Assessing the suitability of 1,2,3-triazole linkers for covalent immobilization of chiral ligands: Application to enantioselective phenylation of aldehydes

Article abstract of DOI:10.1021/jo0624952

Alkynyl-functionalized amino alcohols have been covalently supported on azidomethylpolystyrene resins with different levels of functionalization through Cu(I)-catalyzed 1,3-dipolar cycloadditions ("click chemistry"). The resulting 1,2,3-triazole-substituted resins, characterized by different levels of ligand loading and, depending on the nature of the alkynyl-functionalized amino alcohol, the presence of a one-carbon, four-carbon, or eight-carbon linear spacer, have been tested as catalysts in the enantioselective phenyl transfer from zinc to aldehydes. High catalytic activities and enantioselectivities (up to 82% ee) have been recorded. The influence of structural characteristics of the resin on enantioselectivity are discussed, and the limitations in enantiocontrol inherent to the use of a 1,2,3-triazole linker have been rationalized with the help of DFT calculations on model systems.

Full text of DOI:10.1021/jo0624952

Assessing the Suitability of 1,2,3-Triazole Linkers for Covalent
Immobilization of Chiral Ligands: Application to Enantioselective
Phenylation of Aldehydes
Amaia Bastero,† Daniel Font,† and Miquel A. Perica`s*,†,‡
Institute of Chemical Research of Catalonia (ICIQ), AV. Pa¨ısos Catalans 16, 43007 Tarragona, Spain and
Departament de Qu´ımica Orga`nica, UniVersitat de Barcelona (UB), 08028 Barcelona, Spain
mapericas@iciq.es
ReceiVed December 5, 2006
Alkynyl-functionalized amino alcohols have been covalently supported on azidomethylpolystyrene resins
with different levels of functionalization through Cu(I)-catalyzed 1,3-dipolar cycloadditions (“click
chemistry”). The resulting 1,2,3-triazole-substituted resins, characterized by different levels of ligand
loading and, depending on the nature of the alkynyl-functionalized amino alcohol, the presence of a
one-carbon, four-carbon, or eight-carbon linear spacer, have been tested as catalysts in the enantioselective
phenyl transfer from zinc to aldehydes. High catalytic activities and enantioselectivities (up to 82% ee)
have been recorded. The influence of structural characteristics of the resin on enantioselectivity are
discussed, and the limitations in enantiocontrol inherent to the use of a 1,2,3-triazole linker have been
rationalized with the help of DFT calculations on model systems.
Introduction
and the catalyst to be anchored, the immobilization strategy is
based on covalent or noncovalent grafting (e.g., electrostatic
interaction, hydrogen bond, adsorption, or entrapment).3 Cova-
lent grafting implies, in general, a more stable interaction and
therefore a broader applicability of the immobilized catalysts.
However, it usually requires greater synthetic effort since the
catalyst has to be chemically modified with a group that enables
anchoring to the polymer.
Catalysts are often the most expensive ingredients in mixtures
leading to chemical reactions and frequently difficult to separate
from the reaction product. Therefore, use of catalysts supported
onto different platforms is of great interest since it can allow
simpler product separation while offering the possibility of
catalyst recycling as an additional bonus.1 Additional attractive-
ness of the immobilized catalysts lies in the possibility of
performing catalytic reactions in continuous mode, and this is
of interest for industrial applications.2 Different types of
polymeric, inorganic, or hybrid materials are widely used as
insoluble supports. Depending on the nature of both the support
We previously reported the immobilization of chiral ligands
to resins following different synthetic routes.4,5 As a guide to
(3) Jacobs, P. A.; Devos, D. E.; Vankelecom, I. F. J. Chiral Catalyst
Immobilization and Recycling; VCH Publishers: Weinheim, 2000.
(4) (a) Vidal-Ferran, A.; Moyano, A.; Perica`s, M. A.; Riera, A. J. Org.
Chem. 1997, 62, 4970-4982. (b) Vidal-Ferran, A.; Bampos, N.; Moyano,
A.; Perica`s, M. A.; Riera, A.; Sanders, J. K. M. J. Org. Chem. 1998, 63,
6309-6318.
(5) (a) Perica`s, M. A.; Castellnou, D.; Rodriguez, I.; Riera, A.; Sola`, L.
AdV. Synth. Catal. 2003, 345, 1305-1313. (b) Castellnou, D.; Sola`, L.;
Jimeno, C.; Fraile, J. M.; Mayoral, J. A.; Riera, A.; Perica`s, M. A. J. Org.
Chem. 2005, 70, 433-438. (c) Castellnou, D.; Fontes, M.; Jimeno, C.; Font,
D.; Sola`, L.; Verdaguer, X.; Perica`s, M. A. Tetrahedron 2005, 61, 12111-
12120.
† Institute of Chemical Research of Catalonia (ICIQ).
‡ Universitat de Barcelona (UB).
(1) Reviews on immobilized catalysts include the following: (a) Im-
mobilized Catalysts: Solid Phases, Immobilization and Applications. In
Topics in Current Chemistry; Kirschning, A., Ed.; Springer GmbH: Berlin,
2004; Vol. 242, pp 1-336. (b) In Polymeric Materials in Organic Synthesis
and Catalysis; Buchmeiser, D. R., Ed.; VCH Publishers: Weinheim, 2003.
(2) For a recent review, see: Jas, G.; Kirschning, A. Chem. Eur. J. 2003,
9, 5708-5723.
10.1021/jo0624952 CCC: $37.00 © 2007 American Chemical Society
Published on Web 03/09/2007
2460
J. Org. Chem. 2007, 72, 2460-2468

1,2,3-Triazole Linkers for Immobilization of Chiral Ligands
FIGURE 1. Some modular amino alcohol ligands anchored to polymers through remote functions for minimal perturbation of catalytic sites.
