Strain-promoted alkyne azide cycloaddition for the functionalization of poly(amide)-based dendrons and dendrimers

Article abstract of DOI:10.1021/ja910581d

Functionalization of a poly(amido)-based dendron with ethylene glycol chains (PEG) using coppercatalyzed alkyne azide cycloaddition (CuAAC) afforded dendrons with significant levels of copper contaminations, preventing the use of such materials for biological applications. We suggest that the presence of amide, PEG, and triazole functional groups allows for copper complexation, thereby preventing the separation of the copper catalyst from the final dendron. To minimize this problem, synthetic variations on CuAAC including the addition of "click" additives for copper sequestering as well as the use of copper wire as the copper source were investigated. None of these strategies, however, resulted in copper-free products. In contrast, we developed a copper-free strain-promoted alkyne azide cycloaddition (SPAAC) strategy that functionalized poly(amide)-based dendrons and dendrimers with PEG chains quantitatively under mild reaction conditions without any metal contamination. The SPAAC products were characterized by ¹H and ¹³C NMR, 2D HSQC and COSY NMR, mass spectrometry, and elemental analysis. This is the first report on the use of SPAAC for dendrimer functionalization, and the results obtained here show that SPAAC is an important tool to the dendrimer and more general biomaterials community for the functionalization of macromolecular structures due to the mild and metal-free reaction conditions, no side products, tolerance toward functional groups, and high yields.

Full text of DOI:10.1021/ja910581d

Published on Web 02/25/2010
Strain-Promoted Alkyne Azide Cycloaddition for the
Functionalization of Poly(amide)-Based Dendrons and
Dendrimers
Ca´tia Ornelas, Johannes Broichhagen, and Marcus Weck*
Molecular Design Institute and Department of Chemistry, New York UniVersity,
New York, New York 10003-6688
Received December 15, 2009; E-mail: marcus.weck@nyu.edu
Abstract: Functionalization of a poly(amido)-based dendron with ethylene glycol chains (PEG) using copper-
catalyzed alkyne azide cycloaddition (CuAAC) afforded dendrons with significant levels of copper
contaminations, preventing the use of such materials for biological applications. We suggest that the
presence of amide, PEG, and triazole functional groups allows for copper complexation, thereby preventing
the separation of the copper catalyst from the final dendron. To minimize this problem, synthetic variations
on CuAAC including the addition of “click” additives for copper sequestering as well as the use of copper
wire as the copper source were investigated. None of these strategies, however, resulted in copper-free
products. In contrast, we developed a copper-free strain-promoted alkyne azide cycloaddition (SPAAC)
strategy that functionalized poly(amide)-based dendrons and dendrimers with PEG chains quantitatively
under mild reaction conditions without any metal contamination. The SPAAC products were characterized
1
by H and 13C NMR, 2D HSQC and COSY NMR, mass spectrometry, and elemental analysis. This is the
first report on the use of SPAAC for dendrimer functionalization, and the results obtained here show that
SPAAC is an important tool to the dendrimer and more general biomaterials community for the
functionalization of macromolecular structures due to the mild and metal-free reaction conditions, no side
products, tolerance toward functional groups, and high yields.
Introduction
dendrimers have a number of advantages including: (i) the
controlled multivalency of dendrimers that can be used to attach
Over the past decades, a variety of drug delivery carriers
including polymer microcapsules, liposomes, and nanoparticles
have been explored for biomedical applications such as drug
delivery, gene delivery, and in vivo imaging.1-7 Among these
materials, dendrimers have emerged as important candidates for
these applications.8-16 In comparison to linear polymers,
several drug molecules, targeting groups and solubilizing
moieties; (ii) the low polydispersity, which provides reproducible
pharmacokinetic behavior; and (iii) the globular shape of higher
generation dendrimers, which has the potential to affect their
biological properties.17
Since the first iteration of a Michael addition followed by
nitrile reduction demonstrated by Vo¨gtle in 1979,18 function-
alization of dendrimers was successfully achieved using this19
and other synthetic methodologies including amide and ester
couplings,20 Williamson reactions,21,22 and cross-metathesis.23-25
(1) Fox, M. E.; Szoka, F. C.; Fre´chet, J. M. J. Acc. Chem. Res. 2009, 42,
1141–1151.
(2) Daniel, M. C.; Astruc, D. Chem. ReV. 2004, 104, 293–346.
(3) van Dongen, S. F. M.; de Hoog, H.-P. M.; Peters, R. J. R. W.; Nallani,
M.; Nolte, R. J. M.; van Hest, J. C. M. Chem. ReV. 2009, 109, 6212–
6274.
(4) Debbage, P. Curr. Pharm. Des. 2009, 15, 153–172.
(5) Biomedical Nanostructures; Gonsalves, K. E., C. R. H., Laurencin,
C. T., Nair, L. S., Eds.; John Wiley & Sons, Inc.: New York, 2008.
(6) Torchilin, V. P. AdV. Drug DeliVery ReV. 2006, 58, 1532–1555.
(7) Uhrich, K. E.; Cannizzaro, S. M.; Langer, R. S.; Shakesheff, K. M.
Chem. ReV. 1999, 99, 3181–3198.
(16) Medina, S. H.; El-Sayed, M. E. H. Chem. ReV. 2009, 109, 3141–
3157.
(17) Gillies, E. R.; Fre´chet, J. M. J. Drug DiscoVery Today 2005, 10, 35–
43.
(18) Vo¨gtle, F.; Weber, E. Angew. Chem., Int. Ed. Engl. 1979, 18, 753–
776.
(8) Helms, B.; Meijer, E. W. Science 2006, 313, 929–930.
(9) Esfand, R.; Tomalia, D. A. Drug DiscoVery Today 2001, 6, 427–436.
(10) Lee, C. C.; MacKay, J. A.; Frechet, J. M. J.; Szoka, F. C. Nat.
Biotechnol. 2005, 23, 1517–1526.
(19) Tomalia, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos, G.; Martin,
S.; Roeck, J.; Ryder, J.; Smith, P. Macromolecules 1986, 19, 2466–
2468.
