Preparation of functional spherical polysilsesquioxane/gold nanoparticle composites and their applications in DNA assay

RESEARCH PAPER

Preparation of functional spherical polysilsesquioxane/goldnanoparticle composites and their applications in DNA assay

Jung A. Jung • Young Baek Kim • Young A. Kim •

Seung Bum Ryu • Veronica Kim

Received: 29 January 2010 / Accepted: 8 June 2010

� Springer Science+Business Media B.V. 2010

Abstract Functional spherical solid and hollow

particles of polysilsesquioxanes (PSQs) containing

amine, thiol, and vinyl groups were prepared by

polymerizing organotrialkoxysilanes (OTASs) con-

taining corresponding chemical groups. Fluorescent

PSQ particles were prepared by physically entrap-

ping Rhodamine 6G, Coumarin 7, and Fluoresce-

ine sodium salts. The intensity of fluorescent light

increased initially with increasing amount of entrapped

fluorophores and then leveled off or decreased slightly

after reaching a maximum value. PSQ particles

containing gold nanoparticles (GNPs), both inside

and on the surface, were prepared by the in situ

reduction of gold ions by the PSQ particles. When the

reduction reaction was carried out for extended

periods of time, the GNP that had formed inside the

poly(3-mercaptopropyl)silsesquioxane (PMPSQ) and

polyvinylsilsesequioxane (PVSQ) particles under-

went interesting morphological changes. PSQ particles

containing amine and thiol groups fixed the GNPs on

the surface, which could be utilized further in binding

amine-modified oligo-DNA strands. The aggregation

of PSQ/GNP particles combined with complementary

oligo-DNA strands was examined to demonstrate

that these particles could be applied to DNA assays

and isolation. The particles were characterized by

scanning electron microscopy, transmission electron

microscopy, solid state nuclear magnetic resonance

spectroscopy, ultraviolet/visible spectroscopy, and

fluorescence microscopy.

Keywords Fluorescent particles �Polysilsesquioxane particles � Gold nanoparticles �DNA assay � DNA–amine interaction �Synthesis

Introduction

Small particles with useful properties, such as fluores-

cence, magnetism, distinctive colors, different densi-

ties, and different sizes, have a range of potential

applications in many fields including biology

Electronic supplementary material The online version ofthis article (doi:10.1007/s11051-010-9997-z) containssupplementary material, which is available to authorized users.

J. A. Jung � Y. B. Kim (&)

Department of Nanotechnology, PaiChai University,

Daejeon 302-735, Korea

e-mail: [email protected]

Y. A. Kim

PolyChrom Inc, 452-6 Moknaedong, Ahnsan,

Kyungkido 425-100, Korea

S. B. Ryu

Alteogen, Inc, 461-8, Jeonmin-Dong, Yusung-Ku,

Daejeon 305-811, Korea

V. Kim

Department of Animal Science, Cornell University,

Ithaca, NY 14853, USA

123

DOI 10.1007/s11051-010-9997-z

J Nanopart Res (2011) 13:2361–2374

/ Published online: 20 June 2010

(Klostranec and Chan 2006; Parak et al. 2003; Pelleg-

rino et al. 2005). The particles would be even more

useful if they could be easily combined with useful

compounds, i.e., stealth compounds, biomarkers, and

probes. For example, gold nanoparticles (GNPs) are

used widely in a variety of applications on account of

their strong color and spontaneous reaction with SH

groups in different types of molecules including

proteins and modified DNA strands. The distinct color

changes caused by the aggregation of GNPs have also

been applied to DNA assays (Eaton et al. 2007;

Niemeyer and Simon 2005; Thaxton et al. 2006).

There are different strategies for preparing parti-

cles with chemical reactivity and useful physical

properties. One of them is to prepare composite

particles of more than one type (Corr et al. 2008;

Heitsch et al. 2008; Liong et al. 2008; Pasqua et al.

2009; Pellegrino et al. 2005; Ren et al. 2008;

Schottner 2001; Smith et al. 2008; Tan and Zhang

2005; Yang et al. 2009; Lin 2009). For example,

reactive fluorescent particles might be prepared by

entrapping fluorescent nanoparticles into other solid

and hollow particles containing reactive functional

groups.

Polysiloxane (PSO) and polysilsesquioxane (PSQ)

nanoparticles from organoalkoxysilanes attracted our

interest because of the excellent biocompatibility of

PSO. PSQ might be as biocompatible as PSO because

it contains identical chemical groups to those of PSO,

Si–R, and SiOH. It is also advantageous that organ-

oalkoxysilanes (OAS) containing different functional

groups are commercially available and many of them

can be polymerized in one step reaction to form

spherical particles. The direct preparation of func-

tionalized particles removes the additional chemical

reactions needed to introduce the functional groups to

non-functionalized particles (Kim et al. 2006; Lee

et al. 2008; Pasqua et al. 2009).

Our earlier results showed that the base-catalyzed

polymerization of pure diorganodialkoxysilane

(DODAS) gave only oligomeric oils and amorphous

monoliths (Lee et al. 2008). However, many organo-

trialkoxysilane (OTAS) formed monodisperse insol-

uble network particles in quantitative yield under

basic conditions. Furthermore, the PSQ particles with

functional groups reduced different metal ions to

produce the corresponding metal nanoparticles and

fixed them inside and on the surface (Kim et al.

2006).

This study examined the nature of PSQ spheres

further with an emphasis on the reduction of gold ions,

binding with GNP, and entrapping organic fluoro-

phores. The possibility of applying these particles to a

DNA assay that can be accomplished with great

convenience was also investigated.

