Key role of Ti(IV) in the selective radical - radical cross-coupling mediated by the ingold-Fischer effect

Article abstract of DOI:10.1021/ja807613q

We report an innovative approach for the selective synthesis of polyfunctional derivatives by cross-combination of different radicals generated under mild conditions. The coordinating effect of Ti(IV) plays a key role in the reaction mechanism: due to its chelating action on the hydroxyl groups, it promotes the homolytic C-C bond cleavage of α,β-dihydroxy ketones by enhancing the captodative effect and the consequent stabilization of the corresponding α-hydroxy-α-carbonyl radicals. When these radicals are generated in the presence of stoichiometric amounts of TiCl₄ and 2,2'-azo-bis-isobutyronitrile (AIBN) is employed as a source of α-cyanoisopropyl radicals, the selective radical-radical cross-coupling is observed, affording the corresponding β-hydroxynitriles in high yields. This innovative methodology allows application of the well-known Ingold-Fischer effect to a wider range of stabilized carbon-centered radicals, whose formation derives from the chelating action of Ti(IV).

Full text of DOI:10.1021/ja807613q

Published on Web 12/02/2008
Key Role of Ti(IV) in the Selective Radical-Radical
Cross-Coupling Mediated by the Ingold-Fischer Effect
Raffaele Spaccini, Nadia Pastori, Angelo Clerici, Carlo Punta,* and
Ombretta Porta*,†
Dipartimento di Chimica, Materiali e Ingegneria Chimica “G. Natta”, Politecnico di Milano,
Via Mancinelli 7, I-20131 Milano, Italy
Received September 30, 2008; E-mail: carlo.punta@polimi.it
Abstract: We report an innovative approach for the selective synthesis of polyfunctional derivatives by
cross-combination of different radicals generated under mild conditions. The coordinating effect of Ti(IV)
plays a key role in the reaction mechanism: due to its chelating action on the hydroxyl groups, it promotes
the homolytic C-C bond cleavage of R,ꢀ-dihydroxy ketones by enhancing the captodative effect and the
consequent stabilization of the corresponding R-hydroxy-R-carbonyl radicals. When these radicals are
generated in the presence of stoichiometric amounts of TiCl4 and 2,2′-azo-bis-isobutyronitrile (AIBN) is
employed as a source of R-cyanoisopropyl radicals, the selective radical-radical cross-coupling is observed,
affording the corresponding ꢀ-hydroxynitriles in high yields. This innovative methodology allows application
of the well-known Ingold-Fischer effect to a wider range of stabilized carbon-centered radicals, whose
formation derives from the chelating action of Ti(IV).
Introduction
The PRE allowed development of a large number of selective
reactions involving radicals both highly stabilized7,8 and bearing
crowded substituents.9
For a long time free radical chemistry was thought to be
useless for selective synthesis of complex molecules in high
yields, because of the erroneous concept that the enthalpic effect
was the main driving force in free radical processes. Thus, high
reactivity had to be always associated to low selectivity. More
recently the synthetic potentiality of radical reactions greatly
increased and a wide range of selective free-radical routes has
been developed by considering also the key role of polar and
steric effects.1-3
A few years ago Porta and co-workers10 reported the
reduction of R,ꢀ-dicarbonyl compounds by TiCl3, in the presence
of an aldehyde, heading the steroselective synthesis of R,ꢀ-
dihydroxy ketones 1 (eq 1).
An intriguing example of the potential of free radical
chemistry is shown by the use of the Ingold-Fischer “Persistent
Radical Effect” (PRE)4,5 as a synthetic tool for the preparation
of clean products in high yields. The PRE principle permits us
to explain the highly specific C-C bond formation by a cross-
coupling between radicals when two species are generated at
similar rates and one is more persistent than the other, that is
with a lifetime significantly greater than that for a methyl radical
under the same conditions.6
More recently, the same research group reported the results
of a study on the reactivity of methyl mandelate 2 in the
presence of a Ti(IV)/base system: the oxidative dimerization
led to 3 in good yields11 (Scheme 1, path a), whereas when the
same reaction was conducted in the presence of an aldehyde,
the aldol addition occurred, leading to 4 in high yields11 (Scheme
1, path b).
Homolytic cleavage of the C-C bond makes compounds
1, 3, and 4 potential good sources of radical species driven
by the captodatiVe effect,12 that is the effect on the stability
of a carbon-centered radical determined by the combined
† Deceased May 3, 2008.
(1) (a) Barton, D. H.; Beaton, J. M.; Geller, L. E.; Pechet, M. M. J. Am.
Chem. Soc. 1960, 82, 2640–2641. (b) Viehe, D. H. G.; Janousek, Z.;
Mere´nyi, R. Substituent Effects in Radical Chemistry; NATO ASI
series C 189; Dordrecht, 1986. (c) Minisci, F. Free Radicals in
Synthesis and Biology; NATO ASI series C 260; Dordrecht, 1988.
