Synthesis of chiral C2-symmetric 1,2-diamines by the addition of organolithium reagents to N,N'-bis[(S)-1-phenylethyl]ethanediimine

Article abstract of DOI:10.1016/S0040-4020(00)00782-1

The additions of alkyl-, phenyl- and vinyllithium reagents to N,N'-bis[(S)-1-phenylethyl]ethanediimine in THF at -78°C and in DME at -60°C gave high yields of 1,2-diamines with low stereocontrol. Care was taken to quench the reaction mixtures with de-aera

Full text of DOI:10.1016/S0040-4020(00)00782-1

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 8367±8374
Synthesis of Chiral C2-Symmetric 1,2-Diamines by the
Addition of Organolithium Reagents to
N,N0-Bis[(S)-1-phenylethyl]ethanediimine
*
Gianluca Martelli, Stefano Morri and Diego Savoia
Á
Dipartimento di Chimica `G. Ciamician', Universita di Bologna, via Selmi 2, 40126 Bologna, Italy
Received 27 June 2000; revised 11 August 2000; accepted 23 August 2000
AbstractÐThe additions of alkyl-, phenyl- and vinyllithium reagents to N,N0-bis[(S)-1-phenylethyl]ethanediimine in THF at 2788C and in
DME at 2608C gave high yields of 1,2-diamines with low stereocontrol. Care was taken to quench the reaction mixtures with de-aerated H2O
to avoid formation of N-alkylidene-1-phenylethylamines which were formed through homolysis of the dilithium 1,2-diamides to give a-
amido radicals which reacted with O2/H2O in the quenching step. Pentadienyl-, cinnamyl- and 1-trimethylsilylallyllithium reagents gave only
the linear 1,2-diamines with high yields and moderate to good diastereoselectivities. q 2000 Published by Elsevier Science Ltd.
²
Introduction
auxiliary used, see the 1,2-bis-imine 1 in Scheme 1,
because it provides stereocontrol at a high degree, and is
inexpensive, available in both enantiomeric forms and
easily removed by hydrogenolysis. However, a limited
number of 1,2-diamines 2 have been prepared by this
route applying Grignard reagents: the reported examples
include the additions of allylmagnesium chloride,8 PhMgCl
and MeMgBr in diethyl ether,9 and t-BuMgCl in n-hexane at
508C.10 In our previous works in this area,11,12 we described
that the additions of allylic zinc reagents in tetrahydrofuran
(THF), occurring with allylic inversion, were very ef®cient
and diastereoselective, and determined by X-ray structure
analysis the con®guration of the newly formed stereocentres,
Enantiomerically pure 1,2-diamines, especially those
having C2-symmetry, have gained considerable importance
in organic synthesis since they are used as reagents, bases or
ligands in a large number of asymmetric reactions, most
importantly transition metal-catalysed ones,1,2 and as
medicinal agents.3±6 Among the several methods allowing
the asymmetric synthesis of these compounds, the double
addition of organometallic reagents to the 1,2-bis-imine 1
derived from glyoxal and 2 equiv. of an optically pure
amine (chiral auxiliary) is particularly appealing. Up to
now, 1-phenylethylamine has been almost the only chiral
Scheme 1.
²
Keywords: amines; diastereoselection; imines; radicals.
* Corresponding author. Tel.: 139-51-2099571; fax: 139-51-2099456;
e-mail: savoia@ciam.unibo.it
The diastereoselective addition of MeLi±CeCl3 (8 equiv.) to the bis-
SAMP-hydrazone of glyoxal, followed by the addition of propionyl chlo-
ride (24 equiv.) afforded the protected 1,2-diamine (yield 50%, d.e. 98%).7
0040±4020/00/$ - see front matter q 2000 Published by Elsevier Science Ltd.
PII: S0040-4020(00)00782-1

8368
G. Martelli et al. / Tetrahedron 56 (2000) 8367±8374
so correcting the previous assignment8 which was based on
a stereochemical model proposed for the allylboration of an
imine.
involved cleavage and oxidation (dehydrogenation) steps.
With this regard, it has been reported that meso-1,2-
diamines were converted to the more stable d,l-isomers
through the following steps: metallation, cleavage of the
C1±C2 bond of the dilithium diamides to give a-amido
radicals, and recombination of these radicals in an equili-
brium process.14,15 Similarly, in our hands, the pure diamine
(R,R)-2a underwent partial loss of diastereomeric purity
(6%) after formation of the 1,2-diamide 4a by treatment
with 2 equiv. n-BuLi at 08C and allowing the temperature
to rise to 208C during 1 h (Scheme 2). The formation of the
diastereomeric mixture of 1,2-diamines 2a probably
occurred via homolytic cleavage of the C1±C2 bond of
(R,R)-4a, followed by recombination of the a-amido radi-
cals 5a to give an equilibrium mixture of the three diaster-
eomers of 4a. Moreover, we supposed that imine 3a, which
was detected by GC±MS analysis in the same experiment,
was formed from the intermediate radical 5a through an
oxidation step.