Experimental section

Materials and reagents

Rhodamine 6G (Rh6G), Fluoresceine sodium salt (Fl),

Coumarin 7, methyltrimethoxysilane (95%, MTMS),

3-mercaptopropyltrimethoxysilane (95%, MPTMS),

3-aminopropyltrimethoxysilane (95%, APTMS), and

Tween 20 were purchased from Aldrich Chemicals,

and used as received. Two sets of complementary

oligo-DNA strands containing C7-amino groups on

the 30-ends, 50-TAGCTATGGAATTCCTCGTAGGC

A-amino(C7)-30 (DNA1), 50-TCGCTACGAGGAAT

TCCATAGCTA-amino(C7)-30 (DNA10), 50-TTTTTT-

aminor(C7)-30(DNA2), and 50-AAAAAA-amino(C7)-

30 (DNA20), were purchased from Xenotech (Daejeon,

Korea) and used after a purity check using 2% poly-

acrylamide gel electrophoresis.

Preparation of PSQ spheres

The production of PSQ particles was examined

by polymerizing OTAS in water using 3-aminopropyl-

trimethoxysilane (APTMS), 3-aminopropyltriethoxysi-

lane (APTES), triethylamine (TEA), 3-(2-aminoethyl)

aminopropyltrimethoxysilane (AEPTMS), and hexyl-

amine (HA) as catalysts. The polymerization mixtures

were heterogeneous because most OTAS were insoluble

in water. In a typical example of polymerizing MTMS

using TEA as catalyst, 1 mL of MTMS was mixed in

20 mL of distilled water. The mixture was hand shaken

until the MTMS layer disappeared, typically for approx-

imately 30 s. TEA (0.1 mL) was then added to the

reaction mixture, which was also hand shaken vigorously

until the reaction mixture turned milky, usually taking

approximately 1 min. The reaction mixture was shaken

every hour at room temperature. The mixtures were

observed by optical microscopy until the solid particles

formed, and were then allowed to stand overnight at room

temperature. The production of solid particles was

confirmed by observing the products recovered by

123

J Nanopart Res (2011) 13: –2361 23742362

centrifugation. The products recovered, if any, were

dispersed in ethanol and acetone to determine their

solubility in organic solvents.

Network PSQ spheres containing amine, vinyl, and

thiol groups were prepared by the polymerization of

MTMS, VTMS, and MPTMS with APTMS. The sizes

of the PSQ spheres were controlled using Tween 80 in

different amounts. A typical example for the prepara-

tion of 2 lm PSQ spheres with mercapto group was as

follows. Four grams (20.4 mmol) of 3-mercaptopro-

pyltrimethoxysilane (MPTMS) was added to a 70 mL

vial containing 50 mL of distilled water. The mixture

was mixed thoroughly by hand shaking for a few

minutes, and then 0.41 g (2.3 mmol) of APTMS was

added once. The reaction mixture was shaken gently

overnight. The particles were recovered by centrifu-

gation and washed thoroughly with water and ethanol.

The particles were finally washed with acetone and

dried at room temperature under vacuum until the

weight did not change after further drying for 24 h.

Small particles with 200 nm diameter were prepared

using the same method except that 10 mg of Tween 20

was added to the distilled water.

The hollow spheres were prepared by adding

100 mL of VTMS and 40 mL of APTMS to w/o

emulsions prepared from mixtures of distilled water

(1 mL), cyclohexane (10 mL), and Span 80 (20 mg).

The reaction mixture was allowed to stand at ambient

temperature for 12 h and then mixed with 20 mL of

1:1 mixture of water and ethanol. The precipitates

were recovered by centrifugation and washed with

ethanol and distilled water.

Reduction of gold ion by functional PSQ particles

Typically, 20 mg of functional PSQ particles were

dispersed in 20 mL of 0.2 mM aqueous solution of

HAuCl4 in a 50 mL centrifuge tube. The tubes were

tightly sealed and annealed in ovens at 70 and 50 �C.

When the color of supernatant HAuCl4 solution

turned noticeably thinner, PSQ particles were centri-

fuged and the old HAuCl4 solution was replaced with

flesh one. The reduction was continued until the

absorbance of supernatant HAuCl4 solution at

214 nm stopped decreasing for 24 h. The PSQ–

GNP particles were recovered by centrifugation and

washed with excess amounts of distilled water at least

three times.

The morphological changes of GNP formed in

PSQ particles were observed using PSQ–GNP parti-

cles that were annealed at 50 �C in 0.2 mM aqueous

solution of HAuCl4 after the absorbance of the

supernatant solution stopped decreasing.

Preparation and characterization fluorescent PSQ

particles

The following summarizes the typical procedure for

preparing fluorescent PSQ particles for PMSQA.

Fifty milliliters of distilled water, 1 mL of a 21 mM

Rhodamine 6G (Rh6G) solution in ethanol, 0.5 mL of

a 1 wt% aqueous solution of Tween 20, and 1 mL of

MTMS were mixed and stirred vigorously using a

magnetic stirring bar. 0.1 mL of APTMS was then

added at once and the resulting mixture was stirred at

room temperature for 24 h. The particles loaded with

Rh6G were recovered by centrifugation and washed

repeatedly with 50 mL of distilled water until the

absorbance of the supernatant solution was\0.001 at

530 nm. The fluorescent particles were redispersed in

50 mL of a pH 7.4 phosphate buffer s.

Preparation of 2.6 nm GNP colloidal solution

GNP colloidal solution was prepared following

method described elsewhere. Ninety milliliters of

deionized distilled water was placed in a 250 mL

beaker and heated to boiling point. One milliliter of a

0.025 M aqueous HAuCl4 solution, 0.034 M aqueous

sodium citrate solution, and 1 mL of a 0.075 wt%

NaBH4 solution in a 0.034 M aqueous sodium citrate

solution, which had been stored at 0 �C, were added

when the distilled water began to boil. The reaction

mixture was boiled for 10 min and allowed to cool to

room temperature to obtain the 2.6 nm GNP colloidal

suspension.

