One-phase synthesis of water-soluble gold nanoparticles with control over size and surface functionalities

Article abstract of DOI:10.1021/la904438s

We report a simple and efficient synthetic method to prepare gold nanoparticles (AuNPs) in aqueous phase using HAuCl₄ and poly(ethylene glycol) (PEG) ligands appended with bidentate anchoring groups. Our approach provides narrow size distribution nanocrystals over the size range between 1.5 and 18 nm; this range is much wider than those achieved using other small molecules and polymer ligands. The NP size was simply controlled by varying the molar ratio of Au-to-PEG ligand precursors. Further passivation of the as-prepared AuNPs permitted in situ functionalization of the NP surface with the desired functional groups. The prepared AuNPs exhibit remarkable stability in the presence of high salt concentrations, over a wide range of pHs (2-13), and a strong resistance to competition from dithiothreitol (DTT). These results are a clear manifestation of the advantages offered by our synthetic approach to prepare biocompatible AuNPs, where modular, multifunctional ligands presenting strong anchoring groups and hydrophilic PEG chains are used.

Full text of DOI:10.1021/la904438s

pubs.acs.org/Langmuir
©
2010 American Chemical Society
One-Phase Synthesis of Water-Soluble Gold Nanoparticles with Control over
Size and Surface Functionalities
†
†
‡
,†,§
Eunkeu Oh, Kimihiro Susumu, Ramasis Goswami, and Hedi Mattoussi*
†
‡
Division of Optical Sciences, Code 5611 and Materials Science Division, Code 6355, U.S. Naval Research
§
Laboratory, Washington, D.C. 20375. Present address: Department of Chemistry and Biochemistry,
Florida State University, 4006 Chemical Sciences Building, Tallahassee, Florida 32306.
Received November 23, 2009. Revised Manuscript Received January 7, 2010
We report a simple and efficient synthetic method to prepare gold nanoparticles (AuNPs) in aqueous phase using
HAuCl and poly(ethylene glycol) (PEG) ligands appended with bidentate anchoring groups. Our approach provides
4
narrow size distribution nanocrystals over the size range between 1.5 and 18 nm; this range is much wider than those
achieved using other small molecules and polymer ligands. The NP size was simply controlled by varying the molar ratio
of Au-to-PEG ligand precursors. Further passivation of the as-prepared AuNPs permitted in situ functionalization of the
NP surface with the desired functional groups. The prepared AuNPs exhibit remarkable stability in the presence of high
salt concentrations, over a wide range of pHs (2-13), and a strong resistance to competition from dithiothreitol (DTT). These
results are a clear manifestation of the advantages offered by our synthetic approach to prepare biocompatible AuNPs, where
modular, multifunctional ligands presenting strong anchoring groups and hydrophilic PEG chains are used.
7
,8
Introduction
(e.g., DNA hybridization), (3) surface-enhanced Raman scat-
9
,10
tering (SERS) nanoparticle tags for in vivo spectroscopy,
4) photoinduced heating of Au nanoparticles and nanorods for
localized photothermal therapy.
dent uptake of AuNPs by live cells was studied for sizes ranging
AuNPs are also efficient quenchers of
dyes and luminescent quantum dots (via resonance energy
transfer) when brought in close proximity.
tions often require that the AuNPs are (1) available in a broad
size range, (2) stable over a broad range of conditions, and (3)
easy to surface functionalize with specific reactive groups for
further coupling to target molecules.
and
Gold nanoparticles (AuNPs) along with an array of other
metallic and semiconducting nanocrystals with their size-depen-
dent physical and chemical properties have generated a tremen-
dous interest in the past decade. This interest has also been driven
by the great potential for applications in optical and electronic
(
1
1,12
Recently, the size-depen-
1
3-15
from 2 to 100 nm.
1
-3
devices and more recently in biological sensing and imaging.
1
6,17
These applica-
AuNPs with size larger than 2 nm (in diameter) exhibit size- and
shape-dependent surface plasmon absorption band (SPB);
AuNPs smaller than 2 nm, in comparison, exhibit photoinduced
4
emission. The SPB is also sensitive to proximity-driven NP-to-
NP interactions and to the surrounding environment, including
surface ligands, solvent, and temperature. Thus, red shift of the
SPB coupled with broadening of the peak width has been
observed in the presence of NP aggregation, which can be
triggered by interactions with a particular target or in response
to an external stimulus (changes in pH or excess ion con-
centrations). Larger size AuNPs and nanorods can also trans-
form laser excitation to local thermal heating. These unique
properties have provided researchers in biology with powerful
methods to develop sensing techniques that include (1) colori-
Since the classic citrate reduction of aurate to prepare citrate-
stabilized AuNPs was pioneered by the groups of Turkevich and
18,19
Frens,
there has been a sustained effort aimed at develop-
ing new chemical routes to prepare AuNPs that are stable and
easily dispersible in water. These efforts have also focused on
designing synthetic schemes to functionalize the AuNP surfaces
2
0
with specific target ligands. Some of the most common methods
(
9) Qian, X. M.; Peng, X. H.; Ansari, D. O.; Yin-Goen, Q.; Chen, G. Z.; Shin,
1
,5,6
metric assays to detect NP assemblies,
(2) molecular rulers
D. M.; Yang, L.; Young, A. N.; Wang, M. D.; Nie, S. M. Nat. Biotechnol. 2008, 26,
8
3–90.
based on plasmon coupling to detect bridging between NP pairs
(10) Qian, X. M.; Nie, S. M. Chem. Soc. Rev. 2008, 37, 912–920.
11) Huang, X. H.; Jain, P. K.; El-Sayed, I. H.; El-Sayed, M. A. Lasers Med. Sci.
(
2
008, 23, 217–228.
*To whom correspondence should be addressed. E-mail: mattoussi@
(12) Jain, P. K.; Huang, X. H.; El-Sayed, I. H.; El-Sayed, M. A. Acc. Chem. Res.
chem.fsu.edu.
2008, 41, 1578–1586.
(
1) Daniel, M. C.; Astruc, D. Chem. Rev. 2004, 104, 293–346.
(13) Jiang, W.; Kim, B. Y. S.; Rutka, J. T.; Chan, W. C. W. Nat. Nanotechnol.
(
2) Zheng, J.; Nicovich, P. R.; Dickson, R. M. Annu. Rev. Phys. Chem. 2007, 58,
2008, 3, 145–150.
4
09–431.
(14) Wang, S. G.; Lu, W. T.; Tovmachenko, O.; Rai, U. S.; Yu, H. T.; Ray, P. C.
Chem. Phys. Lett. 2008, 463, 145–149.
(15) Perrault, S. D.; Walkey, C.; Jennings, T.; Fischer, H. C.; Chan, W. C. W.
(
3) De, M.; Ghosh, P. S.; Rotello, V. M. Adv. Mater. 2008, 20, 4225–4241.
4) Schaeffer, N.; Tan, B.; Dickinson, C.; Rosseinsky, M. J.; Laromaine, A.;
(
McComb, D. W.; Stevens, M. M.;Wang, Y. Q.; Petit, L.; Barentin, C.;Spiller, D. G.;
Nano Lett. 2009, 9, 1909–1915.
Cooper, A. I.; Levy, R. Chem. Commun. 2008, 3986–3988.
(16) Dulkeith, E.; Morteani, A. C.; Niedereichholz, T.; Klar, T. A.; Feldmann,
J.; Levi, S. A.; van Veggel, F. C. J. M.; Reinhoudt, D. N.; Moller, M.; Gittins, D. I.
Phys. Rev. Lett. 2002, 89, 203002.
