Salisbury Screen Terahertz Absorber Formed with an Insulator: 4-N,N-Dimethylamino-4'-N'-methyl-stilbazolium Tosylate (DAST).

A novel Salisbury screen absorber (SSA) based on a 4-N,N-dimethylamino-4'-N'-methyl-stilbazolium tosylate (DAST) crystalline film was designed and fabricated. Different from the conventional SSA, an insulating organic DAST film replaced the normally highly conductive top surface metal layer. The absorption spectra and the structure of this SSA were optimized with a transmission line model to correspond to the DAST absorption peak at 1.1 THz, whose results were further verified by numerical simulations. If the thickness of the DAST film is 4 μm, a nearly perfect terahertz (THz) absorption is possible with this strategy, whereas the absorption of a bare 4 μm thick DAST film would be 4 times lower. The design allows the terahertz response of this DAST system to be tuned by adjusting either the thickness of the DAST film or the spacer, both of which remain in deep sub-wavelength to broaden the range of applications.

Introduction

Terahertz (THz) waves in the frequency range of 0.1–10 THz, between infrared and millimeter waves, have many applications, such as thermal emitters,[1] terahertz imaging,[2] communication,[3] detectors,[4] and filters.[5] For detection, imaging, and especially nonlinear optics, spectral-selective THz absorption needs to be maximized. Many designs, such as metamaterial absorbers,[6−10] resistive frequency selective surface absorbers,[11] and Salisbury screen absorbers (SSAs),[12,13] demonstrate the general feasibility of enhanced absorption but not necessarily the general suitability for tasks with different constraints. For example, metamaterial absorbers, constructed by planar metallic structures, can achieve perfect absorption near their resonance frequencies but often come with the cost of a complicated fabrication process and limited scalability. In contrast, SSAs are relatively simple and readily scalable, excellent in performance, and relatively easy in their manufacturing process. In terms of fabrication process, typical structures of SSAs are similar to the three-layer metamaterial structures but without the sub-wavelength lithographic patterning. They come with a particular feature, namely, their top Salisbury surface is made of highly conductive metal,[13] CNTs,[16] graphene,[12] or other materials, such as VO and a-Si,[14] for impedance matched to free space (e.g., 377 Ω).[12] However, in the terahertz regime, the thickness of the surface metal in a SSA should be controlled to be at the nanometer scale for impedance matching, which is challenging in large-scale fabrication of devices. Therefore, study results on novel and feasible SSAs operating at terahertz frequencies will be considerably attractive.

On the other hand, 4-N,N-dimethylamino-4′-N′-methyl-stilbazolium tosylate (DAST) has become one of the most important and successful organic nonlinear optical materials since it was introduced in 1989.[15] Although DAST is not conductive, like most organic crystals, hence seemingly out of Salisbury Screen’s reach, its equivalent optical conductivity that can be derived from its refractive index and extinction coefficient[16] is finite. At its strong terahertz absorption frequency of 1.1 THz, corresponding to the anion–cation pair vibration in the optical phonon mode,[17] the optical conductivity of DAST is expected to be high. This suggests the possibility of replacing the conductive materials on the top of SSAs by DAST films and thus leading to a possible enhanced absorption without increasing the thickness in a Salisbury configuration. Increasing absorption without increasing thickness could be particularly meaningful in the cases of nonlinear optics and thin-film flexible electronics, e.g., applications of the DAST-based SSAs in rationally enhancing the efficiency of the second-harmonic generation of DAST or in enhancing the light absorption for the terahertz detectors, etc. Although graphene-based SSAs have been shown to achieve near-perfect absorption in the terahertz range,[12] graphene is not known for its optical nonlinearity. But according to the equation σt = 2ωε0nDκDt described in ref (16), it can be estimated that the equivalent optical conductance of a 1 μm thick DAST film at 1.1 THz is around 6.47 × 10–5 S, close to that of a single-layer graphene (6.605 × 10–5 S).[18] It therefore seems rational to attempt to construct a DAST-based SSA to operate at around 1.1 THz, as we set out to explore the feasibility in this work. However, in such an attempt, one must also deal with the extra phase associated with the likely finite DAST thickness and its anisotropy.