(20) Newkome, G. R.; Kim, H. J.; Moorefield, C. N.; Maddi, H.; Yoo,
K. S. Macromolecules 2003, 36, 4345–4354.
(11) Svenson, S.; Tomalia, D. A. AdV. Drug DeliVery ReV. 2005, 57, 2106–
2129.
(21) Hawker, C.; Fre´chet, J. M. J. J. Chem. Soc., Chem. Commun. 1990,
1010, 1013.
(12) Majoros, I. J.; Williams, C. R.; Baker, J. R. Curr. Top. Med. Chem.
2008, 8, 1165–1179.
(22) Ruiz, J.; Lafuente, G.; Marcen, S.; Ornelas, C.; Lazare, S.; Cloutet,
E.; Blais, J. C.; Astruc, D. J. Am. Chem. Soc. 2003, 125, 7250–7257.
(23) Liang, C. O.; Fre´chet, J. M. J. Macromolecules 2005, 38, 6276–6284.
(24) Ornelas, C.; Mery, D.; Blais, J. C.; Cloutet, E.; Aranzaes, J. R.; Astruc,
D. Angew. Chem., Int. Ed. 2005, 44, 7399–7404.
(25) Ornelas, C.; Mery, D.; Cloutet, E.; Aranzaes, J. R.; Astruc, D. J. Am.
Chem. Soc. 2008, 130, 1495–1506.
(13) Agarwal, A.; Asthana, A.; Gupta, U.; Jain, N. K. J. Pharm. Pharmacol.
2008, 60, 671–688.
(14) Wolinsky, J. B.; Grinstaff, M. W. AdV. Drug DeliVery ReV. 2008, 60,
1037–1055.
(15) Tekade, R. K.; Kumar, P. V.; Jain, N. K. Chem. ReV. 2009, 109, 49–
87.
9
10.1021/ja910581d  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 3923–3931 3923

A R T I C L E S
Ornelas et al.
None of these reactions, however, can be carried out easily in
vitro or in vivo. Furthermore, in some cases, nonquantitative
functionalization steps as well as functional group compatibility
have limited the success of these strategies. One synthetic
methodology that has the potential to overcome these limitations
is the use of “click chemistry”, which was introduced by
Sharpless and co-workers in 2001.26 Click reactions have to
meet several criteria including high yields, tolerance toward
functional groups, and virtually no side-reactions. Among the
“click” reactions, the Cu(I)-catalyzed alkyne azide 1,3-dipolar
cycloaddition (CuAAC) is the most popular.27 CuAAC was
successfully applied in different areas of materials chemistry28
including, but not limited to, polymers,29 nanoparticles,30
interlocked molecules,31-33 and dendrimers.34 In dendrimer
chemistry,CuAACwasusedfortheconvergent35 anddivergent36,37
synthesesofdendrimersaswellasintheirfunctionalization.34,38-43
The cytotoxicity of copper, however, is a potential drawback
of the CuAAC reaction for the synthesis and/or functionalization
of nanomaterials such as dendrimers for biomedical applications.
In cases where the final structure contains numerous functional
groups able to bind copper ions, the removal of the copper
catalyst can be problematic, limiting the use of copper-
contaminateddrugdeliverycarriersforpharmaceuticalpurposes.44,45
Therefore, copper-free “click” strategies that combine the
advantages of CuAAC without the use of a toxic transition metal
would be highly desirable for biomedical applications.44
Recently, an interesting copper-free reaction based on strained
cyclooctynes and azides was developed by Bertozzi and co-
workers for in vivo applications.46-48 This reaction goes back
to the late work of Georg Wittig who described the exothermic
cycloaddition of cyclooctyne with phenyl azide leading to the
corresponding triazole.49 The Bertozzi group has applied this
strain-promoted alkyne azide cycloaddition (SPAAC) in covalent
modifications of living systems50 and in vivo imaging of
membrane-associated glycans in developing Zebrafish using a
multicolor detection strategy.51 This reaction was then applied
by other groups to label peptides52 and lipids,53 to cross-linked
hydrogels,54 photodegradable star polymers,55 and to function-
alize polymers.56,57
In this contribution, we describe the limitation of copper-
based 1,3-dipolar cycloaddition functionalization strategies in
poly(amide)-based dendrons and present SPAAC as an efficient
copper-free strategy to functionalize dendrons and dendrimers.
We utilize poly(amido)-based dendrons that are known to be
biocompatible and therefore are an interesting starting point for
novel biomaterials. Furthermore, they can be multifunctionalized
through the use of either bifunctional dendrons40 or a trifunc-
tional core,58 bringing us closer to our long-term goal, the
construction of well-defined dendrimers for theranostics that
require several functionalities including (i) hydrophilic groups
such as PEG chains to increase water solubility and biocom-
patibility, (ii) imaging agents to monitor the trajectory of the
dendrimer in vitro or/and in vivo, (iii) targeting groups such as
folic acid, biotin, or specific antibodies, and (iv) the desired
drugs. Herein, we report the use of SPAAC as a highly
advantageous and efficient functionalization strategy toward
multifunctional dendrimers.
Results and Discussion
Synthesis of Poly(amide)-Based Dendrons and Dendrimers
with Azide Termini. The key to the development of a 1,3-dipolar
cycloaddition-based functionalization strategy in dendrimer
chemistry is the synthesis of the azide-containing dendrons and
dendrimers.59 We employed poly(amide)-based dendrons that
follow the 1f3 connectivity pioneered by Newkome.60 Den-
drons containing three and nine azide termini and a dendrimer
containing 18 azide termini were synthesized in high yields from
commercial dendrons nitrotriester 1 and aminotriester 5 as
shown in Scheme 1.
(26) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed.
2001, 40, 2004–2021.
(27) Meldal, M.; Tornoe, C. W. Chem. ReV. 2008, 108, 2952–3015.
(28) Iha, R. K.; Wooley, K. L.; Nystro¨m, A. M.; Burke, D. J.; Kade, M. J.;
Hawker, C. J. Chem. ReV. 2009, 109, 5620–5686.
