Low cost anti-soiling coatings for CSP collector mirrors and heliostats

Scott R. Hunter*a, D. Barton Smitha, Georgios Polizosa, Daniel A. Schaeffera, Dominic F. Leea,
Panos G. Datskosa
a
Oak Ridge National Laboratory, 1 Bethel Valley Rd., Oak Ridge, TN, USA 37831

ABSTRACT

Most concentrating solar power (CSP) facilities in the USA are located in the desert southwest where open land and
sunshine are abundant, but airborne dust is prevalent. The accumulation of dust, sand and other natural pollutants on
collector mirrors and heliostats presents a significant operational problem and M&O cost for the CSP facilities in this
region. The optical performance of the CSP collectors is key to achieving low electricity costs, where a 1% decrease in
reflectance directly leads to a 1% increase in the levelized cost of electricity (LCOE) generated by these facilities.
In this paper we describe the development of low cost, easy to apply anti-soiling coatings based on superhydrophobic
(SH) functionalized nano silica materials and polymer binders that posses the key requirements necessary to inhibit
particulate deposition on, and adhesion to, CSP mirror surfaces, and thereby significantly reducing mirror cleaning costs
and facility downtime. The key requirements for these coatings are excellent optical clarity with minimal diffuse
reflectance, and coating mechanical and exposure durability in harsh desert environments while maintaining SH and dirt
shedding properties. The coatings developed to date have excellent SH properties with water contact angles > 1650 and
rolling angles < 50. The solar weighted optical reflectance of the anti-soiling coating over the wavelength range 250 nm
to 3µm is >99% that of uncoated mirror surfaces with coating diffuse reflectance being <1% over this wavelength range.
Ongoing mechanical and accelerated solar UVA exposures also indicate these coatings will meet the required durability
goals.
Keywords: superhydrophobic, anti-soiling, optical coatings, CSP, heliostats, collectors

1. INTRODUCTION
Concentrated solar power (CSP) electricity generation is emerging as one of the most efficient, reliable and cost
competitive means for the conversion of sunlight into electricity1. However, the achievement of the DOE Sunshot
program goal of reducing the total installed cost for utility-scale solar electricity generation using CSP technologies to
roughly $.06/kWh by 2020 without subsidies, will require new, and in some cases, highly innovative technical
developments to enable rapid cost reductions in component materials, manufacturing techniques and plant operating and
maintenance procedures2. Baseline levelized cost of electricity (LCOE) generated by CSP facilities, without economic
support, was approximately 21¢/kWh in 2010 and has been reduced to 13 ¢/kWh in 20133. However, maintenance and
operating (M&O) costs remain around 4¢/kWh3. As CSP facility manufacturing, installation and operating costs are
reduced, system maintenance costs assume an ever increasing percentage of the cost of electricity. Consequently, to
achieve an LCOE of 6¢/kWh will require M&O costs of < 1¢/kWh3.
One of the most significant maintenance problems and cost
associated with CSP solar collectors is the soiling of the first
surface of the solar radiation reflectors by the accumulation of
sand, dust and other polutants4. According to Solar Power
World, “a dust layer of one-seventh of an ounce per square
yard decreases solar power conversion by 40 percent. In
Arizona, dust is deposited each month at about 4 times that
amount.”5. Participants from solar and glass industries at the
DOE workshop “Specialty glass needs for the U.S. solar
industry”6, prioritized the factors that affect the development
and availability of high-performance solar glass coatings, and
“self-cleaning coatings” ranked second among the key
technical needs for coatings. CSP electricity generation Figure 1. Dust buildup on the heliostats at the Ivanpah
facilities are, by necessity, located in the southwest region of CSP facility in the Mojave Desert in California.
*Email: [email protected]; Telephone: (865) 576-4422; Fax (865) 574-9407
High and Low Concentrator Systems for Solar Energy Applications IX, edited by Adam P. Plesniak, Candace Pfefferkorn,
Proc. of SPIE Vol. 9175, 91750J · © 2014 SPIE · CCC code: 0277-786X/14/$18 · doi: 10.1117/12.2061845

Proc. of SPIE Vol. 9175 91750J-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 the U.S. in semi-arid and arid areas where there is abundant sunlight and land available, but airborne sand and dust are
very prevalent, and clean water to periodically clean the reflector surfaces is very scarce and expensive. The optical
performance of the CSP collectors is key to achieving low electricity costs, where a 1% increase in reflectance directly
leads to a 1% reduction in LCOE2. Figure 1 shows the dust loading (average mirror reflectivity < 85%) on a heliostat
mirror at a CSP facility located in the Mojave Desert in California just before the mirrors in this field were to be hand-
washed using a brush and deionized water. Self-cleaning (anti-soiling) and anti-reflection (AR) first surface reflector
coatings are a viable solution to achieving high performance, low maintenance cost CSP electricity generation.
The objectives of our DOE Sunshot funded project are to develop, test and implement low-cost durable multifunctional
(self-cleaning and anti-reflecting) nanostructured collector surface coatings that significantly enhance the reliability and
efficiency of CSP collectors, while drastically reducing collector cleaning and maintenance costs. In this paper, we
describe the development of low cost, easy to apply anti-soiling coatings based on superhydrophobic (SH) functionalized
nano silica materials and polymer binders that posses the key requirements necessary to inhibit particulate deposition on,
and sticking to, CSP mirror surfaces, and thereby significantly reducing mirror cleaning costs and facility downtime.
The key requirements for these coatings are excellent optical clarity with minimal diffuse reflectance, and coating
mechanical and exposure durability in harsh desert environments while maintaining SH and dirt shedding properties.