Moreover, it was stated that the reactions of Grignard
reagents and lithium alkyls in THF gave complex mixtures
of diastereomeric products.9 Despite the fact that these early
results did not look promising, we decided to carefully
investigate the addition of organolithiums to the 1,2-diimine
1, because the success of these reactions would enormously
widen the scope of this route to enantiopure 1,2-diamines 2
(Scheme 1), considering that diverse, highly reactive orga-
nolithium reagents can be prepared by a wider range of
methods, as compared to Grignard reagents.
Results and Discussion
The additions of PhLi and n-BuLi were preliminarily carried
out in THF at 2788C and in all cases GC analyses of the
reaction mixtures revealed the formation of the 1,2-
In order to shed light on this point, we repeated the prepara-
tion of 2a (1-PhLi-THF) taking more care to completely
avoid air contamination in the reaction apparatus and
using thoroughly de-aerated THF as the solvent, but the
outcome of the reaction was the same. Finally, we observed
that quenching the reaction mixture at 2788C by the addi-
tion of de-aerated H2O in N2 atmosphere the imine 3a disap-
peared (Table 1, entry 3).§ This result led us to repeat the
additions of t-BuLi and n-BuLi to the 1,2-bis-imine 1 taking
care to quench the reaction mixtures with de-aerated H2O.
Using this approach the yields of 1,2-diamines 2b,c were
slightly improved at the expense of the imines 3b,c (entries
6 and 11). It should be observed that the diastereomeric
ratios (d.r.) of the products 2a±c were not affected by the
quenching procedure.
³
diamines 2a and 2b, respectively, with good yields,
together with relevant amounts of the imines 3a and 3b
(Table 1, entries 1 and 4) which were identi®ed by compar-
ison of their GC retention times and mass spectral fragmen-
tations with those of authentic compounds. Column
chromatography (SiO2) of the products allowed the isolation
of the main C2-symmetric diastereomers (R,R)- and (S,S)-2a
and -2b with low yields. It was then observed that by carry-
ing out the reactions in 1,2-dimethoxyethane (DME) at
2608C was slightly more ef®cient, because the yields of
the 1,2-diamines 2a,b increased at the expense of the bypro-
ducts 3a,b, but the diastereomeric ratios (d.r.) were lower
than in THF (entries 2 and 5).
Successively, we examined the reactions of t-BuLi in the
same solvents as well in diethyl ether and n-hexane (entries
7±11). In the less polar solvents, Et2O and especially n-
hexane (entries 7 and 8), the temperature had to be increased
to achieve complete conversion of the starting 1,2-bis-imine
1, and the amounts of the imine by-product 3c reached
maximum levels. It is also noteworthy that the d.r. of 2c
prepared in n-hexane changed slightly when the reaction
mixture was not quenched at the right time: in fact, by
stirring the mixture one further hour at 208C the second-
eluted diastereomer (R,S)-2c almost disappeared and the
third-eluted one (S,S)-2c became prevalent in the mixture.
This observation indicated that an equilibration between the
three diastereomeric 1,2-bis-adducts had occurred. On the
other hand, the yield and diastereoselectivity of the 1,2-
diamine 2c were higher in DME (entry10), although very
similar results were obtained in THF (entry 9); moreover,
the d.r. of 2c did not change signi®cantly with increasing
time in DME at 2608C (entry 10).
We envisaged a mechanism, described in Scheme 3, which
allows a reasonable explanation of the results obtained. The
a-amido radicals 5 formed from the dilithium diamides 4
react with H2O in the quenching step to give a-amino radi-
cals 7, whose dimerization can afford the 1,2-diamines 2. In
alternative pathways, both the a-amido and a-amino radi-
cals 5 and 7 react with the oxygen dissolved in water to give
the a-amido- and a-aminoalkyl peroxy radicals 6 and 8,
respectively, which are the precursors of the imine 3 through
elimination or hydrolysis processes. It should be observed
that the possibility for the imine 3 to be formed by dispro-
portionation of the a-amino radical 7 can be rejected, since
we never detected any of the saturated amines which would
have been concomitantly formed. It should also be observed
that the relative amounts of imines 3 and 1,2-diamines 2, as
determined by GC±MS analysis of the quenched samples,
roughly re¯ect the corresponding amounts of the radical-
anions 5 and 1,2-diamides 4 which are present at the equili-
brium in solution; however, the precise determination of the
equilibrium constants is presumably hampered by the possi-
bility for the imines 3 to undergo partial hydrolysis during
the quenching/extraction procedure.
