- Email: [email protected]
New perspectives in the formation of the Grignard reagent
Temzhedron Lmn.
oo4o-4o39(94)0119
New Perspectives
Labormike
I-5
in the Formation
Eric P&ale% Jean-Claude
Vol. 35. No. 32. pp. 5857-5860. t994 Elscvier Science Lfd Printed in Great Britain oLMo-4039/94 $7.al+o.00
of the Grignard
NCgreI+, Michel
Reagent
C&anon
1411). FaculCedes Sciencesde St Jtrtlme universiti d’Aix-Mameiiem, 13013rvkseik (Fnmce)
AM3. asso& au CNRS (URA
Abstractz Inhibitors.in very low comxntratkm,inhibit the tea&m between I-bimno-3-methylderivative,the butaneandmagncaium&tabwJbyvaptxizationofmetal.FasachamagtMom inhibitors~tplaythcroltof”IriUcrs”for~tivcsitckAcbainreactionfuthcformationofthe Grignardreageatisproposed.
The general agreement for intcrmcdky
alkyl radicals’”
during Grignard reagent fortnation (Scheme 1)
is founded on the alkyl group isomriza tion obscrvcd in the reaction of magnesium metal with alkyl halides in diethylether and on CIDIW studics4. RX +b& R’ +MgX’lltuc
is tiisagrccment. however, concunin
whether these radicals =
adsorb4
_c
R’
+M@C’
Scheme
RMgX
1
g the mobility of these radicals during the reaction. The question of
on the mkgnesium surfa&
a dif%c
fi&y
in solution6 has cumntly
been
hiSCUSSC4L
Two classes of mechanisms
fa the Grignard reagent formation are proposed: D-(Diffusion) ‘Iand A-
(Adsorption) mod&. The first mechani~&-~~ follows from a tithcmatical
model based on a kinetic analysis of the product
dish’ibution. This model supposes that “all the radicals leave the surface and diffuse freely in solution”, because it uses the eki&&
kinetic data in the litcratu#
mechanism is based on expaimntal
obtained under homownmus
solution conditions. The scumd
pmduck distPfIn~ti&nsand sten?och&iatiies
&served
during the’famation
ofGiignard~t.Thismoddsrtpposcs~the~~~ir;~~~~the~~ofmagnwium to explain why the radicals generat lzl’.
fmm optically active al&l halides partially maintain their wnfiguratio&
A strongest support for the A-model comes from the using of a pcrdcutcrated ether solvent or a radical
trap dcutcratcd dicyclohexylpbosphine.
In all cases, only a small pacentagc
Inagncsium bo yield the ckumt2. 5857
of the radicals leave the surface of
5858
We now wish to report further results of our study designed to suggest steps which will account for the formation of RMgX from the radical R. These steps are in agreement with the A-model. However, once adsorbed, these radicals will produce an intermediate RMg(I) that could be a better reducing agent than the mclgnesiwn su&ace. Therefore a radicai chain process isproposed in place of the stoichiometric one described in model A. This mechanism is based on the following evidence: l
The reaction between organic halides and magnesium often shows an induction period which can be shortened by the use of “activator@. This period ofinitiation18*1g is characterized by the formation of isolated corrosion pits while the solution become turbid. When the reaction proceeds, the turbidity disappears
l
and the corrosion pits grow in size. This observation
might suggest a step of chain
propagation. A low concentration of inhibitors inhibits the Grignard reagent formation’.
The aim of our work is to csrry out the Grignard reaction with catalytic quantities of inhibitotx. Indeed, the radical chain inhibition can be used as a diagnostic test for reactions which involve radicals or radicalanions20. We used an active magnesium obtained by vaporization of metal in a rotary metal atom reactogl. at -110 “C in THF. This method allows the formation of clean, alkali halide free, and extremely reactive magnesium. This clean magnesium excludes the presence of MgO which could poison the active sites and the important number of active sites excludes the possibility that small quantities of inhibitors might inhibit all the sites. The first piece of evidence (Table 1) consists in a test of reactivity of I-bromo-3-methylbutane
and
magnesium. Experiments give identical yields (NH%) of 2-methylbutane (RH). Table 1. Reactivities of 1-Bromo-3-methylbutane
and Mga
a Magnesiumactivatedby vacuixatton; b addirion+ stimng + hydrolysis cW&tenninatedhyCGwithintanaiatanclatd.
