Defining the structural parameters of triazole ligands in the templated synthesis of silver nanoparticles

Article abstract of DOI:10.1002/ejic.201402593

This manuscript describes a one-pot method for the synthesis of size- and shape-selected silver nanoparticles (AgNPs) using Tollens' reagent [Ag(NH₃)₂OH] as the silver source. Sugar triazole ligands facilitate the formation of monodi

Full text of DOI:10.1002/ejic.201402593

FULL PAPER
DOI:10.1002/ejic.201402593
Defining the Structural Parameters of Triazole Ligands in
the Templated Synthesis of Silver Nanoparticles
[
a]
[b]
[c]
Heba A. Kashmery, Alasdair W. Clark, Ruggero Dondi,
[
d]
[d]
[a]
Andrew J. Fallows, Paul M. Cullis, and Glenn A. Burley*
Keywords: Template synthesis / Nanoparticles / Silver / Click chemistry / Nitrogen heterocycles
I
This manuscript describes a one-pot method for the synthesis
of size- and shape-selected silver nanoparticles (AgNPs)
that the Ag -binding affinity of these triazole ligands deter-
mines their capacity to tune the size of the resultant AgNPs
I
using Tollens’ reagent [Ag(NH
3
)
2
OH] as the silver source.
formed. Weaker Ag -binding ligands can be used to form
Sugar triazole ligands facilitate the formation of monodis-
perse AgNPs in which the size and shape can be controlled
according to the reaction conditions. Increasing the size of
the ligand reduces size tunability but enhances colloidal sta-
bility in high-salt buffers. A key conclusion from this study is
monodisperse, angular AgNPs over a wider range of sizes
[(12Ϯ3) to (33Ϯ7) nm], whereas triazole ligands that exhibit
I
a higher Ag -binding affinity produce monodisperse, spheri-
cal AgNPs of a single size [(18Ϯ5) nm].
colloidal stability.[17–22] Despite the simple method of prep-
aration of AgNPs and mild reaction conditions of this pro-
cess, an in-depth understanding of the mechanistic factors
that control AgNP formation using Tollens’ reagent is lack-
ing.
Introduction
Silver nanoparticles (AgNPs) are of considerable impor-
tance for the development of highly sensitive plasmon-
based diagnostic platforms.[1] By virtue of their strong sur-
face plasmon resonance that can extend throughout the
To address this shortfall, we have explored the use of
visible and near-infrared regions,[
2,3]
AgNPs are superior
I
Ag -binding triazole ligands to direct the synthesis of size-
plasmonic candidates relative to other noble-metal nano-
[23]
and shape-selected AgNPs using Tollens’ reagent.
In an
particles such as gold and copper.[
4–8]
Although a variety
I
earlier study, we reported the utility of the Ag -binding li-
gand 1 [Figure 1 (a)] to form monodisperse and stable sus-
pensions of spherical AgNPs [i.e., AgNP@(1)]. Our current
mechanistic rationale for the formation of AgNP@(1) in-
of methods exist for the preparation of AgNPs,[
9,10]
very
few of these can control both the size and shape of AgNPs
under mild reductive conditions and at room tempera-
[
11–13]
ture.
AgNP formation using Tollens’ reagent
I
a
volves the initial coordination of Ag to the 3-position (N )
[
Ag(NH ) OH] in the presence of aldehydes is one such
3
2
of both the resorcinol triazole nitrogen atoms [red in Fig-
[
11,14–16]
method that has been underutilised in the literature.
Monosaccharides and disaccharides have been used pre-
viously to form AgNPs with varying levels of dispersity and
b
ure 1 (a)] and the 2-position (N ) of the southern-most tri-
azole nitrogen atom [pink in Figure 1 (a)]. We surmised that
I
the chelating effect of 1 positions Ag atoms in close prox-
imity to the aldehyde groups of the galactose sugars. Once
[
a] Department of Pure & Applied Chemistry, University of
I
in position, these sugars facilitate the reduction of Ag to
Strathclyde,
[
24]
2
95 Cathedral Street, Glasgow, G1 1XL, UK
putatively form nanoclusters of Ag nuclei. These chelated
nanoclusters can then form size-selected AgNPs by an ex-
E-mail: glenn.burley@strath.ac.uk
http://www.burleylabs.co.uk
http://www.strath.ac.uk/chemistry/staff/academic/glennburley/
b] Division of Biomedical Engineering, School of Engineering,
University of Glasgow,
ponential growth phase followed by termination and ligand
[
[
[
[16,25]
capping [Figure 1 (a)].
What was unclear from our pre-
vious work was how each of the constituent structural ele-
ments (i.e., sugar unit, triazole, aromatic core) influenced
the size and dispersity of AgNP formation.
Oakfield Avenue, Glasgow, G12 8LT, UK
c] Department of Pharmacy and Pharmacology,
University of Bath, Claverton Down,
Bath, BA2 7AY, UK
Kvitek et al. proposed a correlation between the molec-
ular structure of the reducing sugars used in the Tollens-
d] Department of Chemistry, University of Leicester,
University Road, Leicester, LE1 7RH, UK
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201402593.
[21]
mediated formation of AgNPs. Disaccharides (e.g., malt-
©
2014 The Authors. Published by Wiley-VCH Verlag GmbH & ose), for example, produced smaller AgNPs with a narrower
Co. KGaA. This is an open access article under the terms of
the Creative Commons Attribution License, which permits use,
distribution and reproduction in any medium, provided the
original work is properly cited.
size distribution relative to monosaccharides (e.g., glucose).
