Mechanistic study of the reaction of 1,1-dihalo-2-methyl-2-phenylpropanes with LDA. Evidence for radical and carbene pathways

Article abstract of DOI:10.1021/jo951299k

An attempt was made to determine the mechanisms involved in the reactions of the model systems 1,1-dichloro-2-methyl-2-phenylpropane (1) and 1,1-diiodo-2-methyl-2-phenylpropane (2) with LDA.These systems were chosen as ones capable of providing evidence for the formation of radical as well as carbene products.The techniques employed in investigating the mechanistic features of these reactions involved studying the effect of the leaving group, the effect of radical and carbene trapping agents on the product distribution, and isotopic tracer experiments using labeled solvent (THF-d8) and nucleophile (LDA-d2).The major product of the reaction of the geminal dichloride (1) is thought to be derived from a chlorocarbene, whereas the geminal diiodide (2) appears to form products derived from both carbene and radical intermediates.On the basis of the results of radical trapping experiments and those of deuterium-labeling experiments, evidence is presented to support the notion that products A, E, and H are derived from a radical precursor.In addition, products A and H are also believed to be formed from the vinilyc halide D (or B) and the monoiodide E, respectively.Reasonable mechanisms for the formation of the other products formed in these reactions have been proposed on the basis of the available data.

Full text of DOI:10.1021/jo951299k

1322
J . Org. Chem. 1996, 61, 1322-1330
Mech a n istic Stu d y of th e Rea ction s of
1,1-Dih a lo-2-m eth yl-2-p h en ylp r op a n es w ith LDA. Evid en ce for
Ra d ica l a n d Ca r ben e P a th w a ys
E. C. Ashby,* Ali Mehdizadeh, and Abhay K. Deshpande
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332
Received J uly 19, 1995 (Revised Manuscript Received November 30, 1995X)
An attempt was made to determine the mechanisms involved in the reactions of the model systems
1,1-dichloro-2-methyl-2-phenylpropane (1) and 1,1-diiodo-2-methyl-2-phenylpropane (2) with LDA.
These systems were chosen as ones capable of providing evidence for the formation of radical as
well as carbene products. The techniques employed in investigating the mechanistic features of
these reactions involved studying the effect of the leaving group, the effect of radical and carbene
trapping agents on the product distribution, and isotopic tracer experiments using labeled solvent
(THF-d8) and nucleophile (LDA-d2). The major product of the reaction of the geminal dichloride
(1) is thought to be derived from a chlorocarbene, whereas the geminal diiodide (2) appears to
form products derived from both carbene and radical intermediates. On the basis of the results of
radical trapping experiments and those of deuterium-labeling experiments, evidence is presented
to support the notion that products A, E, and H are derived from a radical precursor. In addition,
products A and H are also believed to be formed from the vinylic halide D (or B) and the monoiodide
E, respectively. Reasonable mechanisms for the formation of the other products formed in these
reactions have been proposed on the basis of the available data.
In tr od u ction
the ones shown below, with such nucleophiles as lithium
aluminum hydride,4a sodium trimethyltin,4b,c and lithium
diisopropylamide (LDA)1a,g have been studied extensively
by this group.
Nucleophilic aliphatic substitution reactions of the SN1
and SN2 type involving aliphatic halides (RX) and nu-
cleophiles (Y-) have been widely studied (eq 1). However,
RX + Y- f RY + X-
(1)
X
X
X
we have been able to demonstrate that some of these
reactions can also proceed via competing carbene, car-
banion, and single electron transfer (SET) pathways
depending upon the particular halide leaving group, as
well as the nature of the nucleophile.1
a
b
c
where X = Cl, Br, I, OTs
LDA has been categorized as a hindered, non-nucleo-
philic strong base;5 however, work carried out by this
group has provided evidence for radical involvement in
the reactions of LDA with polynuclear hydrocarbons (eq
3),6 with aromatic ketones,7 and with the cyclizable probe,
6-iodo-5,5-dimethyl-1-hexene (eq 4).1a,g,6
Since 1960, when Arai2 reported cyclization of the
5-hexenyl radical to the methyl cyclopentyl radical (eq
2), cyclizable radical probes have been used for synthetic
or mechanistic purposes, initially by Beckwith,3a Garst,3b
Ingold,3c and more recently in this laboratory.1,3d,4
•–
LDA
(3)
(SET)
Kc = 105/s
(2)
•
LDA
(SET)
(4)
•–
I
I
In order to verify a radical pathway, the reactions of
certain alkyl halide “cyclizable radical probes”, such as
In addition to these reactions, other reactions have also
shown that LDA can function as a one-electron donor
toward heterocyclic compounds,8 R-bromo imines,9 some
conjugated acetylenic compounds,10 and benzophenone.11
X Abstract published in Advance ACS Abstracts, February 1, 1996.
(1) (a) Ashby, E. C.; Park, B.; Patil, G. S.; Gadru, K.; Gurumurthy,
R. G. J . Org. Chem. 1993, 58, 424. (b) Ashby, E. C.; Deshpande, A. K.;
Patil, G. S. J . Org. Chem. 1995, 60, 663. (c) Ashby, E. C.; Deshpande,
A. K. J . Org. Chem. 1994, 59, 7358. (d) Ashby, E. C.; Deshpande, A.
K.; Doctorovich, F. J . Org. Chem. 1994, 59, 6223. (e) Ashby, E. C.;
Deshpande, A. K. J . Org. Chem. 1994, 59, 3978. (f) Ashby, E. C.;
Deshpande, A. K.; Doctorovich, F. J . Org. Chem. 1993, 58, 4205. (g)
Ashby, E. C.; Park, B. Acta. Chem. Scand. 1990, 44, 291. (h) Ashby,
E. C.; Deshpande, A. K. J . Org. Chem. 1995, 60, 4530. (i) Ashby, E.
C.; Deshpande, A. K. J . Org. Chem. 1995, 60, 7117.
(5) (a) House, H. O. Modern Synthetic Reactions, 2nd ed.; Ben-
jamin: Menlo Park, 1972, Chapter 9. (b) Fraser, R. R.; Bresse, M.;
Mansour, T. S. J . Chem. Soc., Chem. Commun. 1993, 620.
(6) Ashby, E. C.; Goel, A. B.; DePriest, R. N. J . Org. Chem. 1981,
46, 2429.
(2) Arai, S.; Sato, S.; Shida, S. J . Chem. Phys. 1960, 33, 1277.
(3) (a) Beckwith, A. L. J .; Phillipou, G. J . Am. Chem. Soc. 1974, 96,
1613. (b) Garst, J . F.; Smith, C. D. J . Am. Chem. Soc. 1976, 98, 1520.