Our strategy is based on building block 1. The deprotection
of the acid groups on 1 was carried out by stirring 1 in formic
acid at room temperature for 24 h, resulting in dendron triacid
(29) Meldal, M. Macromol. Rapid Commun. 2008, 29, 1016–1051.
(30) Boisselier, E.; Diallo, A. K.; Salmon, L.; Ruiz, J.; Astruc, D. Chem.
Commun. 2008, 4819–4821.
(31) Megiatto, J. D.; Schuster, D. I. J. Am. Chem. Soc. 2008, 130, 12872–
12873.
(32) Megiatto, J. D.; Schuster, D. I. Chem.-Eur. J. 2009, 15, 5444–5448.
(33) Dichtel, W. R.; Miljanic, O. S.; Spruell, J. M.; Heath, J. R.; Stoddart,
J. F. J. Am. Chem. Soc. 2006, 128, 10388–10390.
(47) Baskin, J. M.; Prescher, J. A.; Laughlin, S. T.; Agard, N. J.; Chang,
P. V.; Miller, I. A.; Lo, A.; Codelli, J. A.; Bertozzi, C. R. Proc. Natl.
Acad. Sci. U.S.A. 2007, 140, 16793–16797.
(34) Franc, G.; Kakkar, A. Chem. Commun. 2008, 5267–5276.
(35) Wu, P.; Feldman, A. K.; Nugent, A. K.; Hawker, C. J.; Scheel, A.;
Voit, B.; Pyun, J.; Fre´chet, J. M. J.; Sharpless, K. B.; Fokin, V. V.
Angew. Chem., Int. Ed. 2004, 43, 3928–3932.
(48) Sletten, E. M.; Bertozzi, C. R. Angew. Chem., Int. Ed. 2009, 48, 6974–
6998.
(49) Wittig, G.; Krebs, A. Chem. Ber.-Recl. 1961, 94, 3260–3275.
(50) Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. J. Am. Chem. Soc. 2004,
126, 15046–15047.
(36) Ornelas, C.; Aranzaes, J. R.; Cloutet, E.; Alves, S.; Astruc, D. Angew.
Chem., Int. Ed. 2007, 46, 872–877.
(51) Laughlin, S. T.; Baskin, J. M.; Amacher, S. L.; Bertozzi, C. R. Science
2008, 320, 664–667.
(37) Joralemon, M. J.; O’Reilly, R. K.; Matson, J. B.; Nugent, A. K.;
Hawker, C. J.; Wooley, K. L. Macromolecules 2005, 38, 5436–5443.
(38) Wu, P.; Malkoch, M.; Hunt, J. N.; Vestberg, R.; Kaltgrad, E.; Finn,
M. G.; Fokin, V. V.; Sharpless, K. B.; Hawker, C. J. Chem. Commun.
2005, 5775–5777.
(52) Kele, P.; Mezo, G.; Achatz, D.; Wolfbeis, O. S. Angew. Chem., Int.
Ed. 2009, 48, 344–347.
(53) Neef, A. B.; Schultz, C. Angew. Chem., Int. Ed. 2009, 48, 1498–1500.
(54) Meredith, C.; Patrick, K. Polym. Int. 2009, 58, 1190–1195.
(55) Johnson, J. A.; Baskin, J. M.; Bertozzi, C. R.; Koberstein, J. T.; Turro,
N. J. Chem. Commun. 2008, 3064–3066.
(39) Yoon, K.; Goyal, P.; Weck, M. Org. Lett. 2007, 9, 2051–2054.
(40) Goyal, P.; Yoon, K.; Weck, M. Chem.-Eur. J. 2007, 13, 8801–8810.
(41) Ornelas, C.; Aranzaes, J. R.; Salmon, L.; Astruc, D. Chem.-Eur. J.
2008, 14, 50–64.
(56) Lallana, E.; Fernandez-Megia, E.; Riguera, R. J. Am. Chem. Soc. 2009,
131, 5748–5750.
(42) Srinivasachari, S.; Fichter, K. M.; Reineke, T. M. J. Am. Chem. Soc.
2008, 130, 4618–4627.
(57) Canalle, L. A.; Berkel, S. S. v.; Haan, L. T. d.; Hest, J. C. M. V. AdV.
Funct. Mater. 2009, 19, 3464–3470.
(43) Malkoch, M.; Schleicher, K.; Drockenmuller, E.; Hawker, C. J.;
Russell, T. P.; Wu, P.; Fokin, V. V. Macromolecules 2005, 38, 3663–
3678.
(58) Ornelas, C.; Weck, M. Chem. Commun. 2009, 5710–5712.
(59) Molecules containing azide groups are potentially explosive and should
be handled with care. For more details about the sensitivity of these
materials, please see: Bra¨se, S.; Gil, C.; Knepper, K.; Zimmermann,
V. Angew. Chem., Int. Ed. 2005, 44, 5188–5240.
(44) Lutz, J. F. Angew. Chem., Int. Ed. 2008, 47, 2182–2184.
(45) van Dijk, M.; Rijkers, D. T. S.; Liskamp, R. M. J.; van Nostrum,
C. F.; Hennink, W. E. Bioconjugate Chem. 2009, 20, ????.
(46) Agard, N. J.; Baskin, J. M.; Prescher, J. A.; Lo, A.; Bertozzi, C. R.
ACS Chem. Biol. 2006, 1, 644–648.
(60) Newkome, G. R.; Moorefield, C. N.; Baker, G. R.; Behera, R. K.;
Escamillia, G. H.; Saunders, M. J. Angew. Chem., Int. Ed. Engl. 1992,
31, 917–919.