(d) Motherwell, W. B.; Crich, D. Free Radical Chain Reactions in
Organic Synthesis; Academic Press: London, 1991. (e) Togo, H.
AdVanced Free Radical Reactions for Organic Synthesis; Elsevier:
2004.
(2) Renaud, P.; Sibi, M. P. Radicals in Organic Synthesis; Wiley VCH:
2001.
(3) Mayr, H.; Ofial, A. R. Angew. Chem., Int. Ed. 2006, 45, 1844–1854.
(4) Fischer, H. J. Am. Chem. Soc. 1986, 108, 3925–3927.
(5) Daikh, B. E.; Finke, R. G. J. Am. Chem. Soc. 1992, 114, 2938–2943.
(6) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1976, 9, 13–19.
(7) Bravo, A.; Bjørsvik, H.-R.; Fontana, F.; Liguori, L.; Minisci, F. J.
Org. Chem. 1997, 62, 3849–3857.
(8) Recent reviews on the subject: (a) Studer, A. Chem.sEur. J. 2001, 7,
1159–1164. (b) Fischer, H. Chem. ReV. 2001, 101, 3581–3610. (c)
Studer, A. Chem. Soc. ReV. 2004, 33, 267–273.
(9) (a) Frenette, M.; Aliaga, C.; Font-Sanchis, E.; Scaiano, J. C. Org. Lett.
2004, 6, 2579–2582. (b) Focsaneanu, K.-S.; Aliaga, C.; Scaiano, J. C.
Org. Lett. 2005, 22, 4979–4982.
(10) Clerici, A.; Porta, O. J. Org. Chem. 1989, 16, 3872–3878.
(11) Clerici, A.; Pastori, N.; Porta, O. J. Org. Chem. 2005, 70, 4174–4176.
9
18018 J. AM. CHEM. SOC. 2008, 130, 18018–18024
10.1021/ja807613q CCC: $40.75
2008 American Chemical Society

Ti(IV)-Mediated Ingold-Fischer Effect
A R T I C L E S
Scheme 1. Reactivity of Methyl Mandelate-Ti(IV)-enediolate:
Oxidative Homocoupling (Path a) versus Aldol
syn-Diasteroselective Condensation (Path b)
Scheme 3. Homolytic Cleavage of the C-C Bond in Dimethyl
2,3-Dihydroxy-2,3-diphenyl Butanedioate 3, Catalyzed by Ti(IV)
Scheme 2. Captodative Effect in R-Hydroxy-R-carbonyl Radicals
Scheme 4. Stereochemical Equilibration of Isomer 3-meso in the
Presence of 2 equiv of TiCl4
action of a captor (electron-withdrawing; in this case the
carbonyl group) and a dative (electron-releasing; in this case
the hydroxyl group) substituent, both attached to the radical
center (Scheme 2).
Here we report the interesting results obtained by employ-
ing derivatives 1 and 3 as sources of stabilized radicals in
the presence of Ti(IV). The coordinating effect of the metal
ion13 induces an increased stabilization onto the new radicals,
due to an enhanced captodative effect. This observation
prompted us to develop a new one-pot multicomponent route
for the synthesis of polyfunctional derivatives of high added
value.
the transition metal (radical B, Scheme 4). The new O-Ti(IV)
bond is significantly more polar than the corresponding O-H
one, and this induces an enhanced captodative effect on the
carbon-centered radical, due to the stabilization of the positive
charge on the oxygen atom, according to the resonance structures
reported in Scheme 2.
In addition, the possible coplanarity of the carbomethoxy
group with the radical center due to Ti(IV) chelation could
facilitate the electron delocalization in the radical.
To verify the stabilizing role of Ti(IV), 2 equiv of TiCl4
were added to a solution of the stereoisomer 3-meso in
benzene and the mixture was heated at 80 °C for 1.5 h. In
this case no disproportion was observed, which is consistent
with the formation of the O-Ti(IV) bond, but the C-C bond
cleavage occurred with stereochemical equilibration of the
isomer, leading to a mixture of meso and dl in a 2.4:1 ratio
(Scheme 4).
Results and Discussion
The effect of Ti(IV) on dimethyl 2,3-dihydroxy-2,3-diphenyl
butanedioate 3 in benzene was first investigated. In the absence
of the transition metal, no cleavage was observed at 80 °C, and
compound 3 was recovered unreacted after 3 h. Nevertheless,
when a catalytic amount (25%) of TiCl4 (1 M solution in
CH2Cl2) was added to the reaction mixture, the homolysis of 3
occurred and the starting derivative was quantitatively converted
to methyl phenylglyoxylate 5 and methyl mandelate 6 within a
few minutes (Scheme 3).