The imines 3a±c were unexpected by-products, since they
were never detected in previous reports dealing with the
organometallic additions to the 1,2-diimine 1.8±12 We
considered that they were produced by a process which
We also carried out the addition of vinyllithium, prepared
³
It is noteworthy that more polar solvents, e.g. THF, favoured the N-
alkylation in the additions of EtLi and n-BuLi to N,N0-di-tert-butyl-bis-
§
Similarly, the formation of oxidized compounds in the reactions of aryl-
lithiums with CO was suppressed by quenching with de-aerated H2O.16,17
1,2-ethanediimine.13

G. Martelli et al. / Tetrahedron 56 (2000) 8367±8374
8369

8370
G. Martelli et al. / Tetrahedron 56 (2000) 8367±8374
Scheme 2.
Scheme 3.
from tetravinyltin and MeLi in THF,18 and obtained, after
quenching with de-aerated H2O, the expected 1,2-diamine
2d with very high yield but low diastereoselectivity (Table
1, entry 12); furthermore, the diastereomers could not be
separated by column chromatography (SiO2). It should be
noteworthy that the reactions were not plagued by the
formation of imines analogous to 3a±c, even if care
was not taken to use de-aerated H2O in the quenching
step. We cannot give a rationale for this different beha-
viour. The pure (R,R)-diastereomers were isolated in
pure state by column chromatography. At our knowl-
edge, additions of these substituted allyllithium reagents
to imines were never reported, and it is noteworthy that
a complete regioselectivity was observed in all the reac-
tions described in Scheme 4, since only the linear
products coming from attack of the g-carbons of the
allylic moieties to the CvN bonds were obtained.k
This should be compared with previously reported reac-
tions of pentadienyllithium27 and cinnamyllithium28 with
aldehydes and ketones, where mixtures of branched and
linear homoallylic alcohols were generally obtained,
contrasting with the reactions of 1-trimethylsilylallyl-
lithium which gave the linear alcohols.25,29
1
underlined that the d.r. could be determined only by H
NMR spectroscopy of the crude product, because 2d decom-
posed during the GC and GC±MS analyses. On the other
hand, all the (R,R)- and (S,S)-diastereomers of the 1,2-
diamines 2a±c were obtained pure, and with reasonable
yields for the main diastereomers, by chromatographic
separations on SiO2 columns (entries 3, 6 and 11).
Aiming to further expand the synthetic scope of the
route described herein to C2-symmetric 1,2-diamines,
we then investigated the additions of a few organo-
lithium compounds which can be prepared by metalla-
tion of compounds having an acidic C±H bond (Scheme
4). Thus we prepared pentadienyllithium19±22 by metal-
lation of 1,4-pentadiene 9 using n-BuLi as the base in
THF±hexane mixtures, as well as phenylallyllithium23,24
from allylbenzene 10 (n-BuLi/TMEDA in THF/n-hex-
ane) and trimethylsilylallyllithium25 from allyltrimethyl-
silane 11 (t-BuLi±TMEDA in THF/n-pentane), slightly
modifying a procedure described for the lithiation of
allyldimethylphenylsilane.26 After the addition of the
It is worth pointing out that higher yields and/or diastereo-
selectivities in the preparation of the 1,2-diamines 2 were
provided by the substituted allyllithium reagents, with
respect to alkyl-, phenyl- and vinyllithium reagents. This
k
Our ongoing studies demonstrate that the retro-allyllithiation reactions of
branched homoallylic lithium amides occur even at 2788C, so that we
cannot exclude that the linear 1,2-diamines 2e±g are derived from the
initially formed branched isomers through a retro-allyllithiation/re-addition
sequence.
1,2-bis-imine 1 to the organolithiums 2788C, the
desired 1,2-diamines 2e±g were obtained with high
yields and moderate to good diastereoselectivities. It is

G. Martelli et al. / Tetrahedron 56 (2000) 8367±8374
8371
Scheme 4.
Scheme 5.
can be explained, in our opinion, by assuming an ionic
mechanism for the carbon±carbon bond formation. First,
it should be observed that the C±Li bond has a greater
ionic character than the C±MgX and C±ZnX bonds. We
suppose that the reactions of covalent organolithiums RLi,
which are dimeric and tetrameric in THF, take place through
the preliminary coordination of lithium by the bidentate 1,2-
bis-imine 1 to give the complex 12, followed by dissociation
to ionic species or triple ions (ion triplets) 13 (Scheme 5).
We previously took into consideration such a mechanism
for organometallic additions to activated imines.30,31 More-
over, we assume that allylic lithium compounds are ionic,
especially those having an electron-withdrawing substituent
at the allylic terminus, so their additions to the 1,2-bis-imine
1 should proceed through the corresponding ionic couples
13 (Scheme 5). Since allylic anions are stabilized by reso-
nance, it is likely that the (diastereo)selectivity in their addi-
tions to electrophiles is more easily controlled with respect
to the reactions of more reactive alkyl, vinyl and phenyl
anions. Finally, the occurrence, at least in part, of a mechan-
ism proceeding through the initial single electron transfer
(SET) from RLi to the 1,2-bis-imine 1 and the subsequent
radical recombination cannot be excluded,³13,32 although no
experimental result pointing to this mechanism was
obtained.
