Dynamic Article LinksC<Nanoscale

Cite this: Nanoscale, 2012, 4, 2326

www.rsc.org/nanoscale PAPER

Publ

ishe

d on

30

Janu

ary

2012

. Dow

nloa

ded

by U

nive

rsity

of

Win

dsor

on

29/1

0/20

14 1

2:19

:38.

View Article Online / Journal Homepage / Table of Contents for this issue

Synthesis of polymer/inorganic nanocomposite films using highly porousinorganic scaffolds†

Huanjun Zhang,a Matthias Popp,b Andreas Hartwigb and Lutz M€adler*a

Received 19th December 2011, Accepted 27th January 2012

DOI: 10.1039/c2nr12029a

Polymeric/inorganic nanocomposite films have been fabricated through a combination of flame-spray-

pyrolysis (FSP) made inorganic scaffold and surface initiated polymerization of cyanoacrylate. The

highly porous structure of pristine SnO2 films allows the uptake of cyanoacrylate and the

polymerization is surface initiated by the water adsorbed onto the SnO2 surface. Scanning electron

microscopy study reveals a nonlinear increase in the composite particle size and the film thickness with

polymerization time. The structural change is rather homogeneous throughout the whole layer. The

composite is formed mainly by an increase of the particle size and not by just filling the existing pores.

High-resolution transmission electron microscopy imaging shows SnO2 nanoparticles embedded in the

polymeric matrix, constituting the nanocomposite material. Thermogravimetric analysis indicates that

the porosity of the nanocomposite films decreases from 98% to 75%, resulting in a significant

enhancement of the hardness of the films. DC conductivity measurements conducted in situ on the

nanocomposite layer suggest a gradual increase in the layer resistance, pointing to a loss of connectivity

between the SnO2 primary particles as the polymerization proceeds.

Introduction

Synthesis of polymeric/inorganic hybrid nanocomposite mate-

rials with structural control of both atomic and macroscopic

dimensions represents a key requirement for obtaining the

increased functionality from their interface. Through proper

combination of the inorganic and polymeric constituents, supe-

rior mechanical strength, thermal stability and electrochemical

reactivity can be achieved over the individual component mate-

rials, thanks to their high degree of hybridization.1

Conventional synthesis of hybrid composite materials gener-

ally relies on physical blending or chemical surface modification

of the constituents.2–4 Very often a chemical surface modification

is carried out first in order to improve particle-matrix compati-

bility, followed by physical mixing. In this regard we have carried

out extensive work on silica/epoxy nanocomposites.5,6 For

nanocomposites with a high inorganic loading, the chemical

surface modification is particularly attractive because it enables

introduction of a second component directly into a pre-formed

hosting matrix, leading to a defined distribution of the

aFoundation Institute of Materials Science (IWT), Department ofProduction Engineering, University of Bremen, Badgasteiner Strasse 3,28359 Bremen, Germany. E-mail: lmaedler@iwt.uni-bremen.de; Fax:+49 421 2185378; Tel: +49 421 21851200bFraunhofer–Institute for Manufacturing Technology and AdvancedMaterials (IFAM), Wiener Strasse 12, 28359 Bremen, Germany.E-mail: andreas.hartwig@ifam.fraunhofer.de; Fax: +49 421 2246430;Tel: +49 421 2246470

† Electronic supplementary information (ESI) available: SEM imagesand powder XRD patterns. See DOI: 10.1039/c2nr12029a