FIGURE 2. Possible pathways in metal catalysis with click-supported ligands.
this work, we designed the monomeric ligands in such a way
that anchoring to the polymer would involve positions remote
from the catalytic center and the transition states involved in
the catalytic cycle would remain essentially unaffected. Use of
polymers functionalized with amino alcohol residues and
designed according to these principles (I-III) as catalysts for
enantioselective carbonyl additions4b,5 has led to very good
results in terms of catalytic activity and enantioselectivity
(Figure 1). However, anchoring these ligands onto polymers
critically depends on the success of nucleophilic substitution
reactions, and this poses some limitation with respect to the
ligand structure and prevents use of the anchoring step as a
source of diversity for the preparation of libraries of supported
ligands.
“Click” chemistry categorizes highly efficient, specific, and
functional-group-tolerant chemical transformations to make
molecular connections.6 The most popular click reaction is the
copper(I)-catalyzed regioselective Huisgen dipolar cycloaddition
reaction between alkynes and azides to yield 1,2,3-triazoles.7,8
This reaction has seen enormous application in the last years
in different fields from drug discovery9 to material science,10
but has been barely used as a tool to support reagents11 or
catalysts on polymers.12,13
Taking into account the limitations associated to ligand
anchoring through nucleophilic substitution, click chemistry
appeared to be a most convenient synthetic alternative. It is to
be noted, however, that the 1,2,3-triazole moiety, conceived as
a mere linker between the chiral ligand and the polymer, could
also interact with the catalytic metal and trigger non-enanti-
oselective reaction pathways (Figure 2).
Whereas formation of chelates like B have been shown in a
recently reported type of achiral phosphines for palladium-
catalyzed processes,14 the coexistence of enantioselective and
non-enantioselective pathways is strongly suggested in click-
polymer-supported chiral aza(bisoxazolines) for Cu(II)-catalyzed
processes since much higher enantioselectivities are recorded
when the 1,2,3-triazole moiety is avoided as a linker.12 In fact,
the only reported example where a click-supported ligand depicts
very high enantioselectivity is in organocatalysis, where the
1,2,3-triazole should act as a mere spectator.13
(6) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed.
2001,40, 2004-2021.
(7) Huisgen, R. In 1,3-Dipolar Cycloaddition Chemistry; Padwa, A., Ed.;
Wiley: New York, 1984; Vol. 1, pp 1-176.
(8) (a) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem., Int. Ed. 2002, 41, 2596-2599. (b) Tornøe, C. W.;
Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67, 3057-3064.
(9) Kolb, H. C.; Sharpless, K. B. Drug DiscoVery Today 2003, 8, 1128-
1137.
(10) (a) Collman, J. P.; Devaraj, N. K.; Chidsey, C. E. D. Langmuir
2004, 20, 1051-1053. (b) Lee, J. K.; Chi, Y. S.; Choi, I. S. Langmuir
2004, 20, 3844-3847. (c) O’Reilly, R. K.; Joralemon, M. J.; Wooley, K.
L.; Hawker, C. J. Chem. Mater. 2005, 17, 5976-5988. (d) Fleming, D. A.;
Thode, C. J.; Williams, M. E. Chem. Mater. 2006, 18, 2327-2334. (e)
Collman, J. P.; Devaraj, N. K.; Eberspacher, T. P. A.; Chidsey, C. E. D.
Langmuir 2006, 22, 2457-2464.
FIGURE 3. Target polymer-supported amino alcohols (1a-c) as-
sembled through click chemistry and containing variable-length spacers
(X).
In order to critically assess the usefulness of click chemistry
as a tool for supporting on polymers chiral ligands for
asymmetric catalysis purposes, we decided to prepare a family
of polymer-supported amino alcohols (1), assembled through
1,2,3-triazole linkers onto slightly cross-linked polystyrene resins
and containing variable length spacers (X) (Figure 3). We hoped
that long-chain spacers could contribute to isolate the amino
(11) Gheorge, A.; Matsuno, A.; Reiser, O. AdV. Synth. Catal. 2006, 348,
1016-1020.
(12) Gissibi, A.; Finn, M. G.; Reiser, O. Org. Lett. 2005, 7, 2325-2328.
(13) Font, D.; Jimeno, C.; Perica`s, M. A. Org. Lett. 2006, 8, 4653-
4655.
(14) Detz, R. J.; Are´valo Heras, S.; de Gelder, R.; van Leeuwen, P. W.
N. M.; Hiemstra, H.; Reek, J. N. H.; van Maarseveen, J. H. Org. Lett. 2006,
8, 3227-3230.
J. Org. Chem, Vol. 72, No. 7, 2007 2461

Bastero et al.
SCHEME 1. Catalytic Asymmetric Phenyl Transfer to
Aldehydes (L*: Chiral Ligand)
alcohol moiety from the triazole one, thus favoring A- over
B-type chelates and facilitating enantioselective pathways.
The amino alcohol moiety present in these polymers is known
to efficiently control the enantioselective arylation of aldehydes
when linked to a Merrifield resin through nucleophilic
substitution,5d and we decided to use the click resins 1a-c as
insoluble catalysts in the asymmetric phenyl transfer to alde-
hydes (Scheme 1).
FIGURE 4. Some pharmaceutically active diarylmethanols.
SCHEME 2. Propargylation of Amino Alcohol 2
This reaction represents the most convenient way to obtain
enantiopure diarylmethanols,15 which are important intermedi-
ates for the synthesis of biologically active compounds.16 For
example, (R)-neobenodine, (R)-orphenadrine, and (S)-carbinox-
amine are compounds of this class that have been used for a
long time as muscle relaxants or antihistaminics (Figure 4).17
Complete catalyst separation from products is one of the most
strict requisites for the catalytic synthesis of pharmaceutically
relevant products, and use of covalently immobilized catalysts
appears to be a convenient strategy toward this goal. In the case
of enantioselective arylation of aldehydes, although many
different homogeneous catalytic systems able to promote the
reaction have been developed,18 the search for polymer-based
catalytic systems has been accompanied by much less success.
Indeed, the first examples of recyclable catalytic systems for
this reaction showed that enantioselectivity was only achieved
for soluble polymeric supports.19 Attempts to use a catalyst
bound to a cross-linked polymer failed and led to racemic
products,19b probably due to an unfavorable balance with the
noncatalyzed background reaction. Only quite recently the first
examples have been reported where high enantioselectivity is
achieved by promoting the reaction with highly active ligands5a,b
anchored to insoluble resins.5c
linking of the polymer chain and the length of the spacer (X)
between the triazole and the amino alcohol moieties on the same
parameter. In addition, DFT calculations on the possible
interference of the triazole moiety in the reaction are presented.