Conclusions
We have demonstrated that the double addition of organo-
lithium reagents to the 1,2-bis-amine 1 is, in certain cases, a
practicable method for the synthesis of enantiopure C2-
symmetric 1,2-diamines. A major limitation is the low dia-
stereoselectivity obtained using alkyl, vinyl and phenyl
reagents. Moreover, in these reactions care must be taken
to quench the mixtures with de-aerated H2O in order to
avoid the formation of imines as by-products coming from
the reaction of oxygen with the a-amido-radicals which are
formed in equilibrium with the main products, i.e. the
dilithium 1,2-diamides. Best results, i.e. high yield and dia-
stereoselectivity and complete regioselectivity, have been
obtained with substituted allyllithium compounds. This is
important because it opens a stereoselective route to novel
1,2-diamines exploiting a variety of substituted or functio-
nalized organolithium compounds, i.e. allylic, benzylic and
a-hetero-substituted reagents, which are easily available by

8372
G. Martelli et al. / Tetrahedron 56 (2000) 8367±8374
metallation of activated C±H bonds. We are actively pursu-
ing this objective in our laboratory.
5(R),6(R)-Di-[(S)-1-phenylethylamino]-decane (R,R)-2b.
Eluted second by column chromatography; yellowish oil,
1
0.740 g (38%); [a]D20ˆ2122.2 (c 1.3, CHCl3); H NMR
(300 MHz, CDCl3±D2O) d 0.76 (t, Jˆ7.2 Hz, and m, 10H,
CH3 and CH2), 1.10 (m, 4H, CH2), 1.37 (m, 4H, CH2), 1.40
(d, Jˆ3.0 Hz, 6H, CHCH3), 2.08 (m, 2H, NCHCHN), 3.81
(m, 2H, CHCH3), 7.10±7.38 (m, 10H, Ph). MS m/e (relative
intensity, %): 105 (100), 190 (85, M1/2), 86 (60). The di-
hydrochloride (R,R)-2b-2HCl was prepared by routine
procedure and puri®ed by several washings with CH2Cl2:
white solid, mp 218±2208C (dec.); [a]2D0ˆ231.9 (c 0.506,
CHCl3); 1H NMR (300 MHz, CDCl3±D2O) d 0.33 and 0.50
(2m, 4H, CH2), 0.67 (t, Jˆ6.9 Hz, 6H, CH2CH3), 1.0 (m,
4H, CH2), 1.64 (m, 2H, CH2), 1.98 (d, Jˆ6.9 Hz, 6H,
CHCH3), 2.11 (m, 2H, CH2), 2.49 (m, 2H, NCHCHN),
4.26 (m, 2H, CHCH3), 7.42 and 7.63 (2m, 10H, Ph), 10.1
and 10.7 (broad, 4H, NH2). Basic treatment of the salt
allowed to recover more pure (R,R)-2b: [a]2D0ˆ2134.6 (c
0.87, CHCl3). Found: C 82.00, H 10.68, N 7.32%; C26H40N2
requires: C 82.04, H 10.59, N 7.36%.
Experimental
General conditions
Melting points are uncorrected. Solvents were distilled over
the appropriate drying agent in N2 atmosphere before use:
THF (sodium benzophenone ketyl, then LiAlH4), Et2O and
DME (Na, then LiAlH4), n-hexane (Na), CH2Cl2 (P2O5).
Optical rotations were measured on a digital polarimeter
in
a
1-dm cell and [a]D-values are given in
1021 deg cm3 g21. H NMR spectra were recorded on a
Varian Gemini instrument at 300 or 200 MHz for samples
in CDCl3 which was stored over Mg: 1H chemical shifts are
reported in ppm relative to CHCl3 (dH 7.27) and J-values are
given in Hz. MS spectra were taken at an ionizing voltage of
70 eV on a Hewlett-Packard 5970 or 5890 spectrometer
with GLC injection. Chromatographic puri®cations were
performed on columns of SiO2 (Merck, 230±400 mesh) at
medium pressure. The following organolithium reagents
and organic compounds were purchased from Aldrich:
n-BuLi, t-BuLi, MeLi, PhLi, glyoxal trimeric dihydrate,
(S)-1-phenylethylamine, tetravinyltin, TMEDA, 1,4-penta-
diene, allylbenzene and allyltrimethylsilane. The 1,2-bis-
imine 1 was prepared according to the already described
procedure.11,12 All the organometallic reactions were
performed in a ¯ame-dried apparatus under a static atmos-
phere of dry N2.