2326 | Nanoscale, 2012, 4, 2326–2332

constituent species. This method has been successfully applied in

the fabrication of nanocomposites based on cyanoacrylate and

silica aerogels that act as the skeleton matrix.7 However, its

extension is rather difficult because of the very limited choice of

the inorganic components in an aerogel form, besides the strin-

gent (therefore costly) conditions required for making the aerogel

itself.8

Here, we present a low-cost and highly efficient method for the

fabrication of hybrid nanocomposite films based on surface-

initiated polymerization (SIP),9,10 which has proven to be

a powerful tool for chemical functionalization of the surface of

a wide range of materials including carbon nanotubes,11 silica,12

metals13 and polymers.14 The motivation of adopting the SIP

method here is to harness the surface polymerization process for

making film-structured nanocomposite materials. A highly

porous inorganic layer was first fabricated using flame spray

pyrolysis15 through thermophoretic deposition of metal oxide

nanoparticles onto a glass substrate. Subsequently, this porous

layer serves as a scaffold, around which polymer components

have been loaded through the polymerization of monomers

introduced via the vapor phase to generate the polymer/inorganic

nanocomposite films. Semiconducting SnO2 is chosen as the

inorganic material to examine the structure evolution by

measuring in situ the electrical conductivity of the composite

films. The vapor-phase introduction of a monomer is necessary

for achieving homogeneous dispersion of the SnO2 nanoparticles

in the polymer matrix, because, if liquid-phase polymerization is

employed, the inorganic scaffold is not mechanically strong

enough to withstand the forces (e.g. the capillary force), which

This journal is ª The Royal Society of Chemistry 2012

Publ

ishe

d on

30

Janu

ary

2012

. Dow

nloa

ded

by U

nive

rsity

of

Win

dsor

on

29/1

0/20

14 1

2:19

:38.

View Article Online

can lead to the collapse of the scaffold and a composite product

with poor particle dispersion similar to composites obtained by

simple blending. A comparison between conventional ‘‘blending’’

and our ‘‘reverse’’ filling method is illustrated in Scheme 1.

Considering the wide spectrum of inorganic materials including

metal oxide semiconductors that can be generated by FSP, the

reverse filling method here should be generalizable to the fabri-

cation of various nanocomposites based on porous, inorganic

scaffolds.

Experimental

Fabrication of porous SnO2 films through FSP

SnO2 films composed of crystalline nanoparticle agglomerates

were deposited onto glass substrates in a flame spray reactor.

Detailed description of the setup can be found elsewhere.15 The

liquid precursor solution containing 0.5 mol L�1 Sn(II) ethyl-

hexanoate (90 wt% in 2-ethylhexanoic acid, STREM Chemicals)

in xylene (98.5%, BDH) was fed by a syringe pump (KD Scien-

tific) at a fixed flow rate of 5.0 mL min�1. The liquid was

dispersed into fine droplets by oxygen gas (5.0 L min�1, 1.5 bar;

99.95 vol%) at the spray nozzle exit. The spray was ignited by

a CH4/O2 (1.5 and 3.2 L min�1, respectively; CH4–99.5 vol%,

Westfalen) flame to form a self-sustained flame. The generated

SnO2 particles deposit thermophoretically onto a water-cooled

glass substrate (20 mm � 20 mm, 1.0 mm thick; Menzel-Gl€aser)

placed 20.0 cm above the nozzle. Before FSP deposition, the glass

substrates were cleaned by acetone (99.5%, Sigma-Aldrich) in an

ultrasonic bath. The flow rates of all gases were controlled by

calibrated mass flow controllers (Bronkhorst High-Tech).

Polymerization of cyanoacrylate in the SnO2 layer

Polymerization of ethyl cyanoacrylate (90 wt%, Henkel Loctite)

was carried out in an oven at 70 �C. The glass supported SnO2

layer was placed closely on top of a vial containing ethyl

cyanoacrylate, allowing for the penetration of the evaporated

cyanoacrylate molecules into the porous space in the SnO2 layer.

Water is present on the surface of SnO2 particles due to handling

under ambient conditions and the adsorbed water should initiate

the polymerization of cyanoacrylate. The duration of polymeri-

zation was controlled in order to study the kinetics of the growth

of the nanocomposite films.

