decrease in the com-
plexation energies. It should be noted that the DG values
differ considerably from the DE and DH values, thus
indicating a significant contribution of enthropy. The
complexes are weak and DG of their formation is
positive. Complexation of an alkyne with bisolvated
magnesium species appeared to be energetically un-
favourable (see also Fig. 1) and should be ruled out.
As a consequence, the reaction proceeds through the
replacement of a solvent molecule by the alkyne and
subsequent rate-limiting conversion of the complex to
products:
combining the data from Table 3 and the Gibbs free
energies of solvation calculated in our previous work.16
The results are presented in Table 4. When the Lewis
acidity of the species decreases, the replacement of the
ligands becomes easier. However, all the equilibria are
shifted far to the left.
The nature of the phenylmagnesium bromide solution
can also be represented by the Schlenk equilibrium:
2PhMgBr „ Ph2Mg ‡ MgBr2
ꢀ8†
shifted towards PhMgBr in diethyl ether ([PhMgBr]/
[Ph2Mg] ꢁ 8, from Refs 17 and 18). A semiquantitative
treatment of the Lewis acidities of Grignard entities7
indicated somewhat lower acidity of PhMgBr and Ph2Mg
species in comparison with corresponding methyl
compounds. Moreover, diethyl ether, as a weaker Lewis
acid due to the steric demands, should enhance the acidity
of the magnesium center compared with dimethyl ether.
Consequently, the equilibrium constants for the complex
formation in the phenylmagnesium bromide–hex-1-yne
system must be greater to some extent than those in Table
4. Nevertheless, the concentrations of the complexes in
K
Z Á 2E ‡ A „ ZEA ‡ E
ꢀ5a†
ꢀ5b†
k
ZEA À! products
‰Z Á 2EŠ‰AŠ
v ˆ k‰ZEAŠ ˆ kK
ꢀ6†
‰EŠ
where Z denotes a magnesium compound, A the alkyne
and E an ether molecule.
The equilibrium constant K can be estimated from the
Figure 1. The optimized (at the B3LYP/6±31 ‡ G* level of theory) structures of the propyne complexes with (a) unsolvated
methylmagnesium bromide, (b) monosolvated methylmagnesium bromide and (c) disolvated methylmagnesium bromide
Copyright  2002 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2002; 15: 701–705

704
A. TUULMETS ET AL.
the reaction solution remain extremely low. This seems to
be the reason for the rather feeble reactivity of alkynes,
although the rate of product formation from an alkyne–
Grignard complex appears to be unexpectedly large.
The equilibrium (5a) is shifted far to the left and the
equilibrium can be taken into account by multiplying the
observed rate constants by molar concentration of the
free ether:
fast rearrangement of the complex to products, the overall
reaction is slow.
Addition of non-donating solvents accelerates the
reaction, presumably by shifting the replacement equili-
brium in favour of the complex formation. In addition, a
susceptibility of the reaction rate to changes in the
solvent polarity and polarizability was revealed by a
correlation analysis. This points to a smaller polarity of
the transition state in comparison with the reagents, thus
presenting evidence in favour of a cyclic structure of the
transition state.
kK
kII ˆ
ꢀ9†
‰EŠ
Rate constants corrected in this way (Table 2) were
correlated according to Eqn. (1) and the following
equation was obtained:
EXPERIMENTAL
Materials. All the procedures with the reagents and
solutions were carried out under dry argon. The Grignard
reagents in diethyl ether were prepared in the conven-
tional manner. The reagents in binary solutions were
obtained by diluting the Grignard reagents with appro-
piate amounts of the ether and/or of the non-donating
solvent.
log ‰kII‰EŠŠ ˆ
À1:02ꢀÆ0:54† À ꢀ2:6 Æ 0:9†Y À ꢀ8:4 Æ 2:0†P ꢀ10†
R ˆ 0:841; SE ˆ 0:101; n ˆ 12:
The concentration of free ether in the solution can be
only roughly estimated, because the contribution of the
Schlenk equilibrium [Eqn. (8)] and of the association
equilibria of the species involved cannot be taken into
account. This seems to be the reason for the relatively
poor correlation of the data. However, the polarity and
polarizability terms in Eqn. (10) are statistically sig-
nificant and their signs are now coherent. Decelerating
effects of solvent polarity and polarizability indicate a
smaller polarity of the transition state in comparison with
the reagents, thus presenting evidence in favour of a
cyclic structure of the transition state.
The fast reactions of ketones can hence be rationalized
on the basis of extensive complex formation with
Grignard reagents. As the values for the equilibrium
constants are large (for a review, see Ref. 19), with an
excess of one of the reagents the other is almost
completely bound and the changes in the solvent
composition, similar to those in the present work, only
slightly affect the positions of the equilibria. This
explains why good linear log k vs Kirkwood function
plots have been obtained for the reactions of ketones.3–6
Moreover, the refractive indexes of ethers and of non-
donating additives used in those studies differed only
slightly and therefore the contribution of solvent
polarizability remained undetected.
Kinetic measurements. The reaction was carried out in a
glass vessel mantled with foam plastic and placed in a
thermostated housing. The equipment was sealed with a
thermostated lid. The reaction cell was provided with a
mechanical stirrer and a thermistor, which was connected
through a bridge circuit to a recording potentiometer.
