Photochemistry and Photobiology, 2005, 81 : 279-290
zyxwvut
Photocatalytic Coatings for Environmental ApplicationsTt
Norman S. Allen*', Michele Edge', Gonzalo Sandoval', Jo Verran',
John Stratton3 and Julie Maltby3
'Department of Chemistry and Materials, Faculty of Science and Engineering,
Manchester Metropolitan University, Manchester, UK
'Department of Biological Sciences, Faculty of Science and Engineering,
Manchester Metropolitan University, Manchester, UK
3Millenium Inorganic Chemicals, Grimsby, UK
zyxwvutsrq
Received 22 June 2004; accepted 19 July 2004
ABSTRACT
A series of nano- and micronparticle-grade anatase and rutile
titanium dioxide pigments have been prepared with various
densities of surface treatments, particle size and surface area.
Their photocatalytic activites have been determined in a series
of paint films by FTIR, chalking, color, gloss change and weight
loss after artifical weathering. The pigments have also been
examined by rapid assessment methodologies using photodielectric microwave spectroscopy, 2-propanol oxidation and
hydroxyl analysis. The microwave response under light and
dark cycles provides an extended timescale probe of chargecarrier dynamics in the pigments. Pigment particle size, surface
area and properties clearly play an important role in dispersion
and any polymer-pigment interactions. Photooxidation studies
on several types of paint films show a clear demarcation
between nanoparticle- and pigmentary-grade titanium dioxide,
with the former being more active because of their greater
degree of catalytic surface activity. The photosensitivity of
titanium dioxide is considered to arise from localized sites on
the crystal surface (i.e. acidic OH), and occupation of these sites
by surface treatments inhibits photoreduction of the pigment
by ultraviolet radiation; hence, the destructive oxidation of the
binder is inhibited. Coatings containing 2-5 % by weight
alumina or alumina and silica are satisfactory for generalpurpose paints. If greater resistance to weathering is desired,
the pigments are coated more heavily to about 7-10% weight.
The coating can consist of a combination of several materials,
e.g. alumina, silica, zirconia, aluminum phosphates of other
metals. For example, the presence of hydrous alumina particles
lowers van der Waals forces between pigments particles by
several orders of magnitude, decreasing particle-particle
attractions. Hydrous aluminum oxide phases appear to
improve dispersibility more effectively than most of the other
hydroxides and oxides. Coated nanoparticles are shown to
exhibit effective light stabilization in various water- and oilbased paint media in comparison with conventional organic
stabilizers. Hindered piperidine stabilizers are shown to
provide no additional benefits in this regard, often exhibiting
strong antagonism. The use of photocatalytic titania nanoparticles in the development of self-cleaning paints and microbiological surfaces is also demonstrated in this study. In the
former case, surface erosion is shown to be controlled by
varying the ratio of admixture of durable pigmentary-grade
rutile (heavily coated) and a catalytic-grade anatase nanoparticle. For environmental applications in the development of
coatings for destroying atmospheric pollutants such as nitrogen
oxide gases (NO,), stable substrates are developed with photocatalytic nanoparticle-grade anatase. In this study, porosity of
the coatings through calcium carbonate doping is shown to be
crucial in the control of the effective destruction of atmospheric
NOx gases. For the development of microbiological substrates
for the destruction of harmful bacteria, effective nanoparticle
anatase titania is shown to be important, with hydrated high
surface area particles giving the greatest activity.
zyxwvutsrqpo
zyxwvu
zyxwvutsrqp
zyxwvutsrqp
INTRODUCTION
Of the many pigments and fillers used commercially in polymers and
coatings applications, one of the most important and complex is
titanium dioxide (1-14). To date, research activities into the
photochemical properties of titanium dioxide relate to a number of
areas. One current area of topical interest deals with studying the
photocatalytic activity of titanium dioxide to convert harmful organic
waste products into harmless environmentallyacceptable materials (9),
whereas another deals with minimizing the photocatalytic activity of
Ti02 to enhance the durability of organic substrates such as polymers
and coatings (1,2). In many plastics and coatings applications,
however, the prime use of titanium dioxide is as an opacifier.
The ability of pigments to catalyze the photooxidation of
polymer systems has also received significant attention in terms of
their mechanistic behavior. In this regard, much of the information
originates from studies carried out on TiOz pigments in both
polymers and model systems (1,2). To date, there are three current
mechanisms of the photosensitized oxidation of polymers by Ti02
l/Posted on the website on 27 July 2004
*To whom correspondence should be addressed: Department of Chemistry
and Materials, Faculty of Science and Engineering, Manchester Metropolitan University, John Dalton Building. Chester Street, Manchester M 1
SGD, UK. Fax: 0044-161-247-6357;e-mail: n.s.allen@mmu.ac.uk
Abbreviufions:HALS, hindered m i n e light stabilizer; NOx, nitrogen oxide
gases; PVC, poly(viny1 chloride); UV, ultraviolet; VOC, volatile organic
compound.
?Presented at VIn ELAFOT Photochemistry Congress in La Plata, Buenos
Aires, 8-1 1 November 2004.
0 2005 American Society for Photobiology 003 1-8655/05
279
zyxwvutsrqp
zyxwvutsrqp
zyxwvutsrqp
zyxwvutsrq
280 Norman S. Allen eta/.
zyxwvuts
zyxwvutsrqponm
zyxwvutsrq
zyxwvu
and for that matter other white pigments such as ZnO. These are
as follows.
The formation of an oxygen radical anion by electron transfer
from photoexcited Ti02 to molecular oxygen (13,14). A recent
modification of this scheme involves a process of ion annihilation
to form singlet oxygen, which then attacks any unsaturation in the
polymer (3).
+ 2 TiOi’ + 0,’ (I)
(I) Ti02 + ‘Oz(ion annihilation)
(I) + H 2 0 + Ti02 + HO’ + H02’
Ti02
0 2
+
2H02’
RCH2
CHR’
1
+
H202
+
0 2
+ ’ 0 2 + RCH = CHCH(O0H)R’
Formation of reactive hydroxyl radicals by electron transfer from
water catalyzed by photoexcited Ti02 (14). The Ti3+ ions are
reoxidized back to Ti4+ ions to start the cycle over again.
Ti02
H20
[Ti4+] e‘
[Ti3+] O2
+
+
hv
+
+
+
+
H+ e’(Aqu)
[Ti3+]
[Ti4’]
+ ‘OH
Irradiation of TiOz creates an exciton (p), which reacts with the
surface hydroxyl groups to form a hydroxyl radical (15). Oxygen
anions are also produced, which are adsorbed on the surface of the
pigment particle. They produce active perhydroxyl radicals.
+
Ti02 2 e’ (p)
OH- (p) -+ HO’
Ti4+ + e’ + Ti3+
+
+ + [Ti4+. . . 02-]adsorbed
[Ti4+. . . 0’-] adsorbed + H2O + Ti4’ + HO- + HO;
Ti3+
0 2
Titanium dioxide pigments (titania) exist in two morphological
crystalline forms that exhibit different photoactivities when
incorporated into a number of commercial polymers
(1,2,10,14,16). Each modification exhibits 6:3 coordination, but
whereas the anatase forms a distorted octahedral structure the rutile
forms slender prismatic crystals that are often twinned. Differences
in the photoactivities of the two modifications of titania depend
markedly on the manufacturing history of the pigment. Anatase
pigments are generally more photoactive than the rutile types
(17,18). To improve pigment dispersion and reduce photoactivity,
the surface of the pigment particles is coated with precipitated
aluminosilicates. Zirconates are also used in some instances,
whereas for other applications such as in nylon polymers and
fibers, the anatase is coated with manganese silicates or
phosphates. Anatase will photosensitize the oxidation of a polymer,
the effect is dependent on the nature and density of the coating and
increases with pigment concentration. Uncoated rutiles are also
photosensitizers, but again, the effect is reduced and proportional
to the effectiveness of the coating. In this case, stabilization
increases with increasing coated rutile concentration. Thus, the
surface characteristics of the titania pigment is an important factor
in controlling photoactivity (19,20). The surface is covered with
hydroxyl groups of an amphoteric character formed by the
adsorption of water. These groups are more acidic in character
on the surface of anatase and less effectively bound than those on
rutile. The surface carriers (excitons), therefore, react more slowly
with the hydroxyl groups in the case of rutile.