1
(S,S)-2b. Eluted ®rst by column chromatography; yellowish
1
oil, 0.292 g (15%); [a]2D0ˆ275.8 (c 1.0, CHCl3); H NMR
(300 MHz, CDCl3±D2O) d 0.82 (t, Jˆ7.2 Hz, and m, 8H,
CH3 and CH2), 1.0±1.4 (m, 12H, CH2 and NH), 1.42 (d,
Jˆ3.0 Hz, 6H, CHCH3), 2.51 (m, 2H, NCHCHN), 3.6 (m,
2H, CHCH3), 7.10±7.38 (m, 10H, Ph). MS m/e (relative
intensity, %): 105 (100), 190 (85, M1/2), 86 (60). (S,S)-
2b-2HCl: white solid, mp 205±2078C (dec.); [a]2D0ˆ19 (c
0.36, CHCl3); 1H NMR (300 MHz, CDCl3) d 0.82 (t,
Jˆ6.7 Hz, 6H, CH3), 1.12 (m, 8H, CH2), 1.59 (m, 2H,
CH2), 1.72 (s, 2H, NH), 1.90 (d, Jˆ6.6 Hz, 6H, CHCH3),
2.20 (m, 2H, CH2), 3.24 (m, 2H, NCHCHN), 4.36 (m, 2H,
CHCH3), 7.42 (m, 6H, Ph), 7.65 (m, 4H, Ph), 10.4 (broad,
2H, NH).
Addition of organolithium compounds to the 1,2-bis-
imine 1. Preparation of 1,2-diamines 2a±c
General procedure: To the solution of the 1,2-bis-imine 1
(1.32 g, 5 mmol) in dry THF (20 ml) cooled at 2788C under
N2 was added the organolithium reagent (12.5 mmol) during
30 min. After stirring for 1 h, the mixture was quenched
with de-aerated H2O (10 ml), and the organic phase was
extracted with Et2O (3£20 ml). The collected ethereal
phase was dried (Na2SO4) and concentrated to leave the
crude 1,2-diamine 2 as an oil. The pure diastereomers
were obtained by chromatography on an SiO2 column elut-
ing with cyclohexane±EtOAc mixtures. The hydrochlorides
were prepared by treatment with excess 37% HCl, evapora-
tion and thorough drying in vacuo.
3(R),4(R)-Di-[(S)-1-phenylethylamino]-2,2,5,5-tetramethyl-
hexane (R,R)-2c. Eluted ®rst by column chromatography;
yellowish oil, 0.936 g (48%); [a]2D0ˆ173.4 (c 1.8, CHCl3);
we determined the same value for the compound prepared
by the reported procedure,10 where a de®nitely lower value
was given for the crude compound: [a]2D5ˆ131.6 (c 0.364,
CHCl3); MS m/z (relative intensity, %) 105 (100), 190 (79,
1
M1/2), 86 (50), 191 (15), 79 (14), 323 (12); the H NMR
spectrum was identical to that described for the known
compound.10
(S,S)-2c. Yellowish oil, 0.487 g (25%), contained ca. 5% of
(R,R)-2c; [a]2D0ˆ275.7 (c 1.2, CHCl3); 1H NMR (300 MHz,
CDCl3) d 0.78 (s, 18H, t-Bu), 1.37 (d, Jˆ6.6 Hz, 6H,
CHCH3), 1.91 (s, 2H, NH), 2.34 (s, 2H, NCHCHN), 3.84
(q, Jˆ6.6 Hz, 2H, CHCH3), 7.18±7.40 (m, 10H, Ph).
1(R),2(R)-Di-[(S)-1-phenylethylamino]-1,2-diphenyl-
ethane (R,R)-2a. Eluted ®rst by column chromatography;
white solid, 0.882 g (42%); [a]2D0ˆ2178.2 (c 0.65, CHCl3);
the compound was recrystallized from pentane: colourless
crystals, mp 108±110, [a]2D0ˆ2213.4 (c 0.32, CHCl3); lit:9
[a]2D0ˆ2205 (c 0.7, CHCl3); MS m/z (relative intensity, %)
105 (100), 106 (90), 210 (85, M1/2), 79 (21), 77 (18), 211
Addition of vinyllithium to the 1,2-bis-imine 1
1
(13); the H NMR spectrum was identical to that of the
compound described in the literature.9
MeLi (1.6 M in hexane, 7.5 ml, 12 mmol) was slowly added
to the solution of tetravinyltin (0.68 g, 3 mmol) in dry THF
(20 ml) cooled at 2208C under N2, and the mixture was
stirred during 2.5 h at 220 to 08C. Then the mixture was
cooled to 2788C and the solution of the 1,2-bis-imine 1
(1.056 g, 4 mmol) in dry THF (10 ml) was added during
20 min. After stirring for 1 h, the mixture was quenched
with de-aerated H2O (20 ml), and the organic phase was
(S,S)-2a.Yellowish oil, 0.210 g (10%), containing an
unidenti®ed impurity (ca. 10%); [a]2D0ˆ256.7 (c 1.6,
CHCl3); 1H NMR (300 MHz, CDCl3) d 1.29 (d, Jˆ6.6 Hz,
6H, CHCH3), 1.85 (br, 2H, NH), 3.59 (q, Jˆ6.3 Hz, 2H,
CHCH3), 3.88 (s, 2H, NCHCHN), 7.0±7.42 (m, 20H, Ph).

