Nano-film aluminum-gold for ultra-high dynamic-range surface plasmon resonance chemical sensor

Abstract

An analytical and experimental study of nano-film aluminum (Al) for ultra-high dynamic range surface plasmon resonance (SPR) biosensor is presented in this article. A thin film of 16 nm Al is proposed for metallic sensing layer for SPR sensor. For the protective layer, a 10 nm of gold (Au) layer was configured on top of Al as a protection layer. This ultra-high dynamic range of SPR biosensor reached the bulk refractive index sample limit up to 1.45 RIU. For the analytical study, with the assumption of anisotropic refractive indices experiment, the dynamic range showed a refractive index value of around 1.58 RIU. The refractive index value limit achieved by the proposed sensing design is potentially implemented in various applications, such as in chemical detection and environmental monitoring study with high refractive index solution sample. The experimental results are presented as a proof-of-concept of the proposed idea.

Keywords

dynamic range / surface plasmon resonance (SPR) / sensor / aluminum (Al) / gold

Introduction

Surface plasmon resonance (SPR) has emerged as one of the eminent biochemical sensor platforms gaining tremendous attention from scientists in recent decades [¹,²]. Numerous potential applications in environmental monitoring, biomedical detection, and food science have been reported to employ and explore the distinctive features of SPR sensor, such as its high sensitivity, real-time and label-free detection [³–⁶]. The surface plasmon phenomena exist due to significant numbers of free electron confinement in the nanoscale thickness of metal film surface interfacing with the dielectric medium, such as water or air. The free electron oscillation in nanoscale can be observed by p-polarized light coupling through high refractive index prism, in particular, incident angle. In the resonance condition, the fraction energy of p-polarized light will be absorbed by the surface plasmon wave and results in a dark band in the reflectivity profile [⁷]. Interestingly, the reflectivity profile will be shifted to the lower or higher angle of incident light with the fluctuation of the refractive index medium value [⁸–¹⁰].

In chemo-sensor technology development, several researchers attempt to improve the sensitivity as optimized as possible along with the smallest limit of detection (LOD) for ultra-low concentration detection. This parameter of sensing performance plays an essential role in early diagnosis or in general experimental laboratory works requiring a small volume of sample with low concentration contained. Therefore, myriads of strategies were applied to improve the low concentration detection using nanosensing platform. Nevertheless, while efforts have been made constantly to achieve the lower LOD, another important sensing parameter such as dynamic range is not adequately explored in biosensing technology development.

In some particular applications, such as, in cancer marker early detection [¹¹], diseases diagnosis [¹²], food contamination [¹³] and heavy metal detection [¹⁴], the sensitive and low detection limit performance in a constructed sensor is critically required. Pertaining to the key roles of the sensitivity and low LOD, a variety of techniques have been proposed in the SPR sensor fabrication including the application of nanomaterial structures [¹⁵–¹⁷], 2D materials [¹⁸–²¹], advance grating structures [²²] and novel metal film structures [²³–²⁶].

On the other hand, in several particular detection setups, high dynamic range performance is necessary, for instance in bio-detection involving high refractive index solution such as Dulbecco’s modified eagle medium (DMEM)[⁶,²⁷], or environmental monitoring for oil, toluene, or benzene contamination [²⁸]. However, the trade-off between sensitivity/detection limit and dynamic range performance is a challenging factor for sensor designers [²⁹–³¹]. A dynamic range refers to the targeted concentration range which can be detected by a sensing platform [³²]. Typically, a platform with ultra-low LOD exhibits a very low dynamic range due to the low saturation point of the detection. In contrast, platforms with high dynamic range sensing usually suffer from the low sensitivity performance. Mishra et al. proposed a GaP-based Kretschmann prism for the SPR sensor to boost the dynamic range performance [³³]. However, the GaP prism cost and availability in the market are disadvantageous factors for this purpose.

In SPR sensor, a thin film gold layer is applied as conventional metallic sensing layer owing to its stability and non-reactive behavior in the ambient environment [³⁴,³⁵]. Besides, the gold dielectric constant value is quite superior to enhance the plasmonic activity. Whereas, silver is also regarded potential as sensing layer due to its higher optical reinforcement to generate plasmonic fields than gold, yet, several drawbacks, such as toxicity issues and oxidation reactivity have made Ag less preferred in SPR sensor development [³⁶]. Bimetallic and multilayer Ag/Au have also been exploited as an alternative sensing layer in SPR sensor configuration demonstrated in a portable platform by our group [²³,²⁴,³⁷]. Using these noble metal sensing layer, the performance of the dynamic range is limited, particularly, in angular interrogation-based SPR sensor.