Surface modifications of the TiOl particles with inorganic
hydrates may reduce the photochemical reactivity of titanium
pigments. This can reduce the generation of free radicals by
physically inhibiting the diffusion of oxygen and preventing the
release of free radicals. The often simultaneous chemical effects of
surface modification can involve provision of hole and electron
recombination sites or hydroxyl-radical recombination sites. In
addition to the latter effects, the surface treatment or coating, as
mentioned above, can improve other properties such as improve
the wetting and dispersion in different media (water, solvent or
polymer) to improve compatibility with the binder and dispersion
stability and color stability. The photosensitivity of titanium
dioxide is considered to arise from localized sites on the crystal
surface, and occupation of these sites by surface treatments inhibits
photoreduction of the pigment by ultraviolet (UV) radiation; hence,
the destructive oxidation of the binder is inhibited. Coatings
containing 2-5% by weight alumina or alumina and silica are
satisfactory for general-purpose paints. If greater resistance to
weathering is desired, the pigments are coated more heavily to
about 7-10% weight. The coating can consist of a combination of
several materials, e.g. alumina, silica, zirconia, aluminum phosphates of other metals. For example, the presence of hydrous
alumina particles lowers van der Waals forces between pigments
particles by several orders of magnitude, decreasing particleparticle attractions. Hydrous aluminum oxide phases appear to
improve dispersibility more effectively than most of the other
hydroxides and oxides.
During the weathering of commercial polymers containing white
pigments such as titania, oxidation occurs at the surface layers of
the material, which eventually erodes away, leaving the pigment
particles exposed. This phenomenon is commonly referred to as
“chalking” and has been confirmed by scanning electron
microscopy (3). Titania pigments absorb strongly in the near UV
region, with anatase having a cut-off point at 340 nm and rutile at
370 nm. Simply on this basis, one may argue that the latter will
screen more of the active UV radiation. Methods of assessing
pigment photoactivities have attracted much interest from both
scientific and technological points of view. Artificial and natural
weathering studies are tedious and very time consuming.
Consequently, numerous model systems have been developed to
rapidly assess their photochemical activities. Most of these systems
undergo photocatalytic reactions to give products, which are easily
determined, usually by UV absorption spectroscopy, highperformance liquid chromatography or gas chromatography etc.
Some examples of predictive methods are listed: (1) isopropanol
oxidation measuring acetone formation (21); (2) oxygen uptake
(13); (3) electron spin resonance for Ti (111) ions (22); (4)
luminescence measurements (4,7,23); (5) hydroxyl analysis (24);
and (6) microwave spectroscopy (25,26).
Microwave photodielectric spectroscopy has been found by far
to be one of the latest and interesting methods for ascertaining
pigment photoactivity and photoconductivity (26-28). Similar to
all other spectroscopic methods, one obtains an absorption
spectrum in the microwave region as return frequency loss, which
is related to the polarizability of the pigment structure. The higher
the initial absorption frequency the more polarizable the structure,
zy
zy
zyxwvutsrqponm
Photochemistry and Photobiology, 2005, 81 281
Table 1. Properties of pigments used in this study
Sample
BET surface area m2/g
zyxwvut
zyxwv
Particle size
Surface treatment
% Surface treatments
~~
A-Anatase normal
B-Rutile normal
C-Rutile normal
D-Rutile normal
E-Nan0 anatase
F-Nan0 anatase
G-Nano anatase
H-Nano anatase
I-Nano rutile
J-Nan0 rutile
K-Nano anatase
L-Nan0 rutile
M-Nan0 anatase
N-Nano anatase
0-Rutile normal
10.1
6.5
12.5
12.5
44.4
77.9
329.1
52.1
140.9
73.0
190.0
73.0
239.0
190.0
12.5
0.24 p
0.28 p
0.25 p
0.29 p
20-30 nm
15-25 nm
5-10 nm
70 nm
25 nm
40 nm
6-10 nm
30-50 nm
71 nm
92 nm
2.50 nm
hence greater activity, with anatase being more active than rutile.
Irradiation of the Ti02 in a microwave cavity causes charge
separation and increased polarization. The absorption frequency
then shifts with time, providing a measurement of the extent of
charge generation. On switching off the light, the charge carriers
recombine. Fast rise and decays indicate shallow traps, whereas the
more slower decays indicate deep traps.
In recent years, however, there has been extensive interest in
ultrafine or nanoparticle fillers and pigments, especially with regard
to their properties as a UV “blocker” in coatings applications
(29,30). Thus, whereas for conventional pigmentary titania light
scattering is governed by the Mie theory, for ultrafine titania
Rayleigh’s scattering applies. Thus, the finer the particle the more
effective will be the scattering of UV light below 400 nm. At an
optimum particle size of 2-50 nm, visible light will be transmitted,
and the system becomes essentially translucent. Recent study
has also indicated that for ultrafine titania, physisorption and
chemisorption process are paramount and can significantly influence the performance characteristics of stabilizers in different
ways (31), and this can in turn control the stability of the matrix.
In this article, we have undertaken a study on a comparison of
the photochemical behavior of a selection of nano- versus micron
(pigmentary)-grade anatase and rutile titania pigments in poly
(vinyl chloride) (PVCtalkyd, acrylic, fluoroacrylic, polyestersiloxane and isocyanate paint films. Rates of oxidation are
measured via infrared (FTIR), gloss, color and changes in mass.
The interrelationship in the photoactivity of the TiOz, as measured
by rapid assessment methods on the basis of microwave photodielectric spectroscopy, hydroxyl content and 2-propanol oxidation,
has been demonstrated. The UV absorption activity of coated
ultrafines of anatase and rutile in acrylic wood coatings and
automotive-grade isocyanate clear over-varnish have also been
compared with conventional organic UV absorbers and hindered
amine light stabilizers (29,30). Techniques based on FTIR, mass
loss and color change are used in this study to follow the
degradation processes. Wood is very sensitive to weathering, and
through the action of sunlight (UV light) and moisture, the
unprotected wood surface will degrade outdoors leading to the
decomposition of lignin and discoloration because of quinone
formation, causing the wood to yellow or darken in color.
Traditional choices for protecting the wood from rapid discoloration have involved the use of organic UV absorbers, hindered
None
A1
Al
Al
None
None
None
Hydroxy apatite
None
Al, Zr
Al, Si, P
Al, Zr
Al, Si, P
Al, Si, P
Al, Si, P
1
2.8
3.4
5
13
20
13
12
20
3.5
amine light stabilizers or transparent iron oxides. The addition of
organic UV absorbers protects the wood from discoloration during
the early stages of the product lifetime; however, the effect is not
permanent, and such stabilizers tend to migrate or decompose
during exposure. On the other hand, transparent iron oxides impart
color to the coating, a feature that is not always desirable. Ultrafine
Ti02 is a pure inorganic UV filter, and therefore, when properly
applied it should be able to provide constant protection to
the substrate.
