Broadband High-Performance Infrared Antireflection Nanowires

Letter pubs.acs.org/NanoLett

Broadband High-Performance Infrared Antireflection Nanowires Facilely Grown on Ultrafast Laser Structured Cu Surface Peixun Fan,*,†,‡ Benfeng Bai,‡ Jiangyou Long,† Dafa Jiang,† Guofan Jin,‡ Hongjun Zhang,† and Minlin Zhong*,† †

Downloaded by CENTRAL MICHIGAN UNIV on September 11, 2015 | http://pubs.acs.org Publication Date (Web): August 18, 2015 | doi: 10.1021/acs.nanolett.5b02141

Laser Materials Processing Research Centre, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, P. R. China ‡ State Key Laboratory of Precision Measurement Technology and Instruments, Department of Precision Instrument, Tsinghua University, Beijing 100084, P. R. China S Supporting Information *

ABSTRACT: Infrared antireflection is an essential issue in many fields such as thermal imaging, sensors, thermoelectrics, and stealth. However, a limited antireflection capability, narrow effective band, and complexity as well as high cost in implementation represent the main unconquered problems, especially on metal surfaces. By introducing precursor micro/nano structures via ultrafast laser beforehand, we present a novel approach for facile and uniform growth of high-quality oxide semiconductor nanowires on a Cu surface via thermal oxidation. Through the enhanced optical phonon dissipation of the nanowires, assisted by light trapping in the micro structures, ultralow total reflectance of 0.6% is achieved at the infrared wavelength around 17 μm and keeps steadily below 3% over a broad band of 14−18 μm. The precursor structures and the nanowires can be flexibly tuned by controlling the laser processing procedure to achieve desired antireflection performance. The presented approach possesses the advantages of material simplicity, structure reconfigurability, and cost-effectiveness for mass production. It opens a new path to realize unique functions by integrating semiconductor nanowires onto metal surface structures. KEYWORDS: Metal, nanowire, infrared, antireflection, ultrafast laser, structuring

A

forest on Si22 or LiTaO323 substrates. Moreover, ultrabroadband AR properties over the spectrum from UV to THz regime have been shown by fabricating nanotip arrays on Si surfaces.4 Broadband infrared light absorption of Si surfaces can also be reached by surface microstructuring or by hyperdoping them with chalcogens via femtosecond laser irradiation.24 Unfortunately, the manufacturing efficiency as well as the cost issue of the nanotubes or nanotips is still an unconquered obstacle hindering their practical usage. Furthermore, these methods suffer from a material constraint and are inconvenient to be implemented on other materials, especially on metal surfaces. Enhanced infrared absorption on Au over a broad spectrum (2.7−15.1 μm) was reported by forming quasi-periodic microstructures.8 However, multiple processes are utilized in the manufacturing procedure, making it not qualified as a general method. Recently, “black metal” surfaces produced through direct femtosecond laser irradiation was reported to realize strong UV−vis-NIR broadband antireflection via a combined effect of macro-, micro-, and nano- surface

ntireflection (AR) surfaces have attracted tremendous attention over the years due to their great potential to be applied in wide fields.1−21 Inspiring progress has been achieved through different strategies including destructive interference coatings,2,3 gradient refractive index or impedance matching coatings,4−9 geometrical light trapping,10−12 resonance-based methods,13−20 etc., with near perfect AR effect being obtained. To date, however, most of the achieved progress focused on the visible spectrum. AR properties in mid-infrared (mid-IR) or lower frequency spectrum, although of great realistic significance for infrared imaging, sensors, thermoelectrics, stealth, artificial blackbody, and many other applications, have not been adequately studied. Even though the resonance absorption strategies can result in efficient elimination of the infrared reflection, the absorption spectral range strongly depends on the geometry and size of their component units and is inherently limited to a single narrow spectrum band, no matter if they are metamaterial based,13 surface plasmon based,18 or resonant optical cavity based. 19 Besides, elaborate structures and complicated fabrication procedures are usually involved in the resonant systems, making them too expensive or unsuitable for large area usage. Several reports have demonstrated broadband infrared AR properties by growing vertically aligned carbon nanotube © 2015 American Chemical Society