The justification given by the authors for this observation
was that disaccharides had an increased number of reducing
Eur. J. Inorg. Chem. 2014, 4886–4895
4886
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Figure 1. (a) Proposed binding model of the templated synthesis of AgNPs using the first-generation ligand 1. (b) Structures of sugar
triazoles 2–7 prepared in this study.
Eur. J. Inorg. Chem. 2014, 4886–4895
4887
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
equivalents relative to monosaccharides.[ This observa- azole ligand structure. The underlying motivation for ex-
tion is in contrast to our previous study,[ in which we ploring this original design was based on a previous obser-
showed that an increase in the number of reducing equiva- vation that this ligand was used to form Ag nanostructures
lents (i.e., increasing the number of reducing sugar units when covalently attached to DNA duplexes and exposed to
20]
23]
[
27]
from two of 1 to four) did not result in the formation of Tollens’ reagent. To gain further insight into how the li-
smaller AgNPs. gand structure could influence AgNP formation, ligands
These somewhat conflicting observations have led us to were prepared in which the resorcinol core (as present in 1)
I
hypothesise that the Ag binding affinity of our triazole li- was replaced with phloroglucinol [Figure 1 (b)]. This
gands rather than the number of reducing sugars could be change subtly alters the triazole regiochemistry and inter-
playing a defining role in influencing the size of the result- triazole distance in the southern part of the ligand. The
ant AgNPs formed.[ In this manuscript, we show that the influence of this change on AgNP formation was investi-
structure of the ligand determines the size, shape, dispersity gated using a three-sugar triazole (2) and two-sugar triazole
and stability of these AgNPs. We also show that weaker (3) ligand system. The role of the central aromatic ring was
26]
I
Ag -binding ligands enable access to monodisperse AgNPs probed by the replacement of the resorcinol core with a
where their size can be tuned according to the reaction con- central nitrogen atom (4). To further explore how sugar
ditions used. Finally, we show evidence that an increase in density plays a role in the size and shape of AgNPs, the
I
the Ag -binding affinity of ligands produces AgNPs smaller hybrid ligand (5) was synthesised. Finally, to investigate
I
in diameter.
how the Ag -binding affinity is influenced by the galactose
sugar units, compounds 6 and 7 were prepared.
Results
Synthesis of Sugar Triazole Ligands 2–7
Design of Sugar Triazole Ligands 2–7
Sugar triazole 2 was prepared from phloroglucinol
In our first-generation ligand design (e.g., 1), a central (1,3,5-trihydroxybenzene) (8) in three steps (Scheme 1).
resorcinol scaffold was used to tether two galactose sugar Alkylation of 8 with propargyl bromide afforded 9 in 40%
units by means of a triazole linkage.[ A third triazole with yield.
23]
[28]
The formation of 11 proceeded smoothly under
different regiochemistry to the two resorcinol triazoles was standard copper-catalysed Huisgen cycloaddition condi-
installed in the southern part of 1 to complete a three-tri- tions using four equivalents of galactose azide 10. Acid de-
Scheme 1. Preparation of 2 and 3. Reagents and conditions: (i) Propargyl bromide (3.5 equiv.), K
CH CN, reflux, 24 h, 40%; (ii) compound 10 (4.0 equiv.) CuSO (0.1 m, 0.32 equiv.), sodium ascorbate (2.0 equiv.), THF/H
iii) trifluoroacetic acid (TFA)/H O (1:1), 70 °C, 3 h, 60%; (iv) compound 10 (2.0 equiv.), CuSO (0.1 m, 0.34 equiv.), sodium ascorbate
2.0 equiv.), THF/H O (3:1), 24%; (v) ligand 13 (5.0 equiv.), CuSO (0.1 m, 1.2 equiv.), sodium ascorbate (2.0 equiv.), THF/H O (3:1);
O (1:1), reflux, 70 °C, 3 h, 70% over two steps.
2
CO
3
(3.5 equiv.), 18-crown-6 (0.2 equiv.),
3
4
2
O (3:1), 67%;
(
(
2
4
2
4
2
and (vi) TFA/H
2
Eur. J. Inorg. Chem. 2014, 4886–4895
4888
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
protection of the isopropylidene groups of 11 afforded 2 prepared in a convergent manner through triazole forma-
after purification by means of reverse-phase (RP)-HPLC. tion by using 9 and 15 (3.3 equiv.) in the presence of cop-
For the synthesis of compound 3, a twofold click reaction per(I) to initially yield 16 in 41% yield, followed by acid
between 9 and 10 afforded 12 in 24% yield. A third click deprotection of the isopropylidine groups to afford 5
reaction using 3-azido-1-propanol (13) followed by acid de- (Scheme 3).
protection of the isopropylidines and purification by RP-
HPLC yielded 3.
Sugar triazole 4 was prepared by a threefold click reac-
tion between commercially available tripropargylamine (14)
and galactose azide (10), followed by acid deprotection and
purification by RP-HPLC (Scheme 2). Compound 5 was
Scheme 2. Preparation of 4. Reagents and conditions: (i) Com-
pound 10 (4.5 equiv.), 2,6-lutidine (1.0 equiv.), [Cu(CH CN) ]PF
0.01 equiv.), CH CN, 0 °C, 36%; and (ii) TFA/H O (1:1), 70 °C, (2.0 equiv.), THF/H
h, 60%. (2 m), reflux, 40 °C, 24 h, 50% over two steps.
