High-performance solution-processed plasmonic Ni nanochain-Al2O3selective solar thermal absorbersXiaoxin Wang, Haofeng Li, Xiaobai Yu, Xiaoling Shi, and Jifeng Liu Citation: Appl. Phys. Lett. 101, 203109 (2012); doi: 10.1063/1.4766730 View online: http://dx.doi.org/10.1063/1.4766730 View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v101/i20 Published by the American Institute of Physics. Related ArticlesEffect of load variations on exergy performance of solar box type cooker J. Renewable Sustainable Energy 4, 053125 (2012) Intergrain variations of the chemical and electronic surface structure of polycrystalline Cu(In,Ga)Se2 thin-filmsolar cell absorbers Appl. Phys. Lett. 101, 103908 (2012) Experimental investigation of various designs of solar flat plate collectors: Application for the drying of green chili J. Renewable Sustainable Energy 4, 043116 (2012) Effect of solar intensity on efficiency of the convection solar air heater J. Renewable Sustainable Energy 4, 042901 (2012) Nusselt number and friction factor correlations for solar air heater duct with broken V-down ribs combined withstaggered rib roughness J. Renewable Sustainable Energy 4, 033122 (2012) Additional information on Appl. Phys. Lett.Journal Homepage: http://apl.aip.org/ Journal Information: http://apl.aip.org/about/about_the_journal Top downloads: http://apl.aip.org/features/most_downloaded Information for Authors: http://apl.aip.org/authors

High-performance solution-processed plasmonic Ni nanochain-Al2O3

selective solar thermal absorbers

Xiaoxin Wang,a) Haofeng Li, Xiaobai Yu, Xiaoling Shi, and Jifeng Liub)

Thayer School of Engineering, Dartmouth College, 14 Engineering Drive, Hanover,New Hampshire 03755-8000, USA

(Received 25 July 2012; accepted 24 October 2012; published online 13 November 2012)

Selective solar thermal absorber coating is an important component of concentrated solar power

systems. It maximizes the absorption of solar spectrum and minimizes thermal radiation losses in

the mid-infrared regime. In this letter, we demonstrate a solution-processed plasmonic Ni

nanochain-Al2O3 selective solar thermal absorber with a high solar absorptance >90% and a low

thermal emittance loss <10%. Unlike conventional graded-index cermet coatings, the spectral

selectivity is tailored by the lengths of Ni nanochains, elimating the requirement of costly vacuum

deposition for stringent thickness control. These results open a path to utilize plasmonics for

low-cost, high-performance solar thermal systems. VC 2012 American Institute of Physics.

[http://dx.doi.org/10.1063/1.4766730]

Solar thermal and solar photovoltaics are two major

approaches to harvest solar energy. Compared to photovol-

taics, solar thermal technology based on concentrated solar

power (CSP) is advantageous in terms of lower cost and

much easier energy storage. Since the heated working fluid,

such as molten salt, can be stored and kept at a high tempera-

ture for an extended period of time, CSP allows stored solar

thermal energy to be dispatched as needed, an attractive so-

lution to the intermittency issue of solar energy.1 A critical

component of CSP systems is selective solar thermal absorb-

ers.2 Ideally, the solar absorber for a CSP system should

absorb all the incident solar energy and convert it to heat

without thermal emittance losses in the mid infrared (MIR)

regime due to black-body radiation. Instead, the MIR ther-

mal radiation is reflected back to the CSP system.

Cermet selective solar absorbers consisting of metal

nanoparticles dispersed in a ceramic matrix have been devel-

oped for applications in CSP systems.3,4 It is well known that

metals have a high absorption coefficient in the solar spec-

trum, yet a continuous layer of metal reflects most of the light

instead of absorbing it due to the refractive index mismatch

with air. Conventionally, cermet structures address this issue

by incorporating very small metal nanoparticles with diame-

ters of 5–10 nm into ceramic layers at different volume frac-

tions to tune the refractive index profile of the coating for

optical impedance matching in the solar spectrum regime. As

illustrated in Fig. 1(a), a graded-index cermet comprises mul-

tiple ceramic layers with increasing volume fractions of small

metal nanoparticles from top to bottom. Sometimes an infra-

red (IR) reflector coating is coupled to enhance spectral selec-

tivity.5 Since the size of metallic nanostructures is much

smaller than the wavelength of incident light, the effective

dielectric function of cermets can be approximated by effec-

tive medium theories (such as Maxwell-Garnet Rule) based

on the dielectric function of metal, ceramic matrix, and metal

nanoparticle volume fraction.6 The design optimization is

based on engineering the effective refractive index and film

thickness of each layer.7 In these structures the cermet film

thicknesses are critical for the optical performance, leading to

stringent requirements on thickness control. Therefore, most

of the existing cermet fabrication techniques rely on vacuum

deposition, such as sputtering, evaporation, or chemical vapor

deposition.3 It was not until recent years that non-vacuum

coating methods, such as electrodeposition8 and solution-

chemical coating,9 have been investigated as less expensive

and more efficient approaches for cermet-based thermal solar

absorbers.