The formation of products 5 and 6 arised from the dispropor-
tion of radicals A, generated in situ by the chelating action of
Ti(IV) on the dimer 3 (Scheme 3).11
The low enthalpy of dissociation of the dimer in the presence
of Ti(IV) appears to result from structural features in the radical,
which stabilize the radical center and lower the energy. The
carboxy group is not a particularly strong radical stabilizing
function in itself,14 and notwithstanding the presence of a
hydroxyl group enhancing the stabilization (Scheme 2), it is
not sufficient to undergo homolytic cleavage at 80 °C. The
higher stability of the radical in the presence of Ti(IV) derives
from the complexation of the hydroxyl group, which leads to
the formation of a covalent bond between the oxygen atom and
The isomerization, which derives from the equilibrium
between dimer 3 and the corresponding chelated radicals,
clearly shows the homolytic nature of the C-C bond
cleavage. Furthermore, this result confirms the stabilizing
effect of Ti(IV), the most stable radical being the least likely
to disproportionate.
This equilibrium warrants a stationary concentration of
radicals B with a relatively high lifetime in solution. Thus,
the higher stabilization indirectly confers an enhanced
persistent character to the new Ti(IV)-complexed radical B
when compared with radical A, as the last one escapes the
equilibrium undergoing fast disproportion (Scheme 3).
This suggested that radical B could behave as a suitable
scavenger of more transient radicals generated in situ, leading
to the formation of C-C bonds according to the Ingold-Fisher
effect and shifting the equilibrium of Scheme 4 toward the
formation of new B radicals.
(12) (a) Stella, L.; Janousek, Z.; Mere´nyi, R.; Viehe, H. G. Angew. Chem.,
Int. Ed. Engl. 1978, 17, 691–692. (b) Viehe, H. G.; Mere´nyi, R.; Stella,
L.; Janousek, Z. Angew. Chem., Int. Ed. Engl. 1979, 18, 917–932. (c)
Mere´nyi, R.; Janousek, Z.; Viehe, H. G. in ref 1b, p 301.
(13) Moreira, I. P. R.; Bofill, J. M.; Anglada, J. M.; Solsona, J. G.; Nebot,
J.; Romea, P.; Urpi, F. J. Am. Chem. Soc. 2008, 130, 3242–3243.
(14) Russell, G. A. Free Radicals, Vol. 1; Kochi, J. K., Ed.; Wiley: New
York, 1973; p 275.
9
J. AM. CHEM. SOC. VOL. 130, NO. 52, 2008 18019

A R T I C L E S
Spaccini et al.
Table 2. Reactivity of 5, 6 and 3 in the Presence of IBNa
Scheme 5. Thermal Decomposition of AIBN
IBN
(equiv)
product 3
(selectivity %)
product 9
(selectivity %)
other products
(selectivity %)
entry substrate
1
2
3
4
5
6
5+6
6
5+6
3
0
0
1
1
2
-
-
-
-
-
-
-
-
-
>98% (6)
>98% (5 + 6)
>98% (6)
>98% (5 + 6)
80% (5 + 6)
17 (meso + dl)
a All reactions were performed under N2 atmosphere for 1.5 h in the
presence of stoichiometric amounts of TiCl4.
Table 1. Effect of Ti(IV) on the Selectivity in the Formation of
Product 9a
was obtained, according to eq 2, when different amounts of
TiCl4 were added.
IBNc
(selectivity %)b
9
7 + 8
(selectivity %)b
(selectivity %)b
[Ti(IV)]/[3]
0
0
25
38
77
76
100
57
50
20
22
0
18
12
3
0.25
0.50
0.75
1.00
2
a All reactions were performed under N2 atmosphere for 1.5 h; see
Experimental Section and Supporting Information for details. b With the
term selectiVity we mean here the percentage of R-cyanoisopropyl
radicals which lead to the formation of the different products. It has
been determined by 1H NMR with 1-biphenyl-4-yl-ethanone as an
A hypothetical ionic mechanism based on the in situ
formation of IBN, followed by a classical retro-aldol reaction,
has to be excluded on the basis of the results reported in Table
2. Under the reaction conditions of Table 1, in the absence of
AIBN neither 6 nor the mixture 5 + 6 afforded dimer 3 (entries
1 and 2, Table 2). These results are consistent with what was
previously reported by our group by working at room temper-
ature.11 Thus, the isomerization of 3 meso (Scheme 4) may only
be ascribed to a radical homocoupling. Furthermore, any attempt
to obtain product 9 in the presence of IBN from 6 and 5 + 6
failed, and the starting materials were collected unreacted at
the end of the reaction (entries 2 and 3, Table 2). Finally, when
a solution containing dimer 3, TiCl4 (1 equiv), and IBN in
benzene was refluxed for 1.5 h, no product 9 was observed,
while 3 underwent disproportion as observed in the absence of
AIBN, this excluding the ionic hypothesis.
internal standard added to the crude reaction mixture. c IBN
isobutyronitrile.
