Salt/ligand-activated low-valent titanium formulations: the 'salt effect' on diastereoselective carbon-carbon bond forming SET reactions

Article abstract of DOI:10.1016/j.tet.2008.05.084

A comprehensive study on the influence of exogenously added electropositive metal salts as promoters/secondary activators on preformed LVT species has resulted in the construction of highly efficient low-valent titanium (LVT) reagents. These salt-activated LVT reagents while exhibiting enhanced chemoselectivity and diastereoselectivity accelerated the reductive olefination rates of aromatic and aliphatic carbonyls under ambient temperature conditions and in much reduced reaction times. The versatility of the salted reagent was further explored in other single electron transfer reactions, namely, imino-pinacol couplings and one-pot synthesis of phenanthrenes from o-alkoxy aromatic carbonyls. We envisage that, in contrast to multiphase heterogeneous colloidal slurries, salt-activated LVT reagents afforded uniformly viscous homogeneous slurries generating a highly reactive monomeric intermetallic LVT complex. Continued judicious exploration of the emerging paradigms by studying the influence of external ligands/auxiliaries/redox agents on LVT reagents, and organometallics in general, will be critical to widen the scope and utility of the classical McMurry reaction and other SET reactions.

Full text of DOI:10.1016/j.tet.2008.05.084

Tetrahedron 64 (2008) 7225–7233
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Salt/ligand-activated low-valent titanium formulations: the ‘salt effect’
on diastereoselective carbon–carbon bond forming SET reactions
,
y
*
*
*
Shyam M. Rele , Sandip K. Nayak , Subrata Chattopadhyay
Bio-Organic Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 April 2008
Received in revised form 14 May 2008
Accepted 16 May 2008
Available online 21 May 2008
A comprehensive study on the influence of exogenously added electropositive metal salts as promoters/
secondary activators on preformed LVT species has resulted in the construction of highly efficient low-
valent titanium (LVT) reagents. These salt-activated LVT reagents while exhibiting enhanced chemo-
selectivity and diastereoselectivity accelerated the reductive olefination rates of aromatic and aliphatic
carbonyls under ambient temperature conditions and in much reduced reaction times. The versatility of
the salted reagent was further explored in other single electron transfer reactions, namely, imino-pinacol
couplings and one-pot synthesis of phenanthrenes from o-alkoxy aromatic carbonyls. We envisage that,
in contrast to multiphase heterogeneous colloidal slurries, salt-activated LVT reagents afforded uniformly
viscous homogeneous slurries generating a highly reactive monomeric intermetallic LVT complex.
Continued judicious exploration of the emerging paradigms by studying the influence of external
ligands/auxiliaries/redox agents on LVT reagents, and organometallics in general, will be critical to widen
the scope and utility of the classical McMurry reaction and other SET reactions.
Keywords:
Low-valent titanium
Activation
Deoxygenation
Electron transfer
Inorganic Grignard
Monomerization
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
dimerization of the starting ketone or aldehyde via electron
transfer (ketyl anion radical) to form a carbon–carbon bond
Designing of reagents with appropriate reactivity or improving
the existing methodologies for selective organic transformations is
an important domain of organometallic chemistry. For the transi-
tion metal-based reagents in general, and for low-valent titanium
(LVT) in particular,1–4 it is widely known that the reactivity, re-
producibility, and stereochemistry of reaction products vary greatly
with the source of the metal, its method of preparation, and the
experimental conditions. Due to the unique features of high oxo-
(titanium pinacolates) and (ii) subsequent deoxygenation of the
1,2-diolate intermediate involving cleavage of the carbon–oxygen
bonds to give the alkene (rate determining step) along with the
thermodynamically stable Ti-oxides (TiO2). While the dimerization
of carbonyls to generate the pinacols (step 1) has been accom-
plished with a variety of reducing metals (Zr, Sn, Sm, Nb, Ce, In, V),8
including LVT,1–7 deoxygenation of pinacols to olefins (step 2) is
rather unique to LVT reagents. The extrusion of oxygen by titanium
(oxophilic) from the pinacolates necessitates the use of solvent-
reflux temperatures and prolonged reaction times affording the
alkene. While many of the otherwise reducible functionalities
survive the McMurry reaction at low temperatures (which pre-
dominantly generates pinacols), the incompatibility of several
functionalities under the refluxing reaction conditions limits its
applications. These drawbacks apparently restrict the utility of LVT
reagents in the case of oxygenated complex natural products with
semicompatible functionalities, where the introduction of olefinic
double bonds is achieved in two steps via initial pinacolization at
lower temperatures followed by deoxygenation through indirect
milder methods.1 For the synthesis of olefins at lower tempera-
tures, the ‘activation’ of LVT species therefore becomes imperative.
Moreover, the intrinsic tendency of the reactive metals toward
deactivation necessitates further depassivation or secondary acti-
vation.2 Thus, the motivation to generate new LVT reagent(s) which
not only possesses enhanced reactivity (due to activation) but is
philicity (
D
H¼ꢀ225.8 kcal/mol) and reducing power (Ti/Tiþ2
þ2e–; E0þ¼1.63 V), low-valent titanium (LVT) reagents have gained
widespread acceptance in organic synthesis.1–7 In particular, the
remarkable scope of LVT reagents to effect reductive coupling of
carbonyls (McMurry reaction) has resulted in a variety of applica-
tions such as synthesis of strained olefins,1 heterocyclic com-
pounds,5 and macrocyclic ring systems6 to complex natural
products including paclitaxel.7
The LVT-induced reductive deoxygenation of carbonyls to ole-
fins takes place in two successive steps1 (Scheme 1): (i) reductive
* Corresponding authors. Tel.: þ1 626 404 6144; fax: þ1 626 305 9094 (S.M.R.).
E-mail addresses: shyamrele@yahoo.com, shyamrele@gmail.com (S.M. Rele),
sknayak@magnum.barc.ernet.in (S.K. Nayak), schatt@apsara.barc.ernet.in (S.
Chattopadhyay).
y
Present address: 129 N. Hill Avenue, Ste 104, Pasadena, CA 91106, USA.
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.05.084

7226
S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
Scheme 1. Classical mechanism of C–C bond formation in LVT-mediated reductive deoxygenation of aldehyde/ketone.
also capable of direct one-step olefination at ambient/low tem-
peratures would serve as ideal low-valent formulations in tita-
nium-induced single electron transfer (SET) reactions.
Besides the native state of the metal, the stability, tendency
for aggregation, and the reactivity of LVT reagents are highly
dependent on the coordinating solvents/auxiliaries and the stabil-
ity of the complexes formed in situ which in turn are influenced by
the steric and electronic factors. Consequently, the addition or
subtraction of electron(s) can dramatically alter the redox potential
of the native titanium species, and therefore, the type of chemistry
the generated active or passive metal center (Ti) might mediate.
To this end, in continuation of our earlier work3 on the rational
design of organometallic reagents in electron transfer processes,
it has been shown by us3 and others4 that the reducing ability of
LVT-based reagents can be rationally tuned by the simple addition
Figure 1. Intermetallic (inorganic Grignard) and metal hydride species generated from
different LVT preparations.