Other applications involving the photocatalytic effect of titania
are the use of self-cleaning paints and antibacterial activity. In the
first case, it is important to be able to control the surface activity of
the paint matrix allowing shedding of micron layers only. In this
study, control of oxidation of the substrate has been investigated
through the use of mixtures of pigmentary and nanoparticle grades
of titania in both siliconized polyester and fluorinated acrylic
paints. In terms of environmental atmospheric pollution, porous
paint coatings have also been developed in this study and shown to
be effective in the destruction of nitrogen oxide gases (NO,). In
this study, the admixture of nanoparticle titania with calcium
carbonate in the matrix plays a crucial role in the rate of NO,
destruction. Finally, the antibacterial role of nanoparticle titania is
also demonstrated on agar gels. The pigments used in this study
have been specially treated by different milling operations to alter
particle size and surface area as well as being subject to different
inorganic and organic surface treatments. Sol-gel anatase has also
been investigated in paint matrices for ease of dispersion and
enhanced activity.
zyxwv
MATERIALS AND METHODS
Materials. The titania pigments used in this study and their associated
properties were all experimental grades, prepared in the laboratories of
Millenium Inorganic Chemicals (Grimsby, UK), and are listed by
alphabetical codes in Table 1. Experimental samples of titania prepared
via the sulfate and chloride routes were milled for different periods of time
to examine the influence of particle size on durability and spectroscopic
properties, as were various untreated and surface-treated grades of
pigments. Various laboratory-grade sol-gel anatase titanias were also
used in the NOx destruction studies for paints and were undried.
Paint formulations. The alkyd paint formulation used is based on an 18%
PVC air drying long oil alkyd based on commercial Sobral P470 supplied by
the Scott Bader Company Ltd. (Wellingborough, Northants, UK). The paint
zyxwvutsrqpo
zyxwvutsr
zyxwvutsrq
zyxwvutsrqp
zyxwvutsrq
282 Norman S.Allen et a/.
zyxwvu
zyxwvuts
Figure 2. Attenuation of power stored in microwave cavity (dQ) with
on-off irradiation cycles (s) for different titania pigments.
Figure 1. Comparison of hydroxyl content and oxygen consumption
rates for 2-propanol oxidation with pigmentary and nanoparticle
titanium dioxide.
Oxygen consitniption and hydroiyl content. Oxygen uptake was
measured before and after irradiation of the pigments in a sealed glass
tube loaded onto a metallic support. The tubes are irradiated via a series of 8
W black light 340 nm bulbs for 5 h at 40°C. Hydroxyl analysis is referred to
elsewhere (24).
Microwave technique. Microwave measurements were undertaken using
a Marconi (6200A) 2-20GHz (Wanvick Rhode Island. UK) programmable
sweep generator and an automatic amplitude analyzer, coupled to a circular
waveguide and cylindrical cavity. Powdered samples (0.2 g) were packed in
the bottom of the cavity on a plastic dish (to ensure reproducibility) and
were irradiated in the cell chamber with an ILC 302UV xenon source via an
optical fiber set-up (Laser Lines Ltd., Banbury, Oxon, UK) switchable
between UV and visible light with a cut-off point at 400 nm. In this study,
only visible light was used because the UV response is weak for the titania
pigments. Changes in microwave cavity resonant frequency and attenuation
of microwave power were monitored during 1800 s of irradiation and for
1800 s after switching off the light source (unless stated otherwise). All the
measurements were carried out in triplicate and at 25°C. The uncertainty in
the values of “shift in microwave cavity resonant frequency” is of the order
?0.0004 GHz; whereas that for “attenuated power” is 20.05 dBm.
Antibacterial studies. Water suspensions of the titania particles were
prepared and placed in petri dishes irradiated with a 100 W high-pressure
Hg lamp. Twenty-five microliters of lo5 bacteria per cm3 (Escherichia roli
strain NCTC 9001) was used as a model microorganism. After different
periods of irradiation, 0.1 cm3 of the suspension was transferred to an agar
medium and then incubated in an oven at 37°C for 24 h. The number of
surviving bacteria on the agar matrix was then counted.
N O , measurements. NOx (NO and NOz) measurements were obtained
using a Signal chemiluminescence detection instrument (Signal Group Ltd.,
Camberley, Surrey. UK). Nitrogen with 100 ppm of NO was used as the
flow gas (I L/min) and subsequently diluted down to 30 ppm and passed
over paint films 10 cm2 applied to Melinex polyester substrates and
irradiated using a 10 W/m2 365 nm fluorescent tube.
zyxwvutsrqponml
drier is a commercial Manosec CD33 supplied by Rhone Poulenc Chemicals
(Manchester, UK), whereas the methyl ethyl ketoxime was supplied by
Banner [Samuel] & Co. Ltd. (Liverpool, UK). The resin and white spirit are
initially mixed and then dispersed at 2000 rpm using a Dispersmat AE3-C
with SO mm diameter impeller for 30 rnin (until no striations are observed).
The pigment is then added to the dispersion that is then ball-milled using
steatite balls in Melinex capped glass jars. The milled base was then let down
using the fomiulation shown and agitated for 10 rnin using the Dipsersmat
and then placed on the rollers for 15 min followed by a 15 h stabilization
period. The paint films (coated on stainless steel plates for accelerated
weathering and aluminum plates for natural exposures) were conditioned for
10 days at 40°C to remove any trapped solvent.
The water-based acrylic wood stain formulation, supplied by Rohm and
Haas (Philadelphia, PA), is based on Part A (Primal AC-337,61.3 g; BYK024, 0.1 g; water, 75 g; Texanol, 2.3 g and ammonia, 0.3 g) and Part B
(water, 95 g; Acrysol 2020, 3.5 g and Acrysol RM-I2W, 12 g). Part A is
mixed with the appropriate amount of nanoparticles and stabilizers using
a laboratory stirrer for 10 min. Part B (50% wt/wt) was then added and
mixed for a further 5 min. The prepared paints were then applied to
pinewood panels for color measurements during aging. Samples were also
applied to aluminum plates via a 15 micron K-bar with a drying period of
24 h between the coats. These test specimens were used for mass loss, gloss
and FTIR analysis during aging.
The isocyanate water-based acrylic paint (clear autofinish) was formulated
and supplied by Bayer (Leverkusen, Germany), siliconized polyester paint
from DSM (Zwolle, The Netherlands) and Lumiflon paint from Asahi Glass
Fluoropolymers (UK) Ltd. (Lancashire, UK). An aqueous UV absorber
dispersion. SanduvorB 3041 (2,4-dihydroxyhenzophenone),and aqueous
HALS (hindered amine light stabilizer, Hostavin N-20), Sanduvor 3051, were
obtained from Clariant (UK) Ltd. (Leeds, UK). The nanoparticles and UV
stabilizers were added to the paint systems at 1%. 2% and 5% (by weight on
total resin solids) along with 0%, 0.5% and 1.0% HALS.
Aging and weathering. The paint films were weathered using an Atlas
Ci65 weatherometer and durability measured via mass loss and gloss loss.
Gloss was measured using a BYK Micro Tri Glossmeter at 60”.
Acrylic coating samples were aged in a QUV, Q-Panel Company
(Bolton, UK). The irradiation conditions used were UV 340 nm fluorescent
lamps for outdoor simulation and black body temperature of 60°C during
irradiation and 45°C during the condensation cycle. The cycle for
condensation was 4 h. Spray cycles in this instance were not used because
of the potential loss of films from the substrates. Mass measurements were
recorded at appropriate time intervals during the irradiation period on
coated substrates. Samples were conditioned to constant weight in air in
a desiccator with silica gel for 4 h before weighing. Infrared spectra were
recorded and analyzed using a Nicolet Nexus FTIR spectrometer. Color
measurements were taken using an XI-800 Tristimulus Colorimeter series
from Gadner Laboralones (TX). Measurements were also made using
a Gretag MacBeth Spectral-Eye (Liverpool, UK).
zyxw
RESULTS AND DISCUSSION
Photoactivity tests for 2-propanol oxidation and
hydroxyl content
These are specific tests to ascertain pigment photoactivity. The
oxidation of 2-propanol to yield acetone is a specific methodology,
and in this study, it has been related to oxygen consumption during
irradiation of the medium in the presence of the titania particles
(22).The hydroxyl content relates to the concentration of hydroxyl
functionalities present on the pigment particles and is often related
to activity (25). The data for both tests are compared in Fig. 1 for
a selection of the titania particles studied. There are a number of
correlations and trends within the data. First, all the nanoparticle
grades exhibit higher photoactivities than the pigmentary grades.