In this work, a DAST-based SSA was designed and fabricated. Starting with the conventional three-layer SSA structure, we investigated and optimized its absorption performance by the transmission line model (TLM). The TLM model provided a parametric insight in to the design optimizations, results of which were further validated by numerical simulations with the Computer Simulation Technology (CST) software. Particularly, the as-simulated absorption spectra were also experimentally compared.

Results and Discussion

Theoretical Model

In transmission line terminologies, the absorption of the wave of a wave vector k can be obtained as followswhere S11 is the reflection coefficient, S21 is the transmission coefficient, R = |S11|2 is the reflectivity, and T = |S21|2 is the transmittance. If the thickness of the substrate metal exceeds the skin depth at terahertz, the transmittance of the terahertz wave is zero, that is, T = |S21|2 = 0.

For broader applications, the DAST-based SSA was designed to be simple in this work, as illustrated schematically in Figure a. More details about the design are shown in the Experimental Section. For the designed DAST-based SSA, the electromagnetic response can be modeled by the equivalent transmission line as shown in Figure b, where ZL and Z are the impedances of the bottom Al layer and medium dielectric spacer layer, respectively; d is the thickness of the intermediate spacer layer; Z0 and Zin′ are the vacuum impedance and the impedance of the whole SSA, respectively; and Rs is the equivalent optical sheet resistance of DAST.

Figure 1

(a) Schematic illustration of the designed DAST-based Salisbury screen absorber and (b) its equivalent circuit model.

In theory, the combined impedance of the medium dielectric layer and the bottom metal layer can be expressed as[19]where is the magnitude of the wave vector, the wavelength λ is the free space wavelength, n is the refractive index of the spacer region, and θ is the incidence angle of light in the spacer region; under normal incidence, θ = 0.

Aluminum (Al) metal is a perfect electric conductor for the terahertz waves. At a thickness much larger than the skin depth (about 120 nm for Al at 1.1 THz),[16] the incident terahertz wave can not penetrate through the metal. In this case, the equivalent impedance of the bottom Al layer for the designed SSA can therefore be taken asThen, the input impedance isThe final equivalent impedance of the Salisbury screen iswhere ZR is the impedance of the top DAST thin film.

Since DAST is an insulator, it has only an equivalent optical conductivity of[16]where ω = 2πf and f is the frequency of the incident electromagnetic wave; ε0 is the vacuum permittivity; and εr″ = 2nDκD, where εr″, nD, and κD are the frequency- and polarization-dependent relative permittivity, refractive index, and extinction coefficient of DAST, respectively. The equivalent optical sheet resistance of DAST is given by[16]where t is the thickness of the top DAST film.

The reflection coefficient of the transmission line and thereby the absorbance of the DAST-based SSA under normal incidence condition can be expressed as[19]where Y0 is the vacuum admittance and k is the wave vector. The absorption spectra as functions of wavelength and thickness were investigated, for polarization along DAST’s a-axis and b-axis, respectively.

Simulation and Calculation

First, we investigated the optimal structural parameters for the designed DAST-based SSA. The simulation results on the effects of the thickness of the space layer on the absorption of the DAST-based SSA are displayed in Figure . The thickness of the top DAST single-crystal film was set at 1 μm, whereas the refractive index of the space layer was fixed at 1.38. The permittivities of DAST film along the a- and b-axes are shown in Figure S1 in the Supporting Information. The method for the simulations of DAST-based SSA along the a- and b-axes by CST is described in the Experimental Section. As shown in Figure , if the thickness of the space layer is smaller than 50 μm, both the absorptions of the SSA along the a- and b-axes increase with the increase of the thickness of the space layer. But if the thickness of the space layer is larger than 50 μm, both the absorptions along the a- and b-axes decrease, suggesting optimal terahertz absorption performance observed at the spacer thickness of 50 μm. This also suggests that the resonant frequency of the F–P cavity with the 50 μm thick space layer corresponds to 1.1 THz. On the other hand, the optimal spacer thickness can also be simply estimated through the lumped-element TLM according to eq , where the maximum absorption is achieved at Y0 – 2ωε0nDκDt = 0 and nY0 cot(kd) = 0. For the intermediate spacer layer, kd = π/2 implies that nd should be a quarter wavelength. When the resonance frequency is set at the targeted frequency of 1.1 THz, an optimal spacer thickness of 49 μm, close to that (50 μm) deduced by CST simulations (Figure ), was thus estimated.

Figure 2

CST-simulated results at different thicknesses of the intermediate spacer along the a-axis (a, b) and along the b-axis (c, d) of the top DAST single crystal, respectively.

Figure also reveals that the second absorption peak at ∼3 THz shifts to lower frequency (red-shifts) with the increase of the thickness d of the intermediate spacer layer. The simulated absorption frequencies near 3 THz under different spacer layer thicknesses are plotted in Figure . When the spacer thickness is increased from 40 to 70 μm, the absorption peak of the SSA near 3 THz red-shifts from 3.7 to 2.3 THz along the a-axis, whereas that along the b-axis red-shifts from 3.6 to 2.3 THz. The red-shift of the peak at ∼3 THz can be attributed to the effects of the F–P resonant absorption mode. Since the thickness (1 μm) of the DAST single-crystal film is much thinner than that (30–70 μm) of the spacer layer, the optical path is mainly derived from the latter. Accordingly, the F–P resonant frequency can be expressed aswhere f is the frequency of the absorption peak and d and n are the thickness and refractive index of the space layer, respectively. Equation reveals that the peak frequency is reciprocal to the thickness d of the intermediate spacer layer, agreeing well with the simulated results in Figure .

Figure 3

Frequency of the absorption peak near 3 THz at different spacer thicknesses along the (a) a-axis and (b) b-axis of the top DAST crystal, respectively.

According to the above CST simulation results, the optimal intermediate spacer thickness was set as 50 μm. Accordingly, the effects of the top DAST layer thickness on the absorption of the DAST-based SSA were further investigated via both the CST simulation and TLM calculation. For comparison, both the CST-simulated and TLM-calculated spectra as a function of the thickness for the top DAST layer are plotted in Figure . Figure reveals that two absorption peaks at 1.1 and 3.1 THz are observed along the a-axis whereas four absorption peaks at 1.1, 1.3, 1.6, and 3.0 THz appear along the b-axis.[20] As the thickness of the DAST top layer is increased from 0.5 to 4 μm, both the absorptions polarized along the a- and b-axes are increased, and particularly, a high absorption of ∼92.9% is seen along the a-axis at 4 μm (TLM results). As shown in Figure , the spectra in the 0–2 THz range calculated by TLM are in good agreement with those simulated by CST. However, the absorption peaks along the a-axis at ∼3 THz simulated by CST significantly shift to lower frequencies as the DAST thickness increases whereas those along the b-axis only slightly shift (Figure ). Rather differently, both the absorption peaks at ∼3 THz along the a- and b-axes calculated by TLM remain unchanged (Figure ).

Figure 4

Absorption spectra of the SSAs under different thicknesses for the top DAST film along the (a) a-axis and (b) b-axis obtained by TLM calculations (dashed lines) and CST simulations (solid lines), respectively.

In the CST simulations, the red-shift of the absorption peaks at ∼3 THz is attributed to two factors: the great anisotropy of the DAST (absorption coefficient of ∼100 cm–1 along the a-axis and that of ∼300 cm–1 along the b-axis near 3 THz)[20] and the second resonant absorption mode at λ2 (∼λ1/3), as emerging from the F–P cavity formed by the top DAST film and the bottom metal layer. This resonance condition can be estimated analytically. Under normal incidence, the phase difference of the round-trip optical path iswhere n and d are the refractive index and thickness of the intermediate spacer layer, nD and t are the refractive index and thickness of the top surface DAST layer, and φ0 is the phase change of light reflected from the metal substrate, respectively. As the thickness t of the DAST film increases, the contribution of the DAST film to the phase change becomes more and more important, especially relative to the higher frequency (shorter wavelength) mode at ∼3 THz. This leads to a significant red-shift of the peak along the a-axis at ∼3 THz with the increase of the thickness t of the DAST layer. However, DAST crystal exhibits a higher absorption coefficient along the b-axis than along the a-axis at ∼3 THz,[20] suggesting that the absorption of the DAST-based SSA at ∼3 THz along the b-axis would be dominated by the intrinsic absorption of DAST crystal, thus leading to only slight red-shifts of the peak at ∼3 THz seen along the b-axis (Figure b). The case is different for the TLM calculations, where the DAST is lumped into a single sheet resistance Rs without a DAST layer thickness-dependent phase, and thus the absorption peaks at ∼3 THz remain unchanged with the increase of the DAST thickness.

For assessing the enhancement extent of the absorption due to the Salisbury strategy, the terahertz spectra simulated by CST for a 4 μm thick DAST single-crystal film and an SSA constructed by a 4 μm thick DAST single-crystal film as the top surface layer are compared in Figure . It shows that the absorptions at 1.1 THz for the as-designed SSA along the a- and b-axes of DAST crystal are 86.9 and 68.9%, about 3.4 and 4.4 times higher than those for the bare DAST film with the same thickness, respectively.

Figure 5

CST-simulated spectra of a SSA constructed by a 4 μm thick DAST single-crystal film and a bare 4 μm thick DAST single-crystal film along the (a) a-axis and (b) b-axis of DAST crystal, respectively.

Experimental Results

According to above TLM calculation and CST simulation results, we fabricated the as-designed DAST-based SSAs. To overcome the difficulty in the preparation of uniform DAST single crystal with a precise and desirable thickness, DAST polycrystalline thin films were prepared in this work. Figure shows the metallography micrographs of the spray-coated DAST polycrystalline films with the thicknesses of 2 and 4 μm, respectively, both of which were used in the fabrication of the designed SSAs. Compared with the 2 μm thick DAST film (Figure a), the 4 μm thick DAST film (Figure b) is denser.

Figure 6

Metallographic micrographs of (a) 2 μm and (b) 4 μm thick DAST polycrystalline films spray-coated on the surfaces of Scotch tape films, respectively.

The crystallinity of the as-prepared films was characterized by X-ray diffraction (XRD). Figure a indicates that four peaks, appearing at the diffraction angles of ∼13.9, 16.8, 18.5, and 25.44°, respectively, were detected from the DAST films. The two main peaks at 13.9 and 16.8° are assigned to the signals from the (1̅11) and (021) crystal planes of DAST, whereas the other two ones at 18.5 and 25.44° correspond to the signals from the (202̅) and (1̅31) crystal planes of DAST, respectively.[21] These XRD patterns confirm that polycrystalline DAST films with high crystallinity can be prepared by the simple spray-coating method.

Figure 7

(a) XRD pattern and (b) Fourier transform infrared (FTIR) spectra of the spray-coated DAST polycrystalline films with various film thicknesses.

Fourier transform infrared (FTIR) was employed to further investigate the chemical structures of the films, results from which are shown in Figure b. In the IR spectra, the absorption peak at 3039 cm–1 is assigned to the C=C vinyl C–H stretching mode, whereas that at 1589 cm–1 corresponds to the vibrational mode of the C=C and C–C.[22,23] The absorption peaks at 1530 and 1371 cm–1 are ascribed to the C=C in the ring stretching mode and the vinyl C–H bending mode, respectively.[21] The absorption peaks at 1181 and 825 cm–1 can be identified as the S=O of the sulfonate group stretching mode and the aromatic ring vibrational mode, respectively. The absorption peak at 561 cm–1 is due to the CH3 twisting mode.[21,22] It is clear that the two films shown in Figure exhibit similar mid-IR features (Figure b), both agreeing well with those previously reported about the DAST-based materials.[21,22] This verifies basic DAST structures for the spray-coated films in this work. Moreover, IR measurements at different spots of the same samples revealed that both the chemical structures and optical responses are uniform, in accordance with the morphologic results (Figure ).

The measured terahertz spectra of the fabricated SSAs with the respective DAST film thickness of 2 and 4 μm are displayed in Figure a. Two absorption peaks at ∼1.1 and 1.4 THz, combined the contributions from both the a-axis (1.1 THz)[20] and the b-axis (1.1 and 1.3 THz)[20] of single crystal in the prepared polycrystalline DAST films, were experimentally observed over the range of 0.9–1.5 THz (Figure a). Particularly, the peak at 1.4 THz in Figure a, assigned to the rotational vibration of anion along the a-axis in the optical phonon mode,[17] corresponds to the calculated peak of DAST crystal along the b-axis at 1.3 THz (Figure b) and close to that (1.46 THz) previously measured by FTIR.[24] Further inspection indicates that when the DAST film thickness is increased from 2 to 4 μm, the measured absorption of the fabricated SSA at 1.1 THz is increased from 29.8 to 66.8% (Figure a). In contrast, the peak absorption of the 4 μm thick polycrystalline DAST film deposited on the Al substrate without the SSA structure was measured to be 15.2% (Figure a). For comparison, the spectra of the SSA simulated under the respective DAST thickness of 2 and 4 μm are shown Figure b, revealing that the averaged absorbance of the SSA along the a-axis and b-axis near 1.1 THz is increased from 58.2 to 77.6% as the DAST thickness is increased from 2 to 4 μm. Both measured (Figure a) and simulated (Figure b) results confirm that the SSA structure indeed can enhance the terahertz absorption even in the nonideal polycrystalline phase of DAST formed by spray-deposition and annealing. These experimental results demonstrate that organic SSA with high terahertz absorption can be realized by a DAST film (Figure a), despite its being an insulator and countering the usual expectation of a conductive surface for a SSA. It however does not counter the underlying physics as a highly absorbing organic crystal has its equivalent high optical conductivity, as expressed in eq . It is noted that perfect 100% absorption at 1.1 THz is not expected in this system as the phase and amplitude conditions for a perfect impedance match are mutually constrained here by the available range of the optical conductivity and that of the DAST film thickness.

Figure 8

(a) Experimentally measured and (b) CST-simulated terahertz spectra of the SSAs with different thicknesses of the top DAST films.

Conclusions

In this work, a novel THz SSA with a counterintuitive or unconventional insulating organic DAST film as the surface absorbing and impedance matching layer was designed and successfully fabricated. The absorption properties of this DAST-based SSA were theoretically investigated by an analytical TLM method and validated by numerical simulations using commercial CST software. From both, one can expect the designed DAST-based SSA to exhibit a strong structurally enhanced absorption at 1.1 THz, which when successful will help broaden the reach of DAST into regimes of applications where both high absorption efficiency and thin and flexible form factors are needed. The TLM calculations along the a-axis show that the SSA formed with 4 μm thick DAST film could operate as a nearly perfect absorber with absorbance of 92.9% at 1.1 THz. The experimental measurements show high absorption of 66.8% at 1.1 THz, in support of the theory but suffering from the limitation of a polycrystalline film with mixed contributions from both the a- and b-polarizations.

The enhanced terahertz absorption of a thin DAST film in a SSA structure comes with the advantage over the bulk crystal counterpart in flexibility, scalability, simpler and easier fabrication, and ductility for handling. To the best of our knowledge, this is the first report about such organic SSA. Although a DAST polycrystalline film is used in the actual test, instead of the simpler and better single crystal for modeling, it is expected that in future tests when precision fabrication of DAST single-crystal film of a predetermined thickness becomes possible, a higher terahertz absorption and better agreement with theory will be achievable.

Experimental Section

Design of the DAST-Based SSA

With the strong terahertz absorption of DAST crystal at 1.1 THz, the optical conductivity of DAST was estimated to be very high. On the basis of this, we designed a novel SSA formed with a DAST film as the top layer, the structure of which is completely different from that of the conventional SSAs formed with highly conductive metals. In the designed SSA, the top surface layer is a DAST single-crystal film, serving as both the terahertz absorber and the impedance matcher with the free space. Moreover, the bottom layer is a 200 nm thick Al film, which is thin enough to remain flexible and thick enough to ensure a complete reflection in terahertz. The intermediate layer is a common polymer, Scotch tape, which is transparent in the terahertz region and serves as the dielectric spacer layer for the SSA. It was reported that the resonance frequency of the Scotch tape is near 1.0 THz, close to the absorption peak of DAST crystal at 1.1 THz.[25] Moreover, measurements revealed that the Scotch tape used in this work has excellent transmittance, exceeding 85% in the range of 0.2–3 THz (see Figure S2 in the Supporting Information). Accordingly, this Scotch tape was chosen as the intermediate dielectric spacer layer for the designed DAST-based SSA (Figure a).

Simulations

The finite-difference time-domain algorithm of the CST Microwave Studio 2014 electromagnetic simulation software was used for the simulations. As shown in Figure a, the incident electromagnetic wave and its electric and magnetic fields are along the z, x, and y axes, respectively. The frequency-dependent permittivities of DAST crystal along the a- and b-axes are shown in the Supporting Information. In the CST simulations, the three-layer SSA structure, constructed by a bottom 200 nm thick Al, an intermediate 50 μm thick spacer layer with a refractive index of 1.38, and a top 1 μm thick DAST layer, was first established. The simulation allows for adjusting either the thickness of the DAST film or that of the spacer. Second, the real and imaginary parts of the frequency-dependent permittivities of DAST along the a-axis and the b-axis (see Figure S1 in the Supporting Information) were imported into the CST software, respectively. Finally, the terahertz absorption spectra for DAST single crystal along the a-axis and the b-axis were simulated by CST, respectively.

Preparation of the DAST-Based SSA

First, 80 mg of DAST powder was dissolved in 10 mL of methanol, and then the solution was stirred with a glass rod and ultrasonically treated for 3 h till a uniform DAST-methanol solution was yielded. Second, a 200 nm thick Al film was deposited by magnetron sputtering on a flexible Scotch tape (3M Utility Tape #522, 3M Corp.) with a refractive index of 1.38 and a thickness of 51 μm, close to the optimum value (50 μm) by simulations (Figure ). Third, the DAST-methanol solution was sprayed on the other side of this Scotch tape under the conditions of 0.2 MPa air pressure, at a height of 20 cm, and spraying for 30 s. After spraying, the samples were heated at 60 °C for 5 min for solidification. The processes were repeated till desirable thickness of the DAST film was achieved. Finally, the samples were annealed at 80 °C for 2 h in highly pure nitrogen gas for removing the solvent and water.

Characterization

The terahertz spectra of the bare DAST film and the DAST-based SSA were characterized by terahertz time-domain spectroscopy (TPS Spectra 3000, TeraView) and Fourier infrared spectroscopy (PerkinElmer Spectrum 400), respectively. The morphologies of the fabricated SSAs were imaged by a metallographic microscope (MV 5000). The crystallinities of the films were characterized by X-ray diffraction (XRD) patterns (Philips X’pertProMPD), whereas the film thicknesses were measured by a step profilometer (Ambios XP-300).