See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/6933919

Anatase TiO 2 Nanocomposites for Antimicrobial

Coatings

Article in The Journal of Physical Chemistry B June 2005

DOI: 10.1021/jp0502196 Source: PubMed

CITATIONS READS

446 1,114

3 authors, including:

Patricia S Vary Chhiu-Tsu Lin

Northern Illinois University Northern Illinois University
33 PUBLICATIONS 1,253 CITATIONS 124 PUBLICATIONS 1,685 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Chhiu-Tsu Lin on 12 September 2017.

The user has requested enhancement of the downloaded file.

 J. Phys. Chem. B 2005, 109, 8889-8898 8889

Anatase TiO2 Nanocomposites for Antimicrobial Coatings

Guifen Fu, Patricia S. Vary, and Chhiu-Tsu Lin*

Department of Chemistry and Biochemistry, Northern Illinois UniVersity, DeKalb, Illinois 60115-2862
ReceiVed: January 12, 2005; In Final Form: March 2, 2005

A sol-gel chemistry approach was used to fabricate nanoparticles of TiO2 in its anatase form. The particle
size is shown to be sensitive to the use of HClO4 or HNO3 as acid catalyst. The gold-capped TiO2
nanocomposites were processed by the reduction of gold on the surface of the TiO2 nanoparticles via a chemical
reduction or a photoreduction method. Different percentages of vanadium-doped TiO2 nanoparticles, which
extended the TiO2 absorption wavelength from the ultraviolet to the visible region, were successfully prepared.
The synthesized nanocomposites have a size of about 12-18 nm and an anatase phase as characterized by
XRD, TEM, AFM, and UV-vis spectroscopy. The TiO2 nanocomposite coatings have been applied on glass
slide substrates. The antibacterial activity of TiO2 nanocomposites was investigated qualitatively and
quantitatively. Two types of bacteria, Escherichia coli (DH 5R) and Bacillus megaterium (QM B1551), were
used during the experiments. Good inhibition results were observed and demonstrated visually. The quantitative
examination of bacterial activity for E. coli was estimated by the survival ratio as calculated from the number
of viable cells, which form colonies on the nutrient agar plates. The antimicrobial efficiency and inhibition
mechanisms are illustrated and discussed.

1. Introduction catalytic efficiency, including increase of the surface area of

TiO2 through tailoring particle size and pore-size distributions,
The TiO2 photocatalysts have been investigated extensively
generation of defect structures to induce space-charge separation
for the killing or growth inhibition of bacteria.1-3 Most works
via metal dopants, and surface modification of the TiO2 with a
have been done by using a fine TiO2 powder and a strong UV
metal or another semiconductor.12-14 Because the majority of
light. Only a little work has been reported in the use of TiO2
photoexcited charge carriers [electrons (e-) and holes (h+)] may
nanoparticle aqueous solutions coated on substrates.4,5 Recently,
undergo a rapid recombination, single component semiconductor
nanosize (<100 nm) TiO2 particles have attracted a lot of
nanoparticles have shown to exhibit relatively poor photocata-
activity from many researchers.6 These nanometer-size TiO2
lytic efficiency. Semiconductor-metal nanocomposites, on the
particles exhibit many special properties due to the fact that
other hand, exhibit increased efficiency of photocatalytic activity
the band gap of the nanoparticles increases with the decrease
because of a reduction in the e--h+ recombination rate due to
of their size, and the small TiO2 particles offer a very large
better charge separation between the electrons, which ac-
surface area.7-9
cumulate on the metal, and holes, which remain on the
When TiO2 catalyst is irradiated with light of energy greater photocatalyst surface. Many investigations have reported that
than or equal to its band gap energy, electron-hole pairs are the addition of noble metals such as gold and silver may enhance
generated that can induce redox reactions at the surface of the the overall photoefficiency of TiO2, and a variety of noble metal
TiO2.10 The general scheme for the photocatalytic damage of capped TiO2 nanocomposite solutions have been synthesized
microorganism cells by TiO2 photocatalytic properties involves in recent years to improve the efficiency of the photocatalytic
several steps:11 (1) the photoexcited TiO2 catalyst produces activity of TiO2.15
electron-hole pairs that migrate to the TiO2 surface; (2) In this paper, an anatase phase TiO2 nanoparticle suspension,
photogenerated holes in TiO2 can react with adsorbed H2O or a gold-capped TiO2 nanocomposite aqueous solution, and
OH- at the catalyst/water interface to produce the highly reactive vanadium-doped TiO2 nanoparticles in aqueous solution are
hydroxyl radicals and the electrons can react with oxygen processed by a sol-gel technique and dispersed to form coatings
vacancies to form superoxide ions; (3) finally, the various highly that can inhibit microbial growth on various substrates. The TiO2
active oxygen species generated can oxidize organic compounds/ nanoparticles and nanocomposites were characterized by X-ray
cells adsorbed on the TiO2 surface, resulting in the death of the diffractometry (XRD), transmission electron microscopy (TEM),
microorganisms. UV-vis absorption spectroscopy, and atomic force microscopy
The efficiency of TiO2 photocatalytic properties and anti- (AFM). The antibacterial effect of those nanoparticle suspen-
bacterial activities should depend on our abilities (i) to make sions and coatings were investigated, both qualitatively and
stable nanostructured TiO2 particles or composites, (ii) to quantitatively, using Escherichia coli (DH 5R), a Gram-negative
generate electron-hole pairs by extending the excitation bacterium, and Bacillus megaterium (QM B1551), a Gram-
wavelength to the visible light region, and (iii) to achieve a positive bacterium.
reduced recombination rate on the newly created electron-hole
carriers. Several methods may be used to improve the photo- 2. Experimental Section
* To whom correspondence should be addressed. Telephone: (815) 753- 2.1. Materials. Titanium(IV) isopropoxide (98%) was pur-
6861. Fax: (815) 753-4802. E-mail: [email protected]. chased from Strem Chemical Company and used as received.
10.1021/jp0502196 CCC: $30.25 2005 American Chemical Society
Published on Web 03/31/2005
8890 J. Phys. Chem. B, Vol. 109, No. 18, 2005 Fu et al.

