Letter pubs.acs.org/ac
Nondestructive Size Determination of Thiol-Stabilized Gold Nanoclusters in Solution by Diffusion Ordered NMR Spectroscopy Kirsi Salorinne,† Tanja Lahtinen,† Jaakko Koivisto,† Elina Kalenius,† Maija Nissinen,† Mika Pettersson,† and Hannu Hak̈ kinen*,†,‡ †
Department of Chemistry, Nanoscience Center, University of Jyväskylä, P.O. Box 35, 40014 JYU, Finland Department of Physics, Nanoscience Center, University of Jyväskylä, P.O. Box 35, 40014 JYU, Finland
‡
S Supporting Information *
ABSTRACT: Diffusion ordered NMR spectroscopy (DOSY) was used as an analytical tool to estimate the size of thiolstabilized gold nanoclusters in solution, namely, phenylethanethiol (PET) stabilized Au25(PET)18, Au38(PET)24, and Au144(PET)60. This was achieved by determining the diffusion coefficient and hydrodynamic radius from solution samples that were confirmed to be monodispersed by electrospray ionization mass spectrometry. The average cluster diameters obtained by this technique were estimated to be 1.7, 2.2, and 3.1 nm for the Au25(PET)18, Au38(PET)24, and Au144(PET)60 nanoclusters, respectively, which were shown to agree well with the average diameters of the corresponding single crystal or theoretical structures reported in the literature. Consequently, the DOSY technique is demonstrated to be a potentially valuable nondestructive tool for characterization of nanoparticle mixtures and verifying the purity of product solutions.
T
diffraction, which has been the stumbling block of utilizing this technique extensively. So far, the structure determination by Xray crystallography has been successful for only a few cluster sizes: Au 25 (SR) 18 in anionic and neutral forms, 10−12 Au38(SR)24,13 and Au102(SR)44.14 Therefore, researchers have taken up on other structural characterization techniques, such as mass spectrometric, electron microscopic, and gel electrophoresis analyses. Although all of these techniques give valuable information about the structure and composition, each of them has also some drawbacks. The challenge in MALDI mass spectrometry is to achieve sufficiently fragmentation-free spectra for reliable determination of the parent cluster size and composition, whereas the more “gentle” ESI method depends critically on the cluster ionization yield. The use of gel electrophoresis on the other hand requires samples in aqueous solutions, and for electron microscopy, high-resolution equipment is needed in order to achieve the near-atomic resolution and accurate determination of the size of the metal core. To date, NMR spectroscopy has not been exploited to its full potential in the structural characterization of monolayer protected gold nanoclusters, although it is routinely used in the characterization of organic and organometallic compounds. In comparison to the other analytical methods described above, the only requirement of using solution NMR spectroscopy is
he robust but modifiable interaction between gold and the thiol group of organosulfur compounds has been for a long time recognized as the key factor for control of size and function of stabilized gold nanoparticles for potential applications in site-specific bioconjugate labeling, sensing, drug delivery, medical therapy, molecular electronics, and nanocatalysis.1−7 Since the first reports of synthesis of relatively monodisperse monolayer protected gold nanoclusters in the few nanometer size range by Brust and co-workers8,9 in 1994− 1995, the field has experienced an immense growth. The recent strong interest in ultrasmall gold nanoclusters, with the diameter of the metal core less than 2−3 nm, stems from their intrinsic molecule-like quantum properties regarding electronic structure, optical absorption and emission, and chemical/catalytic reactivity. Although the synthesis methodology for the ultrasmall thiol-stabilized gold nanoclusters is by now well-known and consists of quite straightforward solution chemistry, obtaining truly atomically monodisperse gold nanocluster samples on the preparative scale still remains a challenge. Therefore, in order to substantiate the product distribution and outcome of the reaction, structural characterization in terms of size and structure of the produced nanoclusters is of utmost importance. The most unambiguous structural characterization method is single crystal X-ray diffraction, which not only gives information about the size of the nanocluster but also information about the exact structural composition. However, this method relies on obtaining good quality single crystals amenable to sufficient © 2013 American Chemical Society
Received: December 18, 2012 Accepted: March 18, 2013 Published: March 18, 2013 3489
dx.doi.org/10.1021/ac303665b | Anal. Chem. 2013, 85, 3489−3492
Analytical Chemistry
Letter
measured from a quartz cuvette with a 1.0 mm optical path length (Hellma-analytics QX). All spectra show the characteristic absorption peaks of the respective clusters. The mass spectrometric experiments were performed with a QSTAR Elite ESI-Q-TOF mass spectrometer equipped with an API 200 TurboIonSpray ESI source from AB Sciex (former MDS Sciex) in Concord, Ontario (Canada). Experimental details are given in the Supporting Information. DOSY NMR Measurements. Diffusion ordered NMR spectroscopy (DOSY) measurements were performed by using Bruker Avance 400 MHz spectrometer equipped with a Great 1/10 pulsed gradient unit and a direct probe at 298 K. A LED29 pulse sequence (ledbpgp2s) was used for the diffusion experiments with a sine-shape pulsed gradient duration δ (P30) of 1.0−1.2 ms incremented from 0.68 to 32.4 G cm−1 in 16 steps. The pulsed gradient separation Δ (D20) was 100 ms, the spoil gradient (P19) was set to 600−800 μs, and the eddy current delay (D21) was 5 ms. The reported diffusion coefficients were obtained using the T1/T2 relaxation module in TopSpin 2.1 software. The samples were prepared either in CDCl3 (Au25 and Au38) or CD2Cl2 (Au144), and the experiments were repeated at least two times. An average value of the determined diffusion coefficient (D, see Table 1) and viscosity (η) values of 0.537 and 0.40 mPa s for CHCl3 and CH2Cl2, respectively, were used for the hydrodynamic radius calculation according to the Stokes−Einstein equation.
that the gold nanoclusters are soluble in a deuterated solvent (aqueous or organic) and the ligand shell has an NMR active atom that can be detected (e.g., 1H, 13C, or 31P). Although a larger amount of sample is needed in order to get sufficient signal-to-noise ratio, the product, however, can be fully recovered after the measurement by simply evaporating the NMR solvent. A few groups have shown that proton NMR spectroscopy has been successively used to characterize the ligand environment15−17 and symmetry18 of thiolate protected gold nanoclusters. Moreover, NMR spectroscopy can additionally be used to determine the size of the nanoclusters by using diffusion ordered NMR spectroscopic (DOSY)19 techniques, which ultimately enable the calculation of the hydrodynamic radius of the nanocluster based on its diffusion coefficient at a given solvent. Diffusion NMR experiments were first used to characterize gold nanoclusters by Kohlmann et al.,15 but it has not been until recently that the technique has gained more interest in the field in terms of substantiating unbound ligands20 or in the estimation of the nanocluster size21−23 and sample monodispersity.24 However, in all of the examples found in the literature using DOSY NMR for the size estimation of gold nanoclusters, the sample was more or less polydisperse. Therefore, in order to emphasize and validate the applicability of the DOSY NMR technique as a reliable and convenient method for the gold nanocluster size estimation, we determined the size, i.e., the hydrodynamic radius of three wellknown monodisperse phenylethanethiol (PET) stabilized gold nanoclusters Au25(PET)18, Au38(PET)24, and Au144(PET)60.25 The hydrodynamic radius was calculated using the diffusion coefficients obtained for each cluster size from the diffusion NMR measurements. A comparison is made with the known crystal structures or theoretically predicted structures found in the literature.
Table 1. Average Diffusion Coefficients D (m2 s−1) and the Hydrodynamic Radius r (nm) of the Aux(PET)y Clusters at 298 K Da Au25(PET)18 Au38(PET)24 Au144(PET)60
■
EXPERIMENTAL SECTION Materials. The thiolate monolayer protected monodisperse Au25(PET)18/Au144(PET)60 and Au38(PET)24 clusters were prepared according to previously published synthesis procedures using phenylethanethiol (PET; SC2H4Ph) as the protecting ligand.26,27 All reagents were commercially available. The as-prepared Au25(PET)18, Au38(PET)24, and Au144(PET)60 clusters were used as obtained according to the synthesis procedures. Of note, the synthesis of Au25(PET)18 initially gives the native negatively charged [Au25(PET)18]− species with tetraoctylammonium (TOA+) as the countercation, which then slowly oxidizes in air to the neutral Au25(PET)18 species.12 Since the air oxidation is unavoidable, both species are therefore likely present, and as no further purification was done the TOA remains in the solution most probably either as a countercation or as TOA+OH− salt, as has been suggested by Jin and co-workers.12 Characterization. 1H NMR spectra were recorded on a Bruker Avance 400 MHz spectrometer at 298 K in either CDCl3 (Au25 and Au38) or CD2Cl2 (Au144) and the chemical shifts were calibrated to the residual proton resonances of the deuterated solvent. The 1H NMR spectra of Au25(PET)18 and Au38(PET)24 nanoclusters were compared with the NMR spectra reported in the literature.17,28 No reference 1H NMR spectrum was found for the Au144(PET)60 cluster. UV−vis spectra were measured in two parts using a Perkin-Elmer lambda 850 and Nicolet Magna-IR 760 -spectrometers for wavelengths 200−880 nm and 880−1100 nm, respectively. All samples were dissolved in CH2Cl2, and absorbance was
−10
4.84 × 10 3.74 × 10−10 3.56 × 10−10
r
solvent
Dsolventb
0.84 ± 0.07 1.09 ± 0.05 1.53 ± 0.06
CDCl3 CDCl3 CD2Cl2
2.09 × 10−9 2.11 × 10−9 3.37 × 10−9
a Standard deviation