(17) Oh, E.; Hong, M. Y.; Lee, D.; Nam, S. H.; Yoon, H. C.; Kim, H. S. J. Am.
Chem. Soc. 2005, 127, 3270–3271.
(
5) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature 1996, 382,
6
5
07–609.
(
6) Lee, J. S.; Ulmann, P. A.; Han, M. S.; Mirkin, C. A. Nano Lett. 2008, 8, 529–
33.
(
7) Sonnichsen, C.; Reinhard, B. M.; Liphardt, J.; Alivisatos, A. P. Nat.
Biotechnol. 2005, 23, 741–745.
8) Reinhard, B. M.; Siu, M.; Agarwal, H.; Alivisatos, A. P.; Liphardt, J. Nano
Lett. 2005, 5, 2246–2252.
(18) Turkevich, J.; Stevenson, P. C.; Hillier, J. Discuss. Faraday Soc. 1951, 55–.
(19) Frens, G. Nat. Phys. 1973, 241, 20–22.
(20) Weisbecker, C. S.; Merritt, M. V.; Whitesides, G. M. Langmuir 1996, 12,
3763–3772.
(
7
604 DOI: 10.1021/la904438s
Published on Web 02/02/2010
Langmuir 2010, 26(10), 7604–7613

Oh et al.
Article
include the following: (1) Growth of 1.5-5.2 nm AuNPs surface-
We have also shown that both TA-PEG- and DHLA-PEG-based
ligands can be used to cap-exchange citrate-stabilized NPs. In
addition, we have demonstrated that DHLA-PEG ligands pro-
vide effective surface functionalization of semiconductor quan-
tum dots and promote their transfer to buffer media. Further-
more, the bidentate nature of the anchoring group provides better
stability to both types of nanocrystals compared with their
functionalized with alkanethiol using two-phase toluene/water
2
1,22
reaction developed by Brust and co-workers;
this route
produced hydrophobic nanoparticles. (2) Use of poly(ethylene
glycol) oligomer (PEG) containing ligands combined with the
two-phase reaction developed by Brust and co-workers, to pre-
2
3
pare Au clusters first reported by Murray et al. and then later by
2
4
32,33
Brust et al. The resulting NPs (water dispersible) are capped
monothiol analogues.
2
3
with either monothiol-terminated PEG (MW 5000) oligomers,
Here, we build on the synthetic rationales reported for the
preparation of thiol stabilized AuNPs combined with the nature
of our TA-PEG ligands to develop a simple one-phase (aqueous)
growth and passivation method to prepare a series of hydro-
philic AuNPs. This new route provides nanocrystals with a
broad range of sizes (1.5-18 nm in diameter), which can also
be readily functionalized with reactive end groups for further
coupling to target bioreceptors. Since these NPs are stabilized
with disulfide anchoring groups, they exhibit remarkable stabi-
lity in the presence of excess counterions, against competi-
tion with dithiothreitol (DTT), and to changes in solution
pH, similar to what was reported for Au nanoparticles cap-
exchanged with TA-PEG ligands (starting with citrate-coated
NPs) or prepared in the presence of thiol-functionalized poly-
2
4
or tetraethylene glycol ligands. (3) Use of dialkyl disulfide
2
5,26
stabilizers to prepare 1.4-3.8 nm diameter AuNPs.
(4)
Synthesis of 1.5-8 nm size of AuNPs using water-soluble alkyl
2
7,28
thioether- and thiol-functionalized poly(methacrylic acid).
(5)
29
Preparation of phosphine-stabilized 1.4 nm AuNPs; 1.4 nm
AuNPs surface-functionalized with maleimide are commercially
available (Nanoprobes, Yaphank, NY). These preparation meth-
ods reflect the tremendous progress made in the past decade. They
provided researchers with an array of AuNPs which have been
employed for developing a variety of applications. Nonetheless,
each of these synthetic schemes has encountered some limitations.
These include (1) reduced stability against excess salts and
changes in solution pH (e.g., citrate-stabilized NPs); (2) the
inability to prepare nanocrystals over a wide size regime (citrate
reduction usually produces AuNPs smaller than 10 nm, but larger
sizes require additional refluxing in the presence of sodium
citrate); (3) most NPs prepared using thiol-alkyl-Au interac-
tions are hydrophobic, and the transfer to aqueous media further
requires postsynthetic processing via ligand exchange, which can
be tedious and requires large amounts of ligands. Murphy and co-
workers combined seeding, growth, and temperature treatment in
the presence of cetyltrimethylammonium bromide (CTAB) to
synthesize larger size AuNPs and Au nanorods (5-40 nm
2
8,35
mers.
We describe the synthetic details developed and
discuss optical and structural characterization of the prepared
AuNPs using UV-vis spectroscopy, high-resolution transmis-
sion electron microcopy (HRTEM), and dynamic light scatter-
ing (DLS).
Results and Discussion
Control of Au Nanocrystal Growth. For this study, we
synthesized TA-PEG-OCH ligands with average poly(ethylene
3
glycol) methyl ether molecular weight (MW) of 550 or 750.
3
0
range). These findings combined clearly indicate that, despite
the tremendous progress, there is still a need to develop new
simple synthetic methods to prepare AuNPs over a wide range of
sizes that exhibit enhanced stability in buffer media (often rich in
ionic complexes) and which can be surface-functionalized.
We have designed and synthesized an array of modular ligands
made of a tunable length PEG segment appended with either a
thioctic acid (TA, which has a terminal disulfide) or a dihydro-
lipoic acid (DHLA, formed by reducing the terminal disulfide to
dithiol) at one end and a potentially reactive group at another end
Ligands with a PEG MW ∼ 550 (TA-PEG550-OCH ) were used
3
throughout this report, unless otherwise noted. For in situ
functionalization, we used TA-PEG600-COOH (PEG MW ∼ 600)
ligand mixed with TA-PEG-OCH during AuNP synthesis.
3
Synthesis of TA-PEG-OCH3 and TA-PEG600-COOH was
carried out following the procedures detailed in our previous
32-34
reports.
For control experiments, we used mercaptohexa-
noic-acid-appended PEG (HS-PEG550-OCH , monothiol ap-
3
pended ligand). The synthetic details are provided in the
Experimental Section. The AuNPs were prepared using a three-
step reaction consisting of (1) precursor formation by reacting the
ligands with tetrachloroauric(III) acid (HAuCl4 3H O); (2)
3
1-34
(
see Figure 1).
The synthetic procedures we have developed
are simple to implement and provide multigram scale of materials.
3
2
growth of the AuNP cores triggered by addition of NaBH
4
reducing agent; and (3) further passivation and functionalization
of the cores by adding extra free ligands, as schematically
diagrammed in Figure 1. Tetrachloroauric acid and the ligands
were first mixed in water to promote the formation of Au-TA-
(
21) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. Chem.
Soc., Chem. Commun. 1994, 801–802.
22) Brust, M.; Fink, J.; Bethell, D.; Schiffrin, D. J.; Kiely, C. J. Chem. Soc.,
Chem. Commun. 1995, 1655–1656.
23) Wuelfing, W. P.; Gross, S. M.; Miles, D. T.; Murray, R. W. J. Am. Chem.
Soc. 1998, 120, 12696–12697.
24) Kanaras, A. G.; Kamounah, F. S.; Schaumburg, K.; Kiely, C. J.; Brust, M.
Chem. Commun. 2002, 2294–2295.
(
(
PEG-OCH metal-ligand precursors. This precursor formation
3
(
manifests in a rapid color change of the original yellow solution to
(
(
(
25) Yonezawa, T.; Yasui, K.; Kimizuka, N. Langmuir 2001, 17, 271–273.
26) Shon, Y. S.; Mazzitelli, C.; Murray, R. W. Langmuir 2001, 17, 7735–7741.