Sodium borohydride (powder, 98%), hydrogen tetrachloroaurate- megaterium was grown in Supplemental Nutrient Broth (SNB)
(III) trihydrate, and ammonium metavanadate (99%) were at 30 C for about 14-16 h. The density of the bacterial cells
obtained from Aldrich Chemical. 1-Propanol (anhydrous, 99.7%) in liquid cultures was estimated by optical density (OD) at 600
was obtained from Sigma-Aldrich. Ethyl alcohol, 200 proof, nm wavelength. The OD was chosen in a range of 0.8-1.0,
was purchased from Pharomco. The water employed in all which is the optimal optical density of the cells for conventional
preparations was purified by a Milli-Q system (Millipore). Two bacterial activity testing. The cell suspensions used for anti-
bacteria strains, Bacillus megaterium (QM B1551) and Escheri- bacterial activity on both solution phase and coated glass
chia coli (DH 5R), were from laboratory stock of the Biological substrate were 108-109 colony-forming units (cfu)/mL.
Sciences Department of Northern Illinois University. Other 2.5. Bacterial Activity Test on TiO2 Nanocomposites in
materials for bacteria cultivation, such as agar, yeast extract, Solution Phase. The following procedures were used for testing
sodium chloride, tryptone, and plastic and Pyrex Petri dishes antibacterial activity of E. coli and B. megaterium using TiO2
were used from conventional laboratory stock. nanoparticle solution, gold-capped TiO2 nanocomposite solution,
2.2. Preparation of TiO2 Nanoparticle/Nanocomposite and vanadium-doped TiO2 nanocolloid. The testing Petri plates
Aqueous Colloidal Solution and Coating. Different concentra- were prepared by LB or SNB medium. Using a homemade glass
tions of TiO2 nanoparticle aqueous solution were prepared by atomizer, the bacteria were sprayed evenly on the top of the
the procedure described previously.16 As an example, for 5 mM medium covering the whole area of the testing Petri plate. After
TiO2 nanoparticle suspensions, 0.5 mL of titanium isopropoxide about 5-10 min to allow the bacteria to dry out, the testing
was added to 4.5 mL of 1-propanol to achieve a 10% titanium solutions of TiO2 nanocomposites were dropped on top of the
isopropoxide solution. In a separate beaker, 200 mL of deionized bacteria. For each testing solution, three marked spots on each
water was adjusted to a pH 1.5 with 6 M HClO4 or HNO3. Petri dish were applied. In this way, the large part of the bacteria
Then, 2.97 mL of 10% titanium isopropoxide was added was on the culture medium, and three marked areas of the
dropwise to the pH-adjusted water with vigorous stirring using bacteria were on the testing solution. Samples of the testing
a magnetic stir bar for about 1-1.5 h, resulting in a transparent Petri plates were kept in the incubator at 37 C for E. coli and
5 mM TiO2 colloidal solution. 30 C for B. megaterium for 20-26 h, and then testing results
The TiO2/Au nanocomposites were prepared by adding the were determined visually and with the aid of an optical
desired amount of 5 mM stock HAuCl4 solution to the 5 mM microscope.
colloidal TiO2 suspension prepared above while stirring vigor- 2.6. Bacterial Activity Test for TiO2 Nanocomposites on
ously. Reduction of [AuCl4]- was achieved by dropwise addition Coated Glass Substrates. The tests of inhibition of bacterial
of sodium borohydride (9 mM, in ethanol) until a wine-red color growth for TiO2 nanocomposites on coated glass substrates were
was observed. Two different TiO2/Au solutions were prepared carried out both qualitatively and quantitatively. In the qualita-
containing a [TiO2]:[Au] ratio of 20:1 (a wine-red color) and tive testing, the testing glass plate was spray-coated and
1:1 (a dark purple color) by keeping the gold concentration thermally cured at 90 C for 30 min on three marked areas using
constant at 0.2 mM while varying the TiO2 concentration from a fixed concentration of TiO2 nanoparticle solution, gold-capped
4 to 0.2 mM. The TiO2/Au nanocomposites were also synthe- TiO2 nanocomposite solution, or vanadium-doped TiO2 nano-
sized by a photoreduction method using a 450 W Xe lamp. colloid. The coated glass slides were sterilized by autoclaving.
Various concentrations (3-13 wt %) of vanadium-doped TiO2 The bacteria were sprayed evenly across the entire testing
nanoparticle solutions were prepared. As an example, for a 3 substrate using a small homemade atomizer and allowed to dry
wt % vanadium-doped TiO2 nanoparticle solution, 0.02 g of for 10 min. The nutrition agar, LB for E. coli and SNB for B.
NH4VO3 was dissolved in 20 mL of deionized water while megaterium, was pipetted onto the top of the bacteria. There
stirring and heating slightly for about 10 min. The prepared are two important steps that need to be done very carefully.
NH4VO3 solution was added dropwise to 20 mL of TiO2 One is to spray the bacteria as evenly as possible; the other is
nanoparticle solution and stirred for about 1 h. to carefully control the temperature of the LB agar when
2.3. Characterization of Nanoparticles/Nanocomposites. pipetting it onto the bacteria. The melting point of LB agar is
UV-vis absorption spectra of colloidal solutions were recorded 100 C and it will become a gel at about 45 C; higher than 55
by a Perkin-Elmer double beam scanning spectrometer, Model C, the bacteria will be killed by the high temperature. Therefore,
Lambda 19. The TiO2 crystal structure of the dry solution the optimal temperature to pipet the LB agar is at about 55 C.
(powder) was investigated by Rigaku Miniflex X-ray powder Samples of the coated glass plates are then transferred to a clean
diffraction (XRD). The white powder was obtained from 5 mM Petri dish and kept in the incubator at 37 C for E. coli and 30
TiO2 solution after slow evaporation of solvents, and was ground C for B. megaterium for 20-26 h. The qualitative testing results
and compressed in a sample holder. The XRD pattern was of bacterial activity were determined visually and with the aid
obtained with Cu KR radiation at 60 keV and 15 mA over the of an optical microscope.
range 20 < 2 < 60. The morphology and size of the For the quantitative testing, a set of quantitative dilutions of
nanoparticles/nanocomposites were studied with a JEOL 2010 bacteria suspensions were used and the number of bacterial
transmission electron microscope (TEM) operating at 200 keV colonies appearing on the coated plates are actually counted
and 110 A equipped with a Gatan CCD camera, and a Quesant under a microscope. Equal aliquots of 1 mL of E. coli suspension
Q-scope 2350 atomic force microscope (AFM). Different area were pipetted into the TiO2/Au nanocomposite coating cham-
sizes of the sample surfaces were scanned with 600 600 pixel bers, and then 50 L of the reaction mixture was immediately
scan resolution to make a complete contour of the sample. taken from the chamber and diluted into 4.95 mL of LB broth
2.4. Bacterial Cultures and Test of Antibacterial Activities. in a centrifuge tube to achieve the 10-2 colony-forming units
Two types of bacteria, B. megaterium (QM B1551), a Gram- (cfu)/mL. The series of dilutions from 10-2 cfu/mL of reaction
positive bacterium, and Escherichia coli (DH 5R), a Gram- mixture solution was conducted by taking out 50 L of the 10-2
negative bacterium, were used as model bacteria in this study. cfu/mL solution and diluting to 10-4 cfu/mL with 4.95 mL of
Luria broth (LB) and agar were used as sources for culturing LB broth in a centrifuge tube, and then mixing completely.
E. coli at 37 C on a rotary platform in an incubator. The B. Similarly, a total of four different concentrations, 10-4, 10-5,
TiO2 Nanocomposites for Antimicrobial Coatings J. Phys. Chem. B, Vol. 109, No. 18, 2005 8891