Coupling PSQ spheres with GNP. Approximately

10 mg of the dry PSQ particles was dispersed in

distilled water and recovered by centrifugation to

prepare the wet PSQ particles. Two milliliters of

aqueous colloidal 2.6 nm GNP solution was added to

the wet PSQ particles, and the resulting mixture was

stirred overnight at room temperature. The PSQ/GNP

composite particles were recovered by quick centri-

fugation and washed repeatedly with distilled water.

The recovered PSQ/GNP composite particles were

123

J Nanopart Res (2011) 13: –2361 2374 2363

redispersed in 10 mL of a pH 7.4 phosphate buffer

solution.

PSQ/GNP/DNA composites

One milliliter of a pH 7.4 phosphate buffer solution,

0.1 mL of 2 lm PMSQA/GNP and 200 nm PMSQA/

GNP fluorescent colloidal solution were mixed

separately in 1.5 mL Eppendorf tubes. Fifteen micro-

liters of the two complementary oligo-DNA strands

were added separately to each Eppendorf tube. The

tubes were agitated at room temperature for 1 h and

the particles were recovered by centrifugation. The

particles were dispersed in 1 mL of a pH 7.4

phosphate buffer solution.

Hybridization of PSQ/GNP/DNA composite

particles

0.1 mL of each of particles conjugated with comple-

mentary DNA were mixed in a 1.5 mL Eppendorf

tube and heated to 90 �C for 1 h. The particles

conjugated with DNA1 and DNA10 were annealed at

60 �C while the particles conjugated with DNA2 and

DNA20 were annealed at 40 �C for 2 h before they

were cooled to room temperature and observed with

an optical microscopy. The control samples were

prepared in the same manner using PSQ/GNP parti-

cles without the DNA strands, and PSQ/GNP parti-

cles conjugated with identical DNA strands.

Intensity of fluorescent light

The intensities of fluorescent light emitted from the

PSQ particles containing fluorophores were measured

from the same direction of the incident light beam

using a Y-shaped optical cable. The one branch of the

Y-optical cable was connected to incident light source

and the other branch was connected to the detector,

Ocean Optics USB 4000-00286. The particles were

packed in black Eppendorf tubes with a thickness of

1 cm and the incident light was then irradiated from

1 cm above the surface of the packed sample. The

spectrum of the wavelength of incident light was

adjusted so that the spectrum of fluorescent light did

not overlap (Supplementary data 1). To minimize

experimental error, the samples were prepared imme-

diately before the measurements. All measurements

were made with a minimal time gap between the

measurements of different samples. The consistency

of the reading values was checked by measuring the

fluorescent intensity of a working standard periodi-

cally during the measurements.

Miscellaneous

Transmission electron microscopy (TEM) and scan-

ning electron microscopy (SEM) images were taken

using a TECNAI G2 T-20S and Philips XL30 FEG,

respectively. The ultraviolet/visible (UV/Vis) spectra

were recorded using a Shimazu UV-2450 spectro-

photometer. An Olympus XI 7 microscope equipped

with fluorescent accessories, including a mercury arc

lamp and appropriate filters, was used to observe the

fluorescent particles. Solid state nuclear magnetic

resonance (NMR) spectroscopy was carried out at

KSBI Dadegu Center in Korea using a 400 MHz

Bruker Solid State NMR DSX.

Results and discussions

Table 1 shows the types of products obtained from the

attempted polymerization of the OTAS using different

amine compounds as catalysts. Only phenyltrimeth-

oxysilane (PMTS) produced insoluble spheres in

sizable yields (ca. 20%) when N-(2-aminoethyl)amino-

propyltrimethoxysilane (AEPTMS) was used as the

catalyst among OTAS containing polar and bulky

groups, glycidoxypropyltrimethoxysilane (GDPMS),

trimethoxysilylpropylisocyanate (TMSPC), trimeth-

oxysilylpropyl-p-methoxycinnamamide (TMSPMCA),

PTMS, and triethoxysilylpropyl-p-methoxycinnama-

mide (TESPMCA). The reactions of GDPMS and

TMSPC were attempted using only TEA because the

other amines would react with oxirane and isocyanate

groups in these compounds.

OTAS that contained small organic groups,

methyltrimethoxysilane (MTMS), vinyltrimethoxysi-

lane (VTMS), and 3-mercaptopropyltrimethoxysilane

(MPTMS), readily gave insoluble monodisperse PSQ

spheres regardless of the catalysts used.

TMSPMCA and TESPMCA produced significant

amount of insoluble spherical powders along with

monolith solid material when strong bases were used

as the catalyst. These results showed that it was not

possible to predict the nature of products based only

on either of the properties of OTAS or the catalyst.

123

J Nanopart Res (2011) 13: –2361 23742364

The type of products from a given OTAS differs

depending on the catalysts used and vice versa.

The formation of PSQ particles from MTMS

droplets dispersed in water was observed by an optical

microscopy. MTMS droplets that were initially stable

in water became unstable and moved around cata-

strophically, making the observation difficult, as soon

as a small amount of 3-aminopropyltrimethoxysilane

(APTMS) was added. The sizes of MTMS droplets

decreased and eventually no MTMS droplets were

observed when the reaction mixture became static.

Then numerous small particles appeared and grew

with time. The whole process took less than 5 min.

The catastrophic movement of the MTMS droplets

caused by the addition of APTMS was attributed to

the production of water soluble molecules from

MTMS. The disappearance of MTMS droplets and

sudden appearance of many particles indicated that

the particles were produced by the nucleation-growth

mechanism. It should be noted that nuclei can only

formed in the aqueous medium because MTMS has to

react with water molecules. Mixtures of water and

droplets of MPTMS and VTMS showed virtually

identical behaviors. These processes were compara-

ble to those in emulsion polymerization where the

nuclei are produced in the aqueous medium and grow

as monomers are provided from the monomer

droplets dispersed in the aqueous medium.