G. Martelli et al. / Tetrahedron 56 (2000) 8367±8374
8373
extracted with Et2O (3£20 ml). The collected ethereal phase
was dried (Na2SO4) and concentrated to leave an oil. Chro-
matography on an SiO2 column eluting with cyclohexane±
EtOAc (5:1) mixture gave 0.970 g (75%) of 2d as an inse-
parable mixture of the (R,R)/(S,S)-diastereomers (56:44 by
1H NMR), followed by (R,S)-2d (0.030 g, 2%).
17.1 Hz, 1H, CHvCH2), 5.52 (m, 1H, CHvCH2), 6.06
(dd, Jˆ15.3 and 15.3 Hz, 1H, CHvCH±CHvCH2), 6.31
(dt, Jˆ16.8 and 10.2 Hz, 1H, CH2CHvCH), 7.18±7.40 (m,
10H, Ph).
Addition of cinnamyllithium to the 1,2-bis-imine 1
3(R),4(R)-/3(S),4(S)-Di-[(S)-1-phenylethylamino]-1,5-
hexadiene (R,R)/(S,S)-2d. Eluted ®rst by column chroma-
tography; colourless oil; 1H NMR (200 MHz, CDCl3) d 1.30
and 1.32 (2 d, Jˆ6.6 Hz, 6H, CHMe), 1.70 (broad, 2H, NH),
2.61 (m, 1H, (R,R)-NCHCHN), 3.17 (m, 1H, (S,S)-NCH±
CHN), 3.79 (2q, Jˆ6.2 Hz, 2H, CHMe), 4.81±5.21 (m, 4H,
CHvCH2), 5.32±5.73 (m, 2H, CHvCH2), 7.18±7.42 (m,
10H, Ph). Found: C 82.50, H, 8.83, N, 8.71%; C22H28N2
required: C 82.45, H 8.81, N 8.74%.
To dry THF (12 ml) cooled at 2788C were added n-BuLi
(1.6 M in hexane, 9.4 ml, 15 mmol), then TMEDA (2.39 ml,
16 mmol) and, after 10 min, allylbenzene 10 (2.07 g,
17.5 mmol). The mixture was stirred for 1 h, meanwhile
the temperature was allowed to reach 08C. A deep red colour
was observed and attributed to phenylallyllithium. The
mixture was cooled to 2788C and the 1,2-bis-imine 1
(1.32 g, 5 mmol) dissolved in THF (7 ml) was added drop-
wise while stirring. After 1 h the reaction was quenched
with H2O (5 ml) and the organic phase was extracted with
Et2O (3£20 ml). The collected ethereal layers were dried
(Na2SO4), and concentrated to leave the crude product 2f
(R,S)-2d. Colourless oil; [a]2D5ˆ291 (c 0.24, CHCl3); H
1
NMR (300 MHz, CDCl3) d 1.20 and 1.34 (2d, Jˆ6.6 Hz,
s H, CHMe), 1.66 (broad, 2H, NH), 2.85 and 2.94 (2m, 2H,
NCHCHN), 3.47 and 3.76 (2q, 2H, CHMe), 4.85±5.20 (m,
4H, CHvCH2), 5.46±5.58 (m, 2H, CHvCH2), 7.05±7.38
(m, 10H, Ph).
1
as a yellow thick oil: d.r. 78:22 by H NMR spectroscopy.
The diastereomers were separated by chromatography on an
SiO2 column eluting with cyclohexane±EtOAc (95:5).
(E,E)-4(R),5(R)-Di-[(S)-1-phenylethylamino]-1,8-diphe-
nyl-1,7-octadiene (R,R)-2f. Eluted second by column chro-
matography; yellowish thick oil; 1.12 g, 45%: [a]2D5ˆ274
Addition of pentadienyllithium to the 1,2-bis-imine 1
1
(c 0.7, CHCl3); H NMR (300 MHz, CDCl3): d 1.24 (d,
To dry THF (30 ml) cooled at 2788C were added n-BuLi
(1.6 M in hexane, 15.9 ml, 25.5 mmol) and, after 10 min,
1,4-pentadiene 9 (2.80 ml, 27.2 mmol, taken by a syringe
from the holder cooled at 08C). The mixture was stirred for
1 h, meanwhile the temperature was allowed to reach 08C.
Two phases were observed, the orange upper phase contain-
ing pentadienyllithium. The mixture was cooled to 2788C
and the 1,2-diimine 1 (2.25 g, 8.5 mmol) dissolved in THF
(10 ml) was added dropwise during 1 h while stirring. After
a further 1.5 h the reaction was quenched with H2O (10 ml)
and the organic phase was extracted with Et2O (3£30 ml).
The collected ethereal layers were dried (Na2SO4), and
concentrated to leave crude diamine 2e as a yellowish
Jˆ6.6 Hz, 6H, CHMe), 1.36 (br, 2H, NH), 2.74 (m, 6H,
NCHCH2), 3.82 (q, Jˆ6.6 Hz, 2H, CHMe), 5.60±5.83 (m,
4H, CHvCH), 7.0±7.35 (m, 20H, Ph). Found: C 86.40, H
8.08, N 5.58%; C36H40N2 requires: C 86.35, H 8.05, N
5.60%.