10-6, and 10-7 cfu/mL, of the original concentration of the

reaction mixture, were pipetted onto the LB agar plates for each
sample, and three replicate plates were used for each dilution.
The sample plates were kept in an incubator at 37 C for about
24 h; the number of viable cells in the sample was determined
by choosing the appropriate dilution of the sample onto LB agar
plates and counting colonies that appeared on the plates under
a microscope. An average number of viable cells was obtained
by averaging the numbers in the three replicate plates.

3. Results and Discussion

3.1. Anatase TiO2 Nanocolloid Suspension. Several methods
for synthesizing TiO2 nanoparticles have been published.17-22
Two major structural forms of titanium dioxide can exhibit
photocatalytic activity: anatase and rutile.11 Anatase (chains of
TiO6 octahedral sharing two edges) has a wider optical band Figure 1. XRD pattern of TiO2 nanoparticles prepared by the
gap of 3.2 eV, and the band gap of rutile (chains of TiO6 hydrolysis of titanium isopropoxide, and catalyzed by either HNO3 or
HClO4. A library spectrum of anatase TiO2 is marked for comparison.
octahedral sharing four edges) is 3.0 eV.19 It is important to
develop methods for the selective preparation of TiO2, because solution, while lowering the temperature with an ice bath (at
anatase and rutile are often formed at the same time during the 4 C) to obtain a transparent solution. In trying to make a
formation of TiO2 and it is desired to synthesize anatase solution concentration higher than 260 mM TiO2 aqueous
nanoparticle TiO2 since the anatase form has a higher photo- suspension, white suspensions were immediately formed upon
catalytic activity than rutile TiO2.23 addition of titanium isopropoxide solution to the water. X-ray
Different preparation methods and conditions yield different diffraction (XRD) was used to investigate the crystalline phases
particle sizes and phase compositions of the nanostructured TiO2 (anatase or rutile) of the TiO2 powder. Two different acids used
that display different morphological, structural, and surface to control the pH in preparing 5 mM TiO2 nanoparticle
physiochemical properties, and also give different photocatalytic suspension were then investigated. Figure 1 shows the XRD
activities.24,25 These conditions include the types of precursors, pattern of TiO2 powder made by hydrolysis of titanium
the concentration and molar ratio of reactants, the polarity of isopropoxide with different acids, either HNO3 or HClO4. The
the solvent, the pH, and the temperature of the reacting solution. XRD shows that both acids yield the anatase phase of TiO2.
In this study, we focused on the preparation of highly The type of acid used in this synthesis gives an identical anatase
photoactive nanosize anatase structured TiO2. The TiO2 nano- phase of titanium dioxide with this preparation method. As
particle aqueous suspension was prepared by employing titanium illustrated, two syntheses gave the same structures in the XRD
isopropoxide as a precursor, water as a solvent, and HNO3 or pattern: broad peaks are observed at values of 2 equal to 25.5,
HClO4 as a catalyst at room temperature for a period of at least 38.5, 48.2, 54.4, and 55.4 that correspond to the anatase
2 h. This allowed for a hydrolysis reaction to take place and structure.18 A small broad signal at 31 is from TiO2 brookite.
ensured the formation of a transparent TiO2 solution. The pH Also, strong peaks are observed at 38.3 and 45 that are
of these solutions is about 1.5. At this pH, nanoparticles are attributed to the diffraction of aluminum, which is the sample
positively charged, which favors the formation of small particles. holder. A good match of peaks from the TiO2 sample spectra
Two different acids, HNO3 and HClO4, were used for controlling with a library spectrum of anatase TiO2 is shown in the bottom
the solution pH. The dropping speed of titanium isopropoxide of Figure 1. It further confirms that TiO2 nanoparticles prepared
solution into water had an important effect on the particle size by this method, with either HNO3 or HClO4 acid, both generate
of TiO2. At room temperature and pH 1.5 and when titanium an anatase crystalline structure without a significant amount of
isopropoxide solution was added dropwise slowly over a period any other phase. From the diffraction patterns, the anatase TiO2
of 1.0-1.5 h, a transparent colloidal solution was obtained. materials prepared are in the form of very small particles, since
When titanium isopropoxide solution was added dropwise the XRD peaks are very broad.
rapidly within 30-40 min, a less transparent or an opaque The morphology and size of the colloidal nanoparticles were
solution was observed. This result indicated that the key examined by TEM. The TEM image of 5 mM TiO2 nano-
parameters for producing small TiO2 particles in aqueous particles is presented in Figure 2. A TiO2 particle size of about
suspension are the pH of the solution and the rate of addition. 12-18 nm was obtained, and a spherical particle shape of
The UV-vis absorption spectrum of a 5 mM TiO2 nano- anatase TiO2 was also observed. To further characterize the TiO2
particle solution is shown to have an absorption edge at 360 nanoparticles, the AFM was used to examine the morphology
nm (i.e., 3.45 eV), indicating that the TiO2 colloidal solution of the coating. A 5 mM TiO2 aqueous solution, prepared with
obtained is anatase and not rutile. The band gap of the prepared either HNO3 or HClO4 acid, was coated on window glasses and
TiO2 nanoparticles is at a higher energy than that determined examined. The left and right portions of Figure 3 show the AFM
for bulk anatase (3.2 eV). Several different concentrations of angle view image and top view image of 5 mM TiO2
TiO2 aqueous suspensions were prepared in an attempt to obtain nanoparticles coated on glass, respectively, when HClO4 acid
a TiO2 nanoparticle solution in a higher concentration. The was used in the synthesis of TiO2 suspension. From these
highest concentration that could be reached in an aqueous images, small particles were observed and their dimension varied
suspension was 260 mM. A key factor for preparing an aqueous in a narrow range (40-100 nm); the shape of the particles was
solution higher than 5 mM TiO2 is controlling the rate of almost spherical. A similar result was also obtained from 5 mM
dropwise addition of the titanium isopropoxide solution into the TiO2 nanoparticles coated on glass prepared with HNO3 acid.
water by slowly adding titanium isopropoxide into water, drop The AFM angle view image (left) and top view image (right)
by drop, much more slowly than when making the 5 mM TiO2 are shown in Figure 4. The AFM images again prove that the
8892 J. Phys. Chem. B, Vol. 109, No. 18, 2005 Fu et al.