(c) Ingold, K. U.; Griller, D. Acc. Chem. Res. 1980, 13, 317. (d) Ashby,
E. C.; Bowers, J .; DePriest, R. N. Tetrahedron Lett. 1980, 21, 3541.
(4) (a) Ashby, E. C.; DePriest, R. N.; Goel, A. B.; Wenderoth, B.;
Pham, T. N. J . Org. Chem. 1984, 49, 3545. (b) Ashby, E. C.; DePriest,
R. N.; Su, W. Y. Organometallics 1984, 3, 1718. (c) Ashby, E. C.; Su,
W. Y.; Pham, T. N. Organometallics 1985, 4, 1493.
(7) Ashby, E. C.; Goel, A. B.; DePriest, R. N. Tetrahedron Lett. 1981,
22, 4355.
(8) Newkome, G. R. and Hager, D. C. J . Org. Chem. 1982, 47, 599.
(9) Kimpe, N. D.; Yao, Z. P.; Schamp, N. Tetrahedron Lett. 1986,
27, 1707.
(10) Shen, C. and Ainsworth, C. Tetrahedron Lett. 1989, 20, 89.
(11) (a) Kowalski, C.; Creary, X.; Rollins, A. J .; Burke, M. C. J . Org.
Chem. 1978, 43, 3101. (b) Kimpe, N. D.; Palamarera, M.; Schamp, N.
J . Org. Chem.1985, 50, 2993.
0022-3263/96/1961-1322$12.00/0 © 1996 American Chemical Society

Reactions of 1,1-Dihalo-2-methyl-2-phenylpropanes with LDA
J . Org. Chem., Vol. 61, No. 4, 1996 1323
Ta ble 1. Effect of Stoich iom etr y in th e Rea ction s of 1,1-Dich lor o-2-m eth yl-2-p h en ylp r op a n e (1) a n d
1,1-Diiod o-2-m eth yl-2-p h en ylp r op a n e (2) w ith LDA a t 0 °C in THF a
% yield of productsb
CH3
CH3
CH3
Ph
H
CH3
CH3
Ph
Cl
CH3
CH3
Ph
CH3
CH3
Ph
CH2I
CH3
Ph
C
C
C
C
C
C
I
CH3
ratio
dihalide
exp no.
dihalide
dihalide:LDA
recovered
A
B
D
H
E
1c
2
1
1
1
2
2
2
1:5
1:2
1:1
1:5
1:2
1:1
5.5
5.8
69
67
9.6
19
trace
trace
15
6.7
trace
3
4d
5d
6d
12
32
17
7.6
1.3
15
24
39
a
b
Reaction times were 10 min. Trace amount of C was formed in experiments 1-3. c Product G was also formed in 3.7% yield in
d
experiment 1. Dimer F was formed in experiments 4, 5 and 6 in yields of 7.2, 7.8, and 4.5%, respectively.
CH3
CH2Cl Ph
CH3
CH CH
CH3
CH3
Ph
CH3
Ph
Ph
CH3
C
F
G
CH3
CHCl2
CH3
LDA has also been shown to form carbenes on reaction
with benzylic halides.12
CH3
CH3
Ph
H
LDA
+
Ph
C
C
THF, 0 °C
In one of our recent works,1a we observed that alkyl
monochlorides react very slowly with LDA by a carbene
pathway, whereas alkyl monoiodides afford products
derived from radical intermediates. It was, therefore, of
interest to determine if introduction of a second halogen
atom on the carbon bearing the first halogen atom would
have an effect on the mechanistic course of the reaction,
especially for the chloride. We decided to employ for our
studies a system (1 and 2) where cyclization of an
intermediate radical (if formed) would not be very likely
(eq 5). In addition, the phenyl group was expected to
facilitate detection of carbene-derived products consider-
ing the relative ease with which phenyl migration occurs
in carbenes.
1
A
(6)
CH3
CH2Cl
CH3
Ph
Cl
CH3
CH3
+
+ Ph
C
C
Ph
B
C
G
CH3
CHI2
CH3
CH3
CH3
Ph
H
LDA
+
Ph
C
C
C
THF, 0 °C
2
A
CH3
CH2I
CH3
CH3
CH3
CH3
Ph
(7)
Ph
Ph
C
+
+
I
CH3
+
CH3
CH3
LDA
D
E
H
products
CHX2
(5)
CH3
CH3
CH3
CH3
Ph
CH CH
CH3
Ph
1, X = Cl
2, X = I
F
The advantages of this system, as well as the neopentyl
system a , are that the complication of an unwanted
dehydrohalogenation pathway is eliminated and the
possibility of a SN2 or SN1 pathway is lessened because
of the primary nature of the alkyl group and the steric
hindrance of the neopentyl system. As a result, the SET
pathway has a better chance of being observed in
reactions of such probes with nucleophiles.
high molecular weight polymers or other products not
detectable by GC. Thus, during this work, the reactions
extensively studied were those in experiments 3 and 6,
where the material balances were satisfactory enough to
enable one to evaluate the reactions mechanistically. In
all cases, experiments were carried out in duplicate and
repeated if the duplicate results were not within 10%.
Initially, it seems reasonable to categorize the products
of experiments 3 and 6 into two classes: carbene products
(products B and D) and reduction products coming from
either single electron transfer (SET) or polar routes
(products A, C, E, and F ).
Effect of Lea vin g Gr ou p . The results in Tables 1
and 2 (experiments 1-8) indicate a significant effect of
the leaving halide on the product distribution. The only
important product in experiment 3 (product B) appears
to be a carbene product, whereas in experiment 6 both
carbene product (D) and reduction product (E) are
present in substantial amounts. The ratio of the carbene
product D to the SET or polar reduction products A, E,
and F in experiment 6 is approximately 1. Furthermore,
there is only a trace amount of monochloride C formed
in experiment 3, while experiment 6 affords a significant
Resu lts a n d Discu ssion
The products formed in the reactions of LDA with the
dihalides 1,1-dichloro-2-methyl-2-phenylpropane (1) and
1,1-diiodo-2-methyl-2-phenylpropane (2) are given in eqs
6 and 7 and the results of varying the reactant stoichi-
ometries are shown in Table 1. As the data in Table 1
indicate, poor material balances are observed at reactant
ratios of 1:5 and 1:2 for both dihalo substrates. One
might speculate that the reason for such low material
balances is that excess LDA converts the radical probe
dihalides, and/or the products formed initially, to some
(12) (a) Dougherty, C. M.; Olofson, R. A. Org. Synth. 1978, 58, 37.
(b) Creary, X. J . Am. Chem. Soc. 1977, 99, 7632.

1324 J . Org. Chem., Vol. 61, No. 4, 1996
Ashby et al.