)
2,2′-Azo-bis-isobutyronitrile (AIBN) resulted in being an ideal
source of radicals from both a mechanistic and a synthetic point
of view, according to Scheme 5.
Because of the importance of azonitriles in initiating radical
reactions, R-cyanoalkyl radicals have received a great deal of
study,15 including the investigation of the cage effect in thermal
and photochemical decomposition.16 In addition, the easy
introduction of cyano substituents, precursors of carboxylic
groups on a wide range of organic derivatives, represents a
synthetic tool for the preparation of polyfunctional deriva-
tives.17-20
AIBN (1.2 mmol) and dimer 3 meso (0.6 mmol) were dissolved
in 10 mL of benzene in the presence of different amounts of TiCl4
(1 M solution in CH2Cl2) under nitrogen atmosphere. The mixture
was heated at 80 °C, and the reaction was allowed to take place
for 1.5 h, which equals a one-half life period for the decomposition
of AIBN at 80 °C. Thus, after 1.5 h the amount of R-cyanoisopropyl
radicals, obtained by decomposition of AIBN, was equimolar with
the quantity of radicals A deriving from dimer 3. The results are
reported in Table 1.
Figure 1 shows the plot of the ratio [Ti(IV)]/[3] versus the
selectivity of product 9, that is the percentage of R-cyanoiso-
propyl radicals indeed trapped by radicals A or B.
The selectivity in 9 linearly increased by increasing the
amount of Ti(IV): for example, when 25% of Ti(IV) was
employed with respect to 3, only 25% of the desired product
was recovered. The remaining starting dimer 3 underwent
quantitative disproportion, affording products 5 and 6. R-Cy-
anoisopropyl radicals in excess afforded 7 and 8 (in minor
yields) by homocoupling (Scheme 5) and isobutyronitrile (IBN)
by disproportion with radical A. When dimer 3 was reacted
with a double excess of AIBN, in the presence of stoichiometric
amounts of TiCl4, product 9 was isolated with 84% yield.
These results clearly demonstrate that the scavenging effect
is due to radical B (eq 3) deriving from the chelating action of
Ti(IV), whereas radical A is not able to capture any radical (eq
4). In fact, the high selectivity for the cross-combination
observed in the presence of stoichiometric amounts of Ti(IV)
cannot be ascribed to radical A, because the value of the kinetic
constant for the termination reaction by homocoupling of two
A radicals is very high (kt ∼5 × 108 M-1 s-1),21 quite similar
to that of two R-cyanoisopropyl radicals (kt ) 9.4 × 108 M-1
s-1, Scheme 5).22 These high values for the dimerization kinetic
constants clearly show that A and the R-cyanoisopropyl radical
are both transient radicals and cannot undergo cross-coupling
In the absence of TiCl4, no C-C bond cleavage was
observed for dimer 3, whereas the cross-coupling product 9
(15) (a) Hammond, G. S.; Sen, J. N.; Boozer, C. E. J. Am. Chem. Soc.
1955, 77, 3244–3248. (b) Hammond, G. S.; Lucas, G. B. J. Am. Chem.
Soc. 1955, 77, 3249–3251. (c) Van Hooh, J. P.; Tobolsky, A. V. J. Am.
Chem. Soc. 1958, 80, 779–782. (d) Tanner, D. D.; Samal, P. W.;
Tomoki, C.-S. R.; Henriquez, R. J. Am. Chem. Soc. 1979, 101, 1168–
1175.
(16) (a) Hammond, G. S.; Fox, J. R. J. Am. Chem. Soc. 1964, 86, 1918–
1921. (b) Waits, H. P.; Hammond, G. S. J. Am. Chem. Soc. 1964, 86,
1911–1918. (c) Nelsen, S. F.; Bartlett, P. D. J. Am. Chem. Soc. 1966,
88, 143–149.
(17) (a) Kamal, A.; Ramesh Khanna, G. B.; Ramu, R. Tetrahedron:
Asymmetry 2002, 13, 2039–2051. (b) Kamal, A.; Ramesh Khanna,
G. B. Tetrahedron: Asymmetry 2001, 12, 405–410.
(18) (a) Corey, E. J.; Wu, Y. J. J. Am. Chem. Soc. 1993, 115, 8871–8872.
(b) Fukuda, Y.; Okamoto, Y. Tetrahedron 2002, 58, 2513–2521.
(19) (a) Suto, Y.; Kumagay, N.; Matsunaga, S.; Kanai, M.; Shibasaki, M.
Org. Lett. 2003, 5, 3148–3150. (b) Mirmashhori, B.; Azizi, N.; Saidi,
M. R. J. Mol. Cat. A: Chemical 2006, 247, 159–161.
(20) Miura, Y.; Muranaka, Y.; Teki, Y. J. Org. Chem. 2006, 71, 4786–
4794.
9
18020 J. AM. CHEM. SOC. VOL. 130, NO. 52, 2008

Ti(IV)-Mediated Ingold-Fischer Effect
A R T I C L E S
Figure 1. Effect of Ti(IV) on the selectivity in the formation of product 9.