27) Hussain, I.; Graham, S.; Wang, Z. X.; Tan, B.; Sherrington, D. C.;
red, yellow, and finally colorless. Addition of NaBH initiates
4
reduction of the Au ions and rapid growth of the Au nanocrystals.
We varied the molar ratio of Au-to-ligand as a means of
controlling the size of the resulting metal core. Once the growth
was complete (usually associated with a saturation in the UV-vis
absorption spectrum), free ligands were further added to the
solution (to a final Au/ligand molar ratio of ∼1:1). This last step
provided additional passivation and functionalization (when
desired) of the AuNPs by filling unoccupied surface sites. None-
theless, when NPs were grown under initial low Au-to-ligand
Rannard, S. P.; Cooper, A. I.; Brust, M. J. Am. Chem. Soc. 2005, 127, 16398–
6399.
28) Wang, Z. X.; Tan, B. E.; Hussain, I.; Schaeffer, N.; Wyatt, M. F.; Brust, M.;
Cooper, A. I. Langmuir 2007, 23, 885–895.
29) Weare, W. W.; Reed, S. M.; Warner, M. G.; Hutchison, J. E. J. Am. Chem.
Soc. 2000, 122, 12890–12891.
1
(
(
(
(
30) Jana, N. R.; Gearheart, L.; Murphy, C. J. Langmuir 2001, 17, 6782–6786.
31) Uyeda, H. T.; Medintz, I. L.; Jaiswal, J. K.; Simon, S. M.; Mattoussi, H.
J. Am. Chem. Soc. 2005, 127, 3870–3878.
32) Susumu, K.; Uyeda, H. T.; Medintz, I. L.; Pons, T.; Delehanty, J. B.;
Mattoussi, H. J. Am. Chem. Soc. 2007, 129, 13987–13996.
33) Mei, B. C.; Susumu, K.; Medintz, I. L.; Delehanty, J. B.; Mountziaris, T. J.;
Mattoussi, H. J. Mater. Chem. 2008, 18, 4949–4958.
34) Susumu, K.; Mei, B. C.; Mattoussi, H. Nat. Protoc. 2009, 4, 424–436.
(
(
(35) Mei, B. C.; Oh, E.; Susumu, K.; Farrell, D.; Mountziaris, T. J.; Mattoussi,
H. Langmuir 2009, 25, 10604–10611.
(
Langmuir 2010, 26(10), 7604–7613
DOI: 10.1021/la904438s 7605

Article
Oh et al.
Figure 1. (A) Schematic representation of the one-phase growth method using TA-PEG-OCH
3
ligands. (B) Normalized UV-vis absorption
spectra for several Au/ligand molar ratios used for the growth (of AuNPs with different sizes), along with a few representative TEM images
collected for selected subsets of these nanocrystals. (inset) Image collected from a series of AuNP dispersions in deionized water. Changes in
solution color from light to dark brown and to red (from left to right) reflect increase in AuNP size.
molar ratio (smaller than 1:1), no extra free ligands were added to
the final sample; the initial molar concentration of the ligands was
high enough to provide sufficient passivation of the nanoparticles
in this case.
vigorous stirring. To prepare different size nanoparticles, we
fixed the molar concentration of HAuCl and varied the amount
4
of the ligands used. Following addition of the reducing agent,
the color of the reaction mixture immediately changed to brown or
red depending on size. The mixture was then left stirring for at least
3 h, as it gradually progressed to its final appearance (when the
absorption spectrum reached saturation). The AuNP dispersions
(as-prepared) were characterized using UV-vis spectroscopy,
TEM, and DLS. For extra passivation (or passivation and
In a typical reaction, the following conditions were used to
-6
prepare our AuNPs. An amount of 156 μL (7.92 ꢀ 10 mol) of
5
0.8 mM tetrachloroauric(III) acid (HAuCl4 3H O) stock solu-
3
2
tion and the desired molar concentration of TA-PEG550-OCH
3
were dissolved in 25 mL of deionized water; the mixture was
then stirred at room temperature for 1 h. The stock solution of
functionalization), TA-PEG550-OCH or a mixture of either
3
-
4
(
HAuCl4 3H O) was prepared by mixing 100 mg (2.54 ꢀ 10
TA-PEG550-OCH /TA-PEG600-COOH or TA-PEG550-OCH /
3 3
3
2
mol) of AuCl in 5 mL of H O. Stirring for 1 h was based on the
TA-PEG600-NH (final ratio of Au to ligand = 1:1) and 18 μL
4
2
2
-5
anticipated minimum time necessary to promote precursor for-
(1.6 ꢀ 10 mol) of 880 mM NaBH were further added to the
4
-
5
mation (see below). An amount of 72 μL (6.3 ꢀ 10 mol) of
AuNP solution, and left stirring for an additional 3 h. The
dispersion was then purified from free ligands by three cycles of
centrifugation using a membrane filtration device (Millipore).
8
80 mM sodium borohydride (NaBH ) stock solution in deio-
4
nized water was added (in aliquots of 18 μL) over 30 min with
7
606 DOI: 10.1021/la904438s
Langmuir 2010, 26(10), 7604–7613

Oh et al.
Article
These AuNPs were then characterized and compared to those
not subjected to the extra passivation/functionalization step. In
our experimental protocol, we used 10ꢀ total molar excess of
the reducing agent during the growth and passivation steps
(
i.e., total NaBH -to-Au molar ratio ∼ 10). This value was
4
based on those used in previous studies and provided the most
effective preparation conditions including those reported using
2
1
the Brust method. Reducing this ratio to 9 or 8 does not change
the results. Lower values (e.g., ∼1ꢀ), however, do not provide
growth of homogeneous AuNP dispersions. As controls, we
synthesized thioctic acid (TA)-AuNPs, mercaptohexanoic acid
(
MHA)-AuNPs, and HS-PEG-OCH -AuNPs using the same
3
method but mixing either TA, MHA, or monothiol-appended
PEG ligands instead of TA-PEG-OCH (Figure 1A).
3
Characterization of the AuNPs. 1. UV-Vis Absorp-
tion Spectroscopy. Figure 1B shows the absorption spectra
collected from a set of AuNP dispersions prepared using an
increased molar ratio of Au to TA-PEG-OCH ; some represen-
3
tative TEM images collected from these dispersions are also
shown. The absorption spectra show that at low Au-to-ligand
molar ratio the characteristic SPB located at 520 nm is very
weak (not discernible for the smallest size NPs). However, the
peak becomes sharper and more defined at Au/ligand molar
ratiosabove 6:1. The well-defined and narrowSPB peaksfor these
dispersions (in particular for those prepared with high Au-to-
ligand molar ratios) indicate that narrow size distribution char-
acterizes these NPs. The inset shows an image of several NP
dispersions in deionized water. The dispersion color changed
from dark yellow to brown and then to red as the Au-to-TA-
PEG-OCH ligand ratio increased, indicating an increase in the
3
NP size. Data clearly indicate that our NP growth method is
capable of providing high quality nanocrystals over a broad size
range, where effective control over the NP dimensions has simply
been achieved by varying the Au/ligand ratio used during growth.
The growth reaction carried out in the presence of TA, MHA
and HS-PEG-OCH yielded only smaller size AuNPs (diameter
3
<
10 nm). In addition, dispersions prepared using high Au/ligand
ratios easily precipitated when monothiol-appended ligands were
used (see Figure S1 in the Supporting Information). For instance,
dynamic light scattering experiments indicated the presence of
aggregate buildup (∼190 nm size) for the dispersions of HS-PEG-
OCH -AuNPs (see Figure S2 in the Supporting Information).
3
The rest of the structural characterization will focus on nano-
crystals prepared using TA-PEG-OCH ligands.
3
2
. Transmission Electron Microscopy (TEM). We relied on
the use of high-resolution TEM to determine nanocrystal sizes
and assess their crystalline quality; only data collected from
TA-PEG-OCH -AuNPs are shown. Figure 2 shows the TEM
3
images collected from six representative nanocrystal dispersions
Figure 2. (A) TEM images of several different size AuNPs (scale
bar =20nm). Averagesizeandstandarddeviationarereportedfor
each sample. Samples for TEM were prepared by spreading a drop
of the AuNP dispersions onto the TEM grid (made of a holey
carbon film on a 400 mesh Cu grid). (B) IFFT of the TEM image of
prepared using a Au/TA-PEG-OCH ratio that varied from 1:10
3
to 4000:1. The images show that the size of the NPs progressively
increased with increasing Au-to-ligand molar ratio, as anticipated
from the absorption properties shown in Figure 1B. The average
diameter extracted from these TEM images together with the
Au/ligand ratios are listed in Table 1. Typically, the measured
diameters were 1.5 ( 0.3 nm (for Au/ligand=1:10), 2.2 ( 0.3 nm
2
1
.2 nm TA-PEG-OCH
3
-AuNPs (inset, upper left) and FFT of
2nmTA-PEG-OCH -AuNPs(inset, lower right).Tindicatesthe
3
twin plane in the lower right inset. Lattice spacing is about 0.24 nm
for both. (C) High-resolution image of a larger size NP showing
multiple twinning, which is characteristic of FCC structure.
(
1
1
for Au/ligand=1.4:1), 4.6 ( 0.4 nm (Au/ligand=30:1), 10 (
.4 nm (Au/ligand=600:1), 12( 1.7 nm(Au/ligand=2000:1), and
8 ( 3.3 nm (Au/ligand=4000:1), respectively. These images do
in Figure 2A). Higher resolution images for a few representative
nanocrystal sizes with different crystal orientations are shown in
Figure 2B together with the corresponding fast Fourier transfor-
mation (FFT) and inversed FFT (IFFT). The lattice spacing along
not show any sign of nanocrystal clumping on the TEM grid,
further confirming the colloidal stability of these dispersions. The
TEM images also show that spherical shapes dominate the disper-
sions of small size nanocrystals. However, slight inhomogeneities in
nanoparticle shape are observed for the larger size NPs (see panel 6
˚
the (111) planes measured for our nanocrystals was 2.4 A, in
˚
agreement with what has been previously reported (2.36 A) for
Langmuir 2010, 26(10), 7604–7613
DOI: 10.1021/la904438s 7607

Article
Table 1. Sizes Extracted from High-Resolution TEM and DLS
Oh et al.
Experiments Shown Together with the Corresponding Au-to-Ligand
Ratios Used throughout the Study
a
b
c
Au/ligand ratio
TEM (nm)
DLS (nm)
DLS (nm)
1
1
.4:1
:10
:1
1.5 ( 0.3
1.7 ( 0.2
2.2 ( 0.3
2.4 ( 0.3
3.0 ( 0.5
4.3 ( 0.6
4.6 ( 0.4
5.0 ( 0.7
7 ( 1.2
1
3:1
1
2
3
0:1
0:1
0:1
11 ( 2.4
13 ( 2.8
13 ( 2.7
19 ( 4.6
1
2
3
6
00:1
50:1
00:1
00:1
000:1
000:1
8 ( 0.9
10 ( 1.4
12 ( 1.7
18 ( 3.3
16 ( 3.6
21 ( 4.0
27 ( 8.4
21 ( 3.8
25 ( 5.3
2
4
33 ( 13
a
b
Precursor ratios used during core synthesis of AuNPs. Sizes
c
measured for as-prepared AuNP dispersions. Sizes measured after
the extra passivation step with additional ligands.
3
6
face-centered cubic (FCC) AuNPs. The TEM images also show
that twinning defects characteristic of FCC crystalline structures
are observed primarily for the larger size nanocrystals (see
Figure 2C). A plot summarizing the progression of particle size
(extracted from TEM) versus Au-to-ligand ratio for the full set of
nanocrystals prepared using our growth method is shown in
Figure 3A. To further test the quality of our AuNPs, we plotted
the progression of the extinction coefficient at 520 nm versus
diameter (based on TEM measurements) together with those
measured for commercially available citrate-stabilized AuNPs
(Ted Pella Inc.); see Figure 3B. The data are essentially identical
for both sets of NPs. The overall trend for the nanocrystal size
versus Au-to-ligand ratio compiled for our materials indicates
a rapid size increase at low ratios (Au/ligand e 10:1, diameter
e 3 nm). This dependence becomes moderate for the medium size
range (diameter ∼ 3-10 nm, Au/ligand ratio ∼ 10-600) and
eventually very weak for large size NPs (diameter > 10 nm, at Au-
to-ligand ratio exceeding ∼600). The rapid size change measured at
low ratios reflects the pronounced effects of surface-to-volume
ratios for small size NPs; this ratio is sensitive to the amounts of
ligands used. As the NP size increases and the relative surface-to-
volume ratios decreases, effects of ligand concentration become less
dominant, which translates into a weaker change in NP size at
rather large Au/ligand ratios. We should emphasize that the size
range described here is not exclusive, as slightly smaller NPs
(
∼1-1.4 nm) and larger size nanocrystals (∼20 nm) can be made.
We were not able to collect high quality TEM images for NPs
smaller than 1.5 nm, due to weaker electronic contrast and
potential interference from the ligands; we also limited our char-
acterization of the largest size to 18 nm NPs. Use of thiol-
terminated poly(methacrylic acid) (as ligand) for the synthesis
of water-soluble AuNPs has been reported by Brust and co-
Figure 3. (A) Plot of the AuNP size measured by TEM versus Au-to-
3
TA-PEG-OCH molar ratio. Inset shows a blow up of the region near
the origin (small ratios). (B) Plot of the extinction coefficients of AuNPs,
calculated from UV-vis absorbance at 520 nm, versus diameter
extracted from TEM. Inset shows a blow up of the section correspond-
ing to small ratios. (C) Intensity versus hydrodynamic diameter profiles
27,28
workers.
the exact polymer structure as well as hydrophobicity and/or
denticity” of the end groups on the size and distribution of
prepared NPs. In particular, they showed that AuNPs with sizes
The authors also explored the effects of changing
“
extracted from DLS for TA-PEG-OCH -AuNPs (as-prepared) with
3
28
different Au-to-ligand molar ratios: 2000:1, 600:1, 250:1, and 30:1. The
corresponding hydrodynamic diameters are 21 ( 4.0 nm, 16 ( 3.6,
ranging from 1.5 to 8 nm can be made.
. Dynamic Light Scattering (DLS). Dynamic light scatter-
ing (DLS) is a powerful technique that has widely been used to
3
1
3 ( 2.8, and 11 ( 2.4, respectively. (D) Intensity versus hydrodynamic
diameter profiles for the same series after the extra passivation.
probe the stability of colloidal dispersions and polymeric
(
36) Kuo, C. H.; Chiang, T. F.; Chen, L. J.; Huang, M. H. Langmuir 2004, 20,
820–7824.
37) Berne, B. J.; Pecora, R. Dynamic light scattering: with applications to
chemistry, biology, and physics; Dover Publications: Mineola, NY, 2000.
38) Brown, W. Dynamic light scattering: the method and some applications;
Clarendon Press: Oxford, New York, 1993.
37-39
solutions.
Due to the extreme sensitivity of the scattered
7
(
(
(39) Pons, T.; Uyeda, H. T.; Medintz, I. L.; Mattoussi, H. J. Phys. Chem. B 2006,
110, 20308–20316.
7
608 DOI: 10.1021/la904438s
Langmuir 2010, 26(10), 7604–7613

Oh et al.
Article
Figure 4. Images of AuNP dispersions under various conditions: (A) TA-PEG-OCH
3
-AuNP dispersions without (-) and with (þ) 2 M
NaCl after 30 min and 4 days at room temperature; (B) images of NPs subjected to refluxing at 100 °C for 30 min; and (C) NP dispersions
mixed with 0.5 M DTT at pH 8 and kept at room temperature for 1 day. (D) pH stability test of AuNPs synthesized with the various ligands
after 9 months of storage at room temperature.
signal to changes in the radius (R) of the scattering objects
addition of extra free ligands (Figure 3D). However, we consis-
tently observed a small increase in the measured hydrodynamic
size following extra passivation (compare data in Figure 3C
and D); the increase was more pronounced for the larger size
nanocrystals. For example, the hydrodynamic diameter increased
from 21 nm for the sample prepared using a Au/ligand ratio of
2000:1 to ∼25 nm for the same one following extra passivation
(see Table 1). In comparison, the diameter measured by TEM was
∼12 nm for both samples (Table 1). We infer from this change that
addition of extra ligands essentially increases the density of the
surface ligands on the nanocrystals (resulting in a more homo-
geneous surface coverage), which slightly changes the contribution of
the hydrodynamic interactions to the measured size. The slightly
larger increase in hydrodynamic diameter for larger size NPs
(prepared using higher Au/ligand ratio) can be attributed to a lower
ligand coverage of the nanocrystals during the initial growth. The
ligand density is higher for their smaller size counterparts because
higher molar concentrations of the ligands are used during the growth.
Colloidal Stability Tests. 1. Stability to Excess Salts,
pH Changes, and Refluxing at Higher Temperature. We first
explored the effects of added excess NaCl on the colloidal stability
6
(
scattered intensity µ R ), DLS can detect even small fractions
of aggregate buildup in NP dispersions and polymer solutions.
DLS also provides a measure for the hydrodynamic diameter
2RH of individual nanocrystals (with RH being the nanocrystal
hydrodynamic radius), which accounts for subtle effects of sur-
3
8
face ligand charges on the hydrodynamic interactions. RH is
extracted from the autocorrelation function of the scattered signal
and is often provided as a number or intensity distribution to
3
7,39
account for size dispersity/ditsribution.
Figure 3C shows the
experimental plots of the intensity versus hydrodynamic diameter
for four representative TA-PEG-OCH -AuNPs dispersions pre-
3
pared using different Au-to-ligand molar ratios. Single narrow
peaks were measured for all four samples, indicating narrow size
distributions in these dispersions (i.e., unimodal distributions
and absence of aggregate formation; see Figure 3C). Table 1
shows the average valuesof the hydrodynamic diameter measured
for these samples side-by-side with those extracted from TEM;
only hydrodynamic diameters for the larger size are provided as
weak scattering prevented the collection of reliable DLS data for
smaller size NPs (2RTEM < 5 nm). A ratio RTEM/R of ∼0.55 was
H
4
0
measured for the various samples. The larger hydrodynamic sizes
were consistently measured for the various dispersions due to
of AuNPs prepared using the various ligands introduced above,
namely, TA-PEG-OCH and TA (both are disulfide-terminated)
3
contributions from the ligand shell (TA-PEG-OCH ) and hydro-
along with HS-PEG-OCH and MHA; the last three constitute
3
3
3
7-39
dynamic interactions.
This confirms our previous DLS
control samples. For consistent comparison, we used similar size
nanoparticles. Figure 4A shows side-by-side images of ∼5 nm
AuNPs grown using TA, HS-PEG-OCH , and TA-PEG-OCH
measurements applied to CdSe and CdSe-ZnS quantum dots,
where the sizes extracted from DLS were also consistently larger
than those derived from TEM; a ratio of 0.5 was measured for
3
3
dispersed in deionized water with and without 2 M NaCl; data
were collected after 30 min and 4 days of storage. We found that
aggregation buildup combined with a clear color transformation
took place for TA- and MHA-capped NPs (data for MHA-
AuNPs are not shown); the dispersions changed color from red
to violet immediately after salt addition and precipitated within
3
9
those materials. The differencebetween RTEM and R cannot be
H
simply accounted for by a geometrical superposition of the
metallic core plus the TA-PEG-OCH ligands as discussed above.
3
DLS and TEM were combined to understand the effects of
adding extra free ligands following complete growth of the
nanocrystals (e.g., effects of extra passivation of the metal cores).
TEM measurements showed that there were no differences
between sizes measured for the as-prepared nanocrystals and
for those subject to extra passivation. The intensity distribution
data extracted from DLS showed single peaks, which indicates
absence of potential etching or aggregation buildup after the
1 h. In comparison, dispersions of TA-PEG-OCH -AuNPs
3
and HS-PEG-OCH -AuNPs were stable and aggregate-free
3
(
40) Levy, R.; Thanh, N. T. K.; Doty, R. C.; Hussain, I.; Nichols, R. J.;
Schiffrin, D. J.; Brust, M.; Fernig, D. G. J. Am. Chem. Soc. 2004, 126, 10076–
10084.
Langmuir 2010, 26(10), 7604–7613
DOI: 10.1021/la904438s 7609

Article
Oh et al.
(
see Figure 4A); TA-PEG-OCH -AuNPs stayed dispersed and
3
aggregate-free in 2 M NaCl solution at least for 8 months. None-
theless, dispersions of HS-PEG-OCH -AuNPs became progres-
3
sively unstable with time, and macroscopic precipitation was
eventually observed after 1 month of storage at room temperature.
In the second stability test, we probed the effects of heating
(
refluxing at 100 °C) the nanocrystal dispersions for 30 min. The
images shown in Figure 4B indicate that while TA-AuNPs
Figure 5. Gel electrophoresis of TA-PEG-AuNPs following ex-
tra passivation withmixed OCH - and COOH- orNH -terminated
3 2
changed color to dark purple, TA-PEG-OCH -AuNPs and HS-
3
PEG-OCH -AuNPs were essentially unaffected, even though
3
ligands; samples shown were treated with TA-PEG550-OCH3/
TA-PEG600-COOH = 0:1, 1:1, 3:1 (lanes 1-3), 100% TA-
HS-PEG-OCH -AuNPs exhibited a slight change in color fol-
3
lowing heat treatment (see Figure 4B). Dispersions of MHA-
3 3
PEG550-OCH (lane 4), and TA-PEG550-OCH /TA-PEG600-
NH = 3:1, 1:1, 0:1 (lanes 5-7).
2
AuNPs changed to colorless and precipitated (data not shown).
The UV-vis absorption spectra recorded for TA-PEG-OCH -
3
AuNP dispersions have identical SPB peaks before and after
heat treatments (Supporting Information, Figure S3). Effects of
heating were also tested in the presence of additional excess
free ligands (30 min refluxing with additional TA-PEG-OCH3,
disulfide bond of proteins, antibodies, and other disulfide con-
41,42
taining biomolecules.
It can displace thiol-terminated ligands
away from the NP surface, which potentially results in aggregate
buildup with time. Ligand displacement by DTT alters the
spectroscopic properties of the AuNPs, including a decrease in
the SPB peak and change in the dispersion color, along with an
increase in the absorbance at longer wavelengths. We compared
the “resistance” of three sets of aqueous AuNP dispersions,
capped with TA, HS-PEG-OCH , and TA-PEG-OCH ligands
1
0 times higher than Au atom in solution) and in the presence of
Dulbecco’s modified Eagle’s growth medium. We found no
change either in NP size (extracted from TEM) or in the absor-
bance spectra collected before and after heat treatment. In
addition, the TA-PEG-OCH -AuNPs exhibited remarkable
3
3
3
stability when dispersed in growth media with no change in the
absorption spectra and no aggregate formation; further details
are provided in the Supporting Information (Figures S3 and S4).