In these reaction systems, smaller particles were

produced when larger number of nuclei was formed.

The number of nuclei can be increased by increasing

the interfacial area between OTAS and water that

allows more OTAS molecules to diffuse into water.

Increasing the amount of catalyst is also expected to

increase the number of nuclei by enhancing the

reactions related to the formation of nuclei.

The interfacial area can be controlled by the

agitation rate, amounts of water and surfactants. Our

results showed that using surfactants was the most

efficient in production of small particles (Kim et al.

2006). For example, a mixture of 50 mL of water,

1 mL of MTMS, 0.1 mL of TEA, and 10 mg of

Tween 80 produced spheres with a mean diameter of

35 nm, as shown in Fig. 1a. The spheres prepared

under identical conditions in the absence of surfactant

typically had diameters approximately 1 lm as seen

in Fig. 1b.

Increasing the amounts of amine catalysts above

certain limits caused formation of irregularly sized

particles in low yields and no solid products. For

example, the SEM images of monodisperse parti-

cles and particles with very different sizes pro-

duced from MPTMS in the presence of smaller and

larger amounts of APTMS are seen in Fig. 1c and d,

which respectively show poly(3-mercaptopropyl)sils-

esquioxane (PMPSQ) spheres prepared from 994:9:1

and 994:9:10 molar mixtures of water, MPTMS, and

APTMS.

The initial decrease in size with the increase of

APTMS concentration was attributed to the faster and

Table 1 Types of products from attempted polymerization of the organotrialkoxysilane using different amine compounds

as catalysts

Organotrialkoxysilane Catalyst Product

Methyltrimethoxysilane,

vinyltrimethoxysilane

3-Aminopropyltri(m)ethoxysilane,

triethylamine, hexylamine, 3-(2-

aminoethyl)propyltrimethoxysilane

Insoluble monodisperse spheres

Glycidoxypropyltrimethoxysilane,

trimethoxysilylpropylisocyanate

Tiethylamine, hexylamine Oily product

Phenyltrimethoxysilane 3-(2-Aminoethyl)propyltrimethoxysilane Mixture of insoluble monodisperse spheres

and soluble spheres

3-Aminopropyltri(m)ethoxysilaen,

triethylamine, hexylamine

Mixture of soluble spheres and oil

Tri(m)ethoxysilylpropyl-p-

methoxycinnamamide

3-Aminopropyltri(m)ethoxysilane,

triethylamine, hexylamine, 3-(2-

aminoethyl)propyltrimethoxysilane

Mixtures of soluble amorphous solid and

insoluble monolith

3-Mercaptopropyltrimethoxysilane 3-Aminopropyltri(m)ethoxysilane,

triethylamine, hexylamine, 3-(2-

aminoethyl)propyltrimethoxysilane

Insoluble monodisperse spheres (agitated),

insoluble spheres with large size distribution

(non-agitated)

123

J Nanopart Res (2011) 13: –2361 2374 2365

more efficient formation of nuclei in an aqueous

medium. The formation of irregularly sized particles

in low yields and no solid products in high concen-

trations of amine catalysts was ascribed to the high

pH of the medium where the hydrolysis and conden-

sation of OTAS was faster. The faster polymerization

can lead to formation of solid shell on OTAS droplets

before OTAS molecules inside diffused into the

aqueous medium. The OTAS inside the shells that

could not diffuse out eventually polymerized to form

large solid particles. The irregularly sized particles

formed most frequently from MPTMS because

MPTMS was significantly more viscous than MTMS

and VTMS that caused slower diffusion into the

aqueous medium. Also, the reaction of MPTMS was

slower than MTMS and VTMS under the identical

reaction conditions. The mechanism for the formation

of large particles is comparable to suspension poly-

merization as the monomer droplets polymerize to

form solid particles in both reaction systems.

The effect of low pH was further elucidated by the

fact that a 9940:9:10 molar mixture of water,

MPTMS and APTMS gave monodisperse spheres.

The ratio between MPTMS and APTMS was the

same while the pH of this reaction system was lower

than that of the 994:9:10 molar mixture, which gave

irregularly sized particles, as described above.

One of the most interesting features of PSQ particles

containing amine, vinyl and mercapto groups was that

in the presence of different metal ions, they produced

metal nanoparticles that were fixed inside and on the

surface (Kim et al. 2006). The reduction of gold ions by

the PSQ particles was examined further using

PVSQT91, PVSQA91, PMPSQT91, PMPSQA91,

and PMSQA91. In these notations, the final letter,

T and A, respectively, indicates that the spheres were

prepared using TEA and APTMS as catalysts. The

number indicates the molar ratio of OTAS and the

catalyst. For example, PVSQA91 indicates PVSQ

spheres prepared from a 9:1 molar mixture of VTMS

and TEA. Quantitative solid state 13C NMR analysis

showed that the amount of 3-aminopropylsilsesquiox-

ane units in PVSQA91, PMPSQA91, and PMSQA91

was commonly approximately 10 mol% (Supplemen-

tary data 2).

Figure 2 shows the quantity of gold produced by a

unit mass of PVSQA91 particles heated in a 0.1 mM

aqueous gold ion solution at 50 and 70 �C as a

Fig. 1 SEM images of polymethylsilsesquioxane spheres

prepared from a 50 mL of water, 1 mL of methyltrimethoxy-

silane, 0.1 mL triethylamine, and 10 mg of Tween 80;

b 50 mL of water, 1 mL of methyltrimethoxysilane, and

0.1 mL triethylamine; SEM image of poly(3-mercaptopro-

pyl)silsesquioxane spheres prepared from c 994:9:1 molar

mixture of water, MPTMs, and APTMS; d 994:9:10 molar

mixtures of water, MPTMS, and APTMS

123

J Nanopart Res (2011) 13: –2361 23742366

function of the reaction time. The amount of gold

produced was calculated from the amount of disap-

peared gold ions that was determined based on the

linear relationship between the concentration of

aqueous HAuCl4 solutions and their absorbance at

214 nm (Supplementary data 3).

