In situ reaction monitoring reveals a diastereoselective ligand exchange reaction between the intrinsically chiral Au38(SR)24 and chiral thiols

Article abstract of DOI:10.1021/ja310330m

The ligand exchange reaction between racemic Au₃₈(2-PET) ₂₄ (2-PET = 2-phenylethylthiolate) clusters and enantiopure 1,1′-binaphthyl-2,2′-dithiol (BINAS) was monitored in situ using a chiral high-performance liquid chromatography app

Full text of DOI:10.1021/ja310330m

Communication
pubs.acs.org/JACS
In Situ Reaction Monitoring Reveals a Diastereoselective Ligand
Exchange Reaction between the Intrinsically Chiral Au (SR) and
38
24
Chiral Thiols
Stefan Knoppe, Raymond Azoulay, Amala Dass, and Thomas Bu
†
†
‡
rgi*,†
̈
†
Department of Physical Chemistry, University of Geneva, 30 Quai Ernest-Ansermet, 1211 Geneva 4, Switzerland
Department of Chemistry and Biochemistry, University of Mississippi, 352 Coulter Hall, Oxford, Mississippi 38677, United States
S Supporting Information
‡
*
15
samples, which was explained by the intrinsic chirality of
these clusters.
ABSTRACT: The ligand exchange reaction between
racemic Au (2-PET) (2-PET = 2-phenylethylthiolate)
38
24
The chirality of Au (SR) is based on the fact that the
3
8
24
11,12
clusters and enantiopure 1,1′-binaphthyl-2,2′-dithiol
protecting staple motifs are arranged in a chiral fashion.
(
BINAS) was monitored in situ using a chiral high-
Two triblade fans, each composed of three dimeric staples,
protect the “poles” of the cluster. The fans have the same
handedness, thus leading to a chiral structure. The remaining
three monomeric staple motifs are arranged around the
equator. The two enantiomers of the clusters have been
performance liquid chromatography approach. In the first
exchange step, a clear preference of R-BINAS for the left-
handed enantiomer of Au (2-PET) is observed (about 4
3
8
24
times faster than reaction with the right-handed
enantiomer). The second exchange step is much slower
than the first step. BINAS substitution deactivates the
cluster for further exchange, which is attributed to
1
1,15−17
labeled as L (left-handed) and D (right-handed),
but we
prefer to use the descriptors A (anticlockwise, left-handed) and
C (clockwise, right-handed), which were proposed for the
(
stereo)electronic effects. The results constitute the first
18
description of chirality in coordination polyhedrons. L and D
example of a ligand exchange reaction in a thiolate-
protected gold cluster with directed enrichment of a
defined species in the product mixture. This may open
new possibilities for the design of nanomaterials with
tailored properties.
are usually used to describe central chirality, but the chiral
feature we discuss is due to the arrangement of the protecting
ligands.
Here we report for the first time a diastereoselective thiolate-
for-thiolate ligand exchange reaction observed when the
enantiomers of Au (2-PET) reacted with enantiopure R-
38
24
BINAS (Figure 1). rac-Au (2-PET) clusters were prepared
38
24
4,19−21
1
igand exchange reactions are commonly used for
and isolated as described previously.
The UV−vis and
1
−3
L
postsynthetic functionalization of gold clusters.
Choice
MALDI-MS spectra are in agreement with Au clusters (not
38
21−23
of ligand and control over the reaction conditions allow
shown).
The chiral high-performance liquid chromatog-
4
,5
tailoring of optical (e.g., fluorescence) and electrochemical
raphy (HPLC) shows only two peaks with the same area at
8.85 and 17.10 min, indicating monodispersity of the clusters
6
1,2
properties, incorporation into liquid crystals, and tuning of
24
7
and a true racemate. We performed ligand exchange using R-
BINAS and monitored the reaction in situ by HPLC. For
convenience, the following notation is used: (A/C-38, 24 −
solubility. Although quite detailed information on ligand
exchange has been gained recently, no attention has been paid
8
until now to the influence of the handedness of the particles or
2
x, R-x), where A/C-38 are A- or C-Au (SR) , 24 − 2x is the
38
24
clusters, despite the fact that intrinsic chirality was found to be a
9
−13
number of 2-PET ligands, and R-x denotes the number of R-
BINAS ligands.
common feature of such clusters.