These results strongly indicate that depending on the reducing
agent used, a significant proportion of the free salt produced in situ
during the generation of the active LVT species is consumed in the
formation of the reactive intermetallic species which is chemically
bound to the low-valent titanium nucleus (Ti–Mg/Ti–H/Ti–Li
bonded bimetallic complex). Significantly, the active titanium
species generated by each method may therefore differ not only in
its reducing and Lewis acid properties but also in the surface bound
ligands, morphology, aggregation and formal oxidation state (redox
state). The low-valent Ti metal center thus emerges as a tunable
functionality whose redox properties and reactivity could therefore
be modulated and/or controlled by the exogenous addition of metal
salts (salt tuning) and/or mixture of solvating ligands (solvent–
ligand tuning). Addition of neutral salts as promoters has been
known to enhance the Lewis acidity of active metal center and the
resultant ionic strength of the medium, thereby influencing the
rates of many organometallic electron transfer reactions (salt
effect).11 All told, since single electron transfer is an elementary
reaction step in LVT-mediated reductive transformations, modula-
tion of the LVT-mediated redox process by judicious incorporation
of metal salts (as promoters) may not only dramatically augment
the reactivity and selectivity of the C–C bond formation in McMurry
olefination and other SET reactions, but may also provide an al-
ternative route to ‘activation’ of LVT reagents under redox potential
control. Despite the significance of in situ generated salts in LVT
preparations, the direct influence of electropositive metal salts
(mono-, bivalent) as promoters on the reactivity of LVT-based re-
agents has not received much attention, and is therefore, long
overdue. In continuation of our preliminary work on salted LVT
reagents,3b the present investigation is a comprehensive study on
the influence of exogenously added metal salts/solvents (redox
of co-solvents, external ligands (p-acid species such as pyridine,
triphenylphosphine, and fullerenes), and chemical redox agents
(arenes, I2, salts).3 While the surrounding electronic environment is
at the heart of LVT chemistry, the actual reactive metal species re-
sponsible for the chemical transformation and its genesis has been
put to considerable debate.
Conventionally, LVT reagents have been prepared using Rieke’s
protocol involving reduction of titanium halides with an alkali
metal (lithium, sodium, or potassium) in an ethereal or hydrocar-
bon solvent.2 Based on ESR studies involving stoichiometric re-
action of TiCl3 with various reducing agents (Li, Mg, LiAlH4) in THF
(Rieke type activation process), Geise et al. proposed the formation
of finely suspended Ti species in zero-valent state adsorbed on the
precipitated inorganic salt byproducts (LiCl, MgCl2).9 However,
Bogdanovic et al. suggested that the actual scenario involving the
generation of activated LVT species in McMurry olefination is far
more complex and involves the stepwise formation of bimetallic
inorganic Grignard reagents as the active species highlighting
the crucial role of in situ released salts.1c,2b,10 For example, studies
involving Tyrlik’s reagent (TiCl3–Mg–THF) produced a highly sol-
uble, covalently bonded paramagnetic Ti–Mg species (Ti–Mg–Clx
bimetallic complex) coordinated to the solvent molecules. The
reaction involved the initial formation of [TiMgCl2$xTHF] which
further reacts with the excess Mg giving 1 mol of Ti(MgCl)2THF (I),
the actual reducing species, along with 0.5 mol of free MgCl2
(Fig. 1). Analogous ESCA studies by Bogdanovic and his co-workers
on McMurry reagent (TiCl3–LiAlH4–THF) have revealed the for-
mation of Ti(II) chlorohydride [HTiCl(THF)w0.5] (II) as the active LVT
species.2b,10

S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
7227
Table 1
agents) as ligand-tuning control elements in modulating and
enhancing the reactivity of LVT reagents. Significantly, the chemi-
cally engineered ‘salt-activated’ LVT reagent(s) performs the dia-
stereoselective C–C bond formation in McMurry’s carbonyl coupling
and other SET-inducedreactions notonlyefficientlyand atenhanced
reaction rates, but also at ambient temperatures and in improved
yields, thereby presenting an emerging paradigm critical to actual-
izing the preparative potential in LVT chemistry in SET reactions.
Influence of salts on the reactivity of LVT reagents and its outcome on the reductive
coupling of acetophenone (1a) to stilbene (3a)
Entry
Reagenta
Saltb
Temp (ꢂC)/
time (h)
Product yieldsc,d (%)
Pinacol 2a
(dl:meso)
Stilbene 3a
(E:Z)3a,c
1
2
3
4
5
6
7
8
A
A
A
A
A
A
A
C
A
B
B
B
B
B
B
D
Nil
Nil
25/16.0
Reflux/16.0
25/16.0
25/16.0
25/16.0
25/2.0
89 (75:25)
d
Trace
87 (75:25)
25 (65:35)
22 (62:38)
Trace
65 (75:30)
82 (85:15)
MgCl2
MgCl2
LiI
CsCl
KCl
KCl
ZnCl2
Nil
62 (70:30)
64 (62:38)
75 (80:20)
d
e
2. Results and discussion
25/2.0
25/2.0
d
2.1. ‘Salt-activated’3b LVT reagents: influence of metal salts on
the preparation and reactivity of activated LVT reagents and
the stereochemical outcome in McMurry olefination
No reaction
9
25/16.0
25/16.0
Reflux/16.0
25/16.0
25/16.0
25/16.0
25/16.0
25/2.0
45 (74:26)
85
32 (32:68)
d
10
11
12
13
14
15
16
Nil
Trace
25
55
18
57
d
85 (76:24)
52 (44:56)
36 (51:49)
65 (40:60)
30 (20:80)
86 (65:35)
LiCl
CsCl
KCl
ZnCl2
KCl
In the present studies, TiCl3–Li–THF (McMurry reagent) and
TiCl3–Mg–THF (Tyrlik’s reagent) were selected as the LVT-source on
the grounds of their unique reactivity toward carbonyl coupling
reactions. A set of ‘activated’ LVT reagents were generated in situ
by external addition of alkali metal/alkaline earth metal salts/
amphoteric salts to the preformed low-valent titanium species
obtained from the above systems and the potential of these salted
LVT reagents in McMurry olefination was then investigated.
Reductive dimerization of acetophenone (1a) to 2,3-diphenyl-2-
butene (3a) (Scheme 2) was chosen as the model reaction to opti-
mize the influence of salted LVT reagent and the results are
summarized in Table 1.
a
Reagent A: TiCl3–Li–THF; reagent B: TiCl3–Mg–THF (Tyrlik’s reagent); reagent C:
TiCl3–Li–DME; reagent D: TiCl4–Li–THF.
b
Amount of salt used: 2 equiv.
c
All yields refer to isolated products (purity >95%, analyzed by 1H NMR).
d
Stereochemical assignments (calculation of E:Z and dl:meso ratio) were made
by 1H NMR and by comparison of spectroscopic data to that reported in the liter-
ature.3a,c,14 For 3a: 1H NMR
d (ppm) 1.86 (E) and 2.16 (Z) (2s, 6H), 7.0–7.3 (m, 10H).
e
Amount of salt used: 8 equiv.
at ambient temperature resulted in a facile dimerization affording
3a within 2 h in very high yield (82%) (Table 1, entry 7) with no
trace of pinacol. Changing the titanium salt from trichloride to
tetrachloride (TiCl4) in the salt-activated LVT preparation (TiCl4–Li–
THF–KCl, reagent D) had marginal influence on the pinacol to stil-
bene product balance with stilbene 3a obtained as the sole product
(Table 1, entry 16). Furthermore, the influence of divalent ampho-
teric salt like ZnCl2 on the activation of LVT preparation was
observed to be less efficient with both pinacol 2a and stilbene 3a
being produced in moderate yields (Table 1, entry 9). While 2 equiv
of salt was sufficient for optimum LVT activation (cf. entries 3 and
4), the LVT reagent was insignificantly affected by the nature of the
anionic part (LiI) of the added salt (Table 1, entry 5). However, the
effect of the reaction medium was more dramatic on the product
outcome. For example, DME-solvated salted LVT (TiCl3–Li–DME–
KCl) rendered the reagent ineffective by completely suppressing
the C–C bond formation (pinacolization) even at room temperature
resulting in quantitative recovery of the unreacted substrate 1a
(Table 1, entry 8). Analogous reactions studying the influence of
salts on reagent B (TiCl3–Mg system) again indicated that the best
results were observed when KCl was used as the metal salt (Table 1,
entry 14), albeit less efficiently in comparison to reagent A–KCl
system. Thus, reaction of 1a with reagent B–KCl while exhibiting
pronounced reactivity furnished 3a (65%) at ambient temperature
(16 h) (Table 1, entry 14); the same reaction when carried out
without the addition of salt (reagent B) afforded solely the pinacol
2a (85%) over the same period of time (Table 1, entry 10).9,13 Based
on all the above observations, reagent A–KCl (2 equiv) was pri-
marily used as the reagent of choice for McMurry olefination and
other SET reactions. Moreover, the above results clearly demon-
strate the dramatic influence of exogenously added metal salts on
the preformed LVT reagents in enhancing their reactivity in
McMurry reaction (C–C bond formation).