2. TECHNICAL APPROACH
The Nanosystems and Structures Research Group at the
Oak Ridge National Laboratory (ORNL) has been (b)
developing a suite of highly water repellent SH coatings for
(a)
a range of applications over the past several years. The
prior work of the Group on the development of optically
transparent SH coatings has been summarized in previous
presentations7-9. Water repellency is usually determined by
measuring the water contact angle (CA, θo – Figure 2a), the
rolling angle (RA) and the coating surface coverage. When Droplet on normal hydrophobic surface Droplet on superhydrophobic surface
the CA is < 900 the surface is hydrophilic, when it is > 900
Figure 2. (a) Water droplets on a low surface energy
the surface is hydrophobic, and when it is > 1500 the surface
hydrophobic surface, and (b) a nanostructured super-
is superhydrophobic. Surface coatings can routinely be hydrophobic surface.
fabricated with CAs > 1700, RAs < 0.50 and with surface
coverage close to 100%7. With very SH surfaces, not only
does water not stick to these surfaces, but the forces binding
dust and other particulates to the surfaces are also reduced
such that it is easy to remove these materials from the
surface as shown schematically in Figure 3.
Figure
Surface 3. Schematic showing dirt
war. particles remaining on a
Fig. 4
roughness and selfcleanlng by rinsing with
normal surface (left) and being collected by a water drop on
a superhydrophobic surface (right).

The main difficulty with essentially all SH surfaces is their

poor mechanical durability, which is due to the inherent
nature of a SH surface. Coatings can be fabricated with low
surface energies (such as fluorinated polymer coatings), but
the maximum CA for these surfaces is in the 110-1150 range
and they are therefore not SH in nature. To fabricate a SH
surface, the surface must have a low surface energy and be
structured on the nano to micro scale range to minimize the
water droplet surface contact area with the underlying surface
(Figure 2b). Figures 4b and 4c are SEM images of an
optically transparent SH coating at different magnifications
showing the surface coverage and particle sizes, with film
Figure 4. (a) Glass slides spray-coated with transparent self- thicknesses of ≈ 100-300 nm. The uniform density and
cleaning and anti-reflecting silica nanoparticles and SEM texture of the coating is evident on the micron (Figure 4b) as
images at low (b) and high (c) magnifications. (d) The water
0 well as in the nanometer (Figure 4c) size scales. The
contact angle of the coated surfaces ≈ 175 .

Proc. of SPIE Vol. 9175 91750J-2

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 multifunctional nature of the nanoparticulate surface coating derives from the modification of the high surface-area
nano-structure with self-assembly of low surface energy monolayers. Specifically, the combination of surface nano-
roughness and surface functionalization with hydrophobic organic monolayers, enhances the hydrophobicity and results
in surfaces with CA values ≈ 1750 (Figure 4d) across the whole surface. Moreover, the particles do not scatter light at
wavelengths > 300 nm because of their small dimensions (≈ 50-100 nm, i.e. much smaller than the wavelength of the
incident solar radiation) and consequently, the coatings are transparent with good optical transmission across the entire
solar spectrum (280 nm – 2.5 microns).
This coating was made from relatively inexpensive fluorinated silica nanoparticles, off-the-self polyurethane binders and
environmentally friendly solvents (mainly IPA and a few percent acetone), and applied using standard paint industry
practices and paint deposition techniques. The major problem with the coatings that were fabricated early in the project
is that although they possessed excellent SH properties and good optical transmission, they were not mechanically
durable as measured using standard Taber rub tests (to be described later). The Taber rub test result for the first
fabricated coating is shown in Figure 14 and is labeled C1. The coating was able to withstand only one rub cycle before
the CA of the coating fell to an unacceptable value.
A systematic study using design of experiments (DOE) techniques was initiated to understand the coating durability
failure mechanisms and devise coating formulations and techniques to enhance the durability of the fabricated coatings.
The main variables in this study were the silica particles (particle type, size, size distribution and concentration), particle
functionalization (paraffinic and fluorocarbon silanes), binders (various polymer and epoxy clearcoats and resins) and
EPA compliant solvents. Coatings fabricated in the study were routinely subjected to a number of tests with the goal of
optimizing the three main requirements of the coatings in this application, namely:
(1) Maintenance of the water repellent, and hence the anti-soiling properties, of the coatings
(2) Maintenance of the optical transmission and specular reflectance properties of the second surface coated
heliostat mirrors
(3) Improve the mechanical and environmental durability of the coatings such that they would be able to survive on
the heliostats and collector mirrors for 10 to 20 years in the field without significant degradation in the optical
performance of the mirrors.
2.1 Nano Silica Particles

In-house colloidal silica (SiO2) synthesis

Scale65bar
nm = 100particles
silica nm Scale225
barnm= silica
1 µmparticles Scale425
barnm= silica
200 particles
nm
EHT= 5.00 kV Signal A =SE2 Date :i 5 Jan 2013 ZEISS