Scheme 4. Preparation of 6. Reagents and conditions: (i) Com-
pound 17 (5.1 equiv.), CuSO (0.1 m, 0.33 equiv.), sodium ascorbate
O (3:1), 24 h; and (ii) ammonia in methanol
3
4
6
4
(
3
3
2
2
Scheme 3. Preparation of 5. Reagents and conditions: (i) Compound 15 (3.3 equiv.), CuSO
4
(0.1 m, 1.48 equiv.), sodium ascorbate
(
2.0 equiv.), THF/H O (3:1), 24 h, 41%; and (ii) TFA/H O (1:1), 70 °C, 24 h, 47%.
2
2
Eur. J. Inorg. Chem. 2014, 4886–4895
4889
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Scheme 5. Preparation of 7. Reagents and conditions: (i) Compound 17 (3.0 equiv.), CuSO
4
(0.1 m, 0.33 equiv.), sodium ascorbate
2.0 equiv.), THF/H O (3:1), 24 h, 76%; (ii) NaN (10.0 equiv.), acetone/H O (4:1), reflux, 3 h, 85%; (iii) ammonia in methanol (2 m),
2 3 2
(
reflux, 40 °C, 24 h; and (iv) compound 23 (3.0 equiv.), compound 24 (1.1 equiv.), DMSO, room temperature, 24 h, 31% over two steps.
Triazole ligand 6 was prepared by a click reaction be- meters of (21Ϯ3) and (18Ϯ5) nm, respectively. AgNP@2
tween azide 17 and alkyne 9. Base-mediated acetyl depro- formed using an intermediate concentration of Tollens’ (i.e.,
tection and purification by RP-HPLC afforded the tri- #19), produced AgNPs [Ø = (17Ϯ3) nm]. These AgNPs
hydroxylated ligand 6 in 50% yield (Scheme 4). Finally, li- were larger in diameter and of different shape than
[
23]
gand 7 was prepared by triazole formation between 19 and AgNP@1 [Ø = (8Ϯ5) nm].
7 (3 equiv.) in the presence of copper(I) to yield 20. Chlor-
1
ide displacement of 20 using NaN yielded azide 21. Depro-
3
tection of the acetyl groups of 21 with ammonia produced
the dihydroxylated compound 22. The installation of the
southernmost triazole group was achieved by a click reac-
I[29]
tion between 22 and butyn-1-ol 23 using resin-bound Cu
(
I
24) in 31% yield (Scheme 5). Immobilisation of Cu was
critical for the isolation of pure ligand 7 as conventional
click chemistry conditions resulted in a significant amount
of 7 complexed with copper.
Ligands 2–5 Templated the Formation of Angular AgNPs
The formation of AgNPs was screened as a function of
[
ligand] and [Tollens’]. A reaction screen of [2] and [Tol-
lens’] produced AgNP@(2) in three regions [yellow boxes
in Figure 2 (a)]: (i) A region of low concentration of Tol-
lens’ (i.e., 1 mm); (ii) a region of intermediate concentration
of Tollens’ (10 mm); and (iii) lastly, a high concentration
region of Tollens’ (20 to 50 mm). The dependence on con-
centration of both ligand and Tollens’ was similar to
AgNPs derived from 1 (i.e., AgNP@1).[ Silver aggregates
were also observed in some reaction vessels [grey boxes,
Figure 2 (a)]. UV/Vis spectra of AgNP@2 formed in these
23]
three exemplar regions all exhibited a diagnostic surface Figure 2. (a) AgNP screening array prepared using sugar triazole 2
plasmon resonance peak at approximately 420 nm [Fig- and Tollens’ reagent. White boxes represent no AgNP formation,
yellow boxes represent AgNP formation and grey boxes represent
the formation of silver mirrors. (b) UV/Vis spectra of reactions
ure 2 (b)]. TEM analysis revealed that AgNP@2 formed
angular AgNPs of similar diameter (Figure 3). For example,
#
15, 19 and 31, which formed AgNPs as observed by a surface
#
15 (lower concentration of Tollens’) and #31 (high con-
plasmon peak at 420 nm. Sample #19 was diluted 1:10 and #31
centration of Tollens’) afforded angular AgNP@2 with dia- was diluted 1:70 prior to the measurements.
Eur. J. Inorg. Chem. 2014, 4886–4895
4890
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
The formation of AgNPs was then investigated using li-
gand 3. AgNP@3 was formed over concentration ranges of
1
to 25 mm [3] and 10 to 50 mm [Tollens’] (Figure S2 in the
Supporting Information). In contrast to AgNP@2, the for-
mation of AgNPs using ligand 3 was dependent on the reac-
tion conditions. TEM analysis of several examples in this
series produced angular AgNP@3 of tuneable sizes that
ranged from (12Ϯ3) nm in diameter when a higher concen-
tration of Tollens’ was used (#31, Figure S3c in the Sup-
porting Information) through to (33Ϯ7) nm when an inter-
mediate concentration of Tollens’ was used (#19, Fig-
ure S3a). Lower concentrations of Tollens’ did not form
AgNPs@3. No silver mirrors or formation of Ag aggregates
were observed when ligand 3 was used. This suggests that
the third galactose sugar present in 2 is not essential for
AgNP formation.