(S,S)-2f. Eluted ®rst by column chromatography; yellowish
1
thick oil; 0.412 g, 16%: [a]2D5ˆ243.7 (c 0.6, CHCl3); H
NMR (300 MHz, CDCl3): d 1.14 (d, Jˆ6.9 Hz, 6H, CHMe),
1.46 (br, 2H, NH), 1.98 (m, 2H, NCHCH2), 2.50±2.64 (m,
4H, NCHCH2), 3.58 (q, Jˆ6.9 Hz, 2H, CHMe), 5.96 (m,
2H, CH2CHvCH), 6.37 (d, Jˆ15.3 Hz, 2H, CHvCH±
Ph), 7.02±7.30 (m, 20H, Ph).
1
thick oil: 3.30 g; d.r. 80:20 by H NMR. The pure dia-
stereomers were obtained by chromatography on an SiO2
column eluting with cyclohexane±AcOEt (15:1).
Addition of 1-trimethylsilylallyllithium to the 1,2-bis-
imine 1
(E,E)-6(R),7(R)-Di-[(S)-1-phenylethylamino]-1,3,9,11-
dodecatetraene (R,R)-2e. Eluted second by column chro-
matography; 2.35 g, 69%: [a]2D5ˆ298.7 (c 0.7, CHCl3); 1H
NMR (300 MHz, CDCl3) d 1.26 (d, Jˆ6.6 Hz, CHMe), 1.40
(s, 2H, NH), 2.14 (m, 6H, NCHCH2), 3.72 (q, 2H, CHMe),
To a solution of allyltrimethylsilane 11 (1.71 g, 15 mmol)
and TMEDA (0.75 ml, 5 mmol) in dry THF (3 ml), cooled
at 2788C, was slowly added t-BuLi (1.7 M in pentane,
2.95 ml, 5 mmol). The mixture was allowed to reach slowly
2158C and further stirred at this temperature for 1 h, so
assuming an orange colour. The mixture was then cooled
to 2788C and the 1,2-bis-imine 1 (0.264 g, 1 mmol)
dissolved in THF (3 ml) was added dropwise while stirring.
After 1 h the reaction was quenched with de-aerated H2O
(5 ml) and the organic phase was extracted with Et2O
(3£10 ml). The collected ethereal layers were dried
(Na2SO4), and concentrated to leave the crude product 2g
as a clear yellowish oil: d.r. 9:68:23 by GC±MS analysis.
Column chromatography of the crude product coming from
the addition of 1-trimethylsilylallyllithium gave the major
diastereomer (R,R)-2g in a pure state; fractions containing
mixtures of the other two linear diastereomers (R,S)- and
(S,S)-2g were also eluted, but no branched compound was
observed.
4.85 (d, Jcisˆ10.2 Hz, 1H, CHvCH2), 4.92 (d, Jtrans
ˆ
16.8 Hz, 1H, CHvCH2), 5.20 (m, 1H, CHvCH2), 5.53
(dd, Jˆ15.3 and 15.3 Hz, 1H, CHvCH±CHvCH2), 6.09
(dt, Jˆ17.1 and 10.5 Hz, 1H, CH2CHvCH), 7.18±7.32 (m,
10H, Ph); the spectrum revealed the absorptions of another
compound, perhaps (R,S)-2e (,10%) Found: C 83.93, H
9.04, N 7.00%; C28H36N2 requires: C 83.95, H 9.06, N
6.99%.
(S,S)-2e was eluted ®rst by column chromatography;
0.204 g, 6%: [a]D25ˆ18.6 (c 0.78, CHCl3); 1H NMR
(300 MHz, CDCl3) d 1.21 (d, Jˆ6.6 Hz, CHMe), 1.43 (s,
2H, NH), 1.92 (m, 2H, NCHCH2), 2.44 (m, 2H, NCHCH2),
2.59 (m, 2H, NCHCHN), 3.63 (q, Jˆ6.6 Hz, 2H, CHMe),
5.01 (d, Jcisˆ10.5 Hz, 1H, CHvCH2), 5.12 (d, Jtrans
ˆ

8374
G. Martelli et al. / Tetrahedron 56 (2000) 8367±8374
11. Alvaro, G.; Grepioni, F.; Savoia, D. J. Org. Chem. 1997, 62,
4180±4182.