Scheme 1 Strategies for fabricating nanocomposite films containing inorga

composites with usually low particle content, where surface treatment on the n

matrix and (b) ‘‘reverse’’ filling method presented here which uses surface-initia

leading to composites largely maintaining the particle connectivity.

This journal is ª The Royal Society of Chemistry 2012

Characterization

Powder XRD analysis. Both the as-deposited SnO2 films and the

filter-collected SnO2 powders were subjected to XRD analysis in

order to study the crystallinity of SnO2. Powder XRD patterns

were measured by a PANalytical X’pert diffractometer with a Cu

Ka radiation source. Rietveld refinement was performed on the

XRD patterns to estimate the average crystallite size of SnO2.

N2 physisorption analysis. The Brunauer–Emmett–Teller

(BET) specific surface area of the filter-collected SnO2 particles

was measured by N2 physisorption at liquid nitrogen tempera-

ture. High-resolution transmission electron microscopy (TEM)

of the SnO2 nanoparticles was conducted on an FEI Titan

microscope under a 300 kV acceleration voltage. Scanning elec-

tron microscopy (SEM, Leo 1530 Gemini, Zeiss) was employed

to investigate the morphology of the SnO2 films and to estimate

the thickness of both the SnO2 films and the polycyanoacrylate/

SnO2 composite films. The morphology of the polymer/SnO2

hybrid particles was investigated using transmission electron

microscopy (Philips CM20).

Thermogravimetric analysis (TGA). The nanocomposite films

were analyzed using a TGA Q5000 analyzer (TA Instruments)

between room temperature and 900 �C at the heating rate of 5 �Cmin�1 with air purging. The measured weight loss due to polymer

degradation was used to calculate the volumetric percentage of

the polymer phase. Combined with the film thickness data, the

porosity of the nanocomposite films was derived. The Martens

micro-hardness tests were performed on a Fischerscope HM2000

micro-indenter (Helmut Fischer).

DC electrical conductivity measurements. The electrical

conductivity of the polymer/SnO2 films was measured to provide

in-depth insights into the growth of the composite films during

polymerization. In particular, DC conductivity measurements

reveal the connectivity between SnO2 nanoparticles and predict

the overall percolating structure in the whole nanocomposite

films. To obtain reliable conductivity data, planar alumina

substrates carrying interdigitated Pt electrodes16 were used for

the deposition of SnO2 nanoparticles and subsequent polymeri-

zation of cyanoacrylate. The DC conductivity of the nano-

composite films was measured in situ with a FLUKE112 digital

multimeter during the polymerization process, where the SnO2-

coated electrode was placed above the monomer reservoir. At

nic nanoparticles and polymers: (a) conventional blending method for

anoparticles is often necessary to achieve good dispersion in the polymer

ted polymerization into porous, FSP-deposited nanoparticulate scaffolds,

Nanoscale, 2012, 4, 2326–2332 | 2327

Publ

ishe

d on

30

Janu

ary

2012

. Dow

nloa

ded

by U

nive

rsity

of

Win

dsor

on

29/1

0/20

14 1

2:19

:38.

View Article Online

room temperature the resistance of the deposited pure SnO2 layer

is higher than 200MU, which is beyond the detection range of the

applied ohmmeter. To reduce its resistance, the SnO2 layer was

doped with Sb during the FSP process, which can render a drop

in resistance by several orders of magnitude, depending on the Sb

concentration.17 In the present work, the Sb concentration is 1%

in the weight ratio between Sb and SnO2.

Fig. 2 Cross-sectional SEM images of the polycyanoacrylate/SnO2

nanocomposite film showing a homogeneous thickness (a) and a porous

structure (b) after 30 minutes of polymerization.