A narrow reaction window was observed for the forma-
tion of poor quality and polydisperse AgNPs using ligand
4
(i.e., AgNP@4). Yellow samples with comparatively weak
plasmon peaks (maximum absorption at 0.1) were formed
at [Tollens’] 100 μm to 1 mm with silver mirrors and aggre-
gates predominating outside the optimal conditions (Fig-
ure S4a in the Supporting Information, grey boxes). The
poor dispersity and size control of AgNP@(4) was con-
firmed by SEM analysis (Figure S5 in the Supporting Infor-
mation).
AgNPs formed using ligand 5 (i.e., AgNP@5) displayed
a significantly weaker dependence on [Tollens’] (Figure S6a
in the Supporting Information) than that observed for 3. A
similarly weak dependence on AgNP formation was pre-
viously observed using ligand 25, which produced angular
(10Ϯ2) nm AgNP@25 over a similarly wide concentration
range of [25] and [Tollens’]. AgNP@5 was formed with dia-
meters that ranged from (12Ϯ1) nm (#22, Figure S7a in the
Supporting Information) to (20Ϯ2) nm (#28, Figure S7d).
The formation of AgNPs was then investigated by using
ligand 6 and Tollens’ reagent in the presence of three equiv-
alents of galactose (Figure S8a in the Supporting Infor-
mation). This experiment was conducted to determine
whether covalent attachment of the reducing sugars to the
central ligand core was essential to facilitate AgNP forma-
Figure 3. TEM images of AgNP@2 prepared using reaction condi-
tions in Figure 2 (a): (a) #15, Ø = (21Ϯ3) nm. (b) #19, Ø =
(17Ϯ3) nm. (c) #31, Ø = (18Ϯ5) nm.
Eur. J. Inorg. Chem. 2014, 4886–4895
4891
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
tion. A ratio of galactose/6 (3:1) was chosen because this observed but with a slower end point (ca. 570 s) using
correlated to the same stoichiometry of reducing sugars in 100 μm [3] and 10 mm [Tollens’]. The most significant differ-
2. No AgNPs were formed over all concentrations of 6 and ence in the kinetics of AgNP formation was observed for
Tollens’ surveyed. We therefore conclude that covalent at- AgNP@5. A slower rate of onset (ca. 192 s) and a signifi-
tachment of galactose sugars to the triazole scaffold is es- cantly slower end point (ca. 1200 s) was observed for
sential for these ligands to facilitate AgNP formation. the formation of AgNP@5 using 100 μm [5] and 10 mm
Table 1 summarises the results of AgNPs formed in this [Tollens’].
study.
Table 1. Summary of AgNPs formed using ligands 2, 3 and 5.
AgNPs Formed Using Ligand 5 are Stable in High-Salt
Ligand
Tollens’/ligand
AgNP size [nm]
Buffers
2
1 mm:1 mm
21Ϯ3
The stability of AgNP@2 and AgNP@5 in aqueous solu-
tions that contain NaCl was then monitored. AgNP@2 re-
mained dispersed in aqueous solutions up to 0.3 m NaCl
(
1
(
5
1:1)
0 mm:25 mm
2:5)
0 mm:25 mm
2:1)
17Ϯ3
18Ϯ5
(
Figure S9a in the Supporting Information), whereas
(
AgNP@5 remained dispersed in 4.5 m NaCl at room tem-
perature over a period of 24 hours (Figure S9b). We con-
clude that the larger ligand (5) significantly enhances the
stability of AgNP colloid in high-salt buffer solutions.
3
5
10 mm:25 mm
2:5)
0 mm:1 mm
20:1)
0 mm:25 mm
2:1)
33Ϯ7
14Ϯ3
12Ϯ3
(
2
(
5
(
Binding Characteristics of Ligands with AgI
20 mm:1 mm
20:1)
0 mm:10 μm
2000:1)
0 mm:5 mm
10:1)
0 mm:100 μm
500:1)
12Ϯ1
15Ϯ2
13Ϯ2
20Ϯ2
(
2
(
5
1H NMR Spectroscopic Titration Studies
1
An H NMR spectroscopic titration study using ligands
(
5
(
1–3 and 6–7 in the presence of up to six equivalents of
I
AgNO was conducted to gain further insight into the Ag -
3
binding characteristics. Previous work revealed a large
c
d
downfield shift (δ = 0.2 ppm) for both H and H in 1 upon
I [23]
the addition of one equivalent of Ag .
This downfield
d
I
shift of H reached a maximum at an Ag /1 ratio of 1:1,
whereas the downfield shift of H reached its maximum at
a 2:1 Ag /1 ratio. This suggested that in the presence of one
equivalent of Ag , all three triazole units were participating
in the coordination of a single Ag cation. A further in-
crease in Ag changed the chelating behaviour of 1 to poss-
The Kinetics of AgNP Formation
c
I
The kinetics of AgNP formation was then explored using
–5 by monitoring the onset of the surface plasmon peak
I
2
I
at 400 nm. The onset of formation of AgNP@(2) was ob-
served at approximately 180 s with an end point at approxi-
mately 420 s using 100 μm [2] and 10 mm [Tollens’] (Fig-
ure 4). A similar kinetic profile was observed for the forma-
tion of AgNP@4 using 500 μm [4] and 5 mm [Tollens’]. A
similar rate of onset of formation of AgNP@3 was also
I
ibly access a putative binuclear metallocyclic species as ob-
I
[30–32]
g
served with previous Ag -chelating triazole ligands.
