- Email: [email protected]
Intramolecular addition of Free radicals to quaternised heterocyclic rings.
oo404039190 $3.00+ 00 Peqmcm Rear pk
Temhebn Letms,VoL31.No.11.pp 1625-1628.1990 PrintedinGmatB~
Intramolecular
Addition of Free Radicals to Quaternised
Heterocyclic
Rings.
John A. Murphy* and Michael S. Sherburn, Department of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD.
The intramolecular addition of free radicals to quaternary pyridinium salts gives good yields of tetrahydroquinolizinium salts.
The addition of carbon free radicals to protonated hetemcyclic bases shown in Qure 1 has been extensively explored by Minisci and coworkers1 in elegant experiments over many yearsThese authors have shown that the heterocyclic ring needs to bear a positive charge for successful attack by nucleophilic carbon radicals. Barton and coworkers2 have also explored this area. All of the examples in the literatures, however, utilise intermolecular additions to protonated bases. Furthermore the methods of generating carbon radicals so far used by Minisci’s group are all oxidative; this is quite logical since the mechanism by which the reactions proceed4 requires an oxidative step to generate the aromaticity in the products. However, the development of free-radical strategies in synthetic organic chemistry has benefitted enormously from the availability of non-oxidative radicals e.g. ttialkylstannyl radicals. Their reactions generally avoid the aggressive hydrogen-atom abstraction pmcesses which feature so prominently in the chemistry of oxidative radicals. Our strategy was to see if carbon radicals generated using tributyltin hydride would lead to dihydmpytidines , and by analogy with NADH to NAD+ conversions, whether a subsequent mikl oxidative step could complete the overall “substitution” process.
-J-l+
\ CL I
0
:+
AK
oxidant
I
R
\
HN
Figure 1 1625
R
In this pap&, we dacribc theflrstexamples ofintramolecuhrradicaladditionsto quatemized pyridinium salts accomplished using the mild and non-oxidativechemisay of uibutyltin hydride There is no ncctssity~ointrodwxascparateoxidativcs~;infactitisnotpossiblctoisolateproducts~thanthc~~ of substitution.
I I
Bu&I-I,
AIBN
w
OXiddOfl
(3)
(1) mgure2
Pyridbe was tnated with diiodobutaueto generate a mixtuxeof the monopyridinium salt (1) and the dipyridinium salt (2). These were easily sepam&dby cohunn chmamtography followed by mrystallisation. When the salt (1) ( 1 quivalmt ), was treatedwith tributykin hydride ( 1.3 quivalam ) under nitrogen and in the presence of ambisisobutymnitrile ( 1.2 quivalents ) a clean cyclisation ensued as judged by n.m.r. The bicycle (3) , rap. 178-1820 C, was isolated in 60% yield simply by perfo-ming a petrol / a&o&rile partitioning oftiLtcNdertaction~tvaparatingtheacetonitrileand~s~gfnwne~l/ethyl~~.Thetin bypmducts fkom uibutyltin hydride mctions frequently cause ptui6caIioo problems, but because of the polarity of the pyridinium products, arganotiu compounds are very easily xxmoved by the above pmcedure. Hence the use of aibutyltin hydride is positively advantageous.l%e clean_fommion of (3) indicates that no polysubstitution is seen in the reaction; this contrastswith the intenmlecular substitution reactions of pyridinium salt&.
There are a numbexof mechanistic points which deserve cmsidemtion. A full quivalent of azobisisobutyronitrile was squired in these nactions; to ensum that the isobutymniaile radicals produced in the themal decomposition of this substance were not responsible for direct attackon the iodocompound. without involvenmt of tht tributyltin hydride in the radical pmcess. the reaction was mcmned in the absence of tributyltin hydride. This led solely to movery of startingmamials, showing that the tributyltin hydride was requimi for reaction. Another mechanistic possibility was thattxibutyltinhydride might be used in catalytic quantityinthenactian.~scouldoccuriftributyltiniodidewerereduced~byanintermediate dihydropyridine Mvative. However we have &monstmted that this was not the case. In the pnsence of less
1627
than one equivalent of tributyltin hydride, the don
did not p io completion. Two possible pathways for the substitution mction am suggested in Pigme 3. Route A is the pathway suggested by h4inki for his substitutions. With uibutyltin hydride as radical source, however, we cannot rule out the altunative possibility B. The nature of the oxidant required as partof this pmcess is under investigation.
Pathway A
BqSnH
H-
I
-HI
-Bu$nl
Pathway B . Bu&’
oxidant
Figure 3
To test if substituted tetrahydroquinolizinium salts could also be synthesised in this manner2-picolinc(4).4-piooline(5)aad35_lutidiae(6)waccomrated~O~g iodobutylquamnatysalts.Thescwcrt~~lyaansfarmedimobicycliccompounds(7)(8) aIKl(9)on matmcnt with tributybin hydride and AIBN. It is intcmting to note that only one pmduct arises kxn the N-iodobutyl-2-picdinium iodide (10). No productresulting from addition to the carbon bearing the ethyl gmup is seen.We are cummtly exploiting this @oselCCtivity in the synthesis ofcomplex subdution pmducts.
v I
N'
6)
I
-
Figure 4
Acknowledgements. We thank the S.E.R.C. for the award of a Quota studentship.
References. 1. For a review see F. Minisci, E. Vimara, F. Fontana, Hererocycles 1989,28,489. 2.D. H. R. Barton, B. Garcia, H. Togo and S. Z. Zard,TetruhedronLetters, 1986,27,1327 3.R. Baur, T. Sugimoto and W. F%leiderer, HeJv. Chim. Acru, 1988,71,531; M. Tada and Y. Yokoi, J. Heterocyclic Chem., 1989,26, 45. 4. F. Mix&i, E. Vismara and F. Fontma, J. Org. Chem. 1989,54,5224. 5. These results wee previously disclosed at the East Midlands Regional Meeting of the Royal Society of Chemistry. Nottingham, 1989.
(Received in UK 21 December 1989)
Temhebn Letms,VoL31.No.11.pp 1625-1628.1990 PrintedinGmatB~
Intramolecular
Addition of Free Radicals to Quaternised
Heterocyclic
Rings.
John A. Murphy* and Michael S. Sherburn, Department of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD.
The intramolecular addition of free radicals to quaternary pyridinium salts gives good yields of tetrahydroquinolizinium salts.
The addition of carbon free radicals to protonated hetemcyclic bases shown in Qure 1 has been extensively explored by Minisci and coworkers1 in elegant experiments over many yearsThese authors have shown that the heterocyclic ring needs to bear a positive charge for successful attack by nucleophilic carbon radicals. Barton and coworkers2 have also explored this area. All of the examples in the literatures, however, utilise intermolecular additions to protonated bases. Furthermore the methods of generating carbon radicals so far used by Minisci’s group are all oxidative; this is quite logical since the mechanism by which the reactions proceed4 requires an oxidative step to generate the aromaticity in the products. However, the development of free-radical strategies in synthetic organic chemistry has benefitted enormously from the availability of non-oxidative radicals e.g. ttialkylstannyl radicals. Their reactions generally avoid the aggressive hydrogen-atom abstraction pmcesses which feature so prominently in the chemistry of oxidative radicals. Our strategy was to see if carbon radicals generated using tributyltin hydride would lead to dihydmpytidines , and by analogy with NADH to NAD+ conversions, whether a subsequent mikl oxidative step could complete the overall “substitution” process.
-J-l+
\ CL I
0
:+
AK
oxidant
I
R
\
HN
Figure 1 1625
R
In this pap&, we dacribc theflrstexamples ofintramolecuhrradicaladditionsto quatemized pyridinium salts accomplished using the mild and non-oxidativechemisay of uibutyltin hydride There is no ncctssity~ointrodwxascparateoxidativcs~;infactitisnotpossiblctoisolateproducts~thanthc~~ of substitution.
I I
Bu&I-I,
AIBN
w
OXiddOfl
(3)
(1) mgure2
Pyridbe was tnated with diiodobutaueto generate a mixtuxeof the monopyridinium salt (1) and the dipyridinium salt (2). These were easily sepam&dby cohunn chmamtography followed by mrystallisation. When the salt (1) ( 1 quivalmt ), was treatedwith tributykin hydride ( 1.3 quivalam ) under nitrogen and in the presence of ambisisobutymnitrile ( 1.2 quivalents ) a clean cyclisation ensued as judged by n.m.r. The bicycle (3) , rap. 178-1820 C, was isolated in 60% yield simply by perfo-ming a petrol / a&o&rile partitioning oftiLtcNdertaction~tvaparatingtheacetonitrileand~s~gfnwne~l/ethyl~~.Thetin bypmducts fkom uibutyltin hydride mctions frequently cause ptui6caIioo problems, but because of the polarity of the pyridinium products, arganotiu compounds are very easily xxmoved by the above pmcedure. Hence the use of aibutyltin hydride is positively advantageous.l%e clean_fommion of (3) indicates that no polysubstitution is seen in the reaction; this contrastswith the intenmlecular substitution reactions of pyridinium salt&.
There are a numbexof mechanistic points which deserve cmsidemtion. A full quivalent of azobisisobutyronitrile was squired in these nactions; to ensum that the isobutymniaile radicals produced in the themal decomposition of this substance were not responsible for direct attackon the iodocompound. without involvenmt of tht tributyltin hydride in the radical pmcess. the reaction was mcmned in the absence of tributyltin hydride. This led solely to movery of startingmamials, showing that the tributyltin hydride was requimi for reaction. Another mechanistic possibility was thattxibutyltinhydride might be used in catalytic quantityinthenactian.~scouldoccuriftributyltiniodidewerereduced~byanintermediate dihydropyridine Mvative. However we have &monstmted that this was not the case. In the pnsence of less
1627
than one equivalent of tributyltin hydride, the don
did not p io completion. Two possible pathways for the substitution mction am suggested in Pigme 3. Route A is the pathway suggested by h4inki for his substitutions. With uibutyltin hydride as radical source, however, we cannot rule out the altunative possibility B. The nature of the oxidant required as partof this pmcess is under investigation.
Pathway A
BqSnH
H-
I
-HI
-Bu$nl
Pathway B . Bu&’
oxidant
Figure 3
To test if substituted tetrahydroquinolizinium salts could also be synthesised in this manner2-picolinc(4).4-piooline(5)aad35_lutidiae(6)waccomrated~O~g iodobutylquamnatysalts.Thescwcrt~~lyaansfarmedimobicycliccompounds(7)(8) aIKl(9)on matmcnt with tributybin hydride and AIBN. It is intcmting to note that only one pmduct arises kxn the N-iodobutyl-2-picdinium iodide (10). No productresulting from addition to the carbon bearing the ethyl gmup is seen.We are cummtly exploiting this @oselCCtivity in the synthesis ofcomplex subdution pmducts.
v I
N'
6)
I
-
Figure 4
Acknowledgements. We thank the S.E.R.C. for the award of a Quota studentship.
References. 1. For a review see F. Minisci, E. Vimara, F. Fontana, Hererocycles 1989,28,489. 2.D. H. R. Barton, B. Garcia, H. Togo and S. Z. Zard,TetruhedronLetters, 1986,27,1327 3.R. Baur, T. Sugimoto and W. F%leiderer, HeJv. Chim. Acru, 1988,71,531; M. Tada and Y. Yokoi, J. Heterocyclic Chem., 1989,26, 45. 4. F. Mix&i, E. Vismara and F. Fontma, J. Org. Chem. 1989,54,5224. 5. These results wee previously disclosed at the East Midlands Regional Meeting of the Royal Society of Chemistry. Nottingham, 1989.
(Received in UK 21 December 1989)