Photochemistry and Photobiology, 2005, 81 283
z
I
zyxwvuts
zyxwvutsrqponmlkjih
zyxwvutsrqponmlkjihgf
zyxwvutsrqpo
M
zyxwvutsrqponm
0
44
oY"""""""'""""''""
0
200
400
~ ~ " ' " " " " " " " " " " " " " " " " " " " ~ '
600
800
'1000
100
148
Figure 3. Weight loss (mgi100 cm2) of PVC alkyd paint films during
irradiation in an Atlas weatherometer containing equivalent amounts of
titania pigments.
Thus, for oxygen consumption, the anatase A is more active than
the rutile Types B and D, the latter being the least active and most
durable pigment. Second, of the nanoparticles, the rutile Grade I is
the most active in both the tests. The three anatase Grades E, F and
G exhibit increasing activity with hydroxyl content, whereas for
oxygen consumption F is greater.
Microwave analysis
In this technique, microwaves are directed, by a waveguide,
through an aperture into a cavity. At certain frequencies the cavity
abstracts appreciable power because the oscillating electric and
magnetic fields of the microwave energy reach a maximum when
they are resonant with the cavity. Typical microwave spectral
changes for a selection of the anatase and rutile pigments used in
this study with respect to that for the cavity are shown in Fig. 2.
The amount of energy stored by a microwave cavity is known as its
quality factor (Q-factor). When a sample is placed in the cavity and
exposed to UV-visible light, the position of the resonant frequency
shifts and microwave power is attenuated.
The attenuation of microwave power is proportional to the
changes in conductivity of the sample as free carriers are produced.
This is because of the creation of large numbers of phonons (heat)
by the free carriers, thus reducing the ability of the microwave
cavity to store energy. The magnitude of the resonant frequency
shift is a measure of the extent to which the electric charge
distribution in the sample can be distorted or polarized by the
electric field. The nature and concentration of both free and trapped
camers in the sample will influence this effect.
The data depicted relate to the exposure and microwave
conductivity response of the titanium dioxide powders. Figure 2
shows that during irradiation, the attenuation in microwave power
is greater in the presence of alumina on the surface of the rutile
pigments B, C and D. This is consistent with a reduction in freecarrier population and enhanced trapping as in the case of the more
heavily coated Grade D. The positive shift in resonant frequency
on irradiation arises from excess carrier formation, after saturation
of traps, during prolonged exposure. The nanoparticles E, F and G
show an interesting behavior. They all exhibit greater carrier
generation than the rutile pigments-hence
greater activity.
However, pigment G after a fast initial rise in carrier production
284
380
428
468
528
588
irradiation Time (h)
1200
IrradiationTime (hrs)
212
Figure 4. Gloss loss of PVC alkyd paint films during irradiation in an atlas
weatherometer containing equjvalent amounts of titania pigments.
displays a strong plateau as a result of strong trapping followed by
a sharp dark decay to the valence state after extinguishing the light
source because of carrier recombination. Pigment G is more finely
milled than either E or F and hence is likely to contain a greater
number of small, shallow defect sites able to trap out the carriers.
No such rapid dark recombinations were evident in the case of the
pigments E and F. This would certainly account, in part, for the
apparent lower photoactivity of pigment G in the polymer study
than E and F. The dark recombination process in pigments E and F
suggests that excess carriers are trapped in deep-lying states giving
rise to the observed long-lived residual frequency component in the
decay process. The hydroxyl apatite-coated pigment H also
exhibits rapid camer generation although less than that for
pigments E and F followed by a gradual dark recombination as
for E and F. Thus, the particular coating used in this study has little
influence on carrier generation and trapping, hence retaining the
activity of the pigment. In all, measurements on microwave
dielectric properties of titania pigments provide a useful and more
versatile tool for the assessment of their photoactivity extended to
a polymer matrix.
zyxw
Paint degradation studies
The relative photoactivities of the pigments and nanoparticles also
have been ascertained by measuring their influence in the first
Figure 5, Transmission spectra of Lowilite 24 and Lowilite 26 compared
with nanoparticles anatase K and rutile L in a cured alkyd resin film (0.85
mm) at 2% wt/wt on resin solids.
zyxwvutsrqp
zyxwvutsr
zyxwvutsrqponm
284 Norman S. Allen eta/.
35
J"
30
25
35 F
F
1
25
15
10
5
J
0
0
50
100
150
IRRADlATlON TIME, HRS
200
o ~ ~ " " " " ' " " " ' ' '
zyxw
zyxwvutsrqponmlk
zyxwvutsrqponm
zyxwvutsrqpon
50
0
100
150
200
zyxwvutsrqponm
IBlank
-X-I%W+05%H4LS
-x- Z%W+I%MLS
+Blank+O 5%MLS
UI%W+l%H4LS
+"Blank+l%WALS
Q2%W
-5 % W
+-5%W+O
IRRADlATlON TIME, HRS
4 1 % W
*2%W+0
5%tJALS
5%MLS +5%W+l%HALS
-C Blank
+BlankcO 5%HALS +Blank+l%MLS
0l%+O S%HALS
0l%+l%W\IS
+-m*
*Z%+I%HALS
-5%
+5%+0
5%HALS
U 1%
-X- Z%+O.S%HALS
-X- 5%+I%MLS
Figure 6. Color change versus irradiation time (hours) in an Atlas
weatherometer for water-based acrylic paint films with UV absorber
(Sanduvor 3041) and HALS (Sanduvor 3051) combinations.
Figure 8. Color change versus irradiation time (hours) in an Atlas
weatherometer for water-based acrylic paint films with the HALS
(Sanduvor 3051) and nanoparticle L combinations.
instance on the durability of an 18% wt/wt PVC-based alkyd paint
matrix. Mass loss and gloss loss are the two industrial parameters
often used and are compared in Figs. 3 and 4, respectively. These
results clearly show that pigment activity is divided into two main
trends. The gloss loss for G, E, F and H diverges after 400 h of
irradiation. For both mass loss and gloss loss, all four nanoparticle
anatase pigments are the most photoactive. The rutile pigments B
and D and the anatase pigment A exhibit similar activity to that of
the nanoparticle rutile Grade I.
In the case of clear paints, ultrafine particles can display useful
properties, especially in terms of stability. This applies notably to
coated nanoparticles as UV blockers. For comparative studies on
transmission spectra, organic benzotriazole and 2-hydroxybenzophenone absorbers were dissolved in alkyd resins and cast as films
both with and without nanoparticles K and L. The alkyd system is
more UV transmissive than the acrylic resins. Transmission spectra
were recorded from both 0.32 and 0.85 mm thick films and are
shown in Fig. 5 for the latter as an example. Both absorbers peak at
around 350 nm and then tail off rapidly to 400 nm. The
nanoparticle rutile absorbs more strongly at higher wavelengths
above 380 nm than the nanoparticle anatase although in the thicker
film both types exhibit strong opacity up to 700 nm.
In the first instance, the UV absorber and HALS systems were
compared. Color changes during the Atlas weathering of acrylic
films are displayed in Fig. 6 and show an initial increase during the
first 100 h before attaining limiting values. From this data,
a number of interesting features are apparent. At all dosages, the
HALS has a minimal effect on the stability of the acrylic base. The
UV absorber is more effective at inhibiting color change, but
combinations with the HALS prove to be antagonistic, especially at
1.0% wt/wt combination. At 5% UV absorber there is effective
stabilization; again, the presence of the HALS provides no
additional benefit.