3.2. Formation of Gold-Capped TiO2 Nanocomposites.

Gold can be deposited on the TiO2 nanoparticle surface by either
cation adsorption or anion adsorption to the TiO2 surface
depending on the pH of the working solution. The TiO2 is an
amphoteric oxide with an isoelectric point26 IEPTiO2 ) 6.
Therefore, when the solution pH is higher than IEPTiO2, the main
surface species is -O-, so the TiO2 surface is negatively charged
and gold can be deposited on TiO2 by cation adsorption. When
the solution pH is lower than IEPTiO2, the main surface species
is -OH2+, so the TiO2 surface is positively charged. Then gold
can be produced on the TiO2 surface by anion adsorption. For
example, gold-capped TiO2 can be prepared by adding the
desired amount of HAuCl4 solution to the colloidal TiO2 solution
made in an acidic medium while stirring vigorously. The TiO2
colloids prepared in acidic media are positively charged. They
act as a support to adsorb [AuCl4]-. The negatively charged
[AuCl4]- adsorbs strongly on the TiO2 nanoparticle surface.
Upon reduction with NaBH4, stable TiO2/Au nanocomposites
Figure 2. TEM image of 5 mM TiO2 nanoparticle suspension. are obtained in aqueous solution. Gold actually covers only a
same structure of TiO2 nanoparticles was obtained by using both small area of the TiO2 surface. Therefore, a large surface area
acids. However, the images show that the particle size is a little of TiO2 nanoparticle is still exposed. Gold reduction carried
smaller when prepared with HNO3 (<50 nm) than that with out in the absence of a TiO2 core using the same experimental
HClO4 (<100 nm) acid. procedure and conditions did not produce stable colloids.27
The TiO2 particle size is smaller according to the TEM results Two methods are used to prepare TiO2/Au nanocomposite
than with the AFM image, which may reflect the drying-induced solutions: chemical reduction and photoreduction. The gold-
particle growth or aggregating. For the AFM imaging, the coated capped TiO2 nanoparticles were prepared by first synthesizing
glasses were dried at room temperature for 1 week, and then at TiO2 nanoparticle suspensions in aqueous solution (stabilized
90 C in an oven for about 1/2 h. For TEM measurement, the at pH 1.5) and then capping these particles with gold nano-
freshly prepared solutions were used; the sample preparation crystallites. In acidic medium, when the noble metal salt
for TEM testing requires an overnight drying process. (HAuCl4) solution was added to the TiO2 aqueous solution, the

Figure 3. (left) AFM angle view image and (right) AFM top view image, of 5 mM TiO2 nanoparticles catalyzed by HClO4 and coated on glass.

Figure 4. (left) AFM angle view image and (right) AFM top view image, of 5 mM TiO2 nanoparticles catalyzed by HNO3 and coated on glass.
TiO2 Nanocomposites for Antimicrobial Coatings J. Phys. Chem. B, Vol. 109, No. 18, 2005 8893

Figure 5. UV-vis absorption spectra of Au-capped TiO2 nanocomposites (a [TiO2]:[Au] ratio of 20:1): (left) reduced by NaBH4 and (right)
reduced by UV irradiation of different exposure times.

Figure 6. TEM images: (left) 4 mM gold-capped TiO2 nanocomposites; (right) TiO2/Au nanocomposites at different [TiO2]:[Au] ratios of (A)
20:1 and (B) 1:1.
TiO2 core acted as a support to adsorb [AuCl4]- ions first. Then, nanocomposites of about 5-10 nm diameter were obtained
upon reduction with a freshly prepared solution of NaBH4, stable (image A). In contrast, the TEM image for the [TiO2]:[Au] ratio
TiO2/Au nanocomposites were obtained and the aqueous solu- of 1:1 nanocomposites showed a larger size of about 10-20
tion was wine-red in color. A broad spectral band at 537 nm, nm (image B). They also tend to close in on each other,
which is attributed to the surface plasmon band of gold indicating the aggregation effect. The fact that the optimal
nanoparticles,27 appears on the UV-vis absorption spectrum amount of Au was obtained at relatively low loading or at
of Au-capped TiO2 suspension as shown in the left portion of relatively low surface coverage of TiO2 may reflect some
Figure 5. recombinative activity of the metallic islands.11 Without Au-
The transmission electron micrograph (TEM) of gold-capped capped particles, TiO2 solutions showed significant turbidity
TiO2 nanocomposite is shown in the left portion of Figure 6. A after several hours at room temperature. In contrast, the TiO2/
well-defined image of spherical shape of TiO2/Au nano- Au solution remained clear, indicating that gold capped the TiO2
composites was obtained. The particle diameter of gold-capped particles and protected them against precipitation. Therefore,
TiO2 nanocomposites was in the range of 10-15 nm, almost the 4 mM TiO2/Au solution with a [TiO2]:[Au] ratio of 20:1
the same size as TiO2 nanoparticles alone. Compared to the was chosen as the one of the testing solutions for antibacterial
TEM graph of TiO2 nanoparticles, shown in Figure 2, there is study in this study.
no significant change in particle size, indicating that the TiO2 The reduction of [AuCl4]- ions absorbed on the surface of
surface was capped with a very thin shell or islands of gold, so TiO2 particles can also be done photochemically. The right
the size was determined by the TiO2 nanoparticles. portion of Figure 5 shows the absorption spectra of Au-capped
Two different ratios of [TiO2]:[Au] nanocomposite solutions TiO2 nanocomposites prepared by using the photoreduction
were prepared and characterized. Transmission electron micro- method at different UV exposure times. No coloration change
graphs of two different [TiO2]:[Au] ratios are shown in the right took place at the zero exposure time, and no Au surface plasmon
portion of Figure 6. For the high core:shell ratio (i.e., a band was observed in the absorption spectra. When the TiO2
[TiO2]:[Au] ratio of 20:1), fairly uniform size gold-capped TiO2 with HAuCl4 mixture solution was exposed to UV light, after
8894 J. Phys. Chem. B, Vol. 109, No. 18, 2005 Fu et al.