Particle Size - 65 nm Particle Size - 225 nm Particle - Size 425 nm

WD =11.6 mm Mag= 30.00 K File Name = 204äf

Figure 5. SEM images of the spherical silica nanoparticles fabricated Figure 6. Highly agglomerated
using an in-house procedure with very uniform particle size fumed silica with individual
distributions over the range from 65 to 425 nm. particles in the 10-20 nm size range.
The anti-soiling coatings fabricated in this study were based on formulations consisting of nanometer sized silica
particles embedded in a polymer matrix, as previous studies had shown that nanosilica based coatings offered the best
route to the formation of reasonably rugged SH coatings10-11. Several types of functionalized and unfunctionalized silica
particles were either purchased or fabricated in-house for this study, including commercially available Aerosil, Aerogel,
colloidal silica and fumed silica from various suppliers. The prime requirement on the size of the particles in this study
was that they be in the size range 10-100 nm, such that they would not optically scatter the UV solar radiation at
wavelengths greater than 280 nm. Previous studies have shown this to be the optimum feature size range for highly
water repellent and condensation resistant coatings with excellent optical transparency and anti-reflection surface
properties12-14.
Figure 5 shows SEM images of three batches of the in-house fabricated spherical silica particles with tight particle size
distributions over the size range from 65 nm to 425 nm. Silica nanoparticles have been fabricated with average sizes

Proc. of SPIE Vol. 9175 91750J-3

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 from approximately 12 nm up to several hundred nm with those fabricated in the 12-100 nm size range investigated in
this study. Although the individual particles in the purchased fumed silica samples are typically quoted with 10-15 nm
particle sizes, actual silica powders are usually highly agglomerated (Figure 6) and must be broken down into smaller
particle size distributions (using high shear mixing and sonication techniques) if the fabricated coatings are not to scatter
the shorter wavelength solar radiation. The most successful anti-soiling coatings were fabricated with a combination of
Aerogel and either the in-house or colloidal silica nanoparticles. It proved difficult to breakup and remove enough of the
larger agglomerated silica in the fumed silica materials to prevent significant optical scattering at the shorter solar
wavelengths.
2.2 Silica Particle Functionalization
Silica nanoparticles are naturally hydrophilic in nature and must be
functionalized by covalently bonding low surface energy self-assembled
monolayers (SAM) onto the surface of the silica nanoparticle. Various paraffinic
and fluorocarbon silanes were examined in this study with the fluoro-silane
functionalized nanosilica particles exhibiting very SH properties with water
contact angles up to 1750. Figure 7 shows a schematic of a silica particle
functionalized using (heptadecafluorochloro-1,1,2,2-tetrahydrodecyl)trichloro-
silane, producing a particle with a very low surface energy. To improve the
bonding of the silica particles in the polymer binder matrix, double silanation
studies are in progress to add hydrophilic coupling agents to the silica particles to
enhance particle bonding to the polymer binders. The hydrophilic agents can
covalently bond to the polymer, considerably enhancing the silica particle
binding to the polymer and improving the mechanical durability of the anti-
Figure 7. Fluorosilane functionalized
soiling coating without significantly compromising the water repellency, and silica particle.
hence anti-soiling properties, of the coatings.
2.3 Polymer Binder Studies
Table 1. Clearcoat Binders and the Concentration Ranges used in this Study
Concentration (wt%)
Low High
Polymer binder 1 0.15 3
Polymer binder 2 4 100
RTV615 0.4 1.3
PDMS 0.1 15
Teflon-AF 1.0 1.0
Several commercially available polymer and epoxy based clearcoat binders were studied with the goal of finding a
binding agent that was optically clear over the entire solar weighted spectrum but also had improved mechanical and
environmental durability in comparison with the initial Teflon-AF binder used to fabricate the C1 coating (Figure 14).
The binders used in this study were mainly polyurethane or polysiloxane based clearcoats used primarily for automotive
and military applications, and contain UV inhibitors to reduce environmental degradation of the coatings. They are
listed in Table 1 along with the range of concentrations used in the mixture studies, where the goal was to optimize the
optical clarity, mechanical durability and SH properties of the resultant coatings.
2.4 EPA Compliant Solvent Studies
Table 2. EPA Compliant Solvents and Concentration Ranges used in this Study
Concentration (wt%)
Low High
MEK 97 99.9
Acetone 32 99.9
IPA 80 99.9
4-Chlorobenzotrifluoride 4 10
Toluene 45 99
FC-40 98 99

Proc. of SPIE Vol. 9175 91750J-4

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 One of the major concerns with spray coating CSP mirrors onsite is emission of VOCs in the ecologically sensitive
environments where a number of these facilities are located. Consequently, we limited our choice of solvents to those
that have minimal health hazards, and are compliant with the product-weighted reactivity limits established in the
National Volatile Organic Compound (VOC) Emission Standards published by the Environmental Protection Agency
(EPA) in 40 CFR Part 59. These solvents are listed in Table 2 and were selected to not only dissolve the polymeric
binders, but also to be compatible with the hydrophobic monolayers used for the nanoparticle functionalization. Proper
solvent selection leads to the formation of homogeneous binder/nanoparticle suspensions that can be applied to a glass
surface and result in a uniform coating. The concentration limits in Table 2 represent the maximum and minimum
solvent concentrations required in order to mix the nanoparticles with the polymeric binder while maintaining their
ability to be deposited on the substrate. Design of experiments (DOE) techniques were used to categorize the most
promising silica nanoparticles and their functionalization, solvents and binders to optimize the durability, and the optical
transparency and anti-soiling properties of the coatings.