Scheme 2. McMurry olefination via salted LVT reagents.
Conventional McMurry olefination using TiCl3–Li–THF (reagent
A)12 and TiCl3–Mg–THF (reagent B)9,13 afforded only the pinacol 2a
(87–89% yield) at ambient temperatures, while the stilbene 3a was
isolated as the predominant product (ꢁ85% yield) only after pro-
longed (16 h) refluxing (Table 1, entries 1, 2, 10, 11).12 In contrast,
reaction of 1a with reagent A–MgCl2 (2 equiv) furnished the stil-
bene 3a (25%) along with the pinacol 2a (62%) even at ambient
temperature in 16 h (Table 1, entry 3), clearly indicating the positive
influence of external salt addition (MgCl2) on the activity of the
reagent A. Addition of excess amount of MgCl2 (8 equiv) did not
show any significant change (Table 1, entry 4) on the diastereomeric
outcome of the products (2a/3a) formed. Based on these pre-
liminary observations, an extensive study on the conversion of 1a
to 3a was carried out using different electropositive mono-/divalent
metal salts with varying size/charge such as CsCl, KCl, LiI and ZnCl2.
Surprisingly, as indicated by the results in Table 1, the more elec-
tropositive metal salts KCl and CsCl enhanced the activity of reagent
A significantly. For example, addition of CsCl to the LVT preparation
(reagent A) led to the formation of 3a in appreciable yield (65%)
(Table 1, entry 6), while addition of KCl under identical conditions
2.1.1. Diastereoselectivity and steric course of stilbene formation
The stereoselectivity during McMurry olefination is a function of
the number of interdependent variables, such as solvent, the nature

7228
S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
of the active reagent (valences etc.), external auxiliaries, tempera-
ture, and steric bulk of the groups.1–3 The study of the stereo-
chemical outcome (E/Z) of the olefination with the above salted LVT
reagents (Table 1) was revealing. In general, use of the salted
reagent A and reagent D (alkali metal Li-based reductant) led to
preferential formation of E-3a isomer as the major product (Table 1,
entries 3, 4, 6, 7, and 16). Surprisingly, a complete reversal in ster-
eoselectivity was observed on addition of a divalent ampholyte
ZnCl2 (reagent A–ZnCl2 combination), affording Z-3a as the pre-
dominant product (Table 1, entry 9). Moreover, the preponderance
of E-3a olefination was augmented in case of reagent A–KCl
(E:Z 85:15) by carrying out the reaction at ambient temperature
(Table 1, entry 7) in comparison to the conversion of 1a to 3a under
prolonged reflux (16 h) when reagent A was solely used (E:Z 75:25)
(Table 1, entry 2). Subtle variation in the preparation of LVT reagent
(TiCl3 replaced TiCl4) although resulted in comparable stilbene
formation, w24% loss in E:Z stereoselectivity (E:Z 85:15 to
E:Z 65:35) was observed in the case of KCl-activated TiCl4 reagent
(Table 1, see entries 7 and 16). In contrast, when the salt-modified
LVT reagents derived from Tyrlik’s reagent (TiCl3–Mg–THF, alkali
earth metal-based reductant) were employed, the Z-3a isomer was
preferentially obtained over the E-3a isomer (Table 1, entries 12–
15). The influence of salts on the diastereoselectivity bias of 3a
when reacted with reagent A (Li as reductant) and reagent B (Mg as
reductant) is consistent with our earlier finding involving genera-
tion of reactive LVT species obtained by reducing TiCl3 with in
situ formed chemically redox metal–arene systems acting as the
soluble organic reductants.3a Specifically, while the LVT reagents
derived from TiCl3–Mg–arenes–THF (bivalent akaline earth metal
reductant) showed similar bias toward the formation of the
Z-3a stilbene, a complete reversal in stereoselectivity (E-3a dia-
stereomer) was observed by switching over to monovalent alkali
metal–arene-based reductants (TiCl3–Li–arenes–THF). Our results
summarized in Table 1 demonstrate that the stereochemical out-
come of the products is amenable to tuning by judicious design of
reagents and reaction conditions (salts, ligands/auxiliaries). The
ratio of E- and Z-3a stilbene was determined by 1H NMR spec-
troscopy,3a,c,14 where the NMR spectrum showed two characteristic
peaks at 1.86 and 2.16 ppm corresponding to the E- and Z-stilbene
stereoisomers,3a,c respectively (m/z¼Mþ 208.09).
7–9), which are otherwise cleaved under refluxing conditions,3f,15
remained unaffected during olefination at ambient (low) tem-
peratures, ensuring excellent chemoselectivity in the McMurry
olefination. Moreover, the activated reagent A–KCl exhibited
diastereoselective bias toward the predominant formation of the
respective E-stilbenes as the major product based on 1H NMR
spectroscopy.3a,c,14
One of the striking limitations of the McMurry reaction has been
its inability to effectively couple aliphatic carbonyls to olefins.16 The
relatively strong alkyl–oxygen bonds in the intermediate pinaco-
lates16 require prolonged refluxing for the required deoxygenation
to occur. Even then, the yields of the product olefins are often
unsatisfactory. However, the low propensity of alicyclic ketones
(1k, 1l) and aliphatic aldehydes (1m, 1n) (Table 3, entries 1–4) to
couple was greatly augmented using reagent A–KCl combination.
As such, reductive dimerization of alicyclic ketones (1k, 1l) and
aldehydes (1m, 1n) to their respective olefins (3k–n) was accom-
plished at ambient temperature in yields almost comparable to
olefins obtained from aromatic substrates (see Scheme 2, Table 3).
2.2. One-pot synthesis of phenanthrenes
Previously, a short one-step synthesis of phenanthrenes (6a)
from o-alkoxy aromatic aldehydes/ketones (4a) was developed in
our laboratory, albeit in poor yield and after prolonged heating
(36%, 16 h reflux).3f Mechanistically, the synthesis involves multiple
steps in tandem, viz., (i) preferential formation of the Z-stilbene
(an apriori condition for phenanthrene formation) followed by
(ii) ortho dealkoxylation (determining factor) of the aryl alkyl ether
functionality in Z-stilbene to generate the aryl radical and its sub-
sequent C–C coupling leading to ring cyclization (Scheme 3). In the
present work, the applicability and efficiency of the various salt-
activated LVT preparations in the synthesis of phenanthrenes were
explored.
20-Methoxypropiophenone (4a) was chosen as the model sub-
strate to investigate the influence of salted LVT reagents. When 4a
was subjected to coupling with Tyrlik’s reagent TiCl3–Mg–THF
(reagent B), stilbene 5a [2,3-bis-(20-methoxyphenyl)-hex-3-ene]
was isolated as the sole product obtained under refluxing condi-
tions (Table 4, entry 1).3f However, coupling of 4a using a salted
TiCl3–Mg–THF (reagent B–LiCl), afforded 9,10-diethyl phenan-
threne 6a in 17–20% yield along with the stilbene 5a (65%) (Table 4,
entries 3 and 4). Interestingly, addition of sub-stoichiometric
amount of Li metal as a co-reductant with Mg (Table 4, entry 5) in
2.1.2. Generality and selectivity
To explore the generality and scope of the reagent A–KCl,
experiments were carried out using a variety of substituted aryl
carbonyls, such as aryl alkyl ketones (Table 2, entries 1–4), aryl
aldehydes (Table 2, entries 6–9), and a diaryl ketone (Table 2, entry
5) (Scheme 2). In all the cases, the respective olefins were obtained
smoothly at room temperature and in good yields (65–85%) as
the exclusive products (see Table 2). Significantly, functional groups
such as halogen, aryl-OMe, and methylenedioxy (Table 2, entries
reagent
B
furnished 6a in 30% yield (TiCl3/Li¼1:1.65, TiCl3/
Mg¼1:0.85). This implies that the McMurry salt component in the
form of the more electropositive Li either as a co-reductant or in its
salt form (LiCl) drives the dealkoxylation–ring cyclization process
post E-/Z-stilbene formation generating phenanthrene 6a. The
proposed hypothesis was corroborated when substituting the re-
ducing metal Mg in reagent B by the more electropositive reductant
Li (reagent A) afforded the phenanthrene 6a as the sole product in
36% yield (Table 4, entry 2).3f Consequently, the salted reagent A, in
comparison to Mg-based LVT reagents, showed more pronounced
effect on the phenanthrene synthesis. Thus, while the ketone 4a
Table 2
Reagent A–KCl (2 equiv) induced facile McMurry olefination of aromatic carbonyls
Entry
Substrate
Temp (ꢂC)/time (h)
Product3a,c,14
Product yields
(%) (E:Z)a,b
1
2
3
4
5
6
7
8
9
1b
1c
1d
1e
1f
1g
1h
1i
25/6.0
25/3.0
25/10.0
25/10.0
25/5.0
25/2.0
25/2.5
25/14.0
25/12.0
3b
3c
3d
3e
3f
3g
3h
3i
68 (72:28)
82 (65:35)
75 (80:20)
78 (75:25)
75
85 (90:10)
65 (92:8)
64 (85:15)
65 (100 E)
Table 3
Low temperature reductive duplication of aliphatic carbonyls with reagent A–KCl
Entry
Substrate
Temp (ꢂC)/time (h)
Product3a,c
Yielda (%)
1
2
3
4
1k
1l
1m
1n
25/2.0
25/2.0
25/6.0
25/18.0
3k
3l
3m
3n
70
68
76
52
1j
3j
Isolated yields of pure products, fully characterized by IR and 1H NMR spectra.
Diastereoisomeric ratios of olefins were determined by 200 MHz NMR and
a
Isolated yields of pure products, fully characterized by IR and 1H NMR (200 MHz)
b
a
comparing with literature data.3a,c
spectra.3a,c

S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
7229
Scheme 3. LVT-mediated one-pot synthesis of phenanthrenes.
stereoselectivity and final product outcome. Notably, the ability of
reagent A–KCl to cleave o-methoxy group in the resultant stilbene
under prolonged refluxing conditions to generate phenanthrenes
6a is in contrast to the reductive coupling of p-methoxy aceto-
phenone using the same reagent (see Table 2, entry 8) where the
methoxy group was retained under milder ambient conditions.
Similarly, the reduction potential of several other LVT formulations
(data not shown) was also screened for different LVT preparations
such as TiCl3–ZnCu–THF, TiCl3–Zn–Li–THF which essentially pro-
duced stilbene 5a as the major product. This was expected con-
sidering the lower reduction potential of Zn (E0þ¼ꢀ0.71 V) in
comparison to that of Li (E0þ¼ꢀ3.047 V) and Mg (E0þ¼ꢀ2.37 V).
In order to examine the influence of solvating medium (ligand
exchange effect) in modulating the reactivity of the LVT prepara-
tions, an unusual solvent/ligand effect was observed. Substituting
the solvating ligand THF in reagent A with slightly stronger elec-
tron-pair donor coordinating solvent DME (reagent C) proceeded
only up to the stilbene stage without any dealkoxylation under
prolonged heating (Table 4, entry 8).3f Surprisingly, the same
reaction when carried out using equimolar amounts of THF as co-
Table 4
Reductive deoxygenation–reductive dealkoxylation–cyclization of o-alkoxy aro-
matic aldehyde–ketone by salt/solvating ligand-activated low-valent titanium
Entry
Reagent
Salt/metal/solvent
Product (yield %)a
E:Zb
Phenanthrene3f
1
2
3
4
5
6
7
8
B
A
B
B
B
A
A
C
C
A
A
A
A
A
Nil
Nil
5a, 85
d
d
6a, 36
6a, 17
6a, 20
6a, 35
6a, 81
6a, 84
LiCl
5a, 65 (80:20)
5a, 60 (68:22)
5a, 65 (55:45)
d
LiClc
Lid
KCl
CsCl
Nil
d
5a, 75
9
THFe
I2
6a, 75
6a, 25
6a, 30
6b, 82
6c, 75
6d, 73
10
11
12
13
14
5a, 35 (60:40)
5a, 33 (58:42)
d
d
d
Naphthalene
CsCl
KCl
KCl
a
All yields refer to isolated products purified by preparative thin layer or column
chromatography (purity >95%, analyzed by 1H NMR).
b
Stereochemical assignments (E:Z) were made by 1H NMR and by comparison of
spectroscopic data to that reported in the literature.3f For 5a3f 1H NMR
d (ppm)
solvent with DME (1:1, v/v) ([Ti ]$(DME)xꢀy(THF)y) furnished the
*
(200 MHz, CDCl3) 1.76 (t, 6H, CH3), 2.36 (q, 4H, CH2), 3.66 and 3.73 (s, 6H, OCH3),
phenanthrene 6a as the sole product in a significantly high yield of
75% (Table 4, entry 9). While the influence of coordinated ligand
exchange effect on the reactivity of the resultant LVT species and
the formation of 6a is inexplicable at this juncture, an observation
reported by Matsubara et al. suggests that addition of THF as a co-
solvent to an inert (TiCl3)n–(amine)m cluster afforded monomeric
[TiCl3(amine)1–2(THF)1–2] particles augmenting its reactivity in
enantioselective pinacol coupling reactions from 40% to 58%.17
Similar reactions using various 1:1 combinations of ethereal and
aprotic solvents like THF/dioxan, THF/CH3CN either resulted in
quantitative recovery of substrate (complete arrest) or afforded
the pinacol as was observed in case of THF/CH2Cl2 (unpublished
6.46–8.1 (m, 8H). For 6a3f 1H NMR
d ppm (200 MHz, CDCl3) 1.32 (t, 6H, CH3), 3.21 (q,
4H, CH2), 7.36–8.6 (m, 8H).
c
Amount of salt used: 8 equiv.
Sub-stoichiometric amounts of Li and Mg were used as co-reductant (TiCl3/
d
Li¼1:1.65, TiCl3/Mg¼1:0.85).
e
1:1 v/v ratio of anhydrous THF and DME.
yielded 6a in 36% yield (Table 4, entry 2) with reagent A alone,3f in
combination with either KCl or CsCl, a 2.5–3.0-fold jump in re-
activity of the LVT reagent was observed furnishing 6a in vastly
improved yields of 81 and 84%, respectively (Table 4, entries 6 and
7) with no trace of 5a or dealkoxylated alkene. Characterization of
6a by proton NMR showed the clear disappearance of the methoxy
signals (3.66 and 3.73 ppm for E-5a:Z-5a) with distinct resonance
signals for CH3 (1.32 ppm, triplet) and CH2 (3.21 ppm, quartet)
along with Ar peaks corresponding to 9,10-diethyl phenanthrene
skeleton (m/z Mþ 234.12).
`
results). The efficacy of the salted reagents vis-a-vis those of
I2-activated LVT and naphthalene-activated LVT reagents developed
earlier from our laboratory was also explored.3a,c As compared to
the present salted LVT formulation (reagent A–KCl), the reactivity of
these two LVT preparations (reagent A–I2, reagent A–naphthalene)
in the synthesis of 6a was less pronounced (25% and 30%) which
also afforded the stilbene 5a in 33% and 38% yield, respectively
(Table 4, entries 10 and 11). Application of this methodology to
From the above results, it is evident that metal salts exert not
only a dramatic influence on modulating the coordination sphere
and the reactivity of the LVT reagents but also on the

7230
S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
other o-alkoxy aromatic ketones and aldehydes (4b–d) as sub-
strates expectedly furnished the corresponding phenanthrenes
6b–d (Table 4, entries 12–14), thereby demonstrating the generality
of the reaction. All the above observations clearly demonstrate the
chemoselective nature of salted/solvated LVT reagents where small
differences in the steric environment (electron density) around the
native titanium can tip the balance in favor of less shielded reaction
site and an energetically favorable product leading to enhanced
selectivities under redox potential control.
carried out using naphthalene-activated LVT reagent developed in
our laboratory afforded the imino-pinacol coupling product 8a,
albeit less efficiently than the salt-activated LVT reagent (Table 5,
entry 4).3a The ratio of dl- and meso-8a isomers was calculated
on the basis of 1H NMR which clearly indicated the presence of
singlet resonance signals at 4.72 and 5.12 ppm corresponding to the
methine (–CH) protons in addition to the aromatic peaks at
6.7–7.3 ppm (20H). Likewise, coupling of the imines 7b, 7c and 7d
in the presence of reagent A–KCl proceeded smoothly producing
the vicinal diamines 8b, 8c and 8d, respectively, in appreciable
yields along with their respective monoamines as byproducts
(Table 5, entries 5, 6, and 7). Thus, while the salted reagent A ac-
celerated the imino-pinacol coupling to vicinal diamines, the high
reactivity of the LVT reagent also resulted in competitive uni-
molecular reductions.
2.3. Imino-pinacol coupling
Vicinal diamines find extensive applications in radiopharma-
ceuticals and as complexing agents and chiral auxiliaries.18 In view
of their frequent occurrence in natural products and medicinal
compounds, we have earlier developed an LVT-mediated synthesis
of them via an imino-pinacol coupling reaction.19 In principle, the
intermediate can undergo dimerization to diamines 8a by bi-
molecular process (path a) or can be quenched by hydrogen from
the medium to give unimolecularly reduced amines 9a (path b). In
this study, the utility of the KCl-activated reagent A for the coupling
of imino substrates was also explored (Scheme 4).
2.4. Mechanistic aspects
A mechanistic hypothesis for the high reactivity exhibited by the
salted LVT formulations described herein is mainly based on the
rational assessment and interpretation of the results obtained by
judicious incorporation of various salts as auxiliaries into the LVT
reagent prepared from TiCl3/1.5 Mg/THF (Tyrlik’s reagent) and
TiCl3/3.3 Li/THF (McMurry’s reagent), during a series of C–C bond
forming SET-induced reactions. Due to the labile nature of Ti spe-
cies, McMurry reagents may exist in structurally different forms
and subtle changes in the reagent preparation of LVT can have
a significant impact on its reactivity, selectivity, and performance.
The ambiguity which surrounds the mechanistic interpretation and
molecular level characterization of the active Ti species, including
its oxidation state, is mainly due to the oxophilic and in situ gen-
erated multiphase and heterogeneous nature of these reagents, and
therefore, still under debate. Conventionally, heterogeneous sus-
pended colloidal slurries of activated Ti are obtained by refluxing
TiCl3/TiCl4 with reductants (Li/Mg/LiALH4/Zn) in an ethereal sol-
vent (THF, DME, dioxan).1,2,9 Studies by Bogdanovic et al. on some
of these systems suggests the formation of bimetallic Ti–metal
bond surrounded by solvent ligands as the active species (I).2,10
Addition of metal salt to this preformed intermetallic LVT suspen-
sion resulted in a slight effervescence indicating the exothermic
nature of the reaction and the possible ligand displacement from
the coordination site of Ti. It is important to note that on reflux
(1 h), the resultant salt-activated LVT system afforded black, uni-
form, and viscous slurries (homogeneous but not transparent),
unlike traditional LVT reagents. While the nature and morphology
of the salted LVT active species present in the intermetallic cocktail
are unknown, the enhanced reactivity of the salted LVT reagent
prepared could be rationalized on the basis of a soluble/homoge-
neous model due to the possible electronic modification and
monomerization of the LVT species.
Scheme 4. Reagent A–KCl-mediated imino-pinacol coupling.
The effect of the reagent A–KCl3b on the imino-pinacol reaction
was noteworthy as coupling of the imine 7a with the salted reagent
was complete in just 20 min at ambient temperature affording the
vicinal diamine 8a (dl:meso¼80:20) in 65% yield. However, com-
petitive unimolecular reduction also produced N-benzyl aniline
(9a) in an appreciable amount (30%) (Table 5, entry 3). A compa-
rable yield (w62–68%) of 8a was obtained with reagent A or re-
agent B alone albeit after a much longer reaction time (w2.5–3.5 h)
(Table 5, entries 1 and 2).19 In comparison, the same reaction when
Owing to the existence of a definite Ti–Mg/Ti–Li bond (as seen
in I), Ti is likely to assume higher electron density (compared to
metallic Ti) when one considers the more electropositive character
of Mg/Li. Exchanging the existing bound MgCl2/LiCl salt component
in the bimetallic Ti–Mg/Ti–Li species by the more electropositive
metal salts with larger ionic size (KCl, CsCl) further augments the
electron density on the active titanium center in Ti–KCl/CsCl re-
agents while exhibiting varying reducing activities. Consequently,
by virtue of their larger ionic size, we anticipate that Kþ/Csþ ions in
comparison to Liþ/Mgþ2 are likely to inhibit the uniform dimeric
intermetallic cluster formation in LVT preparations (as seen in I)
affording the monomeric LVT species (like III) with greater
reactivity and selectivity. Presumably, the formation of highly re-
active, viscous, and more homogeneously soluble LVT slurries
observed could therefore be a direct manifestation of the mono-
merization and electronic modification of the LVT center caused
Table 5
Reagent A–KCl-mediated dimerization of aldimines at 25 ꢂC to generate vicinal
diamines
Entry Reagent
Timea (h) Product(s)19 (% yields, dl:meso)b
1
2
3
4
5
6
7
TiCl3–Mg–THF (B)
3.5
8a (62)
TiCl3–Li–THF (A)
A–KCl
A–naphthalene (0.25 equiv) 0.5
A–KCl
A–KCl
A–KCl
2.5
0.33
8a (68, 75:25)
8a (65%, 80:20), 9a (30%)
8a (45%), 9a (30%)
8b (70%), 9b (18%)
8c (68%, 78:22), 9c (20%)
8d (72%), 9d (15%)
1.0
1.5
2.0
a
All the reactions were carried out at ambient temperature (25 ꢂC).
All yields refer to isolated products (preparative thin layer or column chroma-
b
tography), analyzed by 1H NMR, IR, and MS. Stereochemical assignments (dl:meso)
were made by 1H NMR and by comparison of spectroscopic data to that reported
in the literature.19 For 8a19
:
1H NMR (200 MHz, CDCl3)
d
(ppm) 4.12 (br s, D2O
exchangeable), 4.72 (s, 1H, dl), 5.12 (s, 1H, meso), 6.64–7.31 (m, 20H).

S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
7231
due to the exogenous addition of salts which possibly segregates/
breaks the bimetallic LVT clusters formed (Scheme 5) to generate
a highly reactive electron rich ‘monomeric ate’ complexes (III). The
following explanation is further reinforced by analogous ‘salt effect’
on activation and solubility of Grignard reagents20,21 and chro-
mium(II) species.22 Furthermore, a recent example involving the
reaction of Fe(II) salt (FeCl2) in THF with inorganic Grignard species
I derived from TiCl3–Mg–THF system resulting in the in situ for-
mation of a highly active intermetallic phase of Ti–Fe validates our
proposed theory.1c While the reactivity and stability of transition
metal complexes are mutually exclusive, these aspects vary con-
siderably in various low-valent metal–ligand/auxiliary systems
depending on the ability of the metal center to accept (activation)
or donate (passive) electrons.3a–d,4 In principle, the apparent lack
of vacant d orbitals in alkali/alkaline earth metal ions restricts
the synergistic ability of electron rich low-valent titanium (more
reactive) to back-donate the excess electrons, as was observed in
case of N-heterocycles such as pyridine (Chatt–Dewar–Duncunson
model),23 wherein the pyridine-deactivated LVT complex com-
pletely suppressed the deoxygenation of the pinacolate inter-
mediate.3d,g Simply put, incorporation of electropositive metal salts
while enhancing the electron density/Lewis acid character of the
native Ti center also alters its redox potential thereby augmenting
the reactivity and accelerating the reaction rates of LVT-mediated
stereoselective C–C bond formation in electron transfer processes.
cyclization. To this end, metal salts as promoters perform the
dual role of not only modulating the activity of the LVT reagents
but also effecting the electron transfer interactions between sub-
strate and the active LVT species thereby governing the product
diastereoisomeric ratio under redox potential control.
3. Conclusion
Discovering new reagents for old reactions is crucial to the de-
velopment of sustainable chemical processes as well as for broad-
ening the spectrum of synthetic methodologies. In situ regulation
of the LVT-mediated electron transfer process allows selective
generation of radicals which have a significant influence on the
reactivity/passivity of these reagents facilitating stereoselective
product formation under redox control. Significantly, the McMurry
reaction when conducted using the salt-activated LVT reagents
augmented the reductive dimerization of carbonyls (aromatic/
aliphatic substrates) to olefins at ambient (low) temperatures. In
addition, one-pot synthesis of biologically relevant vicinal diamines
and phenanthrenes in good yields enhanced the scope and utility
of the protocol. The accelerated reaction rates observed in the
various SET transformations unequivocally establish the profound
influence of metal salts as one of the critical parameters for the
‘activation’ of LVT reagents under redox conditions. Mechanisti-
cally, we envisage that the reaction of metal salts with the LVT re-
agents possibly results in the formation of highly reactive and
homogeneous monomeric intermetallic Ti complex. The synergism
between the additives, along with the unique properties exhibited
by LVT species has thus culminated in the strategical design and
development of new titanium formulations possessing varying
degrees of reducing activity. The exceptional performance of some
of the salt-activated LVT reagents while facilitating significant im-
provement in the old McMurry reaction and other C–C bond
forming SET transformations provides new impetus for organo-
metallic preparative chemistry. This work may therefore redefine
the rationale for fine-tuning the reactivity of LVT reagents (and
organometallic reagents, in general) to generate soluble reductant
systems with well-defined oxidation states, thereby enhancing the
scope of the SET reactions leading to the discovery of new and
selective synthetic transformations.
Scheme 5. Monomerization of clustered bimetallic LVT species (inorganic Grignard)
on external addition of metal salts.
This clearly suggests that important structural differences likely
exist between the mechanism of action of salt-activated uniformly
homogeneous LVT complexes and conventional heterogeneous LVT
suspensions. The smooth accelerated stereoselective dimerization
of carbonyls/imines under facile conditions to the respective ole-
fins/vicinal diamines (3a–n/8a–d) even at ambient temperatures or
formation of phenanthrenes (6a–d) in high yield could only pos-
sibly be explained by considering activation of the LVT species
through monomerization and ligation with the metal salts as pro-
posed above. This argument is also substantiated by the fact that
the conventional McMurry reagents (reagents A and B) produce
only the 1,2-diols at room temperature. Moreover, the diaster-
eoselectivity bias toward a particular product (E/Z or dl/meso) in
titanium-mediated SET reactions is greatly influenced by modifi-
cation of steric and electronic properties of ancillary ligands/
auxiliaries. Specifically, while reagent A predominantly produced
E-stilbene (electropositive alkali metal reductant, Li), a complete
reversal in diastereoselectivity (Z-stilbene as major product) was
observed when reagent B (bivalent alkaline earth metal reductant,
Mg) was employed (Tables 1 and 2). This pronounced diaster-
eoselectivity bias observed in case of reductive deoxygenation of
aromatic ketones/aldehydes could be a direct result of combined
ligand exchange effect of auxiliary (salt), reductant, and solvent
coordinated to the metal center. Additionally, the high phenan-
threne yield obtained from these salted reagents is directly related
to the stereochemical outcome of the most favored (sterically and
energetically) intermediate Z-stilbene (over E-stilbene) thereby
promoting the process of demethoxylation and subsequent ring
4. Experimental section
4.1. General methods
General information regarding instruments, techniques, and
source of chemicals used is the same as mentioned in our previous
publications.3,5a,15a,b Lithium rods cut into small pieces were used
for the reduction of titanium chlorides. All the products 3a–n, 6a–d,
8a–d, 2a, 5a, 9a–d are known compounds and all yields refer to
isolated products purified by preparative thin layer or column
chromatography (purity >95%) analyzed by mp, 1H NMR, IR in
comparison with known literature data (for product details see
Supplementary data).
4.2. McMurry olefination via reductive deoxygenation
4.2.1. General procedure for metal salt-activated LVT-induced
reductive deoxygenation of aromatic and aliphatic carbonyls (1a–n)
to corresponding olefins (3a–n)
To a mixture of anhydrous TiCl3 (10 mmol, 1.55 g) in dry THF or
DME (50 ml) in a three-necked flask, reducing metal (Li: 33 mmol,
0.231 g/Mg: 17 mmol, 0.408 g) was added and refluxed (3 h) under
Ar. With time, the violet color of the reaction suspension changed
and gradually got converted to a black colloidal slurry. (Note: in
case of Mg as reducing agent, a brownish black homogeneous slurry

7232
S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
of the LVT reagent was obtained. With Li, black, finely particulated
heterogeneous slurry of LVT was produced along with a few sus-
pended pieces of unreacted Li on the surface of the slurry.) After
attaining room temperature, addition of annhydrous metal halide
(2.0–8.0 equiv of TiCl3) to this preformed LVT reagent resulted
in slow effervescence indicating the exothermic nature of the re-
action. The resultant formulation was then refluxed for an addi-
tional 1 h and then cooled to room temperature affording viscous,
non-transparent homogeneous slurry. (With anhydrous MgCl2,
LiCl, CsCl, KCl, black slurries were obtained, while annhydrous LiI
and ZnCl2 provide brownish black slurry. Use of anhydrous LiCl,
MgCl2, ZnCl2 resulted in brisk effervescence in comparison to CsCl
and KCl salts.) Carbonyl substrates 1a–n (2.5 mmol) in dry THF or
DME (5 ml) were added to activated LVT reagent with continued
stirring at room temperature. The reaction was monitored at reg-
ular intervals (TLC) and on completion; the reaction was quenched
with water and diluted with hexane. The mixture was thoroughly
extracted with hexane/ethylacetate (70:30) mixture and the extract
passed through Celite. After repeated washings (five times), the
organic portion was pooled together, washed with water and brine,
and dried (Na2SO4). Removal of the solvent under reduced pressure
yielded the crude product which was subjected to preparative TLC
(SiO2 gel, 5% EtOAc/hexane) furnishing the olefins 3a–n (mixture of
cis and trans isomers) along with vicinal diols 2a (pinacols) wher-
ever reported. All the product stilbenes 3a–n and pinacol 2a
reported are known compounds for which references have been
cited in the text and tables.
activated slurry. An appropriate amount of aldimine 7a–d was
added (2.5 mmol, 5 ml THF) and was stirred at room temperature
till all the starting compound disappeared (monitored by TLC).