2 min exposure time, the gold was reduced on the surface of

TiO2, which was confirmed by UV-vis absorption spectra. The
max values were observed in the range 537-543 nm, indicating
the formation of Au(0) nanoparticles. Under UV irradiation,
TiO2 nanoparticles undergo charge separation,27 and the gener-
ated electron-hole carriers can act to reduce [AuCl4]- ions that
produce TiO2/Au nanocomposites via the following mecha-
nisms:

TiO2 + h f TiO2 (h+ + e-) (1)

TiO2 (3e-) + Au3+ f TiO2 + Au0 (2)

Chlorine atoms produced during the reactions presumably Figure 7. Absorption spectra of different percentages of V-doped TiO2
reacted with water to form hydrochloric acid and hypochlorous nanoparticle suspension.
acid.
3.3. Generation of Vanadium-Doped TiO2 Nanocolloid. thick wall composed of many layers of peptidoglycan polymer,
To initiate photocatalytic reaction, a light source with energy and only one membrane (plasma membrane). The Gram-
greater than the TiO2 band gap energy is required. The band negative bacteria have only a thin layer of peptidoglycan and a
gap energy of TiO2 is high in the UV range of less than 385 more complex cell wall with two cell membranes, an outer
nm. This limitation has been overcome by doping various membrane, and a plasma membrane. The addition of the outer
transition metal ions into TiO2, to extend the excitation membrane of the Gram-negative bacteria cells influences the
wavelength range of TiO2 from the UV into the visible region permeability of many molecules. Under certain conditions, the
where convenient light sources are available.28,29 The primary Gram-negative bacteria are more resistant to many chemical
process is that the transition metal ions absorb light in the visible agents than Gram-positive cells.31
region and transfer the excited electron to the TiO2 conduction The photocatalytic process of anatase TiO2 particles includes
band. For example, in one study,28 titanium tetraisopropoxide chemical steps that produce highly reactive species such as
was used as a titanium dioxide source; NH4F was used as a hydroxyl radical, hydrogen peroxide, and superoxides that, in
precursor of the dopant. The F--doped TiO2 with anatase and principle, can cause fatal damage to microorganisms.4 Among
brookite phases was synthesized by hydrolysis of titanium these reactive oxygen species, the hydroxyl radicals are highly
tetraisopropoxide in an NH4F aqueous solution. The result reactive and therefore short-lived. The superoxide ions are
showed that the F--doped TiO2 samples exhibited strong relatively longer lived. Due to their negative charge they cannot
absorption in the UV-visible range and a red shift in the band penetrate the cell membrane. They must contact directly the
gap transition. Another research group, Klosek et al., studied outer surface of bacteria unless the TiO2 particle has penetrated
TiO2 particles doped with V4+ and found that V4+-doped TiO2 the cell. Hydrogen peroxide is less harmful compared to
has the effect of extending the wavelength range of the TiO2 hydroxyl radicals and superoxide ions, but it can enter the cell.32
into the visible region (396-450 nm).30 The advantages of However, among these reactive oxygen species, which species
transition metal doping include extending the wavelength to the are directly involved in the damage of bacterial cells or which
visible region and also improving the photocatalytic activity of species contribute more to the oxidative reactions with organic
the TiO2. It is noted29 that some research results showed certain compounds is still subject to investigation.
metals such as Fe3+ and V4+ in Fe3+- and V4+-doped TiO2 Several proposed mechanisms for cell killing by the TiO2
actually reduced the electron-hole pair recombination rate and photocatalytic processes were reported.4,33 The Sunada research
improved the photocatalytic activity over the undoped TiO2. group reported32 direct evidence of cell membrane damage by
Some metal ions such as Co3+ and Al3+ in Co3+- and Al3+- the irradiation of a thin transparent TiO2 film to examine the
doped TiO2 either showed no improvement or even decreased photocatalytic degradation of endotoxin from E. coli. The
the photocatalytic efficiency of TiO2.29 endotoxin is a component of the outer membrane of Gram-
A series of vanadium-doped TiO2 nanocolloids have been negative bacteria and is released only when the cellular structure
synthesized and investigated. The absorption spectra of TiO2 is destroyed. The results indicated that the TiO2 photocatalyst
particles doped with different percentages of vanadium are destroys the outer membrane of the E. coli cell and causes the
shown in Figure 7. Without vanadium doping, the UV-vis death of the bacteria.4 Recently, the mechanism of cell killing
absorption spectrum of a 5 mM anatase TiO2 nanoparticle of E. coli on TiO2 thin films has been investigated by atomic
solution is shown to have an absorption edge at 360 nm (i.e., force microscopy (AFM).33 The damage process of the cell wall
3.45 eV). The absorption edge of vanadium-doped TiO2 and the cell membrane was observed by AFM imaging. Also,
nanocolloids is progressively red shifted with vanadium con- the permeability of the cell membrane was examined by K+
centration. A red shift to the visible region at 450 and 500 nm leakage from E. coli. Results showed that the intracellular K+
is shown in Figure 7 for a 3% and 7% V-doped TiO2 leaked out from E. coli very quickly after TiO2 thin film was
nanocolloid, respectively. There was no further red shift at irradiated by UV. The researchers believe the cell death was
concentrations higher than 7% V-doped TiO2 solution. caused by the decomposition of the cell wall first, and then
3.4. Bactericidal Activity Effect on TiO2 Nanocolloid subsequent decomposition of cell membrane.33 Damage of the
Solutions. The bactericidal activities of the anatase TiO2 cell membrane directly leads to leakage of minerals, proteins,
nanocomposites were evaluated by the inhibition of bacterial and genetic materials, causing cell death.
growth of B. megaterium and E. coli. B. megaterium is one of In this paper, the antibacterial effect of nanocomposites on
the Gram-positive bacteria and E. coli is one of the Gram- E. coli and B. megaterium was first tested for TiO2 suspensions
negative bacteria. The Gram-positive bacteria have a relatively and gold-capped TiO2 solutions. A control solution was also
TiO2 Nanocomposites for Antimicrobial Coatings J. Phys. Chem. B, Vol. 109, No. 18, 2005 8895