f
Triazole protons (H ) and aromatic protons (H ) of 2
were used as diagnostic markers of Ag coordination (Fig-
I
ure S10 in the Supporting Information). Titration of up to
I
one equivalent of Ag resulted in a significantly smaller
f
downfield shift (δ = 0.04 ppm) of H relative to that ob-
c
d
served for H /H in 1 (Figure S10d). A further marked di-
I
vergence in the Ag -chelating behaviour between 1 and 2
was the observation of a further δ = 0.08 ppm downfield
f
I
shift of H in 2 upon the addition of six equivalents of Ag .
f
In contrast to the behaviour of H in 2, a δ = 0.16 ppm
upfield shift of the aromatic H was observed upon the ad-
dition of up to three equivalents of Ag (Figure S10f). This
g
I
behaviour is divergent from the initial upfield shift of the
e
aromatic H in 1, which was followed by the chemical shift
e
of H returning to its original position after the addition of
I
I
six equivalents of Ag . We therefore conclude that the Ag -
binding affinity of 2 is weaker and markedly different to 1.
Figure 4. Kinetics of formation of AgNP using 2 (green), 3 (pur-
ple), 4 (red) and 5 (blue) as monitored by the formation of the
surface plasmon peak at 400 nm.
1
An H NMR spectroscopic titration was then conducted
I
using ligand 3. A similar Ag -chelating behaviour to 2 was
Eur. J. Inorg. Chem. 2014, 4886–4895
4892
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
h
j
i
I
observed using triazole H /H and aromatic protons H of Table 2. Ag -binding affinity of ligands 1–3 and 6–7 calculated by
3
1
I
nonlinear curve fitting of H NMR spectroscopic chemical-shift
data of the triazole protons using the WinEQNMR2 software pro-
gram.
as diagnostic markers of Ag coordination (Figure S11 in
the Supporting Information). Titration of up to one equiva-
lent of Ag resulted in a downfield shift (δ = 0.04 ppm) of
H /H , similar to that observed for H in 2 (Figure S11d,f).
A further δ = 0.1 ppm downfield shift was observed up to
the addition of six equivalents of Ag . A δ = 0.18 ppm up-
field shift of the aromatic H was observed upon the ad-
dition of up to three equivalents of Ag (Figure S11h), again
consistent with that observed for H in 2.
To compare the differences in Ag binding of the two
core scaffolds in more detail, H NMR spectroscopic ti-
trations were conducted by using 6 and 7. Ligand 6 is repre-
sentative of the basic core of the second-generation ligand
I
–
1
h
j
f
Compound
K [m ]
logKa
a
1
2
3
6
7
72393Ϯ2
533Ϯ1
743Ϯ1
794Ϯ1
489Ϯ1
4.859Ϯ0.219
2.727Ϯ0.018
2.871Ϯ0.040
2.900Ϯ0.057
2.689Ϯ0.012
I
i
I
g
I
1
I
Mass Spectrometric Analysis of the Ag Binding Using
Ligands 2, 3 and 5
2
, whereas 7 is representative of the first-generation ligand
ESI-MS analysis of 2 in complex with one equivalent
I
(1). Figures S12 and S13 in the Supporting Information of Ag resulted in the predominant formation of the [M +
I
+
I
show a comparative analysis of the Ag -binding characteris- Ag] ion (m/z 962). A minor amount of the 2:1 2/Ag com-
tics of ligand 6 and 7 by H NMR spectroscopic titration plex [2M + Ag + H] at m/z 1819 (Figure S15a in the Sup-
with AgNO . Titration of up to one equivalent of Ag re- porting Information) was observed, which was indicative of
sulted in a δ = 0.04 ppm downfield shift of H in 6. A fur- the formation of a metallocyclic species.
ther δ = 0.1 ppm downfield shift was observed upon the Ag] ion still predominated in the presence of two equiva-
addition of a further three equivalents of Ag (Figure S12d). lents of Ag ; however, the formation of a small amount of
For ligand 7, titration of up to one equivalent of Ag , a δ [M + 2Ag] at m/z 1070 was also observed (Figure S15b).
=
1
+
I
3
k
[30–32]
The [M +
+
I
I
I
+
m
I
0.06 ppm downfield shift of H and a δ = 0.02 ppm Increasing the number of equivalents of Ag up to four
o
+
downfield shift of H was observed. A further downfield afforded a range of ions that corresponded to [M + Ag] ,
shift of δ = 0.1 ppm in H and δ = 0.07 ppm in H occurred [M + 2Ag] and [M + 3Ag] , but with the 1:1 complex still
m
o
+
+
I
upon the addition of a further three equivalents of Ag
Figure S13d,f). A δ = 0.2 ppm upfield shift of the aromatic equivalent of Ag , ligand 2 predominantly forms the 2·AgI
protons of both ligands was observed upon the addition of complex with a 1:1 stoichiometry. Increasing the number of
predominating (Figure S15d). Thus, in the presence of one
I
(
I
I
up to three equivalents of Ag (Figure S12f and S13h). We equivalents of Ag resulted in the formation of a range of
conclude that 6 and 7 bind to Ag in a similar manner.
I
I
other minor Ag -chelating species.