Received: June 1, 2015 Revised: August 6, 2015 Published: August 17, 2015 5988

DOI: 10.1021/acs.nanolett.5b02141 Nano Lett. 2015, 15, 5988−5994

Letter


Nano Letters

Figure 1. Schematic of the strategy for facile growth of NWs on ultrafast laser structured surfaces. a, Original bulk Cu sample. b, Integrated microcone arrays formed after ultrafast laser structuring. c, Single unit of the microcone arrays showing the multi-internal reflection effect. d, Single unit of the microcone arrays with NWs (the slender red rods around the cones) grown after thermal oxidation process; the different colors in the shells of the microcones represent the different degree of oxidation, more specifically, purple for cuprous oxide and red for cupric oxide. The blue arrows in c and d represent the traveling of incident light among the microcones.

structures.25,26 But the performance of the combined effect in the mid-IR and longer wavelength regime was less than satisfactory.26 Therefore, the efficient infrared AR property still remains a major challenge, especially on metal surfaces which are usually inherent excellent reflectors at the infrared wavelength regime. Until now, few reports of near unity infrared AR property reaching reflectance below 1% and fewer reports of broadband near unity AR strategies which keep effective through the spectrum of 5−25 μm on metal surfaces have been demonstrated. Besides, for the sake of practical usage, convenient and cost-effective implementation of the AR properties must be fulfilled. It is known that the high optical impedance between metal and the free space is the culprit accounting for the severe apparent reflection of metal surfaces. An effective method for alleviating the optical impedance is to introduce a transitional medium. The semiconductor materials, metal oxide in particular, are expected to be good candidates to realize such a purpose. However, it is technically difficult to steadily produce a uniform as well as highly efficient oxide layer on metal surfaces. Here, a novel ultrafast laser hybrid processing approach is proposed and experimentally realized to tackle this fundamental challenge. By forming precursor micro/nano structures via ultrafast laser, oxide in the unique structural form of nanowires (NWs) is facilely and uniformly grown on Cu surfaces after a subsequent simple thermal oxidation process. Based on the enhanced optical phonon dissipation effects of the oxide NWs, assisted by the ultrafast laser fabricated micro structures’ light trapping, ultralow hemispherical reflectance of 0.6% is achieved at the wavelength around 17 μm and keeps steadily below 3% over a wide spectrum from 14 to 18 μm. Surface reflection over the entire mid-IR spectrum range of 5−25 μm is also dramatically suppressed. To date, to our knowledge, this is the broadest near

unity AR property in the mid-IR regime obtained by structures in situ formed on one single metallic material system. Both the ultrafast laser structuring and the thermal oxidation are flexible techniques which are not only convenient but also cost-effective for scaling up production; thus, our approach demonstrates a promising candidate for practical applications. Figure 1 shows the scheme and strategy used in our approach. Thermal oxidation has been reported to be an effective and efficient method for growing CuO NWs on Cu substrates with particular micro scale dimensions, e.g., micro Cu grids,27 micro Cu wires,27 micro Cu powders,28 and so forth. However, it is still an unsolved technical problem to grow CuO NWs uniformly on arbitrary bulk Cu substrates without exfoliation. Here, we use an ultrafast laser to produce specific micro and nano structures on Cu surfaces first (Figure 1b,c). Specifically, an Edgewave Nd: YVO4 laser system, which can generate 10 ps pulses at a central wavelength of 1064 nm, was utilized for structuring Cu samples with a dimension of 25 × 25 × 3 mm3. An x−y galvo was used to scan the laser beam on Cu surfaces in a pattern of cross lines with varied scanning intervals in atmospheric environment. After laser processing, the Cu samples with micro/nano structures were taken in a quartz boat and heated in a horizontal-tube furnace in static air for thermal oxidation. For one hand, the produced micro/nano structures serve as the precursor which can drive the subsequent growth of oxide NWs under thermal oxidation. For another hand, the micro/nano structures can provide a pinning and protection effect for the oxide NWs to prevent them from exfoliation. Cracking or even exfoliation of the oxide NWs from Cu substrates is a common problem during thermal oxidation process. The cracking and exfoliation is mainly caused by the thermal stress induced by the volume misfit between the oxide layer and the metal substrate.29,30 Through introducing precursor micro/nano structures beforehand, the thermal stress 5989


Letter


Nano Letters

the top surface.29 For another aspect, the limited lateral dimension of the Cu microstructure results in limited Cu atom source to diffuse to the lateral surface, whereas the bulk Cu substrate offers abundant Cu atom source to diffuse to the top surface; thus the amount of Cu atoms or ions available on the lateral surface is relatively more insufficient, which can favor the transition of cuprous oxide to cupric oxide. It is reported that any factor that favors the formation of CuO phase may enhance the growth of NWs.32 Comprehensively, the formation of oxide NWs preferentially takes place on the periphery of the Cu microstructures. The bulk Cu substrate, the laser fabricated microcones and induced nanoscale features, together with the oxide NWs, construct a unique macro−micronano−nanowire hierarchical structure. Similar hierarchical structures have been prepared on Si33 and Pt34 surfaces via two-step or three-step femtosecond laser irradiation in water. Specifically, a unique hierarchical structure with nanometer-scale rods of 50 nm width grown normally to the surface of micrometer-scale spikes of 5−10 μm width has been presented on Si surface.33 Different from those previous studies, the hierarchical structure in the present research is composed of metallic microstructures as well as semiconductor NWs, making it a good candidate for the applications demanding any performance combination between metal and semiconductor. Figure 3a shows the clear contour and growth orientation of the NWs, where composition examination was conducted. Only

gets more free space to release during thermal oxidation. Consequently, the cracking and exfoliation of the oxide NWs from the Cu substrate can be prevented. Through this oxidation process, oxide NWs are expected to grow along the skins of the laser produced micro/nano structures as shown Figure 1d. Suggested thermal oxidation parameters can be 500 °C for 2 h. The relatively large scale Cu micro structures, e.g., microcones, serve as the skeleton of the whole surface architectures. The microcones can play the role of geometrical light trapping and induce multi internal reflections to some extent, while the NWs together with the oxide layer surrounding the microcones constitute the transition medium from Cu substrate to the free space. Comprehensively, high-performance AR effect can be achieved, as sketched schematically by the differences in the thickness of the light arrows in Figure 1c,d. Under optimized ultrafast laser processing parameters and with a scanning interval of 40 μm, uniformly distributed microcone arrays with a periodicity of about 40 μm were fabricated on Cu surface, among which were regular microholes (Figure 2a). All of the cone and the hole features have clear

Figure 2. Microstructure characterization of the micro/nano structures as well as the NWs. a,b, SEM and 3D-OM images for ultrafast laser structured Cu surface. c,d, SEM images for structured Cu surface with NWs in low and high magnification. Scale bars in a, c, and d equal 20, 20, and 5 μm, respectively.

contour profiles which can be identified more clearly from the 3D digital optical microscope image (Figure 2b). Plenty of nanoscale features, e.g., nanoparticles and nanocorrugations, covering the surfaces of the microcones have been found during a detailed SEM examination, which are not shown here for the sake of briefness. Nanoscale feature formation is a typical result of the ultrafast laser interaction with metals.31 After thermal oxidation, the microcones turn to be fluffy, while the microholes turn to be blurry, as shown in Figure 2c. Through observation of the fluffy cones in higher magnification (Figure 2d), NWs are found to indeed exist around the cone surfaces, just as depicted in Figure 1d. The NWs radially grow from the outer surface of the microcones to the free space with a dense and uniform distribution, reaching a length over 10 μm. At the lateral surface of the cones, the NWs are more developed than at the top skin of the cones. A similar phenomenon has also been observed during thermal oxidation of Cu grid27 as well as porous Cu substrate.29 For one aspect, the lateral surface of the Cu microstructure can get more stress accumulation and correspondingly more driving force for the NWs’ growth than

Figure 3. Composition detection of the NWs. a, High magnification SEM image for NWs growing around the Cu microcones. b, Compositional line profiles of Cu and O recorded at the yellow point shown in a. c, d, Elemental mapping of Cu Lα 1 and O Kα 1 contained in the NWs. All scale bars equal 500 nm.

two kinds of elements, i.e., Cu and O, were detected via the energy-dispersive X-ray spectroscopy (EDS) point analysis of a typical NW and the atomic percentage of both elements are approximately equal (Figure 3b). Based on the location and intensity of the signal in the elemental mapping, the distribution of element Cu fits in well with that of O, and both of the two elements spread uniformly along the entire skeleton of the NWs (Figure 3c,d). No obvious contrast 5990


Letter

Nano Letters

at the infrared regime, this corresponds to a 98% absorbance. Thus, a near-unity infrared AR property has been achieved. The spacing between the adjacent microcones is larger than the incident wavelength investigated in this research; therefore, geometrical light trapping and multi-internal reflections can be aroused. The number of the internal reflections before the light escapes from the sample surfaces can be estimated using the same method described in detail in ref 35. Then the reflectance for one single bounce of light at the structure surfaces, namely, R0 in this research, was calculated through exponent arithmetic. The R0 reflectance curves can give us more clues about the mechanisms how the AR properties are induced. As shown in Figure 4a, both the calculated R0 curves for the microcone arrays without and with NWs exhibit a similar evolution trend with the reflectance curves directly measured from their corresponding surfaces. The difference between the R0 curve of the microcone arrays without NWs and the reflectance curve of the polished Cu surface results from the sub-nanoscale features covering the microcones as mentioned previously. Meanwhile, the further decrease of reflectance from the R0 curve of the microcone arrays without NWs to that with NWs is attributed to the enhanced AR effect triggered by the oxide NWs. Rigorous simulation has been conducted to further reveal the AR mechanisms of the NWs with the Chandezon method.36 Two-dimensional sinusoidal gratings were adopted to represent the microcone arrays. In order to achieve good convergence, simulation parameters have been optimized to be as follows: grating period d = 40 μm, grating depth h = 30 μm, CuO NW layer thickness t = 5 μm, incident wavelength λ = 7−25 μm, normal incidence. Although the dimensions of the gratings and the NW layer thickness may have a departure from those of the real fabricated structures, we believe that the evolution tendency should be analogous. On the whole, the NW layer is treated as an effective medium. From the perspective of the structural feature of the NW first, the refractive index of this effective medium is estimated by neff = f + (1 − f)nCuO, where f is a filling factor of air ranged in [0, 1]. Different filling factors have been tested, with a typical spectrum shown in Figure 4b (the magenta line with star symbols). It can be seen that, when the microcones are coated with oxide layer, obvious reflection reduction can only appear at a specific wavelength band in the simulated spectrum, which is close to the intrinsic absorption peaks of CuO.37,38 Little changes in the reflection spectra have been observed when f varied (not shown here). Thus, it is indicated that the gradient refractive index was not the main role the CuO NWs played for infrared AR performance. This is an intelligible result given that the refractive indices of copper oxides themselves, i.e., cupric oxide and cuprous oxide, are already much lower than that of Cu in infrared regime39−41 (see Supporting Information, Figure S2), and further reducing it by filling air can exert no significant influence. In other words, the gradient refractive index profile from the Cu substrate to the free space mainly relies on the intrinsic physical properties of the copper oxides rather than the NW structure form. Next, when setting f = 0.2, we regard the function of the oxide NWs as a scattering layer which may dissipate partial of the energy of incident photons by exciting phonons. This effect can be dealt with in the simulation by assuming an imaginary part k (i.e., the extinction coefficient) to neff so that neff = 0.2 + 0.8nCuO + ik. Different values of k from 0.02 to 0.06 have been considered for simulation. It can be seen from Figure 4b that the simulated reflection spectra decrease gradually with the


between Cu and O was observed, which further indicates an equivalent content of both elements in the NWs. Thus, a main phase constitution of cupric oxide is confirmed in the NWs. Xray diffraction characterizations also certify the appearance of CuO phase after thermal oxidation (Supporting Information, Figure S1). It is speculated that the newly appeared CuO phase is mainly contained in the NWs. The antireflection properties of the Cu samples in the mid-IR region (2.5−25 μm or 4000−400 cm−1) were measured by a Bruker Tensor-Fourier transform infrared (FTIR) spectroscope with an A562 integrating sphere, as shown in Figure 4a. The

Figure 4. Broadband near unity infrared antireflection via the NWs grown on micro/nano structures. a, The red and blue solid lines correspond to the measured hemispherical reflectance as a function of wavelength for structured Cu surfaces without and with NWs, respectively. For comparison, the measured hemispherical reflectance of polished Cu surface is also shown as the black solid line. The red and blue dashed lines plot the derived reflectance for one single bounce of light at the structure surfaces without and with NWs, respectively. b, Simulated reflection spectra for uncoated microcones and those coated by oxide layers with different artificial set increments in the imaginary part of refractive index.

polished Cu surface without any AR structures displays high reflectance throughout the mid-IR regime. With ultrafast laser fabricated microcone arrays, the hemispherical reflectance of Cu surface shows a 20−30 percentage-point decrease compared to the polished one. Yet the surface reflection still keeps at an approximately equal level within the whole studied spectrum range, with the reflectance curve remaining to be a steady line. After further growing oxide NWs, the hemispherical reflectance of the sample surface declines abruptly over the whole studied spectrum range, and some deep reflectance valleys emerge. In particular, we focus on the feature at around 15−16 μm where the reflectance drops to a minimum value of ∼2%, corresponding to a reduction by a factor ∼38 with respect to the microcones without NWs and by a factor ∼48 with respect to the polished Cu surface. Since the bulk Cu sample is opaque 5991


Letter


Nano Letters increase in k and get more and more analogous with the measured spectrum in Figure 4a. Thus, the contribution of the NWs to the near-unity AR performance gets clarified. It is also suggested that only the presence of the oxide layers is not enough, and the NW structure form of the oxide is fatal to satisfying AR performance. Actually, the phonon-assisted optical absorption is a general phenomenon for semiconductors, and the interaction of photons with phonons usually just occurs in the infrared region.42,43 For the semiconductor NWs, in particular, the surface effect, small size effect, and quantum size effect can increase the rate of phonon scattering and help the energy transfer from the photons to the phonons as well as the dissipation of the energy from the phonons to random vibrations of lattice. As a result, better infrared AR properties can get realized compared to the bulk or dense oxide layers. As a comparison, the reflection spectra of planar Cu surfaces were also simulated (Supporting Information, Figure S3). Overall, the reflection spectra from planar Cu surfaces present similar evolution trends with those in Figure 4b, supporting the above analysis on the AR mechanism of the oxide NWs. Besides, all of the reflection spectra from planar Cu surfaces are higher than their counterparts on structured Cu surfaces, indicating that there is an amplification effect of the microcones with regard to the AR function of the NWs. Indeed, it is reasonable to regard the whole macro−micronano−nanowire hierarchical structure as an effective medium between the Cu substrate and the free space, which can not only significantly diminish the surface reflection but also broaden the AR wavelength band, thus giving rise to a remarkably improved AR performance. The macro−micronano−nanowire hierarchical structures as well as their antireflection properties can be facilely tuned through simply control the laser processing procedure. For instance, different laser scanning intervals, e.g., 30 and 50 μm, have been used to produce microstructure arrays with varied periodicities. From Figure 5a,b, it is shown that not only the dimensions and the periodicities of the micro structure arrays but also the morphology of the single structure unit are

changed as the laser scanning interval varies. Microcones with denser distribution in contrast with those in Figure 2a and micro petals are formed under laser scanning intervals of 30 and 50 μm, respectively. The change in the structure morphology is ascribed to the property of the Gaussian laser beam and the change in the area of laser directly processed zones on sample surfaces. Encouragingly, NWs grow up on both kinds of microstructure arrays after thermal oxidation regardless of the structure morphology, as shown in Figure 5c,d. Similar to the condition under the scanning interval of 40 μm, the NWs grow radially out of the outer profiles of the laser-produced microstructures and reach a considerable length (Supporting Information, Figure S4). The reflection spectra measurements for the denser microcones and the micropetals in Figure 6a show analogue results as

Figure 6. Broadband near unity infrared antireflection via the modified structures. a, Hemispherical reflectance as a function of wavelength for structured Cu surfaces with periodicity of 30 μm (magenta lines) and 50 μm (olive lines). The solid, dashed, and dash−dot lines correspond to the measured reflectance of structures without NWs, the calculated reflectance of structures with NWs, and the measured reflectance of structures with NWs, respectively. b, The average reflectance in different wavelength ranges of structured Cu surfaces with NWs with a periodicity of 30 μm. Inset shows the measured minimum reflectance which locates in the range of 15−20 μm.

in Figure 4, in consideration of the difference between the structures without and with NWs. Concretely, the reflectance curves of the microstructures without NWs perform steady evolution trends, while those with NWs drop deeply to show a couple of reflectance valleys. However, the reflection spectra for both the denser microcones and the micro petals are apparently lower than those in Figure 4. Particularly, the reflection spectra for both structures integrated with NWs descend to below 10% in a broad wavelength range. The denser microcones and the micropetals perform decreased surface reflection compared to the microcones in Figure 2 through similar mechanisms: the denser microcone arrays have a smaller spacing between the adjacent cones, while the micropetals have a unique morphology; both structure forms can render the sample

Figure 5. Microstructure characterization of the modified structures. a,c, SEM images for structured Cu surfaces without and with NWs with a periodicity of 30 μm. b,d, SEM images for structured Cu surfaces without and with NWs with a periodicity of 50 μm. All scale bars equal 20 μm. 5992


Letter


Nano Letters

outer NWs to the surface micro/nanostructures and finally to the original bulk substrate was formed. Therefore, the hierarchical structure wholly built up a more impactful effect medium, which can significantly enhance the AR properties. Moreover, due to the isotropic and symmetric distribution of the large-scale Cu micro structure arrays, this broadband nearunity infrared AR property shall be free from any angle or polarization sensitivity.16 Correspondingly, the hemispherical reflectance at oblique incident angles of the Cu microcone structures with a periodicity of 40 μm with NWs has been simulated. Besides, specular reflectance of the same Cu sample in the mid-IR region has been experimentally measured by a PerkinElmer Frontier FT-IR/FIR spectrometer, whose specular reflectance accessories allow the measurement at variable incidence angles above 30°. It can be found that neither the simulated hemispherical reflectance spectra nor the measured specular reflectance spectra have shown any apparent change as the incident angle increases to even 60° (Supporting Information, Figure S6). Thus, it demonstrates that the fabricated hierarchical structure in present study can show unique infrared AR property independent of the angles of incidence. The achieved AR performance through our approach covers the working wavelength band of infrared horizon sensors (14−16 μm) and right adjacent to the atmospheric transparency window (8−14 μm), making it potentially useful for infrared detection, sensing as well as imaging applications. The macro−micronano−nanowire hierarchical AR structure demonstrated here requires no conventional nanofabrication techniques (e.g., lithography, wet etching) and thus can easily be implemented to cover a large area. Furthermore, the NWs are under protection of the microcone skeleton, avoiding any durability concerns. Both the highly ordered large-scale microcones and the in situ grown NWs can be reproduced conveniently, enabling its use as a reconfigurable system. The material simplicity, the structure robustness as well as the architecture reconfigurability represent significant advantages for practical usage. Additionally, the combination of the microcone skeleton with the NWs provides us more tunable variances for achieving desired antireflection performance in desired spectral ranges. Therefore, we believe the presented methodology should also be effective for the THz, microwave, and other low-frequency regimes. Finally, our approach provides a general and convenient route for assembling different kinds of materials to a variety of building blocks with novel structures and functions. In particular, it should be universal to the integration of other kinds of semiconductors with metal substrates for wide applications in light management.

surfaces stronger geometrical light trapping effects and thus realize lowered reflectance. For the denser microcone arrays with NWs, a minimum reflectance value of ∼0.6% at around 17 μm has appeared, which is the least experimentally realized result reported to date on metal surfaces, to the best of our knowledge. More importantly, this near-unity infrared AR property is not obviously wavelength sensitive; namely, it is a broadband spectrum feature with the hemispherical reflectance kept steadily below 3% over the range of 14−18 μm. From Figure 6b, it can be further found that the arithmetical average of the measured reflectance within the 10−15 μm, 15−20 μm, and 20−25 μm regimes are all lower than 10%. Even from the perspective of the entire studied spectrum range from 5 to 25 μm, the broadband infrared AR properties maintain effectiveness, with a notable low value of 12.5% being reached. This indicates that nearly 90% of the surface reflection has been successfully eliminated via the unique macro-micronano− nanowire hierarchical AR structure. To better understand the conditions leading to broadband near unity antireflection, a set of calculations to get the estimated reflectance of the denser microcones and micropetals with NWs are performed. Specifically, the R0 for the micro structures without NWs (R0withoutNW) and with NWs (R0withNW) in Figure 4 were adopted. The numbers of the internal reflections were estimated by the logarithm of the measured reflectance of both micro structures without NWs (RwithoutNW) with respect to base R0withoutNW. Then, the calculated reflectance of both micro structures with NWs shall be the product of R0withNW to the power of the estimated internal reflection number. It can be seen that the calculated spectra match well with the experimental data across the entire 5−25 μm range for both structures. Especially, the calculated hemispherical reflectance stay steadily below 1% in the 14−18 μm range for the denser microcone arrays with NWs, which further approves the broadband near unity infrared AR performance. Therefore, the proposed mechanism gets confirmed: the oxide NWs can induce phonon dissipation and eliminate the incident photons’ energy to bring out a lower R0; then the reduced R0 gets multiplied through multi-internal reflection among the laser produced microstructures, with a much lower overall surface reflectance being attained. The differences between the reflection spectra of the microstructures with and without NWs give us more evidence for understanding the mechanism. The calculated three differential curves for all the three conditions with structure periodicities of 30, 40, and 50 μm match well with each other (Supporting Information, Figure S5), demonstrating that the NWs play a same role in the three surface AR architectures. The macro−micronano−nanowire hierarchical AR structure reported here combines the phonon dissipation effect of the oxide semiconductor NWs with the light trapping approach of the relatively large-scale microstructures, while entirely fulfilled on one single metallic substrate without any additional materials. The integration of the metal microstructures with the semiconductor NWs in our strategy is clearly distinct from the conventionally reported gradient refractive index coatings where the transition layers, e.g., Si nanotips,4 Au microstructures,8 TiO2/SiO2 nanoporous,5 take effects isolatedly, and thus their AR spectrum is usually restricted to a limited band. Here in our strategy, the semiconductor NWs are in situ bonded to the bulk substrate via the ultrafast laser produced micro/nanostructures. A gradually structure transition from the