Figure 8. Antibacterial effect on B. megaterium. Left: (A) control

solution without TiO2 nanoparticles; (B) 5 mM TiO2 nanoparticle Figure 9. Antimicrobial activity testing with different concentrations
suspension. Right: 4 mM TiO2 nanoparticle suspension, (A) dark of gold-capped TiO2 nanocomposite solutions. Left: on B. megaterium
control covered with an aluminum foil and (B) under room light. at (A) 4, (B) 1, (C) 0.5, and (D) 0.1 mM. Right: on E. coli at (A) 4,
(B) 1, and (C) 0.5 mM.
TABLE 1: Antibacterial Effects with Different
Concentrations of TiO2 Suspension on B. megaterium and E. TABLE 2: Antibacterial Effects with Different
coli Concentrations of TiO2/Au Suspension on B. megaterium and
E. coli
5 mM 1 mM 0.5 mM 0.1 mM 0.05 mM 0.01 mM
5 mM 1 mM 0.5 mM 0.1 mM 0.05 mM 0.01 mM
E. coli ++ - - - - -
B. megaterium + + ++ + - - - E. coli ++ + - - - -
B. megaterium + + ++ + - - -
prepared that contains all other chemical compositions as a 5
mM TiO2 suspension, but no TiO2 nanoparticles. The left portion membranes. It is noted that the biological organisms are killed
of Figure 8 shows the antibacterial effect on B. megaterium by various highly reactive oxygen species (hydroxyl radical,
using a control solution without TiO2 nanoparticles (top) and a hydrogen peroxide, or superoxides) generated upon the photo-
5 mM TiO2 solution (bottom). The result shows that there was catalytic process of TiO2 nanoparticle suspension. The antibac-
no inhibition on bacterial growth when a control solution terial effects of TiO2 suspension on both B. megaterium and E.
contains no TiO2 nanoparticles. On the bottom plate where three coli should also be related to the particle size (10-15 nm),
spots of 5 mM TiO2 solution were applied, bacteria grew on surface area, band gap energy (3.45 eV), and coordination of
the entire plate except those three TiO2 spots, which display Ti4+ ions (tetrahedral) of anatase nanoparticles prepared.
bacteria-free areas compared to the surrounding areas. The Similar antimicrobial activity testing with different concentra-
antibacterial effect of nanocomposites on E. coli and B. tions of gold-capped TiO2 solutions and V-doped TiO2 solutions
megaterium was then tested for TiO2 suspensions with and was also performed. The test results with TiO2/Au nano-
without room light (fluorescent discharge tubes). The dark composite solutions on B. megaterium and E. coli are shown
sample was carried out by covering the testing plate with in the left portion and right portion of Figure 9, respectively.
aluminum foil. The right portion of Figure 8 shows the The good antibacterial growth of B. megaterium is observed
antibacterial activity of 4 mM TiO2 solution on B. megaterium, for 4 mM (left A) and 1 mM (left B) TiO2/Au nanocomposite
where the top plate was covered with aluminum foil and the solutions. Some inhibition effect is still observed at a low
bottom plate was not. The results indicate that the antibacterial concentration of 0.5 mM (left C), and no noticeable inhibition
effects observed in Figure 8 are due to the photocatalytic activity effects are seen in the left D plate of Figure 9. For the
of the TiO2 nanoparticles in suspension.4,32 antibacterial activity of E. coli, the TiO2/Au nanocomposite
Under room light condition, a series of dilutions with 5 mM solutions show good, medium, and no inhibition effects at 5
TiO2 solution was used to examine the concentration effect of mM (right A), 1 mM (right B), and 0.5 mM (right C),
TiO2 suspensions on the antibacterial activity for both B. respectively. The testing results for a wide range of concentra-
megaterium and E. coli. Six different concentrations of TiO2 tions of TiO2/Au nanocomposite solutions both bacteria are
suspension were tested, and the results are summarized in Table summarized in Table 2. The antibacterial effects on B. mega-
1, where + + indicates the best inhibition results, + means terium and E. coli by TiO2/Au nanocomposite solutions (Table
relatively less inhibition effect, and - represents no inhibition 2) are shown to be similar to those by TiO2 nanoparticle
at all. At a 5 mM concentration, TiO2 suspension inhibits solutions (Table 1). However, a slightly better inhibition of
effectively the bacteria growth for both B. megaterium and E. bacterial growth is noticed for TiO2/Au nanocomposite solutions
coli. The inhibition of B. megaterium growth activity by TiO2 than for TiO2 nanoparticle solutions. A clear inhibition effect
suspension is effective in the following order: 5 mM 1 mM on E. coli is observed for the 1 mM TiO2/Au solution, but not
> 0.5 mM. No significant antibacterial activity was observed for the 1 mM TiO2 solution. Moreover, the bacterial inhibition
for B. megaterium and E. coli at of TiO2 suspension concentra- spots on TiO2/Au nanocomposite testing plates in Figure 9 are
tions of 0.1 and 1 mM, respectively. A concentration of TiO2 much clearer than those spots on TiO2 nanoparticle testing
suspension an order of magnitude higher is required to kill E. plates. Table 3 shows the summary of antibacterial activity of
coli than to kill B. megaterium, probably related to their different B. megaterium and E. coli by V-doped TiO2 solutions that are
cell wall structures and cell membranes. B. megaterium is a quite similar to those displayed in Tables 1 and 2.
Gram-positive bacterium with only one membrane, so B. 3.5. Bactericidal Activity Effect on TiO2 Nanocoatings. As
megaterium is more sensitive to the chemical agents than E. shown in the Experimental Section, the TiO2 nanocomposite
coli, which has a more complex cell wall with two cell coatings were prepared on glass slides for both qualitative and
8896 J. Phys. Chem. B, Vol. 109, No. 18, 2005 Fu et al.

Figure 10. Inhibition growth of E. coli (left plate) and B. megaterium (right plate). On each plate, the top three circles are coated with 5mM TiO2
nanoparticles and the bottom three circles are coated with 4 mM TiO2/Au nanocomposites.