In the third test, we probed the effects of pH changes on the
stability of nanocrystals prepared using the same set of ligands.
The dispersion pH was controlled by adding a few drops of 2 M
NaOH (basic) or HCl (acidic) solution to the NP dispersions. We
against 0.5 M DTT at pH 8 (see Figure 4C). The three sets of NPs
have ∼5 nm core size and were prepared using Au/ligand molar
ratios of 100:1 (for TA-PEG-OCH ), 15:1 (for HS-PEG-OCH ),
and 100:1 (for TA). The Au/ligand ratio varied from one set
to another due to difference in the NP growth rate for each
ligand. The absorption spectra measured for these samples
had identical absorption features prior to addition of DTT (see
Supporting Information, Figure S5). The data showed that TA-
3
3
found that dispersions of TA-PEG-OCH -AuNPs stayed stable
3
(
no color change) and aggregate-free over the pH range of 2-13
PEG-OCH -AuNPs were essentially unaffected by the pre-
3
for at least 9 months (see Figure 4D). HS-PEG-OCH -AuNPs
also stayed well dispersed at the same pH range, though there was
3
sence of DTT. In comparison, TA-AuNPs experienced a slight
color change to purple, while HS-PEG-OCH -AuNPs comple-
3
a slight color change from reddish to bright pink at lower pH
tely precipitated after 1 day of storage at room temperature
(
Figure 4D). Dispersions of TA-AuNPs and MHA-AuNPs, in
comparison, aggregated at pH e 4 and pH e 5, respectively, after
day of storage (see Figure 4D). We should stress that the
(
Figure 4C). This test is simple yet very informative, and it
indicates that there is an added benefit to using surface ligands
that have both the bidentate anchoring group and a PEG
segment compared with those missing either module. The stron-
ger resistance to competition from other thiol containing com-
pounds can drastically enhance the colloidal stability in
biological environments (often rich in sulfur containing proteins)
and expand the opportunity for using these NPs in biological
1
stability of AuNPs synthesized using our one-phase growth
method in the presence of TA-PEG-OCH ligands was similar
3
to that observed for AuNPs ligand-exchanged with TA-PEG-
3
3,35
OCH from citrate-stabilized nanocrystals.
3
These results indicate that the colloidal stability of AuNP
dispersions to added excess ions, changes in solution pH, and heat
treatment was enhanced when the NP growth was carried out using
PEGylated ligands. The terminal PEG promotes better affinity to
the surrounding aqueous media by reducing interparticle attrac-
tions, which results in substantially enhanced steric stabilization of
the NPs against added salts, changes in pH, and heat treatment.
Importantly, the stability was further improved for NPs capped
32
targeting and imaging.
The stability tests described here are very informative and very
relevant to applications in biology. For example, assays based on
DNA hybridization are often carried out under harsh conditions
(
such as high concentrations of salts and relatively high tempera-
tures, ∼60 °C).
In Situ Surface-Functionalization of AuNPs. One addi-
tional advantage offered by the present synthetic approach is the
ability to control the surface functionalities of the AuNPs. This
can be achieved by mixing end-functionalized TA-PEG ligands
with disulfide-terminated ligands (TA-PEG-OCH ) compared
3
with single thiol-terminated ligands (MHA and HS-PEG-OCH3).
2
. Dithiothreitol Competition. Dithiothreitol (DTT) is a
small molecule known to have high affinity to Au surfaces
due to its dithiol function. Thus, stability test of TA-PEG-
(
e.g., carboxylic acid, amine, biotin, azide) together with the inert
TA-PEG-OCH ligands during the final passivation step. In this
3
OCH -AuNPs against competition from DTT will be very
3
study, we used TA-PEG600-COOH or TA-PEG600-NH in the
2
informative, as it permits one to assess the ligand binding affinity
to the nanocrystals. It also provides information on how AuNPs
potentially behave in biological media which are often rich in
sulfur containing biomolecules, such as cysteine and glutathione.
DTT is one of the most abundant chemicals used for reducing
mixture at different OCH /COOH (or NH ) ratios (i.e., 0:1, 1:1,
3
2
3:1, 1:0). Figure 5 shows the gel electrophoresis image collected
from 5 nm AuNPs prepared using the above mixing ratios;
experiments were run on 1% agarose gel for 10 min at 7 V/cm.
The gel image clearly shows that while AuNPs passivated with
(
41) Agasti, S. S.; You, C. C.; Arumugam, P.; Rotello, V. M. J. Mater. Chem.
(42) Wiita, A. P.; Koti Ainavarapu, S. R.; Huang, H. H.; Fernandez, J. M.
2008, 18, 70–73.
Proceed. Nat. Acad. Sci. 2006, 103, 7222–7227.
7
610 DOI: 10.1021/la904438s
Langmuir 2010, 26(10), 7604–7613

Oh et al.
Article
TA-PEG-OCH3 (OCH /COOH = 1:0) did not exhibit any
3
mobility shift under applied voltage, NPs functionalized with
OCH /COOH and OCH /NH mixtures experienced a sizable
3
3
2
and ratio-dependent mobility shift. Furthermore, the sign of the
mobility shift confirmed the presence of negative charges for
carboxyl-terminated ligands (AuNPs migrated toward the anode)
and positive charges for amine-terminated ligands (AuNPs mi-
grated toward the cathode). This demonstration clearly indicates
that our synthetic method can allow in situ functionalization of
the AuNPs with an array of reactive end groups.
Wewould like to stress the factthat surface functionalizationof
the AuNPs with the desired reactive groups using our synthetic
design is effective and simple to implement, because the mixed
ligands used in the passivation step are essentially identical (the
same anchoring groups and same PEG) and thus have the same
affinity to the nanocrystal surfaces. Varying the fraction of the
reactive ligands in the mixture would ultimately allow control
over the number and type of reactive groups per NP. This concept
has been tested, even though via cap exchange with the reduced
formof the ligands (DHLA-PEG), with CdSe-ZnS quantum dots;
nanocrystals have been separately functionalized with discrete
3
3
fractions of amines and carboxylic acid groups. A variety of
other end groups could be used.
In situ functionalization with the desired terminal group can
substantially simplify sample manipulation by avoiding addi-
tional ligand exchange step(s) often required with commercially
3
3,40
available citrate-stabilized AuNPs.
trolled coupling to target molecules such as biological receptors
proteins, peptides, DNAs) and dyes. Further targeted biological
It also facilitates con-
(
studies using AuNPs that have been in situ functionalized with
additional reactive groups are underway, including assay deve-
lopment intracellular uptake. Results will be presented in future
reports.
Elucidation of Interactions between Aurate and Disulfide
Groups: Precursor Formation. To gain additional insight into
what may potentially initiate and affect the NP growth, weprobed
Figure 6. Time-dependent UV-vis absorption spectra for HAuCl
and ligand mixtures (Au/ligand = 3:1, concentration of ligand =
06 μM). (A) TA-PEG-OCH and (B) HS-PEG-OCH . Changes in
the spectra are indicative of ligand to gold complex (precursor)
formation. The spectra for other ligands including TA and DHLA-
4
1
3
3
3
PEG-OCH are provided in the Supporting Information. Excess of
10ꢀ corresponding to 1.06 mM was used for the pure ligands.
the interactions between TA-PEG-OCH and Au ions by measur-
3
ing changes in the UV-vis absorption properties of aurate-plus-
nm, and saturation was reached within 15 min. Furthermore,
as with the HS-PEG-OCH , there is no new peak appearing at
ligand mixtures in the absence of NaBH reducing agent. For
3
4
comparison, we carried out similar experiments using mixtures of
aurate with TA, DHLA-PEG-OCH , and HS-PEG-OCH li-
∼360 nm (see Supporting Information, Figure S6).
We attribute the changes in the absorption features to trans-
3
3
-
gands. The time-trace curves of the absorption spectra collected
from mixtures of TA-PEG-OCH (and HS-PEG-OCH ) with
formation of AuCl
4
ions in the presence of the thiol or dithiolane
containing ligands, where chlorides are replaced with “thiol(s)” or
3
3
4
3
aurate are shown in Figure 6; additional spectra are provided in
the Supporting Information (Figure S6). There are two distinct
peaks at ∼220 and ∼290 nm for the pure aurate solutions (see
Figure 6 and Supporting Information, Figure S6), which have
“disulfide” groups. This Au-ligand precursor formation follows
different “rates” for the various ligands used; rates are slow for
HS-PEG-OCH , moderate for TA-PEG-OCH , and fast for
3
3
DHLA-PEG-OCH . (We do not specify whether it is Au(III) or
3
-
43
been assigned to the presence of AuCl4 ions. Upon mixing with
the ligands, one can isolate two distinct trends in the absorption
Au(II) that is involved in the precursor formation.) We assign the
new peak at 360 nm to a progressive Au-TA-PEG-OCH (specific
3
spectra. In the presence of TA-PEG-OCH the peaks at ∼220 and
3
for disulfide) precursor formation with time. Free disulfides (TA-
2
3
90 nm decreased with time, while an additional broad peak at
60 nm progressively built up; saturation was rapidly reached
PEG-OCH alone) exhibit a very weak absorption peak at ∼340 nm,
3
but their contribution to the measured spectra is negligible at the
(
within ∼40 min; see Figure 6A). In the mixtures with HS-PEG-
44
concentrations used (106 μM) (see Figure 6A). The different
OCH , the decrease in the peak amplitudes at 220 and 290 nm was
3
behavior observed with TA may be attributed to a rather poor
solubility of the ligand in water (TA is usually dissolved in a small
amount of methanol first before it is diluted in water), combined
with a potential interference from the COOH group with com-
slower (saturation was reached after ∼140 min), and there was no
additional peak at 360 nm. With aurate/TA mixtures, the change
in the absorption spectrum (namely, the peaks at 220 and 290 nm)
were much smaller and rapidly reached saturation. Furthermore,
there was no additional absorption peak at 360 nm as observed
-
plexation between AuCl4 and TA. Finally, the data shown in
Figure 6 also indicate that overall complexation between Au and
ligand (i.e., formation of precursors) takes place prior to, and
independent of, addition of the reducing agent.
for TA-PEG-OCH (see Supporting Information, Figure S6).
3
In contrast, mixing aurate with DHLA-PEG-OCH3 ligands
produced a very rapid decrease in the peaks at ∼220 and ∼290
(
43) Torigoe, K.; Esumi, K. J. Phys. Chem. B 1999, 103, 2862–2866.
(44) Bucher, G.; Lu, C. Y.; Sander, W. ChemPhysChem 2005, 6, 2607–2618.
Langmuir 2010, 26(10), 7604–7613
DOI: 10.1021/la904438s 7611

Article
A close look at these results provides insight into an under-
Oh et al.
Experimental Section
standing of the different growth properties observed in our
investigation: (1) The absence of a clear Au-TA complex for
mixtures with TA leads to inconsistent growth, providing only a
1
H NMR, UV-Vis Spectroscopy, and Transmission
Electron Microscopy. The ligands used were characterized by
H NMR. The spectra were recorded on a Bruker SpectroSpin
1
narrow size range. (2) With DHLA-PEG-OCH ligand, the rapid
1
00 MHz spectrometer. Chemical shifts for H NMR spectra are
3
4
formation of the Au-dithiol-PEG produces a strongly bound
precursor, which may interfere with the Au reduction in the
reported relative to tetramethylsilane (TMS) signal in the deuter-
ated solvent (TMS, δ=0.00 ppm). All J values are reported in
hertz. Electronic absorption spectra were recorded using an HP
8453 diode array spectrophotometer (Agilent Technologies,
Santa Clara, CA). The spectra were collected using quartz cuv-
ettes (Spectro Cells) with 0.5 or 1 cm optical path length.
Structural characterization of the prepared AuNPs was carried
out using a JEOL 2200-FX analytical high-resolution transmis-
sion electron microscope, with a 200 kV accelerating voltage.
Samples for TEM were prepared by spreading a drop of the
AuNP dispersion onto the holey carbon film on a fine mesh Cu
grid (400 mesh) and letting it dry. Individual particle sizes were
measured using Olympus measureIT; average sizes along with
standard deviation were extracted from analysis of ∼100 nano-
particles on average.
presence of NaBH ; this would eventually produce a smaller size
4
for comparable Au-to-ligand ratios. (3) Conversely, with HS-
PEG-OCH , the Au-to-ligand binding in the complex is also
3
strong, and this produces a narrower size range though the
stability of the AuNPs is enhanced. (4) Finally, in the presence
of TA-PEG-OCH , the issues associated with TA and the strong
3
binding exhibited by DHLA-PEG-OCH and HS-PEG-OCH3
3
become less important. The slightly weakerinteractionsin the Au-
TA-PEG-OCH precursor allow controlled reduction of the Au
3
ions and facilitate a more controlled NP growth, especially for the
large size NPs prepared using large Au/ligand ratios. This
produces growth of a more homogeneous nanocrystal population
over a broader size range, which could not be achieved using
the other ligand configurations. We should emphasize that
NaBH4 also initiates reduction of the disulfide end groups
Dynamic Light Scattering. DLS measurements were carried
out using a CGS-3 goniometer system equipped with HeNe laser
illumination at 633 nm and a single-photon counting avalanche
photodiode for signal detection (Malvern Instruments, South-
corough, MA). The autocorrelation function was performed by
using an ALV-5000/EPP photon correlator (ALV, Langen,
Germany) and analyzed using Dispersion Technology software
3
3,34
(
because 10ꢀ excess of NaBH4 was used),
which then
produces dithiol-driven anchoring of the ligands with the AuNP
surface; this strong anchoring produces enhanced colloidal stabi-
lity of the resulting nanoparticles. The PEG tail provides an inert
segment that reduces undesired interference with the precursor
formation and further provides better affinity to the surrounding
water phase. The dithiol-driven coordination combined with the
presence of PEG on the ligand are at the origin of the better
(
DTS) (Malvern Instruments). All AuNP solutions were filtered
through 0.2 or 0.45 μm syringe filters (Millipore). The sample
temperature was maintained at 20 °C. For each sample, the
autocorrelation function was the average of three runs of 10 s
each, and then repeated at different scattering angles from 50° to
110°. Then contin analysis was applied to extract intensity versus
hydrodynamic size profiles for the dispersions studied.
colloidal stability observed for TA-PEG-OCH -AuNPs in the
3
presence of excess salts, changes in solution pH, and their resistance
to competition by DTT.
Gel Electrophoresis. Gel electrophoresis experiments were
conducted using AuNPs functionalized with (i) pure TA-PEG-
OCH , (ii) mixture of TA-PEG-OCH /TA-PEG-COOH, and
Conclusion
3
3
We developed a water-based synthesis and growth method
of a series of AuNPs using a modular bidentate PEG ligand.
Our synthetic method relies on the use of modular ligands made
of PEG appended with a disulfide anchoring group. The
present approach allows us to control the nanocrystal size over
the range between 1.5 and 18 nm which is a much wider size
regime compared with previous reported methods using strong
Au-S interaction. The nanocrystal size is simply controlled
by varying the molar ratio of Au to the PEG ligand. TA-PEG-
(iii) mixture of TA-PEG-OCH
crystal dispersions were run on a 1% agarose gel using tris
borate EDTA buffer (TBE, 100 mM Tris, 83 mM boric acid,
3
2
/TA-PEG-NH . The nano-
1
mM EDTA, pH 8.3). AuNP dispersions were prepared with
∼
0.5-1 μM concentrations in a 10% glycerol TBE loading
buffer prior to use. The experiments were conducted using a
voltage of 7-8 V/cm, and the images were captured using the
white light mode of a Kodak Gel Logic 2200 imaging system.
Synthesis of TA-PEG Series. TA-PEG550-OCH
PEG-OCH , TA-PEG600-NH and TA-PEG600-COOH were
synthesized, purified, and characterized following the procedures
3
, TA-
3
2
OCH -AuNPs showed dramatically enhanced colloidal stabi-
3
3
2-34
lity over a wide range of pHs (from 2 to 13), under high salt
concentrations, and against DTT competition/displacement.
Such superior stability results from the combined effects: (1)
hydrophilic nature of long PEG chain and (2) tight binding
offered by the bidentate dithiol anchoring group on Au
surfaces.
detailed in our previous reports.
Synthesis of MHA-PEG550-OCH (HS-PEG550-OCH ).
A. Synthesis of 6-(acetylthio)hexanoic acid (AcS(CH ) -
3
3
2
5
-
2
COOH). To 5.00 g (2.56ꢀ10 mol) of 6-bromohexanoic acid in
dimethylformamide (100 mL) was added potassium thioacetate
-2
(8.6 g, 7.5 ꢀ 10 mol) in one portion at ∼0 °C. The deep red
mixture was stirred for 30 min at room temperature, diluted with
Our growth method also permits in situ functionalization of the
AuNP surface with ligands presenting the desired end groups
CH
layer was dried over Na
yellow crude product was purified by vacuum distillation. TLC
Cl
2 2
(250mL), and washedthree timeswithwater. The organic
2
SO , and the solvent was evaporated. The
4
(
such as amine, carboxylic acid, and biotin) to promote biocom-
patibility. These features combined with the remarkable colloidal
stability of these nanocrystals open up new opportunities for use
in biological applications. For example, coupling of the present
AuNPs to specific peptides and other biological receptors could
allow the development of assays that can be carried in harsh
conditions (e.g., in vivo) where the remarkable stability offered by
these materials could be verybeneficial. Attaching cell penetrating
peptides and proteins onto these NPs could allow guided delivery
of such conjugates inside live cells where targeted photothermal
1
(
CH
CDCl
.33 (s, 3H), 1.55-1.71 (4H), 1.38-1.46 (m, 2H).
B. Synthesis of AcS-PEG550-OCH (acetyl-S-(CH ) CONH-
2
Cl
2
/MeOH=10:1 v/v) R
f
∼ 0.57. H NMR (400 MHz, in
3
): δ (ppm) 2.87 (t, 2H, J=7.2 Hz), 2.36 (t, 2H, J=7.4 Hz),
2
3
2 5
(
CH CH O) -CH CH OCH , ATA-PEG550-OCH ). Amounts
2
2
n
2
2
-2
3
3
O-PEG550-NH
23
,
of 5.5 g (∼1.0 ꢀ 10 mol) of CH
3
2
0.53 g
-3
(
1
2.6 ꢀ 10 mol) of 4-(dimethylamino)pyridine (DMAP), and
-2
0
.2 g (1.0 ꢀ 10 mol) of N,N -dicyclohexylcarbodiimide (DCC)
2 2
were added to 70 mL of CH Cl . An amount of 2.0 g (1.1 ꢀ
3
2
-2
therapy can be applied.
2 5 2 2
10 mol) of AcS(CH ) COOH dissolved in 20 mL of CH Cl
7
612 DOI: 10.1021/la904438s
Langmuir 2010, 26(10), 7604–7613

Oh et al.
Article
was added dropwise to the reaction mixture in an ice bath
under N . Then the reaction mixture was gradually warmed up to
room temperature and further stirred overnight. The mixture was
filtered through Celite, and the precipitate was rinsed with
times). The combined organic layers were dried over Na
2
SO
/MeOH = 10:1 v/v) R
): δ (ppm) 6.29 (br s, 1H),
3
4
, and
the solvent was evaporated. TLC (CH
∼ 0.5. H NMR (400 MHz, in CDCl
Cl
2 2
f
2
1
3.58-3.71 (m), 3.53-3.57 (m, 4H), 3.46 (t, 2H, J = 5.2 Hz), 3.38
(s, 3H), 2.5-2.56 (dt, 2H), 2.19 (t, 2H, J = 7.2 Hz), 1.6-1.67 (m,
4H), 1.38-1.46 (m, 2H), 1.35 (t, 1H, J = 7.8 Hz).
2 2
CH Cl . After evaporating solvent, 1 M of HCl was slowly added
to the residue. The aqueous mixture was washed three times with
ether and saturated by NaHCO . The product was extracted with
3
CH
over Na
on silica gel with 10:1(v/v) CH
TLC (CH Cl /MeOH = 10:1 v/v) R ∼ 0.55. H NMR (400 MHz,
2
Cl
2
(three times). The combined organic layers were dried
SO and evaporated. The residue was chromatographed
Cl /MeOH to collect the product.
Acknowledgment. The authors acknowledge NRL, Office of
Naval Research (ONR), the Army Research Office for financial
support. E.O. was supported by a fellowship from the Korea
Research Foundation (D00089). Some of the data were col-
lected using the HRTEM characterization facility provided at
the Nanoscience Institute (NRL). We also thank Igor Medintz
and Dorothy Farrell for assistance with some experimental
details.
2
4
2
2
1
2
2
f
in CDCl
3
): δ (ppm) 6.14 (br s, 1H), 3.58-3.71 (m), 3.53-3.58
(
m, 4H), 3.46 (t, 2H, J=5.2 Hz), 3.38 (s, 3H), 2.86 (t, 2H, J=
7
1
.6 Hz), 2.32 (s, 3H), 2.18 (t, 2H, J = 7.2 Hz), 1.54-1.71 (m, 4H),
.38-1.46 (m, 2H).
C. Synthesis of HS-PEG550-OCH (MHA-PEG550-OCH ).
3
3
-3
An amount of 1.0 g (1.4 ꢀ 10 mol) of ATA-PEG550-OCH3
AcS-PEG550-OCH ) was dissloved in 50 mL of MeOH and
purged with N
. A total of 0.055 g (1.4 ꢀ 10 mol) of NaOH in
mL of water was injected into the reaction, and the mixture was
(
3
Supporting Information Available: UV-vis absorption
-3
2
3
spectra collected from dispersions of monothiol-PEG-OCH -
6
3
AuNPs, TA-PEG-OCH -AuNPs before and after refluxing,
stirred overnight at room temperature under N . HCl (0.5 M) was
2
added to the reaction mixture for neutralization. After evapora-
tion of MeOH, the product was extracted with CH Cl (three
2 2
DLS data; UV-vis time trace for ligand-plus-aurate mixtures
collected for TA and DHLA-PEG ligands. This material is
available free of charge via the Internet at http://pubs.acs.org.
Langmuir 2010, 26(10), 7604–7613
DOI: 10.1021/la904438s 7613