3. RESULTS AND DISCUSSION

The anti-soiling coating mixtures discussed above were initially applied to glass slides and small glass coupons using
spin and spray coating techniques to facilitate easy analysis of the resultant anti-soiling coatings. These coatings were
then subjected to a range of analyses to determine their durability, and optical and SH properties. These analyses
included surface characterization studies using AFM, SEM and surface profilometry, optical transmission and specular
and diffuse reflectance measurements over the entire solar spectrum, water contact and rolling angle measurements to
determine the SH and anti-soiling properties, Taber rub measurements to measure the coating durability, and accelerated
UV measurements to understand any possible environmental degradation mechanisms. Selected results from these
studies are shown in the next section, primarily for the best coating (C7) fabricated to date.

3.1 Surface Characterization

The most important requirement on the anti-soiling coatings in order to maintain excellent optical transparency, with
negligible diffuse reflectance, is for the particles in the coating to be relatively uniform in size with a particle size
distribution in the range 10-100 nm. But just as importantly, the particles must be uniformly distributed throughout the
coating, the coating must cover 100% of the surface and coating thickness and surface roughness must be minimized.
Several surface characterization techniques are used to optimize these properties. SEM and AFM imaging were used to
primarily understand surface roughness and particle distribution issues within the coatings.
A'lk...- .. (--". .
i
`r t ` . , 4)..71.
t" ' T ! ._
. .4 .
.

.-. . 4
.á
'
`
'
I- rr .X'. ."'"('''' ..7 .
y .....

\ ' .!'
..

;` t"
,f
'
ae,

1
Date :12 Aug 2013 100 nm EHT= 5.00 kV Signal A= lnLens Date :12 Aug 2013
Time :16:30:20 WD= 5.9 mm Mag= 150.00 K Time :16:39:06
.7., ëv +4r

Figure 8. SEM images at two different magnifications showing the size and distribution of the silica particles
in an anti-soiling coating.

Figure 8 shows typical SEM images of a glass slide coated with the C7 anti-soiling coating. Note the size of the silica
particles is in the range 20-100 nm, and the particles are uniformly distributed throughout the coating. AFM images of
the C7 coating are shown in Figure 9 and again show excellent surface coverage with an average surface roughness of 21
nm. Previous studies have shown that optimal surface SH and water repellency occurs when the surface roughness is in
the range 20-100 nm13. Earlier coating formulations resulted in much poorer surface coverage and increased surface
roughness, which were the result of particle agglomeration when the sprayed mixture dried on the substrate. Particle
agglomeration is the result of using thermodynamically immiscible functionalized nanoparticles and solvents, and leads

Proc. of SPIE Vol. 9175 91750J-5

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 to phase separating morphologies where the silica particles clump together on the surface. These coating formulations
were characterized by micron-size defects that led to detrimental effects on both the water contact angle measurements
and mechanical durability of the coatings. By contrast, coatings consisting of silica nanoparticles and compatible
solvents, used for their dispersion, exhibit excellent dispersion properties over large surface areas.
11.11.1 1a1a.re w.n.u..
s..
.... ....
a..r..
... r1.
M.
d ..... ..
1mar
i as
Y1
A.1 ...

Figure 9. AFM surface characterization of spray-coated C7 anti-soiling on a glass substrate surfaces at two
different magnifications. The AFM measurements reveal evenly distributed functionalized silica nanoparticles
and uniform coatings with average roughness of 21 nm.

Surface profilometry studies were also performed on these coatings over larger surface areas to understand surface
uniformity, surface coverage and any possible large particle surface contamination issues, which would have detrimental
effects on the specular reflectance of the surface. A KLA-Tencor Alpha-Step profilometer was used in these studies,
with typical 1.4 mm line scan across the C7 anti-soiling coating shown in Figure 10a, where the vertical scale is in
Angstroms and the horizontal scale is in microns. The line plot shows a very uniform surface with a 100% surface
coverage and a larger 20-30 nm silica particle projecting from the surface. Figure 10b is a histogram plot of the average
surface roughness from a large number of line plots, and shows that the average surface roughness for this coating was
22.5 nm, in excellent agreement with the AFM estimate shown in Figure 9 for the same coating.
50 100
Length = 1425.5 pm R=315.71 A Scale= 500 A
(a) A
(b)
300 40 80

200

80 ¿
100
Nrs,1/4..
GW
LL
o
>
j w Ñ
-100 - 1

7
> E
0 100 200 300 400 500 500 700 800 900 1000 1100 1200 1300 1400 pm U
10 zo

Figure 10. (a) A representative 1.4 mm line scan across the surface of a C7
anti-soiling coating, and (b) a surface roughness histogram from a number of
line scans showing that the average surface roughness was 22.5 nm. 100 150 200 250 300
o

Surface Roughness (A)

Coating depth profile studies were also performed on multi layer C7 anti-soiling coatings to determine the thickness of
the coatings, and to understand the location and distribution of the silica particles within the coatings. This information
is important in determining the improvement in abrasion resistance of a multi layered coating. The depth profiles
obtained from AFM measurements of one, two and three layer C7 coatings are shown in in Figure 11 (a), (b) and (c)
respectively, with the distributions being very narrow with average values and centered around approximately 80 nm for
all three layers. The average surface roughness Ra from these measurements was < 30nm in all cases, in excellent
agreement with the AFM and surface profilometry measurements given in Figures 9 and 10 on separate C7 coated
substrates. These measurements show that the superposition of alternating C7-CeraKote layers produces well defined
and very uniform structured layers that lead to the enhancement of the anti-reflection properties of the coatings shown in
Figure 12.