The amounts of produced gold in Fig. 2 should not

be taken as the true amounts of GNP fixed inside and

on the surface of the PSQ particles because GNP

aggregates that were not bound to PSQ particles were

often found in the TEM images. In addition, it should

be noted that the concentration of HAuCl4 in solution

was not constant during the experiment because

HAuCl4 solutions were added when the color of the

supernatant solutions turned pale. However, every set

of reduction experiment was carried out under virtu-

ally identical conditions and the results could be used

to draw general conclusions. Figure 2 shows that the

overall amount of gold ion reduced by PVSQA91 was

not affected by the reaction temperature but the

reduction rate at 70 �C was significantly higher than

that at 50 �C. Similar results were obtained from

experiments carried out using other PSQ particles

although the amounts of reduced gold ion were

different for different particles (Supplementary data

4). These results indicated that the amount of reduced

gold ion was determined by the amount of functional

groups available for the reduction.

The GNPs on the surface and inside PVSQT91

underwent interesting changes when they were kept

in the HAuCl4 solutions for extended periods of time.

Figure 3a and b show TEM images of the PVSQT91

particles heated at 50 �C in a 0.2 mM aqueous

HAuCl4 solution for 7 and 20 days, respectively.

These figures show that many GNPs produced inside

the particles and the sizes of GNPs increased as the

reaction period increased from 7 to 20 days. Figure 3c

and d, respectively, shows TEM and SEM images of

the identical particles recovered from the same

reaction mixtures that had been heated at 50 �C for

40 days. The number of GNPs inside the particle

further decreased while the sizes of them increased.

The SEM image shown in Fig. 3d was obtained using

particles that were not sputtered with gold for SEM

analysis but no serious whitening effect took place

indicating that these particles might be conducting. It

is noteworthy that some of PVSQT particles were

cracked and some GNP particles produced inside were

exposed as seen in Fig. 3c. The development of cracks

was ascribed to the stress that was built up around the

GNPs while they became larger.

Particles recovered from the same reaction mixture

after 50 days showed more striking changes, as

shown in Fig. 3d and e. The number of GNPs inside

the PSQ particles decreased further. Furthermore, the

GNPs that were produced originally on the surface

and inside, but close to the surface, disappeared

leaving large holes. Considering that the amount of

gold ions reduced during this period of time was very

small, these results indicated that the large GNPs

formed as the original smaller GNPs produced inside

PSQ particles merged. The large GNPs formed on the

surface fell off as the binding force between PVSQT

and GNPs per unit mass of GNPs decreased with the

surface area per unit mass of GNPs. The UV/Vis

spectroscopy analysis showed that kmax of GNP

shifted from 560 to 628 nm as the sizes increased as

shown in Fig. 4.

For a better understanding, the behavior of the GNPs

formed in the smaller PVSQT91 spheres was examined

using PVSQT91 particles with diameters of approxi-

mately 100 nm. The spheres annealed in a 0.2 mM

HAuCl4 solution for 15 days contained several

GNPs with diameters of approximately 15 nm, as

shown in Fig. 5a. Figure 5b shows particles that had

been recovered from the same reaction mixture after

50 days. Only a few PSQ particles contained one large

GNP (ca. 40 nm) inside, probably when the initial

GNPs formed in the center of the particles. Most of the

Fig. 2 The quantity of gold produced by a unit mass

of PVSQA91 particles heated in a 0.1 mM aqueous gold ion

solution at 50 (circle) and 70 �C (square) plotted against the

reaction time

123

J Nanopart Res (2011) 13: –2361 2374 2367

particles contained no GNPs on the surface or inside,

but had large holes on the surface, indicating that GNPs

had merged together and then fell off. These results

showed that the GNPs underwent identical morpho-

logical changes regardless of the PSQ particle size.

Figure 3d and e showed that pores with tens of

nanometers were clearly observed by TEM. Since

there were no PSQ particles that originally contained

such large pores as observed with TEM, the pores

where the large GNPs formed should have been

produced during the formation of large GNPs. Exclud-

ing the possibility of cleavage and reformation of

Si–O–Si bonds, these results suggest that PVSQT

particles were soft and porous enough for smaller

GNPs to move around and fuse together. The porous

nature of these particles was proved by efficient

entrapment of small fluorescent molecules as

described below.

The change of the GNPs produced inside the

PMPSQT91 particles was also unusual. The GNPs

produced by PMPSQT91 appeared to be very small as

seen in Fig. 6a for PMPSQT91 particles that had been

placed in a 0.2 mM HAuCl4 solution for 20 days.

Fig. 3 TEM images of particles obtained by heating PVSQT91 at 50 �C in a 0.2 mM aqueous solution of HAuCl4 for a 7 days,

b 20 days, c 40 days, d SEM image of particles shown in c, e 50 days, f SEM image of particles shown in e

Fig. 4 UV/Vis spectra of colloidal solutions of PVSQ91

particles heated at 50 �C in a 0.2 mM aqueous solution of

HAuCl4 for a 7 days, b 20 days, c 40 days, d 50 days

123

J Nanopart Res (2011) 13: –2361 23742368

The presence of GNP in these particles was revealed in

the TEM image of the particles calcined at 1200 �C

that is presented in Fig. 6b (Supplementary data 5).

The much darker layer closer to the surface in Fig. 6a

indicated that larger amounts of GNPs formed near the

surface. A close comparison of Fig. 6a and b shows

that the GNPs in the shell layer disappeared after

calcination. The disappeared GNPs should have

formed larger masses of gold shown in Fig. 6b.

The TEM image of PMPSQT91 recovered after

50 days from the identical reaction mixture appeared

to be smaller, i.e., had no GNP rich shell layer, and an

uneven surface, as shown in Fig. 6c. Figure 6d and e

shows the TEM and EDX of shell-like debris

containing GNPs that were found near the spheres

shown in Fig. 6c. The debris should be the GNP rich

layer that had ripped off as stress was built up at the

boundary between the GNP rich layer and the core.

Fig. 5 TEM image of 100 nm PVSQT heated at 50 �C in a 0.2 mM aqueous solution of HAuCl4 for a 15 days, b 50 days, c 50 days

that contained a large GNP in the center

Fig. 6 TEM image of PMPSQT91 heated at 50 �C in a

0.2 mM aqueous solution of HAuCl4 for a 20 days, b calcined

product of particles shown in a at 1200 �C, c 50 days, d debris

found near particles shown in b in the same TEM sample,

e EDX spectrum of debris shown in d

123

J Nanopart Res (2011) 13: –2361 2374 2369

All PSQ spheres containing functional groups

prepared in this study produced GNPs both inside and

on the surface. The GNPs initially produced by the

particles produced using APTMS as the catalyst also

grew in size when they were placed in aqueous

solution of HAuCl4 for extended periods of time, but

none underwent such dramatic changes as those

observed in the PVSQT91 and PMPSQT91 particles

(Supplementary data 6).

Further investigation is needed to explain the

unusual and interesting phenomena observe in PVSQT

precisely.

The production of GNPs in thin PSQ layers was

examined by immersing the hollow spheres made of

PVSQA91 with a 43 nm thick wall in a 0.02 M

HAuCl4 solution for 1 week. Figure 7 shows GNPs

were produced inside and surfaces of the shell. These

results showed that production of GNP took place

regardless of the shape and dimension of PSQ

materials.

The porous nature of the PSQ particles was

expected to be useful for encapsulating other useful

molecules. The encapsulation of fluorophores, Rho-

damine 6G, Coumarin 7, and Fluoresceine sodium

salt was examined simply by preparing PSQA91

particles in their presence. The fluorophores did not

affect the formation of PSQ particles in terms of their

size, morphology and ability to reduce gold ions. The

PMPSQA91 particles containing these fluorophores

were unusual because they had less intensive colors

than the other PSQA91 particles containing identical

fluorophores even in less amounts. The lighter color

of PMPSQ91 might be explained by the smaller sizes

of the voids in PMPSQA91 particles than those in

other particles. Our attempts to assay the quantity of

fluorophores entrapped was unsuccessful due to

experimental difficulties, however, particles prepared

in the presence of larger amount of fluorophore had

stronger color suggesting larger amounts of fluoro-

phores were entrapped.

The intensity of fluorescent light emitted from the

PSQA particles was estimated by irradiating the

particles with an incident light with a spectrum of

wavelengths that did not overlap with that of the

fluorescent light (Supplementary data 1). Figure 8

shows that the intensity of fluorescent light increased

initially with increasing amount of entrapped

Fig. 7 a SEM image of hollow PVSQA82 spheres containing

GNPs, b TEM image of hollow PVSQA82 spheres containing

GNPs

Fig. 8 Intensity of fluorescent light plotted against the

concentration of fluorophores, right tilted triangle, PMSQA91

particles containing Rhodamine 6G, down triangle, PMSQA91

particles containing Fluoresceine sodium salt, diamond,

PVSQA91 particles containing Fluoresceine sodium salt, uptriangle, PMPSQA91 particles containing Fluoresceine sodium

salt, filled square, PMPSQA91 particles containing Coumarin

7, open square, PVSQA91 particles containing Rhodamine 6G,

left tilted triangle, PMPSQA91 particles containing Rhoda-

mine 6G, circle, PVSQA91 particles containing Coumarin 7

123

J Nanopart Res (2011) 13: –2361 23742370

fluorophores and then leveled off or decreased slightly

after reaching a maximum value. These results corre-

spond to the well-known fact that the fluorophores emit

weaker fluorescent light when they are concentrated

due to self-quenching. The intensity of fluorescent light

was the strongest when the fluorophores were

entrapped in PMSQA than PVSQA and PMPSQA.

Hollow PVSQA82 prepared in the presence of fluoro-

phores also gave particles with strong fluorescence

(Supplementary data 7).

It is known that amines bind to gold surface very

weakly and amine groups have been rarely used to fix

biological molecules to gold. The color of PSQ

particles prepared in this study commonly turned dark

purple when they were stored in colloidal solutions of

GNPs of different sizes. TEM images of the colored

particles revealed that significant amounts of GNPs

were bound. PSQA/GNP (2.8 nm diameter) particles

that were subjected to heating in 95 �C water for 6 h,

vortexing (5 min at maximum speed), and ultrasonic

treatment (10 min, 5 W) were virtually identical to the

original ones as observed with TEM. The supernatant

water was colorless as observed with naked eyes when

the PSQA/GNP particles described above were cen-

trifuged down. These results corresponded to those

reported earlier in that GNP bind quite strongly to

amine compounds in contrast to bulk gold (Pong et al.

2005).

Figure 9a and b shows TEM images of PMPSQT

and PMPSQA91 particles, respectively, that were

treated with 30 nm GNPs. Figure 9 shows that the

number of GNPs attached to the PMPSQA91 particles

were much larger than that of PMPSQT. These results

were unexpected because PMPSQT particles con-

tained only 3-mercaptosilsesquioxane units while

PMPSQA91 particles contained 10 mol% of amino

and 90 mol% of mercapto groups, and mercapto group

was stronger binder to gold than amine group. These

results can be explained that hydrophilic 3-aminopro-

pylsilsesquioxane groups assisted the mercapto group

in 3-mercaptopropyl group be exposed to surface of

the particles.