Results and Discussion
(R)-2-Piperazino-1,1,2-triphenylethanol (2), whose N-substi-
tuted derivatives depict an optimal catalytic profile in solution,20
has a NH moiety that can allow anchoring onto polymers with
minimal steric perturbation of the relevant â-amino alcohol
unit.5b,c In the present instance, the alkyne handle required for
click chemistry was introduced by alkynylation at that site. Thus,
treatment of 2 with propargyl bromide in acetonitrile in the
presence of cesium carbonate allowed isolation of the desired
alkyne-tagged monomer 3a in 91% yield (Scheme 2).
The complementary azido group for the dipolar cycloaddition
was installed onto commercial Merrifield resins (1% DVB,
f ) 0.74-2.5 mmol Cl/g resin)21 by reaction with sodium
azide.22 In this way, a set of azido resins with different levels
of functionalization (f ) 0.74-2.25 mmol N3/g resin) was
prepared.23 The Cu(I)-catalyzed reaction between 3a and the
polymer-bound azide took place smoothly in 1:1 DMF/THF at
35 °C to lead to resins 1a (Scheme 3).24
Herein we report the use of click chemistry as a most suitable
strategy for the preparation of polymer-supported amino alcohols
1a-c. We also describe the use of these insoluble materials as
catalysts in the asymmetric phenylation of aldehydes and the
results of a systematic investigation of the influence on
enantioselectivity of the degrees of functionalization and cross-
The progress of the cycloaddition reaction could be easily
controlled by withdrawing resin samples from the reaction vessel
at different times and following the disappearance of the azide
function by IR spectroscopy (ca. 2100 cm-1) (Figure 5). As a
general rule, stirring the reaction mixture (shaker) at 35 °C for
(15) The asymmetric reduction of a suitable diaryl ketone precursor is
usually hampered by low ee’s due to the steric and electronic similarities
between both aryl groups, see, for instance: Ohkuma, T.; Koizumi, M.;
Ikehira, H.; Yokozawa, T.; Noyori, R. Org. Lett. 2000, 2, 659-662 and
references therein.
(16) (a) Bolshan, Y.; Chen, C.-Y.; Chilenski, J. R.; Gosselin, F.; Mathre,
D. J.; O’Shea, P. D.; Roy, A.; Tillyer, R. D. Org. Lett. 2004, 6, 111-114.
(b) El Ahmad, Y.; Maillet, P.; Laurent, E.; Talab, A.; Teste, J. F.; Ce´dat,
M. J.; Fiez-Vandal, P. Y.; Dokhan, R.; Ollivier, R. Eur. J. Med. Chem.
1997, 32, 205-218. (c) Botta, M.; Summa, V.; Corelli, F.; Di Pietro, G.;
Lombardi, P. Tetrahedron: Asymmetry 1996, 7, 1263-1266. (d) Stanchev,
S.; Rakovska, R.; Berova, N.; Snatze, G. Tetrahedron: Asymmetry 1995,
6, 183-189. (e) Toda, F.; Tanaka, K.; Koshiro, K. Tetrahedron: Asymmetry
1991, 2, 873-874. (f) Meguro, K.; Aizawa, M.; Sohda, T.; Kawamatsu,
Y.; Nagaoka, A. Chem. Pharm. Bull. 1985, 33, 3787-3797.
(17) (a) Harms, A. F.; Nauta, W. T. J. Med. Pharm. Chem. 1960, 2,
57-77. (b) Sperber, N.; Papa, D.; Schwenk, E.; Sherlock, M. J. Am. Chem.
Soc. 1949, 71, 887-890.
(20) Sola`, L.; Reddy, K. S.; Vidal-Ferran, A.; Moyano, A.; Perica`s, M.
A.; Riera, A.; Alvarez-Larena, A.; Piniella, J. F. J. Org. Chem. 1998, 63,
7078-7082.
(21) Product specifications provided by the vendor (Novabiochem).
(22) Harju, K.; Vahermo, M.; Mutikainen, I.; Yli-Kauhaluoma, J. J.
Comb. Chem. 2003, 5, 826-833.
(23) For the nitrogen-containing resins, the degree of functionalization,
f [mmol of functional fragment/g of resin] is calculated from the results of
elemental analysis with the formula f ) (0.714/n)%N, where n is the number
of nitrogen atoms in the functional unit and %N is the percent of nitrogen
provided by the elemental analysis. For a given resin, the maximum possible
substitution level fmax [mmol of functional fragment/g of resin] can be
calculated with the formula fmax ) fo/[1 + fo(∆Mw) 10-3], where fo is the
functionalization of the starting resin and ∆Mw is the difference in molecular
weight between the final and the initial functional fragments. The yield
(%) of the considered step can then be evaluated as 100f/fmax.
(24) Lo¨ber, S.; Gmeiner, P. Tetrahedron 2004, 60, 8699-8702.
(18) For a review, see: Schmidt, F.; Stemmler, R. T.; Rudolph, J.; Bolm,
C. Chem. Soc. ReV. 2006, 35, 454-470.
(19) (a) Huang, W.-S.; Hu, Q.-S.; Pu, L. J. Org. Chem. 1999, 64, 7940-
7956. (b) Bolm, C.; Hermanns, N.; Claben, A.; Mun˜iz, K. Bioorg. Med.
Chem. Lett. 2002, 12, 1795-1798.
2462 J. Org. Chem., Vol. 72, No. 7, 2007

1,2,3-Triazole Linkers for Immobilization of Chiral Ligands
FIGURE 5. Comparison of IR spectra of an azido resin (red line; azide band at 2108 cm-1) and the corresponding click resin 1a (blue line).
SCHEME 3. Synthesis of Supported Catalysts 1a from Commercial Merrifield Resins
1 day led to complete conversion. Using this procedure a set of
resins 1a with different degrees of ligand functionalization
(f ) 0.59-1.16 mmol/g resin) could be easily prepared.