TABLE 3: Antibacterial Effects with Different

Concentrations of V-Doped TiO2/Au suspension on B.
megaterium and E. coli
5 mM 1 mM 0.5 mM 0.1 mM 0.05 mM 0.01 mM
E. coli ++ + - - - -
B. megaterium + + ++ - - - -
quantitative testing of antimicrobial activity of E. coli and B.
megaterium. For the qualitative testing, the coated glass slide
on the left of Figure 10 shows inhibition of E. coli growth with
three 5mM TiO2 nanoparticle coating spots on the top and three
4 mM TiO2/Au nanocomposite coating spots on the bottom.
The circles were marked to indicate where the spray-coat spots
were made. As can be seen around the coated spots, there is
less growth or no growth of bacteria than in the surrounding
uncoated area. Similar results were also obtained for B.
megaterium, shown on the coated glass slide on the right of
Figure 10. Moreover, Figure 11 shows the antibacterial testing
for E. coli on coated glass slide sprayed by different nanocom-
posites. The top-left two spots were coated by 5 mM TiO2
nanoparticle solution, the top-right two spots were made by 4 Figure 11. Antibacterial testing for E. coli on coated glass slide sprayed
mM TiO2/Au nanocomposite solution, the middle-left two spots with different nanocomposites. See text for the conditions of coating.
were coated by 260 mM TiO2 nanoparticle solution, the middle-
right two spots were made by 260 mM TiO2/Au nanocomposite number of survival cells; in set II, the bacterial suspensions of
solution, the bottom-left two spots were identical to (or a repeat E. coli cells in concentrations of 10-2, 10-3, 10-4, and 10-5
of) those on the middle-left two spots made by 260 mM TiO2 cfu/mL were used for counting a larger number of survival cells.
nanoparticle solution, and the bottom-right two spots were For the experimental set I obvious changes in survival cells were
coated by 260 mM TiO2/Ag nanocomposite solution. The results not observed on the bare glass chamber with E. coli suspension
may be summarized as follows: (1) The two spots on the top under either room light (sample IA) or UV LED irradiation
right, which are coated by 4 mM TiO2/Au nanocomposite (sample IB). This indicates that the weak UV light alone used
solution, show excellent inhibition results. (2) Comparing the in this experiment did not affect the bacteria growth. In contrast,
5 mM TiO2 coated spots (top-left two spots) and the 260 mM the survival cells (using 10-7 cfu/mL sample) decreased on the
TiO2 coated spots (middle-left or bottom-left two spots), the TiO2/Au coating under room light (most of the time is in a dark
lower concentration (or smaller particle size) of TiO2 nanopar- incubator, sample IC) or UV LED light (sample ID). The
ticle exhibits better bacterial inhibition results. (3) Both the 260 efficiency is remarkably high and fast, as most of the bacteria
mM TiO2/Au nanocomposite solution (middle-right two spots) were killed in 5 min. There are not enough survival cells for an
and the 260 mM TiO2/Ag nanocomposite solution (bottom-right accurate counting after 5 min. In the second set of experiments
two spots) show a similar bacterial inhibition effect. (set II), a higher concentration of bacterial suspensions of E.
For the quantitative testing, the survival ratio of E. coli cells coli cells in 10-5 cfu/mL was used. Ideally, for having an
dropped onto a TiO2/Au coating was examined as a function accurate colony counting, the plate that is chosen for counting
of UV illumination time. The light exposure was done either the survival cells should be in the range of 30-300 colonies
under room light or with a low power LED device emitting at on the plate, and be expected to have no survival cells in 60
360, 390, or 470 nm. Two sets of experiments were made, and min of illumination. Similar results on the two control samples
the results are plotted in Figure 12. In set I, the bacterial were obtained: no obvious changes in cell survival for glass
suspensions of E. coli cells in concentrations of 10-2, 10-3, 10-4, slides without TiO2/Au coatings and under either room light
10-5, 10-6, and 10-7 cfu/mL were used for counting a smaller (sample IIA) or UV LED irradiation (sample IIB). However,
TiO2 Nanocomposites for Antimicrobial Coatings J. Phys. Chem. B, Vol. 109, No. 18, 2005 8897