Sizable amounts of GNPs bound to PMSQT

particles that contained no organic functional groups

probably due to the interaction between GNPs and

silanol units and non-specific interaction between

small particles. The number of GNPs bound to

PMSQA91 particles was also largely enhanced due to

the amine group (Supplementary data 8).

The formation of stable bonds between GNP and

functional PSQ particles, especially those containing

amine groups, suggested that particles with DNA

strands would be easily prepared by mixing amino-

DNA strands with the PSQ particles where GNPs

have already been bound. The composite particles of

PSQ, GNP, and DNAx would be useful to detect and

isolate DNA strands that contain fractions with

complementary base sequence of DNAx.

The applicability of these particles to detecting

oligo-DNA strands was examined. The strategy was to

prepare two sets of PSQ/GNP composite particles and

two complementary DNA strands with amino termi-

nal group. Each of DNA strands is combined with

PSQ/GNP via interaction between GNP and amine

group. Annealing the two sets of particles would

aggregate if the complementary DNA strands on the

particles hybridize. This method is more advanta-

geous over other methods that need chemical reac-

tions to bind biological molecules to solid supporters;

chemical reactions, especially carried out in coupling

in aqueous systems, are often difficult to control. It

should also be noted that amino-DNA can be used

instead of mercapto-DNA; mercapto-DNA needs to

be reduced prior to use that require additional non-

quantitative process.

Applying these particles to the detection of oligo-

DNA was impeded by two problems. The first

problem was caused by the attractive interactions

between small particles with and/or without attached

DNA stands. The second problem was distinguishing

aggregates of particles with two complementary

DNAs from those formed non-specifically.

The first problem regarding the aggregation of small

particles could be resolved by adding detergent such as

Tween 20. Attempts to resolve the second problem was

made using fluorescent particles and non-fluorescent

particles. Results showed that the fluorescent light

emitted by the fluorescent particles was so strong that it

was not possible to distinguish between non-fluores-

cent particles and fluorescent particles in aggregated

bodies. This problem was resolved using small fluo-

rescent particles and much larger non-fluorescent

particles. For example, 200 nm PMPSQA91 particles

containing Rhodamine 6G were invisible by a micros-

copy in normal mode but were visible in fluorescence

mode. On the other hand, non-fluorescent 2000 nm

PMSQA91 particles were invisible in fluorescent mode

but were visible in normal mode.

123

J Nanopart Res (2011) 13: –2361 2374 2371

Figure 10 shows SEM and TEM images of the non-

fluorescent 2000 nm PMPSQA91/GNP/DNA1 parti-

cles and fluorescent 200 nm PMPSQA91/GNP/DNA2

particles used in this study, respectively. These PSQ

particles behaved precisely as expected. Figure 11

shows microscopy images of control mixture of non-

fluorescent 2000 nm PMPSQA91/GNP/DNA1 parti-

cles and fluorescent 200 nm PMSQA91/GNP/DNA1

particles that were heated at 90 �C and cooled to room

temperature. The large particles and small particles

were, respectively, indicated by thick and thin arrows.

Figure 11 shows that the small fluorescent particles

were virtually invisible in normal mode and the large

non-fluorescent particles were invisible in fluores-

cence mode. Figure 11 also shows that the particles

did not aggregate seriously under these conditions.

The mixtures of the colloidal solutions of fluores-

cent 200 nm PMSQA91/GNP/DNA1 particles and

2000 nm PMPSQA91/GNP/DNA2 particles treated in

the same manners formed large aggregates as seen in

Fig. 11c and d. The presence of large non-fluorescent

and small fluorescent particles in the aggregates were

confirmed by observing them in normal and fluores-

cence modes. The mixtures of PMSQA91/GNP/DNA3

and PMSPSQA91/GNP/DNA4 gave similar results.

These results clearly showed PSQ particles were

Fig. 9 TEM image

of a PMPSQT particles

treated with 30 nm GNPs,

b PMPSQA91 particles

treated with 30 nm GNPs

Fig. 10 a SEM image of

2000 nm PMPSQA91/GNP/

DNA1 particles, b SEM

image of 200 nm

PMPSQ91/GNP/DNA2/

Rhodamine 6G particles,

c TEM image of 200 nm

PMPSQ91/GNP/DNA2/

Rhodamine 6G particles,

d TEM image of 200 nm

PMPSQA91/GNP/DNA1

particles

123

J Nanopart Res (2011) 13: –2361 23742372

aggregated by hybridization of the complementary

DNA strands that were fixed on the PSQ particles via

GNPs. TEM analysis also showed the presence of

aggregates of small and large particles (Supplemen-

tary data 9).

This method is convenient compared to other

methods that involve chemical reactions to couple

biological molecules to solid supports because the

chemical reactions involves multiple steps and are

often difficult to control. Also, it is advantageous that

amino-DNA strands can be used that are simpler to

handle than mercapto-DNA that usually needs to be

pre-treated before use. It is expected that these

particles should be useful for assaying and isolating

DNA strands with specific sequences as demonstrated

above.

Conclusion

PSQ particles containing different functional groups

with a narrow size distribution with controlled diam-

eters ranging from tens of nanometers to thousands of

nanometers could be prepared in one pot reaction

simply by adjusting the composition of the reaction

mixtures. These particles contained pores inside that

be altered afterwards, particularly for PVSQT parti-

cles. The pores inside the PSQ particles could be used

to encapsulate fluorophores to obtain PSQ particles

with different sizes with tunable fluorescence intensi-

ties. The GNPs bound strongly to the amine and thiol

containing PSQ particles, and the PSQ/GNP particles

could be applied to a DNA assay after binding the

amine-modified DNA strands. The results highlight

the potential applications of functional PSQ particles

in a range of fields, including biomedical and cosmetic

applications.