Elemental analysis of the final resins23 in conjunction with
high-resolution magic angle spinning (HRMAS) 13C NMR
spectroscopy25 allowed us to establish that the amino alcohol
anchors to the resins in quantitative yields. A comparison
between the 13C NMR spectra of resin 1a (HR-MAS) and the
model compound 4a, prepared from benzyl azide and alkyne
3a, is shown in Figure 6.
effects: first, it allows reducing the amount of diphenylzinc
employed in the reaction since both phenyl groups in this reagent
can be equally transferred to the carbonyl substrate through the
mixed species. On the other hand, since PhZnEt is less reactive
than ZnPh2, the non-enantioselective background reaction is
efficiently suppressed with its use.
We then tested our novel resins 1a as catalysts in the
asymmetric phenylation of p-tolualdehyde and p-anisaldehyde
using PhZnEt as the phenyl-delivering species. The resins were
first allowed to swell in toluene and then treated with the ZnEt2/
ZnPh2 (2:1) mixture. After 1 h, the reaction mixture was cooled
to 10 °C and the aldehyde was then added. Table 1 shows that
high conversions are obtained after 2 h using only 5 mol %
catalyst with enantioselectivities up to 74% (p-tolualdehyde)
and 77% (p-anisaldehyde). Interestingly, a clear dependence of
enantioselectivity on the degree of functionalization of resin 1a
is observed: for the phenylation of p-tolualdehyde, the corre-
sponding diarylcarbynol is obtained with 74% ee when using a
resin with low functionalization (f ) 0.59 mmol/g) and with
68% ee for a resin with slightly higher functionalization (f )
0.74 mmol/g). In the limit, when resin 1a with a larger level of
functionality (f ) 1.16 mmol/g) was used, the alcohol was
obtained in racemic form.30
With resins 1a in hand, their use as ligands in the catalytic
enantioselective phenylation of aldehydes was explored. Several
protocols have been developed for the homogeneous version
of the reaction using as aryl transfer agent either ZnPh2,26
a
27
mixture of ZnEt2/ZnPh2 (presumably giving rise to PhZnEt
as the actual phenyl delivering agent),28 or phenylboronic acid.29
Use of a mixed ethylphenylzinc reagent, formed in situ from
diethylzinc and diphenylzinc (in a 2:1 ratio), has two beneficial
(25) The resins were allowed to swell in C6D6 overnight before
acquisition of the NMR spectra with the HRMAS probe.
(26) Dosa, P. I.; Ruble, J. C.; Fu, G. C. J. Org. Chem. 1997, 62, 444-
445.
(27) (a) Bolm, C.; Hermanns, N.; Hildebrand, J. P.; Muniz, K. Angew.
Chem., Int. Ed. 2000, 39, 3465-3467. (b) Bolm, C.; Kesselgruber, M.;
Hermanns, N.; Hildebrand, J. P.; Raabe, G. Angew. Chem., Int. Ed. 2001,
40, 1488-1490.
(28) Fontes, M.; Verdaguer, X.; Sola`, L.; Perica`s, M. A.; Riera, A. J.
Org. Chem. 2004, 69, 2532-2543.
(29) (a) Bolm, C.; Rudolph, J. J. Am. Chem. Soc. 2002, 124, 14850-
14851. (b) Rudolph, J.; Schmidt, F.; Bolm, C. Synthesis 2005, 840-842.
(30) On the opposite side, when a low-loading resin (f ) 0.29) prepared
from a commercial Merrifield resin by reaction with defect sodium azide,
capping with MeOH, and cycloaddition with amino alcohol 3a was used in
the reaction, the enantiomeric purity of the resulting carbynol was 60 %. It
is thus suggested that functionalization of the resin has an optimal value in
the vicinity of f ) 0.6 for this particular reaction.
J. Org. Chem, Vol. 72, No. 7, 2007 2463

Bastero et al.
FIGURE 6. Controlling the anchoring of amino alcohol 3a to an azido resin through click chemistry by HRMAS 13C NMR (1a, top). The spectrum
of triazole 4a is included for comparison (bottom). PS: polystyrene.
TABLE 1. Enantioselective Phenylation of Aldehydes with Resins
1a with Variable Degree of Functionalizationa
a Reaction conditions: aldehyde (0.25 mmol); ligand (0.0125 mmol; 5
mol %); Ph2Zn (0.16 mmol); Et2Zn (0.33 mmol); 7.4 mL of toluene; 10
°C; 2 h. b Calculated from %N determined by elemental analysis. c Deter-
FIGURE 7. Possible non-enantioselective pathways in the phenylation
of aldehydes with high-loading resins 1a.
1
mined by H NMR. d Measured by chiral HPLC.35 The product had the R
configuration when a ligand of R configuration was anchored to the resin.
e 3.6% catalyst. f 10% catalyst.
insert a linear hydrocarbon spacer between the triazole moiety
and the amino alcohol fragment which should coordinate to zinc
Although the origin of this behavior has not been investigated
in detail, it can be assumed that a high local concentration of
Lewis acid/Lewis base sites on the catalytically activated
polymer effectively triggers a non-enantioselective reaction
pathway (Figure 7). In this context, it is interesting to point out
that the catalytic efficiency and enantioselectivity of amino
alcohols similar to I (see above) anchored to Merrifield resins
is strongly dependent on the functionalization level of the resin.4b
as a chelate. Two alkyl chains of different length (C4 and C8)
were used as spacers in an attempt to determine the optimal
separation between the two considered fragments. For this
purpose, (R)-2-piperazino-1,1,2-triphenylethanol (2) was reacted
with commercially available 6-chloro-1-hexyne and 10-bromo-
1-decyne32 in the presence of cesium carbonate to yield 3b and
3c, respectively. The subsequent Cu(I)-mediated cycloaddition
of these alkynyl-functionalized amino alcohols with polymer-
bound azides led to resins 1b and 1c in quantitative yield
(Scheme 4). According to the results recorded with 1a, low-
loading azido resins were used in these experiments.