To alleviate the stringent film thickness constraint of tra-

ditional cermet solar selective absorbers and facilitate

solution-chemical processing, we explore metal nanochains

that exhibit inherent spectral selectivity for CSP applications

instead of relying on cermet thickness design. In this letter

we demonstrate a solution-processed plasmonic Ni nano-

chain-Al2O3 selective solar thermal absorber with a high so-

lar absorptance >90% and a low thermal emittance <10%

for low-cost, high performance CSP systems. As shown in

Fig. 1(b), the Ni nanochains consist of Ni nanoparticles,

�100 nm in diameter, an order of magnitude larger than

nanoparticles in traditional cermets. This nanoparticle size is

chosen such that absorption and scattering in the solar spec-

trum can be significantly enhanced by optical excitation of

surface plasma polaritons (SPPs) in metal nanostructures.

Strong SPP scattering from Au nanoparticles (50–100 nm in

diameter) towards Si has been utilized to increase the photo-

current in Si photodiodes and improve the energy conversion

efficiency in thin film photovolatic devices.10–12 A challenge

for solar absorber applications, though, is that the plasmonic

resonances in noble metals (like Au and Ag) are too narrow

to cover the entire solar spectrum. For broadband plasmonic

absorbers, Ag crossed gratings have been designed and fabri-

cated by e-beam lithography.13 In contrast to noble metals,

ferromagnetic metals such as Ni and Fe have a higher damp-

ing coefficient so that the SPP resonance appears broader in

spectrum.14 This feature is highly advantageous for solar

thermal absorbers. Furthermore, the optical response can be

a)Electronic mail: Xiaoxin.Wang@dartmouth.edu.b)Electronic mail: Jifeng.Liu@dartmouth.edu.

0003-6951/2012/101(20)/203109/5/$30.00 VC 2012 American Institute of Physics101, 203109-1

APPLIED PHYSICS LETTERS 101, 203109 (2012)

optimized by tailoring the length of Ni nanochains. The

nanochains can also form a 3D network, which offers even

stronger overall solar absorption due to multiple scatterings.

On the other hand, the �100 nm diameter Ni nanoparticles

are small enough such that long wavelength MIR photons

from thermal radiation cannot resolve them. Consequently,

these MIR photons see the Ni nanochain network as a contin-

uous metal sheet and are reflected back to the CSP system,

minimizing thermal emittance losses. The optical perform-

ance is mainly determined by the structure of plasmonic Ni

nanochains instead of layer thicknesses, which greatly facili-

tates solution-chemical processing.

In order to understand the plasmonic effect of Ni nano-

chains, the absorption, scattering, and extinction efficiency

factors of individual Ni nanochains are calculated by a 3D

finite-element method (FEM) and shown in Fig. 2. The nano-

chains are composed of a series of connected Ni nanospheres

with a diameter of 80 nm, and the chains are 1-, 2-, 6-, and

10-nanosphere long. The efficiency factors presented here are

absorption/scattering/extinction cross-sections normalized by

the geometric area of the nanostructures and thereby unit-

less. The dielectric functions of Ni and Al2O3 are obtained

from Ref. 15. To improve the computation efficiency based

on the symmetry of the structure, only one quarter of the

domain is calculated using adequate boundary conditions. To

check the validity of FEM method, the absorption/scattering/

extinction efficiency factors of one Ni nanosphere are calcu-

lated and found to be highly consistent with analytical solu-

tions based on Mie scattering theory (Fig. 2(a)).16 We have

analyzed incident plane waves propagating in the z direction

with linear x-polarization (along the nanochain) and y-

polarization (perpendicular to the nanochain), respectively.