metal salt (HAuCl4) by both a chemical reduction method and

a photoreduction method. The TEM images showed the gold-
capped TiO2 nanoparticles produced in this work had a very
well-defined shape and small size. The ratio of [TiO2]:[ Au]
directly affected the particle size and stability of the nanocom-
posites. Aggregation effects were observed with a [TiO2]:[Au]
ratio of 1:1 as indicated by TEM imaging. In contrast, the
nanocomposites with a [TiO2]:[Au] ratio of 20:1 were stable
and a thin gold shell was formed with no significant change in
particle size compared to the TiO2 nanoparticles. Different
percentages of vanadium-doped TiO2 nanoparticles were also
successfully prepared to extend the wavelength of TiO2 from
the UV to the visible region. The V-doped TiO2 samples showed
a red shift in the UV-vis absorption spectrum.
The antibacterial effect of those TiO2 nanoparticles, gold-
capped TiO2 naonocomposites, and V-doped TiO2 nanoparticles
evaluated on two types of bacteria, E. coli and B. megaterium,
were shown visually for the first time. The method to demon-
strate the antibacterial effect of those nanoparticles in two forms,
an aqueous phase and thin coating substrates, were successfully
developed. Results showed good inhibition on two types of
bacteria, E. coli and B. megaterium, even under room light. A
decrease in the survival ratio of E. coli and a fast killing effect
were observed. The results indicated the TiO2 and gold-capped
TiO2 nanocomposites prepared in this study possess strong
oxidizing ability and photocatalytic activity. The good antibacte-
rial effect (60-100% killing efficiency as observed in Figure
12) may be due to those particles small size, large surface area,
large band gap energy, and more active sites for carrying out
catalytic reactions. Further testing on the survival ratio of E.
coli and B. megaterium (or other microorganisms) with the TiO2,
gold-capped TiO2, and V-doped TiO2 nanoparticles dispersed
Figure 12. Changes in the survival of E. coli bacterial colonies on a
TiO2/Au nanocomposite coating. Set I and set II experiments were in coating formulations and coated glasses (or other substrates)
conducted by using a concentration of bacterial suspensions of E. coli is in progress.
cells in 10-7 and 10-5 cfu/mL, respectively.
References and Notes
the survival rate on the TiO2/Au coating under a UV intensity (1) Sunada, K.; Watanabe, T.; Hashimoto, K. EnViron. Sci. Technol.
of 1 mW/cm2 LED source (sample IID) was almost the same 2003, 37, 4785-4789.
as that on the coating under room light (sample IIC). This result (2) Jacoby, W. A.; Maness, C. P.; Wolfrum, E. J.; Blake, D. M.; Fennel,
J. A. EnViron. Sci. Technol. 1998, 32, 2650-2653.
is the same as the previous testing in aqueous solutions. This (3) Linkous, A. C.; Carter, G. J.; Locuson, D. B.; Ouellette, A. J.;
demonstrates that the nanosize anatase TiO2 with Au-capped Slattery, D. K.; Simitha, L. A. EnViron. Sci. Technol. 2000, 34, 4754-
composites made in an acidic medium possesses photocatalytic 4758.
activity even under room light or very weak UV illumination. (4) Sunada, K.; Kikuchi, Y.; Hashimoto, K.; Fuhishima, A. EnViron.
Sci. Technol. 1998, 32, 726-728.
The problem in this experiment was that, using a concentration (5) Wolfrum, E. J.; Huang, J.; Blake, D. M.; Maness, P.; Huang, Z.;
of 10-5 cfu/mL dilution, there were over 800 colonies on each Fiest, J. EnViron. Sci. Technol. 2002, 36, 3412-3419.
plate; a counting error could be the major problem responsible (6) Hagfeldt, A.; Gratzel, M. Chem. ReV. 1995, 95, 49-68.
(7) Murakami, Y.; Matsumoto, J.; Takasu, Y. J. Phys. Chem. B 1999,
for the inaccurate data. However, the results showed that the 103, 1836-1840.
gold-capped TiO2 composite coating exhibited good photo- (8) Alivisatos, A. P. J. Phys. Chem. 1996, 100, 13226-13239.
induced antibacterial effect for the sterilization of E. coli. (9) Henglein, A. J. Phys. Chem. 1993, 97, 5457-5471.
(10) Kamat, V. P. Chem. ReV. 1993, 93, 267-300.
(11) Linsebigler, A. L.; Lu, G.; Yates, J. T. Chem ReV. 1995, 95, 735-
4. Conclusion 738.
(12) Dawson, A.; Kamat, P. V. J. Phys. Chem. B 2001, 105, 960-966.
In this study, a novel and simple method was developed for (13) Yu, C. J.; Zhang, L.; Zheng, Z.; Zhao, J. Chem. Mater. 2003, 15,
the synthesis of highly photocatalytic, nanostructured, anatase 2280-2286.
TiO2 particles in aqueous solution by hydrolysis of titanium (14) Martin, S. T.; Morrison, C. L.; Hoffman, M. R. J. Phys. Chem.
isopropoxide in an acidic environment. Two different acids, 1994, 98, 13695-13704.
(15) Haick, H.; Paz, Y. J. Phys. Chem. B 2003, 107, 2319-2326.
HClO4 and HNO3, used as catalysts in the hydrolysis process (16) Hoffman, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W. Chem.
were investigated. The results of characterization of the TiO2 ReV. 1995, 95, 69-96.
nanoparticles by UV-vis absorption spectroscopy, XRD, TEM, (17) Stallings, W. E.; Lamb, H. Langmuir 2003, 19, 2989-2994.
(18) Peiro, A. M.; Peral, J.; Domingo, C.; Domenech, X.; Ayllon, J.
and AFM demonstrated that fairly uniform sized nanoparticles Chem. Mater. 2001, 13, 2567-2573.
of 10-15 nm diameter with spherical-shaped anatase form were (19) Mogyorosi, K.; Dekany, I.; Fendler, H. J. Langmuir 2003, 19,
successfully obtained. There was no influence on the TiO2 phase 2938-2946.
by the acid type used in this preparation method. (20) Hamid, A. M.; Rahman, A. I. Malaysian J. Chem. 2003, 5 (1),
086-091.
Moreover, the gold-capped TiO2 nanocomposites were suc- (21) Park, I.; Jang, S.; Hong, S. J.; Vittal, R.; Kim, K. Chem. Mater.
cessfully synthesized with a mixture of TiO2 solution and a noble 2003, 15, 4633-4636.
 8898 J. Phys. Chem. B, Vol. 109, No. 18, 2005 Fu et al.

(22) Tsevis, A.; Spanos, N.; Koutsoukos, P. G.; Van, A. L. J.; Lyklema, (28) Yu, C. J.; Yu, J.; Ho, W.; Jiang, Z.; Zhang, L. Chem. Mater. 2002,
J. J. Chem Soc., Faraday Trans. 1998, 94 (2), 295-300. 14, 3808-3816.
(23) Scalfani, A.; Palmisano, L.; Schiavello, M. J. Phys. Chem. 1990, (29) Choi, W.; Termin, A.; Hoffmann, M. R. J. Phys. Chem. 1994, 98,
94, 829-832. 13669-13679.
(24) Andersson, M.; Osterlund, L.; Ljungstrom, S.; Palmqvist, A. J. Phys.
Chem. B 2002, 106, 10674-10679. (30) Klosek, S.; Raftery, D. J. Phys. Chem. B 2001, 105, 2815-2819.
(25) Addamo, M.; Augugliaro, V.; Paola, A. D.; Garcia-Lopez, E.; (31) Tortora, G.; Funke, R. B.; Case, L. C. Microbiology; An Introduc-
Loddo, V.; Marci, G.; Molinari, R.; Palmisano, L.; Schiavello, M. J. Phys. tion; Addison-Wesley Longman, Inc.: New York, 2001.
Chem. B 2004, 108, 3303-3310.
(26) Zanella, R.; Giorgio, S.; Henry, R. C.; Louis, C. J. Phys. Chem. B (32) Blake, M. D.; Manesss, P.; Huang, Z.; Wolfrun, J. E.; Huang, J.
2002, 106, 7634-7642. Sep. Purif. Methods 1999, 28 (1), 1-50.
(27) Subranmanian, V.; Wolf, E. E.; Kamat, V. P. Langmuir 2003, 19, (33) Lu, Z.; Zhou, L.; Zhang, Z.; Shi, W.; Xie, Z.; Xie, H.; Pang, D.;
469-474. Shen, P. Langmuir 2003, 19, 8765-8768.

View publication stats