Results and discussion

Pristine SnO2 scaffolds

Themorphology of the pristine SnO2 scaffolds deposited viaflame

spraypyrolysis on a glass substrate is shown inFig. 1.Thefilms are

highly porous with an even distribution of the pores over milli-

metre dimension (Fig. 1a). The film porosity was estimated to be

ca. 98% both experimentally and theoretically.15,18 A closer

inspection (Fig. 1b) of the SnO2 film reveals that it consists of

aggregates of SnO2 nanoparticles, forming a three-dimensional,

highly porous network. The SEM images here show that both

meso- andmacropores are present in the SnO2 film. PowderX-ray

diffraction combined with Rietveld refinement (Fig. 1c) was

conducted on the filter-collected particles, showing that the crys-

tallites have an average size of 8.7 nm.This value is consistentwith

the results from BET specific surface area analysis, which give

a specific surface area of 97.8 m2 g�1 and indicate an average grain

size of 8.8 nm as derived from the equation dBET ¼ 6/(r � SSA)

(where r is the density of bulk SnO2, 6.95 g cm�3; SSA stands for

the specific surface area of the powder, m2 g�1). The dBET error is

within�2% based on multiple measurements. These particles are

therefore individually single-crystalline, which is further sup-

ported by TEM characterization (Fig. 1, inset).

Polymer/SnO2 nanocomposite films

The morphology and structure of the SnO2 film after being

subjected to polymerization of cyanoacrylate are shown in Fig. 2.

Fig. 1 Characterization of SnO2 layers and powders obtained from FSP: (a)

layer showing inter-connected strands of nanoparticles; and (c) Rietveld refi

8.7 nm—in black is the raw data, in red is the calculated pattern and in blue is t

the diffraction peaks of SnO2). The inset shows a high-resolution TEM of th

between 5 and 10 nm, in consistence with the XRD results.

2328 | Nanoscale, 2012, 4, 2326–2332

The cross-sectional image (Fig. 2a) indicates a highly even

thickness of the composite film, i.e. 54.2 mm, which is 35% thicker

than the starting pristine SnO2 films (40 mm) after 30 min of

vapor phase polymerization. Higher-resolution imaging of the

particles at the cross-section region (Fig. 2b) reveals an average

particle size of 23.0 nm, based on the measurement of at least 200

particles. Considering the initial average particle size of SnO2, i.e.

8.8 nm, nearly a three-fold increase in particle size resulted from

polymerization for 30 minutes. The composite film remains

porous and contains both meso- and macropores as shown in

the SEM image (Fig. 2b). The morphological resemblance of the

composite film to the pristine SnO2 film suggests that the

SEM top-view of the SnO2 layer; (b) higher-resolution SEM of the SnO2

nement of the powder XRD profile giving an average crystallite size of

he difference between them (the vertical bars below denote the positions of

e SnO2 nanoparticles with high crystallinity and an average particle size

This journal is ª The Royal Society of Chemistry 2012

Publ

ishe

d on

30

Janu

ary

2012

. Dow

nloa

ded

by U

nive

rsity

of

Win

dsor

on

29/1

0/20

14 1

2:19

:38.

View Article Online

polymerization of cyanoacrylate within the porous film should be

initiated on the surface of SnO2, i.e. by the surface adsorbed

water on SnO2 nanoparticles, similar to the room-temperature

chemical vapor deposition of cyanoacrylate in silica aerogels.7

The water content in the collected SnO2 powders was estimated

as 3.7% (weight percentage) from thermogravimetric analysis.

This value corresponds to a 1.4 monolayer of water molecules on

the surface of SnO2 particles, which give rise to –OH groups or

OH� ions as the initiator.

Growth of the nanocomposite films

In order to reveal how the nanocomposite film grows during the

polymerization process, we investigated the morphology and

structure evolution of the composite films that have been sub-

jected to polymerization for various durations. The cross-

sectional SEM images were recorded for polymerization times of

10, 30, 60, 120 and 180 minutes corresponding to Fig. 3a–e,

respectively. At the same polymerization temperature (70 �C),apparently the average size of the polymer/SnO2 composite

particles is increasing as the reaction time increases. The porous

structures within the films seem also preserved within 180

minutes. Fig. 3f shows the structure of a pure polycyanoacrylate

layer grown in 120 minutes on glass at the same temperature. The

layer possesses a distinctly different morphology and much lower

porosity as compared with the nanocomposite counterparts

(Fig. 3a–e). The scale bar of 500 nm applies to Fig. 3a–f.