In the water-based acrylic film, the nanoparticles M, N and L
have been examined at concentrations of I%, 2% and 5% wt/wt.
Color changes in the paint films containing these particles are
shown in Fig. 7 together with those from the control containing the
UV absorber (Sanduvor 3041). In terms of color change measurements, the UV absorber was more effective than the rutile
nanoparticles when used at a loading of 1% wt/wt; however, at
higher nanoparticle loadings the color stability was equal to that
afforded by the UV absorber. The anatase nanoparticles also
offered good color stability, with the smaller 70 nm particles giving
the better protection.
The color stabilization of the L rutile nanoparticles, both with
and without HALS, is shown in Fig. 8. As in the previous case,
increasing the concentration of rutile nanoparticles enhances color
stability, whereas the addition of HALS has a negative effect,
especially at the higher 1% wt/wt concentration. The effect of color
35
30
k
0
140
50
100
150
200
IRRADWTION TIME, HRS
~
0
-CONlRCL
FILM
Q PIGMENT L1%
+PIGMENT N9Onm 5%
-+-PiGMENTM7Onm2%
-x- W absorber 2%
0PIGMENT M70nml%
-x- W absorber 1%
+PIGMENT
L5%
6PIGMENTNSOnm 2%
-PIGMENT
N9Onml%
--t PIGMENT M7Onm 5%
6W absorber5%
+PIGMEML
2%
Figure 7. Color change versus irradiation time (hours) in an Atlas
weatherometer for water-based acrylic paint films with UV absorber
(Sanduvor 3041) and nanoparticle combinations.
~
'
'
200
"
'
400
' ~' ' ' " " ' " " " ' ' ' ~
600
800
1000
" '
" " " " " " '
1200
1400
1600
1800
IRRADIATION TIME, HRS
+Blenk
U 1% N (20% ST nanoanata~e)
t 1% M (12% ST nanoanateae)
+I%L(13%STnanorutile)
Figure 9. Weight loss (mg/100 cm') for acrylic paint films during
irradiation in a QUV weatherometer containing equivalent amounts of
titania nanoparticles at 1% wt/wt.
zy
zyxw
Photochemistry and Photobiology, 2005, 81 285
180
160
140
0
120
E 100
s3+
80
zyxwvutsrqponmlkji
zyxwvutsrqponmlkjih
'+-+-+--+-+
60
40
$
s
20
0
500
0
1500
I000
0
2000
346
-X-Blank
-X- BlankrO5% W L S
+ 1% M(12% STnanoanatase)
0 1 % MtO596WLS
+2%M+O 5%WLS
+5%M+0 5%WLS
+2%M
-5XM
-+-Blank+l%H4LS
9 1 % M+I%MLS
Conlrol
+5%
Ruble J
+-1% PALS + 2% Absorber
+5%AnataaeK
Q
*2%Mil%n4LS
+5%M+l%MLS
Figure 10. Weight loss (mg/100 cm2) for acrylic paint films during
irradiation in a QUV weatherometer containing increasing amounts of
titania nanoparticles M at 1%, 2% and 5% wt/wt with and without the
HALS Sanduvor 3051 (0).
change on actual pinewood panels is noted visibly. In this study,
the irradiated portion of the panel for the control had turned a redbrown color because of the well-established formation of quinone
products produced from the photolysis of the phenolic lignin in the
wood (10,29,30).
Mass loss experiments have also been undertaken (Fig. 9) with
the M-, L- and N-type nanoparticles (1% wtlwt loading). This
study, undertaken in the QUV, shows that unlike the color change
data, the smaller particle size and least coated grade anatase is the
most powerful photosensitizer, an effect consistent with observations in thermoplastics (14). Again, this effect supports the theory
that the photoactivity of the nanoparticles enables them to play an
active role in bleaching, therefore, counteracting the color
development and hence giving rise to the superior color
stabilization activity. The L-type rutile nanoparticles offer a degree
of stability compared with the control film.
The powerful photosensitizing activity of the M-type nanoparticles is demonstrated in Fig. 10, both with and without HALS
present for mass loss. Increasing the concentration of H from 1%to
2% wt/wt reduces the rate of degradation, whereas increasing the
concentration further to 5% wtlwt enhances the degradation rate.
35
30
25
w 20
U
15
769
1143
1644
1897
2236
2657
IRRADIATIONTIME, HRS
IRRADIATION TIME, HRS
-0-BM b J (13% ST)
+5%
Ruble J + 1% HlyS
+2%
Anabre K (2W ST)
+5%AnabseK+l%MS
- t B M l e J +1%WS
~ 2 Absorber
%
- X - B Anatase K + 1%H U S
Figure 12. Gloss change versus irradiation time in an Atlas weatherometer
for a series of isocyanate acrylic clear overbase auto finish varnishes
containing rutile J and anatase K nanoparticles with and without HALS.
Clearly, photocatalytic activity is greater at the higher loading
because of the enhanced particle-coating contact. The M-type
anatase at a 5% wt/wt loading and HALS at 1% wt/wt loading were
highly detrimental to the coating and caused the highest mass loss.
The addition of 0.5% HALS also gives rise to strong antagonism.
Although the presence of the UV absorber is beneficial, it
nevertheless has a limited lifetime because of the inherent
instability of the chromophore. Color changes in these films are
also monitored and shown in Fig. 11. The beneficial effect of the
anatase in bleaching the color is illustrated; but again, the presence
of a HALS has little additional effect, especially at 5% wt/wt of M.
Analysis of degradation rates has also been undertaken on
a series of isocyanate acrylic overbase auto finishes for
consistency. In this case, gloss change was also measured, and
the effects are shown in Fig. 12. The J-type Rutile and K-type
anatase particles are used in this application at 2% and 5% wt/wt
concentration. In this study, there is a clear differential between the
two forms of titania with the rutile form exhibiting high stability.
The contribution of HALS is also not evident in this case because
of the high performance of J. In the case of anatase, there is
zyxwvut
zyxwvutsrqpo
zyxwvutsrqponmlkj
Table 2. Lumiflon pain pigmented with RCL-969Bano Titanium, 546
hours Atlas Exposure
zyxwvutsrqponm
10
5
0
0
50
100
150
200
IRRADIATIONTIME, HRS
-m- blank
4z- 1%
-2%
-x-
5%
tblank+O 5%WLS
t 1%+05%MLS
-0 2%+0 5%HALS
+5%+0 S%HALS
~
+blank+l%WLS
~
~
*1%+1%n4LS
-X-2%+1%WLS
-+-S%+l%MLS
Figure 11. Color change versus irradiation time (hours) in an Atlas
weatherometer for water-based acrylic paint films with the HALS
(Sanduvor 305 1) and nanoparticle M combinations.