Table 3. Scavenger Effect of Radical B Determined by Studying
the Cage Effect in Thermal Decomposition of AIBN at 80 °Ca
9
7 + 8
IBN
(selectivity %)b
[3]/[AIBN]
(selectivity %)b
(selectivity %)b
0
0
50
76
85
83
100
45
22
15
17
0
5
2
0
0
0.25
0.50
1.00
2.00
a All reactions were performed under N2 atmosphere for 1.5 h; see
Experimental Section and Supporting Information for details. b With the
term selectiVity we mean here the percent number of R-cyanoisopropyl
radicals which lead to the formation of the different products. It was
determined by 1H NMR with 1-biphenyl-4-yl-ethanone as an internal
standard added to the crude reaction mixture.
Figure 2. Possible chelating action of Ti(IV) on AIBN.
since one important condition for the Ingold-Fisher effect is
the simultaneous presence of a transient and a persistent radical
in solution.
By considering that homolytic decomposition of 3 affords 2
equiv of radical A, the observed linear dependence of selectivity
in product 9 on the amount of Ti(IV) would suggest that each
transition metal is actually able to coordinate two A radicals
(Figure 2), both deriving from the same chelated molecule, as
shown in Scheme 3.
decomposition of AIBN. In fact, when a solution of AIBN in
benzene was heated at 80 °C, under N2 atmosphere, in the
presence of a stoichiometric amount of TiCl4, a white solid
complex precipitated and, after 1.5 h, 83% of unreacted AIBN
was recovered. This represents an important exception to the
Kochi and Mog23 statement, according to which the formation
of radicals from the decomposition of AIBN at 80 °C should
be independent of the presence of metal salts. Indeed, this
unexpected behavior may be simply ascribed to the coordinating
action of Ti(IV) on the cyano groups. The resulting complex
has a low solubility in benzene, thus decreasing the concentra-
tion of AIBN in solution.
However, when 1 equiv of dimer 3 was added to the mixture,
the complex was completely destroyed, as evidenced by the
instantaneous disappearance of the white solid, and the reaction
occurred affording 9 in good yields.
To investigate the scavenging power of radical B, we
determined the cage effect of AIBN in the presence of the
Ti(IV)-3 complex. To study the cage effect it is important to
have a scavenger which would trap all radicals that escape the
cage and no others. A scavenger is presumed to meet this
condition if the same cage effect is found over a range of
scavenger concentrations and also with different scavengers.
Thus, the thermal decomposition of AIBN was conducted in
the presence of different amounts of 1:1 Ti(IV)-3 complex.
The results, reported in Table 3 and shown in Figure 3, point
out the high persistent character of radical B deriving from its
An excess of Ti(IV) with respect to 3 surprisingly decreased
the yield of 9. This is due to the fact that, in spite of what is
usually reported for other transition metal salts, like cupric23
and stannous24 salts, Ti(IV) seems to inhibit or slow the
(21) Fujisawa, T.; Monroe, B. M.; Hammond, G. S. J. Am. Chem. Soc.
1970, 92, 542–544.
(22) Weiner, S. A.; Hammond, G. S. J. Am. Chem. Soc. 1969, 91, 986–
991.
(24) Tabka, T. T.; Chenal, J.-M.; Widmaier, J.-M. Polym. Int. 2000, 49,
(23) Kochi, J. K.; Mog, D. M. J. Am. Chem. Soc. 1965, 87, 522–528.
412–416.
9
J. AM. CHEM. SOC. VOL. 130, NO. 52, 2008 18021

A R T I C L E S
Spaccini et al.
Figure 3. Scavenger effect of radical B determined by studying the cage effect in thermal decomposition of AIBN at 80 °C (2 ) 9; 9 ) 7 + 8).
Scheme 6. Effect of Ti(IV) on the Synthesis of ꢀ-Hydroxynitriles
On the basis of these intriguing results, the effect of Ti(IV)
was tested on a wider range of nonsymmetric substrates (Scheme
6), with the aim to synthesize a wider range of products and to
better clarify the reaction mechanism leading to the homolytic
cleavage of the C-C bond.
The results are summarized in Table 4.
Substrates 1a and 1f underwent homolytic decomposition at
80 °C in benzene even in the absence of Ti(IV), meaning that
the resulting radicals are more stable than radical A. This is
consistent with what is described for diphenyl methyl radicals
Ph2-C•-R by considering the effect of R on the bond
dissociation enthalpy (BDE) of the corresponding dimers. The
BDE values of the dimers are 17.7, 20.0, and 21.8 kcal/mol
when R ) COPh, COMe, and CO2Me, respectively.25
Table 4. Reactivity of R,ꢀ-dihydroxy ketones versus AIBN in the
a
Presence of Stoichiometric Amounts of TiCl4
However, the addition of Ti(IV) to solutions containing 1a
or 1f led to cleavage of the C-C bond even at room temperature
(eq 5).
entry
substrate
[1]/[AIBN]
10 (yield %)b
11 (yield %)b
1c
2
3
4
5
1a
1a
1a
1b
1b
1c
1c
1d
1e
1f
0.50
0.50
0.25
0.50
0.25
0.50
0.25
0.50
0.50
0.50
0.50
0.25
10a: -
11a: -
10a: >99 (96)
10a: >99
10a: 87
11a: 48
11a: 74 (71)
11b: 52
11b: 56 (52)
11c: 51
10a: >99
10a: 89
10a: >99
10a: 82
6
7
8
9
11c: 58 (54)
11d: -
10e: -
11e: -
10d
11
12
10f: -
11a: -
1f
1f
10f: 54
11a: 43
The results reported in Table 3 confirmed the homolytic
nature of the C-C bond cleavage. In the absence of Ti(IV) no
cross-coupling was observed (entries 1 and 10). Even radical
D, when coordinated by Ti(IV), underwent cross-combination
in the presence of AIBN affording ꢀ-hydroxynitrile derivatives
in good yields. Also in this case Ti(IV) plays a key role in the
stabilization of the radical, though radical D resulted in being
less stable, due to the lack of the captodative effect. In fact, by
operating in stoichiometric amounts of R-cyanoisopropyl radi-
cals, radical C always resulted in being a better scavenger as
compared with radical D, affording the corresponding products
in higher yields (entries 2, 4, 6, and 11). This was particularly
true for radicals Ca and Da-c, whereas the difference between
radicals Cf and Da was limited, due to the lower stabilization
of Cf as compared with Ca. The yields could be increased by
working with an excess of AIBN (entries 3, 5, 7, and 12).
10f: 69 (63)
11a: 60
a All reactions were performed under N2 atmosphere at 80 °C for
1.5 h; see Experimental Section and Supporting Information for details.
b Yields have been determined by 1H NMR with 1-biphe-
nyl-4-yl-ethanone as an internal standard added to the crude reaction
mixture. Isolated yields are reported in brackets. c The reaction was
conducted in the absence of Ti(IV). Benzoin and benzaldehyde were
recovered as the sole products. d The reaction was conducted in the
absence of Ti(IV), affording a mixture of disproportion products.
enhanced stabilization: with an equimolar concentration of B
only 24% of R-cyanoisopropyl radicals were not trapped. A
small excess of B was sufficient to capture all the radicals
escaping from the cage, obtaining a value of 15% which is
consistent with the data reported in the literature for different
scavengers in benzene.19
9
18022 J. AM. CHEM. SOC. VOL. 130, NO. 52, 2008

Ti(IV)-Mediated Ingold-Fischer Effect
A R T I C L E S
gas. Flash column chromatography was performed by using a
40-63 µm silica gel packing.
General Procedure for the Determination of Ti(IV) Role in
the C-C Bond Cleavage of 1 and 3. A well-stirred solution of
compound 1 or 3 (0.6 mmol) in benzene (10 mL) was heated at
80.0 °C for 1.5 h under nitrogen atmosphere and in the presence
of different amounts of TiCl4 solution. The reaction mixture was
quickly cooled at room temperature, quenched with water (20 mL),
and extracted with ethyl acetate (3 × 10 mL). The organic phase
was washed with brine, dried (Na2SO4), filtered, and concentrated
1
under vacuum. The crude reaction mixture was analyzed by H
NMR upon addition of 1-biphenyl-4-yl-ethanone as an internal
standard.
General Procedure for the Determination of the Cage
Effect. A well-stirred solution of AIBN (1.2 mmol) in benzene
(10 mL) was heated at 80.0 °C for 1.5 h under nitrogen atmosphere
and in the presence of different amounts of TiCl4 and 1 or 3 in a
1:1 ratio. The reaction mixture was quickly cooled at room
temperature, quenched with water (20 mL), and extracted with ethyl
acetate (3 × 10 mL). The organic phase was washed with brine,
dried (Na2SO4), filtered, and concentrated under vacuum. The crude
reaction mixture was analyzed by 1H NMR upon addition of
1-biphenyl-4-yl-ethanone as an internal standard.
Figure 4. Scavenger effect of 1a determined by studying the cage effect
in thermal decomposition of AIBN at 80 °C (b ) 10a; 9 ) 11a; 2 ) 7
+ 8).
When R3 was an aliphatic group (1d), radical Dd did not
afford any coupling and the only product was 10a (entry 8).
1e, obtained by substitution of the phenyl group in R2 with a
hydrogen atom, did not undergo any C-C bond cleavage (entry
9). In both cases, the behavior has to be ascribed to the lower
stabilization of the corresponding radicals due to the lack of
the aromatic ring, which usually plays a well-known key role
in the stabilization of carbon-centered radicals.