After completion, the reaction mixture was quenched with satu-
rated NH4Cl solution and diluted with hexane/ethylacetate mixture
(60:40) and passed through a Celite bed. The collective organic
fractions so obtained were washed with water, brine and dried
(Na2SO4). Concentration of the dried organic solvent afforded the
crude reaction product which was further purified using pre-
parative TLC (SiO2 gel, 5% EtOAcþhexane) which furnished the
respective vicinal diamines 8a–d along with the mono-reduced
amines (9a–d). The products are known compounds and were
characterized by IR, NMR, MS, physical constants, and also by
comparison with authentic samples.
Acknowledgements
S.M.R. is grateful to the Department of Atomic Energy, Govern-
ment of India, for a Senior Research Fellowship (1998–2000).
Supplementary data
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.tet.2008.05.084.
References and notes
1. For excellent reviews on the synthetic applications of low-valent titanium re-
agents, see: (a) McMurry, J. E. Chem. Rev. 1989, 89, 1513; (b) Lenoir, D. Synthesis
1989, 883; (c) Furstner, A.; Bogdanovic, B. Angew. Chem., Int. Ed. 1996, 35, 2442;
(d) Dushin, R. G. Comprehensive Organometallic Chemistry II; Hegedus, L. S., Ed.;
Pergamon: Oxford, 1995; Vol. 12, p 1071; (e) Ladipo, F. Curr. Org. Chem. 2006, 10,
965; (f) Ephritikhine, M.; Villiers, C. Modern Carbonyl Olefination; Takeda, T., Ed.;
Wiley-VCH: Weinheim, 2004; p 223.
2. For different recipes for the generation of activated LVT reagents, see: (a) Cintas,
P. Activated Metals in Organic Synthesis; CRC: Boca Raton, 1996; (b) Lectka, T.
Active Metal – Preparation, Characterization, Applications; Fu¨rstner, A., Ed.; VCH:
Weinheim, 1996; pp 85–131.
3. (a) Rele, S.; Talukdar, S.; Banerji, A.; Chattopadhyay, S. J. Org. Chem. 2001, 66,
2990; (b) Rele, S.; Chattopadhyay, S.; Nayak, S. K. Tetrahedron Lett. 2001, 42,
9093; (c) Talukdar, S.; Nayak, S. K.; Banerji, A. J. Org. Chem. 1998, 63, 4925; (d)
Balu, N.; Nayak, S. K.; Banerji, A. J. Am. Chem. Soc. 1996, 118, 5932; (e) Nayak, S.
K.; Banerji, A. J. Org. Chem. 1991, 56, 1940; (f) Banerji, A.; Nayak, S. K. J. Chem. Soc.,
Chem. Commun. 1991, 1432; (g) Talukdar, S.; Nayak, S. K.; Banerji, A. Fullurene
Sci. Technol. 1995, 3, 327; (h) Nayak, S. K.; Kadam, S.; Talukdar, S.; Banerji, A.
J. Indian Inst. Sci. 1994, 74, 401; (i) Kadam, S. M.; Nayak, S. K.; Banerji, A.
Synth. Commun. 1995, 25, 135; (j) Mayekar, N. V.; Chattopadhyay, S.; Nayak, S. K.
Synthesis 2003, 2041; (k) Mayekar, N. V.; Chattopadhyay, S.; Nayak, S. K. Lett. Org.
Chem. 2004, 1, 203.
4.3. Phenanthrene synthesis via reductive deoxygenation and
reductive dealkoxylation
4.3.1. Representative procedure for one-pot synthesis of 9,10-
dialkylphenanthrenes (6a–d) by salt-activated LVT reagent
Titanium trichloride (10 mmol, 1.55 g) was added to a dry flask
containing Li pieces (33 mmol, 0.231 g)/Mg (17 mmol, 0.408 g) in
dry THF (50 ml). The mixture was refluxed for 3 h, cooled to room
temperature, and anhydrous metal halides like LiCl, KCl, CsCl (2–8
equiv) were added. The mixture was then refluxed for an additional
1 h, o-alkoxy aromatic aldehydes/ketones 4a–d (2.5 mmol, 410 mg)
in 5 ml THF (and/or DME) were added to the activated LVT species
and refluxed for additional 16 h. After completion (TLC), it was
allowed to attain room temperature, diluted with hexane/ethyl-
acetate (70:30) mixture, quenched with saturated solution of
NH4Cl, and passed through Celite. The collective organic eluent was
washed with water and brine, and dried. Concentration of the or-
ganic extract afforded the crude product which was subsequently
purified by preparative TLC (SiO2 gel, 2.5% EtOAc/hexane) furnish-
ing the phenanthrene derivatives 6a–d and/or the olefins 5a–d.
Comparison of the spectral data of the reaction products with those
of authentic samples confirmed the presence of these compounds.
All the product phenanthrenes 6a–d3f and stilbene 5a3f reported
are known compounds for which references have been cited in the
text in the Supplementary data.
4. (a) Lipski, T. A.; Hilfiker, M. A.; Nelson, S. G. J. Org. Chem. 1997, 62, 4566; (b)
Mukaiyama, T.; Kagayama, A.; Shiina, I. Chem. Lett. 1998, 1107; (c) Ganauser, A.;
Pierobon, M.; Bluhm, H. Angew. Chem., Int. Ed. 1998, 37, 101.
5. (a) Nayak, S. K.; Banerji, A. J. Chem. Soc., Chem. Commun. 1990, 150; (b) Furstner,
A.; Jumbam, D. N. Tetrahedron 1992, 48, 5991; (c) Furstner, A.; Ernst, A.; Krause,
H.; Ptock, A. Tetrahedron 1996, 52, 7328; (d) Furstner, A.; Hupperts, A. J. Am.
Chem. Soc. 1995, 117, 4468; (e) Furstner, A.; Hupperts, A.; Ptock, A.; Janssen, E.
J. Org. Chem. 1994, 59, 5215.
6. (a) Furstner, A.; Seidel, G. Synthesis 1995, 63; (b) Furstner, A.; Seidel, G.;
Kopiske, C.; Kruger, C.; Mynott, R. Liebigs Ann. 1996, 655.
7. (a) Nicolaou, K. C.; Yang, Z.; Liu, J. J.; Ueno, H.; Nantermet, P. G.; Guy, R. K.;
Claiborne, C. F.; Renaud, J.; Couladouros, E. A.; Paulvannnan, K.; Sorensen, E. J.
Nature 1994, 367, 630; (b) Nicolaou, K. C.; Liu, J. J.; Yang, Z.; Ueno, H.; Sorensen,
E. J.; Claiborne, C. F.; Guy, R. K.; Hwang, C. K.; Nakada, M.; Nantermet, P. G. J. Am.
Chem. Soc. 1995, 117, 634; (c) Nicolaou, K. C.; Yang, Z.; Liu, J. J.; Nantermet, P. G.;
Claiborne, C. F.; Renaud, J.; Guy, R. K.; Shibayama, K. J. Am. Chem. Soc. 1995, 117,
645.