For example, the crystal
structures of Au102(p-MBA) (p-MBA = para-mercaptobenzoic
44
At short reaction times (up to ca. 15 h), two new peaks
evolve (Figure 2, left), which are assigned to the exchange
products with x = 1, (A‑38, 22, R‑1) and (C‑38, 22, R‑1), in
agreement with the finding that no x = 2 species are found in
acid) and Au (2-PET) (2-PET = 2-phenylethylthiolate)
38
24
reveal a chiral arrangement of the protecting ligands on the
1
0,11
cluster surface.
racemic in the crystals.
We studied the ligand exchange reaction between Au (2-
Since the ligands are achiral, the clusters are
1
4,15
mass spectra at rather short reaction times.
Evolution of
38
peak areas with time allows us to assign the absolute
configuration of the exchanged products since we know the
handedness of the starting material. The A-enantiomer
decreases faster than the C-enantiomer. Accordingly, the peak
area of the third species, eluting at 24 min, increases faster than
the area of the fourth band, at 33 min. Thus, we assume that
PET)24 and Au (2-PET) clusters and the bidentate, axially
40
24
1
4,15
chiral 1,1′-binaphthyl-2,2′-dithiol (BINAS) ligand
circular dichroism and matrix-assisted laser desorption/
using
14
ionization mass spectrometry (MALDI-MS). It was found
that two 2-PET ligands are replaced by one BINAS ligand.
Furthermore, exchange reactions on pure Au38 and Au40
clusters revealed strong nonlinearities between arising optical
activity and the average number of chiral ligands in the cluster
Received: October 19, 2012
Published: December 6, 2012
©
2012 American Chemical Society
20302
dx.doi.org/10.1021/ja310330m | J. Am. Chem. Soc. 2012, 134, 20302−20305

Journal of the American Chemical Society
Communication
Figure 1. Schematic representation of the ligand exchange reaction between racemic Au (SCH CH Ph) and R-BINAS. The −SCH CH Ph
38
2
2
24
2
2
groups were removed for clarity. Colors: green, AuCore; yellow, AuAdatom; orange, sulfur; gray, carbon; white, hydrogen. It is assumed that BINAS
reacts selectively with the monomeric staples at the equator of the cluster (regioselectivity). The structures in the left equation represent the left-
handed A-enantiomer (A = anticlockwise), and the structures in the equation on the right represent the right-handed C-enantiomer (C = clockwise).
The products on the right-hand side are diastereomeric: A-Au (SCH CH Ph) (R-BINAS) and C-Au (SCH CH Ph) (R-BINAS) . The
38
2
2
22
1
38
2
2
22
1
structures of the Au38 core and staples were extracted from the crystal structure data in ref 11. Note that the structures containing BINAS were not
optimized and are shown here for illustration purposes.
Figure 2. (Left) Exemplary HPLC (0−72 h) of the ligand exchange reaction between rac-Au (2-PET) and R-BINAS. The detector wavelength
38
24
was set to 630 nm to avoid contribution from BINAS to the overall absorbance. Chromatograms were normalized to the peak at 8.96 min, which is
(A‑38, 24, 0). In comparison, the signal of (C‑38, 24, 0) at 17.08 min increases in intensity over time. This indicates that (A‑38, 24, 0) is consumed
faster. The peak at 24.02 min is assigned to (A‑38, 22, R‑1), and the peak at 35.53 min is expected to be (C‑38, 22, R‑1). It is obvious that
(
(
A‑38, 22, R‑1) is produced much faster than (C‑38, 22, R‑1). The signal at 63.22 min appears at longer reaction times and is tentatively assigned to
A‑38, 20, R‑2). (Right) Chromatograms prior to and at 18 and 45 h of reaction times for reaction between R-BINAS and the pure enantiomers.