I
ESI-MS analysis of the Ag -chelating properties was then
I
Calculation of the Ag -Binding Constant
+
investigated using ligand 3. A molecular ion [M + Na]
I
The binding affinity of Ag for ligands 1–3 and 6–7 was with m/z 774 was the predominant species when one equiva-
I
then calculated by nonlinear least-squares fitting (Table 2). lent of Ag was added to 3. A minor amount of [M +
The acquired H NMR spectroscopic data of the downfield Ag] at m/z 860 was also observed (Figure S16a in the Sup-
shift observed for the triazole protons of 1–3 and 6–7, and porting Information). The [M + Ag] peak was the pre-
the concentration of the Ag was used to calculate the Ag
1
+
+
I
I
I
dominant species in the presence of two equivalents of Ag
I
-binding constants using the WinEQNMR2 software (Fig- (Figure S16b). As the number of equivalents of Ag was
[33]
+
ure S14 in the Supporting Information). The ligand that increased up to four, the [M + Ag] ion still predominated
exhibited the highest affinity when one equivalent of Ag
was added was 1 with a binding constant of (72393Ϯ2) m . 966 and a complete loss of [M + Na] at m/z 774 (Fig-
I
+
with the formation of a small amount of [M + 2Ag] at m/z
–1
+
This is two orders of magnitude higher than the binding ure S16d).
I
constants calculated for ligands 2–7. An interesting obser-
ESI-MS analysis of the Ag -chelating properties was then
vation was the third galactose sugar in the southernmost carried out by using ligand 5. The major molecular ion was
–
1
2+
part of ligand 2 [(533Ϯ1) m ] has a slightly detrimental m/z 1205, which corresponded to [M + 2Ag] when one
I
–1
I
effect on Ag binding relative to 3 [(743Ϯ1) m ]. The equivalent of Ag was added to 5. A series of other minor
higher binding constant of 6 [(794Ϯ1) m ] relative to 7 adducts was also observed in the mixture, such as m/z 2303
(489Ϯ1) m ] was also unexpected and shows that the core [M + Ag] and 839 [M + 3Ag] (Figure S17a in the Sup-
–
1
–
1
+
3+
[
structure of the second-generation ligands (i.e., 2 and 3) porting Information). The wide range of molecular ions ob-
I
have a higher affinity for Ag than the core of 1. When served in this ligand complex was in stark contrast to the
+
I
the galactose sugars replaced aliphatic alcohol chains (i.e., predominant [M + Ag] observed in the 1:1 2·Ag complex.
I
7
Ǟ1), the binding affinity increased massively, whereas As the number of equivalents of Ag was increased from
only a moderate increase in binding affinity was observed one up to six, the [M + 3Ag]3 peak became more pro-
+
I
3+
when the structure was changed from 6Ǟ3. The significant nounced (Figure S17d). At 7.5 equiv. of Ag , [M + 3Ag]
reduction in Ag -binding affinity of 7 relative to ligand 1 was the dominant peak in the spectrum with a complete
suggests that the two northernmost galactose sugars in 1 loss of [M + Ag] at m/z 2303 (Figure S17e). In summary,
I
+
I
I
collectively enhance Ag binding.
the Ag -binding behaviour of ligand 5 diverges quite mark-
Eur. J. Inorg. Chem. 2014, 4886–4895
4893
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
edly from 2 and 3 by its ability to chelate a much wider with 7. Both ligands contain the same core structure. When
I
I
dynamic range of Ag ions. This Ag binding behaviour also the northeast and northwest triazoles comprise an alkyl
varies according to the stoichiometric ratio of ligand 5.
chain and a primary alcohol (i.e., 7), there is a significant
reduction in Ag -binding affinity than when a galactose is
I
installed in this position. This result suggests that these
sugars contribute to Ag binding although the nature of this
I
Discussion
interaction is unclear at present. We showed that ligand 26
prepared in our previous study did not facilitate the forma-
tion of AgNPs and produced Ag aggregates instead.[ This
indicates that the southern triazole is essential in facilitating
AgNP formation. Further structural studies will be re-
quired to understand the role of these galactose sugars in
These experiments were designed to test how the struc-
tural features of triazole ligands influenced the size, mor-
phology, stability and dispersity of AgNPs. We discuss here
several conclusions that have emerged from our results.
23]
I
enhancing Ag binding.
The Size of AgNPs Formed Using Ligand 3 is Inherently
Tunable
I
Ligand 1 has an Ag -binding affinity 136 times higher
than ligand 2 and 97 times higher than ligand 3 (Table 2).
I
As a consequence of this significantly higher Ag -binding
affinity, the rate of nucleation of 1 (120 s)[ is faster than
in 2 (180 s). We speculate that while this increase in binding
affinity in 1 increases the rate of Ag nucleation, the rate
of growth of these nuclei into AgNP@1 slows relative to
23]
Changing the southern ligand core from 1 to 2/3 also
resulted in a significant reduction in Ag binding. This was
I
[
23]
AgNP@2 as reflected by the longer endpoint of 1 (588 s)
unexpected as ligands 2 and 3 retain the same features in
the northern part of the structure as 1 and provides further
justification that the southern half of the structure plays
just as significant a role in tuning the Ag -binding affinity
as the galactose sugars in the northern part of the structure.
relative to 2 (420 s). In the case of AgNP@1, we assume
that the increase in Ag -binding affinity would produce a
larger population of Ag nuclei in complex with 1 relative to
I
I
2
in the initial nucleation stages; however, once these nuclei
are formed, the rate of growth of these nuclei into AgNPs
slows. It is not clear at present if this slower growth rate is
directly related to the binding affinity of Ag clusters to
these ligands, and this aspect will be a subject of further
investigation.