Nan0 titanium dioxide
Pigmentary TiOz
Weight loss
10% wt PC500
20% Wt PCSOO
10% Wt PC105
20% Wt PC105
10% wt PCSO
20% wt PC50
10% wt Showa Denko
20% wt Showa Denko
10% wt AT1
20% wt AT1
20% Wt PCSOO
20% wt PC10.5
20% Wt PCSO
20% wt Showa Denko
20% wt AT1
None
Clear resin bank
RCL-696
RCL-696
RCL-696
RCL-696
RCL-696
RCL-696
RCL-696
RCL-696
RCL-696
RCL-696
None
None
None
None
None
RCL-696
None
19.0
66.5
31.0
62.8
30.4
39.0
77.0
10.5.4
16.6
43.2
97.6
128.7
121.4
146.8
138.7
4.1
5.4
zyxwvutsrqp
zyxwvutsrqp
zyxwvutsr
zyxwvutsrq
zyxwv
286 Norman S. Allen et a/.
2 100 1
(3
40
20
5
0
4
0
346
0
769
1143
$544
1887
2236
\
0
500
t
1000
~
1500
~
2000
-
'
2500
"
'
I
~
3000
~
~
IRRADIATION TIME. HRS
2851
IRRADIATION TIME, HRS
+Resin
no pigmen1
0 2 0 % PVC RUTILE 0
020% PVC RUTILE 0 + 10% ANATASE F
+20%
PVC RUTILE 0 + 20% ANATASE F
+20% PVC RUTILE 0 + 30% ANATASE F
zyxwvutsrqponm
zyxwvutsrqponm
zyxwvutsrqpo
U 20% PVC RUTILE 0 + 5% ANATASE F
-C
* 2% RUTlLE J+ 1% W
+20% PVC RUTILE 0 + 15% ANATASE F
2% RUTlLE J (13% ST)
-x-20% PVC RUTILE 0 + 25% ANATASE F
-x-20% PVC RUTILE 0 + 35% ANATASE F
"2%
-*t5%RUTlLEJ+l%HPLS
-+- 1% HALS + 2% ABSORBER
+2% ANATASE K (20% ST)
+5% ANATASE K
ABSORBER
+-1% RUTILE
+Z%
ANATASE K + 1% HALS
5% AN4TASE K + 1% HALS
-
Figure 13. Weight loss (mg/100 cm2) versus irradiation time in an Atlas
weatherometer for a series of isocyanate acrylic clear overbase auto finish
varnishes containing rutile J and anatase K nanoparticles with and without
HAW.
a strong sensitizing effect, again with the HALS offering no
increased stability. The interesting feature of this data is the poor
performance of the absorber and HALS combination. Mass loss for
same samples is shown in Fig. 13. The same contrasting effects are
observed between anatase and mile nanoparticles, with the latter
offering stability compared with that of the control film in all cases.
In this case, the HALS offers slightly increased performance with
rutile J. The anatase nanoparticles are strongly photocatalytic
showing considerable mass loss after 300 h of irradiation in the
Atlas weatherometer.
Self-cleaning paints
Figure 15. Mass loss versus irradiation time in a QUV weatherometer for
a DSM siliconized polyester resin with 20% wtiwt rutile pigment 0 plus
increasing levels of 5%, lo%, 15%, 20%, 25%, 30% and 35% wt/wt of
nanoparticle anatase F.
Compound) it produces smog and contributes to acid rain causing
damage to buildings.
In terms of self-cleaning paints, the idea is to limit the oxidation
and chalking of the paint film to the very near surface layers such
that over time with weathering rain water will wash the top layer
leaving an underlying clean fresh surface. In this regard, mixtures
of pigmentary mtile 0 and nanoparticle anatase F pigments appear
to provide the best option, with the former inducing some level of
base stability, whereas the presence of the latter gives rise
to surface activity. Figures 14 and 15 illustrate this effect for
a siliconized polyester coating exposed in a QUV weatherometer
for gloss and mass loss, respectively. Gloss loss is observed to be
gradually reduced with time; the effect increasing with increasing
loading of anatase nanoparticle F. Mass loss is also observed to
increase gradually with increasing levels of the same nanoparticle.
In this case, it is evident that only low levels of shedding and/or
chalking occur with time such that the paint film retains some level
of durability except for the very near surface layer. A similar but
perhaps more extreme effect is shown in Table 2 for a Lumiflon
fluorinated acrylic paint film. At 10% and 20% concentrations of
the nanoparticles G, F, E and H, chalking is quite high, whereas the
pigmentary rutile 0 at 20% wt/wt only gives a 4.7 mass loss value.
The pigmentary uncoated anatase A is also an option giving high
levels of chalking at 10% and 20% wt/wt. Thus, control of pigment
type and particle size as well as their concentrations is a critical
zyxwvutsrqpo
Environmental issues also play an important role in the
applications of titania fillers. These include the use of their
photocatalytic behavior in the development of self-cleaning
surfaces for buildings i.e. antisoiling and antifungal growth and
NOx reduction (emissions) (32). The latter can cause lung damage
by lowering resistance to diseases such as influenza and
pneumonia, whereas in combination with VOC (Volatile Organic
I
020% PVC RUTILE 0
U Z O % PVC RUTlLE 0 f
5% AAATME F
Table 3. % weight lost after 567 hours Atlas exposure. Various polymers
plus 5% Anatase sol 10-20 nm
02OXPYCRUTlLEO+
10% AAATASE F
+IOXPVCRUTILEO+
PVC RUTlLE 0 f
20% IUIATME F
~
Details
% Wt loss
Styrene acrylic
Styrene acrylic Anatase sol
PVA copolymer
PVA copolymer Anatase sol
Acrylic copolymer
Acrylic copolymer Anatase sol
Polysiloxane BS45
Polysiloxane BS45 + Anatase sol
12.2
97.3
11.4
97.9
1.4
101.0
23.3
13.6
-205
-X-IO% PVC RUTlLE 0
25% ANLITME C
I
+20%
PVC RUTILE 0 1
30% M A T M E F
-X-20%PVCRUllLEOr
0
540
lo00
1500
2000
2500
3000
irradiationtime (h)
Figure 14. Percentage concentration of NO, removed versus the
concentration of anatase sol-gel particles (1G20 nmj at 5 % wt/wt in
a polysiloxane Wacker BS 45 paint system.
+
+
+
zyxwvutsrqponml
zyxwvutsrq
zyxwvutsrqpon
zyxwvutsrqpo
zyxwvutsrqpo
zyxwvutsrqponm
zyxwvutsrqpon
zyxwvutsrqponm
Photochemistry and Photobiology, 2005, 81 287
7000
I
Q
5000
E
4000
* 3000
g
14 E
l2
10
E 8
p ! 6
+UV LIGHT
tANATASE G
+ANATASE G WASHED
+DEGUSSA P-25
+ANATASE
*d
I0
A
20
0
40
60
80
Ti02 PVC
>
20
30
50
60
IRRADIATION TIME, MlNS
76
Figure 16. Number of E . coli colony forming units versus irradiation time
(minutes) in the presence of pigmentary anatase A and nanoparticle anatase
G (washed and unwashed) as well as Degussa nanoparticle P-25.
area of development for effective self-cleanable paint surfaces, the
effect varying also with paint formulation.
Antibacterial effect
The ability of the nanoparticles to destroy bacteria is also
demonstrated in Fig. 16. In this study, E. coli are used, where
their destruction (measured in terms of colony forming units) after
irradiating with UV light in the presence of the titania particles is
plotted against irradiation time. The UV light itself has little effect
on the bacteria, whereas the pigmentary grade of anatase A has
a small effect and the nanoparticle G has a somewhat greater effect.
However, the most interesting feature of this data is the very high
destructive effect of the mixed-phase nanoparticle grade made by
Degussa (P-25). This nanoparticle grade of titania is well
established in the literature in terms of its high photoactivity
(32). In this study, a grade of nanoparticle anatase G was prepared
in the laboratory whereby the particles were seeded from solution
and then dried but not subsequently oven fired. This so-called
washed form of titania is observed in the data to be higher in
activity than the Degussa material. This effect is currently being
investigated further in terns of hydroxyl content and hydrogen
peroxide generation.
E
Figure 18. Percentage concentration of NOx removed versus the
concentration of anatase sol-gel particles (1G20 nm) at 5% wt/wt in
a polysiloxane Wacker BS 45 paint system.
NOx removal
In this part of the research work, it was important to be able to
develop coatings that remove NOx, VOC and potentially ozone.