Ta ble 2. Rea ction s of 1-Ha lo-2-m eth yl-1-p h en yl-1-
quently undergoes a phenyl migration to form the vinylic
halide B or D.
p r op en es (B a n d D) w ith LDA a t 0 °C in THF a
% yield of products
Carbene rearrangements involving hydrogen, methyl,
or phenyl migration to form olefins have previously been
reported by several groups, and the order of migratory
aptitude is reported to be H > Ph > Me.14 Phillip and
co-workers, for instance, have reported the formation of
A, 2b, and 2c (Scheme 2) in the thermal decomposition
of 2-methyl-2-phenyl-1-diazopropane.15 The two major
Ph
X
CH3
CH3
Ph
CH3
C
C
C
C
Ph
H
CH3
halide
exp no.
recovered
B or D
A
G
7
8
B, X ) Cl
D, X ) I
77
9.3
1.5
87
3.2
trace
a
The ratio of B or D to LDA was 1:2, and the reaction time
Sch em e 2
was 10 min.
CH3
CH3
CH3
CHN2
CH3
Ta ble 3. Rea ction s of 1-Ha lo-2-m eth yl-2-p h en ylp r op a n es
∆
Ph
C
H
Ph
(C a n d E) w ith LDA a t 0 °C in THF a
% yield of products
CH3
CH2X
CH3
CH3
Ph
H
CH3
CH3
Ph
Ph
C
C
H
CH3
CH3
Ph
Ph
CH3
CH3
Ph CH3
halide
recovered
+
C
C
+
C
C
exp no.
A
H
C or E
H3C
CH3
H
9
10
C, X ) Cl
E, X ) I
100
71
0
7.5
0
12
2c (41%)
2b (9%)
A (50%)
a
The ratio of C or E to LDA was 1:2, and the reaction time
products A and 2c are phenyl migration and carbene
insertion products, respectively, while product 2b is the
methyl migration product.
was 10 min.
amount of monoiodide E. These observations are con-
sistent with the order of ease of reduction of halo
compounds, i.e., I > Br > Cl.13 In addition, as the results
in Table 2 (experiments 7 and 8) indicate, when the
vinylic halides B and D were independently allowed to
react with LDA, reduction product A was formed in
greater amount from the vinylic iodide D than from the
corresponding vinylic chloride B (87% vs 1.5%). This
observation is again in agreement with the trend of
reduction potential of the halo compounds. Finally, the
monochloride C is essentially unreactive toward LDA
while its monoiodide counterpart E reacts with LDA to
form tert-butylbenzene (H) (experiments 9 and 10, Table
3).
P r od u cts Su ggested To Be F or m ed fr om Ca r -
ben oid /Ca r ben e In ter m ed ia tes. The major product
formed in the reaction of 1 with LDA is B (experiment
3), and the major product formed in the reaction of 2 with
LDA is D (experiment 6). Both products vary only in the
nature of the halogen and appear to be formed by a
mechanism illustrated in Scheme 1.
On the basis of the results reported by Phillip,15 one
might also expect cyclopropane formation from similar
carbenes such as 1b (Scheme 3). However, no cyclopro-
Sch em e 3
CH3
CH3
Ph
~Ph
C
C
Ph
C
X
X
CH3
CH3
B or D
1b
C–H insertion
Ph CH3
X
pane was formed from 1 or 2 in their reactions with LDA.
This can be explained by arguing that the carbene
precursor to B or D is a halocarbene, whereas in Scheme
2 the carbene is a protiocarbene (we have recently
observed that the halocarbene generated in the reactions
of 6,6-dihalo-5,5-dimethyl-1-hexenes with LDA prefer-
entially adds intramolecularly across the terminal CdC
bond, and there is no product derived from insertion of
the halocarbene into a C-H bond1b). Also, one cannot
be certain as to whether the reactive intermediate here
is the carbenoid 1a or the free carbene 1b (Scheme 1).
Therefore, the ambiguity associated with the carbenoid/
carbene mechanism shown in Scheme 1 makes it prob-
lematical to expect the products that would normally be
formed when a free carbene is present. One might also
argue that even if the free carbene 1b existed as the
reactive intermediate the rate of phenyl migration in this
carbene to form the vinylic halides B or D could be
greater than that of C-H insertion to form the cyclopro-
pane compared to the carbene formed in Scheme 2, and
hence, one would not see any cyclopropane.
Sch em e 1
CH3
CHX2
CH3
X
X
LDA
Ph
Ph
C
Li
CH3
CH3
1 or 2
1a
–LiX
CH3
CH3
CH3
Ph
X
~Ph
Ph
C
X
C
C
CH3
B or D
1b
According to this proposed mechanism, LDA acts as a
strong base and abstracts the R-proton of the dihalo
compound to form the carbenoid intermediate 1a , which
then generates the carbene intermediate 1b by the loss
of lithium halide. The carbene intermediate subse-
In order to trap the carbene intermediate believed to
be involved (1b ), a 10-fold excess of a known carbene
(13) Pine, S. H.; Hendrickson, J . B.; Cram, D. J .; Hammond, G. S.
Organic Chemistry, 4th ed.; McGraw Hill: New York, 1980; p 379.
(14) Sargeant, P. B.; Shechter, H. Tetrahedron Lett. 1964, 3957.
(15) Phillip, B. H.; Keating J . Tetrahedron Lett. 1961, 15, 523.

Reactions of 1,1-Dihalo-2-methyl-2-phenylpropanes with LDA
J . Org. Chem., Vol. 61, No. 4, 1996 1325
Ta ble 4. Effect of Ad d itives in th e Rea ction s of 1,1-Dih a lo-2-m eth yl-2-p h en ylp r op a n es (1 a n d 2) w ith LDA
a t 0 °C in THF , Eth er , a n d TMEDAa
% yield of productsc
CH3
CH2I
CH3
CH CH
CH3
CH3
Ph
CH3
Ph
Cl
CH3
CH3
Ph
CH3
CH3
Ph
Ph
C
C
C
C
I
CH3
dihalide
recovered
exp no.
dihalide
solvent/additiveb
B
D
E
F
11
12
13
6
1
1
1
2
2
THF/DMBd
ether/DMB
TMEDA/DMB
none
19
51
72
39
50
67
31
14
24
24
17
9.6
4.5
1.8
14
THF/DBNOe
a
b
Ratio of the dihalide to LDA was 1:1. DMB ) 2,3-dimethyl-2-butene, DBNO ) di-tert-butyl nitroxide. c Trace amounts of A were
also formed in all these experiments, and a trace amount of C was formed in experiments 11-13. Ratio of 1 to DMB was 1:10. e Ratio
of 2 to DBNO was 1:0.1.