Another metal holding a potency to generate plasmonic field is Al. Previous studies reported the structural sensing configuration of Al film for the adhesion layer of the SPR metallic sensing layer [²³,²⁴]. Still, only few studies reported the application of Al for SPR biosensor configuration because of its limitation and the stability issues [³⁸,³⁹]. First, Al is reactive in oxidation and has the tendency to form an Al₂O₃ thin layer in the surface when it is exposed to ambient environment. Second, Al is reactive in the fluidic sample with acidic pH and thus, make it difficult when applied in several biomedical applications where the pH solution is present in major procedure such as the pH solution, such as acetate (pH 5.0) for facilitating surface chemistry and thiol activation [⁴⁰]. Hence, a pure Al sensing layer is not feasible for practical use of SPR sensor applications; although, in theory, the pure Al has potential optical characteristics to trigger further dynamic range of SPR sensing platform and offers the benefit of low-cost material in comparison with the noble metals, such as gold and silver.

In this article, we proposed an Al film sensing layer protected by thin film gold in angular interrogation-based SPR sensor for ultra-high dynamic range performance and studied the numerical and analytical study. The experimental results are presented as a proof-of-concept of the proposed idea. The application of Al for the first layer of the sensing metal has shown the advantages over the use of a typical adhesion layer such as chromium (Cr). An analytical approach to this study was simulated and compared to the other sensing structures for the references [²³,²⁴]. It is found that by using the proposed SPR sensing layers, a dynamic range up to 1.45 RIU solution was successfully achieved. This high value of bulk refractive index solution is comparable to the 63.5 wt% of sucrose water concentration and 22% benzene on methanol solution.

Materials and methods

Simulation setup

The simulation setup was constructed by Kretschmann configuration of SPR sensor using an angular interrogation method. The optical parameters of the materials in this study were obtained from the published articles. The real and imaginary part of the dielectric constant (e_R,e_i), refractive indices (n) and extinction coefficient (k) parameters of Al and Au were obtained from the report of McPeak et al. [⁴¹], respectively. Subsequently, the optical parameters of H₂O as the reference medium were obtained from the article published by Hale and Querry [⁴²]. The n and k of BK7 glass as the Kretschmann prism were obtained from Schott™ optical glass datasheet; while the shape of half cylindrical prism was taken into account. The utilization of the half cylindrical prism instead of the triangle or trapezoid prism was preferred to omit the refraction variable in the incident angle interface between free space and the prism. For the anisotropic refractive indices analysis, the assumption of 60 nm thickness for the refractive index value in the sensing surface was applied considering the surface chemistry of the thiol length (6 to 8 nm), the IgG protein size of around 13.7 nm [⁴³] in some in biomedical applications and the target size such as viral particles of approximately 40 nm [⁶]. In this study, an SPR configuration approach using angular interrogation method in Kretschmann coupler configuration and red monochromatic wavelength of 600 nm was utilized. The simulation description of the experimental setup is demonstrated in Fig. 1.

Experimental setup

The platform was constructed using half cylindrical BK7 prism with a diameter of 80 mm (HiLite Optronics Inc., Hsinchu, Taiwan, China). For the optical convergence, the collimating lens was using a Fresnel lens with a focal length of 5.08 cm (Edmund Optics Inc., New Jersey, USA). The polarizer film using DBEF (3M, Minnesota, USA) was attached in the Fresnel lens and a rotating plate with the angular scale (Chief SI Co., Hsinchu, Taiwan, China). The light source used for the whole experiment was a red laser diode (WelTek Co. Ltd., Taoyuan, Taiwan, China). The metallic sensing layer was deposited by thermal evaporation deposition with a very low deposition rate (0.5 Å/s) in vacuum pressure below 1 × 10⁻⁶ Torr. The PDMS flow cell was a homemade design using molding techniques and constructed with peristaltic tubes (Gilson, Middleton, USA). The peristaltic pump (Shishin Tech., Taipei, Taiwan, China) was utilized to flow the liquid samples. The refractive indices sample was prepared by sucrose (Sigma Aldrich, Missouri, USA) diluted in deionized (DI) water by weight percentage to obtain particular refractive index values. The sucrose water concentrations were 0, 18, 34.8, 45.1, 54.6, 63.5 wt%, which corresponded to the refractive index (RI) value 1.33, 1.36, 1.39, 1.41, 1.43, and 1.45 RIU, respectively.