(E,E)-1(R),2(R)-Di-[(S)-1-phenylethylamino]-1,8-di-(tri-
methylsilyl)-1,7-octadiene (R,R)-2g. Yellowish oil;
0.210 g (43%); [a]D20ˆ297.5 (c 0.9, CHCl3); 1H NMR
(300 MHz, CDCl3): d 20.086 (s, 18H, SiMe3), 1.25 (d,
Jˆ6.6 Hz, 6H, CHMe), 1.39 (br, 2H, NH), 2.12 (m, 6H,
NCHCH2), 3.71 (q, Jˆ6.6 Hz, H, CHMe), 5.19 (d,
Jˆ18.6 Hz, SiCHvCH), 5.54 (dt, Jˆ18.6 and 6.2 Hz,
SiCHvCH), 7.10±7±38 (m, 10H, Ph); MS m/z 105 (100),
246 (75, M1/2), 73 (22), 142 (21), 247 (17). Found: C 73.05,
H 9.84, N 5.69%; C30H48N2Si2 requires: C 73.10, H 9.82, N
5.68%.
12. Alvaro, G.; Grepioni, F.; Grilli, S.; Martelli, G.; Savoia, D.
Synthesis 2000, 581±587.
13. Stamp, L.; tom Dieck, H. J. Org. Chem. 1984, 277, 297±309.
14. Alexakis, A.; Aujard, I.; Mangeney, P. Synlett 1998, 875±876.
15. Smith, J. G.; Ho, I. J. Org. Chem. 1972, 37, 653±656.
16. Sbarbati Nudelman, N.; Doctorovich, F. J. Chem. Soc., Perkin
Trans. 2 1994, 1233±1238.
17. Sbarbati Nudelman, N.; Outumuro, P. J. Org. Chem. 1982, 47,
4347±4348.
18. Amurrio, D.; Khan, K.; KuÈndig, E. P. J. Org. Chem. 1996, 61,
2258±2259.
Acknowledgements
19. Bates, R. B.; Gosselink, D. W.; Kaczynski, J. A. Tetrahedron
Lett. 1967, 199±204.
Á
Financial support from the Ministero dell'Universita e della
20. Bates, R. B.; Gosselink, D. W.; Kaczynski, J. A. Tetrahedron
Lett. 1967, 205±210.
Ricerca Scienti®ca e Tecnologica, Roma, and The Univer-
sity of Bologna (National Project `Stereoselezione in Sintesi
Organica. Metodologie ed Applicazioni') is gratefully
acknowledged.
21. Yasuda, H.; Nakamura, A. J. Organomet. Chem. 1985, 285,
15±29.
22. Pratt, L. M.; Streitwieser, A. J. Org. Chem. 2000, 65, 290±
294.
References
23. Sandel, V. R.; McKinley, S. V.; Freedman, H. H. J. Am. Chem.
Soc. 1968, 90, 495±497.
1. Lucet, D.; Le Gall, T.; Mioskowski, C. Angew. Chem., Int. Ed.
Engl. 1998, 37, 2580±2627.
24. Balzer, H.; Berger, S. Chem. Ber. 1991, 125, 733±737.
25. Corriu, R. J. P.; Guerin, C.; M'Boula, J. Tetrahedron Lett.
1981, 31, 2985±2986.
2. Bennani, Y. L.; Hanessian, S. Chem. Rev. 1997, 97, 3161±3195.
3. Witiak, D.; Wei, Y. J. Org. Chem. 1991, 56, 5408±5417.
4. Jones, D. S.; Srinivasan, A.; Kasina, S.; Fritzberg, A.;
Wilkening, D. W. J. Org. Chem. 1989, 54, 1940±1943.
5. Pasini, A.; Zunino, F. Angew. Chem., Int. Ed. Engl. 1987, 26,
615±624.
26. Uno, H. Bull. Chem. Soc. Jpn 1986, 59, 2471±2479.
Â
27. Gerard, F.; Miginiac, P. Bull. Soc. Chim. Fr. 1974, 1924±
1930.
Â
28. Gerard, F.; Miginiac, P. Bull. Soc. Chim. Fr. 1974, 1924±
2533.
6. Sherman, S. E.; Lippard, S. J. Chem. Rev. 1987, 87, 1153±1181.
7. Enders, D.; Meiers, M. Angew. Chem., Int. Ed. Engl. 1996, 35,
2261±2263.
29. Katritzky, A. R.; Pif¯, M.; Lang, H.; Anders, E. Chem. Rev.
1999, 99, 665±722.
30. Alvaro, G.; Savoia, D.; Valentinetti, M. R. Tetrahedron 1996,
52, 12571±12586 (and references cited therein).
31. Alvaro, G.; Pacioni, P.; Savoia, D. Chem. Eur. J. 1997, 3,
726±731.
8. Neumann, W. L.; Rogic, M. M.; Dunn, T. J. Tetrahedron Lett.
1991, 42, 5865±5868.
9. Bambridge, K.; Begley, M. J.; Simpkins, N. S. Tetrahedron
Lett. 1994, 35, 3391±3394.
32. Alvaro, G.; Martelli, G.; Savoia, D. J. Chem. Soc., Perkin
Trans. 1 1998, 775±783 (and references cited therein).
10. Roland, S.; Mangeney, P.; Alexakis, A. Synthesis 1999, 228±
230.