d
CH3
Ph
H
CH3
CH3
Ph
CH2Cl
CH3
C
C
A
C
trapping agent, 2,3-dimethyl-2-butene,16,17 was used in
the reaction of 1 with LDA under normal experimental
conditions (experiment 11, Table 4). However, no trace
of the expected cyclopropane that would have provided
evidence for the existence of carbene 1b was found
(Scheme 4). Given the result of experiment 11, it seems
2 with LDA in THF at 0 °C (experiment 14, Table 4), the
yield of product E dropped 44% and the yield of F
dropped 60%. Furthermore, a new product, which was
determined to be the corresponding hydroxide of the
radical scavenger as judged by its molecular ion and mass
spectral fragmentation pattern, was also formed. As
Scheme 5 below shows, this hydroxide could have been
Sch em e 4
Sch em e 5
CH3
CH3
Ph
CH3
~Ph
(k1)
•
CH3
C
C
O
Ph
C
C
X
•–
CHI2
(a)
X
Ph
CH3
+
N
B or D
CH3
1b
CH3
CH3
5a
(b)
LDA
C
H3C
X
CH3
Ph
CHI2
LDA
(SET)
CH3
CH3
intermolecular
(k2)
+
O–
Ph
2
CH3
CH3
OH
N
(k1 >>> k2)
H+
N
that if indeed such a free carbene exists, the rate of its
intramolecular reaction (k1) to form the olefin must be
much faster than that of intermolecular trapping of the
carbene (k2) due to the steric hindrance involved in
forming the product. Also, the possibility exists that a
carbenoid intermediate (1a ) rather than a free carbene
(1b) is involved as an intermediate, and therefore, the
formation of 4a would not be expected.
The carbene trapping experiment was also attempted
in two additional solvents, namely ether and TMEDA
(experiments 12 and 13, Table 4), in the hope that the
free carbene 1b would be more readily formed from the
carbenoid 1a in these solvents. Once again, no cyclopro-
pane was found and the rate of reaction decreased.
Red u ction P r od u cts. Products A, C, E, and F
appear to be reduction products. Compared to products
A and C in experiments 3 and 6, products E and F are
formed in considerably large amounts in experiment 6.
It is possible for E and F to be formed from either a polar
or a radical intermediate. A preliminary investigation
employing a known radical scavenger16 provided some
evidence for the involvement of a radical precursor in the
formation of E and F . When 10 mol % of di-tert-butyl
nitroxide (DBNO) was added to the reaction mixture of
5c
5b
formed by electron transfer from the radical anion (5a )
to DBNO (pathway a ) to form the anion (5b) or by
electron transfer from LDA to DBNO (pathway b) to form
5b, followed by subsequent abstraction of a proton from
the medium. The first suggestion assumes that the
radical anion (5a ) has sufficient lifetime to transfer its
electron to the radical scavenger (it was not possible to
isolate 5c and further characterize it by NMR spectros-
copy because it was formed in such small quantity). The
conclusion of this trapping experiment (experiment 14,
Table 4) is that products E and F are formed from radical
precursors.
As far as the formation of E is concerned, the mecha-
nism in Scheme 6 is proposed which involves a SET
route. There are two places from which a hydrogen atom
can be abstracted by the radical intermediate 6b, namely
the solvent and the amide. By use of the solvent THF-
d8 and analysis of the deuterium content of the product,
it was determined that there was no deuterium incorpo-
ration in any of the products in the reaction of 2 with
LDA in THF-d8 (experiment 15, Table 5). However, since
C-H bonds in THF are strengthened by replacing the
hydrogen atoms with deuterium in THF-d8, one cannot
rule out, with absolute certainty, the possibility that THF
is the hydrogen atom source. This point will be estab-
lished later.
(16) Goldstein, M. J .; Dolbier, W. R. J . Am. Chem. Soc. 1965, 87,
2293.
(17) Kirmse, W. Carbene Chemistry, 2nd ed.; Academic Press: New
York, 1971; p 294.

1326 J . Org. Chem., Vol. 61, No. 4, 1996
Ashby et al.
Ta ble 5. Effect of Isotop ica lly La beled Solven t a n d Nu cleop h ile in th e Rea ction of
1,1-Diiod o-2-m eth yl-2-p h en ylp r op a n e (2) w ith LDA a t 0 °C
% yield of productsc
(% d1 incorporation)
CH3
CH3
CH3
Ph
CH3
CH3
Ph
H
CH3
CH3
Ph
CH2I
Ph
CH CH
CH3
Ph
CH3
CH3
C
C
CH3
dihalide
recovered
exp no.
solvent
amidea,b
A
H
E
F
6
15
THF
THF-d8
LDA
LDA
39
40
trace
trace
(0)
trace
(0)
trace
(0)
15
17
14
(0)
8.8
(70)
7.3
(81)
12
7.6
4.5
3.6
(0)
2.0
(0)
1.8
(0)
7.2
3.3
16
17
THF
LDA-d2
LDA-d2
45
47
THF-d8
4
18
THF
THF
LDA
LDA-d2
7.6
6.4
(42)
5.9
(43)
18
(39)
16
(42)
19
THF-d8
LDA-d2
8.1
3.6
a
b
The ratio of 2 to LDA was 1:1 in experiments 6, 15, 16, and 17. The ratio of 2 to LDA was 1:5 inexperiments 4, 18, and 19. c Product
D was also formed in experiments 6, 15, 16, and 17 in yields of 24, 28, 28, and 22%, respectively.
Ph
CH3
CH3
C
C
I
D
Sch em e 6
either experiment 16 or 17. A suggested mechanism in
full accord with such high deuterium incorporation in
product E and that is also consistent with the results
obtained from experiments 14 and 15 is shown in Scheme
7.
CH3
CH3
•–
CHI2
LDA
Ph
CHI2
Ph
(SET)
CH3
CH3
5a
2
–I–
Sch em e 7
CH3
CH3
CH3
CH3
•
LDA
Ph
CHI
Ph
CHI2
•
(SET)
Ph
CH2I
Ph
CHI
CH3
CH3
CH3
CH3
6b
2
E
6b
LDA
THF
1
When the H NMR spectrum of the product mixture
in experiment 15 was evaluated, the signals correspond-
ing to isopropylideneisopropylamine were detected. This
imine is a byproduct of either a polar reduction pathway
such as the one reported in the reduction of benzophe-
none by LDA7 or one formed by SET reduction. The
observation in experiment 14 that the formation of
reduction products E and F was inhibited by using a
radical scavenger leads one to believe that a radical
pathway is operative in the formation of E and F . In
such SET reductions, the R-hydrogens of LDA are the
more probable source of hydrogen atoms.
In order to confirm that the R-hydrogens of LDA are
the major source of hydrogen atoms in the reaction of 2
with LDA in THF, LDA-d2 was prepared from diisopro-
pylamine-d2 which had been found to contain 94%
deuterium by 1H NMR analysis. This labeled amide was
then allowed to react with 2 using protio THF (experi-
ment 16, Table 5) and THF-d8 (experiment 17) as the
solvent. The percentages of deuterium incorporated in
product E in experiments 16 and 17 were 70% and 81%,
respectively. The yield of E decreased as did the rate of
reaction in both experiments. These results are not too
surprising since LDA and LDA-d2 should not be expected
to have identical one-electron donor capabilities, and
indeed the rate decrease involving LDA-d2 was small. In
addition, the yield of F decreased by a factor of 2, and
there was no deuterium incorporation in this product in
•–
N
CH3
2
Ph
CH2I
CH3
E
CH3
•–
CHI2
Ph
+
N
CH3
7a
It is proposed that, in the formation of E, the hydrogen
atom comes from both the LDA and the THF. In
experiment 16, the LDA-d2 is preferred over the protio
THF. However, the 30% protium in product E (experi-
ment 16) can originate from THF or the 6% hydrogen
contained in the LDA-d2. On labeling both the solvent
and the LDA with deuterium (experiment 17), there was
an 11% increase in the deuterium content of E compared
to when only the LDA was labeled, thus showing that
hydrogen can also be abstracted from THF to form E.
The 19% protium content of E (experiment 17) can again
be attributed to the 6% protium in the LDA-d2. Thus, it
appears that even though both sources of hydrogen atom
are labeled here, the radical intermediate 6b preferen-
tially abstracts a hydrogen atom from LDA.
The following mechanism (Scheme 8) is proposed for
the formation of F on the basis of the results of experi-
ments 14-17 (Tables 4 and 5), whereby F seems to be

Reactions of 1,1-Dihalo-2-methyl-2-phenylpropanes with LDA
J . Org. Chem., Vol. 61, No. 4, 1996 1327
Sch em e 8
Sch em e 9
•–
I
CH3
Ph
CH3
CH3
CH3
CH3
CH3
Ph
Ph
H3C
Ph
H3C
LDA
•
C
C
C
C
Ph
CHI
(SET)
I
I
CH3
CH3
I
D
9a
6b
8a
–I–
LDA
(Li–I exchange)
CH3
CH3
Ph
Ph
SH
Li CH3
Ph
C
C
C
•
C
H3C
Ph
H3C
–LiI
H
CH3
CH3
CH3
A
9b
I
The low deuterium content of product A in experiments
18 and 19 (Table 5) can partly be explained by the same
reasons previously discussed for the formation of product
E, namely that the LDA-d2 is only 94% d2, and since the
LDA-d2 is employed in 5:1 ratio with respect to 2, there
is sufficient protium present to give a high protium
content because of the primary deuterium kinetic isotope
effect.
Combining the results of experiments 10, 18, and 19
(Tables 3 and 5) suggests that product H is a product of
further reduction of the monoiodide E with LDA (Scheme
10) in which the R-hydrogens of LDA are almost the
CH3
CH CH
CH3
CH3
CH3
Ph
Ph
F
formed by a radical mechanism that does not involve
hydrogen atom abstraction. In this proposed mechanism,
the radical intermediate 6b dimerizes to form the vicinal
diiodide (8a ) which undergoes iodine-lithium exchange
followed by loss of lithium iodide to form F .
It was not possible to isolate the vicinal diiodide 8a .
However, evidence supporting the last step in this
mechanism was provided by the reaction of the simplest
vicinal diiodide, 1,2-diiodoethane, with LDA. When 1
equiv of LDA was treated with 1,2-diiodoethane at 0 °C
in THF, a gas was generated which was shown to be
ethylene. The reaction was instantaneous and the yield
quantitative.
Sch em e 10
CH3
CH3
•–
CH2I
LDA
Ph
CH2I
Ph
(SET)
CH3
CH3
10a
E
P r op osed Mech a n ism s for th e F or m a tion of Mi-
n or P r od u cts A a n d H. The results of experiments 7
and 8 (Table 2), 10 (Table 3), and 18-20 (Tables 5 and
6) provide evidence that supports reasonable mechanisms
for the formation of A and H. The observation that the
reduction of the vinylic iodide D with LDA in experiment
8 (Table 2) gives an almost quantitative yield of product
A indicates that D is the precursor to A in the reactions
of 2 with LDA. This conclusion is further supported by
the fact that D, which is a major product in exp 6 (Table
1), is not observed when 2 is treated with excess LDA
(experiment 4, Table 1) and that there was 15% of A
formed in exp 4. Even though the vinylic chloride B is
not quite as reactive with LDA as the corresponding
iodide, it seems reasonable to assume that it is reactive
enough to yield small quantities of A in experiments 1-3
(Table 1).
The observation that no deuterium was incorporated
in product A in experiments 15-17 (Table 5) can be
explained by considering that, in these experiments,
product A is formed only in trace quantities so that it
becomes very difficult to obtain an accurate analysis for
the deuterium incorporation in this product. On the
other hand, when 5 equiv of LDA-d2 (prepared from
diisopropylamine-d2, 94% deuterium content) was al-
lowed to react with 2 in protio THF (experiment 18, Table
5) and THF-d8 (experiment 19, Table 5), 39% and 42%
deuterium, respectively, was incorporated in product A.
Consequently, on the basis of these results, it is reason-
able to assume that the R-hydrogen atoms of LDA are
the only source of hydrogen. However, whether the
reduction of D to A occurs via a radical pathway (Scheme
9) or a polar one cannot be verified on the basis of the
results of experiments 18 and 19 since both radical and
polar pathways would utilize the same hydrogen atom
in LDA.
–I–
CH3
CH3
•
SH
Ph
CH3
Ph
CH2
CH3
CH3
H
10b
exclusive source of hydrogen. Furthermore, the low
deuterium content of H (42% in experiment 18 and 43%
in experiment 19, Table 5) can be partly accounted for
by the same arguments as those presented for the
formation of A.
The observation that the deuterium incorporated in
products A and H is almost the same in experiments 18
and 19 (Table 5) speaks against a radical intermediate
abstracting a hydrogen atom from the solvent. In order
to provide evidence for the involvement of radical inter-
mediates in the conversion of D to A (Scheme 9) or E to
H (Scheme 10), a known radical scavenger was employed.
In the reaction of 2 with LDA in a 1:1 ratio, in the
presence of 10 mol % of DBNO, the formation of products
E and F was greatly reduced (Table 4, experiment 14);
however, when 10 mol % of DBNO was added to the
reaction of 2 with LDA in a 1:5 ratio in THF at 0 °C
(experiment 20, Table 6), there was no significant change
in the yield of any of the products except that of the
dimeric radical product F which dropped 60% as it had
in experiment 14 (Table 4). The reason why the yield of
products A, E, and H did not change in experiment 20
(Table 6) could be that LDA is present in a much greater
amount than the radical trap. This causes the radical
intermediates 6b (Scheme 6), 9b (Scheme 9), and 10b
(Scheme 10) to be preferentially trapped by LDA than
by the radical scavenger. Also, it is possible that the
excess LDA reacts with the radical scavenger, DBNO.

1328 J . Org. Chem., Vol. 61, No. 4, 1996
Ashby et al.
Ta ble 6. Effect of Ra d ica l Tr a p DBNO in th e Rea ction of 1,1-Diiod o-2-m eth yl-2-p h en ylp r op a n e (2)
w ith LDA in a 1:5 Ra tio a t 0 °C
% yield of products
CH3
CH CH
CH3
CH3
CH2I
CH3
Ph
CH3
CH3
CH3
Ph
H
CH3
CH3
Ph
Ph
Ph
C
C
CH3
CH3
exp no.
additive
none
DBNO (10 mol%)
A
H
E
F
4
20
15
14
12
14
7.2
4.4
7.6
8.3
Ta ble 7. Rea ction of 1,1-Diiod o-2-m eth yl-2-p h en ylp r op a n e (2) w ith Lith iu m Tetr a m eth ylp ip er id id e in a 1:1 Ra tio a t 0 °C
% yield of products
(% deuterium incorporation)
CH3
CH2I
CH3
CH CH
CH3
CH3
Ph
CH3
Ph
H
CH3
CH3
Ph
CH3
CH3
Ph
Ph
C
C
C
C
I
CH3
exp no.
condns
A
D
E
F
21a
LTMP/THF
LTMP/THF-d8
2.3
2.1
(20)
9.5
22
(0)
6.5
1.4
(14)
21
10
(0)
22b
a
b
In experiment 21, 47% of 2 was recovered after 10 min. In experiment 22, 55% of 2 was recovered after 10 min.
By using a lithium amide with no R-hydrogen atoms
such as lithium tetramethylpiperidide (LTMP), the radi-
cal intermediates 6b, 9b, and 10b were expected to
abstract hydrogen atoms from the solvent. When 2 was
allowed to react with 1 equiv of LTMP in THF-d8 at 0 °C
(experiment 22, Table 7), the deuterium incorporation in
products A and E was 20% and 14%, respectively. This
result provides strong evidence for the involvement of
radical intermediates in the formation of A and E.
However, since H was not formed as a product of the
reaction of 2 with LTMP, data are not available to
support a radical mechanism for the formation of H.
P r op osed Mech a n ism s for th e F or m a tion of Mi-
n or P r od u cts C a n d G. Product C can arise via SET
or polar reduction of 1 by LDA. The available results do
not allow one to differentiate between these two path-
ways. Product G is only formed when an excess of LDA
is used (experiment 1, Table 1) or when products B or D
are allowed to react with LDA (experiments 7 and 8,
Table 2). G is thought to be formed from B or D via a
mechanism (Scheme 11) in which LDA abstracts the
allylic hydrogen of the vinylic halide to form the carban-
ion 11a .
SET process. G could not have originated from C or E
since reaction of C or E with LDA did not produce product
G (Table 3).18
Effect of Solven t. When 2 was allowed to react with
LDA at 0 °C in diethyl ether, there was no appreciable
change in the product distribution or the reaction rate
compared to the same reaction in THF (experiments 6
and 23, Table 8). However, when 1 was allowed to react
with LDA in diethyl ether at 0 °C, the rate of the reaction
decreased substantially compared to the same reaction
in THF (experiments 3 and 24, Table 8). The lower rate
of the reaction of 1 with LDA in diethyl ether compared
to THF can be explained by the lower solvating power of
diethyl ether compared to THF. As a result of this, the
lithium in LDA is not as strongly coordinated to diethyl
ether as it is to THF; therefore, LDA is expected to be
less ionic and more sterically hindered in diethyl ether
than in THF. This result should cause LDA to be a
weaker base in removing the R-hydrogen atoms of the
probe compounds. The same solvent effect was studied
concerning the reactions of 2 with LDA; however, in this
case, the reduction potential of 2 is so much more
favorable to electron transfer from LDA than 1 that SET
is the major reaction pathway that is not very sensitive
to steric effects. On the other hand, the reaction of 1 with
LDA proceeds predominantly through a carbene inter-
mediate that proceeds initially by a proton abstraction
reaction which is sensitive to steric effects.
Sch em e 11
–
CH3
CH2
Ph
X
Ph
X
LDA
C
C
C
C
CH3
CH3
B or D
11a
Exp er im en ta l Section
Ma ter ia ls. Reagent grade diethyl ether, dimethoxyethane
(DME), tetrahydrofuran (THF), and p-xylene were purchased
intramolecular SN2
displacement
Ph
(18) A reviewer has suggested a reasonable alternate pathway for
the formation of G from anion 11a in Sch em e 11, as follows:
11b
–
Ph
CH2
CH2
CH3
Ph
X
Ph
– C
X
C
C
C
G
CH3
This carbanion then cyclizes by an intramolecular SN2
process to form the strained three-membered ring 11b
which subsequently isomerizes rapidly to form G. Since
G was not formed from D (Table 2) and since D is more
likely to undergo a SET process than B, it is unlikely
that G was formed by the reaction of LDA with B by a
11a
–X–
CH2
Ph
Ph
C
C
CH3
11b

Reactions of 1,1-Dihalo-2-methyl-2-phenylpropanes with LDA
J . Org. Chem., Vol. 61, No. 4, 1996 1329
Ta ble 8. Role of Eth er a n d THF a s Solven ts in th e Rea ction s of 1,1-Dih a lo-2-m eth yl-2-p h en ylp r op a n es (1 a n d 2)
w ith LDA a t 0 °C in THF a
% yield of productsb,c
CH3
CH2I
CH3
CH CH
CH3
CH3
Ph
CH3
Ph
Cl
CH3
CH3
Ph
CH3
CH3
Ph
Ph
C
C
C
C
I
CH3
dihalide
exp no.
dihalide
solvent
recovered
B
D
E
F
6
23
3
2
2
1
1
THF
ether
THF
ether
39
43
19
51
24
22
17
14
4.5
7.3
67
31
24
a
b
The ratio of the dihalide to LDA was 1:1. A trace amount of A was also formed in all experiments. c A trace amount of C was also
formed in experiments 3 and 24.