Results and discussion

The following equation describes the wave vector of the surface plasmon (K_SP) along the interface (x-direction) [⁴⁴]:

(1)K_{\mathrm{SP}}={\displaystyle \frac{\mathrm{2}\mathrm{\pi }}{\lambda }}\sqrt{{\displaystyle \frac{{\epsilon }_{\mathrm{M}}{\epsilon }_{\mathrm{D}}}{{\epsilon }_{\mathrm{M}}+{\epsilon }_{\mathrm{D}}}}},

where l is the wavelength of the incident light, e_M, and e_D are real parts of the dielectric constant of the metallic sensing layer and medium, respectively. If we consider the Drude model [⁴⁵] where the e_M of the thin film metal is a negative value, there is a possibility that the value of K_SP is imaginary when |e_M|<e_D. In this condition, surface plasmon wave in the metal-dielectric interface does not exist. Thus, only several metals have potential properties to generate surface plasmon. In addition, in the case of too high dielectric constant value of the medium, the K_SP value will be imaginary as well. In this case, the resonance condition is limited in a particular point of the refractive index (or dielectric constant) of the medium. In the sensing technology perspective, this point is the dynamic range limit of the sensing performance.

While in Kretschmann-based SPR configuration, the resonance condition can be obtained when the incident light in particular angle (q) through high refractive index prism excites the metal sensing as described by

(2)\sin \theta ={\displaystyle \frac{\mathrm{1}}{n_{\mathrm{P}}}}\sqrt{{\displaystyle \frac{{\epsilon }_{\mathrm{M}}{\epsilon }_{\mathrm{D}}}{{\epsilon }_{\mathrm{M}}+{\epsilon }_{\mathrm{D}}}}},

where n_p is the refractive index value of the prism.

As the resonance condition is reached, the reflectivity curve can be seen in the reflectance light profile along the incident angle axis. The reflectivity of the attenuated total reflectance (ATR) through high refractive index prism around the resonance angle can be explained by [⁴⁶]

(3)R=|r_{\mathrm{pmd}}{|}^{\mathrm{2}},

where

(4)r_{\mathrm{pmd}}={\displaystyle \frac{r_{\mathrm{pm}}+r_{\mathrm{md}}\exp (\mathrm{2}\mathrm{i}{k}_{\mathrm{m}x}t)}{\mathrm{1}+r_{\mathrm{pm}}{r}_{\mathrm{md}}\exp (\mathrm{2}\mathrm{i}{k}_{\mathrm{m}x}t)}},

where t is the metallic sensing layer thickness, and

(5)k_{ix}=\sqrt{{\left({\displaystyle \frac{\mathrm{2}\mathrm{\pi }}{\lambda }}\right)}^{\mathrm{2}}\epsilon i-k_z^{\mathrm{2}}},

and for the p-polarized light:

(6)r_{ij}={\displaystyle \frac{{\epsilon }_j{\epsilon }_{ix}-{\epsilon }_i{k}_{jx}}{{\epsilon }_j{k}_{ix}+{\epsilon }_i{k}_{jx}}},

where the subscripts i and j are p, m, or d indicating for prism, metal, and dielectric, respectively.

The reflectivity profile of the Al sensing layer compared to the other noble metals, such as Au and Ag, is demonstrated in Fig. 2(a). In 600 nm wavelength light, based on McPeak et al.’s report [⁴¹], the e_Mreal value for Au, Ag, and Al are -10.47, -16.34, and -39.58, respectively. The substantial negative value of Al e_Mreal is an exciting property to be utilized for sensing layer in SPR sensor; in contrast, Al also exhibits high k-value of 6.36 while Au is 3.24 and Ag is 4.04, This property is responsible for the light absorption significantly.

For further detail analysis, the penetration depth (d) of the evanescent wave inside the medium or metal by each of metal structure can be calculated by

(7)d={\displaystyle \frac{\mathrm{1}}{K_{yi}}},

where K_yi is the surface plasmon wave vector in the y-direction described by

(8)K_{yi}={\displaystyle \frac{\mathrm{2}\mathrm{\pi }}{\lambda }}\sqrt{{\displaystyle \frac{{\epsilon }_i^{\mathrm{2}}}{{\epsilon }_{\mathrm{M}}{\epsilon }_{\mathrm{D}}}}},

where e_i can be e_M or e_D for the penetration inside the metal or medium, respectively. From Eq. (4) when K_SP value in Eq. (1) is real, the K_yi value will be imaginary due to |e_M|>e_D. This mathematical approach indicates that the evanescent wave exponentially decays in its penetration into the medium or metal (inset Fig. 2(b)) [⁴⁷]. The calculated penetration depth inside the medium and metals, where |e_M| is taken into Eq. (4) were plotted in Fig. 2(b). From this analytical calculation, Al performed the deepest penetration inside the measured medium around 346 nm in 600 nm of wavelength in comparison with the penetration depth of Au (188 nm) and Ag (229 nm), respectively.

In the results displayed in Figs. 2(a) and 2(b), water as the dielectric medium is taken into account. As postulated in Eqs. (1) and (2), the dielectric constant of the medium (e_Dreal) is the predominant factor for surface plasmon generation besides the dielectric constant of the metal (e_Mreal). Consequently, in the case of the different refractive index value of the medium, the penetration depth could be different due to the different wave vector of surface plasmon wave. Therefore, in this analytical study, as the application is dedicated to the water and environmental monitoring, the water as the medium of the basic analytical and design is applied.