The reactions l-3 are carried out under the same experimental conditions as the one later used for the reaction with the inhibitors. A complete reaction occurs, whatever the temperature of experiment. Reaction 4 is csrried out with a minimum time of addition of the allcyl halide and quenched immediately after this addition. The reaction is also complete. These results permit to exclude the hypothesis that a poor reactivity of akyl halide aud magnesium could be the cause of the later studies concerning the inhibition of Grigmud -gent formation. Table 2 sum~zes the effect of inhibitors on the reaction between the activated magnesium and lbromo-3methylbutane. The reactions were carried out by star&d Schlenk techniques, in THE, at -80” C and under purified Ar. The values gathered in Table 2 correspond to a threshold determined by 5-6 different experiments with adding decmas ing quantities of inhibitors to the reactive magnesium-alkyl halide mixture. The
5859
consequences
of adding these inhibitors are saiking. Indeed, in all cases, the inhibition of the Grignard reagent
formation is shown by the total absence of consumption ofthe alkyl halide.
Table 2. The Effect of Inhibitors on the Reaction between RDraandMgb
a 1-B~nnc+3-methylta~aae; h Msgmsium activatedby vaporization c %MeIJlyltlutane.
The m-dinitmbenzene
concentration
is l/100 in comparison with the concentration
famed electron transfer induced substitution Tanner2*.
Recently,
bromocamphor
this author proposed
with amines 23, following
reaction. A concentration an electron
used by Komblumn
in the
25 times more important was used by
transfer chain process
for the reduction
the inhibition of the reaction with 4% of p-dinitrobenzene.
of aThough
these inhibitors are present in very low concentration, they are able to suppress the radical or/ and radical-anion and the magnesium is probably a chain process.
reaction. The reaction between the 1-bmmo-3-methylbutane
The metallic surface would play the role of a genemlized base or nucleophile2A. This process could permit a generalization of mechanisms
worked out by Kornbhnn2~~-27 and Russellzs.
Indeed, in 1964. Kornblum proposed (Scheme 2) that carbon alkylation was a radical-anion
process for
the reaction between the lithium salt of 2-nitropropane and the nitrobenzyl hahdeP. A
D
RCHzX
+ Nu- --_)
[RcH2xl-
RCHa’+ Nu’ -r) The nucleophile
iRCH$l’
RCHsNu
(the anion derived from Znitropropane)
halide, the role of electron acceptor A. This mechanism reagent formation (Scheme
+Nu’
RCHz’+ X-
-
Scheme
2
plays the role of electron donor D whereas the alkyl parallels the one generally accepted for the Grignard
1).It was shown in 1966 to be incomplete and a chain reaction was introduced to
explain all the observed facts”
(Scheme A
3).
D
RCHtX + Nu- ;
CRCHzXj- +Nu’ Scheme
RCH2’ + Nu- -, IRCHaNul -
IRCBzNW
+ RCHaX -
RCHaNu + lRCHzXIT
3
RX +I@
-
R’+Mgox
R’+h@RMgo+RX
We proposed
comparison
-
R’ +X-
earlier such a chain mechanism,
with the sc+me
prefer this chain mechanism
Scheme
RM8(l)
4
+fRMBo]+
voi@ut
proposed by Hush and 0m2?
wt
expairaqmal~eyidence.
for tW a&$
to the one proposed on the baskof-kinetic
in a review”
by
mercuric halides formation. We
studies by H&O,
but we have no
compelling evidefKx to discard the latter.