The coating should in this regard be durable and show little or no
loss in activity with aging as well as having the ability to inactivate
nitric acid reaction products. Also, as above, it should be selfcleaning. The coating needs to be translucent so that existing
coatings or stonework can be overcoated without any change in
appearance. To some extent the coating must be photoresistant to
the effects of the nano TiO, and would probably need to be porous
to allow contact between the TiOz surface and the NO,. Nan0 TiO,
is an excellent scatterer of light, and if the coating is porous, this
further increases light scatter. Some potential problems in the
design of such coatings have been circumvented such as poor
adhesion and poor durability. Also, the nitric acid formed in the
reaction could damage the substrate or poison the NOx reaction. A
suitable test method was developed to measure the efficacy of the
coatings studied via a “Signal” detection system.
In this study, the effectiveness of three commercial products E-,
F- and G nanograde anatase particles were evaluated. Also, the
paints chosen must be stable to flocculation and viscosity changes,
cure or dry at ambient temperature and ideally water based to avoid
further environmental problems. Most polymers are carbon based
and are unlikely to be photoresistant, but water-based acrylic latex
paints have been evaluated. In the first instance, four types of
zyxwvu
c 40
0
5 30
0
2
20
X
g
s
10
0
0
0
100
200
300
400
Surface Area m‘lg
Figure 17. Percentage concentration of NOx removed versus the surface
area of anatase sol-gel particles at 5% wt/wt in a polysiloxane Wacker BS
45 paint system.
5
10
15
% T i 4 Volume
I
0%caC03
2.5%CaC03 A ~ % c ~ c o ~ I
Figure 19. Percentage NOx reduction versus volume of titania (anatase
10-20 nm) for a polysiloxane BS 45 paint substrate with 0%, 2.5% and
5.0% wt/wt of nanoparticle calcium carbonate.
zyxwvutsrqp
zyxwvutsrq
6olzyxwvutsrqponmlk
288 Norman S. Allen et a/.
U 10:2.5
50
Table 4. Percentage reduction in NOx gases during irradiation for
zyxwvu
Poiysiloxane paint systems with and without 5% w/w anatase sol particles
in comparison with sol particles alone.
-h- 5.5
NOx reduction by percentage
BS 45 Latex
BS 45 Latex
Sol
+ 5 % Sol
?hNO
% NO2
0
84.9
84.9
0
9.3
55.8
zyxwvutsrqpo
zyxwvutsrqpo
zyxwvutsr
NO, reduction as pg/m2s
J
250
500
750
1000
1250
EXPOSURETIME (hrs)
1500
NO pg/m2s
NO2 W m Z s
0.000
0.060
0.320
0.000
0.055
0.409
1750
Figure 20. Percentage weight loss versus exposure time in an Atlas Ci65
weatherometer for polysiloxane paint films (BS 45) containing different
ratios of nano anatase (10-20 nm) sol-gel titania (5/7.5/10):calcinm
carbonate (0/2.5/ 1O)particles.
BS 45 Latex
BS 45 Latex
Sol
+ 5% Sol
concentration. Porosity can also be introduced by using other
materials other than TiOz. Nan0 calcium carbonate offered the
possibility of high translucency and the ability to react with nitric
acid. The results are confirmed in Fig. 19 where it is observed that
NOx is reduced not only with increasing titania doping but also
with increasing levels of calcium carbonate addition.
The most interesting feature of the results, however, is the
influence of titania and calcium carbonate loading on the extent of
degradation of the polysiloxane paint films, as measured by percent
weight loss. The data shown in Fig. 20 shows that in the absence of
calcium carbonate the extent of degradation is low, as indicated
above, whereas in its presence the rate of degradation increases
with concentration from 2.5% to 5.0% by weight. At 10% by
weight of titania, the extent of degradation is significant in the
presence of the calcium carbonate. In this case, the access of both
moisture and oxygen through the film matrix will be enhanced.
Film translucency also decreases with increasing loadings of titania
and calcium carbonate particles, as shown by the data in contrast
ratio in Fig. 21.
Measurements on NOx reductions have also been obtained in
terms of NO and NOz gases where it is observed that the rate of
NOx destruction is clearly greater in the presence of the
nanoparticles alone, whereas the paint matrix gives rise to a barrier
effect as might be expected (Table 4). Nevertheless, the efficacy of
the paint films in destroying the NOx gases is high.
zyxwvutsrqpo
acrylic water-based paints were evaluated in terms of relative
stability toward photoactive nanoparticles. In this study, a special
sol-gel grade of anatase was prepared in the laboratory with no
postfiring. Particles of varying sizes were also prepared via this
route. The relative paint stabilities with and without the anatase sol
particles (10-20 nm) at 5% wt/wt are shown in Table 3 after 567 h
of weathering. Of these paint formulations only the Polysiloxane
BS 45 (Wacker) proved to be resistant to the photocatalytic effects
of the titania particles. The styrene-acrylic, poly(viny1 acetate) and
acrylic copolymers all showed high degrees of chalking (weight
loss).
Commercial dry nano TiOz products with a range of particle size
and surface area were available from Millennium, with surface
areas ranging from 20 to 300 m2/g for evaluation.
Even with the smallest crystallite size, it is difficult to eliminate
light scattering at levels above 5% at conventional coatings
thickness (25 pm) because of aggregation. With special nondried
sol-gel nano Ti02, there is less light scattering because of reduced
particle aggregation. It appeared that the coatings had to be porous
before there was a significant activity toward NOx reduction. From
the data in Figs. 17 and 18, the efficacy of NOx removal increases
significantly with both an increase in particle surface area and
CONCLUSIONS
0
5
10
15
% TiOz Volume
Figure 21. Translucency (contrast ratio) for BS 45 paint films with volume
addition if sol-gel anatase titania (10-20 nm) particies versus percent
calcium carbonate addition.
Photooxidation studies on paint films show a clear demarcation
between nanoparticle and pigmentary-grade titanium dioxide, with
the former being more active. Model system studies based on 2propanol oxidation and hydroxyl analysis go some way to
predicting pigment activities, but precise correlations do not exist.
On the other hand, the real-time microwave cavity perturbation
method can be used as a gauge of photoactivity in titanium dioxide
pigments. This, we believe, is a useful application, which could
form the basis of an effective characterization technique for
commercial systems. Using this method, a photoconductor may be
continuously irradiated with visible or UV polychromatic light and
the resulting changes to the electric susceptibility of the sample
monitored in terms of its microwave response. This approach
zy
zyxwvutsrqpo
zy
zyxwvuts
Photochemistry and Photobiology, 2005, 81 289
provides an extended timescale probe of charge-canier dynamics in
photoconductors, complementary to the numerous time-resolved
studies published to date (25,26).In this study, carrier generation
and recombination processes can be related to the effects of surface
treatments and milling operations.
The high UV opacity of the nanoparticle rutile form of titanium
dioxide, coupled with its lower photoactivity due to an effective
coating system, has been shown to be an effective UV protector
for acrylic coating applications. Anatase also opacifies but to
a lesser extent above 380 nm and operates primarily as an
effective photocatalyst. Mass loss and color change data show that
anatase is a photosensitizer although the effect is dependent on
surface modifications, whereas rutile is an effective stabilizer that
in some cases offered superior performance to the organic UV
absorbers and HALS. Anatase in combination with HALS was
found to be strongly antagonistic, whereas no significant synergy
was observed between rutile and HALS. The outstanding
performance of both anatase and rutile nanoparticles is visibly
and colorimetrically evident on clear acrylic wood coatings, with
rutile being the more effective. In terms of color stability, the 70
nm particles are more effective than the 90 nm particles. This
trend is, however, reversed for chemical and structural changes in
the film. In this study, the difference may be because of the color
bleaching effect on the lignin products through the photosensitizing activity of the anatase. In general, the data is clearly of
commercial significance for many coatings applications, where
translucent fillers of this type offer significant cost-effective
benefits relative to conventional organic UV absorbers and HALS
in terms of UV stabilization activity. The use of nanoparticle
anatase in conjunction with pigmentary rutile grades is also
a viable option for the development of self-cleaning paint
surfaces. For antibacterial surfaces, nanoparticles are effective,
whereas pigmentaq grades are ineffective. Highly effective
photocatalytic grades of nanoparticles also can be prepared
through control of the preparation and subsequent drying
operations.