Proc. of SPIE Vol. 9175 91750J-6

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 Depth 1st C7-CK Depth 2nd C7 -CK Depth 3rd C7 -CK

6 i.'m 2 .áu a.'m d 0 0.75 1.50 2.00 4.00 0 0.50 1.00

CD rrrrrr inn Hist. i Hist. i Correlation

Figure 11. Depth profiles of the alternating C7-CK layers obtained from AFM measurements for one, two
and three layered anti-soiling coatings.

3.2 Optical Performance

A major goal of this project has been to develop anti-soiling coatings that do not significantly impact the optical
transmission and specular and diffuse reflectance properties of the mirrors on which the anti-soiling coatings are
deposited. The project benchmarks were (1) that the coatings would not decrease the solar weighted transmission
through the mirror glass by more than 1% and (2) that the coatings would not increase the diffuse reflectance from the
coated mirror surface by more than 1%. Optical transmission measurements over the ultraviolet-visible spectral region
(250 - 1100 nm) were performed to evaluate the optical clarity of the coatings. The measurements for a one layer C7
coating and for several multilayer C7 coatings are shown in Figure 12 (a) and (b) respectively in comparison with the
transmission through an uncoated glass substrate. Figure 12(a) shows that we have met the transmission requirements

100
(a) 93- (b)
.-. 80-
e. - Bare Glass
- Coating 1 92-
1v 60-
U
C
tß
p 40- 91
E
rn - Borosilicate Slide
-3 Layers
;c 12 20- 90 4 Layers
H -6 Layers
o - -9 Layers
89
460 660 800 1000 1200 460 660 860 1000
Wavelength (nm) Wavelength (nm)
Figure 12. (a) Optical transmission spectra of a C7 coated borosilicate glass substrate in comparison with an uncoated
glass substrate in the UV-Vis spectral region (250 nm – 1100 nm). The difference measurement in the inset shows a
maximum 0.5% loss in transmission due to the coating. (b) Multilayers coatings can increase the optical transmission
due to optical cavity and refractive index gradient effects in the coatings leading to significant anti-reflection properties.

Proc. of SPIE Vol. 9175 91750J-7

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 for the anti-soiling coating in that the maximum reduction in optical transmission occurs around 500 nm being ≈ 0.5%.
Figure 12(b) shows that the coating transmission actually increases with a multilayered coating (with the maximum
increase of ≈ 2% over the 400-600 nm range) occurring with 4-6 coating layers. Optical cavity and refractive index
gradient effects are responsible for the observed antireflection properties, and hence increased transmission of the
multilayer coatings. The optical transmission decreases with further increases in the number of layers of the C7 coating,
presumably due to the collapse of the coating structure and possible agglomeration of the silica particles on the top
surface of the coating. The abrasion measurements presented in the next section also support this hypothesis in that the
coating durability increases with the number of layers until the coating is around 4 layers thick, where further increases
in the number of layers leads to a decrease in abrasion resistance, presumably due to and increased loss of silica particles
from the coating.
Not only are the coatings required to possess excellent optical transmission over the entire solar spectrum, they must also
not adversely increase the diffuse reflectance of the coated mirrors, which is particularly important for power tower CSP
facility heliostat mirrors. The mirrors in these facilities are focused on a blackbody absorber structure on the tower,
which is heated and used to produce supercritical steam or high temperature molten salts in a boiler located at the top of
the tower. The mirrors may be located up to 1 km from the absorber in these facilities, and any surface defects on the
heliostat mirrors due to scratches, surface irregularies and particulates will increase the mirror diffuse reflectance,
reducing the amount of reflected light falling on the power tower black body absorber. Consequently, a tight specular
reflectance benchmark (deviation <12 mrad) is required to ensure maximum absorbance of the reflected radiation.

Specular Reflectance Diffuse Reflectance

100 100

90 90
SPECULAR REFLECTANCE (%R)

DIFFUSE REFLECTANCE (%R)

80 80

70 70

60 Uncoated mirror (manufacturer #1) 60 Uncoated mirror (manufacturer #1)

50 Spin-coated mirror (CA=158°) (manufacturer #1) 50 Spin-coated mirror (CA=158°) (manufacturer #1)
Spray-coated mirror (CA=152°) (manufacturer #1) Spray-coated mirror (CA=152°) (manufacturer #1)
40 40

30 30

20 20

10 10

0 0
0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500
WAVELENGTH (nm) WAVELENGTH (nm)

Figure 13. (a) Specular and (b) diffuse reflectance measurements of C7 anti-soiling coated mirror
substrates obtained from a CSP mirror manufacturer and CSP facility operator.