9
3924 J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010

Strain-Promoted Alkyne Azide Cycloaddition
A R T I C L E S
Scheme 1. Synthesis of Dendrons Tris-azide 4 and Nona-azide 8, and Dendrimer Eighteen-azide 13
2.60,61 Dendron tris azide 4 was synthesized by the coupling
reaction between the acid groups on 2 and 3-aminopropyl azide
3 using HATU as the coupling agent. After purification by
column chromatography, 4 was obtained as a colorless oil in
93% yield (for NMR and mass spectra of 4, see the Supporting
Information). Peptide coupling between 2 and 5 afforded
dendron nitro-nonaester 6 that after deprotection of the acid
termini using formic acid yielded dendron nitro-nonaacid 7. The
coupling reaction between the acid groups on 7 and 3 afforded
dendron nonaazide 8 that after purification by column chroma-
Reduction of the nitro group on 6 was carried out at 45 psi
H2 and 55 °C in the presence of Raney-Ni yielding dendron
amino-nonaester 9 as a white powder in 89% yield.61,62 The
coupling reaction between the amine group at the focal point
of 9 and the acid groups of the diacid-PEG linker 10 afforded
dendrimer 11 that was purified by dialysis against methanol
using a MWCO (molecular weight cutoff) 2000 dialysis
membrane. Dendrimer 11 was obtained as a colorless waxy
product in 85% yield. The 1H NMR spectrum shows all of the
expected signals including the OCH2CONH group of the PEG
linker at 3.97 ppm, suggesting the completion of the coupling
of the dendrons to the linker (no signal was observed for free
CH2COOH). The MALDI-TOF mass spectrum of 11 shows the
molecular ion peak (M + H)+ at 3070.2 m/z (calcd for
C160H278N8O47: 3065.0). Deprotection of the acid termini on 11
was achieved by stirring in formic acid at room temperature
for 48 h. Completion of the deprotection reaction was verified
by 1H NMR spectroscopy by the disappearance of the
COO(CH3)3 signal at 1.45 ppm. The mass spectrum of the
resulting dendrimer 12 containing 18 acid termini shows the
1
tography was obtained as a colorless oil in 87% yield. The H
NMR and 13C NMR spectra of 8 show all of the expected signals
including the CH2N3 at 3.36 ppm, the CH2CH2CH2N3 at 3.24
ppm, and the CH2N3 at 50.2 ppm, NO2Cq at 94.5 ppm, and
NHCdO at 175.8 and 173.7 ppm. The typical band for azide
groups at 2098.30 cm-1 was observed by IR spectroscopy, and
the mass spectrum shows the molecular ion peak of 8 (M +
1
Na)+ at 1732.9 (calcd for C67H114N40O14Na: 1725.9) (for H
NMR and mass spectra of dendrons 6, 7, and 8, see the
Supporting Information).
(62) Newkome, G. R.; Nayak, A.; Behera, R. K.; Moorefield, C. N.; Baker,
(61) Brettreich, M.; Hirsch, A. Synlett 1998, 1396–1398.
G. R. J. Org. Chem. 1992, 57, 358–362.
9
J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010 3925

A R T I C L E S
Ornelas et al.
Scheme 2. CuAAC between Dendron 8 and Alkyne Derivative 14a
a (a) CH2Cl2/H2O (1:1), CuSO4 ·5H2O, sodium ascorbate (SA), room temperature, 24 h; (b) CH2Cl2/H2O/EtOH (1:1:1), CuI, sulfonated bathophenanthroline
(SBP), sodium ascorbate, room temperature, 24 h.
molecular ion peak (M + Na)+ at 2078.5 (calcd for
C88H134N8O47Na: 2077.8). The azide termini were then intro-
duced into the dendrimer by the coupling reaction between the
acid termini on 12 and 3 affording dendrimer eighteen-azide
13. After purification by dialysis against methanol using a
MWCO 2000 dialysis membrane, 13 was obtained as a colorless
oil in 76% yield. The 1H NMR spectrum of 13 shows all of the
expected peaks including the signals that correspond to the PEG
linker (OCH2CONH at 3.91 ppm and OCH2CH2O at 3.65 ppm)
and the triplet that corresponds to the CH2N3 group at 3.35 ppm.
The IR spectrum shows a band at 2098.30 cm-1, confirming
the presence of azide groups.
Functionalization of Poly(amide)-Based Dendrons and
Dendrimers with PEG Chains Using CuAAC. To obtain water-
soluble dendrons, we carried out the functionalization of 8 with
alkyne derivative 14 that contains a PEG chain, which often
endows materials with water solubility, biocompatibility, and
avoids attachment to nonspecific proteins. We first investigated
the functionalization of 8 using CuAAC (Scheme 2).
extraction of the reaction with methylene chloride. Analysis of
the organic and aqueous phases showed that dendron 15
remained in the aqueous phase together with the copper catalyst
and sodium ascorbate. The aqueous phase was evaporated to
dryness, redissolved in methanol, and filtered. After evaporation
of the solvent, a light green waxy product was obtained. The
green coloration suggested contamination of dendron 15 with
copper, making this workup procedure not usable.
Purification of 15 by column chromatography was attempted
using silica gel (SiO2) and C18-silica gel reverse-phase column
chromatography. In both cases, a green layer remained on top
of the columns, suggesting the removal of a significant amount
of copper. The products obtained, however, also had a light
green coloration, still suggesting the contamination with copper.
We also tried to purify 15 by dialysis against water and
against an aqueous solution of ammonium hydroxide or EDTA.
Again, the green coloration of the product inside the dialysis
bag remained (for detailed information about the workup
procedures, see the Experimental Section).
Our first choice for the CuAAC reaction (method A) was a
mixture of CH2Cl2/H2O (1:1) as solvent and CuSO4/sodium
ascorbate (SA) as the Cu(I) source, and the reaction was stirred
at room temperature for 24 h. To remove the copper catalyst
from the target dendron 15, we investigated several workup
procedures typical for “click” dendrimers including extraction,35,41
chromatographic methods,63,64 and dialysis.65
Despite our efforts to remove the copper catalyst from 15, a
green waxy product was obtained for all workup procedures,
suggesting the contamination with copper. CHN elemental
analysis of 15 synthesized by method A (CuSO4/SA) confirms
the nonpurity of the final product (Anal. Calcd for
C157H276N40O50: C, 53.51; H, 7.89; N, 15.90. Found: C, 44.54;
H, 6.75; N, 9.77). The lower amount of carbon in the elemental
analysis suggests the presence of inorganic impurities. To
confirm the presence of copper, ICP-MS analysis of 15 was
performed, and 0.50% of copper was found.
