Resin-supported catalysts for CuAAC click reactions in aqueous or organic solvents

Article abstract of DOI:10.1021/co300076k

The copper-catalyzed azide-alkyne cycloaddition click reaction is a valuable process for the synthesis of libraries of drug candidates, derivatized polymers and materials, and a wide variety of other functional molecules. In some circumstances, the removal of the copper catalyst is both necessary and inconvenient. We describe here two immobilized forms of a Cu-binding ligand that has been shown to accelerate triazole formation under many different conditions, using different resin supports that are appropriate for aqueous or organic solvents. Copper leaching from these resins was modest, allowing them to be reused in many reaction/filtration cycles without recharging with metal ion. The utility of this catalyst form was demonstrated in the convenient synthesis of 20 N-acetylgalactosamine derivatives for biological testing.

Full text of DOI:10.1021/co300076k

Letter
pubs.acs.org/acscombsci
Resin-Supported Catalysts for CuAAC Click Reactions in Aqueous or
Organic Solvents
Stanislav I. Presolski, Sreeman K. Mamidyala, Florian Manzenrieder, and M.G. Finn*
Department of Chemistry, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037, United States
S
* Supporting Information
ABSTRACT: The copper-catalyzed azide−alkyne cycloaddi-
tion click reaction is a valuable process for the synthesis of
libraries of drug candidates, derivatized polymers and
materials, and a wide variety of other functional molecules.
In some circumstances, the removal of the copper catalyst is
both necessary and inconvenient. We describe here two
immobilized forms of a Cu-binding ligand that has been shown
to accelerate triazole formation under many different
conditions, using different resin supports that are appropriate
for aqueous or organic solvents. Copper leaching from these
resins was modest, allowing them to be reused in many
reaction/filtration cycles without recharging with metal ion.
The utility of this catalyst form was demonstrated in the convenient synthesis of 20 N-acetylgalactosamine derivatives for
biological testing.
KEYWORDS: resin-supported catalysts, CuAAC click reactions, copper catalyst
chelating ligand is covalently attached to a resin material
commonly used for solid-phase organic and peptide synthesis.8
We have described a highly versatile mono-
(benzimidazoylmethyl)-bis(pyridylmethyl) amine ligand (des-
ignated BimPy2) that performs well in both aqueous and
coordinating organic solvents.6 Its ease of synthesis and readily
functionalized single attachment point made it attractive for
immobilization, which we performed on two different solid
supports as shown in Figure 1. The undecyl alcohol BimPy2
derivative 1 was attached to standard 2-chlorotrityl polystyrene
resin (1% divinylbenzene cross-linked, available from several
suppliers, 1.04 mmol/g in the exemplary case here, but a variety
of loadings between 0.5 and 2 mmol/g work equally well),
followed by capping with methanol, to provide the immobilized
PS-BimPy2. The mass increase of the resin beads indicated a
loading of 290 μmol/g. Treatment of the resin with
trifluoroacetic acid to cleave the trityl connector, followed by
MS analysis, revealed that only the desired ligand had been
attached. A negative control resin (PS-OMe) was prepared
using methanol in place of 1. To provide immobilized catalyst
capability for polar and hydrophilic media, the BimPy2 ligand
was also attached to Nova-PEG-amine resin (Novabiochem no.
855126) via an 11-carbon aldehyde (2) by reductive amination,
to give the NovaPEG-BimPy2 ligand. The unreacted primary
amines and the secondary amines resulting from the coupling
were capped with acetic anhydride. A loading of 230 μmol/g
he copper-catalyzed azide−alkyne cycloaddition
1,2
T(CuAAC) reaction
is well suited to many solution-
phase applications with the simple use of Cu(I) salts, either
added directly or generated by reduction of Cu(II) or oxidation
of Cu(0). This process is approximately one million times faster
than the very sluggish uncatalyzed [3 + 2] cycloaddition with
unactivated alkynes and azides.3 When even faster reaction rates
are needed, certain heterocyclic chelating agents have been
shown to provide strong ligand-accelerated catalysis.4−6 The
intrinsically tolerant nature of Cu catalysts to variations in
reaction conditions has allowed the CuAAC process to be
applied in an extraordinarily wide range of situations.
Reliable methods for the removal of the catalyst from the
reaction mixture are desirable. In the case of CuAAC, which is
commonly applied to the synthesis of materials intended for
biological application, this can be of great concern because of
the perception that copper ions are toxic to cells and organisms.
In most cases, we find that washing with or dialysis against
solutions of EDTA removes even small traces of copper ions
from CuAAC reaction mixtures and isolated products.
However, this procedure is not always convenient, especially
in parallel or combinatorial synthesis.
A common strategy is to immobilize the catalyst on a support
that allows access to reactants and can be easily separated from
the reaction mixture upon completion. A handful of reports
have appeared of heterogeneous CuAAC catalysts, but this field
is not yet fully developed.7−17 (See Supporting Information for
an annotated list of these contributions.) We describe here the
synthesis and reactivity of two versions of a system first
described by Chan and Fokin, in which an accelerating
Received: July 12, 2012
Revised: August 18, 2012
Published: September 4, 2012
© 2012 American Chemical Society
527
dx.doi.org/10.1021/co300076k | ACS Comb. Sci. 2012, 14, 527−530