Acknowledgments This study was supported by a grant of the

Korea Healthcare technology R&D Project, Ministry for Health,

Welfare & Family Affairs, Republic of Korea (A084333).

References

Corr SA, Rakovich YP, Gun’ko YK (2008) Multifunctional

magnetic-fluorescent nanocomposites for biomedical

applications. Nanoscale Res Lett 3:87–104. doi:10.1007/

s11671-008-9122-8

Eaton P, Doria G, Pereira E, Baptista PV, Franco R (2007)

Imaging gold nanoparticles of DNA sequence recognition

Fig. 11 Microscopic image

of annealed mixtures of

2000 nm PMPSQA91/GNP/

DNA1 and 200 nm

PMPSQA91/GNP/DNA1

particles obtained

in a normal mode,

b fluorescence mode;

Microscopic image of

annealed mixtures of

2000 nm PMPSQA91 GNP/

DNA1 and 200 nm

PMPSQA91/GNP/DNA2/

Rhodamine 6G particles

observed in c normal mode,

d fluorescence mode

123

J Nanopart Res (2011) 13: –2361 2374 2373

in biomedical application. IEEE Trans Nanobiosci 6:282–

288. doi:10.1109/TNB.2007.908985

Heitsch AT, Smith DK, Patel RN, Ress D, Korgel BA (2008)

Multifunctional particles: magnetic nanocrystals and gold

nanorods coated with fluorescent dye-doped silica shells.

J Solid State Chem 181:1590–1599. doi:10.1016/j.jssc.

2008.05.002

Kim Y, Kim YB, Yoon K (2006) Preparation of functionalized

polysilsesquioxane and polysilsesquioxane-metal nano-

particle composite spheres. Macromol Rapid Commun

27:1247–1253. doi:10.1002/(ISSN)1521-3927

Klostranec JM, Chan WCW (2006) Quantum dots in biological

and biomedical research: recent progress and present

challenges. Adv Mater 18:1953–1964. doi:10.1002/adma.

200500786

Lee D, Kim Y, Kim YB, Kim J, Han Y (2008) One step

preparation of spherical silicon resins from immiscible

reaction. Mixtures Macromol Res 16:353

Lin YS, Wu SH, Tseng CT, Hung Y, Chang C, Mou CY (2009)

Synthesis of hollow silica nanospheres with a micro-

emulsion as the template. Chem Commun 3542–3544.

doi:10.1039/b902681a

Liong M, Lu J, Kovochich M, Xia T, Ruehm SG, Nel AE,

Tamanoi F, Zink JI (2008) Multifunctional inorganic

nanoparticles for imaging, targeting, and drug delivery.

ACS Nano 2:889–896. doi:10.1021/nn800072t

Niemeyer CM, Simon U (2005) DNA-based assembly of metal

nanoparticles. Eur J Inorg Chem 3641–3655. doi:10.1002/

ejic.200500425

Parak WJ, Gerion D, Pellegrino T, Zanchet D, Micheel C,

Williams SC, Boudreau R, Le Gros MA, Larabell CA,

Alivisatos AP (2003) Biological applications of colloidal

nanocrystals. Nanotechnology 14:R15–R27. doi:10.1088/

0957-4484/14/7/201

Pasqua L, Cundari S, Ceresa C, Cavaletti G (2009) Recent

development, applications, and perspectives of mesoporous

silica particles in medicine and biotechnology. Curr Med

Chem 16:3054–3063. doi:10.2174/092986709788803079

Pellegrino T, Kudera S, Liedl T, Javier AM, Manna L, Parak

WJ (2005) On the development of colloidal nanoparticles

towards multifunctional structures and their possible use

for biological applications. Small 1:48–63. doi:10.1002/

smll.200400071

Pong BK, Lee JY, Trout BL (2005) First principles computa-

tional study for understanding the interactions between

ssDNA and gold nanoparticles: adsorption of methylamine

on gold nanoparticulate. Langmuir 21:11599–11603. doi:

10.1021/la052116k

Ren CL, Li JH, Liu Q, Ren J, Chen XG, Hu ZD, Xue DS (2008)

Synthesis of organic dye-impregnated silica shell-coated

iron oxide nanoparticles by a new method. Nanoscale Res

Lett 3:496–501. doi:10.1007/s11671-008-9186-5

Schottner G (2001) Hybrid sol–gel-derived polymers: applica-

tions of multifunctional materials. Chem Mater 13:3422–

3435. doi:10.1021/cm011060m

Smith AM, Duan HW, Mohs AM, Nie SM (2008) Bioconjugated

quantum dots for in vivo molecular and cellular imaging.

Adv Drug Deliv Rev 60:1226–1240. doi:10.1016/j.addr.

2008.03.015

Tan WB, Zhang Y (2005) Multifunctional quantum-dot-based

magnetic chitosan nanobeads. Adv Mater 17:2375–2380.

doi:10.1002/adma.200401650

Thaxton CS, Geroganopoulou DG, Mirkin CA (2006) Gold

nanoparticle probes for the detection of nucleic acid tar-

gets. Clin Chim Acta 363:120–126. doi:10.1016/j.cccn.

2005.05.042

Yang S, Liu HR, Zhang ZC (2009) Design and fabrication of

hollow, magnetic and fluorescent CdS-magnetite-poly

(styrene-co-methyl methacrylate) microspheres. New J

Chem 33:620–625. doi:10.1039/b811418h

123

J Nanopart Res (2011) 13: –2361 23742374