To extract dendron 15 from the reaction mixture, an aqueous
solution of ammonium hydroxide was added followed by the
(63) Lee, J. W.; Kim, J. H.; Kim, B.-K.; Kim, J. H.; Shin, W. S.; Jin, S.-
H. Tetrahedron 2006, 62, 9193–9200.
To minimize the copper contamination on 15, we varied the
CuAAC reaction conditions. Several reports in the literature
suggest that in the CuAAC reaction so-called “click additives”
can be used to stabilize the Cu(I) and even accelerate the reaction
(64) Lee, J. W.; Kim, B. K.; Kim, J. H.; Shin, W. S.; Jin, S.-H. Bull. Korean
Chem. Soc. 2005, 26, 1790–1794.
(65) Ornelas, C.; Ruiz, J.; Salmon, L.; Astruc, D. AdV. Synth. Catal. 2008,
350, 837–845.
9
3926 J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010

Strain-Promoted Alkyne Azide Cycloaddition
A R T I C L E S
Table 1. Reaction Conditions Used in CuAAC between Dendron 4
and Alkyne Derivative 14 Using Copper Wire as the Copper
Sourcea
reducing agent and the use of “click” additives such as SBP
and pentamethyldiethylenetriamine (PMDETA) (Scheme 3 and
Table 1).
reducing
agentb
reaction
time (h)
reaction
conversiond
ligandc
solution color
When solely Cu wire was used (entry 1, Table 1), no CuAAC
reaction was observed, and the solution presented a blue/greenish
coloration, suggesting the presence of Cu(II) in solution (Cu(II)
is not active in CuAAC). The reaction was then carried out in
the presence of SA (entry 2, Table 1) to ensure the presence of
Cu(I) in solution; however, despite the brown coloration of the
solution, the CuAAC reaction did not proceed. To stabilize the
Cu(I) species in solution, PMDETA was used (entries 3 and 4,
Table 1). While this ligand has been shown to be efficient when
used in combination with Cu wire,72 it did not work for the
poly(amide)-based dendrons. From all of the investigated
CuAAC reaction conditions, only the use of SBP as ligand and
SA as the reducing agent (entry 6, Table 1) resulted in the
complete conversion between 4 and 14. In this case, dendron
tris-ETG 16 was isolated easily from the copper catalyst by
removing the pieces of copper wire from the solution, evapora-
tion of the tBuOH/H2O solvent mixture, followed by dissolving
the residue in methylene chloride, filtering it through Celite,
and concentrating it under vacuum to yield a colorless waxy
product in 92% yield. The relative high purity found for dendron
tris-ETG 16 (for elemental analysis results, see the Experimental
Section) encouraged us to apply the same conditions to the
analogue dendron with nine branches.
1
2
3
4
5
6
blue/greenish
brown
yellow
yellow
yellow
24
24
24
48
24
48
0%
0%
0%
0%
0%
SA
SA
SA
PMDETA
PMDETA
SBP
SBP
dark red
100%
a Reactions were carried out in tBuOH/H2O (1:1) at room
temperature. b The amount of sodium ascorbate (SA) added was 40 mol
% per mol of azide groups. c The amount of ligand added was 20 mol
% per mol of azide groups. d The reaction conversions were obtained by
1H NMR spectroscopy after workup.
kinetics.66-69 In particular, sulfonated bathophenanthroline
(SBP) was shown to be a very useful additive because it
enhances significantly the reaction rate and it is efficient on
copper sequestering precluding protein degradation caused by
Cu(I).70 Therefore, we investigated the synthesis of 15 using
different reaction conditions that include SBP. The CuAAC
reaction was carried out using a previously prepared “click”
solution32 that consisted of a deep red water/ethanol (1:1)
solution of CuI, SA, and SBP (method B). The “click” solution
was added to a methylene chloride solution of 8 and 14, and
the mixture was stirred at room temperature for 24 h. However,
all attempts to isolate 15 from the “click” solution including
the above-described extraction, column chromatography, reverse-
phase column chromatography, and dialysis methods resulted
in a green waxy product, also suggesting the presence of copper.
CHN elemental analysis of 15 synthesized by method B also
confirms the nonpurity of the final product (Anal. Calcd for
C157H276N40O50: C, 53.51; H, 7.89; N, 15.90. Found: C, 40.92;
H, 6.11; N, 9.57). The contamination of 15 with copper was
also confirmed by ICP-MS analysis, showing a copper content
of 0.007% in the final dendron. This result demonstrates that
the ability of SBP to sequester copper decreased the amount of
copper contamination, but it did not result in complete removal
of copper from 15.
We then applied the same reaction conditions (Cu wire/SA/
SBP) to the CuAAC reaction between 8 and 14 (method C).
However, the separation of 15 from the copper catalyst was as
difficult as described for methods A and B. Despite the use of
extensive washing, precipitation, and chromatography, the final
dendron 15 synthesized by method C also presented a green
coloration, suggesting contamination with copper. CHN el-
emental analysis of 15 synthesized by method C confirmed the
nonpurity of the final product (Anal. Calcd for C157H276N40O50:
C, 53.51; H, 7.89; N, 15.90. Found: C, 49.50; H, 7.16; N, 18.71).
Contamination of 15 with copper was confirmed by ICP-MS
analysis that showed a copper impurity of 0.004%.
For all synthetic CuAAC reactions studied, method A
(CuSO4/SA), method B (CuI/SA/SBP), and method C (Cu wire/
SA/SBP), complete removal of the copper catalyst from 15 was
not possible. While the use of SBP reduced the amount of copper
contamination (0.007% Cu on method B and 0.004% Cu on
method C) in comparison to method A (0.50% Cu), in all cases,
the amount of copper present in dendron 15 is still significant,
and the CHN elemental analysis suggests the presence of other
unknown impurities. We suggest that the fact that 15 presents
several groups capable of binding copper ions including amides,
triazole, and PEG, together with the encapsulation ability of
the dendritic structure, precludes the separation of 15 from
copper residues. Indeed, it was shown that poly(amido amine)
(PAMAM) dendrimers74,75 and triazole-based dendrimers36 bind
different metal ions including CuI and CuII and that the dendritic
structure plays in important role in the stability of the copper
complexes formed inside the dendrimers. Because the final
dendron nona-ETG 15 synthesized by CuAAC is a combination
of poly(amide)-based and triazole-based dendritic structure, the
inclusion of copper in its structure is not astonishing.