Figure 4 shows the scavenger effect of compound 1a on the
decomposition of AIBN. Also in this case the cage effect was
analogous to those reported in the literature and demonstrated
the high stabilized character of radicals Ca and Da. The
difference between Ca and Da in trapping R-cyanoisopropyl
radicals is also emphasized.
General Procedure for the Synthesis of ꢀ-Hydroxynitriles.
A well-stirred solution of compounds 1a-f (Table 3) or compound
3 (0.6 mmol), TiCl4 (0.6 mL), and AIBN (1.2 or 2.4 mmol) in
benzene (10 mL) was heated at 80.0 °C for 1.5 h under an
atmosphere of nitrogen. The reaction mixture was quickly cooled
at room temperature, quenched with 20 mL of water, and extracted
with ethyl acetate (3 × 10 mL). The organic phase was washed
with brine, dried (Na2SO4), filtered, and concentrated under vacuum.
The desired products were isolated by flash column chromatography
and purified by crystallization (see below). The yields reported in
1
Table 4 refer to H NMR with 1-biphenyl-4-yl-ethanone as an
internal standard.
Characterization of the Products. 3-Cyano-2-hydroxy-3,3-
dimethyl-2-phenyl-propionic Acid, Methyl Ester (9). White solid,
purified by crystallization from ethyl acetate, mp 139-140 °C. 1H
NMR (CDCl3): δH 1.29 (3H, s), 1.45 (3H, s), 3.91 (1H, OH, s,
D2O exchange), 3.96 (3H, s), 7.36-7.38 (3H, m), 7.81-7.84 (2H,
m) ppm. 13C NMR (CDCl3): δC 22.4 (CH3), 23.1 (CH3), 40.5 (C),
53.8 (CH3), 79.9 (C), 123.5 (CN), 126.7 2CH), 128.2 (2CH), 128.8
(CH), 136.7 (C), 172.8 (CO) ppm. IR (liquid film): υmax 3411, 2995,
2955, 2243, 1736, 1255 cm-1. MS (m/z): 233 (M+, <1), 165
(M-C(CH3)2CN, 12), 105 (M-C(CH3)2CN-COOMe, 100).
3-Hydroxy-2,2-dimethyl-4-oxo-3,4-diphenyl-butyronitrile (10a).
White solid, purified by flash column chromatography (eluent
hexane/ethyl acetate 7:3), crystallized from ethyl ether, mp 164-166
Conclusions
We have developed an innovative methodology for the
synthesis of ꢀ-hydroxynitriles by cross-combination between
stabilized radicals and the R-cyanoisopropyl radical. Our results
emphasize the key role of Ti(IV) both in the production of
stabilized radicals generated in situ from R,ꢀ-dihydroxy ketones
and in the induction of a persistent character onto carbon-
centered radicals, which are able to trap the R-cyanoisopropyl
radicals generated by thermal decomposition of AIBN.
This methodology represents an intriguing route to the one-
pot synthesis of polyfunctional derivatives in good to high yields
and provides a new indication of the important role that titanium
salts can play in organic synthesis.
1
°C. H NMR (CDCl3): δH 1.40 (3H, s), 1.47 (3H, s), 3.18 (1H,
OH, s, D2O exchange), 7.26-7.28 (2H, m), 7.30-7.40 (4H, m),
7.60-7.64 (4H, m) ppm. 13C NMR (CDCl3): δC 23.1 (CH3), 23.3
(CH3), 40.8 (C), 84.1 (C), 124.0 (CN), 126.5 (2CH), 128.3 (CH),
128.5 (2CH), 128.9 (CH), 129.9 (2CH), 132.8 (2CH), 134.8 (C),
136.1 (C), 197.8(CO) ppm. IR (liquid film): υmax 3396, 3020, 2401,
2212,1606,1215cm-1.MS(m/z):279(M+,<1),147(M-CN-PhCO,
7), 105 (PhCO, 100).
Experimental Section
General. All materials were purchased from commercial
suppliers. TiCl4 was provided in CH2Cl2 (1 M solution). Benzene
was distilled and dried on molecolar sieves prior to use. R,ꢀ-
Dihydroxy ketones were prepared according to the literature.13,14
All reactions were performed in benzene at 80 °C under nitrogen
atmosphere. NMR spectra were recorded at 500 MHz for 1H
and 125 MHz for 13C, measured in CDCl3. Chemical shifts (δ)
were presented in ppm, using the CDCl3 peak (δ ) 7.26 ppm
3-Hydroxy-2,2-dimethyl-4-oxo-3-phenyl-pentanenitrile (10f).