Although the images in Fig. 3a–e were recorded for the top-layer

Fig. 3 Cross-sectional SEM images of the polycyanoacrylate/SnO2

nanocomposite films generated after different polymerization times: (a)

10 min, (b) 30 min, (c) 60 min, (d) 120 min and (e) 180 min. A pure

polycyanoacrylate layer grown on the glass substrate is shown in (f). The

scale bar representing 500 nm applies to all images.

This journal is ª The Royal Society of Chemistry 2012

region of the films, it is important to note that the morphological

feature for each sample is homogeneous throughout its cross-

sectional region (ESI, Fig. S1†).

The evolution of the average composite particle size is shown

in Fig. 4a. After 10 minutes of polymerization the composite

particles (see Fig. 3a) were found to be slightly larger than the

starting SnO2 particles (Fig. 1b), indicative of the formation of

a thin polymer layer coating individual SnO2 particles. As the

polymerization time increases, the particle size grows further.

After 120 minutes the measured average particle size reached

�150 nm, nearly 18 times that of the pristine SnO2 particles.

However, it seems rather unlikely that each SnO2 particle has

been coated by a thick layer (�70 nm) of polycyanoacrylate.

Such a huge increase in the particle size indicates that several

SnO2 particles are now engulfed in a single ‘‘composite’’ particle,

accounting for the steep increase of the particle size after 30

minutes of polymerization (Fig. 4a), which suggests a coating

mechanism different from that occurred during the initial poly-

merization stage (i.e., within 10 minutes).

Fig. 3d and e show that the composite particle size ceases to

grow after 120 minutes of polymerization, suggesting a termina-

tion of the polymeric chain growth. A similar trend was also

observed in the thickness change of the composite films. As

shown in Fig. 4b, the film thickness increases in a nonlinear

fashion, and it almost doubles the original SnO2 layer thickness

after polymerization for 180 minutes (�75 mm). Actually after

120 minutes, the thickness of the composite film appears to reach

a plateau. A control experiment was conducted by refreshing the

monomer reservoir during the polymerization process, which

Fig. 4 Variation of the average size of the nanocomposite particles (a)

and the film thickness (b) during polymerization. A sudden rise in the

slope of the particle size curve suggests a change in the mode of the

composite particle growth.

Nanoscale, 2012, 4, 2326–2332 | 2329

Publ

ishe

d on

30

Janu

ary

2012

. Dow

nloa

ded

by U

nive

rsity

of

Win

dsor

on

29/1

0/20

14 1

2:19

:38.

View Article Online

resulted in a thicker composite film than that obtained without

refreshing the monomer reservoir (Fig. S2†). This indicates that

the slowing down of the nanocomposite film growth should be

caused by the decrease in the monomer supply from the reservoir,

because the moisture inside the reaction chamber could also

initiate polymerization within the reservoir body, thus reducing

the evaporation rate of the monomer into the vapor phase.

While SEM provides valuable information on the thickness

and composite particle size of the nanocomposite films, it is also

important to closely study the interface between the polymer and

the semiconductor particles. Considering Fig. 2b, apparently the

SnO2 nanoparticles are embedded in the polymer surroundings.