A solar mirror’s performance can be quantified by measuring its specular reflectance across the terrestrial solar spectrum
(λ = 280 to 2500 nm). For the present measurements, the mirror was irradiated with light that was incident at near-
normal angles θi, and the intensity of the reflected light measured with a detector. The specular reflectance ρs(λ,θ,φ) is
defined as the intensity of light at wavelength λ and incidence angle θ that is reflected within a cone of divergence with
half-angle φ and vertex sited at the point of incidence. The hemispherical reflectance ρh(λ,θ,π) of a mirror is the total
radiation reflected from the mirror surface, consisting of the combination of the specular and diffuse reflected and
scattered light. The specular reflectance is calculated by subtracting the diffuse reflectance from the hemispherical
reflectance. Optical losses due to absorption in the present measurements were negligibly small, and not considered in
calculating the specular reflectance. The angles of incidence for the hemispherical and diffuse reflectance measurements
were 3.3° and 0° respectively. A Spectralon® (polytetrafluoroethylene, or PTFE) reflectance standard was used to
calibrate the spectrophotometer detection efficiency, enabling absolute reflectance measurements of the mirror surfaces
to be made. The Spectralon® standard is identical in composition to the broadband reflectance lining in the diffuse
reflectance accessory.
Several small mirror coupons were obtained from various CSP mirror manufacturers and CSP facility operators, and the
reflectance measurements made on uncoated and anti-soiling coated mirror samples. Figure 13 shows the measurements
performed on a set of mirrors from one of the CSP mirror manufacturers, where the anti-soiling coatings were applied
using both a spin coater and a semi-automated spray system. Similar measurements were made on a number of mirrors

Proc. of SPIE Vol. 9175 91750J-8

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 obtained from different sources, and the solar weighted average specular and diffuse reflectance measured for each of
these anti-soiling coated mirrors. Several of these measurements are summarized in Table 3 and indicate that the spin
coated mirrors are meeting the diffuse reflectance benchmark, being ≈ 0.5% larger than the uncoated mirror diffuse
reflectance, and the specular reflectance for this mirror was also only ≈ 0.5% less than for the uncoated mirror. We are
still developing and optimizing the semi-automated spray coating system, and expect the 5% diffuse reflectance
measured for the two mirror samples to decrease to that of the spin coated mirrors shortly.
Table 3. Specular and diffuse measurements for uncoated reference mirrors and SH coated mirrors.
Solar-weighted Reference Spray-coated Reference Spray-coated Spin-coated
reflectance Mirror #1 Mirror #1 Mirror #2 Mirror #2 Mirror #2
Specular 94.8 89.0 95.4 90.1 94.8
Diffuse 0.60 5.65 0.51 4.85 1.06

3.3 Mechanical and Environmental and Mechanical Durability

All the coatings developed in this project were subjected to Taber abrasion testing. This is a standard abrasion test
where two weighted 2” wheels coated with calibrated abrasive materials are rotated over the surface under test, and the
water contact angle measured after the coating had been exposed to a given number of abrasion cycles. A standard CS-0
wheel with 750 gm weights, specifically designed for testing optical coatings, was used in these tests. Figure 14 shows
the measured water contact angle for various anti-soiling coatings as a function of the number of abrasion cycles in
comparison with a polymer binder coating that does not contain any silica particles.

3 Layers C7-PB 40
C7 Layered C7 (3 Layers)
170 Polymer Binder (PB) 36
6 Layers C7-PB
160
@ 10% decrease in CA
Single Abrasion Cycles

C1 32
4 Layers C7-PB
150 5 Layers C7-PB
3 Layers C7-PB Cured
28
140
Contact Angle

24
130
120 20 C6 double silane on polymer binder slide

110 16
C6 double silane
100 12
90 8
80
4
70
0
0 20 40 60 80 100 C1 C2 C3 C6 C7 C6D C6DPB C7L
Single Abrasion Passes Coatings
Figure 14. The measured water contact angle for Figure 15. The number of Taber abrasion cycles
several of the coatings developed in this project as a required for the anti-soiling coatings to fall beneath the
function the number of Taber abrasion cycles. 90% pass/fail contact angle measurement criteria.

These measurements show that the water contact angle of the pure binder coating (PB in Figure 14) remains unchanged
after 35 Taber cycles, being approximately 78-800, whereas all the anti-soiling coatings in the test start out with contact
angles in the range 155-1670. The pass/fail metric for this test was that the coating was deemed to have failed after the
contact angle for the abraded surface fell to 90% of its original value. A graph of the number of abrasion cycles the
coatings survived before failing the 90% pass/fail criteria for the various coatings listed in Figure 14, is shown in Figure
15. The first coating shown in this graph (C1) abraded very quickly and failed the test after 1 abrasion cycle, whereas
the most recent coating given in these figures (3 layers C7) passed 30 abrasion cycles before falling below the 90% level.
This coating was still very hydrophobic even after 100 Taber abrasion cycles. Surface analyses show that the coatings
fail by loss of the silica particles from the coating (rather than the coating itself being abraded away). Techniques to
improve silica particle binding to the polymer binder are in progress to further improve the mechanical durability of
these coatings.