Lowering the amount of 1a employed in the reaction (for
entry 3 only 3.6% catalyst was used) provoked a decrease in
both conversion and enantioselectivity. On the other hand, an
increase to 10 mol % in the amount of 1a (entry 6) did not lead
to any improvement in the enantioselectivity of the process.
Table 2 shows the results of the comparative phenylation of
p-substituted benzaldehydes with the polymeric catalysts 1a-c
As mentioned above, we envisaged that the presence of the
triazole ring near the amino alcohol could give rise to some
interaction with ethylphenylzinc leading to a detrimental effect
in enantioselectivity.31 To avoid this problem, we decided to
(31) Bolm, C.; Rudolph, J. J. Am. Chem. Soc. 2002, 124, 14850-14851.
(32) For example, phenylation of p-tolualdehyde with a Merrifield resin
substituted with (R)-2-piperazino-1,1,2-triphenylethanol as the ligand (5 %)
furnishes the alcohol with 94% ee under similar conditions; see ref 5c.
2464 J. Org. Chem., Vol. 72, No. 7, 2007

1,2,3-Triazole Linkers for Immobilization of Chiral Ligands
TABLE 3. Enantioselective Phenylation of Aldehydes with Resin
SCHEME 4. Preparation of Polymer-Supported Ligands
1b,c
1b (f ) 0.72 mmol/g; 1% DVB)a
TABLE 2. Catalytic Phenylation of Aldehydes with Resins 1a-c:
Effect of the Spacer and Resin Reticulationa
a Reaction conditions: 0.25 mmol aldehyde; 5% catalyst; 10 °C.
1
b Determined by H NMR. c Determined by chiral HPLC.31
SCHEME 5. Phenylation of p-Tolualdehyde with Reference
Ligands
a Reaction conditions: aldehyde (0.25 mmol); ligand (0.0125 mmol; 5
mol %); Ph2Zn (0.16 mmol); Et2Zn (0.33 mmol); 7.4 mL of toluene; 10
°C; 2 h. b Calculated from %N obtained by elemental analysis. c Determined
1
by H NMR. d Determined by chiral HPLC.31 e Resin with higher cross-
linking (2% divinylbenzene). f 10 mol % catalyst used.
containing the different spacers. As it can be readily observed,
resins 1a and 1b exhibit a higher catalytic activity than 1c
(entries 6 and 10). Since separation of the active center from
the polymeric backbone through a longer linker should favor
catalytic activity through a better contact with the reagents,
observation of an opposite behavior is in favor of some
participation of the triazole moiety in the catalytic process.
Regarding the enantioselectivity, slight differences were en-
countered depending on the length of the spacer. As a general
trend resin 1b, containing a four-carbon linker, affords the best
results with the two studied aldehydes.
The possibility of resin recycling and reuse was tested with
the optimal resin 1b on p-anisaldehyde. After three consecutive
runs no change in conversion was detected while the enantio-
meric purity of the resulting carbynol showed a very slight
decrease (78% vs 82% ee). This is indicative of good stability
of the resins under the reaction conditions used.
Once the nature of the optimal resin had been established,
polymer 1b (f ) 0.72 mmol/g, 1% DVB) was tested in the
phenylation of a set of aromatic and aliphatic aldehydes
(Table 3). Using a 5 mol % amount of catalyst and performing
the reaction in toluene at 10 °C for 2-4 h the starting aldehydes
were cleanly converted in an almost quantitatively manner into
the corresponding secondary alcohols with the only exception
As a final step in this scrutiny, to test the influence of the
degree of cross linking of the parent resin on the catalytic
behavior of the supported ligands, two types of 1b resin were
prepared with the same functionalization (f ) 0.66) and different
levels of cross-linking (1% and 2% DVB). Both resins behaved
similarly in terms of enantioselectivity, but the more reticulated
one showed lower catalytic activity as indicated by a lower
conversion in the studied reaction (entries 2 and 3).
According to these results, resin 1b containing a C4 spacer,
prepared from a Merrifield resin of low reticulation (1% DVB),
and possessing a rather low level of functionalization (f ) 0.72
mmol/g) appears to be the optimal one (entries 4 and 9).
(33) Patwardhan, A. P.; Thompson, D. H. Langmuir 2000, 16, 10340-
10350.
J. Org. Chem, Vol. 72, No. 7, 2007 2465

Bastero et al.
FIGURE 8. DFT-calculated complexation energies for the interaction of dimethylzinc with N-2 and N-3 in the triazole ring of model compounds
5-7.
of 2-naphthaldehyde (entry 6), which was only 46% converted
after a 4 h reaction. Interesting to note, 2-ethylbutyraldehyde
and R-methylcinnamaldehyde (entries 4 and 5), which showed
lower reactivity in reactions mediated by an amino alcohol with
the same basic structure supported on a Merrifield resin by
nucleophilic substitution,5c were completely phenylated with
resin 1b.
In terms of enantioselectivity, the reaction delivers all
diarylcarbynol products in enantioenriched form. Over the
family of six studied aldehydes, the mean enantiomeric excess
of the carbynol products is 70%. This value is ca. 20% lower
than that obtained with a similar supported ligand lacking the
1,2,3-triazole moiety,5c and this suggests once again that this
unit can be interfering with the asymmetric process by promot-
ing non-enantioselective pathways like those depicted in
Figures 2 and 7.
This suggestion is reinforced by the fact that the model
compound 4b, prepared from 3b by cycloaddition with benzyl
azide (66% yield), when used as the ligand under the same
reaction conditions yields the alcohol in 94% conversion and
69% ee, this enantioselectivity being much lower than that
recorded with simple derivatives of 2-amino-1,1,2-triphenyle-
thanol28 (Scheme 5).
the suitability of 1,2,3-triazoles as linkers for immobilization
onto polymers of amino alcohols, designed to act as ligands in
the considered process, can critically depend on the thermody-
namics of the interaction of the zinc reagent with the different
electron-rich sites on the ligand molecule.
To gain some insight into these complexation processes, we
decided to study complexation of a prototypical zinc reagent
(dimethylzinc) with the different Lewis base sites in a molecular
system containing the characteristics of ligands 1 from a
theoretical point of view.