A critical question is, how are the semiconductor particles

spatially arranged inside individual nanocomposite particles? We

address this question by examining the samples using high-

resolution transmission electron microscopy (HRTEM). As

shown in Fig. 5, the nanocomposite particle consists of an

ellipsoidal-shaped polymer matrix which embeds several SnO2

nanocrystals (Fig. 5a, the low contrast in the amorphous region

is expected between the polycyanoacrylate and the amorphous

carbon from the TEM grid). The crystal lattice fringes of SnO2

can be clearly identified in Fig. 5b. These observations suggest

that the polymerization propagates by filling the small pores by

capillary condensation with a concomitant polymerization of the

cyanoacrylate. As a result, several neighboring particles are

‘‘bound’’ to form virtually one bigger particle, which accounts for

the sudden increase in the average particle size as shown in

Fig. 4a.

Fig. 5 HRTEM images of polycyanoacrylate/SnO2 nanocomposite

particles: (a) one composite particle composed of several SnO2 nano-

crystals (darker particles) embedded in polycyanoacrylate and (b) the

lattice fringes of the SnO2 crystals are observable with higher magnifi-

cation at the framed region in (a).

2330 | Nanoscale, 2012, 4, 2326–2332

Thermogravimetric analysis, porosity and Martens hardness

The weight percentage of PCA in the nanocomposite films was

measured by TGA, in order to estimate the porosity of the

products. Since the density of PCA in this study is not exactly

known, we assume a density of 1 g cm�3 in order to calculate the

volume ratio between PCA and SnO2. The time-dependent

variations of both the PCA weight percentage and the film

porosity are shown in Fig. 6. The PCA weight% increases with

the polymerization time and it reaches 73% after 180 min, which

is nearly three times that of the inorganic phase. Considering the

larger density difference between PCA (assuming 1 g cm�3) and

SnO2 (6.95 g cm�3), the volume ratio between the two phases is

significant.

The porosity is calculated from the solid volume percentage

(i.e., summation of both the polymer and inorganic phases),

which is estimated by combining TGA and thickness data. The

porosity drops from 98% for the pure SnO2 layer (data not

shown here) to 75% after 180 minutes of polymerization.

Therefore, the nanocomposite films remain rather porous after

polymerization. As the porosity decreases, an increase in the

Martens hardness is observed: the hardness after 180 min is 22

times that after 10 minutes of polymerization, suggesting that the

mechanical strength of the scaffold has been significantly

improved after incorporating the polymer phase. However, the

hardness of the nanocomposite films is still much lower than that

of pure PCA, 141.4 N mm�2. This can also be explained by the

porosity difference, because the pure PCA layer is quite compact

and shows very low porosity (Fig. 3f). The results shown here are

consistent with the reported dependence of the mechanical

properties on porosity for a wide range of materials, being

micro-, meso- or macro-porous.19–21

DC conductivity and particle connectivity

Another important issue concerns the connectivity between the

SnO2 nanoparticles after being immersed in or surrounded by the

polymer matrix. TEM encounters great difficulties in resolving

this question because: (1) TEM images are 2D projections of 3D

structures and (2) TEM provides only local information. To

study on a macroscopic level the connectivity between the SnO2

Fig. 6 Variation of the polymer weight percentage, film porosity and

micro-hardness of the layers as the polymerization time increases.

This journal is ª The Royal Society of Chemistry 2012

Publ

ishe

d on

30

Janu

ary

2012

. Dow

nloa

ded

by U

nive

rsity

of

Win

dsor

on

29/1

0/20

14 1

2:19

:38.

View Article Online

particles, we measured in situ the conductivity of the nano-

composite films. Because polycyanoacrylate is non-conductive,

the conductivity of the nanocomposite film solely depends on the

conductivity of the SnO2 particles that form the percolating

conductive pathways.22,23

We have used such an interdigital electrode-carrying substrate

as shown in Fig. 7a to measure the in situ variation of the film

conductivity. Taking the advantage of FSP that can easily dope

a semiconductor material, we doped SnO2 with antimony to

obtain a higher electrical conductivity of the deposited films.24

One percent (wt%) of Sb renders an SnO2 layer resistance low

enough for us to precisely measure the resistance of the films

during polymerization using an ohmmeter. The pure and 1 wt%

Sb-doped SnO2 samples give almost identical powder XRD

patterns (Fig. S3†). Their BET surface areas are 97.8 and 96.7 m2

g�1 for the pure and doped SnO2, respectively. These data suggest

that Sb doping should bring negligible change to the SnO2

nanoparticles and the structure of the SnO2 layers during the

FSP process. Hence, the polymerization results obtained using

the Sb-doped SnO2 scaffolds should reflect insights that are

transferrable to the pure SnO2 system.

Fig. 7b schematically illustrates the experimental setup, where

the semiconductor layer should cover only the region of the elec-

trode arrays and leave the contact pads exposed. As the polymer-

ization proceeds, an increase in the resistance is observed (Fig. 7c,

which plots the reciprocal of resistance vs. polymerization time),

largely because of the rupture of the percolating pathways.

In a very simplified model, which assumes that each perco-

lating pathway is a single strand of particles in connection, the

reciprocal of resistance values are proportional to the specific

Fig. 7 Electrical conductivity measurement on polycyanoacrylate/SnO2

nanocomposite films: (a) a substrate carrying electrode arrays for Sb-

doped SnO2 scaffold deposition, (b) schematic illustration of the

conductivity measurement setup and (c) the reciprocal resistance varia-

tion with the polymerization time.

This journal is ª The Royal Society of Chemistry 2012

conductivity and hence the number of the undisturbed perco-

lating pathways: after 30 minutes, more than 50% of the perco-

lating pathways remains; even after 2 hours, there is still about

20% of the pathways undisturbed by the polymerization. From

a more realistic point of view, neighboring particles coated with

a very thin layer of PCA (hence separated by a rather short

distance) might still constitute a conductive pathway because of

the tunneling effect. Considering that it takes only disconnecting

two neighboring SnO2 particles within a percolating strand

(typically consisting of tens of particles) to shut down its

conductivity, we expect that the average connectivity of SnO2

particles should be still higher than estimated by the conductivity

measurement. This conclusion is highly encouraging for the

design and fabrication of inorganic/polymer nanocomposite

materials. But on a level more sophisticated than this simplified

model here, further work needs to be done to reveal the details of

the nanocomposite growth.

Conclusions

We demonstrate a facile and a highly efficient method of fabri-

cating polymeric/inorganic nanocomposite films by making use

of the highly porous nature of the inorganic films generated from

flame spray pyrolysis. Polycyanoacrylate was successfully loaded

into the inorganic SnO2 layer through surface-initiated anionic

polymerization. The dispersion of SnO2 nanoparticles was

largely maintained during the polymerization process, leading to

a nanocomposite material with an inorganic constituent

throughout the polymeric matrix forming a percolation struc-

ture. Such a structure cannot be obtained by incorporation of

comparably low amounts of particles into the polymer. Because

flame spray pyrolysis is capable of generating a wide range of

metal-oxide materials,25 the method presented here can have

widespread applicability in the fabrication of nanocomposites

covering a large variety of polymer/metal-oxide constituents.

From our measurement results, we propose the following

model for the formation of the composite with low inorganic-

particle loading but proper inter-particle contact. The cyanoac-

rylate vapor is adsorbed on the surface of the SnO2 particles

which bear the adsorbed water to initiate the polymerization of

the cyanoacrylate and lead to an increase of the composite

particle diameter. Consumption of the cyanoacrylate by poly-

merization maintains the monomer concentration gradient. This

is supported by capillary condensation, leading to quick filling of

the small pores and the formation of composite particles con-

taining SnO2 particles separated by a polymer layer. In general,

the anionic polymerization of the cyanoacrylate is a living kind

of polymerization, but contamination on the particle surfaces

can lead to termination. And if all water on the surface is

consumed, no further polymerization is initiated. Reduced

supply of the monomer from its reservoir can also lead to the

saturation in the composite particle size and the film thickness

while the film is still porous.