Proc. of SPIE Vol. 9175 91750J-9

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 170 AA-
90
160

-en 150
-a
y 140 ó 80
- M- C6 Ú
Ty)

Q 130
f C7 C
(6
- Sodalime
f Binder 2 E
- C7
f C7-A c 70
f C7-B <a
- C7-B
-A- C7-C H - C7-C
- 41- C7-E - C7-D
- C7-E
100 60 - C6
- Binder 2
90
100 ' 200 300 400 ' 500 600 ' 700 300 450 600 750 900 1050 1200

UV Exposure (hrs) Wavelength, X (nm)

Figure 16. Hydrophobicity of several anti-soiling coatings Figure 17. Optical transmittance of the anti-soiling
as a function of accelerated UVA radiation exposure in a coated coupons after 300 to 700 hours exposure in the
QUV weatherometer. Results are equivalent to several weatherometer. Optical transmittance of the coatings is
years exposure in a desert environment. unchanged after the UV exposure.
In addition to mechanical abrading the coated surfaces, several glass coupons coated with anti-soiling coatings were
subjected to simulated solar UVA exposure, and the water contact angles monitored as a function of UVA exposure
time. The accelerated UV exposures are approximately 12-17 times greater than that of normal solar UVA exposures in
the field. Figure 16 shows several coatings along with a polymeric binder with up to 700 hours of simulated UV
exposure. Some coatings still maintain their superhydrophobicity with a CA >150°. The binder-only coating follows a
similar trend, however at a significantly lower CA, as expected. Coatings based on prior formulations initially followed
the same trend as the top performing coatings, but after 200 hours of UV exposure, the CA values fall rapidly and the
coatings become marginally hydrophobic.
The optical transmittance of the coatings was also measured on a regular basis as the coatings were exposed to the UVA
radiation, and selected measurements are shown in Figure 17 for exposure times ranging from 300 to 700 hours. These
measurements show that the coatings maintained their transmission properties over the entire exposure period with
transmittances ± 2% of that of the uncoated sodalime glass reference coupon. These simulated environmental tests result
in the coatings exposure being equivalent to 8000-12000 hours of real world solar UV exposure. This is on the order of
nearly two to three years in the field, assuming 12 hours exposure per day.

3.4 Observed Anti-Soiling Properties

Y

Figure 18. Several frames from a video clip showing silica powder easily being washed from a glass coupon
coated with the C7 anti-soiling coating.
Finally, although the optical and SH water repellency properties of the coatings may be excellent with good coating
durability, if the coatings do not show good anti-soiling behavior, then they are of little value as coatings for CSP
mirrors. We have performed qualitative demonstrations of their value as anti-soiling coatings as shown in Figures 18
and 19. Figure 18 shows several frames from a video where silica powder was sprinkled on the surface of a glass
coupon coated with a C7 anti-soiling coating, where the measured water contact angle was approximately 1650 across
the entire surface. Water was then dribbled onto the surface, and the very loosely bound silica particles were picked up
by the water droplets and transported off the coupon. These demonstrations have been performed with various other
powders including pollen, soot, powdered fire embers and desert sand with the same result – none of these materials was
bonded to the surface and was easily removed by a gentle stream of water flowing across the surface. When these

Proc. of SPIE Vol. 9175 91750J-10

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 materials were scattered on a regular hydrophilic glass surface where the measured contact was >300, the powders
wetted and smeared across the surface as the water drops spread over the surface. Much larger volumes of water were
required to remove the powders, and in most cases, a film of material remained on the surface even after extensive
washing. Sand, dust and other materials can be easily and completely washed or blown from surfaces coated with our
anti-soiling formulations.

(a) (b)
Sand pile

Back side of slide (not coated with the Back side of glass slide coated with
superhydrophobic coating) is clean at dust while the front coated surface is
the beginning of the experiment clean (no dust accumulation)
Figure 19. (a) A glass coupon with and without an anti-soiling coating on the upper and lower surfaces respectively.
After the sand drop experiment, (b) the lower uncoated surface of the coupon is coated with fine dust while the upper
coated surface remains clean.
Even more intriguingly, we have shown that these coatings do not allow finely charged dust particles to stick to and
accumulate on the surface of the coatings. Figure 19 shows a glass coupon coated on the upper surface with an anti-
soiling coating, while the lower back side of the coupon remained uncoated. The measured water contact angles for
these two surfaces was again approximately 1650 and 300 for the upper and lower surfaces respectively. The coated
coupon was mounted on a 450 stand and exposed to falling sand in a standard sand drop experiment. Figure 19(a) shows
the coupon before the sand exposure and 19(b) after exposure to the falling sand. A pile of sand can be seen under the
coupon in Figure 19(b). Notice that the coupon is very transparent in (a) but is opaque in (b) due to the accumulation of
fine sand particles, which we noticed were being actively attracted to the back side of the coupon when the sand was
falling. The front side of the coupon remained clear, and the supposition is that the charged sand particles were attracted
to the lower surface, but not the upper surface of the coupon. These studies are continuing.
4. CONCLUSION
The installation and recent operation of large CSP facilities in the desert southwest of the US has shown the need for
innovative solutions to maintain collector and heliostat cleanliness, and for techniques to quickly and easily remove dust
and other debris that may accumulate on the surface of these mirror during normal facility operation, while minimizing
plant down time and the use of expensive cleaning water and washing procedures. In this paper we have shown that
transparent SH coatings based on functionalized nano SiO2 particles, imbedded in a polymer binder with sub 100 nm
surface roughness and a high water contact angle (>160°), can be fabricated with minimal optical transmission and
diffuse reflectance losses. The optical transmittance of the anti-soiling coating can be improved due to its antireflection
properties, and the coating durability increased when several layers of coating are applied to the glass substrates.
Accelerated UV life testing of the coatings indicates that the coatings maintain superhydrophobicity (and anti-soiling
properties) and excellent optical transmission after exposure to extreme UVA conditions for a simulation period of
several years. Very importantly, the anti-soiling coatings demonstrate moderate to high mechanical durability under
standard Taber abrasion testing, with recently developed coatings surviving 30 times longer than those developed at the
beginning of the project. These coatings are more robust than all SH coatings reported previously in the literature.
Finally, these SH coatings show great promise for a number of other important applications where excellent optical
transmission and anti-soiling are a necessity (e.g. PV panels and large scale installations, windows, windscreens,
waterproof optics, etc.). They are also an excellent choice for a range of non optical applications, including anti-icing
coatings for power lines and icing prone structures, anti-condensation coatings for air conditioning, desalination plants
and windows, and various other anti-corrosion and anti-biofouling applications. These coatings can be tailored to adhere
to a range of substrates using low cost coating application techniques, and applied to large surface areas.