References:
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Kharash. M. S.: Rcinnu~tb. 0. Gr&ymrfReuctions of Non-Met&c .WMmce~ PrmticaHall: New-York.1954. pp l-91. GoMu&!. w Bxhmam~. W. 6. J. Am. Ckm. Sot. 1927.49,236-257i7. Brycc-Smith, D.; Ch. G. F. /. Chem. Sot. 1958. 105@1053. Badewia H. W. I-I.J.; Blombag, C.: Bidtdhaqvt, Pi Teb’okh L&r. &975.24,2003-2006 and dcxences cited tkcin. Hamdouchi. C.; Topolski, ltd.; Godken. V.: Whirsky. H. Ih. J: Org. Chem. 1993.58.3148-3155. Garst, J. F.: Mgv&y. F.: Bath. R.; Lad. K. 6. J. AWL 0kwt. Sot. 1991,113. 5392-5397: Garst, J. F. Act. Chum. Res. 1991.21. 95-97. Garsi J. F.: Swift. B. L. 3. Am. &em: Sot. 19119. Iii. 241-250. Root. K. s.; Hill. c. L.: Lawralce. L. hf.: White&Jes. G. M. J. Am. Chem. Sot. 1989.111. 5405-5412. Wall&. C. Acc.~Chem~ Res. 199i. 24. 255-256. Griller, D.; Ingold, K. U. Act. Chem. Res. 1980,p3. 317-323. waRlo&y, H. rd.: 2 immumann, C. J. Am. Ch.em. Sot. 1992.114, 4996-5000. W&or&y, H. hi. Act. C’hcm. Res. 1990.23. 286-293. Foe-in. P.; Pommier, J.-C. J. Organome tal. Chats. M7B. 150. 187-201. Walbcuaky, H. M.; Banks. B. Bull. Sot. Chim. Belg. 1980,89. 849-868. Dubois, J.-E.: Molle, G.; Towilh, G.; Baw. P. Tefruhdon Left. 1979,52,506%5072. Lag. L. M.: Wtlitesi6es. G. M. J. Am. chew sac. 19da. 102.‘2493-2494. Hill. C. L.; V& Smdc, J. E.; Whitddes, 0. M. 1. Org. Chem. 1980.45, lm1028. ICoon, S. E.; Gyler, C. E.; Hill, J. H M.; Bowyer, W. J. J. Or . Chm. 1993.58, 3225-3226. Tanner, D. D.; B&skbm~. E. V.: Diaz, 0. E; J. Am. Ckm. Sf .198l.t03. 1557-1559. Ntgrel, J.-C.: Gony. M; Chanon, hf.: Lai. R. Itwrg Ctim Acta l99~.207,59-63 and r&exams citedthckn. Kcxbcz. R. C.; Uny. G. W.; Kcnnblum. N. J. Am. Chem. Sot. lM5.87.45204528. Chcn. J. J.; T-, D. D. Can. J. Ckm. 1992.70. X73-176. Chanm, M. Bull. Sot. Chh. Fr. 1982,7-S. 197-238. Kerbw. R. C.: Umy, G. W.; Komblum.N. J. Am. Chmi. Sot. -1964.86, 3904-3905. Kombk. N.: Mi&l.,R. E.: Kcrber, R. C. J. Am. Chem. See. 1%4.88.566O%fj2. Kombhm. N.; Michcl. R. 8.: Kwbcr, R. C. J. Am. Chun Sot. 1966.88. 56624663. Rus&ll. 01 A.; Danen; W. C. J. Am. &em. Sec. X966.88, 5663-5665. Hush, N. S.; Oldham, K. M. J. 8kctmanai. Chem. 1943,6,34. Hbrak, M.; Palm. V.: Soogenbitr. U. Reakfs. Sposobn. Organ. So&in. 1975.709-719.Idem. Chem.Absrr., 1976, 84, ref. 7333lnJ
(Received in France 26 April 1994; accepted 15 June 1994)
oo4o-4o39(94)0119
New Perspectives
Labormike
I-5
in the Formation
Eric P&ale% Jean-Claude
Vol. 35. No. 32. pp. 5857-5860. t994 Elscvier Science Lfd Printed in Great Britain oLMo-4039/94 $7.al+o.00
of the Grignard
NCgreI+, Michel
Reagent
C&anon
1411). FaculCedes Sciencesde St Jtrtlme universiti d’Aix-Mameiiem, 13013rvkseik (Fnmce)
AM3. asso& au CNRS (URA
Abstractz Inhibitors.in very low comxntratkm,inhibit the tea&m between I-bimno-3-methylderivative,the butaneandmagncaium&tabwJbyvaptxizationofmetal.FasachamagtMom inhibitors~tplaythcroltof”IriUcrs”for~tivcsitckAcbainreactionfuthcformationofthe Grignardreageatisproposed.