White solid, purified by flash column chromatography (eluent
hexane/ethyl actetate/methanol 7.5:2:0.5), crystallized from ethyl
1
acetate, mp 94-96 °C. H NMR (CDCl3): δH 1.31 (3H, s), 1.47
(3H, s), 1.84 (3H, s), 3.84 (1H, OH, s, D2O exchange), 7.44-7.48
(2H, m), 7.56-7.60 (1H, m), 7.98-8.00 (2H, m) ppm. 13C NMR
(CDCl3): δC 22.5 (CH3), 22.9 (CH3), 23.3 (CH3), 40.5 (C), 81.4
(C), 123.7 (CN), 128.4 (2CH), 129.8 (2CH), 133.1 (CH), 136.1
(C), 203.1 (CO) ppm. IR (liquid film): υmax 3444, 3020, 2235, 1670,
1597, 1459, 1216 cm-1. MS (m/z): 217 (M+, <1), 174 (M-CH3CO,
19), 148 (M-C(CH3)2CN, 7), 105 (PhCO, 100).
1
for H and 77.00 ppm for 13C) as an internal standard. Mass
spectra were performed with a GLC-MS instrument, using a gas
chromatograph equipped with an SBP-1 fused silica column (30
m × 0.2 mm i.d., 0.2 µm film thickness) and helium as a carrier
3-Hydroxy-2,2-dimethyl-3-phenyl-proprionitrile (11a). White
solid, purified by flash column chromatography (eluent chloroform/
(25) Neumann, W. P.; Stapel, R. in ref 1b, pp 219-222.
9
J. AM. CHEM. SOC. VOL. 130, NO. 52, 2008 18023

A R T I C L E S
Spaccini et al.
1
hexane/ethyl ether 6:3.5:0.5), mp 70-73 °C. H NMR (CDCl3):
δH 1.23 (3H, s), 1.43 (3H, s), 2.25 (1H, d, OH, J ) 3.28 Hz D2O
exchange), 4.56 (1H, d, J ) 3.28 Hz, s after D2O exchange),
7.32-7.42 (5H, m) ppm. 13C NMR (CDCl3): δC 23.22 (CH3), 24.26
(CH3), 39.33 (C), 79.14 (CH), 123.89 (CN), 127.63 (2CH), 128.68
(CH)139.20 (C) ppm. IR (liquid film): υmax 3454, 2982, 2238, 1683,
1455 cm-1. MS (m/z): 175 (M+, <1), 107 (M-C(CH3)2CN, 100).
3-Hydroxy-2,2-dimethyl-3-p-tolylpropanenitrile (11b). Oil,
D2O excange), 7.57 (2H, d, 2CH), 7.66 (2H, d, 2CH) ppm. 13C
NMR (CDCl3): δC 23.46 (CH3), 23.63 (CH3), 39.31 (C), 78.12 (CH),
112.99 (CN), 118.73 (CN) 123.34 (C), 128.61 (2CH),132.41 (2CH),
144.53 (C) ppm. IR (liquid film): υmax 3454, 2985, 2880, 2232,
1666, 1610, 1467, 1196, 1065 cm-1. MS (m/z): 181 (M-H2O, <1);
132 (M-C(CH3)2CN, 100); 104 (70); 77 (Ph•, 49); 69(•C(CH3)2CN,
90).
1
purified by flash column chromatography (eluent chloroform). H
Acknowledgment. Financial support from MURST (Prin 2006)
and Politecnico di Milano is gratefully acknowledged. We thank
Professor Francesco Minisci for fruitful chemical discussions. The
authors dedicate this article to the memory of their coauthor,
Professor Ombretta Porta.
NMR (CDCl3): δH 1.22 (3H, s), 1.43 (3H, s), 2.16 (1H, s, OH,
D2O exchange), 2.36 (3H, s), 4.52 (1H), 7.18 (2H, d, 2CH), 7.32
(2H, d, 2CH) ppm. 13C NMR (CDCl3): δC 21.07 (CH3), 22.90 (CH3),
23,94 (CH3), 39.10 (C), 78.72 (CH), 123.7 (CN), 123.34 (C), 127.17
(2CH),129.02 (2CH), 135.97(C) ppm. IR (liquid film): υmax 3460,
2982, 2926, 2877, 2239, 1683, 1516, 1471, 1448, 1260, 1182, 1061,
1019, 818, 747, 706, 515 cm-1. MS (m/z): 189 (M+, <1); 121
(M-C(CH3)2CN, 100); 93 (55); 77 (Ph•, 41).
4-(2-Cyano-1-hydroxy-2,2-dimethyl-ethyl)benzonitrile (11c).
Oil, purified by flash column chromatography (eluent chloroform).
1H NMR (CDCl3): δH 1.27 (3H, s), 1.41 (3H, s), 2.45 (1H, d, OH,
J ) 3.21 Hz D2O exchange), 4.64 (1H, d, J ) 3.21 Hz, s after
Supporting Information Available: Experimental procedures
and spectral data of products 9, 10a, 10f, and 11a-c. This
material is available free of charge via the Internet at http://
pubs.acs.org.
JA807613Q
9
18024 J. AM. CHEM. SOC. VOL. 130, NO. 52, 2008