Proc. of SPIE Vol. 9175 91750J-11

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms

 ACKNOWLEDGEMENTS
Oak Ridge National Laboratory is operated for the U.S. Department of Energy by U.T.-Battelle under Contract No. DE-
AC05-00OR22725. This work was supported by the U.S. Department of Energy, SunShot Program of the Office of
Energy Efficiency and Renewable Energy. A portion of this research was conducted at the Center for Nanophase
Materials Sciences, which is sponsored at Oak Ridge National Laboratory by the Scientific User Facilities Division,
Office of Basic Energy Sciences, U.S. Department of Energy.

REFERENCES
[1] http://www1.eere.energy.gov/solar/sunshot/csp.html
[2] Sunshot Vision Study, U.S. DOE, Chpt. 5, (2012) – report may be obtained from the following website -
http://www1.eere.energy.gov/solar/pdfs/47927.pdf
[3] http://energy.gov/eere/sunshot/concentrating-solar-power
[4] C. Kennedy, “Advanced Reflectors,” Solar Energy Technologies Program Review Peer Review, U.S. DOE EERE
Program (2010).
[5] http://www.solarpowerworldonline.com/2010/08/nasa-technology-translates-into-self-cleaning-solar-
panels/
[6] Brindle, B, P. de los Reyes and K. Jamison, “Specialty Glass Needs for the U.S. Solar Industry Workshop,” U.S.
DOE EERE, Golden, CO (2008).
[7] Hunter, S. R. and J. T. Simpson, “Large area, clear, transparent superhydrophobic coatings,” Sensors, and
Command, Control, Communications, and Intelligence (C3I) Technologies for Homeland Security and Homeland
Defense XI: Defense Science and Security Symposium 2012, Baltimore, MD (2012).
[8] Schaeffer, D. A., G. Polizos, D. B. Smith, D. F. Lee, S. Rajic, P. G. Datskos and S. R. Hunter, “Spray-on anti-
soiling coatings that exhibit high transparency and mechanical durability,” Sensors, and Command, Control,
Communications, and Intelligence (C3I) Technologies for Homeland Security and Homeland Defense XIII, edited
by Edward M. Carapezza, Proc. of SPIE Vol. 9074, 90740C (2014).
[9] Polizos, G., D. A. Schaeffer, D. B. Smith, D. F. Lee, P. G. Datskos and S. R. Hunter, “Enhanced durability
transparent superhydrophobic anti-soiling coatings for CSP applications,” Proc. of the ASME 2014 8th Int. Conf.
on Energy Sustainability & 12th Fuel Cell Science, Engineering and Technology Conference, ES FuelCell 2014
June 29-July2, 2014, Boston, Massachusetts.
[10] M. A. Sarshar, C. Swarctz, S. R. Hunter, J. T. Simpson and C.-H. Choi, “Effects of contact angle hysteresis on ice
adhesion and growth on superhydrophobic surfaces under dynamic flow conditions,” Colloid. Polym. Sci., 291
427–435 (2013).
[11] Polizos, G., K. Winter, M. J. Lance, H. M. Meyer, B. L. Armstrong, D. A. Schaeffer, J. T. Simpson, S. R. Hunter
and P G. Datskos, “Scalable superhydrophobic coatings based on fluorinated diatomaceous earth: Abrasion
resistance versus particle geometry,” Appl. Surf. Sci. 292, 563-569 (2014).
[12] Park, K. C., H. J. Choi, C.-H. Chang, R. E. Cohen, G. H. McKinley and G. Barbastathis, “Nanotextured silica
surfaces with robust superhydrophobicity and omnidirectional broadband supertransmissivity,” ACS Nano 6, 3789
(2012).
[13] Cao, L. L., A. K. Jones, V. K. Sinod J. Wu and D. Gao, “Anti-Icing Superhydrophobic Coatings,” Langmuir 25,
12555-12448 (2009).
[14] Gao, X. F., X. Yan, X. Yao, L. Xu, K. Zhang, J. Zhang, B. Yang and L. Jiang, “The dry-style antifogging
properties of mosquito compound eyes and artificial analogues prepared by soft lithography,” Adv. Mater. 19,
2213 (2007).

Proc. of SPIE Vol. 9175 91750J-12

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/15/2015 Terms of Use: http://spiedl.org/terms