The general agreement for intcrmcdky
alkyl radicals’”
during Grignard reagent fortnation (Scheme 1)
is founded on the alkyl group isomriza tion obscrvcd in the reaction of magnesium metal with alkyl halides in diethylether and on CIDIW studics4. RX +b& R’ +MgX’lltuc
is tiisagrccment. however, concunin
whether these radicals =
adsorb4
_c
R’
+M@C’
Scheme
RMgX
1
g the mobility of these radicals during the reaction. The question of
on the mkgnesium surfa&
a dif%c
fi&y
in solution6 has cumntly
been
hiSCUSSC4L
Two classes of mechanisms
fa the Grignard reagent formation are proposed: D-(Diffusion) ‘Iand A-
(Adsorption) mod&. The first mechani~&-~~ follows from a tithcmatical
model based on a kinetic analysis of the product
dish’ibution. This model supposes that “all the radicals leave the surface and diffuse freely in solution”, because it uses the eki&&
kinetic data in the litcratu#
mechanism is based on expaimntal
obtained under homownmus
solution conditions. The scumd
pmduck distPfIn~ti&nsand sten?och&iatiies
&served
during the’famation
ofGiignard~t.Thismoddsrtpposcs~the~~~ir;~~~~the~~ofmagnwium to explain why the radicals generat lzl’.
fmm optically active al&l halides partially maintain their wnfiguratio&
A strongest support for the A-model comes from the using of a pcrdcutcrated ether solvent or a radical
trap dcutcratcd dicyclohexylpbosphine.
In all cases, only a small pacentagc
Inagncsium bo yield the ckumt2. 5857
of the radicals leave the surface of
5858
We now wish to report further results of our study designed to suggest steps which will account for the formation of RMgX from the radical R. These steps are in agreement with the A-model. However, once adsorbed, these radicals will produce an intermediate RMg(I) that could be a better reducing agent than the mclgnesiwn su&ace. Therefore a radicai chain process isproposed in place of the stoichiometric one described in model A. This mechanism is based on the following evidence: l
The reaction between organic halides and magnesium often shows an induction period which can be shortened by the use of “activator@. This period ofinitiation18*1g is characterized by the formation of isolated corrosion pits while the solution become turbid. When the reaction proceeds, the turbidity disappears
l
and the corrosion pits grow in size. This observation
might suggest a step of chain
propagation. A low concentration of inhibitors inhibits the Grignard reagent formation’.
The aim of our work is to csrry out the Grignard reaction with catalytic quantities of inhibitotx. Indeed, the radical chain inhibition can be used as a diagnostic test for reactions which involve radicals or radicalanions20. We used an active magnesium obtained by vaporization of metal in a rotary metal atom reactogl. at -110 “C in THF. This method allows the formation of clean, alkali halide free, and extremely reactive magnesium. This clean magnesium excludes the presence of MgO which could poison the active sites and the important number of active sites excludes the possibility that small quantities of inhibitors might inhibit all the sites. The first piece of evidence (Table 1) consists in a test of reactivity of I-bromo-3-methylbutane
and
magnesium. Experiments give identical yields (NH%) of 2-methylbutane (RH). Table 1. Reactivities of 1-Bromo-3-methylbutane
and Mga
a Magnesiumactivatedby vacuixatton; b addirion+ stimng + hydrolysis cW&tenninatedhyCGwithintanaiatanclatd.
The reactions l-3 are carried out under the same experimental conditions as the one later used for the reaction with the inhibitors. A complete reaction occurs, whatever the temperature of experiment. Reaction 4 is csrried out with a minimum time of addition of the allcyl halide and quenched immediately after this addition. The reaction is also complete. These results permit to exclude the hypothesis that a poor reactivity of akyl halide aud magnesium could be the cause of the later studies concerning the inhibition of Grigmud -gent formation. Table 2 sum~zes the effect of inhibitors on the reaction between the activated magnesium and lbromo-3methylbutane. The reactions were carried out by star&d Schlenk techniques, in THE, at -80” C and under purified Ar. The values gathered in Table 2 correspond to a threshold determined by 5-6 different experiments with adding decmas ing quantities of inhibitors to the reactive magnesium-alkyl halide mixture. The
5859
consequences
of adding these inhibitors are saiking. Indeed, in all cases, the inhibition of the Grignard reagent
formation is shown by the total absence of consumption ofthe alkyl halide.
Table 2. The Effect of Inhibitors on the Reaction between RDraandMgb
a 1-B~nnc+3-methylta~aae; h Msgmsium activatedby vaporization c %MeIJlyltlutane.
The m-dinitmbenzene
concentration
is l/100 in comparison with the concentration
famed electron transfer induced substitution Tanner2*.
Recently,
bromocamphor
this author proposed
with amines 23, following
reaction. A concentration an electron
used by Komblumn
in the
25 times more important was used by
transfer chain process
for the reduction
the inhibition of the reaction with 4% of p-dinitrobenzene.
of aThough
these inhibitors are present in very low concentration, they are able to suppress the radical or/ and radical-anion and the magnesium is probably a chain process.
reaction. The reaction between the 1-bmmo-3-methylbutane
The metallic surface would play the role of a genemlized base or nucleophile2A. This process could permit a generalization of mechanisms
worked out by Kornbhnn2~~-27 and Russellzs.
Indeed, in 1964. Kornblum proposed (Scheme 2) that carbon alkylation was a radical-anion
process for
the reaction between the lithium salt of 2-nitropropane and the nitrobenzyl hahdeP. A
D
RCHzX
+ Nu- --_)
[RcH2xl-
RCHa’+ Nu’ -r) The nucleophile
iRCH$l’
RCHsNu
(the anion derived from Znitropropane)
halide, the role of electron acceptor A. This mechanism reagent formation (Scheme
+Nu’
RCHz’+ X-
-
Scheme
2
plays the role of electron donor D whereas the alkyl parallels the one generally accepted for the Grignard
1).It was shown in 1966 to be incomplete and a chain reaction was introduced to
explain all the observed facts”
(Scheme A
3).
D
RCHtX + Nu- ;
CRCHzXj- +Nu’ Scheme
RCH2’ + Nu- -, IRCHaNul -
IRCBzNW
+ RCHaX -
RCHaNu + lRCHzXIT
3
RX +I@
-
R’+Mgox
R’+h@RMgo+RX
We proposed
comparison
-
R’ +X-
earlier such a chain mechanism,
with the sc+me
prefer this chain mechanism
Scheme
RM8(l)
4
+fRMBo]+
voi@ut
proposed by Hush and 0m2?
wt
expairaqmal~eyidence.
for tW a&$
to the one proposed on the baskof-kinetic
in a review”
by
mercuric halides formation. We
studies by H&O,
but we have no
compelling evidefKx to discard the latter.
References:
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Kharash. M. S.: Rcinnu~tb. 0. Gr&ymrfReuctions of Non-Met&c .WMmce~ PrmticaHall: New-York.1954. pp l-91. GoMu&!. w Bxhmam~. W. 6. J. Am. Ckm. Sot. 1927.49,236-257i7. Brycc-Smith, D.; Ch. G. F. /. Chem. Sot. 1958. 105@1053. Badewia H. W. I-I.J.; Blombag, C.: Bidtdhaqvt, Pi Teb’okh L&r. &975.24,2003-2006 and dcxences cited tkcin. Hamdouchi. C.; Topolski, ltd.; Godken. V.: Whirsky. H. Ih. J: Org. Chem. 1993.58.3148-3155. Garst, J. F.: Mgv&y. F.: Bath. R.; Lad. K. 6. J. AWL 0kwt. Sot. 1991,113. 5392-5397: Garst, J. F. Act. Chum. Res. 1991.21. 95-97. Garsi J. F.: Swift. B. L. 3. Am. &em: Sot. 19119. Iii. 241-250. Root. K. s.; Hill. c. L.: Lawralce. L. hf.: White&Jes. G. M. J. Am. Chem. Sot. 1989.111. 5405-5412. Wall&. C. Acc.~Chem~ Res. 199i. 24. 255-256. Griller, D.; Ingold, K. U. Act. Chem. Res. 1980,p3. 317-323. waRlo&y, H. rd.: 2 immumann, C. J. Am. Ch.em. Sot. 1992.114, 4996-5000. W&or&y, H. hi. Act. C’hcm. Res. 1990.23. 286-293. Foe-in. P.; Pommier, J.-C. J. Organome tal. Chats. M7B. 150. 187-201. Walbcuaky, H. M.; Banks. B. Bull. Sot. Chim. Belg. 1980,89. 849-868. Dubois, J.-E.: Molle, G.; Towilh, G.; Baw. P. Tefruhdon Left. 1979,52,506%5072. Lag. L. M.: Wtlitesi6es. G. M. J. Am. chew sac. 19da. 102.‘2493-2494. Hill. C. L.; V& Smdc, J. E.; Whitddes, 0. M. 1. Org. Chem. 1980.45, lm1028. ICoon, S. E.; Gyler, C. E.; Hill, J. H M.; Bowyer, W. J. J. Or . Chm. 1993.58, 3225-3226. Tanner, D. D.; B&skbm~. E. V.: Diaz, 0. E; J. Am. Ckm. Sf .198l.t03. 1557-1559. Ntgrel, J.-C.: Gony. M; Chanon, hf.: Lai. R. Itwrg Ctim Acta l99~.207,59-63 and r&exams citedthckn. Kcxbcz. R. C.; Uny. G. W.; Kcnnblum. N. J. Am. Chem. Sot. lM5.87.45204528. Chcn. J. J.; T-, D. D. Can. J. Ckm. 1992.70. X73-176. Chanm, M. Bull. Sot. Chh. Fr. 1982,7-S. 197-238. Kerbw. R. C.: Umy, G. W.; Komblum.N. J. Am. Chmi. Sot. -1964.86, 3904-3905. Kombk. N.: Mi&l.,R. E.: Kcrber, R. C. J. Am. Chem. See. 1%4.88.566O%fj2. Kombhm. N.; Michcl. R. 8.: Kwbcr, R. C. J. Am. Chun Sot. 1966.88. 56624663. Rus&ll. 01 A.; Danen; W. C. J. Am. &em. Sec. X966.88, 5663-5665. Hush, N. S.; Oldham, K. M. J. 8kctmanai. Chem. 1943,6,34. Hbrak, M.; Palm. V.: Soogenbitr. U. Reakfs. Sposobn. Organ. So&in. 1975.709-719.Idem. Chem.Absrr., 1976, 84, ref. 7333lnJ
(Received in France 26 April 1994; accepted 15 June 1994)