A significant difference in AgNP formation that we ob-
served in this study relative to our previous study was the
inherent tunability of the size of AgNP@3 relative to
AgNP@1. In the case of AgNP@3, monodispersed angular
AgNPs were produced with diameters that ranged from
The installation of a third galactose in 2 resulted in a
I
–1
slightly weaker Ag -binding affinity (533 m ) relative to the
I
–1
Ag -binding affinity of 3 (743 m ). There was some Ag mir-
ror formation observed when the reaction conditions of
AgNP@2 were surveyed (e.g., #21, #27 and #33; Figure 2).
The formation of Ag mirrors was not observed when ligand
3
was used, thus suggesting that the covalent attachment of
a third galactose in the southern part of these ligands actu-
ally has a detrimental effect on the formation of AgNP@3.
Kinetic studies revealed a similar rate of onset of AgNP@2
and AgNP@3, but the end point for 3 was slower (570 s)
than 2 (420 s; Figure 4). Thus we infer that the southern-
most galactose does increase the rate of AgNP formation.
However, at certain [2] and [Tollens’], the southernmost ga-
lactose sugar destabilises the growth of AgNP formation
and can result in Ag mirrors under certain conditions.
Taken collectively, our results suggest that merely increasing
the number of reducing equivalents (i.e., the number of ga-
lactose units) in the system does not necessarily equate to
the formation of monodispersed AgNPs with smaller dia-
meters.
(
12Ϯ3) up to (33Ϯ7) nm (Figure S3), whereas AgNP@2
produced angular NPs that were typically approximately
8 nm in diameter (Figure 3). Crucially, the diameters of
1
these AgNPs were dependent on the concentration of the
Tollens’ reagent and the ligand used. This behaviour was
not observed when ligand 1 was used to form AgNP@1. In
this case, only spherical AgNPs with a significantly smaller
diameter of (8Ϯ5) nm were formed over a wide range of
[
ligand] and [Tollens’]. We attribute the lack of tunability of
the size of AgNP@1 and the differences in shape to the
I
I
higher Ag -binding affinity of 1 and the different Ag -che-
lating behaviour of this ligand than AgNP@3. An unexpec-
ted result was the ability to tune the size of AgNP@2 rela-
tive to AgNP@3. A possible explanation for this is that the
presence of the third galactose sugar in the southern part
of ligand 2 perturbs size tunability.
Larger Triazole Ligand 5 Formed Highly Stable
Monodisperse AgNPs
Out of all the ligand systems surveyed, ligand 5 produced
the most stable suspensions in high-salt buffers. Further-
more, this ligand facilitated the formation of angular
I
The Ag Binding Affinity is Sensitive to Ligand Structure
Another unexpected result that arose from this study was AgNP@5 over much wider concentration ranges of [ligand]
I
the significant difference in Ag -binding affinity of 1 than and [Tollens’] than to ligands 2 and 3 (Figure S6a). A lack
Eur. J. Inorg. Chem. 2014, 4886–4895
4894
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[
[
2] T. Shegai, S. Chen, V. D. Miljkovic, G. Zengin, P. Johansson,
M. Kall, Nature Commun. 2011, 2.
3] M. D. Malinsky, K. L. Kelly, G. C. Schatz, R. P. Van Duyne, J.
Am. Chem. Soc. 2001, 123, 1471.
of size tunability in the formation of AgNP@5 was also
observed in the formation of AgNP@25 and suggests that
the addition of the first-generation triazole ligands on the
distal ends in 5 tends to override the influence of the central [4] C. Dose, D. Ho, H. E. Gaub, P. B. Dervan, C. H. Albrecht,
phloroglucinol core. An extremely long end point of
AgNP@5 formation relative to 2 and 3 was also observed.
This could be a contributing factor to the slightly larger
observed size of AgNP@5 than AgNP@1 and AgNP@25.
Angew. Chem. Int. Ed. 2007, 46, 8384; Angew. Chem. 2007, 119,
536.
5] J. Zheng, P. R. Nicovich, R. M. Dickson, Annual Review of
Physical Chemistry, Annual Reviews, Palo Alto, 2007, vol. 58,
p. 409.
8
[
Taken collectively, compound 5 is a highly versatile ligand [6] S. Lal, N. K. Grady, J. Kundu, C. S. Levin, J. B. Lassiter, N. J.
Halas, Chem. Soc. Rev. 2008, 37, 898.
that can be used to form size-selected, angular and stable
suspensions of AgNPs in high-salt buffers.
[
[
7] C. L. Nehl, J. H. Hafner, J. Mater. Chem. 2008, 18, 2415.
8] A. W. Clark, J. M. Cooper, Angew. Chem. Int. Ed. 2012, 51,
3562; Angew. Chem. 2012, 124, 3622.
[
9] M. Rycenga, C. M. Cobley, J. Zeng, W. Li, C. H. Moran, Q.
Zhang, D. Qin, Y. Xia, Chem. Rev. 2011, 111, 3669.
10] L. S. Nair, C. T. Laurencin, J. Biomed. Biotechnol. 2007, 3, 301.
11] K. Belser, T. Vig Slenters, C. Pfumbidzai, G. Upert, L. Mirolo,
K. M. Fromm, H. Wennemers, Angew. Chem. Int. Ed. 2009, 48,
3661; Angew. Chem. 2009, 121, 3715.
Conclusion
[
[
In this study, we have shown for the first time that the
I
structure and the Ag -binding affinity of triazole ligands
[
12] H. Wennemers, J. Pept. Sci. 2012, 18, 437.