All parts of the equipment and the reagents were
thermostated. The reaction vessel was purged throughly
with pure argon, 15 ml of the Grignard reagent were
cannulated into the cell and the stirring was started. After
thermal equilibrium, 0.92 ml of hex-l-yne was introduced
and the temperature change of the reaction solution
(usually 0.1–0.2°C) was recorded as a plot of temperature
versus time. Because the system was nearly adiabatic, the
heat exchange with the internal part of the calorimeter
caused only a little heat loss.
The zero-order initial rate constants in mol lÀ1 sÀ1
were obtained dividing the rate constants, determined as
the slope of the tangent to the kinetic curve at the initial
point, by the molar temperature rise of the reaction. The
latter was determined from a separate run with the same
reagent containing about 10 mol% triethylamine. The
rapid and complete reaction of 30 ml of hex-l-yne raised
the temperature by about 0.6°C and provided the total
heat of the reaction under the experimental conditions.
The second-order rate constants were obtained dividing
the initial rate constants by initial concentrations of hex-l-
yne and the Grignard reagents.
CONCLUSIONS
The reaction between an alkyne and Grignard reagent
involves two consecutive steps, the first consisting in the
replacement of a coordinated solvent molecule by the
alkyne, followed by a unimolecular reaction of the
complex. The complex formation equilibrium is shifted
far towards the initial reagents and therefore, despite a
CALCULATION METHODS
All calculations were carried out using the Gaussian 98
program package.20
Copyright  2002 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2002; 15: 701–705

SOLVENT EFFECTS IN GRIGNARD ALKYNE REACTION
705
9. Handbook of Chemistry and Physics (79th edn) Lide DR (ed). CRC
Press: Boca Raton, FL, 1998–99.
10. Dessy RE, Salinger RM. J. Org. Chem. 1961; 26: 3519–3520.
11. Koppel IA, Palm VA. In Advances in Linear Free Energy
Relationships, Chapman NB, Shorter J (eds). Plenum Press: New
York, 1972; chapt. 5.
12. Reichardt C. Solvents and Solvent Effects in Organic Chemistry.
VCH: Weinheim, 1988.
13. Dessy RE, Wotiz JH, Hollingsworth CC. J. Am. Chem. Soc. 1957;
79: 358–362.
14. Dessy RE, Paulik E. J. Am. Chem. Soc. 1963; 83: 1812–1819.
15. Abraham MH, Hill JA. J. Organomet. Chem. 1967; 7: 11–21.
16. Tammiku J, Burk P, Tuulmets A. J. Phys. Chem. A 2001; 105:
8554–8561.
Optimizations and vibrational analysis were done
using DFT with hybrid B3LYP functional and the
6–31 ‡ G* basis set. All stationary points were found
to be true minima (number of imaginary frequencies,
NImag = 0). The calculated frequencies were also used
for calculations of enthalpies and Gibbs free energies.
The stability of the complexes was calculated considering
the basis set superposition error estimated according to
the counterpoise correction method of Boys and Bernar-
di.21
17. Smith MB, Becker WE. Tetrahedron 1966; 22: 3027–3036.
18. Smith MB, Becker WE. Tetrahedron 1967; 23: 4215–4219.
19. Holm T, Crossland I. In Grignard Reagents: New Developments,
Richey HG (ed). Wiley: New York, 2000; chapt. 1.
REFERENCES
20. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA,
Cheeseman JR, Zakrzewski VG, Montgomery Jr. JA, Stratmann
RE, Burant JC, Dapprich S, Millam JM, Daniels AD, Kudin KN,
Strain MC, Farkas O, Tomasi J, Barone V, Cossi M, Cammi R,
Mennucci B, Pomelli C, Adamo C, Clifford S, Ochterski J,
Petersson GA, Ayala PY, Cui Q, Morokuma K, Malick DK,
Rabuck AD, Raghavachari K, Foresman JB, Cioslowski J, Ortiz
JV, Stefanov BB, Liu G, Liashenko A, Piskorz P, Komaromi I,
Gomperts R, Martin RL, Fox DJ, Keith T, Al-Laham MA, Peng
CY, Nanayakkara A, Gonzalez C, Challacombe M, Gill PMW,
Johnson B, Chen W, Wong MW, Andres JL, Gonzalez C, Head-
Gordon M, Replogle ES, Pople JA. Gaussian 98, Revision A7.
Gaussian: Pittsburgh PA, 1998.
1. Pa¨llin V, Otsa E, Tuulmets A. J. Organomet. Chem. 1999; 590:
149–152.
2. Pa¨llin V, Tuulmets A. Main Group Met. Chem. 2000; 23: 179–182.
3. Koppel J, Tuulmets A. Reakts. Sposobn. Org. Soedin. 1970; 7:
911–918.
4. Koppel J, Loit J, Luuk M, Tuulmets A. Reakts. Sposobn. Org.
Soedin. 1971; 8: 1155–1164.
5. Koppel J, Tuulmets A. Reakts. Sposobn. Org. Soedin. 1972; 9:
399–411.
6. Viirlaid S, Kurrikoff S, Tuulmets A. Org. React. (Tartu) 1974; 11:
73–80.
7. Tuulmets A. Org. React. (Tartu) 1974; 11: 81–100.
8. Pa¨llin V, Tuulmets A. J. Organomet. Chem. 1999; 584: 185–189.
21. Boys SF, Bernardi F. Mol. Phys. 1976; 19: 325–330.
Copyright  2002 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2002; 15: 701–705