8. (a) Wang, L.; Zhang, Y. Synth. Commun. 1998, 28, 3991; (b) Nishiyama, Y.;
Shinomiya, E.; Kimura, S.; Itoh, K.; Sonoda, N. Tetrahedron Lett. 1998, 39, 3705;
(c) Wang, L.; Zhang, Y. Tetrahedron 1998, 37, 11129; (d) Ascham, F. R.; Carroll,
K. M. J. Org. Chem. 1993, 58, 7328; (e) Hays, D. S.; Fu, G. C. J. Am. Chem. Soc. 1995,
117, 3867; (f) Akane, N.; Hayano, T. H.; Kusui, H.; Nishiyama, Y.; Ishii, Y. J. Org.
Chem. 1994, 59, 7904; (g) Szymoniak, J.; Besancon, J.; Moise, C. Tetrahedron
1992, 48, 3867; (h) Szymoniak, J.; Besancon, J.; Moise, C. Tetrahedron 1994, 50,
2841; (i) Immamoto, T.; Kusumoto, T.; Hatanaka, Y.; Yokoyama, M. Tetrahedron
Lett. 1982, 23, 1353; (j) Wang, C.; Pana, Y.; Wu, A. Tetrahedron 2007, 63, 429; (k)
Pederson, S. F. J. Am. Chem. Soc. 1989, 111, 8014; (l) Xu, X. L.; Hirao, T. J. Org.
Chem. 2005, 70, 8594; (m) Hirao, T. Synlett 1999, 175; (n) Ueda, T.; Kanomata,
N.; Machida, H. Org. Lett. 2005, 7, 2365; (o) See Ref. 17.
4.4. Imino-pinacol coupling reaction
4.4.1. Typical procedure for reductive dimerization of aldimines to
vicinal diamines using reagent A–KCl
A mixture of TiCl3 (6.25 mmol, 964 mg) and freshly cut Li pieces
(20.6 mmol, 144 mg), in dry THF was refluxed (3 h) in an inert
atmosphere of Ar. The reduced LVT reagent so prepared was then
allowed to cool to room temperature. Addition of KCl (2 equiv of
TiCl3) to this active mixture resulted in very slow effervescence
indicating the slightly exothermic nature of the reaction. The
resulting slurry was then further refluxed for an additional 1 h
which on cooling to the room temperature afforded a thick black

S.M. Rele et al. / Tetrahedron 64 (2008) 7225–7233
7233
9. Dams, R.; Malinowski, M.; Westdorp, I.; Geise, Y. H. J. Org. Chem. 1982, 47, 248.
10. (a) Aleandri, L. E.; Bogdanovic, B.; Gaidies, A.; Jones, D. J.; Liao, S.; Michalowicz,
A.; Roziere, J.; Schott, A. J. Organomet. Chem. 1993, 459, 87; Related references:
(b) Bogdanovic, B.; Bolte, A. J. Organomet. Chem. 1995, 502, 109; (c) Aleandri, L.;
Becke, S.; Bogdanovic, B.; Jones, D. J.; Roziere, J. J. Organomet. Chem. 1994,
472, 97.
11. The concept of salt effect on reaction rates is explained in Sykes, P. The Search for
Organic Reaction Pathways; Longman: London, 1972; p 31. Addition of neutral
salts is known to influence the resultant ionic strength of the medium (positive
salt effect) significantly modulating the rates of many reactions in which the
charge is concentrated or dispersed. Also see Refs. 20 and 21.
12. (a) McMurry, J. E.; Fleming, M. P.; Kees, K. L.; Krepski, L. R. J. Org. Chem. 1978, 43,
3255; (b) McMurry, J. E.; Rico, J. G. Tetrahedron Lett. 1989, 30, 1169; (c) McMurry,
J. E.; Dushin, R. G. J. Am. Chem. Soc. 1990, 112, 6942.
13. Tyrlik, S.; Wolochowicz, I. Bull. Soc. Chim. Fr. 1973, 2147.
14. Isomeric ratio of 3a is established using spectroscopic data, see: Andersson,
P. G. Tetrahedron Lett. 1994, 35, 2609.
15. (a) Talukdar, S.; Banerji, A. Synth. Commun. 1996, 26, 1051; (b) Talukdar, S.;
Banerji, A. Synth. Commun. 1995, 25, 813; (c) Tyrlik, S.; Wolochowicz, I. J. Chem.
Soc., Chem. Commun. 1975, 781.
16. McMurry, J. E.; Silvestri, M. G.; Fleming, M. P.; Hoz, T.; Grayston, M. W. J. Org.
Chem. 1978, 43, 3249.
like KI, prior to the reduction of anhydrous MgCl2 with K in refluxing THF,
afforded a reactive black activated Mg (Mg $KCl).
) powder (MgCl2þKIþK/Mg
Scanning electron microscopy (SEM) has shown this Mg to be a sponge-like
material with much smaller particle size. The apparent association of KCl and
Mg metal and its subsequent reaction with the organic halide (RX) was
*
*
*
reported to afford an ‘ate’ complex (R–XþMg
*
$KCl/[R–Mg–(Cl)X]Kþ). The
reactivity of the Mg so produced was assessed by addition of a number of metal
salts (LiCl, LiF, NaCl, NaI, CsI, KCl, KI, MgSO4, CuSO4, NaBr, ZnBr2, etc.). Salts like
NaI, KI, ZnBr2, etc., were found to exhibit an activating effect on the reactivity of
Mg prepared from MgCl2–K–THF. In contrast, other salts (KF, CuSO4, etc) had
a deactivating effect, while the use of NaF, NaCl, LiF, LiCl, KCl essentially gave
the same results as the control reaction.
21. Transition metal Grignards with generalized formula [M1(MgCl)m] are known
to react with metal halides M2X in a molar ratio of n:m in coordinating solvents
generating highly reactive alloys and intermetallics of composition M1nM2m. In
a similar fashion the reaction of [M1(MgCl)m] with its corresponding metal
chloride (M1Cl) can be employed as a method for preparation of finely divided
metal powders (mþn)M1. We anticipate a similar mechanism taking place
on exogenous administration of metal salts to the preformed LVT reagents
(reagents A and B) which breaks the uniform and ordered dimeric intermetallic
cluster resulting in the formation of a highly reactive intermetallic monomeric
‘ate’ donor complex.
17. Hashimoto, Y.; Mizuno, U.; Matsuoka, H.; Miyahara, T.; Takakura, M.; Yoshi-
moto, M.; Oshima, K.; Utimoto, K.; Matsubara, S. J. Am. Chem. Soc. 2001, 123,
1503.
18. Lucet, D.; Gall, T. L.; Mioskowski, C. Angew. Chem., Int. Ed. 1998, 37, 2580.
19. Talukdar, S.; Banerji, A. J. Org. Chem. 1998, 63, 3468 and references therein.
20. Rieke, R. D.; Bales, S. F. J. Am. Chem. Soc. 1974, 96, 1775. Many organic halides
due to their low propensity for reaction with ordinary Mg turnings use
22. See: (a) Wessjohann, L.; Gunter, S. Synthesis 1999, 1; (b) Wessjohann, L.;
Gabriel, T. J. J. Org. Chem. 1997, 62, 3772. The overall reactivity of such organo-
metallic preparations, in general, is dependent on their solubility in the
reaction medium. For example, CrCl2 is poorly soluble in THF and tends to
associate in this medium and shows poor reactivity which can be augmented
by the addition of LiI. The increased reactivity has been attributed to the
enhanced solubility and possibly electronic modification and monomerization
of CrCl2.
activated [Mg ] during formation of Grignard reagents. Rieke and co-workers in
*
their pioneering work have suggested that alkali salts play an important role in
inducing high reactivity of Mg. The addition of simple alkali salts or Lewis bases
23. (a) Chatt, J.; Duncanson, L. A. J. Chem. Soc. 1953, 2939; (b) Dewar, M. J. S. Bull.
Soc. Chim. Fr. 1951, C71.