In the literature, several reports suggest that the use of copper
wire as the Cu(I) source minimizes copper contamination in
CuAAC reaction.71-73 Therefore, we investigated the use of
copper wire for the CuAAC reaction between the poly(amide)-
based dendrons and 14. Because the use of Cu wire was only
demonstrated for monomeric species71 or poly(styrene) and
poly(acrylate)star polymers,72 we set up a series of reactions to
optimize reaction conditions for the poly(amide)-based dendrons
(Table 1). Because of the simplicity of the synthesis, 4 was
used as the model for these studies, and variations on the
reaction conditions include the use of sodium ascorbate as the
(66) Chan, T. R.; Hilgraf, R.; Sharpless, K. B.; Fokin, V. V. Org. Lett.
2004, 6, 2853–2855.
(67) Gupta, S. S.; Kuzelka, J.; Singh, P.; Lewis, W. G.; Manchester, M.;
Finn, M. G. Bioconjugate Chem. 2005, 16, 1572–1579.
(68) Link, A. J.; Tirrell, D. A. J. Am. Chem. Soc. 2003, 125, 11164–11165.
(69) Rodionov, V. O.; Presolski, S. I.; Gardinier, S.; Lim, Y.-H.; Finn,
M. G. J. Am. Chem. Soc. 2007, 129, 12696–12704.
(70) Lewis, W. G.; Magallon, F. G.; Fokin, V. V.; Finn, M. G. J. Am.
Chem. Soc. 2004, 126, 9152–9153.
(71) Appukkuttan, P.; Dehaen, W.; Fokin, V. V.; Van der Eycken, E. Org.
Lett. 2004, 6, 4223–4225.
(72) Urbani, C. N.; Bell, C. A.; Whittaker, M. R.; Monteiro, M. J.
Macromolecules 2008, 41, 1057–1060.
(74) Zhao, M. Q.; Sun, L.; Crooks, R. M. J. Am. Chem. Soc. 1998, 120,
4877–4878.
(73) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596–2599.
(75) Crooks, R. M.; Zhao, M. Q.; Sun, L.; Chechik, V.; Yeung, L. K. Acc.
Chem. Res. 2001, 34, 181–190.
9
J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010 3927

A R T I C L E S
Ornelas et al.
Scheme 3. CuAAC Reaction between Dendron Tris-azide 4 and Alkyne Derivative 14 Using Copper Wire as the Cu(I) Sourcea
a The reaction conditions are described in Table 1.
Scheme 4. Reduction of the Nitro Group at the Focal Point of Dendron 15
As a final attempt to remove the copper residues from 15,
we carried out the reduction of the nitro group at the focal point
to an amine, using a combination of CuSO4/NaBH4 in ethanol.76
The reduction was carried out using 15 obtained from the three
different methods described above. In all cases, during reflux
the green colored solution became clear, and a black precipitate
formed (presumably Cu0). After removal of the black precipitate
by filtration, dendron 17 was obtained as a light yellow waxy
product (Scheme 4). The 1H NMR spectrum of 17 shows all of
the expected peaks including the triazole proton at 8.04 ppm
and both CH2-triazoles at 4.65 and 4.46 ppm. Formation of 17
was further confirmed by MALDI-TOF, showing its molecular
ion peak (M + Na)+ at 3517.1 m/z (calcd for C157H278N40O48Na:
3516.6 m/z). Despite the separation of a significant amount of
copper by its precipitation as Cu0, CHN elemental analysis of
17 still suggests the presence of a significant amount of copper
and/or inorganic impurities (Anal. Calcd for C157H278N40O48:
C, 54.97; H, 8.02. Found: C, 51.23; H, 7.21; the presence of
Cu was confirmed by ICP-MS showing a copper content of
0.001%). Moreover, the harsh reactions conditions for the
reduction step (NaBH4/reflux) cannot be applied, generally
limiting the potential of this strategy.
tures, in particular amide and PEG-based materials, which are
highly desirable for biological applications. While CuAAC was
shown to be a very useful tool in materials chemistry,27,29 the
significant copper contamination demonstrated in this contribu-
tion severely limits the use of CuAAC as the synthetic strategy
of choice for the functionalization of poly(amide)-based materi-
als. Incomplete copper removal prevents the use of these systems
in biological applications due to the toxicity of copper, and this
represents a step back for poly(amido)-based materials such as
PAMAM.
Besides the CuAAC reaction conditions used here, procedures
that include other “click” additives such as TBTA (tris[(1-
benzyl-1H-1,2,3-triazol-4-yl)methyl]amine) or benzimidazole
ligands could have been attempted; however, these ligands were
shown to be efficient in accelerating CuAAC kinetics and in
copper sequestering when used in monomeric species.66-69 The
fact that the copper wire/SA/SBP procedure afforded pure
dendron tris-ETG but yielded copper contaminated dendron
nona-ETG shows that the efficiency of “click” additives on
monomeric species does not guarantee their success on dendritic
molecules. Thus, our efforts were then focused on using a
copper-free synthetic strategy: SPAAC.
These results clearly raise a problem with Cu-catalyzed 1,3-
dipolar cycloaddition when applied to macromolecular struc-
Functionalization of Poly(amide)-Based Dendrons and
Dendrimers with PEG Chains Using SPAAC. As an alternative
to CuAAC, we investigated the copper-free SPAAC, recently
popularized by the Bertozzi group,48 to functionalize our
(76) Sung-eun Yoo, S.-h. L. Synlett 1990, 41, 9–420.
9
3928 J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010

Strain-Promoted Alkyne Azide Cycloaddition
A R T I C L E S
Scheme 5. Synthesis of Cyclooctyne-ETG 20
Scheme 6. SPAAC Reactions between Cyclooctyne-ETG 20 and Dendron Tris-azide 4, Dendron Nona-azide 8, and Dendrimer
Eighteen-azide 13a
a For clarity, only one of the triazole regioisomers is represented (for representation of both regioisomers, see Figure 1).
poly(amide)-based dendrons and dendrimers. Crucial to the use
of SPAAC as a synthetic strategy is the preparation of
functionalized cyclooctynes such as a PEG containing version
that can be synthesized in three steps from commercially
available cycloheptene.50
m/z (calcd for C15H27BrO4Na: 373.1) with an isotopic distribu-
tion typical for bromine containing compounds. Cyclooctyne-
ETG 20 was obtained in DMSO at 60 °C using DBU as the
base (Scheme 5). Product 20 was characterized by mass
spectrometry showing its molecular ion peak (M + Na)+ at
1
Bicyclic 18 (8,8-dibromobicyclo[5.1.0]octane) was synthe-
sized in close analogy to the procedure described by Skattebol
and Solomon for the synthesis of 9,9-dibromo[6.1.0]nonane.53,77
Silver perchlorate was then used to carry out the electrocyclic
ring-opening of 18 to the trans-allylic cation, which was
captured with triethylene glycol monomethyl ether affording
293.3 m/z (calcd for C15H26O4Na: 293.2) and by H and 13C
NMR spectroscopies showing the characteristic Ct C signals
at 100.8 and 93.9 ppm (for NMR and mass spectra of 19 and
20, see the Supporting Information).
We initially investigated the SPAAC reaction between 20 and
4, 8, and 13 (Scheme 6). All three reactions were carried out in
an EtOH/H2O (1:1) mixture using 1.5 equiv of 20 per azide
group. At the beginning of each reaction, the solution was
cloudy, probably due to the low solubility of the starting
materials in the solvent mixture. However, after 30 min, the
1
bromo-trans-cyclooctene 19. The H NMR spectrum of 19
clearly shows a signal at 6.06 ppm characteristic of the vinyl
proton. Compound 19 was also characterized by mass spec-
trometry showing its molecular ion peak (M + Na)+ at 373.2
1
solutions became clear, and the H NMR spectra of the three
(77) Skattebol, L.; Solomon, S. Org. Synth. 1969, 49, 35.
reactions showed significant amounts of triazole products,
9
J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010 3929

A R T I C L E S
Ornelas et al.
4.23 ppm. The assignment of the two different signals to the
CH2-triazole protons suggests the presence of the two expected
regioisomers. Analyzing the HSQC NMR spectra of dendrimer
23 allows for the assignment of the signals at 4.77 and 4.70
ppm to the CH protons of cycloctyne adjacent to triazole moiety,
again suggesting the presence of both regioisomers.
Once the proton signals for both triazole regioisomers were
assigned, their ratio was calculated on the basis of integrations.
The CH2-triazole signals (labeled a and a′ in Figure 1) appear
as a triplet at 4.24 ppm for isomer 1 and as a multiplet at
4.30-4.55 ppm for isomer 2. The signal of the allyl protons
adjacent to the ETG chain (labeled b and b′ in Figure 1) appears
as a triplet at 4.79 ppm for isomer 1 and as a multiplet at
4.63-4.70 ppm for isomer 2. Previous studies about SPAAC
kinetics and isomerization ratios in monomeric units showed
that both isomers are formed in approximately 1:1 ratios.50 In
our case, the ratio of isomer 1/isomer 2 for dendron tris-ETG
21 is 1:1.2 (55% isomer 1 and 45% isomer 2), which is
comparable to the results obtained for the monomeric species.
Dendron 22 and dendrimer 23 present ratios of approximately
1:1.7 (63% isomer 1 and 37% isomer 2), suggesting that the
steric hindrance favors the formation of isomer 1 (Figure 1).
To study the influence of the temperature on regioisomers
formation, the SPAAC reaction was also carried out at 45 and
80 °C. At these temperatures, nearly quantitative yields were
also obtained, but no difference was observed in the final
regioisomer ratio.
The presence of both isomers on the SPAAC products was
also detected by 13C NMR spectroscopy where four different
signals for the two quaternary carbons of the triazole ring (145.2,
144.5, 134.1, and 133.4 ppm) were obtained (see the Supporting
Information).
Mass spectrometry was used to further verify the products
of the SPAAC reactions. Dendron tris-ETG 21 was character-
ized by ESI mass spectrometry, and its molecular ion peak
Figure 1. Partial 1H NMR spectra of the SPAAC products showing signals
of both regioisomers.
(M
+
Na)+ was observed at 1356.9 (calcd for
C64H111N13O17Na: 1356.8). Dendron nona-ETG 22 and den-
drimer eighteen-ETG 23 presented their molecular ion peaks
by MALDI-TOF mass spectrometry at (M + H)+ at 4135.8
(calcd for C202H349N40O50: 4134.6) and 8396.1 (calcd for
C412H711N80O101: 8396.3), respectively (Figure 2). Because
of the higher molecular weight of dendrimer 23, higher laser
power was needed to obtain the molecular ion peak, and thus
significant fragmentation was observed. The difference
between the fragmentation peaks is 335, which corresponds
to the loss of a triazole-ETG fragment (see the Supporting
Information for details on fragmentation).
suggesting that the inclusion of some ETG chains into the
dendrons and dendrimer increases their solubility in the aqueous
solvent mixture. In all three cases, the solvents were removed
under vacuum after 24 h, and the products were precipitated
1
from CH2Cl2/hexane (Scheme 6). The H NMR spectra of the
final products suggest completion of the SPAAC reaction after
24 h by complete disappearance of the CH2N3 signals at 3.36
ppm. The consumption of the azide termini was also confirmed
by IR spectroscopy via the disappearance of the N3 band at 2098
cm-1. These results demonstrate the complete conversion of the
azide groups to the triazole moieties.
In stark contrast to the CuAAC products, all SPAAC
products were obtained analytically pure when submitted to
elemental analysis (For dendron 21, Anal. Calcd for
C64H111N13O17: C, 57.59; H, 8.38. Found: C, 57.68; H, 8.42.
For dendron 22, Anal. Calcd for C202H349N40O50: C, 58.63;
H, 8.48. Found: C, 58.77; H, 8.51. For dendrimer 23:, Anal.
Calcd for C412H710N81O101: C, 58.91; H, 8.52. Found: C,
59.11; H, 8.59). These results clearly demonstrate the
superiority of SPAAC in the functionalization of amide-based
materials including dendrimers over CuAAC, allowing the
use of such materials in biological applications. The SPAAC
reaction showed to be efficient on dendrimers up to 18 azide
1
The H NMR spectra of the SPAAC products present four
new signals at 4.2-4.8 ppm. Assignment of the new signals
was elucidated by COSY and HSQC NMR spectroscopies (for
COSY and HSQC NMR spectra of dendrons 21 and 22 and
dendrimer 23, see the Supporting Information). By comparison
with the azide derivatives, the signal corresponding to
NHCH2CH2CH2-triazole protons can be identified easily at 3.22
pm. Analyzing the 2D COSY NMR spectrum, we observed the
coupling between the NHCH2CH2CH2-triazole protons with the
NHCH2CH2CH2-triazole protons that appear at 2.03 ppm. The
coupling between NHCH2CH2CH2-triazole and the CH2-triazole
protons shows that the CH2-triazole protons appear at 4.39 and
9
3930 J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010

Strain-Promoted Alkyne Azide Cycloaddition
A R T I C L E S
dendrimer chemistry due to the mild and metal-free reaction
conditions, no side products, tolerance toward functional groups,
and high yields. Application of this synthetic tool toward
multifunctional dendrimers to build well-defined drug delivery
carriers for theranostics is currently under investigation in our
group.
Experimental Section
General. Dendrons nitrotriester 1 and aminotriester 5 were
purchased from Frontier Scientific. Dendron triacid 2 was obtained
by deprotection of the acid groups on commercial dendron
nitrotriester 1 as previously described.61,62,78 Dendrons 6, 7, and 9
were prepared following literature procedures (for detailed experi-
mental procedures and characterization data of all compounds, see
the Supporting Information).61,62,78 3-Aminopropyl azide 3 was
synthesized according to literature procedures.79
Synthesis of Dendrimer 23. Dendrimer eighteen-azide 13 (0.013
g, 0.0037 mmol) and cyclooctyne 20 (0.027 g, 0.100 mmol) were
dissolved in 1 mL of ethanol. Water was added (1 mL), and the
solution became cloudy. The solution was stirred at room temper-
ature for 24 h, resulting in a colorless solution. The solvent was
removed, and the product was washed twice with hexanes and
precipitated from CH2Cl2/hexanes to remove the cyclooctyne that
was added in excess. Dendrimer 23 was obtained as a colorless
waxy product in 95% yield (0.030 g). 1H NMR (CDCl3, 400 MHz,
δ
ppm vs TMS): 7.76 (s, CONH, 26H), 4.77 (s, CdCCH-O, isomer
1), 4.70 (m, CdCCH-O, isomer 2), 4.39 (m, CH2-triazole, isomer
2), 4.23 (t, J ) 6.84 Hz, CH2-triazole, isomer 1), 3.89 (s,
OCH2CONH, 4H), 3.65-3.45 (m, OCH2CH2O, 224H), 3.34 and
3.32 (s, OCH3, 54H), 3.22 (m, CONHCH2, 36H), 3.06-2.62 (m,
CH2CdCCHCH2 of cyclooctene ring, 72H), 2.25, 2.16, and 1.96
(m, CH2CH2NHCO, 96H), 2.03 (m, CONHCH2, 36H), 1.77-0.97
(m, CH2 of cyclooctene ring, 108H). 13C NMR (CDCl3, 100 MHz,
δ
ppm vs TMS): 174.2-173.4 (CdO), 145.2 and 144.5 (CH2NCq of
Figure 2. MALDI-TOF mass spectra of (a) dendron 22 and (b) dendrimer
23 (due to the higher molecular weight of 23, higher laser power was needed
to obtain the molecular ion peak, which is probably the cause of the
significant fragmentation peaks).
triazole ring, both isomers), 134.1 and 133.4 (CH2N3Cq of triazole
ring, both isomers), 74.7 and 72.0 (CH-O of cycloctene ring),
70.4-70.6 and 68.0 (OCH2CH2O and OCH2NHCO), 59.2 (OCH3),
46.7 and 45.5 (CH2-triazole, both isomers), 37.0 (CH2CH2CH2-
triazole), 35.6 (CH2CH2CH2-triazole), 31.7-20.2 (CH2 of cy-
clooctene ring and CH2CH2CONH). MALDI-TOF (M + H)+ m/z
calcd for C412H710N80O101: 8396.1, found 8395.3. Anal. Calcd for
termini, which makes this reaction a new useful synthetic
tool for dendrimer functionalization.
C412H710N81O101: C, 58.91; H, 8.52. Found: C, 59.11; H, 8.59.
Conclusion
Acknowledgment. We thank New York University for financial
support of this research. M.W. gratefully acknowledges a Camille
Dreyfus Teacher/Scholar Award and an Alfred P. Sloan Fellowship.
This contribution describes the advantage of strain-promoted
alkyne azide 1,3-dipolar cycloaddition in comparison to the
copper-catalyzed version for the functionalization of poly(a-
mide)-based dendrons and dendrimers, promising materials for
a wide variety of biological applications. In particular, mild
reaction conditions, complete conversions, and the lack of
residual copper make SPAAC the functionalization strategy of
Supporting Information Available: Detailed synthetic pro-
cedures, and NMR and mass spectra for all compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
choice. The SPAAC products were characterized by 1H and 13
C
NMR, 2D HSQC, and COSY NMR spectroscopies, mass
spectrometry, and elemental analysis. All characterization
methods demonstrate clean reactions and complete conversions.
This is the first report on the use of such a reaction on
dendrimers functionalization, and the results obtained here
demonstrate that this reaction can be an important tool in
JA910581D
(78) Newkome, G. R.; Behera, R. K.; Moorefield, C. N.; Baker, G. R. J.
Org. Chem. 2002, 56, 7162–7167.
(79) Vercillo, O. E.; Andrade, C. K. Z.; Wessjohann, L. A. Org. Lett. 2007,
10, 205–208.
9
J. AM. CHEM. SOC. VOL. 132, NO. 11, 2010 3931