CH3
Ph
H
CH3
CH3
Ph
CH2Cl
CH3
C
C
A
C
from Fisher and distilled from a deep blue solution of sodium
benzophenone ketyl under N2 prior to use. Chloroform,
methanol, and absolute ethanol were purchased from Fisher
and used as received. Diisopropylamine, cyclohexylamine,
triethylamine, and 2,3-dimethyl-2-butene were purchased from
Aldrich and distilled from CaH2. Cumene, dl-2-phenylpropa-
nal, N,N,N′,N′-tetramethylethylenediamine (TMEDA), and
2-butanol were purchased from Aldrich and distilled from
CaH2 at reduced pressure. Iodomethane was purchased from
Aldrich and distilled from CaCl2. Methyllithium was pur-
chased as a 1.30 M solution in ether from Aldrich. 2-Methyl-
1-phenyl-1-propene (A), tert-butylbenzene (H), 1-chloro-2-
methyl-2-phenylpropane (C), triethylbenzylammonium chloride,
hydrazine hydrate, iodine, sodium cyanoborodeuteride, 1,2-
diiodoethane, ammonium acetate, CH3OD, di-tert-butyl ni-
troxide, and acetone were also purchased from Aldrich and
used without further purification. The integrity of all starting
materials was checked by GLC prior to use.
under General Procedures, and its mass spectral data were
matched with literature values.20
1-Iod o-2-m eth yl-2-p h en ylp r op a n e (E). This compound
was synthesized and purified according to a literature method21
in 24% yield, and its mass spectral data were matched with
those reported.
2,5-Dim eth yl-2,5-d ip h en yl-3-h exen e (F ). This compound
was isolated from the product mixture of the reaction of LDA
with 2 by flash column chromatography, and its mass spectral
data were matched with those reported.22
2-P h en ylm eth ylen ecyclop r op a n e (G). This compound
was isolated from the product mixture of the reaction of LDA
with 1 by preparative GLC (tR ) 28 min), and its mass spectral
data were matched with those reported.23
Eth ylen e. This compound was generated quantitatively in
the reaction of 1,2-diiodoethane with LDA under the usual
experimental conditions and its spectral data were matched
with those reported.24 IR: 946, 1418, 1461, 1868, 1888, and
3084 cm-1. High resolution mass spectrum: obsd 28.031 479,
calcd 28.031 300 2.
P r epar ation s. 1,1-Dich lor o-2-m eth yl-2-ph en ylpr opan e
(1). This compound was prepared and purified according to a
literature method25 in 25% yield, and its spectral data were
matched with those reported. 1H NMR (CDCl3): δ 7.4 (5 H,
m), 5.95 (1 H, s), and 1.56 (6 H, s). Mass spectrum: (relative
intensity) 202 (M+, 3), 167 (7), 119 (100), and 91 (30).
2-Meth yl-2-p h en ylp r op a n a l. This compound was pre-
pared by a literature method reported for a similar com-
pound.26 To a three-necked round-bottom flask equipped with
a reflux condenser and dropping funnel was added cyclohexy-
lamine (44.6 g, 0.5 mol). Next, dl-2-phenylpropanal (67.1 g,
0.5 mol) was added dropwise, with stirring, over a period of 1
h. The temperature rose during the addition, and a cloudy
mixture formed. The organic layer was separated from the
aqueous layer and treated with 7.5 g of anhydrous K2CO3,
stirred at room temperature for 17 h, and decanted into 6.0 g
of BaO. The mixture was stirred for 10 h, and the organic
layer was filtered to give 96.1 g (89% yield) of the crude imine
which was used in the subsequent step without purification.
Mass spectrum of the imine: m/e (relative intensity) 215 (M+,
9), 133 (13), 91 (20), 83 (40), and 40 (100).
Gen er a l P r oced u r es. All glassware was dried in an oven
at 150 °C and cooled under N2 before use. Glassware in which
the reactions were carried out were flame dried under vacuum
and cooled under N2 three times prior to use.
1H and 13C NMR spectra were recorded on a Varian Gemini
300 MHz instrument using tetramethylsilane (TMS) as refer-
ence (0.0 δ). Mass spectral analyses were performed using a
VG 70-SE mass spectrometer. Gas liquid chromatographic
(GLC) analyses were performed using a Varian Model 3700
instrument equipped with a flame ionization detector (FID)
and a fused silica DB5 (30 m) column. The conditions for
analyses were as follows: 70 °C for 3 minutes, followed by 10
°C per minute to 250 °C for 15 min. Preparative GLC
separations were performed using an F & M Model 720
instrument equipped with a thermal conductivity detector
(TCD) and a Carbowax 20 M (6 ft) column under the following
conditions: column temperature, isothermal at 100 °C; He flow
rate of 50 mL/min. Separations by flash column chromatog-
raphy employed silica gel 200-400 mesh, 60 Å (Aldrich), as
the stationary phase and hexane (Fisher) as the mobile phase.
Ch a r a cter iza tion of P r od u cts. Products 2-methyl-1-
phenyl-1-propene (A), 1-chloro-2-methyl-2-phenylpropane (C),
and tert-butylbenzene (H) were identified by matching their
mass spectra with those of commercially available authentic
samples.
A portion of the above imine (10.75 g, 50 mmol) was added
to a solution of 50 mmol of LDA in 35 mL of THF in a three-
1-Ch lor o-2-m eth yl-1-p h en yl-1-p r op en e (B). This com-
pound was isolated from the product mixture of the reaction
of LDA with 1,1-dichloro-2-methyl-2-phenylpropane (1) by
preparative GLC according to the conditions described under
General Procedures (tR ) 56 min), and its mass spectral data
were matched with literature values.19
1-Iod o-2-m et h yl-1-p h en yl-1-p r op en e (D). This com-
pound was isolated from the product mixture of the reaction
of LDA with 1,1-diiodo-2-methyl-2-phenylpropane (2) by flash
column chromatography according to the conditions described
(20) Campbell, J . R.; Pross, A.; Sternhell, S. Aust. J . Chem. 1971,
24, 24.
(21) Collman, J . P; Brauman, J . I.; Madonik, A. M. Organometallics
1986, 2, 310.
(22) Akermanc, B.; Ljungquist, A. J . Organomet. Chem.1978, 149,
97.
(23) Hsiao, C. N.; Shechter, H. J . Org. Chem. 1988, 53, 2688.
(24) Grasseli, J . G. Atlas of Spectral Data and Physical Constants
for Organic Compounds; CRC Press: Boca Raton, 1973; p 518.
(25) Dehmlow, E. V. Tetrahedron 1971, 27, 4071.
(26) House, H. O.; Lang, W. C.; Weeks, P. D. J . Org. Chem. 1974,
39, 21.
(19) Siehl, J .; Hanack, M. J . Am Chem. Soc. 1980, 102, 2686.

1330 J . Org. Chem., Vol. 61, No. 4, 1996
Ashby et al.
necked round-bottom flask under a stream of N2 at -78 °C.