Molecule 5 could allow evaluation of the interaction of
dimethylzinc with the nitrogen atoms in the 1,2,3-triazole moiety
exhibiting basic character. Introduction of gem-dimethyl sub-
stitution on the two methylene groups adjacent to the triazole
(in 6 and 7) could give an indication of the possibility of
modifying the thermodynamics of the complexation with
introduction of steric congestion near the 1,2,3-triazole linkers.
On the other hand, 5 could also allow evaluation of the stability
of chelates involving the participation of nitrogen atoms in the
triazole (N-3) and piperazino rings.
All available experimental and theoretical evidence on alkyl
and aryl transfer from zinc to carbonyl groups suggests that
the process is favored by coordination of a Lewis base to the
zinc atom from which the group is transferred.33 Accordingly,
The energy gain associated with all complexation possibilities
would be then compared with the one associated with the
process shown in Scheme 6, which is representative for one of
the pre-equilibrium steps in the catalytic cycle in the amino-
alcohol-promoted alkylation/arylation of aldehydes with dior-
ganylzinc reagents.34
The calculations were performed using density functional
theory (DFT) with the B3LYP functional, as implemented in
the Spartan02 suite of programs.34 Geometry optimization and
energy calculations were run using the all-electron 6-31G* basis
set.
(34) See, for instance: (a) Kitamura, M.; Okada, S.; Suga, S.; Noyori,
R. J. Am. Chem. Soc. 1989, 111, 4028-4036. (b) Yamakawa, M.; Noyori,
R. J. Am. Chem. Soc. 1995, 117, 6327-6335. (c) Kitamura, M.; Oka, H.;
Noyori, R. Tetrahedron 1999, 55, 3605-3614. (d) Rosner, T.; Sears, P. J.;
Nugent, W. A.; Blackmond, D. G. Org. Lett. 2000, 2, 2511-2513. (e)
Goldfuss, B.; Steigelmann, M.; Khan, S. I.; Houk, K. N. J. Org. Chem.
2000, 65, 77-82. (f) Va´zquez, J.; Perica`s, M. A.; Maseras, F.; Lledo´s, A.
J. Org. Chem. 2000, 65, 7303-7309. (g) Rudolph, J.; Rasmussen, T.; Bolm,
C.; Norrby, P. O. Angew. Chem., Int. Ed. 2003, 42, 3002-3005. (h) Jimeno,
C.; Vidal-Ferran, A.; Perica`s, M. A. Org. Lett. 2006, 8, 3895-3898. (i)
Rudolph, J.; Bolm, C.; Norrby, P. O. J. Am. Chem. Soc. 2005, 127, 1548-
1552.
(35) Spartan 02; Wavefuction, Inc.: Irvine, CA.
2466 J. Org. Chem., Vol. 72, No. 7, 2007

1,2,3-Triazole Linkers for Immobilization of Chiral Ligands
SCHEME 6. Pre-equilibrium in the Amino
Alcohol-Mediated Addition of Diorganylzincs to Aldehydes
We present graphically in Figure 8 results on the complex-
ation of dimethylzinc with N-2 and N-3 in 5-7. Full details on
the complexation partners and complexes are provided in the
Supporting Information.
FIGURE 9. Most stable complexes of dimethylzinc with the studied
According to the DFT calculations, interaction of dimeth-
ylzinc with the triazole ring leads to a significant stabilization.
From a general perspective, two aspects of the complexation
process should be highlighted: First, complexation of dimeth-
ylzinc with N-3 is more favorable than with N-2, in agreement
with a higher electron density at that position. Second, introduc-
tion of steric congestion near N-3 (compare 3-DMZ-5 and
3-DMZ-7 in Figure 8) does not represent any significant
variation in the complexation energy.
For reactivity purposes and enantioselectivity discussions, it
is important to compare the stability of these complexes with
that of the N,N chelate 9, derived from 5, with the usual oxygen-
bound complex of dimethylzinc 8 (Scheme 6). Their optimized
structures and complexation energies are given in Figure 9.
Further details can be found in the Supporting Information.
As expected, the calculations predict that complexation of
dimethylzinc with the oxygen atom in the chelated methylzinc
alkoxide derived from the amino alcohol moiety (8) represents
the most favorable situation. However, the difference with the
N,N-chelate 9 that could originate from the simultaneous
interaction of dimethylzinc with the triazole ring and the
dimethylamino substituent (i.e., the piperazino group in ligands
like 1a) is small (ca. 0.5 kcal‚mol-1). In addition, complexation
in 8 is only slightly more favorable (ca. 3 kcal‚mol-1) than the
simple interaction of dimethylzinc with the triazole ring (see
Figure 8). It is important to note that this last complexation
mode is available for every click-supported ligand.
model systems.
to high levels of enantioselectivity (up to 82%). In spite of a
thorough optimization of the structure of the polymeric catalyst
involving resin parameters (functionalization and reticulation)
and ligand parameters (length of a linear hydrocarbon spacer
between the triazole linker and the amino alcohol moiety),
the levels of enantioselectivity previously achieved with the
same amino alcohol anchored on the same resin type by
nucleophilic substitution have not been attained. DFT calcula-
tions on model compounds have shown that complexation of
diorganylzinc species with electron-rich nitrogen atoms in the
triazole is energetically favorable and thus suggest that non-
enantioselective reaction pathways can be available for aryl
transfer.
For future applications the present results should provide
caution on the use of click-supported ligands in connection with
azaphilic cations.12 Alternatively, use of this strategy for
organocatalysis13 or in situations where the triazole linker cannot
likely interact with the involved metal offers clear promise in
view of the ease of ligand design, the ready availability of azido-
functionalized resins, and the efficiency of the cycloaddition
leading to ligand immobilization.
Experimental Section
1,1,2-Triphenyl-2-(4-(prop-2-ynyl)piperazin-1-yl)ethanol (3a).
Amino alcohol 25b (1.00 g, 2.79 mmol) was reacted with propargyl
bromide (0.24 mL, 2.79 mmol) in the presence of Cs2CO3 (1.82 g,
5.58 mmol) in acetonitrile (15 mL) at 55 °C for 39 h. After this
time the reaction mixture was cooled and washed with water (3 ×
5 mL). The organic phase was dried with MgSO4, and the solvent
was removed at reduced pressure. The residue was purified by
column chromatography on silica gel using hexane/ethyl acetate
(4:1) as eluent (Rf ) 0.2). The product was obtained as a white
solid (1.01 g, 91% yield). 1H NMR (400 MHz, CDCl3) δ: 7.68 (d,
2H), 7.29-6.89 (m, 13H), 4.56 (s, 1H), 3.16 (d, 4J ) 2.4 Hz, 2H),
Although the stability of the considered complexes does not
necessarily have to be translated into a parallel order of reactivity
for the alkyl or aryl transfer to an aldehyde (in fact, evolution
of species like 8 after complexation of the reacting aldehyde to
the ring zinc atom should be favored by entropy effects) the
predicted tendency of the diorganylzinc reagent to interact with
the 1,2,3-triazole moiety in click-supported amino alcohol
ligands provides a clue for understanding the limits of enan-
tiocontrol when these ligands are used to promote aryl transfer
from zinc to aldehydes.
2.42 (m, 6H), 2.20 (t, J ) 2.4 Hz, 1H), 2.15 (m, 2H). 13C NMR
4
(100.5 MHz, CDCl3) δ: 149.4, 146.0, 137.5, 131.2, 128.7, 128.3,
128.1, 127.9, 127.6, 127.4, 126.7, 126.6, 125.9, 125.7, 79.0, 78.6,
76.9, 73.3, 52.6, 46.7. IR (ATR) υmax: 3287, 3059, 2802 cm-1
.
Conclusions
HRMS (ESI): m/z calcd for C27H28N2NaO, 419.2099; found,
419.2119 [M + Na]+.
In summary, the Cu(I)-mediated cycloaddition of terminal
alkynes with azides (click chemistry) has been used as an
efficient strategy for immobilization of a chiral, properly
functionalized amino alcohol ligand onto azidomethyl PS-DVB
resins. FTIR spectroscopy can be diagnostically used to follow
the progress of the anchoring reaction, while HRMAS 13C NMR
spectroscopy allows the unambiguous characterization of
the insoluble functional polymers. These polymer-supported
ligands have been used as catalysts in the asymmetric pheny-
lation of aldehydes showing high catalytic activity and moderate
2-(4-(Hex-5-ynyl)piperazin-1-yl)-1,1,2-triphenylethanol (3b).
Amino alcohol 25b (0.5 g ,1.39 mmol) was reacted with 6-chloro-
1-hexyne (0.17 mL, 1.4 mmol) in the presence of Cs2CO3 (1.36 g,
4.2 mmol) in acetonitrile (10 mL) at 70 °C for 68 h. After this
time the reaction mixture was cooled and washed with water (3 ×
5 mL). The organic phase was dried with MgSO4, and the solvent
was removed at reduced pressure. The residue was purified by
column chromatography on silica gel with CH2Cl2/MeOH (9:1) as
the eluent (Rf ) 0.7). The product was obtained as a white solid
(0.64 g, 48% yield). 1H NMR (400 MHz, CDCl3) δ: 7.69 (d, 2H),
J. Org. Chem, Vol. 72, No. 7, 2007 2467

Bastero et al.
7.31-6.90 (m, 13H), 5.57 (br, 1H), 4.58 (s, 1H), 2.43-2.13 (br,
aqueous solution of CuSO4‚5H2O (0.73 mg, 2.94 µmol) and 0.5
mL of an aqueous solution of sodium ascorbate (5.82 mg, 29.4
µmol) were added. The reaction mixture was stirred vigorously for
68 h. The suspended product was filtrated and washed with water.
The crude product was purified by column chromatography on silica
gel, eluting with hexane/ethyl acetate mixtures of increasing polarity
to afford the triazole products (4a or 4b) as white solids.
4
12H), 1.90 (t, J ) 2.1 Hz, 1H), 1.47 (br, 4H). 13C NMR (100.5
MHz, CDCl3) δ: 149.4, 145.9, 137.5, 131.2, 128.6, 128.2, 128.1,
127.8, 127.6, 127.5, 127.2, 126.7, 126.6, 126.5, 125.8, 125.7, 84.4,
78.8, 76.9, 68.6, 57.9, 26.5, 25.9, 18.5. IR (ATR) υmax: 3286, 2945,
2800 cm-1. HRMS (ESI): m/z calcd for C30H35N2O, 439.2749;
found, 439.2762 [M + H]+.
2-(4-(Dec-9-ynyl)piperazin-1-yl)-1,1,2-triphenylethanol (3c).
Amino alcohols 25b (0.5 g, 1.39 mmol) were reacted with 10-bromo-
1-decyne33 (0.36 g, 1.67 mmol) in the presence of Cs2CO3 (0.54 g,
1.67 mmol) and NaI (0.01 g, 0.07 mmol) in acetonitrile (10 mL) at
55 °C for 24 h. After this time the reaction mixture was cooled
and washed with water (3 × 5 mL). The organic phase was dried
with MgSO4, and the solvent was removed at reduced pressure.
The residue was purified by column chromatography using hexane/
ethyl acetate (4:1) as eluent (Rf ) 0.2). The product was obtained
as a white solid (0.39 g, 57% yield). 1H NMR (400 MHz, CDCl3)
δ: 7.70-6.90 (m, 15H), 5.62 (br, 1H), 4.57 (s, 1H), 2.44-2.14
(R)-2-(4-((1-Benzyl-1H-1,2,3-triazol-4-yl)methyl)piperazin-1-
yl)-1,1,2-triphenylethanol (4a): 0.090 g, 59% yield. H NMR
1
(CDCl3) δ: 7.67-6.90 (m, 21H), 5.45 (s, 2H), 4.55 (s, 1H), 3.54
(br, 2H), 2.44-2.08 (m, 8H). 13C NMR (CDCl3) δ: 149.3, 145.9,
144.9, 137.4, 134.8, 131.3, 129.3, 128.9, 128.3, 128.2, 127.7, 127.5,
126.7, 126.5, 125.9, 125.7, 122.6, 78.9, 77.4, 54.3, 53.8, 53.3.
HRMS-ESI: m/z calcd for C34H35N5NaO, 552.2739; found, 552.2750
[M + Na]+.
(R)-2-(4-(4-(1-Benzyl-1H-1,2,3-triazol-4-yl)butyl)piperazin-1-
yl)-1,1,2-triphenylethanol (4b): 0.11 g, 66% yield. 1H NMR
(CDCl3) δ: 7.69-6.88 (m, 21H, 5.45 (s, 2H), 4.55 (s, 1H), 2.63
(t, J ) 7.6 Hz, 2H), 2.39-2.39 (m, 14 H). 13C NMR (CDCl3) δ:
149.4, 148.7, 131.2, 129.2, 128.8, 128.2, 128.1, 127.8, 127.5, 127.2,
126.7, 126.5, 125.8, 125.7, 120.7, 78.8, 77.4, 76.9, 58.3, 54.2, 54.1,
27.4, 26.6, 25.8. HRMS-ESI: m/z calcd for C37H41N5ONa, 594.3209;
found, 594.3183 [M + H]+.
Typical Procedure for Enantioselective Phenyl Transfer to
Aldehydes Catalyzed by Resins 1a-c. A mixture of 293 mg
(1.33 mmol) of ZnPh2 and 333 mg (2.7 mmol) of pure ZnEt2 was
dissolved in 25 mL of anhydrous toluene. Then, the corresponding
weight of resin (according to f) was allowed to swell in 1 mL of
toluene and then reacted with 6.4 mL of the ZnPh2/ZnEt2 solution
under argon for 1 h at room temperature. After cooling to 10 °C,
0.25 mmol of aldehyde was added. After reacting at 10 °C for 2 h
the reaction was quenched with aqueous NH4Cl. The aqueous phase
was extracted with CH2Cl2, the combined organic phases were dried
(MgSO4), and the solvent was eliminated under reduced pressure.
Conversions were determined by 1H NMR, and enantiomeric excess
was determined by HPLC.
4
(br, 10H), 1.90 (t, J ) 2.7 Hz, 1H), 1.56-1.24 (br, 14 H). 13C
NMR (100.5 MHz, CDCl3) δ: 149.2, 145.8, 137.4, 131.1, 128.4,
128.0, 127.9, 127.5, 127.3, 127.0, 126.8, 126.5,126.3, 125.6, 125.5,
78.7, 76.8, 68.0, 58.6, 53.9, 29.4, 28.9, 28.6, 28.4, 27.5, 26.8, 18.4.
IR (ATR) υmax: 3420, 3155, 2927, 2853 cm-1. HRMS (ES+): m/z
calcd for C34H42N2NaO, 517.3195; found, 517.3201 [M + Na]+.
Synthesis of Click Resins (1a-c). The N3-functionalized resin24
was reacted with the corresponding alkyne 3a-c (1.3 equiv), CuI
(0.004 equiv), and DIPEA (0.77 equiv) in a 1/1 mixture of DMF
and THF at 35 °C. The progression of the reaction was monitored
by IR spectroscopy. After disappearance of the azide signal (t )
20-68 h), the resin was filtrated and successively washed with
DMF, water, MeOH, and CH2Cl2 to give yellowish materials. The
resins were dried under vacuum overnight (50 °C). The cycload-
dition yield was calculated from the results of nitrogen elemental
analysis (see ref 23).
Resin 1a (f ) 0.71 mmol/g) was obtained from a N3-PS resin
(f ) 0.97 mmol/g). A 100% yield of functionalization was calculated
on the basis of nitrogen elemental analysis: %Nfound ) 4.95; %Ncalcd
) 4.91. 13C NMR (gel, 125. 7 MHz, C6D6) δ: 150.0, 146.6, 145.7,
138.2, 131.4, 127.3, 127.0, 126.6, 126.1, 79.5, 76.8, 53.9, 41.0.
Resin 1b (f ) 0.72 mmol/g) was obtained from a N3-PS resin
(f ) 1.03 mmol/g). A 100% yield of functionalization was calculated
on the basis of nitrogen elemental analysis: %Nfound ) 5.01; %Ncalcd
) 4.97. 13C NMR (gel, 125. 7 MHz, C6D6) δ: 150.3, 146.8, 138.5,
131.5, 128.3, 127.4, 127.0, 126.6, 126.1, 77.0, 71.1, 58.2, 54.1,
41.1, 27.7, 26.9, 26.2.
Acknowledgment. WethankDGI-MCYT(grantno.CTQ2005-
02193/BQU), DURSI (grant no. 2005SGR225), Consolider
Ingenio 2010 (Grant No. CSD2006-0003), and the ICIQ
Foundation for financial support. A.B. is grateful to the Spanish
Ministerio de Educacio´n y Ciencia for a “Juan de la Cierva”
research contract and D.F. for a predoctoral fellowship. We also
thank G. Gonza´lez and K. Go´mez (ICIQ support unit) for their
help with HRMAS 13C NMR measurements.
Resin 1c (f ) 0.67 mmol/g) was obtained from a N3-PS resin
(f ) 1.04 mmol/g). A 97.5% yield of functionalization was
calculated on the basis of nitrogen elemental analysis: %Nfound
)
Supporting Information Available: General experimental
methods, conditions for HPLC analysis, and NMR spectra of new
compounds; Cartesian coordinates and energies for the DFT-
calculated compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
4.69; %Ncalcd ) 4.81. 13C NMR (gel, 125. 7 MHz, C6D6) δ: 150.2,
146.7, 138.4, 131.4, 128.3, 127.3, 126.9, 126.5, 126.0, 125.0, 79.3,
76.9, 58.5, 54.1, 41.0, 29.9, 27.8, 27.2, 26.3.
Synthesis of Model Compounds (4a, 4b). To a mixture of the
corresponding alkyne 3a or 3b (0.29 mmol) and benzyl azide
(0.037 µL, 0.29 mmol) in tert-butyl alcohol (1 mL), 0.5 mL of an
JO0624952
2468 J. Org. Chem., Vol. 72, No. 7, 2007