As summarized in Fig. 2, the optical response of the individ-

ual nanochain is dependent on the polarization direction. For

incident light polarized in x-direction along the nanochain,

the optical response spectrum is extended from k� 1000 nm

to k� 2500 nm with the increase of nanosphere numbers,

covering >99% of optical energy in the solar spectrum. Such

spectrum broadening saturates when the number of nanopar-

ticles in a nanochain is large enough, as evidenced by the

similarities of optical response from 6- and 10-nanosphere

chains in Fig. 2(a). To further elucidate the mechanism of

extended absorption spectrum with the number of spheres in

a nanochain, we investigated the optical field distribution at

k¼ 2500 nm under x-polarization. The optical field is found

to be strongly enhanced at the narrow gaps between nano-

spheres due to near-field plasmonic effect, an advantage over

nanorod structures in terms of SPP enhancement. The 6-

nanosphere chain shows a stronger field enhancement than

the 2-nanosphere one at k¼ 2500 nm, confirming that the

increase in absorption at longer wavelengths (k> 2000 nm) is

indeed related to the SPP effect. It is likely that plasmonic

coupling among a larger number of nanospheres leads to a

stronger field enhancement at the gaps between nanospheres.

When the number of nanospheres becomes large enough the

plasmonic coupling among them reaches a steady state, which

qualitatively explains the saturation in spectral broadening.

On the other hand, for incident light polarized in y-direction

perpendicular to the nanochain, Fig. 2(b) shows that the

absorption/scattering/extinction efficiency factors are not sen-

sitive to the nanochain length, as one would expect. In con-

trast to x-polarization, the optical field distribution between

nanospheres is almost the same for 2-nanosphere and 6-

nanosphere chains. In the real cermet structures with ran-

domly oriented nanochain networks formed in 3D, the optical

response is polarization-independent. It is an average of the

two cases in Fig. 2. The multi-scattering among nanochains

may also enhance the overall absorption. The broad tunable

optical response in the solar spectrum (300–2500 nm) is par-

ticularly beneficial to solar selective absorbers. Unlike con-

ventional graded-index cermets, it is not necessary to use

multilayers with precise thickness control of each layer.

These advantages greatly facilitate large-scale fabrication of

high-performance solar selective absorbers by solution-

chemical techniques.

To demonstrate this concept, Ni nanochain-Al2O3 cermet

coatings are fabricated by a solution-chemical process. First

Ni nanochains with a diameter of �80 nm and a length of

2–3 lm are synthesized by reducing Ni2þ with N2H4,17 as

shown by the scanning electron microscopy (SEM) image in

Fig. 3(a). It is observed that Ni nanochains do not separate

into individual nanoparticles even under strong ultrasonic

vibration. Then the Ni nanochains are dispersed in an Al2O3

sol for spin-coatings on 25-lm-thick, 20� 20 mm2 stainless

steel substrates.18 Finally, the samples are annealed at 400 �Cfor 1 h in N2 to form �1-lm-thick cermet coating. More

FIG. 1. Schematic structures of solar selective absorbers. (a) Conventional

graded-index Ni-Al2O3 cermet composed of 5–10-nm-diameter Ni nanopar-

ticles, with increasing volume fractions of nanoparticles from top to bottom.

(b) Proposed coatings with Ni nanochains embedded in Al2O3 matrix. Nano-

chains comprise a series of connected Ni nanospheres with a diameter of

�100 nm.

203109-2 Wang et al. Appl. Phys. Lett. 101, 203109 (2012)

details about the fabrication process are provided in the sup-

plementary material.19 The surface of the Ni nanochain-

Al2O3 cermet layer visually looks black, as expected for solar

thermal coatings. The SEM image in Fig. 3(b) shows that

Al2O3 nicely covers Ni nanochains.

Fig. 4 shows the x-ray diffraction data of the Ni nano-

chain-Al2O3 layer in comparison with as-synthesized Ni

nanochains alone. The bump at �20� corresponds to amor-

phous Al2O3 matrix, while the three peaks correspond to Ni

(111), Ni (200), and Ni (220), respectively. Compared to as-

sythesized Ni nanochains, it is clear that there is no modifica-

tion in crystal structure or formation of Ni oxide during the

annealing process.

The reflection of the Ni nanochain-Al2O3 cermet coating

in the wavelength range of 0.2–2.5 lm is measured using a

UV-VIS-NIR spectrometer with an integrating sphere to col-

lected both specular and diffuse reflection,20 and the reflec-

tion in the infrared range of 2.5–20 lm is obtained using a

Fourier transform infrared spectroscopy (FTIR) equipment.

Fig. 5 shows the reflectance spectrum of the Ni nanochain-

Al2O3 cermet coating in the wavelength range of 0.3–15 lm.

The AM 1.5 solar spectrum and black-body radiation spec-

trum at 400 �C are also shown as a reference. According to

Kirchhoff’s law, at each wavelength k the absorptance aðkÞis equal to the emittance eðkÞ under thermal equilibrium.4

Since the transmittance through the stainless steel substrate

FIG. 2. 3D finite-element calculation of

absorption (magenta line), scattering (red

line), and extinction (green line) efficiency

factors vs. wavelengths for individual Ni

nanochains with 1, 2, 6, and 10 nanospheres.

(a) Incident light with x-polarization (along

the nanochain). The dotted dark gray, gray,

and light gray lines correpond to the calcu-

lated absorption/scattering/extinction effi-

ciency factors of one Ni nanosphere based on

the analytic Mie scattering theory, respec-

tively. (b) Incident light with y-polarization

(perpendicular to the nanochain). The insets

show the magnitude of the electric field jExjfor x-polarization and jEyj for y-polarization

at k¼ 2500 nm in the cases of 2-nanosphere

and 6-nanosphere chains.

FIG. 3. SEM images of Ni nanochains: (a)

as-synthesized; (b) as-coated and annealed

Ni-Al2O3 composite at an annealing temper-

ature of 400 �C.

203109-3 Wang et al. Appl. Phys. Lett. 101, 203109 (2012)

is 0, the absorptance/emittance at each wavelength is equal

to one minus the reflectance at that wavelength4

aðkÞ ¼ eðkÞ ¼ 1� RðkÞ � TðkÞ ¼ 1� RðkÞ: (1)

Here RðkÞ and TðkÞ are reflectance and transmittance at

wavelength k, respectively. Consequently, a low reflectance

is desirable in the solar spectrum regime of k¼ 0.3–2.5 lm

for high solar absorptance, while a high reflectance is desira-

ble in the mid infrared regime at k> 3 lm for low thermal

emittance. The reflectance of the Ni nanochain-Al2O3 cermet

coating is 0.06–0.09 in the wavelength range of k¼ 0.3–

2.1 lm, leading to high solar absorptance. The reflectance

starts to increase significantly at k> 2.1 lm, indicating a

drastic decrease in absorption. This transition point corre-

sponds well to the roll-off in the calculated absorption/scat-

tering/extinction spectra of Ni nanochains at k> 2 lm, as

shown earlier in Fig. 2(a). Therefore, the optical performance

of this Ni nanochain-Al2O3 cermet structure is consistent

with theoretical predictions. The reflectance is >0.9 at

k> 3.5 lm, leading to low thermal emittance in the MIR re-

gime. The overall solar absorptance, asol, and overall thermal

emittance, etherm, are derived from the reflectance spectrum

using the following equations:9

asol ¼

ð2:5lm

0:3lm

IsolðkÞaðkÞdk

ð2:5lm

0:3lm

IsolðkÞdk

¼

ð2:5lm

0:3lm

IsolðkÞ½1� RðkÞ�dk

ð2:5lm

0:3lm

IsolðkÞdk

;

(2a)

etherm ¼

ð20lm

2lm

IPðkÞeðkÞdk

ð20lm

2:5lm

IPðkÞdk

¼

ð20lm

2lm

IPðkÞ½1� RðkÞ�dk

ð20lm

2:5lm

IPðkÞdk

: (2b)

Here IsolðkÞ is the radiation intensity at wavelength k in

AM 1.5 solar spectrum and IpðkÞ is the radiation intensity at

wavelength k in 400 �C black-body radiation spectrum. The

overall solar absorptance is determined to be 93%, and the

overall thermal emittance is 9% for the current structure. A

small amount of solar energy may be scattered by Ni nano-

chains and absorbed by the stainless steel substrate, which

also contributes to the overall solar-thermal energy conver-

sion. This optical performance is comparable to vacuum de-

posited multilayer cermets,7 while the fabrication process is

much less expensive. The performance can be further opti-

mized by fine tuning the nanoparticle sizes to better match

the solar spectrum and the 400 �C black-body radiation

spectrum.

In conclusion, plasmonic Ni nanochain-Al2O3 cermet

structures are fabricated by cost-effective solution-chemical

approach for solar thermal applications. Unlike conventional

multilayer graded-index cermet coatings, SPP enhanced so-

lar absorption in these nanostructures are tailored by the

lengths of Ni nanochains instead of cermet layer thicknesses,

elimating the requirement of costly vacuum deposition for

stringent thickness control. High solar absorptance >90%

and low thermal emittance losses <10% have been demon-

strated in these Ni nanochain-Al2O3 cermet coatings, compa-

rable to the performance of vacuum deposited selective

cermet absorbers. These results open a path to utilize plas-

monics for low-cost, high-performance performance solar

thermal systems.

This work was supported by New Hampshire Innovation

Research Center (NHIRC) and Axisol, Inc. We would also

like to thank Dr. Xing Sheng at Massachusetts Institute of

Technology for assistance with some optical measurements.

1A. Gil, M. Medrano, I. Martorell, A. L�azaro, P. Dolado, B. Zalba, and L.

F. Cabeza, Renewable Sustainable Energy Rev. 14, 31 (2010).2S. Chaudhuri, D. Bhattacharyya, A. B. Maity, and A. K. Pal, Mater. Sci.

Forum 246, 181 (1997).3J. Gordon, Solar Energy: The State of the Art (James & James Ltd., Lon-

don, 2001).4C. E. Kennedy, Review of Mid- to High Temperature Solar SelectiveAbsorber Materials (National Renewable Energy Laboratory (NREL),

Washington, DC, 2002).

FIG. 4. X-ray diffraction (XRD) data of (a) as-synthesized Ni nanochains

(magenta line); (b) as-annealed Ni nanochain embedded in amorphous

Al2O3 matrix (purple line).

FIG. 5. Reflection spectrum of the solution-chemical derived Ni nanochain-

Al2O3 cermet coating (blue line). Normalized AM 1.5 solar spectrum (olive

line) and 400 �C black-body radiation spectrum (magenta line) are also

shown as a reference.

203109-4 Wang et al. Appl. Phys. Lett. 101, 203109 (2012)

5C. G. Granqvist, Materials Science for Solar Energy Conversion Systems(Pergemon, Oxford, 1991).

6O. P. Agnihotri and B. K. Gupta, Solar Selective Surfaces (John Wiley &

Sons, New York, 1981).7Q. C. Zhang, K. Zhao, B. C. Zhang, L. F. Wang, Z. L. Shen, Z. J. Zhou, D.

Q. Lu, D. L. Xie, and B. F. Li, Sol. Energy 64, 109 (1998).8R. L. P. Teixeira, R. A. Sim~ao, B. Coelho, and A. C. Oliveir, Int. J. Energy

Environ. 5, 266 (2011), available at http://www.naun.org/multimedia/

NAUN/energyenvironment/19-787.pdf.9T. Bostr€om, E. W€ackelgard, and G. Westin, Sol. Energy 74, 497 (2003).

10S. H. Lim, W. Mar, P. Matheu, D. Derkacs, and E. T. Yu, J. Appl. Phys.

101, 104309 (2007).11H. A. Atwater and A. Polman, Nat. Mater. 9, 205 (2010).12D. Derkacs, S. H. Lim, P. Matheu, W. Mar, and E. T. Yu, Appl. Phys.

Lett. 89, 093103 (2006).

13K. Aydin, V. E. Ferry, R. M. Briggs, and H. A. Atwater, Nat. Commun. 2,

517 (2011).14J. N. Chen, P. Albella, Z. Pirzadeh, P. Alonso-Gonz�alez, F. Huth, S. Bone-

tti, V. Bonanni, J. Akerman, J. Nogu�es, P. Vavassori, A. Dmitriev, J. Aiz-

purua, and R. Hillenbrand, Small 7, 2341 (2011).15E. D. Palik, Handbook of Optical Constants of Solids (Academic, Boston,

1985).16MiePlot 4.0 Software, Philip Laven, Switzerland.17M. S. Hussain and K. M. A. Haque, J. Sci. Res. 2, 313 (2010), available at

http://www.banglajol.info/index.php/JSR/article/download/3261/3918.18S. Saha, J. Sol-Gel Sci. Technol. 3, 117 (1994).19See supplementary material at http://dx.doi.org/10.1063/1.4766730 for the

details of the fabrication process.20P. Nostell, A. Roos, and D. Ronnow, Rev. Sci. Instrum. 70, 2481

(1999).

203109-5 Wang et al. Appl. Phys. Lett. 101, 203109 (2012)