1,3-Stereoinduction in radical reactions: Radical additions to dialkyl 2-alkyl-4-methyleneglutarates

Article abstract of DOI:10.1021/ja003235j

Tin hydride-mediated radical additions to a series of α-methylene-glutarates 1, furnishing 2;4-dialkyl-substituted glutarates 3 are reported. The diastereoselectivity of hydrogen transfer to the intermediate adduct radicals 2, possessing a stereogenic center in γ-position, was disappointing in the temperature range of -78 to 80 °C. However, the reactions proved to be able to proceed with excellent 1,3-diastereoselectivities under chelation-controlled conditions, depending on the steric impacts of 2- and 4-alkyl substituents as well as on the ester-alkyl moiety and choice of Lewis acid. Using MgBr₂·OEt₂ as additive, syn-selectivities of 98:2 were achieved upon initial tert-butyl radical addition at -78 °C. High anti-diastereoselectivities were observed in the MgBr₂·OEt₂-controlled pathway at 70 °C when smaller alkyl radicals such as cyclohexyl, ethyl, and methyl were applied. Interesting and uncommon temperature dependences were observed in the temperature range of -78 to 100 °C, revealing strong entropic effects in the transition states. A model that accounts for the opposed stereochemical outcomes under chelation-controlled conditions is presented.

Full text of DOI:10.1021/ja003235j

12458
J. Am. Chem. Soc. 2000, 122, 12458-12468
1,3-Stereoinduction in Radical Reactions: Radical Additions to
Dialkyl 2-Alkyl-4-methyleneglutarates
Ahlke Hayen, Rainer Koch, Wolfgang Saak, Detlef Haase, and Ju1rgen O. Metzger*
Contribution from the Department of Chemistry, UniVersity of Oldenburg, P.O. Box 2503,
26111 Oldenburg, Germany
ReceiVed August 31, 2000
Abstract: Tin hydride-mediated radical additions to a series of R-methylene-glutarates 1, furnishing 2,4-dialkyl-
substituted glutarates 3 are reported. The diastereoselectivity of hydrogen transfer to the intermediate adduct
radicals 2, possessing a stereogenic center in γ-position, was disappointing in the temperature range of -78
to 80 °C. However, the reactions proved to be able to proceed with excellent 1,3-diastereoselectivities under
chelation-controlled conditions, depending on the steric impacts of 2- and 4-alkyl substituents as well as on
the ester-alkyl moiety and choice of Lewis acid. Using MgBr2‚OEt2 as additive, syn-selectivities of 98:2
were achieved upon initial tert-butyl radical addition at -78 °C. High anti-diastereoselectivities were observed
in the MgBr2‚OEt2-controlled pathway at 70 °C when smaller alkyl radicals such as cyclohexyl, ethyl, and
methyl were applied. Interesting and uncommon temperature dependences were observed in the temperature
range of -78 to 100 °C, revealing strong entropic effects in the transition states. A model that accounts for
the opposed stereochemical outcomes under chelation-controlled conditions is presented.
Introduction
year, Porter published his results on allyltributylstannane-
mediated acyclic radical addition reactions of alkyl iodides to
oxazolidinone acrylamides to give 1,3-disubstituted products
with good diastereoselectivity.7 Hanessian then reported some
results on remote stereocontrol in free-radical C-allylation
reactions including 1,3-stereocontrol due to hydrogen bonding.8
An example for radical 1,3-stereoselective Michael analogous
additions, where facial discrimination is provided by hydroxy-
alkyl radicals bearing a chiral tetrahydropyranyl or glucosyl
moiety, was given by Garner.9 In another approach, 1,3-
stereoinduction in radical additons to imines was examined by
Bertrand.10 These previous examples represent a variety of
approaches, employing different elements for 1,3-stereocontrol.
Lewis acids have proven to be valuable auxiliaries, being able
to expand the scope of many stereoselective reactions, but for
now, only two examples of chelation-controlled acyclic 1,3-
induction in radical additions to carbon-carbon bonds have been
reported.11 In these examples dealing with hydrogen and allyl
transfer, respectively, given by Nagano, a γ-situated alkoxy or
hydroxy group in addition to a carboxylic ester moiety is used
as the second binding site for Lewis acids, a strategy that has
often been successfully applied in 1,2-diastereoselective reac-
tions.12 The best selectivity in the hydrogen-transfer reaction,
achieved in the presence of La(fod)3, is 11:1.11a We now report
our study, aimed to obtain more knowledge about factors
The control of stereoselectivity in radical reactions still is an
increasing research field,1 and in analogy to nucleophilic
reactions, the use of Lewis acid control is gaining more and
more attention.2 While diastereoselective radical reactions
involving cyclic systems are well established, acyclic stereo-
control still remains an often challenging subject and most of
the research in this field has been concentrated on 1,2- and 1,4-
asymmetric induction.1 Fleming et al. systematically studied the
potential of 1,3-stereoinduction in nucleophilic addition reac-
tions, and from the wide variety of stereoselectivities observed
for these reactions with participants of different steric demand,
they were able to derive some general tendencies for stereo-
control in those particular reactions.3 The potential of 1,3-
stereoinduction in radical reactions, however, has not received
much consideration yet; only a few examples involving cyclic
stereocontrol have been published 4 and investigations dealing
with acyclic radicals are sparsely found. Both with respect to
the synthesis of natural products bearing stereogenic centers in
the 1,3-position and the free-radical polymerization of vinyl
monomers, the possibility of steering the stereoselectivity of
the trapping of acyclic radicals possessing a stereogenic center
in the 3-position is of great importance.5
In 1995, we reported on relative 1,3-asymmetric induction
in iodine transfer to secondary alkyl radicals,6 and in the same
(6) (a) Metzger, J. O.; Mahler, R. Angew. Chem. 1995, 107, 1012-1014;
Angew. Chem., Int. Ed. Engl. 1995, 34, 902-904. (b) Metzger, J. O.; Mahler,
R.; Francke, G. Liebigs Ann./Recl. 1997, 2303-2313.
* To whom correspondence should be addressed: (fax) (+49) 441-798-
3329; (e-mail) juergen.metzger@uni-oldenburg.de.
(1) (a) Porter, N. A.; Giese, B.; Curran, D. P. Acc. Chem. Res. 1991, 24,
296-304. (b) Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of
Radical Reactions; VCH: Weinheim, 1996.
(7) (a) Radinov, R.; Mero, C. L.; McPhail, A. T.; Porter, N. A.
Tetrahedron Lett. 1995, 36, 8183-8186. (b) Mero, C. L.; Porter, N. A. J.
Org. Chem. 2000, 65, 775-781.
(2) Renaud, P.; Gerster, M. Angew. Chem. 1998, 110, 2704-2722;
(8) Hanessian, S.; Yang, H.; Schaum, R. J. Am. Chem. Soc. 1996, 118,
2507-2508.
Angew. Chem., Int. Ed. Engl. 1998, 37, 2562-2579.
(3) Barbero, A.; Blackmore, D. C.; Fleming, I.; Wesley, R. N. J. Chem.
Soc., Perkin Trans. 1 1997, 1329-1352.
(9) Garner, P.; Anderson, J. T. Tetrahedron Lett. 1997, 38, 6647-6650.
(10) (a) Bertrand, M. P.; Feray, L.; Nouguier, R.; Stella, L. Synlett 1998,
780-782. (b) Bertrand, M. P.; Coantic, S.; Feray, L.; Nouguier, R.; Perfetti,
P. Tetrahedron 2000, 56, 3951-3961.
(4) For recent examples, see: (a) Urabe, H.; Kobayashi, K.; Sato, F. J.
Chem. Soc., Chem. Commun. 1995, 1043-1044. (b) Pyne, S. G.; Schafer,
K. Tetrahedron 1998, 54, 5709-5720. (c) Mase, N.; Watanabe, Y.; Toru,
T. J. Org. Chem. 1998, 63, 3899-3904.
(11) (a) Nagano, H.; Toi, S.; Yajima, T. Synlett 1999, 53-54. (b) Nagano,
H.; Hirasawa, T.; Yajima, T. Synlett 2000, 1073-1075.
(12) Guindon, Y.; Rancourt, J. J. Org. Chem. 1998, 63, 6554-6565.
(5) See ref 1b, pp 203-208.
10.1021/ja003235j CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/29/2000

1,3-Stereoinduction in Radical Reactions
J. Am. Chem. Soc., Vol. 122, No. 50, 2000 12459
Table 1. Screening of Diastereoselectivity and Lewis Acid
Activity of the Tin Hydride-Mediated Addition Reaction of
tert-Butyl Iodide and Dimethyl 2-Methyl-4-methylenepentanedioate
(1a) at -78 °Ca
governing 1,3-diastereoselective reactions. We performed a
systematic investigation of the diastereoselectivity of hydrogen
transfer to chiral radicals 2 formed by radical additions to
2-alkyl-4-methylene pentanedioates 1, examining temperature,
substituent, and Lewis acid effects on the stereochemical
outcome. This system offers several advantages. Two ester
groups allow a sufficient complexation and, if suitable, chelation
of any supplied Lewis acid. Furthermore, a variation of the alkyl
moiety of the ester groups enables further insight into structural
factors governing stereoselectivity. In addition, by changing the
steric impact of the alkyl group in the γ-position relative to the
arising radical center, and by varying the initially attacking alkyl
radicals, the overall condition of the adduct radical, the trapping
agent is faced with, can be systematically varied. By combining
new experimental results with the results of semiempirical
calculations, we were able to reasonably evaluate the stereo-
chemical outcome.13
Lewis acid
(2 equiv)
yield
(%)b
conversion
entry
of 1a (%)b
[anti-3a]:[syn-3a]c
1e
2
3
4
5
6
7
8
9
10
11d,e
12d,e
13e
14d,e
15d,e
16e
81
3
90
67
5
5
7
17
95
90
93
58
60
83
77
70
82
95
92
100
96
65
72
31
96
99
100
100
97
100
100
78
52:48
15:85
50:50
42:58
34:66
23:77
38:62
51:49
44:56
52:48
18:82
1:99
AlCl3
Al(iPrO)3
AlMe3
Et2AlCl
Et3Al2Cl3
TiCl4
Ti(iPrO)4
InCl3
CeCl3
LiClO4
Sc(OTf)3
ZnBr2
MgBr2‚OEt2
MgI2
45:55
2:98
2:98
Results
Mg(OEt)2
52:48
The required alkenes 1a-d were synthesized according to a
four-step sequence depicted in Scheme 1 in an overall yield of
62 (Me, steps iii and iv), 20 (nPr), 25 (iPr), and 4% (cHex).14
a Reactions were carried out in CH2Cl2 on analytical scale (1a 1
equiv, Lewis acid 2 equiv, Bu3SnH 3 equiv, tert-butyl iodide 3 equiv,
Et3B 1 equiv/O2) at an average reaction time of 4-6 h. b Yield and
conversion of 1a were determined by GC analysis using dodecane as
internal standard. c Diastereomeric ratio of 3a was determined by GC.
d Carried out on preparative scale; yields are isolated yields. e Et2O was
used as cosolvent.