Acknowledgements

The authors would like to thank Deutsche For-

schungsgemeinschaft (DFG) for funding this project within the

Research Training Group 1375 ‘‘Nonmetallic Porous Structures

Nanoscale, 2012, 4, 2326–2332 | 2331

Publ

ishe

d on

30

Janu

ary

2012

. Dow

nloa

ded

by U

nive

rsity

of

Win

dsor

on

29/1

0/20

14 1

2:19

:38.

View Article Online

for Physical–Chemical Functions.’’ The support from the Univer-

sity of Bremen under the Initiative ‘‘Func-Band’’ is also gratefully

acknowledged.

Notes and references

1 P. Gomez-Romero, Adv. Mater., 2001, 13, 163.2 T. Lan and T. J. Pinnavaia, Chem. Mater., 1994, 6, 2216.3 W. U. Huynh, X. G. Peng and A. P. Alivisatos, Adv. Mater., 1999, 11,923.

4 G. Kickelbick, Prog. Polym. Sci., 2003, 28, 83.5 A. Hartwig, A. Luehring and J. Trautmann, Macromol. Mater. Eng.,2009, 294, 363.

6 A. Hartwig, M. Sebald, D. Putz and L. Aberle, Macromol. Symp.,2005, 221, 127.

7 D. J. Boday, K. A. DeFriend, K. V. Wilson, D. Coder and D. A. Loy,Chem. Mater., 2008, 20, 2845.

8 A. C. Pierre and G. M. Pajonk, Chem. Rev., 2002, 102, 4243.9 S. Edmondson, V. L. Osborne and W. T. S. Huck, Chem. Soc. Rev.,2004, 33, 14.

10 W. Senaratne, L. Andruzzi and C. K. Ober, Biomacromolecules, 2005,6, 2427.

11 C. Y. Hong, Y. Z. You and C. Y. Pan, Chem. Mater., 2005, 17, 2247.12 H. Zou, S. S. Wu and J. Shen, Chem. Rev., 2008, 108, 3893.

2332 | Nanoscale, 2012, 4, 2326–2332

13 H. W. Ma, J. H. Hyun, P. Stiller and A. Chilkoti, Adv. Mater., 2004,16, 338.

14 A. Bhattacharya and B. N. Misra, Prog. Polym. Sci., 2004, 29,767.

15 L. M€adler, A. Roessler, S. E. Pratsinis, T. Sahm, A. Gurlo, N. Barsanand U. Weimar, Sens. Actuators, B, 2006, 114, 283.

16 N. Barsan and U.Weimar, J. Phys.: Condens. Matter, 2003, 15, R813.17 K. Großmann, K. E. Kovacs, D. K. Pham, L. M€adler, N. Barsan and

U. Weimar, Sens. Actuators, B, 2011, 158, 388.18 L. M€adler, A. A. Lall and S. K. Friedlander, Nanotechnology, 2006,

17, 4783.19 H. Sehaqui, M. Salajkova, Q. Zhou and L. A. Berglund, Soft Matter,

2010, 6, 1824.20 J. C. Tan, T. D. Bennett and A. K. Cheetham, Proc. Natl. Acad. Sci.

U. S. A., 2010, 107, 9938.21 A. Tricoteaux, E. Rguiti, D. Chicot, L. Boilet, M. Descamps,

A. Leriche and J. Lesage, J. Eur. Ceram. Soc., 2011, 31,1361.

22 N. Riefler and L. M€adler, J. Nanopart. Res., 2010, 12, 853.23 N. Barsan, D. Koziej and U. Weimar, Sens. Actuators, B, 2007, 121,

18.24 V. Geraldo, V. Briois, L. V. A. Scalvi and C. V. Santilli, J. Eur.

Ceram. Soc., 2007, 27, 4265.25 W. Y. Teoh, R. Amal and L. Madler, Nanoscale, 2010, 2,

1324.

This journal is ª The Royal Society of Chemistry 2012