The optimization of aluminum layer thickness in the SPR configuration is presented in Fig. 3(a) where the optimum reflectivity dip was obtained at the thickness of 16 nm. In addition, as a protection layer, the Au thickness optimization was simulated from 0 to 16 nm with 2 nm of step as shown in Fig. 3(b). The minimum reflectivity dip values are demonstrated in the inset of Fig. 3(b). Interestingly, the Au protection layer thickness up to 16 nm insignificantly changes the value of minimum reflectivity dip as presented in the inset. The standard deviation of the minimum reflectivity dip was about 0.58% calculated. However, in the real manufacturing process, the formation of Au film below 8 nm is not yet naturally constructed. In this condition, the Au material tends to form nano-islands morphology [⁴⁸–⁵¹]. With this fundamental finding, a 10 nm of Au protection layer was carried out throughout the further analytical study.

The simulated measurement of a series of the solution with different bulk refractive index values was plotted in Fig. 4. The SPR reflectivity shifted from lower to higher resonance angle as the refractive index solution increased (Fig. 4(a)). The correlation of n and k of the medium to the e_D can be expressed by

(9){\epsilon }_{\mathrm{D}}=n_{\mathrm{D}}^2-k_{\mathrm{D}}^{\mathrm{2}},

where n_D and k_D are refractive index and extinction coefficient of the medium, respectively. Moreover, when the refractive index solution escalated, the e_D exponentially increased, and the resonance angle as described in Eq. (2) shifted to the higher incident angle. Finally, with the refractive index value above 1.45, the resonance of surface plasmon did not occur. This dynamic range of Al/Au sensing is highly superior as compared to other sensing structures, such as Au, Ag, or bimetallic Ag/Au as presented in Fig. 4(b).

The experimental work demonstrated the proof-of-concept of the proposed sensing platform with optical configuration for the various measurement of RI samples prepared by sucrose-water solution (Fig. 1). First, the laser diode was passed through the Fresnel lens [⁵²] to maintain the collimation of the laser beam. The flat geometry of Fresnel lens has an advantage that the reflective polarizer film [³⁷] can be attached in the smooth surface of the lens to obtain the p-polarized light as the primary mechanism to couple the SPR. Later, the laser beam was passed through the half cylindrical prism to avoid the refraction phenomena due to the air and prism interface. The sensing metals were located in the center origin of the prism geometry, and the PDMS fluidic cell was attached to flow the sample. The PDMS chamber was designed in the center origin of the prism as well. The total internal reflection (TIR) of the laser beam was observed on the screen. The measurements were done by flowing the measured sample then rotating the protractor plate until the dark band appeared in the laser spot on the screen (see supplementary material video). The presence of the dark band indicates that the SPR condition has been obtained and the protractor angle position is recorded. The results were plotted in Fig. 4 which confirms a good agreement with the simulation results.

In the case of anisotropic surface refractive index solution [⁴⁶], the simulation results were shown in Fig. 5. Anisotropic surface refractive index solution is a typical case for the target sample diluted in buffer. This scenario mostly happens in the bio-detection instances. For example, when antibodies are diluted in the phosphate buffer saline (PBS), the immobilized antibodies near the metal surface give a slightly higher n compared to the n of solution buffer with the assumption that the thickness of the IgG antibody and its antigen are around 28−30 nm [⁵³], respectively. The assumption of 60 nm of the surface refractive index around the metal sensing was configured. As the results, the potential dynamic range of the sensing performance in this anisotropic measurement was prominently improved to 1.58 RIU. Nonetheless, the sensitivity performance implies a trade-off in the lower slope of resonance angle response compared to the bulk refractive index measurement. As a matter of fact, in the bio-detection, a sensitivity performance is more valuable than that of the dynamic range. Based on this analysis, this sensing structure is more suitable for measurement in the environment monitoring application, such as for high bulk refractive index sample like in oil contamination control.

Finally, the summary of related work of dynamic range performance in SPR sensor platform is presented in Table 1. It is revealed that the dynamic range performance of the proposed platform is comparable to the previously published reports.

Conclusion

A comprehensive characterization of Al/Au-based SPR sensing is analytically demonstrated and followed up by the proof-of-concept results experimentally. The detection range up to 1.45 RIU for bulk refractive index measurement was achieved. This high value of bulk refractive index solution is comparable to the 63.5 wt% of sucrose water concentration and 22% benzene on methanol solution. While for the anisotropic refractive index solution in the thickness of 60 nm, the detection range reached up to 1.58 RIU. From this study, we proposed the Al-based sensing platform for ultra-high dynamic range SPR sensor for environmental applications requiring high refractive index analyte, for example, in salt water monitoring, oil contamination study, or other high refractive indices fluidic measurement.

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature