Click chemistry with lanthanide complexes: A word of caution

Article abstract of DOI:10.1039/b918891f

A propargyl-appended Eu(III) complex shows unanticipated reactivity towards NaN₃ in the Cu(I) catalysed click reaction yielding an unsubstituted 1,2,3-triazole. The resulting complex exhibits pH-responsive ¹H NMR and Eu(III) luminesc

Full text of DOI:10.1039/b918891f

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
Click chemistry with lanthanide complexes: a word of caution†
Graeme J. Stasiuk and Mark P. Lowe*
Received 10th September 2009, Accepted 16th September 2009
First published as an Advance Article on the web 24th September 2009
DOI: 10.1039/b918891f
A propargyl-appended Eu(III) complex shows unanticipated
reactivity towards NaN3 in the Cu(I) catalysed click reac-
tion yielding an unsubstituted 1,2,3-triazole. The resulting
complex exhibits pH-responsive H NMR and Eu(III) lumi-
nescence behaviour in the physiological pH range consistent
with deprotonation of the NH-acidic bound triazole.
triazole 3 and the unsubstituted NH-triazole 4 (Scheme 1).
Whilst reports of the synthesis of unsubstituted triazoles have
appeared, these tend to be formed at elevated temperatures,
not under ambient conditions.8 There is currently much interest
in the generation of unsubstituted 1,2,3-triazoles due to their
topological and electronic similarities to amides; their biological
activity is being explored as peptide bond mimics.4d Synthesis of
NH-triazoles usually requires the use of a ‘protected’ organic azide,
such as azidomethyl pivalates and carbamates that can be easily
removed once the triazole is formed.9 The alternative is the highly
undesirable reaction with hydrazoic acid at high temperature.10
1
In order to overcome the inherent sensitivity problem of Magnetic
Resonance (MR) imaging,1 our recent studies have been geared
towards developing Gd(III) complexes as contrast agents that
respond to physiological changes such as enzyme activity to
amplify the signal.2 We are now looking at ways to improve the
sensitivity of MR contrast agents by ‘increasing the payload’,
i.e. by delivering multiple Gd(III) chelates. We were attracted to
the apparent perfection of the Cu(I) catalyzed azide–alkyne cy-
cloaddition reaction in order append multiple lanthanide chelates
from a central ‘hub’. In recent years this ‘click’ reaction to form
1,2,3-triazoles3 has been applied to a plethora of applications.4
Recent reports document the use of this synthetic strategy to attach
Eu(III) chelates to dansyl chromophores and Gd(III) chelates to
e.g. a cyclodextrin hub.5 In both cases an N-(2-propynyl)acetamide
appended from a DOTA-based chelate was utilized. In order to
avoid the use of amides due to slowing of the water exchange rate
on Gd(III) complexes, we chose to use the DOTA-based proligand
1. As a Ln(III) complex we have demonstrated facile conversion
of 2 to the corresponding triazole via the Cu(I) catalyzed click
reaction with organic azides.6 Due to the number of azides per
molecule required to synthesize multimeric assemblies of Gd(III)
complexes, we were wary of isolation of the potentially explosive
azide precursors. Conventional wisdom suggests that reaction of
alkynes with in situ generated azides (via alkyl halides in the
presence of NaN3) is a clean, facile reaction and residual NaN3
does not interfere with the cycloaddition reaction.7 We believed
that this attractive synthetic trick would allay our safety concerns,
circumventing the need to isolate compounds containing multiple
azides. Herein, we report the unexpected reactivity of 2 with
NaN3, to yield a Eu(III) complex containing an unsubstituted
NH-triazole 4, and detail its pH responsive 1H NMR and
luminescent behavior.
Scheme 1 Synthesis of 4.
Due to the unexpected reactivity of 2 with NaN3, we attempted
the click reaction of 2 in water in the presence of CuSO4, sodium
ascorbate and NaN3. Under these conditions complete conversion
to 4 was seen after stirring at room temperature for 12 hours. It is
evident that Ln(III) complexes based on proligand 1 are unsuitable
for use in click reactions where the azide is generated in situ, due
to their propensity to react with NaN3. We have observed this
effect with a variety of lanthanides (e.g. Sm–Dy) and suspect that
it is linked to the close proximity of the alkyne to the Ln(III)
ion. There are numerous postulated intermediates in the proposed
mechanism of the Cu(I) catalyzed 1,3 Huissgen cycloaddition
reaction, in all cases an organic azide is deemed essential for the
reaction to proceed.4,7 One mechanism suggests the Cu-acetylide
that is initially formed, and the azide, are bound to different copper
atoms in a doubly bridged assembly. This may offer some clues to
the unusual reactivity observed in 2 as the copper acetylide must
2 was synthesized via standard methods.† Our initial attempts
to react 2 with simple in situ generated alkyl azides (e.g. via
ethyl or benzyl bromide) under standard room temperature click
conditions did not give the clean high-yielding reaction that
we expected and two products were formed: the desired alkyl
-
be in close proximity to the central Eu(III) ion. It is likely that N3
bound to Eu(III) is amenable to cycloaddition to the activated
alkyne.11
4 was characterized by 1H NMR and luminescence spectroscopy
and displays interesting pH-responsive behaviour in the physio-
1
Department of Chemistry, University of Leicester, University Road, Leices-
ter, UK LE1 7RH. E-mail: mplowe@le.ac.uk
† Electronic supplementary information (ESI) available: Experimen-
tal details, 1H NMR spectra and luminescence spectra. See DOI:
10.1039/b918891f
logical pH range. The H NMR spectrum (400 MHz, pD 9.0,
278K) is typical of a rigid 8-coordinate unsymmetrical Eu(III)
complex exhibiting a broad range of resonances from -18 to
33.5 ppm. This is indicative of triazole coordination to Eu(III).
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 9725–9727 | 9725
©

View Article Online
The restricted arm rotation in 8-coordinate (cf. 7-coordinate)
complexes slows the exchange between the various stereoisomers,
resulting in resolution of the axial and equatorial cyclen ring
hydrogens. The four axial resonances shifted to high frequency
are typical of a square antiprismatic coordination geometry. The
presence of the minor twisted square antiprismatic isomer is also
demonstrated by four resonances at slightly lower frequency (9.9
to 13.6 ppm).
protonation, but via the pyridinic nitrogen, indeed the form of
the spectrum in acid media is very similar to related coordinated
N-substituted triazoles.6
The luminescent emission spectrum of 4 is typical of an
8-coordinate DOTA-based Eu(III) complex. At pH 5.5 the form
of the spectrum is reminiscent of DOTA-based Eu(III) com-
plexes bearing pendant coordinated pyridyls14 (and indeed of the
N-substituted triazoles)6 as are the relative intensities of the DJ =
7
7
At basic pH the triazole is bound to Eu(III) as a monoanionic
donor i.e. as the 1,2,3-triazolide (Scheme 2).12 NH-triazoles are
acidic; their pKas typically ~ 5–9.13 This anionic coordination
mode is different to the coordination behavior of N-substituted
triazoles that coordinate via a ‘pyridinic’ nitrogen atom6 (the pKa
1 (5D0→ F1) and DJ = 2 (5D0→ F2) transitions. On raising
the pH of the solution, one of the transitions in the DJ = 1
manifold shifts from 592 to 597 nm. The ratio of intensities of
these peaks enables a protonation constant of 6.9 ( 0.1) to be
extracted from the data (an excited-state pKa recorded at 298 K),
i.e. ratiometric concentration independent pH response. This is
1
of this nitrogen is typically < 1). The H NMR spectrum of 4
1
exhibits pH-dependent behavior (Fig. 1). On titrating from base
to acid, three of the four most shifted axial resonances (33.4,
33.1 and 32.6 ppm) broaden, moving to lower frequency, the
fourth (31.1 ppm) moves to higher frequency. On moving to acidic
media, the resonances are once more resolved. The resonance that
shifts to higher frequency now appears at 32.1 ppm; the three
resonances that shift to lower frequency now appear at 29.8, 29.2
and 28.6 ppm. This is a clear indication of a protonation event,
i.e. of one of the triazolide nitrogens. The subtle change from
effectively N- coordination to pyridinic N-coordination results in a
marked change in spectral form. Similar movements can be traced
for other resonances in the spectrum for both the square and the
twisted square antiprismatic isomers. The change in chemical shift
of these resonances Dd enable the determination of a protonation
constant (logKMLH or pKa) of 7.5 0.1 i.e. 4 exhibits pH-responsive
behavior in the physiological pH range. Given the apparent rigidity
consistent with the changes observed in the H NMR spectrum
(pKa = 7.5, a ground-state pKa recorded at 278 K). This behavior
supports the idea of a subtle change in coordination mode of the
triazole from neutral ‘pyridyl’ to anionic triazolide. The titration
is complicated by a second deprotonation event occurring at pH
>8. This is deprotonation of the bound water molecule in the
9th coordination site of 4 and is characteristic of such complexes:
there is a change in spectral form of DJ = 1 and an increase in
intensity of the hypersensitive DJ = 2 transition. This transition
is particularly sensitive to the polarisability of the axial donor
atom.
Luminescent lifetime measurements confirm these conclusions.
The lifetime of the Eu(III) excited state is susceptible to quenching
from O–H oscillators. Luminescent lifetime measurements on 4 in
H2O and D2O enable the determination of the hydration state (q) of
the complex. At pH 10.0 q = 0.4 (kH = 1.81, kD = 1.20 ms-1),
O
O
2
2
1
demonstrated in the H NMR spectrum after protonation, we
consistent with deprotonation of the bound water molecule, i.e.
conclude that the triazole remains coordinated to Eu(III) after
OH- is bound. At pH 8.5, before hydrolysis of the bound water,
when the triazole is bound in an anionic manner, q = 1.1 (kH2
=
O
2.27, kD = 1.11 ms-1). At pH 5.5 when bound in a pyridinic
O
2
manner, the apparent hydration state is slightly higher than unity
q = 1.4 (kH = 2.51, kD = 1.06 ms-1).15
O
O
2
2
Conclusions
In conclusion this study has demonstrated that a seemingly
small alteration, substituting an N-(2-propynyl)acetamide for
propargyl as a means of coupling lanthanide chelates via click
chemistry is only suitable if the organic azide involved in the
click reaction is pre-formed and the solution is free of NaN3. The
in situ generation of azides to avoid the isolation of potentially
hazardous multiple azide-containing precursors is incompatible
with 2 due to the unprecedented reactivity of NaN3 with propargyl-
appended Eu(III) complex. This has implications if complexes
such as these are to be used to make multimeric contrast agents
(as Gd(III)) where precursors containing multiple azides are
required.
Scheme 2 Protonation sequence for 4.
The unexpected product of this click reaction 4 showed in-
teresting protonation behavior, switching from triazole bound to
Eu(III) in an N-deprotonated anionic form in basic media, to an
N–H protonated pyridyl form in acidic media. This pH-dependent
switch, demonstrated by both 1H NMR and luminescence studies
occurs in the physiological pH range. We are currently probing
the origins of this unusual reactivity and exploring the potential
Fig. 1 1H NMR vs. pD for 4 (400 MHz, 278K).
9726 | Dalton Trans., 2009, 9725–9727
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
applications of this pH-responsive behavior for a variety of Ln(III)
ions.
5 (a) R. F. H. Viguier and A. N. Hulme, J. Am. Chem. Soc., 2006, 128,
11370–11371; (b) Y. Song, E. K. Kohlmeir and T. J. Meade, J. Am.
Chem. Soc., 2008, 130, 6662–6663.
6 M. Jauregui, W. S. Perry, C. Allain, L. R. Vidler, M. C. Willis, A. M.
Kenwright, J. S. Snaith, G. J. Stasiuk, M. P. Lowe and S. Faulkner,
Dalton Trans., 2009, 6283–6285.
7 P. Wu and V. V. Fokin, Aldrichimica Acta, 2007, 40, 7–17.
8 (a) J. Andersen, S. Bolvig and X. F. Liang, Synlett, 2005, 2941–2947;
(b) A. K. Feldman, B. Colasson and V. V. Fokin, Org. Lett., 2004, 6,
3897–3899.
Acknowledgements
The authors thank the EPSRC National Mass Spectrometry
Service Centre, Swansea for high resolution ESMS and Dr Gerry
Griffiths for NMR spectra.
9 J. C. Loren, A. Krasinski, V. V. Fokin and K. B. Sharpless, Synlett,
2005, 2847–2850.
Notes and references
10 T. Jin, S. Kamijo and Y. Yamamoto, Eur. J. Org. Chem., 2004, 3789–
3791.
1 P. Caravan, Chem. Soc. Rev., 2006, 35, 512–523.
2 M. Giardiello, M. P. Lowe and M. Botta, Chem. Commun., 2007, 4044–
4046.
3 V. V. Rostovtsev, L. G. Green, V. V. Fokin and K. B. Sharpless, Angew.
Chem., Int. Ed., 2002, 41, 2596–2599.
11 Preliminary studies suggest that under typical click conditions 3–5% of
-
2 will have N3 bound.
12 4 is observed in the negative ESMS, [M - H]- m/z = 574.1060 (151Eu).
13 Y. Tanaka, S. R. Velen and S. I. Miller, Tetrahedron, 1973, 29, 3271–
3283.
4 For a selection of recent reviews: (a) J. E. Moses and A. D. Moorhouse,
Chem. Soc. Rev., 2007, 36, 1249–1262; (b) M. V. Gil, M. J. Are´valo and
14 S. Aime, A. S. Batsanov, M. Botta, J. A. K. Howard, M. P. Lowe and
D. Parker, New J. Chem., 1999, 23, 669–670.
´
O. Lo´pez, Synthesis, 2007, 1589–1620; (c) V. D. Bock, H. Hiemstra and
J. H. van Maarseveen, Eur. J. Org. Chem., 2006, 51–68; (d) H. C. Kolb
and K. B. Sharpless, Drug Discovery Today, 2003, 8, 1128–1137.
15 q = 1.2(kH - kD - 0.25) does not correct for quenching by the
2 O
2 O
triazole NH.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 9725–9727 | 9727
©