Scheme 1. Synthesis of 2-Alkyl-4-methylenepentanedioates
1a-da
disappointing diastereomeric ratio of [anti-3a]:[syn-3a] ) 52:
48 (Table 1, entry 1). To test the potential of a variety of
different Lewis acids for our system, we conducted a screening
at -78 °C on an analytical scale for the addition of tert-butyl
iodide to 1a with 2 equiv of Lewis acid previously mixed with
the alkene (Table 1). High yielding, but almost unselective
reactions took place in the presence of Al(iPrO)3, InCl3 and
CeCl3 (entries 3, 9, and 10); a comparable outcome was
observed for Mg(OEt)2 (entry 16). Very good and excellent syn-
selectivities were obtained with LiClO4, MgBr2‚OEt2, MgI2, and
Sc(OTf)3 (entries 11, 12, 14, and 15). Aluminum chloride and
alkylaluminum chlorides (entries 2, 5, and 6) as well as titanium
tetrachloride and isopropylate (entries 7 and 8) were unsatisfying
with regard to selectivity and yield. Trimethylaluminum and
zinc bromide showed moderate yields and low selectivities
(entries 4 and 13). Overall, the screening revealed MgBr2‚OEt2
as being the most suitable additive, considering selectivity,
conversion, yield, and cost (entry 14), while aluminum and
titanium organyls and halides, respectively, were unsatisfying
with regard to selectivity and yield (entries 2-8).
The influence of different solvents on the diastereoselectivity
of the chosen test reaction at -78 °C with MgBr2‚OEt2
supplement was examined. As expected, diethyl ether and
dichloromethane, which are mostly employed in analogous low-
temperature studies, proved to be the most eligible mediums,
providing a 1,3-diastereoselectivity of [anti-3a]:[syn-3a] of 2:98
and 4:96 (Table 2, entries 1 and 3), respectively, whereas THF
unexpectedly led to a reversed and low selectivity of 58:42
(entry 2). Toluene emerged as being slightly less suitable than
dichloromethane with regard to selectivity (9:91, entry 4), and
the most unpolar pentane finally gave a less selective perfor-
mance of 26:74 (entry 5). Considering conversion, dichlo-
romethane and toluene apparently slowed the reaction; yields
a Conditions: (i) Na, MeOH, R1Br. (ii) (1) KOH, HCl, (2) pyridine,
piperidine, paraformaldehyde, reflux. (iii) Na, MeOH, dimethyl mal-
onate.
Tin hydride-mediated reaction of alkyl bromides or iodides
with alkenes 1a-d gave adduct radicals 2. Hydrogen transfer
from tributyltin hydride yielded new relatively syn- and anti-
configured 2,4-dialkyl-substituted dialkyl pentanedioates 3a-k
as outlined in Scheme 2.
Scheme 2. General Outline of the Examined Tin
Hydride-Mediated Radical Addition Reaction of Alkyl
Halides and Alkenes 1a
a Racemic products were obtained. For convenience, only one
enantiomer is depicted.
(13) For a preliminary account of our results, see: Hayen, A.; Koch, R.;
Metzger, J. O. Angew. Chem. 2000, 112, 2898-2900; Angew. Chem., Int.
Ed. 2000, 39, 2758-2761.
(14) (a) Stetter, H.; Kuhlmann, H. Synthesis 1979, 29-30. (b) Stetter,
H.; Marten, K.; Liebigs Ann. Chem. 1982, 240-249. The synthetic approach
to 1a started with commercially available methyl 2-methylacrylate.
Reaction of tert-butyl iodide with alkene 1a at -78 °C gave
the addition products in good yields of 81% but with a

12460 J. Am. Chem. Soc., Vol. 122, No. 50, 2000
Hayen et al.
Table 2. Screening of the Influence of Solvents on the
Diastereoselectivity of the Tin Hydride-Mediated Addition Reaction
of tert-Butyl Iodide (3 Equiv) and Dimethyl
Table 4. Diastereoselectivity of the Tin Hydride-Mediated Radical
Addition Reaction of tert-Butyl Halides and Dimethyl
2-Alkyl-4-methylenepentanedioates 1b-d. Influence of Temperature
a
2-Methyl-4-methylenepentanedioate (1a) in the Presence of 2 Equiv
and MgBr2‚OEt2
of MgBr2‚OEt2 at -78 °Ca
MgBr2‚OEt2
(equiv)
yield
yield
(%)b
conversion
entry alkene
T (°C) (%)b [anti-3]:[syn-3]c
entry
solvent
of 1a (%)b
[anti-3a]:[syn-3a]c
1
2
1b
1b
1c
1c
1d
1d
1b
1b
1c
1c
1d
1d
-78
80
-78
80
-78
80
-78
70
-78
70
-78
70
67
66
74
88
94
72
92
50d
89
65
75d
50d
48:52
54:46
45:55
54:46
42:58
51:49
2:98
24:76
5:95
35:65
3:97
1
2
3
4
5
Et2O
THF
CH2Cl2
toluene
pentane
100
100
79
59
100
100
100
79
68
100
2:98
58:42
4:96
9:91
26:74
3
4e
5
6e
7
1
2
1
2
1.5
2
a The same reaction procedure was used as for reactions of Table 1.
b Yield and conversion of alkene were determined by GC analysis using
dodecane as internal standard. c Diastereomeric ratio of 3a was
determined by GC.
8
9
10
11
12
32:68
Table 3. Temperature Dependence of the Diastereoselectivity of
the Tin Hydride-Mediated Addition Reaction of tert-Butyl Halides
and Dimethyl 2-Methyl-4-methylenepentanedioate (1a).a Variation
of the Concentration of MgBr2‚OEt2
a Conditions: 1b-d 1 equiv, Bu3SnH 3 equiv, alkyl halide 3 equiv.
Alkyl iodides were employed at -78 °C, alkyl bromides at 70 and 80
°C. As solvent, benzene was used at 70 and 80 °C (AIBN as initiator);
at -78 °C, dichloromethane was used (Et3B as initiator). Diethyl ether
was added as cosolvent, if MgBr2‚OEt2 was employed. b Yields are
isolated yields. c Diastereomeric ratio was determined by GC. d Incom-
plete conversion of alkene at the time of analysis. e An excess of 1
was used. The conversion of 1 was not determined.
MgBr2‚OEt2 yield conversion
entry T (°C)
(equiv)
(%)b
(%)c
[anti-3a]:[syn-3a]d
1
2
3
4
-78
-20
0
82
55
100
90
80
83
100
82
85
96
44
82
e
100
92
100
100
100
100
95
52:48
51:49
52:48
51:49
2:98
2:98
1:99
20
the series of MgBr2‚OEt2-controlled reactions, selectivities
increased by lowering the temperature, ranging from an almost
unselective reaction in toluene/Et2O at 100 °C with an [anti] to
[syn] ratio of 46:54 (entry 12) to a ratio of 2:98 at -78 °C
(entry 5). The amount of MgBr2‚OEt2 equivalents applied was
varied in entries 5-7; 1 or 2 equiv with respect to alkene
afforded anti- and syn-products 3a in a ratio of 2:98, and 5
equiv led to a once more improved ratio of 1:99, as measured
by GC analysis. Although the results of the temperature series
seem to be following a reasonable linear behavior, by analyzing
this temperature range (entries 5 and 8-12), one has to keep in
mind that not all the applied reaction conditions (i.e., solvent)
are directly comparable.
5
-78
-78
-78
-20
0
20
70
1
2
5
2
2
2
1
2
6g
7f
8
5:95
9
10
11
12
10:90
15:85
39:61
46:54
96
e
70
100
36
a Conditions: 1a 1 equiv, Bu3SnH 3 equiv, alkyl halide 3 equiv. Alkyl
iodides were employed at -78, -20, 0, and 20 °C; alkyl bromides at
70 and 100 °C. As solvent, benzene was used at 70 and 80 °C, toluene
at 100 °C (AIBN as initiator); at all other temperatures, dichloromethane
was used (Et3B as initiator). Diethyl ether was added as cosolvent, if
MgBr2‚OEt2 was employed. b Yields are isolated yields. c Conversion
of alkene was determined by GC analysis using dodecane as internal
standard. d Diastereomeric ratio of 3a was determined by GC. e An
excess of 1a was used. The conversion of 1a was not determined. f The
reaction was performed on analytical scale and analyzed by GC using
dodecane as internal standard. g Entry 14, Table 1.
Table 4 deals with the effect of substituent R1 in the γ-position
to the radical center in 2b-d on the diastereoselectivity for two
different temperatures of -78 and 70/80 °C. In contrast to the
results displayed in Table 3 (2a, R1 ) Me), a remarkable
temperature dependence was observed with reactions in the
absence of MgBr2‚OEt2. At -78 °C, in all three cases, low syn-
selectivities were observed, becoming larger with increasing
steric effect of the substituent R1 (nPr < iPr < cHex), going
from [anti-3]:[syn-3] ) 48:52 (1b, entry 1) through 45:55 (1c,
entry 3), to 42:58 (1d, entry 5). However, by rising the reaction
temperature to 80 °C, inversion to low anti-selectivities occurred
in all cases (entries 2, 4, and 6). In the chelation-controlled
reaction pathway, the general tendencies observed for alkene
1a (Table 3) were maintained, that is, a very high syn-preference
at -78 °C and a moderate syn-selectivity at 70 °C. There seems
to be only a minor effect of the γ-alkyl substituent on the
diastereoselectivity: at -78 °C a variation of [anti]:[syn]
between 2:98 (1a, Table 3, entries 5 and 6; 1b, Table 4, entry
7), 3:97 (1d, Table 4, entry 11), and 5:95 (1c, Table 4, entry 9)
was observed.
were satisfying to quantitative in all cases. As a conclusion, a
solvent mixture composed of diethyl ether and dichloromethane
was chosen for further studies when a Lewis acid was involved.
A solvent composition with diethyl ether as a key role additive
has been recognized before as being necessary for optimal results
in Lewis acid-controlled enantioselective radical-mediated al-
lylations.15
In Table 3, the results of the temperature influence on the
diastereoselectivity of the same reaction as chosen above are
listed in detail. High-temperature reactions (70 and 100 °C) were
performed in benzene or toluene with a small amount of diethyl
ether as additive if MgBr2‚OEt2 was present and were initiated
by AIBN, reactions at 20, 0, and -78 °C were carried out in
dichloromethane/diethyl ether, as initiator the Et3B/O2 system
was used, and reactions at -20 °C were performed in a
photoreactor using a 5-W low-pressure mercury lamp and, again,
with dichloromethane/diethyl ether as solvent. When the reac-
tions were performed at 70, 80, or 100 °C, alkyl bromides were
used, and at other temperatures, iodides. In the case of
nonchelation, an unselective reaction was observed in the
temperature range of -78 to 20 °C (entries 1-4). However, in
Surprising results were obtained by applying cyclohexyl
halide instead of tert-butyl halide as precursor for the primarily
attacking alkyl radical, providing a cyclohexylmethyl instead
of the previously built up neopentyl moiety in the R-position
to the arising radical center (Table 5). Evidently, minimizing
the steric impact of the R-substituent of the adduct radical had
a major effect on the hydrogen-transfer step determining relative
product configuration, an effect that was not observed by
(15) Murakata, M.; Jono, T.; Mizuno, Y.; Hoshino, O. J. Am. Chem.
Soc. 1997, 119, 11713-11714.