[13] G. Upert, F. Bouillere, H. Wennemers, Angew. Chem. Int. Ed.
plays a defining role in controlling the size and shape of
I
AgNPs formed using the Tollens’ reagent as an Ag source.
2012, 51, 4231; Angew. Chem. 2012, 124, 4307.
I
Ligands that display weaker Ag -binding affinity (i.e., 2 and
[
14] M. Hanisch, M. Mackovic, N. Taccardi, E. Spiecker, R. N. K.
3) produced angular AgNPs that were larger in diameter
Taylor, Chem. Commun. 2012, 48, 4287.
than spherical AgNPs prepared by using ligand 1 with high [15] R. S. Patil, M. R. Kokate, C. L. Jambhale, S. M. Pawar, S. H.
I
Han, S. S. Kolekar, Adv. Nat. Sciences Nanosci. Nanotechnol.
Ag -binding affinity. The methodology described herein
2012, 3, 015013.
could therefore facilitate the application of size- and shape-
tunable AgNPs in areas such as drug delivery, imaging,
[
16] Y. D. Yin, Z. Y. Li, Z. Y. Zhong, B. Gates, Y. N. Xia, S. Venka-
teswaran, J. Mater. Chem. 2002, 12, 522.
diagnostics and optoelectronics. Efforts are now underway [17] L. Kvitek, M. Vanickova, A. Panacek, J. Soukupova, M. Dit-
trich, E. Valentova, R. Prucek, M. Bancirova, D. Milde, R.
Zboril, J. Phys. Chem. C 2009, 113, 4296.
18] J. Soukupova, L. Kvitek, A. Panacek, T. j. Nevecna, R. Zboril,
Mater. Chem. Phys. 2008, 111, 77.
to investigate the application of this methodology in these
areas and will be reported in due course.
[
[
19] L. Kvitek, A. Panacek, J. Soukupova, M. Kolar, R. Vecerova,
R. Prucek, M. Holecova, R. Zboril, J. Phys. Chem. C 2008,
Experimental Section
112, 5825.
[
20] A. Panacek, L. Kvitek, R. Prucek, M. Kolar, R. Vecerova, N.
Pizurova, V. K. Sharma, T. j. Nevecna, R. Zboril, J. Phys.
Chem. B 2006, 110, 16248.
For detailed experimental information see the Supporting Infor-
mation.
[
[
[
[
[
21] L. Kvitek, R. Prucek, A. Panacek, R. Novotny, J. Hrbac, R.
Zboril, J. Mater. Chem. 2005, 15, 1099.
Supporting Information (see footnote on the first page of this arti-
cle): Details on the synthesis of triazole ligands, protocols for the
synthesis and characterisation of AgNPs, kinetics studies of AgNP
formation and the calculation of Ag-binding affinities of triazole
ligands.
22] Y. Saito, J. J. Wang, D. N. Batchelder, D. A. Smith, Langmuir
2
003, 19, 6857.
23] R. Dondi, W. Su, G. A. Griffith, G. Clark, G. A. Burley, Small
012, 8, 770.
2
24] W. E. Benet, G. S. Lewis, L. Z. Yang, D. E. P. Hughes, J. Chem.
Res. 2011, 675.
Acknowledgments
25] V. K. Sharma, R. A. Yngard, Y. Lin, Adv. Colloid Interface Sci.
2009, 145, 83.
The Engineering and Physical Sciences Research Council (EPSRC)
is thanked for an Advanced Research Fellowship (to G. A. B., grant
number EP/E055095/1) and for a postgraduate studentship (to
A. J. F.). H. A. K. thanks the University of King Abdulaziz in
Saudi Arabia for a postgraduate studentship. A. W. C. thanks the
Royal Academy of Engineering (grant number 10216/103), the
James Watt Nanofabrication Centre, and the Kelvin Nanocharact-
erisation Centre at the University of Glasgow. A. J. F. also thanks
the University of Leicester for a postgraduate studentship. The au-
thors thank Ishwar Singh for assistance in the analysis of NMR
spectra.
[26] P. Wu, A. K. Feldman, A. K. Nugent, C. J. Hawker, A. Scheel,
B. Voit, J. Pyun, J. M. J. Frechet, K. B. Sharpless, V. V. Fokin,
Angew. Chem. Int. Ed. 2004, 43, 3928; Angew. Chem. 2004, 116,
4018.
[
27] G. A. Burley, J. Gierlich, M. R. Mofid, H. Nir, S. Tal, Y.
Eichen, T. Carell, J. Am. Chem. Soc. 2006, 128, 1398.
28] I. Mallard-Favier, P. Blach, F. Cazier, F. Delattre, Carbohydr.
Res. 2009, 344, 161.
[
[29] P. Li, L. Wang, Y. Zhang, Tetrahedron 2008, 64, 10825.
[30] M. L. Gower, J. D. Crowley, Dalton Trans. 2010, 39, 2371.
[31] J. D. Crowley, P. H. Bandeen, L. R. Hanton, Polyhedron 2010,
2
9, 70.
[
[
32] J. D. Crowley, P. H. Bandeen, Dalton Trans. 2010, 39, 612.
33] M. J. Hynes, J. Chem. Soc., Dalton Trans. 1993, 311.
Received: June 25, 2014
[1] J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao, R. P.
Van Duyne, Nature Materials 2008, 7, 442.
Published Online: August 15, 2014
Eur. J. Inorg. Chem. 2014, 4886–4895
4895
© 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim