Light absorption is a fundamental optical process playing significantly important role in wide variety of applications ranging from photovoltaics to photothermal therapy. Semiconductors have well-defined absorption bands with low-energy edge dictated by the band gap energy, therefore it is rather challenging to tune the absorption bandwidth of semiconductors. However, resonant absorbers based on plasmonic nanostructures and optical metamaterials emerged as alternative light absorbers due to spectrally selective absorption bands resulting from optical resonances. Recently, a broadband plasmonic absorber design was introduced by Aydin et al. with a reasonably high broadband absorption. Based on that design, here, structurally tunable, broadband absorbers with improved performance are demonstrated. This broadband absorber has a total thickness of 190 nm with 80% average measured absorption (90% simulated absorption) over the entire visible spectrum (400 - 700 nm). Moreover, the effect of the metal and the oxide thicknesses on the absorption spectra are investigated and results indicate that the shorter and the longer band-edge of broadband absorption can be structurally tuned with the metal and the oxide thicknesses, as well as with the resonator size. Detailed numerical simulations shed light on the type of optical resonances that contribute to the broadband absorption response and provide a design guideline for realizing plasmonic absorbers with structurally tunable bandwidths.
© 2014 Optical Society of America
Manipulation of light matter interactions using nanostructured materials have gained vast amount of attraction recently. Plasmonic nanostructures and optical metamaterials provide enhanced and improved performance in established technologies such as photovoltaics [1–5], surface enhanced Raman spectroscopy, fluorescence spectroscopy and IR absorption spectroscopy [6–9], thermal detectors , and also enable novel devices like superlenses [11–13], left handed materials and invisibility cloaks [14, 15]. By introducing resonant optical response via nanostructured plasmonic materials and optical metamaterials, refractive index can be engineered to have wide range of values from negative values, to n = 0 and even high refractive index values. Specifically, the absorption coefficient is of great importance in designing metamaterial (MM) absorbers and widely investigated throughout the past decade [16–21]. It has been shown both theoretically and experimentally that, by using specific building blocks of certain geometries, one can design MMs that absorb light perfectly at resonant frequencies [22–26]. The possibility of utilizing such plasmonic structures as nanoscale heat concentrators has opened new ways in photothermal cancer therapy [27–30] and thermo-photovoltaic applications [31–33]. It is very challenging to increase the bandwidth of resonant plasmonic and metamaterial absorbers due to the resonant behavior. Near unity (almost perfect) absorption have been realized using highly symmetrical nanostructures at the cost of bandwidth [24, 34], whereas by using complicated anisotropic nanostructures one can increase the bandwidth in some cases to almost entire visible region to create a so called black material by using ordinarily transparent dielectrics and/or reflective metals [35–39]. Introducing anisotropy helps eliminating the polarization dependence by breaking the symmetry. The major challenge in designing anisotropic nanostructures is to identify the nature of many different types of optical resonances, and utilize this information for maximizing the absorption over a broad wavelength range. In a recent study, Aydin et al. demonstrated the effectiveness of a specific cross-trapezoid shape in terms of both ultra-broad absorption and polarization independence . However, resonances that are contributing to the observed broadband absorption behavior was not explored in detail, therefore a design guideline is not available for tuning and engineering the absorption bands in such ultrathin broadband plasmonic absorbers. In this study, we investigate the effect of geometrical parameters such as the metal thickness, dielectric thickness and the size of nanostructures on the overall absorption spectra of broadband absorbers and experimentally demonstrate highly absorptive (~80%) ultrathin broadband absorbers at the visible frequency regime, which is performing better than previously reported (71%) broadband absorbers with broader wavelength range and higher overall absorption . Moreover, our present study clearly indicates that the bandwidth and the intensity of the absorption band can be controlled by controlling the material thicknesses, therefore providing great flexibility in designing broadband absorbers with various bandwidths. In particular, we highlight that the absorption band edges can be tuned both at the short and long wavelength edges which cannot be obtained with semiconductor based absorbers.
2. Results and discussion
Proposed ultra-thin metal-insulator-metal (MIM) absorber structure consists of an oxide (SiO2) layer sandwiched between two metal (Ag) layers. The top metal layer is nanostructured to have a specific cross-trapezoid shape while the bottom metal and the oxide layers remain flat as shown in Fig. 1(a). The cross-trapezoid shape is formed by simply crossing two trapezoids at their geometric center. The short and the long edges of the trapezoid are referred in this paper as a and b, respectively.
A commercial-grade simulator based on the finite-difference time-domain (FDTD) method was used for parameter optimization of the MIM absorbers . The parameter set included period (P), bottom metal layer thickness (tbottom), oxide layer thickness (toxide), top metal layer thickness (ttop) and trapezoid parameters a and b as illustrated in Fig. 1(a). The thickness of the bottom metal layer thickness is fixed at 100 nm to suppress the transmission through the MIM multilayer film. The period is at liberty apart from the other parameters. Because, for every period value there is an optimized set of parameters that give the best absorption profile. Therefore, after performing the periodicity sweep as shown in Fig. 1(b), we set the period to 350 nm. Furthermore, the evolution of the bands as a function of the period reveals the nature of the resonances. A significant shift of a resonance as a function of period implies the periodic coupling to propagating modes in the structure such as surface plasmon polariton (SPP) or waveguide modes, whereas a relatively weak dependence generally suggests a localized resonance like localized surface plasmon resonance (LSPR). Observed major modes are traced and indicated in Fig. 1(b). Modes m1-m3 exhibit a rather weak dependence on the periodicity. Therefore we conclude that these modes are either LSPR modes aroused by the distinct nature of the unit cell or Fabry Pérot like modes which do not depend on periodicity. However, modes m4-m6 are a strong function of periodicity. This suggests propagating modes in the structure.
We optimized the geometrical parameters of the trapezoid unit cell after fixing the periodicity at 350 nm. The detailed simulation results and an extensive discussion on the nature of resonances are provided in the appendix A. Here, we present the effect of overall size of the cross-trapezoid on the absorption spectra. This approach is more compatible to the e-beam lithography technique that we used to fabricate our samples. One can easily create different unit cells by simply varying the e-beam exposure using a single mask. It enables us to study disconnected (non-touching) structures, as well. We have defined a size parameter, Δd, as the isotropic length difference relative to a base structure as shown in Fig. 2(a). The base structure has trapezoid height equal to the periodicity. Therefore, when ∆d<0 structures become disconnected and when ∆d>0 we have isotropic expansion of the base structure in xy plane as if we are increasing the e-beam exposure.
Figure 2(b) shows the evolution of the absorption band as a function of Δd. A red reflection band forms between 600 nm to 750 nm gradually as trapezoids become larger (∆d>0). When ∆d-P⁄2 ~0, trapezoids become large enough to fill the entire unit cell and form a continuous metal film. Consequently, the structure becomes a 2D Fabry-Pérot resonator. Guided modes, i.e. mode around 410 nm, disappear when the structures are disconnected (∆d<0). Also, the absorption band gets narrower from both blue and red end of the spectrum. For ∆d + P⁄2 ~0, disconnected structures behave like nanoparticles therefore we observe narrow double resonance which are peaked around 520 nm and 540 nm. Note that, in this limit, nanoparticle is highly anisotropic (not symmetrical as in the case of nanodisks), therefore a complicated resonance is expected.
Figure 3 exhibits a series of micrographs, scanning electron microscopy (SEM) images and corresponding absorption spectra of one of the fabricated sets of the MIM arrays. P = 350 nm for all of the structures. Various sized cross-trapezoids were obtained by using the same mask and varying the e-beam exposure. Corresponding close-up SEM images in Fig. 3(b) reveal the actual structure of each unit cell. There is an inevitable rounding of corners due to the methods used in the fabrication. We performed spectral reflectivity measurements in order to evaluate the absorption characteristics of each structure. Note that, the bottom metal layer is optically thick enough, ensuring that there is no transmission through the MIM film, therefore absorption can be calculated using the formula A = 1-R, where A is the absorbance and R is the reflectance. Measured spectral absorption curves are presented in Fig. 3(c). As the cross-trapezoid unit cell structures get smaller, the absorption tends to increase. There is a dip around 580 nm consequently optical images look greenish. It disappears later as the cross-trapezoids become disconnected. Counter-intuitively, absorption increases with the decreasing fill factor. That means, the incident illumination couples strongly to the MIM structure and suppresses the reflection, therefore optical images look black. Absorption band gets narrower on both short and long wavelength edges when trapezoid nanostructures become smaller. The same trend is apparent in the FDTD calculations shown in Fig. 2(b), as well. The slight mismatch can be attributed to the fabrication imperfections and surface roughness.
We have also investigated the effect of the top metal and the dielectric layer thicknesses on the absorption profile and results are provided in Figs. 4(a) and 4(b) respectively. The top metal thickness has a rather strong effect on localized resonances which heavily depend on size and geometry. It further has an indirect effect on propagating modes due to the change in the effective refractive index. It is evident that the top metal thickness mostly affects bands on the NIR side of the spectrum, thereby allowing a passive way for controlling the long-wavelength edge of the absorption band. The increase of the top metal thickness shifts resonances towards shorter wavelength, an observed behavior for nanodisk arrays ; however the resonance strength is not affected significantly, suggesting an effective refractive index change in the MIM structure. When the top metal is thin enough, i.e. ttop<20 nm, resonances get sparse, spectrally. This causes reflection bands to emerge within the absorption band. Depending on the application, having multiple absorption bands can be an advantage. For a structure with the widest and the highest absorption band, we conclude a top metal thickness of 25 nm. Figure 4(b) shows results of the oxide thickness sweep with a top metal thickness of 25 nm. When the intermediate oxide layer is very thin, there are no waveguide modes, so most of the light is reflected back. The absorption around 400 nm is due to the fundamental Fabry-Pérot mode as can be seen in transfer matrix method calculations of an MIM structure of the same metal and dielectric layer thicknesses given in appendix A Fig. 8. This fundamental mode redshifts with the increased oxide thickness and hybridize with lower energy modes. When the thickness of the oxide layer increases, more waveguide modes are allowed, therefore the resonances become much apparent. For larger than 80 nm thickness of the oxide layer, the high energy modes red shifts due to the increase of the effective refractive index of the MIM structure. On the other hand, the blue-shift in the lower energy modes that are excited by the various parts of the unit cell can be explained by the LC resonator model. The capacitance decreases with the increase of the separation between plates, i.e. oxide thickness. However, the LC resonance frequency increases with the reduction of the capacitance . Hence we observe the blue-shift in lower energy modes. This blue shift in low energy modes along with the red-shift in the Fabry-Pérot mode allows the construction of the broad band response of the plasmonic structure by selecting an appropriate oxide layer thickness.
We have studied the absorption as a function of the top metal and the dielectric thickness experimentally, as well. Results are summarized in Fig. 5(a). Here, we compare the effect of thicknesses of the top metal and the intermediate oxide layer for a certain trapezoid geometry. The black and the green curves represent the measurements from structures of 20 nm and 30 nm top metal thicknesses with a SiO2 intermediate insulating layer thickness of 70 nm. Whereas the blue curve represents a structure of 20 nm top metal thickness and 110 nm SiO2 thickness. All three of the structures have the same unit cell and a bottom metal layer thickness of 100 nm. We observed several peaks in the absorption spectrum which are coalesced to form a single broad absorption band. The strong anisotropy of the structure gave rise to these peaks of different origins, such as localized surface plasmon, propagating surface plasmon polariton and waveguide modes of the MIM structure. As predicted by the FDTD simulations, the thinner top metal (black curve) broadens the absorption band relative to a thicker one (green curve). This relatively higher absorption can be attributed to the increase in the transmission of the top metal thereby enabling more light to couple to the waveguide modes of the MIM structure. In the case of the thicker oxide layer (blue curve), we saw a red shift and reduction of the absorption bandwidth. The thicker oxide layer increases the effective refractive index of the MIM structure. This difference in the effective refractive index causes a red shift in the short wavelength end of the absorption band. In addition, we have performed simulations using digitized SEM images of the fabricated MIM structures to confirm the measurements as well as to account for the imperfections aroused during the fabrication process. The results are illustrated in Fig. 5(b). The unit cell used in these simulations is displayed on the inset of Fig. 5(b) and the area which it was selected is indicated on the SEM image in Fig. 5(d). A mesh override region of 1 x 1 x 1 nm3 on the trapezoid section of the structure was utilized to fine resolve fabrication defects. There is a rather good agreement between the measured and simulated results. The shift in the peak positions are attributed to the difference in material data which is used in the calculations to the actual material data of the fabricated structure. If we consider the visible region (400-700nm) the optimized MIM structure has average absorption of around 0.8 within a 190 nm thickness. We have measured local absorption peaks of over 0.9 at different wavelengths. Whereas the simulations predicted local absorption peaks of over 0.99. Optical microscope images of these three structures also support the measured data. Thick oxide layer structure appears vivid blue color because the rest of the visible spectrum is absorbed in the structure. However, other structures appear much darker. One of them has a greenish hue which is due to a relative decrease in absorption around 550 nm (see green curve in Fig. 5(a)).
In conclusion, we have designed a passive ultrathin metamaterial absorber by using a reflective metal and a transparent dielectric in a certain cross-trapezoid shape. The proposed structure has only 190 nm thickness. It exhibits an absorption bandwidth of larger than 350 nm covering the entire visible spectrum with an average absorption of 80%. We investigated the complex nature of optical resonances occurring upon broad band illumination in the structure using extensive optical simulations. Our results indicate that one can design structurally tunable resonant absorbers with narrow and broadband absorption profiles. Thicknesses of materials used in the absorber such as the top metal and the middle dielectric thickness, as well as the filling factor of trapezoid nanostructures significantly change the absorption spectra. Design guidelines outlined in this study will be quite useful for engineering complex plasmonic absorbers for applications in thermophotovoltaics, photothermal therapy, hot-electron collection devices, thermal emitters, and absorption filters.
Appendix A: Supporting information
A.1 Trapezoid parameter sweep
The absorption as a function of trapezoid edge lengths a and b are represented in Figs. 6(a) and 6(b), respectively. There are two modes which explicitly depend on the respective parameter in each case in the NIR region (bold dashed lines). Two modes are merging together as a becomes larger and as b becomes smaller. This agrees well with the limiting case when a and b is equal there should only be one single resonance. Other modes exhibit relatively weak dependence on a or b which can be attributed to the structural disturbance of the unit cell.
A.2 Absorption maps and the nature of resonances
Full field electromagnetic simulations allow constructing absorption maps and calculating the charge distribution in the structure which is a very effective way to gain an insight to the nature of resonances. Figure 7 exhibits the 3D absorption maps of four major resonances along with the corresponding charge distribution at the oxide - top metal interface. For absorption maps only top 20 nm portion of the bottom metal where almost all of the light absorption occurs is included.
We calculated the absorbed power with the formula in the entire volume of the unit cell and the charge distributions by using Gauss’ Law, at the metal boundaries. There is a dipole like charge distribution and absorption profile at 430 nm. As parameter sweeps shown in Fig. 4 pointed out, the intensity of this mode strongly depends on the thickness of the oxide layer and show rather weak dependence on the other parameters. In addition, the spectral position is almost constant. This concludes that this is a waveguide mode propagating along the MIM structure. The second mode at 520 nm has more complex profile. This mode is aroused by hybridization of different resonances as can be seen in Fig. 4(b). There are two resonances one of which depends on period and one which does not. This suggests that one of them is an SPP mode along the top metal – oxide interface, whereas the other one is a complicated localized mode around wider section of the trapezoid which lies along the polarization direction. The third mode at 670 nm is again a superposition of two resonances. The first one is an SPP resonance which is period dependent, and the second one is a localized resonance which strongly depends on the trapezoid parameter b. Finally, the last major mode at 760 nm has similar characteristics with the third mode. It is composed of one SPP mode plus a localized mode which highly depends on the trapezoid parameter a.
A.3 A snapshot of light interaction with the structure
Figure 9 compares how light interacts with a blank MIM and the trapezoid shaped MIM structures. We have incorporated the intensity and the phase information to calculate wave propagation upon a continuous wave excitation. For a more comprehensible picture, we have suppressed the incident wave. Here only one of the resonances (520 nm) is presented. However similar behavior is also true for other resonances. The plain MIM structure is basically a metallic mirror coated with two extra thin film layers. Therefore it reflects most of the incident light back. Whereas, highly anisotropic and periodically arranged trapezoid patterns couple light to several available modes in the MIM structure. Therefore, the reflection is significantly suppressed. The complexity of the coupled field pattern is another indication of the superposition of numerous resonances in the MIM structure.
This material is based upon work supported by the AFOSR under Award No. FA9550-12-1-0280. KA acknowledges financial support from the McCormick School of Engineering and Applied Sciences at Northwestern University and partial support from the Institute for Sustainability and Energy at Northwestern (ISEN) through ISEN Equipment and Booster Awards. This research was also partially supported by the Materials Research Science and Engineering Center (NSF-MRSEC) (DMR-1121262) of Northwestern University. This research made use of the NUANCE Center at Northwestern University, which is supported by NSF-NSEC, NSF-MRSEC, Keck Foundation, and the State of Illinois and the NUFAB cleanroom facility at Northwestern University.
References and links
4. L. Novotny and N. van Hulst, “Antennas for light,” Nat. Photonics 5(2), 83–90 (2011). [CrossRef]
5. M. E. Solano, M. Faryad, P. B. Monk, T. E. Mallouk, and A. Lakhtakia, “Periodically multilayered planar optical concentrator for photovoltaic solar cells,” Appl. Phys. Lett. 103(19), 191115 (2013). [CrossRef]
7. N. A. Cinel, S. Bütün, G. Ertaş, and E. Özbay, “SERS:‘fairy chimney’‐shaped tandem metamaterials as double resonance SERS Substrates,” Small 9(4), 489 (2013). [CrossRef]
8. R. Adato and H. Altug, “In-situ ultra-sensitive infrared absorption spectroscopy of biomolecule interactions in real time with plasmonic nanoantennas,” Nat Commun 4, 2154 (2013). [CrossRef] [PubMed]
9. A. A. Yanik, A. E. Cetin, M. Huang, A. Artar, S. H. Mousavi, A. Khanikaev, J. H. Connor, G. Shvets, and H. Altug, “Seeing protein monolayers with naked eye through plasmonic Fano resonances,” Proc. Natl. Acad. Sci. U.S.A. 108(29), 11784–11789 (2011). [CrossRef] [PubMed]
10. F. Yi, H. Zhu, J. C. Reed, and E. Cubukcu, “Plasmonically enhanced thermomechanical detection of infrared radiation,” Nano Lett. 13(4), 1638–1643 (2013). [PubMed]
11. Z. Liu, E. Li, V. M. Shalaev, and A. V. Kildishev, “Near field enhancement in silver nanoantenna-superlens systems,” Appl. Phys. Lett. 101(2), 021109 (2012). [CrossRef]
12. H. Liu, B. Wang, L. Ke, J. Deng, C. C. Chum, S. L. Teo, L. Shen, S. A. Maier, and J. Teng, “High Aspect Subdiffraction-limit photolithography via a silver superlens,” Nano Lett. 12(3), 1549–1554 (2012). [CrossRef] [PubMed]
16. Y. Avitzour, Y. A. Urzhumov, and G. Shvets, “Wide-angle infrared absorber based on a negative-index plasmonic metamaterial,” Phys. Rev. B 79(4), 045131 (2009). [CrossRef]
18. H. Tao, N. I. Landy, C. M. Bingham, X. Zhang, R. D. Averitt, and W. J. Padilla, “A metamaterial absorber for the terahertz regime: Design, fabrication and characterization,” Opt. Express 16(10), 7181–7188 (2008). [CrossRef] [PubMed]
19. T. V. Teperik, F. J. Garcia de Abajo, A. G. Borisov, M. Abdelsalam, P. N. Bartlett, Y. Sugawara, and J. J. Baumberg, “Omnidirectional absorption in nanostructured metal surfaces,” Nat. Photonics 2(5), 299–301 (2008). [CrossRef]
20. J. S. White, G. Veronis, Z. Yu, E. S. Barnard, A. Chandran, S. Fan, and M. L. Brongersma, “Extraordinary optical absorption through subwavelength slits,” Opt. Lett. 34(5), 686–688 (2009). [CrossRef] [PubMed]
21. C. Hägglund, G. Zeltzer, R. Ruiz, I. Thomann, H.-B.-R. Lee, M. L. Brongersma, and S. F. Bent, “Self-assembly based plasmonic arrays tuned by atomic layer deposition for extreme visible light absorption,” Nano Lett. 13(7), 3352–3357 (2013). [CrossRef] [PubMed]
22. J. Hao, J. Wang, X. Liu, W. J. Padilla, L. Zhou, and M. Qiu, “High performance optical absorber based on a plasmonic metamaterial,” Appl. Phys. Lett. 96(25), 251104 (2010). [CrossRef]
26. R. Adato, A. Artar, S. Erramilli, and H. Altug, “Engineered absorption enhancement and induced transparency in coupled molecular and plasmonic resonator systems,” Nano Lett. 13(6), 2584–2591 (2013). [CrossRef] [PubMed]
28. A. M. Gobin, M. H. Lee, N. J. Halas, W. D. James, R. A. Drezek, and J. L. West, “Near-Infrared Resonant Nanoshells for Combined Optical Imaging And Photothermal Cancer Therapy,” Nano Lett. 7(7), 1929–1934 (2007). [CrossRef] [PubMed]
29. Y. Ma, X. Liang, S. Tong, G. Bao, Q. Ren, and Z. Dai, “Gold nanoshell nanomicelles for potential magnetic resonance imaging, light-triggered drug release, and photothermal therapy,” Adv. Funct. Mater. 23(7), 815–822 (2013). [CrossRef]
30. J. F. Hainfeld, M. J. O’Connor, P. Lin, L. Qian, D. N. Slatkin, and H. M. Smilowitz, “Infrared-transparent gold nanoparticles converted by tumors to infrared absorbers cure tumors in mice by photothermal therapy,” PLoS ONE 9(2), e88414 (2014). [CrossRef] [PubMed]
31. C. H. Wu, B. Neuner III, J. John, A. Milder, B. Zollars, S. Savoy, and G. Shvets, “Metamaterial-based integrated plasmonic absorber/emitter for solar thermo-photovoltaic systems,” J. Opt. 14(2), 024005 (2012). [CrossRef]
32. C. Simovski, S. Maslovski, I. Nefedov, and S. Tretyakov, “Optimization of radiative heat transfer in hyperbolic metamaterials for thermophotovoltaic applications,” Opt. Express 21(12), 14988–15013 (2013). [CrossRef] [PubMed]
33. S. Molesky, C. J. Dewalt, and Z. Jacob, “High temperature epsilon-near-zero and epsilon-near-pole metamaterial emitters for thermophotovoltaics,” Opt. Express 21(S1Suppl 1), A96–A110 (2013). [CrossRef] [PubMed]
34. N. Mohammadi Estakhri and A. Alù, “Minimum-scattering superabsorbers,” Phys. Rev. B 89(12), 121416 (2014). [CrossRef]
35. C. Argyropoulos, K. Q. Le, N. Mattiucci, G. D’Aguanno, and A. Alù, “Broadband absorbers and selective emitters based on plasmonic Brewster metasurfaces,” Phys. Rev. B 87(20), 205112 (2013). [CrossRef]
36. A. S. Hall, M. Faryad, G. D. Barber, L. Liu, S. Erten, T. S. Mayer, A. Lakhtakia, and T. E. Mallouk, “Broadband light absorption with multiple surface plasmon polariton waves excited at the interface of a metallic grating and photonic crystal,” ACS Nano 7(6), 4995–5007 (2013). [CrossRef] [PubMed]
37. E. E. Narimanov and A. V. Kildishev, “Optical black hole: Broadband omnidirectional light absorber,” Appl. Phys. Lett. 95(4), 041106 (2009). [CrossRef]
38. A. C. Atre, A. García-Etxarri, H. Alaeian, and J. A. Dionne, “A broadband negative index metamaterial at optical frequencies,” Advanced Optical Materials 1(4), 327–333 (2013). [CrossRef]
39. K. Aydin, V. E. Ferry, R. M. Briggs, and H. A. Atwater, “Broadband polarization-independent resonant light absorption using ultrathin plasmonic super absorbers,” Nat Commun 2, 517 (2011). [CrossRef] [PubMed]
40. Lumerical Solutions, Inc.”, retrieved http://www.lumerical.com/tcad-products/fdtd/.
41. Y. B. Zheng, B. K. Juluri, X. Mao, T. R. Walker, and T. J. Huang, “Systematic investigation of localized surface plasmon resonance of long-range ordered Au nanodisk arrays,” J. Appl. Phys. 103, 014308 (2008).