1,3-Stereoinduction in Radical Reactions
J. Am. Chem. Soc., Vol. 122, No. 50, 2000 12461
Table 5. Diastereoselectivity of the Tin Hydride-Mediated Radical
Addition Reaction of Cyclohexyl Halides (R2 ) cHex) and
Dimethyl 2-Alkyl-4-methylenepentanedioates 1a-d. Influence of
Table 6. Diastereoselectivity of the Tin Hydride-Mediated Radical
Addition Reaction of Methyl, Ethyl, and Isopropyl Halides (R2 )
Me, Et, iPr) and Dimethyl 2-Methyl-4-methylenepentanedioate (1a)
as Well as Reduction of Dimethyl 2-Iodo-2,4-dimethylpentanedioate
a
Temperature and MgBr2‚OEt2
a
(4). Influence of Temperature and MgBr2‚OEt2
MgBr2‚OEt2
(equiv)
yield
(%)b
entry alkene
T (°C)
[anti-3]:[syn-3]c
MgBr2‚OEt2
(equiv)
yield
entry substrate R2
T (°C) (%)b [anti-3]:[syn-3]d
1e
2e
3
1a
1a
1b
1b
1c
1c
1d
1d
1a
1a
1a
1a
1a
1a
1b
1b
1c
1c
1d
1d
-78
80
-78
80
-78
80
-78
80
-78
-20
0
20
70
100
-78
70
-78
70
60
98
96
80
98
60:40
52:48
63:37
54:46
59:41
54:46
59:41
53:47
53:47
71:29
69:31
70:30
65:35
62:38
53:47
73:27
56:44
78:22
52:48
81:19
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
4
iPr
iPr
Et
Et
Et
Et
Et
Et
Et
Me
Me
Me
Me
Me
Me
Me
H
-78
-78
-78
63:37c
52:48c
57:43
47:53
85:15
85:15
83:17
81:19
71:29
55:45
46:54
85:15
85:15
86:14
85:15
68:32
59:41
45:55
87:13
80:20
2
4e
5
70
80 100
6e
7
8
9
10
11
12
13
14
15
16
17
18
19
20
87
2
2
2
2
2
-78 100
0
20
40
70
-78
80
-78
0
20
40
70
-78
80
-78
70
82d
68d
64d
38
76
48
53
58
23
85
58
46
42
50
40
76
82
55
70
1
1.5
2
2
2
2
1
2
1.5
2
100f
68d
58d,f
75d,f
98
2
2
2
2
2
95
88
90
62d
58d
4
4
4
H
H
H
1
2
-78
70
2
2
a Conditions: 1 1 equiv, Bu3SnH 3 equiv, alkyl halide 3 equiv. Alkyl
iodides were employed at -78, -20, 0, and 20 °C, alkyl bromides at
70, 80, and 100 °C. As solvent, benzene was used at 70 and 80 °C,
toluene at 100 °C (AIBN as initiator); at all other temperatures,
dichloromethane was used (Et3B as initiator). Diethyl ether was added
as cosolvent, if MgBr2‚OEt2 was employed. b Yields are isolated yields
unless specified otherwise. c Diastereomeric ratio was determined by
GC. d Incomplete conversion of 1 at the time of analysis. e An excess
of 1 was used. The conversion of 1 was not determined. f Yield was
determined by GC analysis using dodecane as internal standard.
a Conditions: 1 1 equiv, Bu3SnH 3 equiv, alkyl halide 3 equiv. Alkyl
iodides were employed at -78, 0, 20, and 40 °C, methyl iodide and
ethyl bromide at 70 and 80 °C. As solvent, benzene was used at 70
and 80 °C (AIBN as initiator); at all other temperatures, dichlo-
romethane was used (Et3B as initiator). Diethyl ether was added as
cosolvent, if MgBr2‚OEt2 was employed. b Yield was determined by
GC analysis using dodecane as internal standard. c Reaction was
performed on analytical scale, product retention time was confirmed
by GC/MS, and relative product configuration was assigned by elution
order. d Diastereomeric ratio was determined by GC.
changing the nature of the γ-substituent upon tert-butyl radical
addition (Table 4).
tion temperatures, benzene and toluene, are less eligible medi-
ums as already indicated by the results presented in Table 2.
Remarkably, the nonchelation reactions again did not follow
this temperature/selectivity profile as selectivity increased by
lowering the temperature to -78 °C in all four cases (Table 5,
entries 1-8). While reactions at 80 °C were comparably
unselective to the ones displayed in Table 4 for tert-butyl radical
addition, at -78 °C a moderately enhanced and, most remark-
ably, higher anti-stereoselectivity than in the chelated case could
be observed, with alkene 1b giving the highest [anti] to [syn]
ratio of 63:37 (entry 3).
Table 6 summarizes the results of some additional reactions
concerning the role of R2, employing isopropyl, ethyl, methyl
(Scheme 2), and hydrogen (Scheme 3) leading to 2,4-dialkyl-
pentanedioates 3. Upon tin hydride-mediated reaction of MgBr2‚
OEt2-complexed alkene 1a with iPrI at -78 °C, an almost
unselective reaction took place, giving an [anti] to [syn] ratio
of 52:48 of 3i (Table 6, entry 2). With no Lewis acid present,
the selectivity increased to 63:37 (entry 1). These findings
resembled those of cyclohexyl addition to 1a (Table 5, entries
1 and 9), as expected from the geometrical similarity of these
alkyl groups; thus only two analytical experiments were
performed. Methyl and ethyl halides, however, again showed a
different behavior and, in general, proceeded in a more sluggish
reaction manner than the preceding examined reactions. In the
MgBr2‚OEt2-controlled reaction pathway, the reactions furnish-
ing dimethyl 2-methyl-4-propylpentanedioate (3j) and dimethyl
2-ethyl-4-methylpentanedioate (3k) did not display a visible
temperature dependence in the range of -78 to 40 °C and led
to high anti-selectivities of ∼85:15 (Table 6, entries 5-8 and
12-15), but when performed at 70 °C, the anti-selectivity
decreased slightly.
Alkenes 1b-d bearing different γ-alkyl substituents were
subjected to the addition reaction at two different temperaturess
with alkene 1a a broader temperature range was examineds
and again the influence of MgBr2‚OEt2 as Lewis acid was
examined. In contrast to the results presented in Tables 3 and
4, the general preference of this addition-transfer sequence
pointed into the direction of anti-selectivity, irrespective of the
presence of a Lewis acid. What was even more striking, an
unexpected temperature behavior was observed. Starting with
the chelation-controlled pathway, at 70 °C the reactions
exhibited an impressive anti-diastereoselectivity in contrast to
the exclusive syn-selectivity demonstrated by the Lewis acid-
controlled hydrogen transfer upon the adduct radicals 2a-d.
The diastereomeric anti/syn-ratio increased clearly visible, going
from R2 ) Me (Table 5, entry 13, 65:35) via nPr (entry 16,
73:27), iPr (entry 18, 78:22), to cyclohexyl (entry 20, 81:19),
thus showing a significant γ-substituent effect in this case.
However, the same series at -78 °C did not demonstrate the
anticipated enhancement of the anti-preference; it rather showed
an unselective reaction (entries 9, 15, 17, and 19). In entries
9-14, the results of a more extended temperature study
employing alkene 1a are displayed. As indicated above,
selectivities increased by raising the reaction temperature. While
the reaction proceeded almost unselectively at -78 °C (entry
9, 53:47), at -20, 0, and 20 °C, moderate anti-preferences of
∼70:30 were observed (entries 10-12). Higher temperatures
resulted in once more declining selectivities of 65:35 (70 °C,
entry 13) and 62:38 (100 °C, entry 14). However, once more
one has to keep in mind that the employed reaction conditions
were not comparable in all cases (i.e., deviations at 70 °C and
100 °C), obviously especially the solvents used for higher reac-

12462 J. Am. Chem. Soc., Vol. 122, No. 50, 2000
Hayen et al.
Table 7. Diastereoselectivity of the Tin Hydride-Mediated Radical
Addition Reaction of tert-Butyl and Cyclohexyl Halides (R2 ) tBu,
cHex) and Di-tert-butyl 2-Alkyl-4-methylenepentanedioate 5.
The nonchelated reactions showed a moderate anti-preference
at -78 °C ([anti]:[syn] ) 57:43, entry 3, R2 ) Et; 55:45, entry
10, R2 ) Me) and a moderate syn-selectivity at 80 °C ([anti]:
[syn] ) 47:53, entry 4, R2 ) Et; 46:54, entry 11, R2 ) Me).
The simplest substitution pattern for R2, which is H, was
examined by first introducing an iodine atom to the R-position
of commercially available dimethyl 2,4-dimethylpentanedioate,
thus providing dimethyl 2-iodo-2,4-dimethylpentanedioate (4)
as a substrate for diastereoselective H-transfer. The stereose-
lectivity of the Lewis acid-supported reduction of 4 (Scheme
3) via radical 2l was comparable to that for radicals 2j and 2k,
a
Influence of Temperature and MgBr2‚OEt2
MgBr2‚OEt2
(equiv)
yield
(%)b
entry
R2
T (°C)
[anti-7]:[syn-7]c
1
2
3
4
5
6
7
8
9
tBu
tBu
tBu
tBu
cHex
cHex
cHex
cHex
cHex
-78
80
-78
70
-78
80
-78
20
65
15
69
43:57
47:53
4:96
2
2
89
49:51
42:58
89:11
88:12
79:21
88d
93
2
2
2
88d
11d
Scheme 3. Reduction of Dimethyl
2-Iodo-2,4-dimethylpentanedioate (4) by Tributyltin Hydride
70
a Conditions: 5 1 equiv, Bu3SnH 3 equiv, alkyl halide 3 equiv. Alkyl
iodides were employed at -78 and 20 °C, alkyl bromides at 70 and 80
°C. As solvent, benzene was used at 70 and 80 °C (AIBN as initiator);
at all other temperatures, dichloromethane was used (Et3B as initiator).
Diethyl ether was added as cosolvent, if MgBr2‚OEt2 was employed.
b Yields are isolated yields unless specified otherwise. c Diastereomeric
ratio was determined by GC. d Yield was determined by GC analysis
using dodecane as internal standard.
although reaching even better anti-selectivity at -78 °C with
[anti-3l]:[syn-3l] ) 87:13 (entry 19). The temperature depen-
dence was solely examined by means of -78 and 70 °C, but
even in the 70 °C case, an only slightly diminished selectivity
of still 80:20 was observed (entry 20). In the absence of a Lewis
acid, stereoselectivities were again comparable to or slightly
better than in the ethyl and methyl addition case, going from
55:45 (entry 10, Me) via 57:43 (entry 3, Et) to 59:41 (entry 17,
H) at -78 °C.
conditions of tin hydride-mediated tert-butyl addition under
chelation control by MgBr2‚OEt2, this resulted in complete
destruction of the educt alkene, without any addition product
being detectable. Without Lewis acid, the high-temperature
reaction was comparably destructive, yet yielding 15% of the
expected addition product 7a in a ratio of [anti-7a]:[syn-7a] )
47:53 (Table 7, entry 2). At -78 °C, 7a was obtained in a ratio
of [anti]:[syn] ) 43:57 in 65% isolated yield (entry 1). A more
satisfying result was achieved by using MgBr2‚OEt2: although
the selectivity did not come up to our expectations of topping
the results achieved with 1a (Table 3, entry 5), an acceptable
isolated yield of 69% (Table 7, entry 3) and a selectivity of
[anti-7a]:[syn-7a] ) 4:96 was obtained. (See Scheme 5.)
Having investigated the influence of the γ-substituent R1 and
sterically different cases of primarily attacking radicals R2 on
the 1,3-stereoinduction under varying conditions, it remained
to determine the impact of the ester alkyl moiety R3. An
enlargement of this group had proven to enhance the 1,3-
stereoselectivity of iodine transfer in a previous study conducted
by our group.6b
Thus, we decided to study the respective di-tert-butyl ester
5 in comparison to dimethyl ester 1a, which was prepared as
shown in Scheme 4 in two steps from ester 1a.16 This synthetic
Scheme 5. Tributyltin Hydride-Mediated Addition Reaction
of Alkyl Halides and tert-Butyl Esters 5
Scheme 4. Synthesis of Di-tert-butyl
2-Methyl-4-methylenepentanedioate (5)a
The cyclohexyl radical additon to 5 was lined up next in order
to complete the picture, and once again, the outcome was
surprising. In the MgBr2‚OEt2-controlled reaction at -78 °C,
the desired product di-tert butyl 2-(cyclohexylmethyl)-4-meth-
ylpentanedioate (7b) was obtained with an encouragingly high
anti-preference of 89:11 and in excellent isolated yield of 93%
(entry 7). By raising the reaction temperature to 20 °C, this
selectivity remained practically unchanged (88:12, entry 8),
while at 70 °C, a diminished selectivity of 79:21 was observed.
Again, a reason for this sudden decrease in selectivity at higher
temperature, which occurred in other cases as well, as reported
above, can be assumed in the change of the solvent. The
nonchelated reaction also furnished the product in a high yield
(89%), but unselectively (entry 5) at -78 °C and with a
moderate syn-preference of 58:42 at 80 °C (entry 6). Compared
to the respective reactions of dimethyl ester 1a (Table 5, entries
1 and 9), these experiments clearly assigned a role to the ester
a Conditions: (i) KOH/EtOH reflux, 2 h. (ii) tert-BuOH, CuCl,
CH2Cl2, room temperature, 3 d. iii) CHCl3, 2 d.
approach was started by saponification of 1a, providing 2-meth-
yl-4-methylenepentanedioic acid, which was then reacted with
a previously prepared solution of tert-butyl N,N′-diisopropyl-
imidocarbamate.
When di-tert-butyl 2-methyl-4-methylenepentanedioate (5)
was subjected at 70 °C to the successfully applied reaction
(16) (a) Vowinkel, E. Chem. Ber. 1967, 100, 16-22. (b) Mathias, L. J.
Synthesis 1979, 561-576.

1,3-Stereoinduction in Radical Reactions
J. Am. Chem. Soc., Vol. 122, No. 50, 2000 12463
alkyl moiety R3 as being able to govern diastereofacial
discrimination.
To be able to directly compare the potential of radical and
ionic reaction pathways, we performed some nucleophilic 1,4-
additions of tert-butylmagnesium bromide to alkenes 1a and
1b and of cyclohexylmagnesium bromide to alkene 1a at 0 °C.
In all cases, the respective anti-products were formed prefer-
entially ([anti]:[syn] ) 70:30 (3a), 83:17 (3b), 74:26 (3e)) in
low to moderate yields. Thus, tert-butyl addition occurred with
reversed diastereoselectivity compared to the MgBr2 chelation-
controlled radical reaction whereas in the case of cyclohexyl
addition the same direction and amount of selectivity was
observed.
Assignment of Relative Stereochemistry. Treatment of
diastereomeric mixtures of 3a and 3e, respectively, with KOH/
EtOH furnished the corresponding pentanedioic acids 8a and
8b as white solids. Recrystallization from ethyl acetate evidently
resulted in crystallization of one configurational form only in
both cases, as deduced from derivatization with diazomethane
and subsequent examination by GC. The relative configurations
of these stereoisomers were assigned by means of single-crystal
X-ray analyses as being syn in either circumstances. Since 3a
and 3e represent the simplest forms of the series of pentanedio-
ates 3, the relative configurations of all other substances could
be deduced based on these results, comparing characteristic
patterns of GC retention times and 1H NMR analysis. GC
retention times of the two diastereomeric forms generally
showed a 2-min delay, with the anti-diastereomers always
eluting first. Another typical feature of the syn-diastereomers
Figure 1. Transition-state models of hydrogen transfer to chelated
radicals 2 and 6 bearing alkyl substituents R1, R2, and R3. Syn-products
can be generated via A‡ and D‡, anti-products via B‡ and C‡.
anti/syn-preferences caused by different substituents R2 in the
tin hydride-mediated addition reactions in the presence of
MgBr2‚OEt2 (Scheme 2). While in the tert-butyl pathway
excellent syn-diastereoselectivities were observed (Tables 3 and
4), preferably anti-configured products were produced with R2
) cHex, iPr, Et, Me, and H (Tables 5 and 6). Apart from the
general selectivity issue, a discussion should also include the
observed diverse and not trivial temperature behavior. The
reactions performed in the absence of a Lewis acid were less
selective in most cases, not showing similar interesting tenden-
cies, which is obviously due to the conformational flexibility
of the alkyl chain.
To develop a simple transition state model of the hydrogen-
transfer step, which explains the various stereoselectivities
observed in the Lewis acid-controlled reactions, our first general
assumption was that the two 1,3-carboxy functionalities of
radicals 2 and 6 are coordinated to the Lewis acid MgBr2‚OEt2,
thus creating an eight-membered-ring system and restraining
the conformational flexibility. Strong support for this was
provided by PM3 calculations, employing [Li(OH2)2]+ as Lewis
acid, which revealed that an eight-membered-ring system indeed
is very stablesirrespective of R1. Although the ring conforma-
tion employed for further consideration (Figure 1) is not the
only possible one, it was calculated to be the most stable eight-
membered one. Further calculations of likely minimum struc-
tures were performed in order to better understand the confor-
mational behavior of the adduct radicals; the resulting energy
differences are compiled in Table 8.
1
was observed in the H NMR spectra: whereas the diaste-
reotopic methylene protons of position-3 of the anti-forms gave
a multiplett at 1.6-1.9 ppm, in the syn-stereoisomers, these
protons provided separated signals; one could be found in the
area of 1.4-1.6 ppm, and the other one exhibited a significant
downfield shift and was observed at 1.9-2.0 ppm. These
downfield-shifted signals could be clearly distinguished in most
cases, because no other signals covered this ppm range. The
diastereotopic protons in the 3′-position (C6H11CHH, C4H9CHH)
did not enable a distinction between syn- and anti-diastereomers.
13C NMR spectra also provided further clues to diastereomerical
assignment: in both the neopentyl- (3a-d) and the cyclohexyl-
methyl series (3e-h), the methylene carbon signals of
(CH3)3CCH2 and C6H11CH2, respectively, appeared at higher
field for the syn-forms. For 3a-d, the respective signals turned
up in the range of 45-48 ppm, with the anti-diastereomers
showing a downfield shift ∆δ relative to syn of 0.41 (3a) to
2.32 ppm (3d). Similarly, for 3e-h downfield shifts ∆δ of 0.27
(3e) to 1.68 ppm (3h) were observed for the methylene signals
of the anti-forms, with signals generally appearing between 39
and 42 ppm. 3i was only treated analytically, so product
stereochemistry was deduced from GC elution order. 3j, 3k,
7a, and 7b were isolated (only anti-forms: 3j, 3k, 7a) and
characterized; GC and NMR analysis here again proved the
previously observed patterns right. syn- and anti-3l were
assigned by applying the NMR and GC characteristics to an
original sample of meso- and (()-3l, which then enabled
comparison of GC elution order and thus determination of
product stereochemistry of reaction material.
Figure 1 shows four possible transition states A‡-D‡ of the
hydrogen-transfer step based on the calculated minimum ground
states, which seem to be most likely to be contributing to product
distribution. Syn-products can be generated via transition states
A‡ and D‡, and anti-products via B‡ and C‡. In A‡ and B‡, the
alkyl ligand R1 at position-3 is pointing away from the radical
center, rendering these less sterically congested than C‡ and D‡,
where R1 is located closer to it. In all chelated structures, the
ring shows a concave and a convex side (upper side and bottom
side in Figure 1, respectively). Hydrogen transfer is supposed
to be directed by the relative weighting of the steric interaction
between the chelate ring system with the present arrangement
of substituents at the γ-position and R-alkyl group R2 on one
hand and, on the other hand, interaction between Bu3SnH and
Discussion
Our experimental results show a remarkable synthetic po-
tential for 1.3-stereoinduction in chelation-controlled radical
trapping reactions. The main selectivity phenomena emerging
from the results described in the preceding section, which should
be posed for a discussion, certainly are the discrepancies in the

12464 J. Am. Chem. Soc., Vol. 122, No. 50, 2000
Hayen et al.
Table 8. Results of PM3 Calculations of Radicals 2a (R2 ) tBu),
2e (R2 ) cHex), 2j (R2 ) Et) 2k (R2 ) Me), 2l (R2 ) H), 6a (R3
) tBu, R2 ) tBu), and 6b (R3 ) tBu, R2 ) cHex). Relative
Energies (kJ/mol) of Conformations A - Da
anti-product is formed. This is supported by the calculations of
radical 2e, with conformation A being by 3.8 kJ/mol more stable
than conformation B. In addition, a strong entropic effect must
be assumed. Judging by the degree of rotational freedom of R2
and on account of the attack of the hydrogen donor from the
free convex side of 2, the transition state exhibiting the highest
activation entropy should be B‡, leading to anti-product. With
∆H‡anti and ∆S‡anti being higher than their syn-counterparts, there
must exist a temperature where the enthalpic term and the
entropic term of eq 1 are compensating each other, leading to
an unselective reaction.17 Evidently, this temperature is located
somewhere about -78 °C in this case. By increasing the
temperature, anti-3e-h are formed with considerable selectivity
due to the now dominating entropic effect, obviously via
transition-state B‡. The increasing anti-selectivity with growing
steric demand of R1 indicates that a participation of transition-
state D‡, furnishing syn-product is involvedsdecreasingly with
increasing steric impact of R1.
R2
R3
A
B
C
D
tBu (2a)
cHex (2e)
Et (2j)
Me (2k)
H (2l)
tBu (6a)
cHex (6b)
Me
Me
Me
Me
Me
tBu
tBu
0
0
0
0
0
0
0
9.2 (38)
3.8 (17)
-0.5 (11)
0.3 (11)
9.6
10.5
8.0
11.0
11.0
10.6
9.5
22.6
17.2
11.0
11.6
6.9 (32)
-3.8 (18)
16.9
9.7
a Conformations A B, C, and D according to the respective transition
states A‡, B‡, C‡, and D‡, depicted in Figure 1. The rotational barrier
for R2 (kJ/mol) is given in parentheses.
this ring system in the transition state. The following will focus
on explaining the stereoselectivities of the trapping of radicals
2 and 6 by using transition-state models A‡-D‡. To rationalize
the complex temperature-selectivity dependence that was
observed upon variation of R2 and R3 in terms of activation
enthalpies and entropies, relative Eyring eq 1 is utilized
The selectivity of H-transfer upon radicals 2j-l does not show
a significant dependence on temperature; thus they have to be
classified as an example for the third case in terms of eq 1. As
no enthalpic effect occurs throughout the whole temperature
range examined, ∆H‡anti and ∆H‡syn have to be assumed as being
equal, leaving the entropic term to direct the stereoselectivity.
Support for this is provided by the calculated energies of Table
8: the energy differences between conformations A and B and
between conformations C and D are becoming smaller with
decreasing steric effect of R2 ) tBu > cHex > Et > Me, a fact
that should also apply to the corresponding transition states.
Accordingly, radicals 2j, 2k, and 2l bearing R2 groups of lesser
steric impact are trapped with negligible (2j, 2l) or without
temperature dependence (2k), but remarkable anti-selectivity
with transition-state B‡ contributing the most to this product
distribution.
Changing the alkyl ester moiety from R3 ) Me in radicals
2a and 2e to R3 ) tBu in radicals 6a and 6b, respectively, clearly
revealed the influence of the steric effect of alkyl-group R3 on
the stereochemical outcome of the reaction. In the case of
radicals 2a and 6a, the slight decrease of syn-selectivity of 2:98
(2a, Table 3, entry 5) to 4:98 (6a, Table 7, entry 3) can be
rationalized by unfavorable 1,4-steric interactions of R2, R3 )
tBu in radical 6a in transition-state A‡, which is supported by
calculations showing that the energetic difference of conforma-
tions A and B of radical 6a amounts to only 6.9 kJ/mol
compared to 9.2 kJ/mol of radical 2a (Table 8). Thus, increasing
the steric effect of the alkyl ester moiety R3 favors the formation
of the anti-product. Further confirmation of this was provided
by the results of radical 6b, which showed the highest and
temperature-independent anti-selectivity of 89:11 measured in
the reaction series also at -78 °C (Table 7, entry 7) in contrast
to the corresponding methyl ester radical 2e, which reacted
unselectively at -78 °C (Table 5, entry 9). Obviously, the
temperature-independent anti-selectivity of the trapping of
radical 6b can be explained analogously as for radicals 2j-l.
The calculations show that the stabilities of conformers A and
B are reversed going from radical 2e to 6b (Table 8); transferred
to transition states, this implies ∆H‡syn g ∆H‡anti for radical 6b
in contrast to radical 2e.
additionally. From the results given in Table 3 for radical 2a,
we can assume that in the temperature range of -78 to ∼80 °C
the structure of the chelated radicals 2 remains essentially
unchanged because of a reasonable Arrhenius-like linear
behavior of stereoselectivity.
With respect to selectivity-temperature behavior, three
different cases that occurred have to be viewed: (1) decreasing
selectivity by performing the reaction at higher temperatures,
(2) increasing selectivity by doing so, and (3) no temperature
dependence of selectivity. Radicals 2a-d and 6a are following
the first case; they are exhibiting a high syn-preference in the
hydrogen-transfer step and a temperature-selectivity profile as
expected, i.e., decreasing selectivity by rising the temperature.
By reference to eq 1, this clearly establishes a dominating
enthalpic effect with activation enthalpy ∆H‡syn obviously being
lower than ∆H‡anti. A comparison of A‡ to the other possible
transition states of H-transfer by a steric viewpoint clearly
reveals that in A‡ steric interactions of the hydrogen donor with
the concave side of the chelate ring system are weaker than
those in C‡, taking into account the additional effect of R1, and
also weaker than the steric repulsion between the tBu group
and the chelate ring system in transition-state B‡ and even more
so in D‡ where R1 enhances steric congestion. Thus, H-transfer
proceeds mainly via transition-state A‡, furnishing syn-product
predominantly. This interpretation also explains the low influ-
ence of R1 (only minor contributions of pathways C‡ and D‡)
on the stereoselectivity (Tables 3 and 4) and is supportedsin
compliance with the Curtin-Hammett-principlesby the results
of the calculations of radical 2a, which reveal A as being the
most stable conformer (Table 8).
Radicals 2e-h (Table 5) with R2 ) cHex of medium steric
impact are an example of the second case. While they are
trapped nearly unselectively at -78 °C, at +70 °C the reactions
take place with remarkable anti-selectivity. Compared to radicals
2a-d and 6a, they are displaying a completely reversed
behavior, with respect to both stereoselective preference and
temperature dependence. An interpretation of these findings in
terms of eq 1 is as follows: similar to the case surveyed above,
Although the reactions performed in the absence of a Lewis
acid did not lead to interesting findings similar to those of the
chelation-controlled ones, some general tendencies appeared that
(17) (a) Giese, B. Angew. Chem. 1977, 89, 162-173; Angew. Chem.,
Int. Ed. Engl. 1977, 16, 125-136. (b) Giese, B. Acc. Chem. Res. 1984, 17,
438-442.
∆H‡ > ∆H‡syn, because with increasing temperature more
anti

1,3-Stereoinduction in Radical Reactions
J. Am. Chem. Soc., Vol. 122, No. 50, 2000 12465
H‡ is entropically more favorable, especially in the case of the
tert-butyl esters.
Conclusions
Our investigation of the tin hydride-mediated radical addi-
tion-hydrogen transfer sequence reactions to 2-alkyl-4-meth-
ylene pentanedioates 1, exhibiting different structural features,
clearly establishes the scope of 1,3-stereoinduction as a powerful
synthetic tool. The findings show that chelation control has a
strong regulating influence. The extensive study reveals a
significant effect of the R-substituent at the radical center and
minor effects of the γ-substituents and the alkyl group of the
ester moieties with respect to selectivity. Apart from the
structural effects, the examination of the temperature dependence
of the selectivity adds another interesting feature to the
performed reaction series, revealing strong entropic influences
on the stereochemical outcome.
Figure 2. Transition-state models of hydrogen transfer to unchelated
radicals 2a-d and 6a. Syn-products are thought to be formed
preferentially via E‡, anti-products via F‡.
Experimental Section
General Methods. All radical reactions were performed under argon
in reaction vessels that were previously evacuated, heated, and flushed
with inert gas. All commercially available reagents were employed as
supplied; solvents were dried and distilled according to standard
procedures. For column chromatography, Merck 60 silica gel, 63-
200 µm, was used. Analytical GC was performed on a Carlo Erba
HRGC with FID detector and fused-silica capillary column DB1 30
m. 1H and 13C NMR spectra were recorded in CDCl3 (unless otherwise
noted) on a Bruker AM 300 or a Bruker AM 500 spectrometer at 20
°C using TMS (1H NMR) and CDCl3 (δ ) 77.0 ppm, 13C NMR) as
internal standard. Mass spectra were recorded on a Finnigan MAT 212.
Elemental analysis was conducted by Mikroanalytisches Labor Beller,
D-37004 Go¨ttingen, Germany.
Calculations. The quantum chemical calculations have been carried
out using the PM3 Hamiltonian18 and the unrestricted Hartree-Fock
(UHF) approach. The size of the investigated systems renders ab initio
calculations not feasible. All structures were minimized with the default
options available within the MOPAC 93 program package.19 Lithium
as Lewis acid was chosen because of the availability of more reliable
Li parameters for PM3,20 compared to Mg parameters.18
Crystallographic Data of 8a and 8b. Crystallographic data (CIF-
files) of 8a and 8b have been deposited with the Cambridge Crystal-
lographic Data Centre as CCDC-137280 and CCDC-137281. Copies
of the data can be obtained free of charge from the CCDC, 12 Union
Rd., Cambridge CB21EZ, U.K.
Dimethyl 2-Alkyl-4-methylenepentanedioates (1a-d). Compounds
1a-d were synthesized by a reaction sequence starting with the
alkylation of commercially available dimethyl malonate, followed by
a saponification/decarboxylation process according to Stetter,14a yielding
R-alkyl acrylic esters. Michael addition of dimethyl malonate to these
esters then led to trimethyl 1,1,3-alkanetricarboxylates, which provided
educts 1a-d upon Stetter reaction.
Figure 3. Transition-state models of hydrogen transfer to unchelated
radicals 2e-k and 6b. Syn-products are thought to be formed
preferentially via H‡, anti-products via G‡.
are worth a short discussion and can reasonably be explained
by considering steric effects in the transition states.6b To start
off with the outcome of the neopentyl series, the observed
diastereoselectivities can be understood by considering transi-
tion-states E‡ and F‡ (Figure 2). While radical 2a reacts
completely unselectively (Table 3, entries 1-4), radicals 2b-d
(Table 4, entries 1-6) are trapped with low syn-selectivities at
-78 °Cswhich become larger with increasing steric demand
of alkyl substituent R1 (nPr < iPr < cHex)sand at 80 °C
inversion to low anti-selectivities takes place. E‡ and F‡ are
based on the assumption that the steric impact of the neopentyl
moiety renders the transition states with the depicted downward
orientation of this group the most likely so that hydrogen transfer
occurs from the upper side of the radical.
∆H‡syn < ∆H‡anti because in transition-state E‡ steric interac-
tions of the R- and γ-substituents are smaller compared to F‡.
When R1 becomes sterically more demanding, transition-state
E‡ gets more favorable, δ∆H‡ increases, and thus more syn-
product is formed. In terms of eq 1, enthalpic and entropic
factors are opposed, which explains the selectivity change from
general syn- to general anti-preference at higher reaction
temperature. The same effect is observed by increasing the steric
demand of alkyl ester group R3. Thus, radical 6a is trapped
with low syn-selectivity (Table 7, entries 1 and 2) in contrast
to the unselective trapping of radical 2a.
Dimethyl 2-Iodo-2,4-dimethylpentanedioate (4). Under argon
atmosphere, n-butyllithium (3.6 mL, 5.8 mmol; 1.6 m solution in
hexane) was added dropwise at -78 °C to a solution of freshly distilled
diisopropylamine (0.85 mL, 6.1 mmol) in 25 mL of dry THF; the
mixture was stirred for 30 min. The temperature was allowed to rise
to 0 °C for 15 min; after that the flask was cooled to -100 °C. A
solution of dimethyl 2,4-dimethylpentanedioate (1 g, 5.18 mmol, 1:1
diastereomeric mixture of meso- and (()-form) in 25 mL of dry THF
was introduced dropwise, and the mixture was stirred for 1.5 h while
the temperature was kept at -100°. The enolate solution was then
decanted into a precooled dropping funnel and by means of this added
to a previously to -78 °C cooled solution of iodine (2 g, 7.9 mmol) in
Trapping of radicals 2e-l occurs with low anti-selectivity,
which decreases with increasing temperature until inversion to
syn-selectivity is observed in the case of radicals 2j-l (Table
6). Interestingly, γ-alkyl substituent R1 exhibits no significant
influence on the selectivity (Table 5), which indicates that only
transition states with R1 pointing backward have to be consid-
ered (Figure 3). Transition-state G‡, providing anti-3e-l is
energetically favored due to lesser steric congestion than in H‡,
while syn-3e-l are formed with higher activation energy via
H‡. This model was applied before for an explanation of 1,3-
stereoinduction of iodine transfer to alkyl radicals.6b
(18) (a) Stewart, J. J. P. J. Comput. Chem. 1989, 10, 209-220. (b)
Stewart, J. J. P.; J. Comput. Chem. 1991, 12, 320-341.
(19) Stewart, J. J. P. QCPE 1993, Program 455 (version 93).
(20) Anders, E.; Koch, R.; Freunscht, P. J. Comput. Chem. 1993, 14,
1301-1312.
Remarkably, in the case of radical 6b (R3 ) tBu), syn-
selectivity increases from -78 °Csalmost unselectivesto [anti]:
[syn] ) 42:58 at 80 °C (Table 7). Obviously, transition-state

12466 J. Am. Chem. Soc., Vol. 122, No. 50, 2000
Hayen et al.
50 mL of dry THF. Stirring was maintained for 2 h, and then HCl
(10% aqueous, 5.3 mL) was added. Workup included addition of diethyl
ether, extraction with aqueous solutions of Na2S2O3, NaHCO3, and
brine, drying with MgSO4, and evaporation, which afforded 4 in a
diastereomeric ratio of 1:1 as an instable light brown liquid, which
decomposed visibly, turning dark brown upon storage under argon at
8 °C. 4 was used without further purification due to the instability:
Workup Procedure. Conditions A-C. GC analysis for determi-
nation of diastereomeric ratios was performed on the crude reaction
mixtures after filtration of the sample through a 1-cm layer of silica
gel. For general workup, the reaction mixture was diluted with 10 mL
of saturated aqueous NaHCO3 and stirred for 2 h. The organic layer
was decanted and the aqueous layer extracted 2× with CH2Cl2 or
benzene, respectively. The organic extracts were combined, successively
washed with water, 1 m HCl, and brine, and dried (MgSO4). After
evaporation, products were purified by chromatography on silica gel
using (1) hexane and (2) petroleum ether 60/80/ethyl acetate (9.5:0.5)
as eluent. All dialkyl 2,4-dialkylpentanedioates were obtained as
colorless liquids.
In an alternative workup procedure, 1 equiv of Me3Al (2 m solution
in hexane, in reference to Bu3SnH) was added to the reaction mixture
and the resultant mixture was stirred for 2 h. After hydrolysis with 15
mL of 2 N NaOH, stirring was continued for another 2 h. The organic
layer was decanted, and the aqueous layer was extracted 2× with the
appropriate solvent. The combined organic extracts were washed with
water, dried (MgSO4), and evaporated. To separate methyltributyltin
from the products, the residue was subjected to a filtration through a
short pad of silica gel, using n-hexane (150 mL) as eluent followed by
n-hexane/ethyl acetate (200 mL, 4:1) to elute the fraction containing
the products. After evaporation, products were purified by chromatog-
raphy on silica gel using (1) hexane and (2) petroleum ether 60/80/
ethyl acetate (9.5:0.5) as eluent.
1
yield 1.13 g (68%); H NMR (500.1 MHz, diastereomer 1) δ 1.18 (d,
3J ) 7.13 Hz, 3H, CH3CH), 2.04 (s, 3H, CH3CI), 2.20 (dd, 3J ) 13.18
Hz, 2J ) 8.78 Hz, 1H, CHCHHCH), 2.43-2.52 (m, 1H, CHCHHCH),
2.73-2.82 (m, 1H, CH3CH), 3.65, 3.75 (2CO2CH3); 13C NMR (125.8
MHz, diastereomer 1) δ 18.89 (CH3CH), 29.88 (CH3CI), 37.49 (CI),
37.98 (CH3CH), 47.97 (CH2), 51.88, 52.98 (2CO2CH3), 172.94, 176.00
1
3
(2CO2CH3); H NMR (500.1 MHz, diastereomer 2) δ 1.24 (d, J )
6.58 Hz, 3H, CH3CH), 2.05 (s, 3H, CH3CI), 2.28 (dd, 3J ) 14.55 Hz,
2J ) 2.47 Hz, 1H, CHCHHCH), 2.38-2.43 (m, 1H, CHCHHCH),
2.60-2.69 (m, 1H, CH3CH); 13C NMR (125.8 MHz, diastereomer 2)
δ 19.08 (CH3CH), 30.31 (CH3CI), 38.66 (CH3CH), 39.18 (CI), 47.42
(CH2), 52.84, 53.00 (2CO2CH3), 173.02, 176.24 (2CO2CH3); MS (CI,
isobutane) m/z (%) 315 (85) [MH+], 283 (100) [MH+ - OCH3], 187
(32) [M+ - I].
Di-tert-butyl 2-Methyl-4-methylenepentanedioate (5). A mixture
of N,N′-diisopropylcarbodiimide (8 mL, 51 mmol), tert-butyl alcohol
(5.6 mL, 59 mmol), and CuCl (0.05 g, 0.51 mmol) was stirred under
argon at room temperature for 3 days. After dilution with CH2Cl2 (30
mL), it was added dropwise to a solution of 2-methyl-4-methylene-
pentanedioic acid (1.34 g, 8.5 mmol) in CHCl3 (50 mL). The pre-
cipitated N,N′-diisopropylurea was filtered off and washed with CH2-
Cl2, and the filtrate was evaporated. Purification by column chroma-
tography afforded 5 as a colorless liquid: yield 31%; Rf ) 0.35
(petroleum ether 60/80/EtOAc 9.5/0.5); 1H NMR (300.1 MHz) δ 1.11
(d, 3J (H,H) ) 6.63 Hz, 3H, CH3), 1.42 (s, 9H, C(CH3)3), 1.50 (s, 9H,
C(CH3)3), 2.20-2.39 (m, 1H, CHCH3), 2.50-2.68 (m, 2H, CH2), 5.48
Reaction of Dimethyl 2,4-Dimethyl-2-iodopentanedioate (4) with
Bu3SnH. Reactions at -78 and 70 °C were carried out according to
conditions A and C, leaving out the addition of alkyl halides.
Dimethyl 2-methyl-4-neopentylpentanedioate (3a): 1H NMR
3
(300.1 MHz, syn-3a) δ 0.86 (s, 9H, C(CH3)3), 1.19 (d, J (H,H) )
6.99 Hz, 3H, CHCH3), 1.10-1.30 (m, 1H, (CH3)3CCHH), 1.40-1.52
(m, 1H, CHCHHCH), 1.72-1.85 (m, 1H, (CH3)3CCHH), 1.95-2.08
(m, 1H, CHCHHCH), 2.35-2.53 (m, 2H, 2CHCO2Me), 3.66 (s, 3H,
CO2CH3), 3.67 (s, 3H, CO2CH3); 13C NMR (75.5 MHz, syn-3a) δ 17.37
(CH3), 29.67 (C(CH3)3), 31.12 (C(CH3)3), 38.00 (CHCH2CH), 37.82,
40.05 (2CHCO2Me), 46.62 ((CH3)3CCH2), 52.03 (2CO2CH3), 176.86,
177.56 (2CO2CH3); 1H NMR (300.1 MHz, anti-3a) δ 0.87 (s, 9H,
2
2
(d, J (H,H) ) 1.58 Hz, 1H, dCHH), 6.19 (d, J (H,H) ) 1.58 Hz,
1H, dCHH); 13C NMR (75.5 MHz) δ 17.05 (CH3CH), 28.02 (C(CH3)3),
36.43 (CH2), 39.21 (CH3CH), 80.06, 80.53 (C(CH3)3), 125.55 (CH2d
C), 139.71 (CH2)C), 166.05 (dCCO2tBu), 175.42 (CHCO2tBu); MS
(CI, isobutane) m/z (%) ) 215 (22) [M+ - C(CH3)3 + 2H], 159 (100)
[M+ - (C(CH3)3)2 + 3H]; C15H26O4 (270.37) calcd C, 66.64; H, 9.69;
found C, 66.97; H, 9.75.
3
C(CH3)3), 1.15 (d, J (H,H) ) 7.0 Hz, 3H, CHCH3), 1.10-1.37 (m,
1H, (CH3)3CCHH), 1.56-1.7 (m, 1H, (CH3)3CCHH), 1.7-1.88 (m,
2H, CHCH2CH), 2.30-2.54 (m, 2H, 2CHCO2Me), 3.66 (s, 3H, CO2-
CH3), 3.69 (s, 3H, CO2CH3); 13C NMR (75.5 MHz, anti-3a) δ 18.50
(CH3), 29.68 (C(CH3)3), 30.07 (C(CH3)3), 39.02 (CHCH2CH), 37.91,
40.30 (CHCO2Me), 47.03 ((CH3)3CCH2), 51.88, 52.01 (2CO2CH3),
176.90, 177.85 (2CO2CH3); MS (CI, isobutane) m/z (%) 245 (100)
[MH+], 213 (68) [MH+ - CH3OH]; C13H24O4 (244.33) calcd C, 63.91;
H, 9.90; found C, 63.90; H, 9.99.
General Procedure for Radical Additions to Alkenes 1a-d and
5 under Chelation-Controlled Conditions. Conditions A (-78, 0,
20, and 40 °C). In a typical run, 5.4 mmol (2 equiv) of MgBr2‚OEt2
was dissolved in 5 mL of dry diethyl ether in an argon-purged reaction
flask. After the mixture was stirred for 15 min at 20 °C, 10 mL of dry
CH2Cl2 and 2.7 mmol of dialkyl 2-alkyl-4-methylenepentanedioate 1
or 5 were added and stirring was continued for at least 30 min. The
solution was cooled to -78 °C, 0 °C, left at ambient temperature, or
heated to reflux and 3 equiv of Bu3SnH and alkyl iodide, respectively,
were added. Via syringe, 5 mL of O2 and 2.7 mmol of Et3B (1 m
solution in hexane) were then injected simultaneously below the surface
of the solution over a time period of 1 h. The reaction mixture was
stirred at the chosen temperature for at least 4 h.
Conditions B (-20 °C). In a typical run, 5.4 mmol (2 equiv) of
MgBr2‚OEt2 was dissolved in 5 mL of dry diethyl ether in an argon-
purged photoreactor. After the mixture was stirred for 15 min at 20
°C, 10 mL of dry CH2Cl2 and 2.7 mmol of dialkyl 2-alkyl-4-methyl-
enepentanedioate 1 or 5 were added and stirring was continued for at
least 30 min. The solution was cooled to -20 °C by means of a cryostat,
and 3 equiv of Bu3SnH and alkyl iodide, respectively, were added.
The reaction mixture was irradiated by a 5-W mercury lamp for 6 h.
Conditions C (70 and 100 °C). In a typical run, 5.4 mmol (2 equiv)
of MgBr2‚OEt2 was dissolved in 2 mL of dry diethyl ether in an argon-
purged reaction flask. After the mixture was stirred for 15 min at 20
°C, 5 mL of dry benzene or toluene and 2.7 mmol of dialkyl 2-alkyl-
4-methylenepentanedioate 1 or 5 were added and stirring was continued
for at least 30 min. The solution was heated to reflux temperature, and
3 equiv of alkyl bromide was added. Via syringe, a solution of 3 equiv
of Bu3SnH and 20 mol % of AIBN in 4 mL of benzene was fed into
the reaction mixture during 2 h. Afterward, reflux was maintained for
another 2-3 h.
Dimethyl 2-neopentyl-4-propylpentanedioate (3b): 1H NMR
(300.1 MHz, syn-3b) δ 0.86 (s, 9H, C(CH3)3), 0.92 (t, 3J (H,H) ) 7.53
Hz, 3H, CH3CH2), 1.12-1.72 (m, 6H, CH3CH2CH2, (CH3)3CCHH,
CHCHHCH), 1.72-1.83 (m, 1H, (CH3)3CCHH)), 1.86-1.99 (m, 1H,
CHCHHCH), 2.32-2.48 (m, 2H, 2CHCO2Me), 3.65 (s, 3H, CO2CH3),
3.66 (s, 3H, CO2CH3); 13C NMR (75.5 MHz, syn-3b) δ 13.82 (CH3-
CH2), 20.35 (CH3CH2), 29.23 (C(CH3)3), 30.61 (C(CH3)3), 34.42,
36.84, (CH3CH2CH2CHCH2), 39.91, 43.16 (2CHCO2Me), 45.78
((CH3)3CCH2), 51.35, 51.47 (2CO2CH3), 176.01, 177.15 (2CO2CH3);
3
1H NMR (300.1 MHz, anti-3b) δ 0.86 (s, 9H, C(CH3)3), 0.88 (t, J
(H,H) ) 7.16 Hz, 3H, CH3CH2), 1.12-1.33 (m, 3H, CH3CH2,
(CH3)3CCHH), 1.33-1.48 (m, 1H, CH3CH2CHH), 1.48-1.63 (m, 1H,
CH3CH2CHH), 1.63-1.82 (m, 3H, (CH3)3CCHH, CHCH2CH), 2.21-
2.34 (m, 1H, CHCO2Me), 2.37-2.49 (m, 1H, CHCO2Me), 3.66 (s, 3H,
CO2CH3), 3.69 (s, 3H, CO2CH3); 13C NMR (75.5 MHz, anti-3b) δ 14.21
(CH3CH2), 20.67 (CH3CH2), 29.68 (C(CH3)3), 31.15 (C(CH3)3), 35.68,
37.58 (CH3CH2CH2CHCH2), 40.47, 43.60 (2CHCO2Me), 47.37
((CH3)3CCH2), 51.77 (2CO2CH3), 176.53, 177.73 (2CO2CH3); MS (CI,
isobutane) m/z (%) 273 (83) [MH+], 241 (100) [MH+ - CH3OH];
C15H28O4 (272.38) calcd C, 66.14; H, 10.36; found C, 66.36; H,
10.12.
Dimethyl 2-isopropyl-4-neopentylpentanedioate (3c): 1H NMR
3
(300.1 MHz, syn-3c) δ 0.85 (s, 9H, (CH3)3C), 0.90 (d, J (H,H) )
3
7.16 Hz, 3H, CH3CHCH3), 0.93 (d, J (H,H) ) 6.78 Hz, 3H, CH3-
CHCH3) 1.18-1.72 (m, 3H, (CH3)3CCHH), CHCHHCH, CH(CH3)2,

1,3-Stereoinduction in Radical Reactions
J. Am. Chem. Soc., Vol. 122, No. 50, 2000 12467
1.72-2.00 (m, 2H, CHCHHCH, (CH3)3CCHH), 2.19-2.38 (m, 2H,
2CHCO2Me), 3.65 (s, 3H, CO2CH3), 3.69 (s, 3H, CO2CH3); 13C NMR
(75.5 MHz, syn-3c) δ 20.35, 20.57 (CH(CH3)2), 29.61 (C(CH3)3), 30.92
(C(CH3)3), 31.14 (CH(CH3)2), 34.37 (CHCH2CH), 40.66, 50.43 (2CHCO2-
Me), 45.48 ((CH3)3CCH2), 51.55, 51.92 (2CO2CH3), 175.64, 177.72
33.27, 35.03, 35.29, 35.44 (cyclohexyl-C, CH3CH2CH2, CHCH2CH),
40.83 (C6H11CH2), 41.04, 43.42 (2CHCO2Me), 51.38 (2CO2CH3),
176.30, 176.67 (2CO2CH3); MS (CI, isobutane) m/z (%) 299 (100)
[MH+], 267 (47) [MH+ - CH3OH]; C17H30O4 (298.42) calcd C, 68.42;
H 10.13; found C, 68.56; H, 10.01.
1
Dimethyl 2-(cyclohexylmethyl)-4-isopropylpentanedioate (3g): 1H
NMR (500.1 MHz, syn-3g) δ 0.75-0.95 (m, 2H, cyclohexyl: H2, H6),
(2CO2CH3); H NMR (300.1 MHz, anti-3c) δ 0.86 (s, 9H, C(CH3)3),
3
3
0.88 (d, J (H,H) ) 6.75 Hz, 3H, CH3CHCH3), 0.90 (d, J (H,H) )
6.57 Hz, 3H, CH3CHCH3), 1.24 (dd, 3J (H,H) ) 14.04 Hz, 2J (H,H) )
3.08 Hz, 1H, ((CH3)3CCHH), 1.64-1.92 (m, 4H, (CHCH2CH, (CH3)2CH,
(CH3)3CCHH), 2.05-2.15 (m, 1H, (CHCO2Me), 2.33-2.44 (m, 1H,
CHCO2Me), 3.65 (s, 3H, CO2CH3), 3.69 (s, 3H, CO2CH3); 13C NMR
(75.5 MHz, anti-3c) δ 19.84, 20.10 (C(CH3)2), 29.29 (C(CH3)3), 30.78
(C(CH3)3), 31.03 (CH(CH3)2), 34.37 (CHCH2CH), 40.33, 49.96 (2CHCO2-
Me), 47.33 ((CH3)3CCH2), 51.18, 51.35 (2CO2CH3), 175.41, 177.40
(2CO2CH3); MS (CI, isobutane) m/z (%) 273 (79) [MH+], 241 (100)
[MH+ - CH3OH]; C15H28O4 (272.38) calcd C, 66.14; H, 10.36; found
C, 66.29; H, 10.40.
3
3
0.91 (d, J (H,H) ) 6.59 Hz, 3H, CH3CHCH3), 0.93 (d, J (H,H) )
7.19 Hz, 3H, CH3CHCH3), 1.08-1.41 (m, 5H, cyclohexyl: H1, H3,
H4, H5, C6H11CHH), 1.50-1.58 (m, 1H, C6H11CHH), 1.58-1.70 (m,
5H, cyclohexyl: H2, H3, H4, H5, H6), 1.70-1.78 (m, 1H, CHCH-
HCH), 1.80-1.89 (m, 1H, (CH3)2CH), 1.90-1.99 (m, 1H, CHCHHCH),
2.18-2.24 (m, 1H, (CH3)2CHCH), 2.35-2.43 (m, 1H, C6H5CH2CHCH2),
3.65 (s, 3H, CO2CH3), 3.67 (s, 3H, CO2CH3); 13C NMR (125.8 MHz,
syn-3g) δ 20.09, 20.20 (CH(CH3)2), 26.08, 26.17, 26.46, 32.22, 32.48,
33.82, (cyclohexyl-C, CHCH2CH), 30.70 (CH(CH3)2), 35.60 (cyclo-
hexyl: C1), 39.41 (C6H11CH2), 41.29 (CH2CHCH2), 50.15 ((CH3)2-
CHCH), 51.55, 51.73 (2CO2CH3), 175.97, 177.03 (2CO2CH3); 1H NMR
(500.1 MHz, anti-3g) δ 0.78-0.93 (m, 2H, cyclohexyl: H2, H6), 0.90
Dimethyl 2-cyclohexyl-4-neopentylpentanedioate (3d): 1H NMR
(300.1 MHz, syn-3d) δ 0.84 (s, 9H, C(CH3)3), 0.84-1.82 (m, 14H,
cyclohexyl-H, C6H11CH2, CHCHHCH), 1.85-1.97 (m, 1H, CHCH-
HCH), 2.21-2.35 (m, 2H, 2CHCO2Me), 3.64 (s, 3H, CO2CH3), 3.66
(s, 3H, CO2CH3); 13C NMR (75.5 MHz, syn-3d) δ 26.66, 31.02
(cyclohexyl-C), 29.61 (C(CH3)3), 30.92 (C(CH3)3), 34.27 (CHCH2CH),
40.71, 49.90 (2CHCO2Me), 40.93 (cyclohexyl: C1), 45.40 ((CH3)3CCH2),
51.55, 51.93 (2CO2CH3), 175.75, 177.74 (2CO2CH3); 1H NMR (300.1
MHz, anti-3d) δ 0.86 (s, 9H, C(CH3)3), 0.85-1.29 (m, 5H, cyclo-
hexyl: H2, H3, H4, H5. H6), 1.4-1.82 (m, 12 H, cyclohexyl: H1,
H2, H3, H4, H5, H6, C6H11CH2CHCH2CHCH2), 2.10-2.12 (m, 1H,
CHCO2Me), 2.29-2.43 (m, 1H, CHCO2Me), 3.65 (s, 3H, CO2CH3),
3.68 (s, 3H, CO2CH3); 13C NMR (75.5 MHz, anti-3d) δ 26.28, 30.38,
30.65, (cyclohexyl-C), 29.30 (C(CH3)3), 30.79 (C(CH3)3), 34.49
(CHCH2CH), 40.39, 49.50 (2CHCO2Me), 40.81 (cyclohexyl: C1),
47.32 ((CH3)3CCH2), 51.18, 51.36 (2CO2CH3), 175.50, 177.34 (2CO2-
CH3); MS (CI, isobutane) m/z (%) 313 (100) [MH+], 281 (98) [MH+
- CH3OH]; C15H28O4 (312.45) calcd C, 69.19; H, 10.32; found C,
69.13; H, 10.25.
3
3
(d, J (H,H) ) 6.58 Hz, 3H, CH3CHCH3), 0.93 (d, J (H,H) ) 7.14
Hz, 3H, CH3CHCH3), 1.06-1.40 (m, 5H, cyclohexyl: H1, H3, H4,
H5, C6H11CHH), 1.48-1.56 (m, 1H, C6H11CHH), 1.59-1.78 (m, 7H,
cyclohexyl: H2, H3, H4, H5, H6, CHCH2CH), 1.79-1.89 (m, 1H,
(CH3)2CH), 2.06-2.13 (m, 1H, (CH3)2CHCH), 2.36-2.45 (m, 1H,
C6H5CH2CHCH2), 3.66 (s, 3H, CO2CH3), 3.67 (s, 3H, CO2CH3); 13C
NMR (125.8 MHz, anti-3g) δ 19.85, 20.24 (CH(CH3)2), 31.01 (CH-
(CH3)2,), 35.42 (cyclohexyl: C1), 26.14, 26.47, 32.15, 33.06, 33.19
(cyclohexyl-C, CHCH2CH), 41.04 (C6H11CH2), 41.27 (CH2CHCH2),
50.30 ((CH3)2CHCH), 51.55, 51.73 (2CO2CH3), 175.97, 177.03 (2CO2-
CH3); MS (CI, isobutane) m/z (%) 299 (42) [MH+], 267 (100) [MH+
- CH3OH]; C17H30O4 (298.42) calcd C, 68.42; H, 10.13; found C,
68.56; H, 10.03.
Dimethyl 2-cyclohexyl-4-(cyclohexylmethyl)pentanedioate (3h):
1H NMR (500.1 MHz, syn-3h) δ 0.72-1.77 (m, 25H, cyclohexyl-H,
C6H11CH2, CHCHHCH), 1.85-1.94 (m, 1H, CHCHHCH), 2.17-2.24
(m, 1H, CHCO2Me), 2.31-2.39 (m, 1H, CHCO2Me), 3.62 (s, 3H, CO2-
CH3), 3.63 (s, 3H, CO2CH3); 13C NMR (125.8 MHz, syn-3h) δ 26.09,
26.18, 26.23, 26.26, 26.47, 29.11, 30.57, 30.65, 32.34, 32.50, (cyclo-
hexyl-C), 33.84 (CHCH2CH), 35.60, 40.44 (2cyclohexyl: C1), 39.35
(C6H11CH2), 41.33, 49.57 (2CHCO2Me), 51.17, 51.44 (2CO2CH3),
1
Dimethyl 2-(cyclohexylmethyl)-4-methylpentanedioate (3e): H
NMR (300.1 MHz, syn-3e) δ 0.73-1.02 (m, 2H, cyclohexyl: H2, H6),
1.03-1.42 (m, 5H, cyclohexyl: H1, H3, H4, H5, C6H11CHH), 1.16
3
(d, J (H,H) ) 7.16 Hz, 3H, CH3), 1.45-1.57 (m, 1H, C6H11CHH),
1
175.48, 176.54 (2CO2CH3); H NMR (500.1 MHz, anti-3h) δ 0.77-
1.57-1.82 (m, 6H, cyclohexyl: H2, H3, H4, H5, H6, CHCHHCH),
1.96-2.08 (m, 1H, CHCHHCH), 2.37-2.60 (m, 2H, 2CHCO2Me), 3.67
(s, 6H, 2CO2CH3); 13C NMR (75.5 MHz, syn-3e) δ 18.02 (CH3), 26.55,
26.86, 33.24, 35.93, 37.90 (CHCH2CH, 4cyclohexyl-C), 40.72
(C6H11CH2), 36.73, 37.96, 41.00 (2CHCO2Me, cyclohexyl: C1), 51.83,
51.93 (2CO2CH3), 176.89, 177.05 (2CO2CH3); 1H NMR (300.1 MHz,
anti-3e) δ 0.73-0.97 (m, 2H, cyclohexyl: H2, H6), 1.03-1.42 (m,
1.81 (m, 26H, cyclohexyl-H, C6H11CH2, CHCH2CH), 2.08-2.14 (m,
1H, CHCO2Me), 2.39-2.42 (m, 1H, CHCO2Me), 3.66 (s, 3H, CO2-
CH3), 3.67 (s, 3H, CO2CH3); 13C NMR (125.8 MHz, anti-3h) δ 26.13,
26.16, 26.23, 26.47, 26.58, 29.11, 30.35, 30.76, 32.16, 33.09, 33.15
(cyclohexyl-C, CHCH2CH), 35.39, 40.70 (2cyclohexyl: C1), 41.03
(C6H11CH2), 41.27, 49.73 (2CHCO2Me), 51.15, 51.34 (2CO2CH3),
175.66, 176.63 (2CO2CH3); MS (CI, isobutane) m/z (%) 339 (76)
[MH+], 307 (100) [MH+ - CH3OH]; C17H30O4 (338.25) calcd C, 70.95;
H, 10.13; found C, 71.25; H, 10.16.
3
5H, cyclohexyl: H1, H3, H4, H5, C6H11CHH), 1.15 (d, J (H,H) )
6.78 Hz, 3H, CH3), 1.45-1.88 (m, 8H, cyclohexyl: H2, H3, H4, H5,
H6, C6H11CHH, CHCH2CH), 2.34-2.58 (m, 2H, 2CHCO2Me), 3.67
(s, 6H, 2CO2CH3); 13C NMR (75.5 MHz, anti-3e) δ 18.40 (CH3), 26.55,
26.87, 33.39, 33.72 (cyclohexyl-C), 36.95 (CHCH2CH), 40.99
(C6H11CH2), 35.85, 38.11, 41.34 (2CHCO2Me, cyclohexyl: C1), 51.78,
51.91 (2CO2CH3), 176.95, 177.07 (2CO2CH3); MS (CI, isobutane) m/z
(%) 271 (60) [MH+], 239 (100) [MH+ - CH3OH]; C15H26O4 (270.37)
calcd C, 66.64; H, 9.69; found C, 66.31; H, 9.46.
Dimethyl 2-ethyl-4-methylpentanedioate (3j): 1H NMR (500.1
3
3
MHz, anti-3j) δ 0.89 (t, J ) 7.41 Hz, 3H, CH3CH2), 1.15 (d, J )
7.13 Hz, 3H, CH3CH), 1.22-1.34 (m, 1H, CH3CHH), 1.48-1.85 (m,
3H, CH3CHH, CHCH2CH), 2.35-2.53 (m, 2H, 2CHCO2Me), 3.67 (s,
6H, 2CO2CH3); 13C NMR (125.8 MHz, anti-3j) δ 11.55 (CH3CH2),
17.61 (CH3CH), 25.94 (CH3CH2), 35.77 (CHCH2CH), 37.73 (CH3CH),
45.07 (CH2CHCH2), 51.35, 51.53 (2CO2CH3), 176.18, 176.64 (2CO2-
1
Dimethyl 2-(cyclohexylmethyl)-4-propylpentanedioate (3f): H
CH3); MS (CI, isobutane) m/z (%) 203 (70) [MH+], 171 (100) [M+
-
NMR (300.1 MHz, syn-3f) δ 0.75-0.98 (m, 2H, cyclohexyl: H2, H6),
0.89 (t, 3J (H,H) ) 7.23 Hz, 3H, CH3CH2), 1.05-1.80 (m, 16H,
cyclohexyl: H1, H2, 2H3, 2H4, 2H5, H6, CHCHHCHCH2CH2,
C6H11CH2), 1.88-2.01 (m, 1H, CHCHHCH), 2.32-2.54 (m, 2H,
2CHCO2Me), 3.66 (s, 6H, 2CO2CH3); 13C NMR (75.5 MHz, syn-3f) δ
13.86 (CH3CH2), 20.39 (CH3CH2), 26.16, 26.46, 32.70, 33.65, 35.09,
35.35 (cyclohexyl-C, CH3CH2CH2, CHCH2CH), 35.55 (cyclohexyl:
C1), 39.98 (C6H11CH2), 40.87, 43.26 (2CHCO2Me), 51.43 (2CO2CH3),
OCH3]; HRMS (CI, isobutane) C10H19O4 [MH+] calcd 203.1283, found
203.1221.
Dimethyl 2-methyl-4-propylpentanedioate (3k): 1H NMR (500.1
3
3
MHz, anti-3k) δ 0.89 (t, J ) 7.41 Hz, 3H, CH3CH2), 1.15 (d, J )
7.13 Hz, 3H, CH3CH), 1.20-1.37 (m, 2H, CH3CH2), 1.37-1.48 (m,
1H, CHCHHCH), 1.51-1.75 (m, 2H, CH3CH2CH2), 1.77-1.85 (m,
1H, CHCHHCH), 2.36-2.48 (m, 2H, 2CHCO2Me), 3.67 (s, 6H, 2CO2-
CH3); 13C NMR (125.8 MHz, anti-3k) δ 13.85 (CH3CH2), 18.02 (CH3-
CH), 20.36 (CH3CH2), 35.06, 36.14 (CH2CHCH2), 37.77 (CH3CH),
43.38 (CH2CHCH2), 51.40, 51.56 (2CO2CH3), 176.42, 176.65 (2CO2-
1
176.13, 176.41 (2CO2Me); H NMR (300.1 MHz, anti-3f) δ 0.74-
3
0.93 (m, 2H, cyclohexyl: H2, H6), 0.88 (t, J (H,H) ) 7.19 Hz, 3H,
CH3CH2), 1.06-1.94 (m, 17H, cyclohexyl: H1, H2, 2H3, 2H4, 2H5,
H6, CHCH2CHCH2CH2, C6H11CH2), 2.28-2.38 (m, 1H, CHCO2Me),
2.38-2.50 (m, 1H, CHCO2Me), 3.65 (s, 6H, 2CO2CH3); 13C NMR (75.5
MHz, anti-3f) δ 13.83 (CH3CH2), 20.36 (CH3CH2), 26.17, 26.48, 33.05,
CH3); MS (CI, isobutane) m/z (%) 217 (46) [MH+], 185 (100) [M+
-
OCH3]; HRMS (CI, isobutane) C11H21O4 [MH+] calcd 217.1440, found
217.1364.

12468 J. Am. Chem. Soc., Vol. 122, No. 50, 2000
Hayen et al.
1
Di-tert-butyl 2-methyl-4-neopentylpentanedioate (7a): H NMR
tane) m/z (%) 355 (36) [MH+], 243 (100) [MH+ - (C4H9)2 + 2H];
C21H38O4 (354.53) calcd C, 71.15; H, 10.80; found C, 71.05; H, 10.91.
2-Methyl-4-neopentylpentanedioic acid (8a): Mp 158 °C; 1H NMR
(500.1 MHz, CD3OD, syn-8a) δ 0.90 (s, 9H, (CH3)3C), 1.18 (d, 3J (H,H)
) 7.13 Hz, 3H, CH3), 1.27 (dd, 2J (H,H) ) 2.19 Hz, 3J (H,H) ) 14.27
Hz, 1H, (CH3)3CCHH), 1.40-1.47 (m, 1H, CHCHHCH), 1.76 (dd, 2J
(H,H) ) 2.19 Hz, 3J (H,H) ) 9.88 Hz, 1H, (CH3)3CCHH), 1.92-1.99
(m, 1H, CHCHHCH), 2.36-2.50 (m, 2H, 2CHCO2Me); 13C NMR
(125.8 MHz, CD3OD, syn-8a) δ 17.29 (CH3), 29.86 (C(CH3)3), 31.65
(C(CH3)3), 38.57, 39.62, 40.99 (CHCH2CH), 47.22 ((CH3)3CCH2),
179.83, 180.74 (2CO2H); MS (CI, isobutane) m/z (%) ) 217 (49)
[MH+], 199 (100) [MH+ - 18].
3
(300.1 MHz, syn-7a) δ 0.88 (s, 9H, C(CH3)3), 1.14 (d, J (H,H) )
6.79 Hz, 3H, CH3), 1.44 (s, 18H, 2CO2(C(CH3)3), 1.03-1.50 (m, 2H,
CHHCHCHHCH), 1.70-1.82 (m, 1H, CHHC(CH3)3), 1.84-1.96 (m,
1H, CHCHHCH), 2.18-2.41 (m, 2H, 2CHCO2tBu); 13C NMR (75.48
MHz, syn-7a) δ 16.76 (CH3), 28.01, 29.47 (3C(CH3)3), 30.77
(CH2C(CH3)3), 38.25, 40.47 (2CHCO2tBu), 38.76 (CHCH2CH), 45.82
((CH3)3CCH2), 79.95 (2CO2C(CH3)3), 175.53, 176.09 (2CO2C(CH3)3);
1H NMR (300.1 MHz, anti-7a) δ 0.88 (s, 9H, C(CH3)3), 1.10 (d, J
3
(H,H) ) 7.16 Hz, 3H, CH3), 1.46 (s, 18H, 2CO2C(CH3)3), 1.10-1.60
(m, 2H, CH2C(CH3)3), 1.63-1.80 (m, 2H, CHCH2CH), 2.23-2.40 (m,
2H, 2CHCO2tBu); 13C NMR (75.48 MHz, anti-7a) δ 18.46 (CH3),
28.01, 29.47 (3C(CH3)3), 30.86 (CH2C(CH3)3), 38.69, 40.98 (2CHCO2-
tBu), 39.46 (CHCH2CH), 46.92 ((CH3)3CCH2), 79.95 (2CO2C(CH3)3),
175.60, 176.55 (2CO2C(CH3)3); MS (CI, isobutane) m/z (%) 329 (26)
[MH+], 217 (100) [MH+ - (C4H9)2 + 2H]; C19H36O4 (328.49) calcd
C, 69.47; H, 11.05; found C, 69.56; H, 11.05.
2-(Cyclohexylmethyl)-4-methylpentanedioic acid (8b): Mp 159 °C;
1H NMR (500.1 MHz, CD3OD, syn-8b) δ 0.81-0.92 (m, 2H,
3
cyclohexyl: H2, H6), 1.15 (d, J (H,H) ) 7.14 Hz, 3H, CH3), 1.10-
1.34 (m, 5H, cyclohexyl: H2, H3, H5, H6, C6H11CHH), 1.44-1.57
(m, 2H, cyclohexyl: H4, C6H11CH2CHCHHCH), 1.61-1.75 (m, 4H,
cyclohexyl: H1, H3, H4, H5), 1.78-1.85 (m, 1H, C6H11CHH), 1.92-
2.02 (m, 1H, C6H11CH2CHCHHCH), 2.38-2.46 (m, 1H, CHCO2H),
2.47-2.55 (m, 1H, CHCO2H); 13C NMR (125.8 MHz, CD3OD, syn-
8b) δ 17.19 (CH3), 27.36, 27.62, 33.90, 34.92, 37.70 (CHCH2CH,
cyclohexyl-C), 36.99 (cyclohexyl: C1), 38.60, 41.80 (CHCH2CH,),
41.47 (C6H11CH2), 179.83, 179.92 (2CO2H); MS (CI, isobutane) m/z
(%) 243 (59) [MH+], 225 (100) [MH+ - 18].
Di-tert-butyl 2-(cyclohexylmethyl)-4-methylpentanedioate (7b):
1H NMR (500.1 MHz, anti-7b) δ 0.77-0.96 (m, 2H, cyclohexyl: H2,
H6), 1.07-1.83 (m, 13H, cyclohexyl: H1, H2, 2H3, 2H4, 2H5, H6,
3
C6H11CH2, CHCH2CH), 1.10 (d, J (H,H) ) 6.59 Hz, 3H, CH3), 1.46
(s, 18H, 2CO2C(CH3)3), 2.26-2.44 (m, 2H, 2CHCO2tBu); 13C NMR
(125.8 MHz, anti-7b) δ 18.32 (CH3), 26.23, 26.52, 26.79, 32.76, 33.57,
(cyclohexyl-C), 28.03 (2CO2C(CH3)3), 35.48 (cyclohexyl: C1), 37.22
(CHCH2CH), 41.00 (C6H11CH2), 38.82, 41.83 (2CHCO2tBu), 79.90,
79.98 (2CO2C(CH3)3), 175.78, 175.80 (2CO2C(CH3)3); MS (CI, isobu-
JA003235J