The resulting solution was warmed to 20 °C over a period of
1.5 h and then treated dropwise with 7.8 mL (0.125 mol) of
CH3I with stirring for 20 min while the temperature of the
reaction mixture was kept in the range 20-40 °C by external
cooling. At the end of this addition, the reaction mixture,
which was a slurry containing solid LiBr, was stirred at room
temperature for 3.5 h. Ice-water was then added to the
reaction mixture, and the organic layer was extracted with
ether. The ether layer was washed with saturated NaCl
solution and distilled water and then dried over anhydrous
MgSO4. The ether was removed using a rotary evaporator to
give 11.0 g (96% yield) of the dimethylimine which was used
in the subsequent step without purification. Mass spectrum
of the dimethylimine: m/e (relative intensity) 229 (M+, 16),
119 (80), 110 (60), 91 (19), and 83 (100).
Hydrolysis of the dimethylimine was then carried out by
stirring 9.16 g (40 mmol) in 50.0 mL of hexane at room
temperature with 120 mL of 8% aqueous acetic acid under a
stream of N2 for 2 h. The aqueous layer was then separated
from the hexane layer and extracted with ether several times.
The combined hexane and ether extracts were then washed
with saturated NaCl solution and distilled water and then
dried over anhydrous MgSO4. Solvents were removed using
a rotary evaporator and the product distilled at reduced
pressure to give 5.30 g (89% yield) of 2-methyl-2-phenylpro-
panal as a colorless liquid (boiling point: 80 °C at 10 mmHg).
The spectral data were matched with those reported in the
literature.27 1H NMR (CDCl3): δ 9.5 (1 H, s), 7.3 (5 H, m),
and 1.45 (6 H, s). Mass spectrum of 2-methyl-2-phenylpro-
panal: m/e (relative intensity) 148 (M+, 7), 119 (100) and 91
(58).
1,1-Diiod o-2-m eth yl-2-p h en ylp r op a n e (2). This com-
pound was prepared according to a literature method reported
for the synthesis of geminal diiodides.28 To 1.42 g (100 mmol)
of 2-methyl-2-phenylpropanal, 12 mL of ethanol, and 6.0 mL
of triethylamine was added 6.0 mL of hydrazine hydrate, and
the reaction mixture was allowed to reflux for 1 hour. The
cooled solution was extracted with CHCl3, and the organic
layer was washed with distilled water and finally dried over
anhydrous MgSO4. The hydrazone was obtained as an oil by
evaporating the solvent using a rotary evaporator, and it was
used in the subsequent step without purification. 1H NMR
(CDCl3): δ 7.2 (5 H, m), 7.05 (1 H, s), 5.1 (2 H, broad s), and
1.4 (6 H, s).
The hydrazone was dissolved in 12 mL of ether and 4.0 mL
of triethylamine and treated with a concentrated solution of
excess iodine in ether until the evolution of N2 gas had ceased
and the iodine color persisted. The ether layer was then
washed consecutively with 5% Na2S2O3, 3 N HCl, 5% Na2CO3,
and distilled water and then dried over anhydrous MgSO4. The
ether was removed using a rotary evaporator, and the crude
material was purified by liquid-solid chromatography using
silica gel 200-400 mesh, 60 Å to obtain 0.77 g (20% yield) of
2 as a yellow solid. 1H NMR (CDCl3): δ 7.32 (5 H, m), 5.7 (1
H, s), and 1.4 (6 H, s). Mass spectrum: (relative intensity)
386 (M+, 8), 259 (27), 132 (100), 117 (49), and 91 (24).
Lith iu m Diisop r op yla m id e (LDA). At -78 °C, 1.0 mL
of a 1.30 M solution of MeLi (1.3 mmol) in ether was added to
0.24 mL (1.7 mmol) of diisopropylamine in 1.0 mL of dry THF
in a round-bottom flask under a stream of N2. The resulting
mixture was stirred at this temperature for 30 min. At the
end of this period, the mixture was warmed to 0 °C and the
solvents and excess diisopropylamine were removed under
reduced pressure. LDA was then obtained as a white powder
which was dissolved in dry THF. The LDA solution was used
immediately after standardization. The standardization was
carried out by adding to the organolithium compound a
standard 0.5 M solution of 2-butanol in p-xylene dropwise
under nitrogen using 2,2′-bipyridyl as an indicator. The purple
color of LDA in 2,2′-bipyridyl changed to pale green at the end
point.29 The yield of LDA was usually 70-90% based on MeLi.
Lith iu m Diisop r op yla m in d e-d 2 (LDA-d 2). This com-
pound was prepared and standardized using the same method
as that for the protio compound employing diisopropylamine-
d2 as the starting amine.
N,N′-Bis(1-d eu ter io-1-m eth yleth yl)a m in e (Diisop r op y-
la m in e-d 2). This compound was prepared according to a
reported method30 in 20% yield based on acetone. The
deuterium content of this compound was found to be about
94% by means of 1H NMR. The 1H NMR of this compound
matched with that reported. 1H NMR: δ 0.967 (12 H, s).
Gen er a l P r oced u r e for Rea ction s of 1,1-Dih a lo-2-
m eth yl-2-p h en ylp r op a n es a n d Oth er Alk yl Ha lid es w ith
LDA. To 1.5 mmol of the alkyl halide in a dry 5 mL round-
bottom flask was added enough solvent to make a solution of
the alkyl halide and then the flask placed in a bath at 0 °C.
Unless otherwise specified, 1 equiv of cold standard LDA was
usually added in a dropwise manner. All the reactions carried
out were complete within 1 min, after which time there was
no change in the product distribution. However, the stirring
was continued for an additional 10 min, after which time the
reaction mixture was quenched and the products were ana-
lyzed by GLC using n-dodecane as an internal standard. The
quenching was usually carried out by adding an aliquot from
the reaction pot to several drops of distilled water at 0 °C under
N2 and extracting the products in ether. In all the reactions
carried out, the concentrations of the halide and LDA were
about 0.5 M. Whenever an additive such as a radical or a
carbene trap was required, the solution of the alkyl halide was
prepared first and the trap was added to it before the addition
of LDA.
Ack n ow led gm en t. We gratefully acknowledge sup-
port of this work by the National Science Foundation,
Grant No. CHE 8914309.
J O951299K
(29) Gall, M.; House, H. O. Organic Synthesis; Wiley: New York,
1983, Collect. Vol. VI, p 125.
(30) Newcomb, M.; Varick, T. R.; Goh, S. H. J . Am. Chem. Soc. 1990,
112, 5186.
(27) Kuntzel, H.; Wolf, H.; Schaffner, K. Helv. Chim. Acta 1971, 54,
882.
(28) Pross, A.; Sternhell, S. Aust. J . Chem. 1970, 23, 984.