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b02141. XRD patterns of ultrafast laser structured Cu with and without NWs; refractive indices (n) and extinction coefficients (k) of Cu, Cu2O, and CuO; simulated reflection spectra of planar Cu surfaces; high-magnification SEM images for NWs growing on the denser microcone and the micropetal structures; differences between the surface reflectance of micro structures with 5993


Letter

Nano Letters

■

(22) Mizuno, K.; Ishii, J.; Kishida, H.; Hayamizu, Y.; Yasuda, S.; Futaba, D. N.; Yumura, M.; Hata, K. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 6044−6047. (23) Lehman, J.; Sanders, A.; Hanssen, L.; Wilthan, B.; Zeng, J.; Jensen, C. Nano Lett. 2010, 10, 3261−3266. (24) Carey, J. E. Ph.D. Thesis, Harvard University, Cambridge, MA, 2004. (25) Vorobyev, A. Y.; Guo, C. J. Appl. Phys. 2008, 104, 053516. (26) Vorobyev, A. Y.; Topkov, A. N.; Gurin, O. V.; Svich, V. A.; Guo, C. Appl. Phys. Lett. 2009, 95, 121106. (27) Jiang, X. C.; Herricks, T.; Xia, Y. N. Nano Lett. 2002, 2, 1333− 1338. (28) Li, X.; Liang, J.; Kishi, N.; Soga, T. Mater. Lett. 2013, 96, 192− 194. (29) Zhang, Q. B.; Xu, D.; Hung, T. F.; Zhang, K. L. Nanotechnology 2013, 24, 065602. (30) Mumm, F.; Sikorski, P. Nanotechnology 2011, 22, 105605. (31) Li, L.; Hong, M.; Schmidt, M.; Zhong, M.; Malshe, A.; Huis In’tVeld, B.; Kovalenko, V. CIRP Ann. 2011, 60, 735−755. (32) Xu, C. H.; Woo, C. H.; Shi, S. Q. Chem. Phys. Lett. 2004, 399, 62−66. (33) Shen, M. Y.; Carey, J. E.; Crouch, C. H.; Kandyla, M.; Stone, H. A.; Mazur, E. Nano Lett. 2008, 8, 2087−2091. (34) Huo, H. B.; Shen, M. Y. J. Appl. Phys. 2012, 112, 104314. (35) Deinega, A.; Valuev, I.; Potapkin, B.; Lozovik, Y. J. Opt. Soc. Am. A 2011, 28, 770−777. (36) Li, L. J. Opt. Soc. Am. A 1999, 16, 2521−2531. (37) Vanithakumari, S. C.; Shinde, S. L.; Nanda, K. K. Mater. Sci. Eng., B 2011, 176, 669−678. (38) Liu, X.; Li, Z.; Zhang, Q.; Li, F.; Kong, T. Mater. Lett. 2012, 72, 49−52. (39) Rakic, A. D.; Djurisic, A. B.; Elazar, J. M.; Majewski, M. L. Appl. Opt. 1998, 37, 5271−5283. (40) Ordal, M. A.; Bell, R. J.; Alexander, R. W.; Long, J. L. L.; Querry, M. R. Appl. Opt. 1985, 24, 4493−4499. (41) Palik, E. D. Handbook of Optical Constants of Solids II; Academic Press: New York, 1991. (42) Cahill, D. G.; Braun, P. V.; Chen, G.; Clarke, D. R.; Fan, S.; Goodson, K. E.; Keblinski, P.; King, W. P.; Mahan, G. D.; Majumdar, A.; Maris, H. J.; Phillpot, S. R.; Pop, E.; Shi, L. Appl. Phys. Rev. 2014, 1, 011305. (43) Cahill, D. G.; Ford, W. K.; Goodson, K. E.; Mahan, G. D.; Majumdar, A.; Maris, H. J.; Merlin, R.; Phillpot, S. R. J. Appl. Phys. 2003, 93, 793−818.

and without NWs; reflection spectra in varied incident angles (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.


■

ACKNOWLEDGMENTS We acknowledge the support by the National Key Basic Research and Development Program of China (Project No.2011CB013000), the National Natural Science Foundation of China (Project No. 51210009, No. 11474180). We are very grateful to Lifeng Li at the Department of Precision Instrument, Tsinghua University for providing the numerical codes of the Chandezon method.

■

REFERENCES

(1) John, S. Nat. Mater. 2012, 11, 997−999. (2) Raut, H. K.; Ganesh, V. A.; Nair, A. S.; Ramakrishna, S. Energy Environ. Sci. 2011, 4, 3779−3804. (3) Kats, M. A.; Blanchard, R.; Genevet, P.; Capasso, F. Nat. Mater. 2012, 12, 20−24. (4) Huang, Y.; Chattopadhyay, S.; Jen, Y.; Peng, C.; Liu, T.; Hsu, Y.; Pan, C.; Lo, H.; Hsu, C.; Chang, Y.; Lee, C.; Chen, K.; Chen, L. Nat. Nanotechnol. 2007, 2, 770−774. (5) Xi, J. Q.; Schubert, M. F.; Kim, J. K.; Schubert, E. F.; Chen, M.; Lin, S.; Liu, W.; Smart, J. A. Nat. Photonics 2007, 1, 176−179. (6) Zhu, J.; Yu, Z.; Burkhard, G. F.; Hsu, C.; Connor, S. T.; Xu, Y.; Wang, Q.; McGehee, M.; Fan, S.; Cui, Y. Nano Lett. 2009, 9, 279−282. (7) Diedenhofen, S. L.; Vecchi, G.; Algra, R. E.; Hartsuiker, A.; Muskens, O. L.; Immink, G.; Bakkers, E. P. A. M.; Vos, W. L.; Rivas, J. G. Adv. Mater. 2009, 21, 973−978. (8) Zhang, S.; Li, Y.; Feng, G.; Zhu, B.; Xiao, S.; Zhou, L.; Zhao, L. Opt. Express 2011, 19, 20462−20467. (9) Rephaeli, E.; Fan, S. Appl. Phys. Lett. 2008, 92, 211107. (10) de Jong, M. M.; Sonneveld, P. J.; Baggerman, J.; van Rijn, C. J. M.; Rath, J. K.; Schropp, R. E. I. Prog. Photovoltaics 2014, 22, 540−547. (11) Escarre, J.; Soederstroem, K.; Despeisse, M.; Nicolay, S.; Battaglia, C.; Bugnon, G.; Ding, L.; Meillaud, F.; Haug, F.; Ballif, C. Sol. Energy Mater. Sol. Cells 2012, 98, 185−190. (12) Tang, G.; Hourd, A. C.; Abdolvand, A. Appl. Phys. Lett. 2012, 101, 231902. (13) Landy, N. I.; Sajuyigbe, S.; Mock, J. J.; Smith, D. R.; Padilla, W. J. Phys. Rev. Lett. 2008, 100, 207402. (14) Teperik, T. V.; Garcia De Abajo, F. J.; Borisov, A. G.; Abdelsalam, M.; Bartlett, P. N.; Sugawara, Y.; Baumberg, J. J. Nat. Photonics 2008, 2, 299−301. (15) Yao, Y.; Yao, J.; Narasimhan, V. K.; Ruan, Z.; Xie, C.; Fan, S.; Cui, Y. Nat. Commun. 2012, 3, 664. (16) Aydin, K.; Ferry, V. E.; Briggs, R. M.; Atwater, H. A. Nat. Commun. 2011, 2, 517. (17) Spinelli, P.; Verschuuren, M. A.; Polman, A. Nat. Commun. 2012, 3, 692. (18) Liu, N.; Mesch, M.; Weiss, T.; Hentschel, M.; Giessen, H. Nano Lett. 2010, 10, 2342−2348. (19) Kats, M. A.; Sharma, D.; Lin, J.; Genevet, P.; Blanchard, R.; Yang, Z.; Qazilbash, M. M.; Basov, D. N.; Ramanathan, S.; Capasso, F. Appl. Phys. Lett. 2012, 101, 221101. (20) Watts, C. M.; Liu, X.; Padilla, W. J. Adv. Mater. 2012, 24, OP98−OP120. (21) Sondergaard, T.; Novikov, S. M.; Holmgaard, T.; Eriksen, R. L.; Beermann, J.; Han, Z.; Pedersen, K.; Bozhevolnyi, S. I. Nat. Commun. 2012, 3, 969. 5994


Broadband High-Performance Infrared Antireflection Nanowires

Recommend Documents