ACS Combinatorial Science
Letter
Figure 1. Synthesis of resin-supported BimPy2 CuAAC-accelerating ligands.
Figure 2. Screening of activity of resin-bound ligands for the model CuAAC reaction shown. (A, B) Amounts of triazole 3 formed (solid circles) and
amount of Cu remaining on the resin (open circles) in sequential cycles of CuAAC reaction using 10 mg (A) PS-BimPy2 or PS-OMe with an initial
charge of 100 μM CuSO4 or (B) NovaPEG-BimPy2 or NovaPEG-Cin with initial charge of 1 mM CuSO4. Reaction conditions: (A) 1.25 mM
BnN3, 2.5 mM PhCCH, 10 mM Na ascorbate, 95:5 DMF:H2O, RT, 12 h; (B) 100 mM BnN3, 100 mM PhCCH, 20 mM Na ascorbate, 90:10
MeOH:H2O, RT, 12 h. Each reaction was drained, washed three times with methanol, air-dried, and recharged with azide, alkyne, and ascorbate to
initiate the next cycle. (C) Time course for reactions mediated by 10 mg of the indicated NovaPEG derivatives; conditions as in the first reaction of
plot (B), except for the omission of resin or CuSO4 as noted.
was indicated by the mass increase from the attachment
reaction, representing occupancy of approximately 30% of the
resin amine groups. Reductive amination with cinnamaldehyde
provide the negative control material in this case (NovaPEG-
Cin). The undecyl hydrocarbon tether was selected in order to
allow the ligand enhanced mobility in the swollen resin (gel)
state, but other chain lengths and types would likely also be
effective. We also explored Sepharose-epoxide as a water-
swellable solid support, but the attachment yields were low and
the material had insufficient mechanical stability to stand up to
the agitation required in repeat reactions.
The polystyrene-based catalysts were initially tested by
adding CuSO4, sodium ascorbate, and small amounts of resin
(10 mg) to CuAAC reactions in 1.5 mL of 95:5 DMF:H2O, a
solvent system that effectively swells the resin. Yields of triazole
3 were determined the following day by HPLC-MS against a
quantitative calibration curve (Figure 2). The reactions were
performed in fritted syringes, allowing for easy separation of the
solution from the solid-phase material. For this length of
528
dx.doi.org/10.1021/co300076k | ACS Comb. Sci. 2012, 14, 527−530

ACS Combinatorial Science
Letter
reaction time, the ligand-free reaction produced relatively high
yield but failed to reach completion, presumably because of
competitive oxidative inactivation of the Cu(I) catalyst. In
contrast, the ligand-bearing resin provided complete reaction,
reflecting an accelerated reaction rate.
azides and alkynes each, including aromatic, aliphatic, and
benzylic examples (Figure 3).19 All triazole products were
isolated in >90% yields after filtration and flash column
1
chromatography, in purities >95% as judged by H NMR.
After removal of the reaction solvent, each of these resins was
recharged with another batch of azide, alkyne, and sodium
ascorbate, but no additional copper source. Only those resins
bearing the BimPy2 ligand produced triazole, showing that the
ligand is required to retain the catalytic metal center. The PS-
BimPy2 catalyst was subjected to 13 cycles of CuAAC reaction,
filtration, and recharge with azide, alkyne, and ascorbate. The
Cu content of the filtered solutions was measured after each
cycle by inductively coupled plasma atomic emission spectros-
copy (ICP-AES). As shown in Figure 2A, all of these reactions
went to completion, and an average of 3% of the Cu was lost in
each cycle (approximately 3 μM of Cu in the solution), leaving
more than 60% of the initial Cu retained on the resin after the
13th reaction. Note that low-micromolar concentrations of
Cu(I) without accelerating ligand do not catalyze the CuAAC
reaction under these conditions (data not shown).
When the solvent was changed from DMF to methanol, the
overnight reactions catalyzed by the polystyrene-immobilized
ligand did not reach completion, probably because of the poor
swelling capacity of the solid support. The ligand-modified
NovaPEG resin served well in this case, being swelled in
methanol (and a wide variety of other solvents) and mediating
complete formation of triazole in two consecutive reactions
(Figure 2B). In the three subsequent runs, the yields were
about 85%, after which performance declined to a greater
degree without Cu recharging. Under these “synthetic”
conditions (100 mM concentrations of substrates, organic
solvent), the greatest amount of Cu contamination was
observed, with approximately 400 μM Cu observed in solution
after the first reaction. This is likely due in part to the high
concentration of triazole formed in this reaction, which
competes for Cu with the immobilized ligand. The
NovaPEG-Cin resin initially induced modest levels of triazole
formation (27% yield with 1 mol % Cu), presumably because of
some acceleration provided by adventitious Cu ligands such as
the amide groups installed on the resin. Following the course of
the NovaPEG-based reactions more closely revealed supported
BimPy2 to be far more active than the control acylated
cinnamoylamine-bearing resin (Figure 2C). When no resin was
employed, the reaction gave only a 5% yield of triazole before
exhaustion of the reducing agent needed to support CuAAC.
The NovaPEG resin has excellent swelling capacity in both
dichloromethane and water. In the former solvent, the use of
CuBr and CuI gave incomplete reactions that contained
substantial quantities of copper, suggesting that counterion
substitution is important in noncoordinating solvents.18 In
aqueous reactions performed under typical bioconjugation
conditions, however, 10 mg of NovaPEG-BimPy2 in 90% H2O,
10% DMSO with 100 μM CuSO4, and 5 mM NaAsc catalyzed
the complete conversion of 100 μM BnN3 to triazole, when 200
μM PhCCH were used. Low levels of Cu were observed in
solution (approximately 5.4 μM, 0.34 ppm, in each cycle). Such
minimal leaching makes the resin-immobilized catalyst an
excellent choice for library syntheses of compounds that can be
used in many biological tests without concern for significant
copper contamination.
Figure 3. Parallel CuAAC reactions performed with NovaPEG-
BimPy2 catalyst.
We demonstrate here that the resin immobilization of the
BimPy2 ligand provides readily accessible and useful catalysts
for the CuAAC reaction in both organic and aqueous solvents.
These systems show convenient properties of rate, metal
retention, and separability from soluble reaction components.
The previously reported and very useful resin of Chan and
Fokin8 employs the tris(triazolylmethyl)amine ligand, which is
a weaker binder of CuI than BimPy2.6 For this reason, this
ligand is not well suited to reactions in strongly coordinating
solvents. If the user intends to perform reactions in DMSO,
DMF, or similar solvents, PS-BimPy2 is an effective and
inexpensive catalyst. When polar protic and aqueous solvent
systems are desired, we recommend NovaPEG-BimPy2 due to
its superior swelling properties. For all other solvents except
acetonitrile (which gives poor results with this class of catalysts;
see Figure 5 of reference 6), both ligand-functionalized resins
should provide good results if the solvent swells the resin to a
sufficient degree.
ASSOCIATED CONTENT
* Supporting Information
■
S
Table showing heterogeneous or supported CuAAC catalysts
with typical reaction conditions, description of the synthesis of
resin-supported ligands, screening of resin−ligand activity, and
use of NovaPEG-BimPy2 in synthesis, and characterization data
The NovaPEG-BimPy2 ligand was tested in preparatory-
scale reactions involving glycan precursors 4 and 6 with ten
529
dx.doi.org/10.1021/co300076k | ACS Comb. Sci. 2012, 14, 527−530

ACS Combinatorial Science
Letter
(16) Buckley, B. R.; Dann, S. E.; Harris, D. P.; Heaney, H.; Stubbs, E.
C. Alkynylcopper(I) polymers and their use in a mechanistic study of
alkyne−azide click reactions. Chem. Commun. 2010, 46, 2274−2276.
(17) He, Y.; Bian, Z.; Kang, C.; Cheng, Y.; Gao, L. Chiral
binaphthylbisbipyridine-based copper(I) coordination polymer gels as
supramolecular catalysts. Chem. Commun. 2010, 46, 3532−3534.
(18) Hein, J. E.; Tripp, J. C.; Krasnova, L. B.; Sharpless, K. B.; Fokin,
V. V. Copper(I)-catalyzed cycloaddition of organic azides and 1-
iodoalkynes. Angew. Chem., Int. Ed. 2009, 48, 8018−8021.
including NMR spectra. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: mgfinn@scripps.edu.
■
Funding
This work was supported by the NIH (R56 AI090832), The
Skaggs Institute for Chemical Biology, and Pfizer Global R&D.
Notes
́
(19) Mamidyala, S. K.; Dutta, S.; Chrunyk, B. A.; Preville, C.; Wang,
C.; Withka, J. M.; McColl, A.; Subashi, T. A.; Hawrylik, S. J.; Griffor,
M. C.; Kim, S.; Pfefferkorn, J. A.; Price, D. A.; Menhaji-Klotz, E.;
Mascitti, V.; Finn, M. G. Glycomimetic ligands for the human
asialoglycoprotein receptor. J. Am. Chem. Soc. 2012, 134, 1978−1981.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank Dr. Vu Hong for helpful discussions.
■
REFERENCES
■
(1) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. A
stepwise huisgen cycloaddition process: Copper(I)-catalyzed regiose-
lective ligation of azides and terminal alkynes. Angew. Chem., Int. Ed.
2002, 41, 2596−2599.
(2) Tornøe, C. W.; Christensen, C.; Meldal, M. Peptidotriazoles on
solid phase: [1,2,3]-Triazoles by regiospecific copper(I)-catalyzed 1,3-
dipolar cycloadditions of terminal alkynes to azides. J. Org. Chem.
2002, 67, 3057−3062.
(3) Lewis, W. G.; Green, L. G.; Grynszpan, F.; Radic, Z.; Carlier, P.
R.; Taylor, P.; Finn, M. G.; Sharpless, K. B. Click chemistry in situ:
Acetylcholinesterase as a reaction vessel for the selective assembly of a
femtomolar inhibitor from an array of building blocks. Angew. Chem.,
Int. Ed. 2002, 41, 1053−1057.
(4) Chan, T. R.; Hilgraf, R.; Sharpless, K. B.; Fokin, V. V.
Polytriazoles as copper(I)-stabilizing ligands in catalysis. Org. Lett.
2004, 6, 2853−2855.
(5) Rodionov, V. O.; Presolski, S.; Díaz, D. D.; Fokin, V. V.; Finn, M.
G. Ligand-accelerated Cu-catalyzed azide−alkyne cycloaddition: A
mechanistic report. J. Am. Chem. Soc. 2007, 129, 12705−12712 and
references therein..
(6) Presolski, S.; Hong, V.; Cho, S.-H.; Finn, M. G. Tailored ligand
acceleration of the Cu-catalyzed azide−alkyne cycloaddition reaction.
J. Am. Chem. Soc. 2010, 132, 14570−14576.
(7) Lipshutz, B. H.; Taft, B. R. Heterogeneous copper-in-charcoal-
catalyzed click chemistry. Angew. Chem., Int. Ed. 2006, 45, 8235−8238.
(8) Chan, T. R.; Fokin, V. V. Polymer-supported copper(I) catalysts
for the experimentally simplified azide−alkyne cycloaddition. QSAR
Comb. Sci. 2007, 26, 1274−1279.
(9) Li, P.; Wang, L.; Zhang, Y. SiO2−NHC−Cu(I): An efficient and
reusable catalyst for [3 + 2] cycloaddition of organic azides and
terminal alkynes under solvent-free reaction conditions at room
temperature. Tetrahedron 2008, 64, 10825−10830.
(10) Dervaux, B.; Van Camp, W.; Du Prez, F. E. Cryogels as novel
solid-supported copper catalyst system for click chemistry and their
use in column reactors. Polym. Prepr. 2008, 49, 692−693.
(11) Jlalia, I.; Elamari, H.; Meganem, F.; Herscovici, J.; Girard, C.
Copper(I)-doped Wyoming’s montmorillonite for the synthesis of
disubstituted 1,2,3-triazoles. Tetrahedron Lett. 2008, 49, 6756−6758.
(12) Namitharan, K.; Kumarraja, M.; Pitchumani, K. CuII-Hydro-
talcite as an efficient heterogeneous catalyst for Huisgen [3 + 2]
cycloaddition. Chem.Eur. J. 2009, 15, 2755−2758.
(13) Chtchigrovsky, I.; Primo, A.; Gonzalez, P.; Molvinger, K.;
Robitzer, M.; Quignard, F.; Taran, F. Functionalized chitosan as a
green, recyclable, biopolymer-supported catalyst for the [3 + 2]
Huisgen cycloaddition. Angew. Chem., Int. Ed. 2009, 48, 5916−5920.
(14) Jlalia, I.; Meganem, F.; Herscovici, J.; Girard, C. “Flash” solvent-
free synthesis of triazoles using a supported catalyst. Molecules 2009,
14, 528−539.
(15) Bonami, L.; Van Camp, W.; Van Rijckegem, D.; Du Prez, F. E.
Facile access to an efficient solid-supported click catalyst system based
on poly(ethyleneimine). Macromol. Rapid Commun. 2009, 30, 34−38.
530
dx.doi.org/10.1021/co300076k | ACS Comb. Sci. 2012, 14, 527−530