The paint coatings are also active to NOx, particularly once
irradiated with UV with high levels of Ti02 and CaC03 enhancing
activity. This effect is associated with increased porosity of the
paint system induced by both the titania and calcium carbonate
particles. Unfortunately, higher levels of Ti02 and CaC03 impart
lower durability to the paint matrix. Higher levels of TiOz and
CaC03 also reduce translucency of the paint films thus increasing
absorptivity. On a positive note, higher levels of CaC03 would
react with more HNO?.
4. Allen, N. S., J. F. McKellar and D. Wilson (1977) Luminescence and
degradation of nylon polymers. Part 11: quenching of fluorescent and
phosphorescent species. J . Photochem. 7, 3 19-324.
5. Allen, N. S., D. J. Bullen and J. F. McKellar (1979) Luminescence
properties and photoactivity of sulphate processed rutile (titanium
dioxide) pigments in commercial polyethylene. J . Muter. Sci. 14,
1941-1944.
6. Day, R. E. (1990) The role of titanium dioxide pigments in the
degradation and stabilisation of polymers in the plastics industry,
Polym. Degrad. Stab. 29, 73-92.
7. Allen, N. S., J. F. McKellar and D. G. M. Wood (1975) Photochemical
processes involving titanium dioxide pigments and titanium dioxidelight stabiliser combinations in commercial polyolefins. J . Polym. Sci.,
Polym. Chem. Ed. 13, 2319-2325.
8. Allen, N. S., J. Luc Gardette and J. Lemaire (1982) Interaction
between titanium dioxide pigments and hindered piperidine/antioxidant combinations in the photostabilisation of polypropylene. Dyes
Pigm. 3, 295-305.
9. Fujishima, A., T. N. Rao and D. A. Tryk (2000) Titanium dioxide
photocatalysis. J . Photochem. Photohiol., Rev. Ed. 1, 1-21,
10. Bryk, M. T. (1991) Degradation of Filled Polymers. Ellis Horwood,
London.
11. Rabek, J. F. (1990) Photostabilisation of Polymers: Principles and
Applicarions. Elsevier Applied Science Publishers, London,
12. Iler, R. K. (1979) The Chemistry ofSilica. John Wiley and Sons, New
York.
13. Hill, G. R. and R. J. Martinovich (1972) The Effect of Pigments on the
Weatherability of HDPE. Technical. Paper. Reg. Techti. Cotzf. Soc.
Plast. Eng., Phil. Sect. 2-3 October, 29.
14. Pappas, S. P. and W. Kuhhirt (1975) Photochemistry of pigments.
J . Paint Technol. 47, 42-47.
15. Sullivan, W. F. (1972) Weatherability of titanium dioxide containing
paints. Prog. Org. Coat. 1, 157-167.
16. Lemaire, J. (1982) The photocatalysed oxidation of polyamides and
polyolefins. Pure Appl. Chem. 54, 1667-1682.
17. Allen, N. S. and M. Edge (1992) Fundameniuls of Polymer
Degradation and Stabilisation. Chapman and Hall, Chichester, UK.
18. Allen, N. S., H. Katami and F. Thompson (1992) Influence of titanium
dioxide pigments on the thermal and photochemical oxidation of low
density polyethylene film. Eur. Polym. J . 28, 817-822.
19. Allen, N. S. and H. Katami (1996) Comparison of various thermal
and photoageing conditions on the oxidation of titanium dioxide
pigmented linear low density polyethylene films. Polym. Degrad.
Stab. 52, 3 11-320.
20. Filimonov, V. N. (1964) Photocatalytic oxidation of gaseous isopropanol on zinc and titanium oxides. Dokl. Akad. Nauk SSSR 154,
151-1 54.
21. Egerton, G. S. (1964) Photosensitising properties of dyes and pigments.
Nature 204, 1153-1 155.
22. Cundall, R. C., R. Rudham and M. S. Salim (1976) Photocatalytic
oxidation of propan-2-01 in the liquid phase by rutile. J . Chem. Soc.
Faraday 72, 1642-1650.
23. Ilenko, V. S., A. V. Uvanov and V. I. Milko (1973) Determining the
photochemical activity of titanium dioxide pigments by EPR. Zuv.
Laboratoria 9, 1089-1095.
24. Allen, N. S., J. F. McKellar, G. 0. Phillips and C. B. Chapman
(1974) The titanium dioxide sensitised degradation of nylon 6,6:
stabilising action of manganese ions. J . Polym. Sci., Polym. Chem.
Ed. 12, 723-727.
25. Boonstra, A. H and C. A. H. A. Mustaers (1973) Relation between the
photoadsorption of oxygen and the number of hydroxyl groups on
a titanium dioxide surface. J . Phys. Chem. 79, 1694-1698.
26. Allen, N. S., Y. S. Chow, F. Thompson, T. S. Jewitt and M. R.
Hornby (1991) The use of photoconductive and microwave
techniques for the evaluation of titanium dioxide pigments. Polym.
Degrad. Stab. 32, 119-129.
27. Allen, N. S., M. Edge, 3. H. He and F. Thompson (1992)
Electrophotographic and microwave photodielectric studies (I) effect
of various transition metal dopants on titanium dioxide pigments in the
solid-state. Dyes Pigm. 20, 21 1-226.
28. Allen, N. S., K. 0. Fatinikun and T. J. Henman (1983) Thermal and
photochemical oxidation of polypropylene: influence of residual
catalyst levels in unstabilised diluent and gas phase polymers. Eur.
Polym. J . 19, 551-554.
zyxwvutsrqpo
zyxwvutsrq
zyxwvutsrqpon
zyxwvutsrqp
Acknowledgements-The authors thank Millenium Inorganic Chemicals
for supplying the pigments and durability data used in this program of
work and also for partial financial support of G.S.
REFERENCES
1. Allen, N. S. (1983) Effects of dyes and pigments. In Degradation and
Stabilisation of Polyolefins, Chap. 8 (Edited by N. S. Allen), p. 337.
Elsevier Science Publishers Ltd., London.
2. Allen, N. S. and J. F. McKellar (1980) Photochemistry of Dyed and
Pigmented Polymers., Applied Science Publishers Ltd. London.
3. Laurenson, P., R. Amaud, J. Lemaire, J. Quemner and G. Roche (1978)
Photothermal degradation of polyethylene films. 11: action of photocatalytic metallic oxides. Eur. Polym. J . 14, 129-137.
zyxwvutsrqp
zyxwvutsr
zyxwvutsrqp
zyxwvutsrq
zyxwvutsrqponm
zyxwvutsr
zyxwvutsr
zyxwv
290 Norman S. Allen ef a/.
29. Allen, N. S., M. Edge, T. Corrales, M. Shah, D. Holdsworth, F.
Catalina, C. Peinado and E. P. Collar (1996) Thermal and
photooxidation of polypropylene: influence of long-tern ambient
oxidation by spectroscopic, thermal and light scattering studies.
Polymer 31, 2323.
30. Clayton, J. (1994) Ultrafine titanium dioxide for wood finishes. Am.
Paint Coat. J., 49-55.
31. Vignolo, C. E. (1995) Some applications of ultrafine titanium dioxide.
Eur Coat. J . Part 5 , 359.
32. Haacke, G., E. Longordo, J. S. Brinen, F. F. Adrawes B. H. Campbell
(1999) Chemisorption and physical adsorption of light stabilisers on
pigment and ultrafine particles in coatings. J . Coat. Technol. 71, 87.
33. Kaneko, M. and I. Okura, Eds. (2003) Photocatalysis: Science and
Technology. Springer-Verlag GmbH & Co. KG, Heidelburg, Germany.