Black traces, A-Au (2-PET) ; red traces, C-Au (2-PET) . The chromatograms are complementary and serve as further proof for the assignment of
38
24
38
24
the reaction products.
the third species is (A‑38, 22, R‑1) and the fourth species is
C‑38, 22, R‑1). At increasing reaction times, a new band
Table 1. Retention Times and Assignment of the Bands
Observed in HPLC of the Ligand Exchange between rac-
Au (2-PET) and R-BINAS
(
centered at 63 min was observed. This band is broad and low in
intensity. We assume that this signal is the second exchange of
the A-enantiomer, (A‑38, 20, R‑2). Since the exchange with the
C-enantiomer is slow, the corresponding x = 2 species is not
observed, even at extended reaction times. In order to validate
the peak assignment made above, the reactions were repeated
3
8
24
peak no.
retention time (min)
assignment
1
2
3
4
5
8.96
17.08
24.02
35.53
63.22
(A-38, 24, 0)
(C-38, 24, 0)
(A-38, 22, R-1)
(C-38, 22, R-1)
(A-38, 20, R-2)
24
starting from the separated enantiomers. Since we do not
expect racemization or interconversion during the reaction, the
chromatograms should be complementary, depending on the
enantiomer of Au used in the reaction. This was indeed found
38
(
Figure 2, right) and thus confirms the assignment of the peaks,
The kinetics can be analyzed quantitatively, assuming that, at
the detection wavelengths of the HPLC (630 nm), the
exchanged and unexchanged species have the same absorption
coefficient. This seems justified, as the incorporation of BINAS
has only minor influence on the UV−vis spectra at longer
which are listed in Table 1.
The system studied can be described by the following two
consecutive independent reactions:
15
(
A‐38, 24, R‐0) + 2R‐BINAS → (A‐38, 22, R‐1)
R‐BINAS → (A‐38, 20, R‐2)
wavelengths. The rate constants k and k for the first and
1 2
second ligand exchanges (x = 0 → x = 1 and x = 1 → x = 2) for
the two reaction pathways were determined. In the following,
the rate constants are highlighted with the handedness of the
+
and
A
/2
C
1/2
two starting enantiomers (k1 and k , respectively). At high
excesses of the BINAS ligand with respect to the clusters, we
can assume its concentration is constant, and the reaction is of
pseudo-first-order with respect to the cluster.
(
C‐38, 24, R‐0) + 2R‐BINAS → (C‐38, 22, R‐1)
R‐BINAS → (C‐38, 20, R‐2)
+
2
0303
dx.doi.org/10.1021/ja310330m | J. Am. Chem. Soc. 2012, 134, 20302−20305

Journal of the American Chemical Society
Communication
The total concentrations in a peak set were normalized to
unity, and the relative concentrations were determined and
used for a fit, shown in Figure 3 (nonlinear curve-fitting, least-
agreement with the observed relative peak areas of
(A‑38, 22, R‑1) and (C‑38, 22, R‑1) in the HPLC, which is
about 4:1. This corresponds to a diastereomeric excess of 60%.
To the best of our knowledge, this is the first time that
diastereoselective ligand exchange in thiolate-protected gold
clusters has been demonstrated. For the faster-reacting isomer
(
left-handed cluster), the rate constants for the second
exchange could also be determined. The first exchange is ca.
5
0
.8 times faster than the second exchange step (0.0514 vs
.0088 h ), which is a surprisingly large difference. One part of
−1
this difference arises due to the fact that, after the first
exchange, one site is already occupied. The number of
equivalent exchange sites for BINAS is three if exchange
takes place at the monomeric staples and six if it takes place at
the dimeric staples. This leads to statistical factors of 2/3 and
5
/6, respectively, for the second exchange with respect to the
first one. Taking this statistical factor into account, the second
exchange constant should be 0.0514 × 2/3 = 0.034 h− for
exchange at the monomeric staples, which is still a factor of 3.9
higher than the measured rate. This factor of almost 4 is
therefore attributed to steric and/or electronic reasons,
meaning that the presence of one BINAS ligand slows down
the intrinsic reaction rate constant considerably.
1
We would like to point out that the measured preference of
R-BINAS toward the left-handed enantiomer of Au (SR) is
3
8
24
based on kinetic analysis of the reaction. The rate constants do
not allow conclusions on the relative stability of the
diastereomeric products formed. However, the observed rates
could be a good test for the recently proposed generalized
8
theoretical model of the ligand exchange.
In conclusion, the ligand exchange between racemic Au (2-
3
8
PET)24 and R-BINAS was shown to be diastereoselective,
leading to preferential formation of (A‑38, 22, R‑1) compared
to (C‑38, 22, R‑1). The exchange is about 4 times faster for the
preferred enantiomer, leading to a diastereomeric excess of
about 60%. The rate of the second exchange step is drastically
lower (ca. 5.8 times). We ascribe this largely to electronic and/
or steric effects. Ligand exchange reactions generally lead to a
distribution of exchange products, but this work shows that
exchange products such as A-Au (SCH CH Ph) (R-BINAS)
1
can be enriched in a mixture, which opens new possibilities for
the design and preparation of atomically precise nanoscale
objects. Future work will be devoted to the characterization of
such species.
Figure 3. Kinetic fits for the A (top) and C (bottom) pathways. The
dots are the values as determined from HPLC experiments, the traces
are the corresponding fits (compare text). Black dots refer to the initial
cluster, whereas red dots represent the first exchange product. The
latter has a maximum at ca. 40 h for the A-pathway, after which the
rate of the second exchange step, leading to (A‑38, 20, R‑2), begins to
dominate the reaction. In direct comparison, it becomes obvious that
the reaction is faster for the left-handed (A) isomer. The
concentrations of (38, 24, 0) and (38, 22, R‑1) are equal at 14.5
and 54 h, respectively (intersection of the curves).
38
2
2
22
squares using MATLAB R2011B from MathWorks, Natick,
MA; the program code is included in the SI). The extracted rate
constants are listed in Table 2. Since the reaction is drastically
ASSOCIATED CONTENT
■
*
S
Supporting Information
−1
Table 2. First-Order Rate Constants (in h ) for the Ligand
Synthesis of R-BINAS, synthesis and size-selection of rac-
Au (2-PET) , HPLC method used for in situ monitoring, and
Exchange between rac-Au (2-PET) and R-BINAS for the
3
8
24
38
24
First and Second Exchange Steps
the program code of the Matlab routine. This material is
available free of charge via the Internet at http://pubs.acs.org.
A pathway
C pathway
k1
k2
0.0514 ± 0.0019 (3.8%)
0.0088 ± 0.0006 (6.9%)
0.0128 ± 0.0003 (2.6%)
not interpretable
AUTHOR INFORMATION
■
Corresponding Author
thomas.buergi@unige.ch
slower for the C-isomer of the cluster, the errors of the
corresponding rate constants are bigger, especially k . The
value for k therefore could not be determined. The other rate
constants (k , k , and k ) can be determined with errors below
C
2
Notes
C
2
The authors declare no competing financial interest.
A
1
A
2
C
1
A
−1
C
−1
7
%. Comparison of k (0.0514 h ) and k (0.0128 h ) shows
1
1
ACKNOWLEDGMENTS
that the reaction is about 4 times faster for the A-isomer,
indicating that the reaction is diastereoselective (one of the
diastereomers of the product is preferably formed). This is in
■
We thank Dr. Igor Dolamic for help with the HPLC work. We
acknowledge financial support from the Swiss National
2
0304
dx.doi.org/10.1021/ja310330m | J. Am. Chem. Soc. 2012, 134, 20302−20305

Journal of the American Chemical Society
Communication
Foundation, the University of Geneva (S.K., R.A., T.B.), and
the University of Mississippi (A.D.).
REFERENCES
1) Cseh, L.; Mehl, G. H. J. Mater. Chem. 2007, 17, 311.
■
(
(
2) Frein, S.; Boudon, J.; Vonlanthen, M.; Scharf, T.; Barbera,
ss-Fink, G.; Burgi, T.; Deschenaux, R. Helv. Chim. Acta 2008, 91,
321.
3) Hostetler, M. J.; Green, S. J.; Stokes, J. J.; Murray, R. W. J. Am.
Chem. Soc. 1996, 118, 4212.
4) Shibu, E. S.; Muhammed, M. A. H.; Tsukuda, T.; Pradeep, T. J.
Phys. Chem. C 2008, 112, 12168.
5) Wang, G.; Huang, T.; Murray, R. W.; Menard, L.; Nuzzo, R. G. J.
Am. Chem. Soc. 2005, 127, 812.
6) Parker, J. F.; Kacprzak, K. A.; Lopez-Acevedo, O.; Hak
Murray, R. W. J. Phys. Chem. C 2010, 114, 8276.
7) Pengo, P.; Polizzi, S.; Battagliarin, M.; Pasquato, L.; Scrimin, P. J.
Mater. Chem. 2003, 13, 2471.
8) Heinecke, C. L.; Ni, T. W.; Malola, S.; Makinen, V.; Wong, O. A.;
Hakkinen, H.; Ackerson, C. J. J. Am. Chem. Soc. 2012, 134, 13316.
́
J.;
Su
̈
̈
2
(
(
(
(
̈
kinen, H.;
(
(
(
9) Hakkinen, H. Nat. Chem. 2012, 4, 443.
(
10) Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.;
Kornberg, R. D. Science 2007, 318, 430.
11) Qian, H.; Eckenhoff, W. T.; Zhu, Y.; Pintauer, T.; Jin, R. J. Am.
Chem. Soc. 2010, 132, 8280.
12) Lopez-Acevedo, O.; Tsunoyama, H.; Tsukuda, T.; Hakkinen,
H.; Aikens, C. M. J. Am. Chem. Soc. 2010, 132, 8210.
13) Tlahuice, A.; Garzon, I. L. Phys. Chem. Chem. Phys. 2012, 14,
737.
14) Knoppe, S.; Dharmaratne, A. C.; Schreiner, E.; Dass, A.; Burgi,
T. J. Am. Chem. Soc. 2010, 132, 16783.
(
(
(
3
(
(
(
15) Knoppe, S.; Dass, A.; Burgi, T. Nanoscale 2012, 4, 4211.
16) Qian, H.; Zhu, M.; Wu, Z.; Jin, R. Acc. Chem. Res. 2012, 45,
1
470.
17) Knoppe, S.; Dolamic, I.; Burgi, T. J. Am. Chem. Soc. 2012, 134,
3114.
18) Brown, M. F.; Cook, B. R.; Sloan, T. E. Inorg. Chem. 1975, 14,
273.
19) Knoppe, S.; Boudon, J.; Dolamic, I.; Dass, A.; Burgi, T. Anal.
Chem. 2011, 83, 5056.
20) Qian, H.; Zhu, M.; Andersen, U. N.; Jin, R. J. Phys. Chem. A
009, 113, 4281.
(
1
(
1
(
(
2
(
21) Qian, H.; Zhu, Y.; Jin, R. ACS Nano 2009, 3, 3795.
(
22) Chaki, N. K.; Negishi, Y.; Tsunoyama, H.; Shichibu, Y.;
Tsukuda, T. J. Am. Chem. Soc. 2008, 130, 8608.
23) Toikkanen, O.; Ruiz, V.; Ronnholm, G.; Kalkkinen, N.;
Liljeroth, P.; Quinn, B. M. J. Am. Chem. Soc. 2008, 130, 11049.
24) Dolamic, I.; Knoppe, S.; Dass, A.; Burgi, T. Nat. Commun. 2012,
, 798.
(
̈
(
̈
3
2
0305
dx.doi.org/10.1021/ja310330m | J. Am. Chem. Soc. 2012, 134, 20302−20305