1. Introduction
Raman spectroscopy is a highly versatile and powerful analytical technique widely employed across physical chemistry, materials science, surface science, and biomedicine due to its unique ability to provide molecular “fingerprint” information through inelastic light scattering [1,2,3,4]. Traditionally, Raman spectroscopy has primarily focused on surface molecular analysis, playing a pivotal role in surface science research [5,6,7]. This capability has enabled researchers to investigate the composition, structure, and reaction mechanisms at surfaces, thereby contributing significantly to the understanding of fundamental physical and chemical processes [8,9,10,11].
However, the growing complexity of scientific and biomedical research increasingly demands access to subsurface molecular information [12,13]. For instance, in biomedicine, many pathological processes originate in subsurface tissue layers. Early-stage tumor development, for example, often occurs below the epidermis [6]. Accurate detection of subsurface analytes with molecular specificity is thus essential for the early diagnosis and treatment of diseases. Conventional subsurface Raman sensing has relied on invasive and destructive methods such as physical cutting and polishing. While these approaches provide access to subsurface analytes, they cause irreversible damage to samples, limiting their application to valuable or living tissues [14]. To address this limitation, minimally invasive techniques—such as microneedle puncture and localized micro-etching—have been introduced. Although these methods reduce the sample trauma, they often introduce interference, compromising the detection accuracy [15,16,17].
Given that light can penetrate subsurface layers, Raman spectroscopy has the inherent potential to enable remote and non-destructive subsurface sensing [10,18,19,20,21]. Efficient coupling of incident light into subsurface regions could unlock the capability for non-invasive Raman detection [22,23]. Recent advances have highlighted the use of microlenses based on photonic nanojets (PNJs) to achieve highly localized, non-evanescent electromagnetic fields in the near-field region [24,25,26]. Numerous studies have focused on optimizing the microlens designs to enhance the subsurface electric field focusing, particularly by improving parameters such as the focal length and full width at half maximum (FWHM) [19,27,28,29,30]. A longer focal length facilitates deeper sensing, while a smaller FWHM enables a high spatial resolution in Raman imaging. Significant progress has been made in resolution optimization. Ling et al. [31] achieved a milestone by designing hollow microsphere lenses, which produced an ultra-narrow focal spot of 140 nm for near-surface nanoimaging. However, this design confines the focal plane to the lens surface, sacrificing the penetration depth—a critical limitation for subsurface applications. Similarly, Heifetz et al. [32] demonstrated a hybrid approach by positioning a gold nanosphere within the nanojet of a Mie-resonant dielectric microsphere. This configuration leveraged the synergy between the plasmonic and photonic effects, enabling a subdiffraction-limited resolution (<λ/3). The integration of the gold nanosphere within the Mie-resonant microsphere’s nanojet not only enhanced the localized electromagnetic fields but also achieved a superior spatial resolution, thereby advancing the potential of such hybrid systems for high-resolution subsurface sensing. While this strategy highlights the potential of combined systems, the fundamental trade-off between the resolution and focal length remains unresolved.
Further research has explored the integration of photonic nanojets with surface-enhanced Raman scattering (SERS) techniques. For instance, Dantham et al. reported a two-step enhancement in Raman spectroscopy by coupling SERS substrates with PNJs, observing the optimal SERS intensity under a 20× objective lens [29]. Chen et al. developed a gel-based 3D SERS substrate incorporating silica beads, which enhanced the SERS sensitivity and reduced the detection limit of malachite green 100-fold using multiple PNJs [22]. Additionally, Laha et al. simulated PNJ-mediated SERS and surface-enhanced fluorescence (SEF), demonstrating single-molecule SERS signal enhancements of three orders of magnitude, with the SEF enhancements exceeding this by at least one order [30]. These studies collectively highlight two critical directions: (1) hybrid systems combining PNJs with plasmonics for multiplicative signal enhancement [32] and (2) geometric innovations in lens architectures for resolution optimization [31]. While these advancements have improved Raman signal collection, challenges remain regarding the trade-offs between the focal length and spatial resolution.
In this study, as shown in Figure 1, we introduce a novel approach to addressing these limitations by designing a non-uniform microlens array that achieves both a long focal length and a small FWHM. The proposed microlens array incorporates central and side cylindrical lenses, systematically optimized for 785 nm and 1550 nm excitation to achieve a superior focusing performance. Our findings demonstrate that the non-uniform microlens array significantly extends the focal length and enhances the spatial resolution compared to these properties in conventional single microlens and uniform microlens array designs. Simulations reveal that introducing a side lens into the design achieves the optimal focusing capabilities. Furthermore, we validated the scalability of our approach by extending the design to a rectangular microlens array. The non-uniform microlens array was successfully fabricated using 3D printing, and the experimental results revealed an 85% increase in the Raman intensity compared to that of single microlens configurations. Additionally, characterization at 1550 nm showed an average experimental focal length of 158 μm, with a deviation of less than 9% from the simulations. This innovative design represents a significant advancement in non-destructive subsurface Raman detection, offering transformative potential for research and applications in this field.
2. Materials and Methods
2.1. Materials
Methylene blue (MB, ≥98.7%, CAS: 64-17-5) and anhydrous ethanol (≥99.9%, CAS: 7220-79-3) were procured from Xiamen Lvyin Reagent Glass Instrument Co. (Xiamen, China). The photosensitive resin (HTL Resin) used for 3D printing was supplied by BMF Precision Tech Inc. (Chongqing, China). Deionized water (18.25 MΩ) was provided by Sichuan UPRIGHT Ultrapure Co., Ltd.
2.2. Finite Element Simulation
Finite element simulations of the electric field distribution in the microlens were conducted using the Wave Optics module in COMSOL Multiphysics (Version 5.5). A two-dimensional model was developed to represent the microlens structure. In the simulations, for the two selected wavelengths, namely 785 nm and 1550 nm, the refractive index of air was set at 1, while that of the photosensitive resin was 1.46, while the refractive index of the microlens material, composed of photosensitive resin, was defined as 1.46. The excitation wavelengths used for the simulation were 785 nm and 1550 nm. The incident light beam was modeled as a Gaussian beam, with its focal plane positioned at the geometric center of the microlens.
2.3. Three-Dimensional Printing of the Microlens
The microlens was fabricated using a MicroArch S230A 3D printer from BMF Precision Tech Inc. (China). Both single microlens (without side lenses) and non-uniform microlens arrays (with side lenses) were designed using three-dimensional modeling software. These models were subsequently imported into the 3D printer for fabrication. The printing resolution was set at 2 μm per pixel, with a layer thickness of 5 μm. The photosensitive resin was cured using a 405 nm wavelength light source during the fabrication process. After the fabrication process was completed, anhydrous ethanol was used to remove any potentially residual photosensitive resin on the surface of the microlenses.
2.4. Raman Spectrum Acquisition Based on Non-Uniform Microlens Array
Raman spectra were acquired using a confocal Raman microscope from Xiamen SHINS Technology Co., Ltd. The confocal Raman microscope (Xiamen SHINS Technology Co., Ltd.) exhibits a frequency stability of 0.5 cm−1, a linewidth stability of 0.2 cm−1, and an intensity stability of 0.01. The excitation wavelength was set to 785 nm, with a laser output power of 50 mW. A 50× objective lens was employed for focusing, and the integration time for each spectrum was set to 10 s. Each Raman spectrum was averaged over 10 repeated measurements to ensure high signal-to-noise ratios. We use a customized platform to ensure that both the non-uniform microlens array and the single microlens were positioned above a prepared 10 mM MB solution, ensuring that the focal point of the microlens was aligned within the solution. The laser beam was focused onto the microlens surface, and the Raman signals were collected for analysis.
2.5. Topological and Optical Characterization of the Microlenses
The topological features of the fabricated microlenses were characterized using a scanning electron microscope (SEM) from Nano Focus (Shanghai) Co., Ltd. (Shanghai, China). The SEM was operated at an accelerating voltage of 10 kV. The optical field distribution of the microlens was analyzed using the MetronLens system from Ideaoptics Instruments Co., Ltd. (Shanghai, China). The optical measurements were conducted at an excitation wavelength of 1550 nm, and a 50× objective lens was employed to focus the laser onto the microlens.
3. Results
3.1. Optimizing the Microlens Design: Performance Enhancement Through Non-Uniform Configurations
Using the Wave Optics module in COMSOL Multiphysics, a two-dimensional simulation model was developed to optimize the microlens design. The radius of the single central lens was systematically optimized within the range of 3 to 6 μm, with increments of 0.5 μm. The simulation results, as shown in Figure S1d, revealed that when the single central lens’s radius was 5 μm, the focal length reached 1.46 μm, and the spatial resolution was 0.796 μm, exceeding the excitation wavelength. This configuration demonstrated a superior overall performance compared to that with the other tested parameters, consistent with findings reported in previous studies [22,33]. Therefore, the optimal size of the single lens for a 785 nm wavelength was determined to be 5 μm.
To improve the lens performance further, side lenses were integrated with the central lens (with a radius of 5 μm), forming a non-uniform microlens array. The radius of the side lenses was optimized within the range of 1 to 5 μm, with 0.5 μm increments. The simulation results for the selected configurations are presented in Figure 2, where the radii of the side lenses are 1 μm, 2.5 μm, and 5 μm in Figure 2a, Figure 2b, and Figure 2c, respectively. Additional simulation results for the other side lens sizes are shown in Figure S2a–f. The corresponding electric field distributions are represented by white dashed lines in Figure 2a–c. In these simulations, the X-axis (horizontal) characterizes the FWHM and the spatial resolution (see Figure 2d), while the Z-axis (vertical) represents focal depth, with further results displayed in Figure S2g. Key performance metrics, such as the FWHM of the focal spot and the focal length, are summarized in Figure 2e,f.
The simulation results revealed that the optimal configuration occurred when the radius of the side lenses was half that of the central lens. Under this condition, the spatial resolution peaked at 780 nm, and the focal length reached 2.22 μm. This spatial resolution represented the highest among those in the ten tested configurations, while the focal length deviated by only 0.01 μm from the maximum achievable value. As a result, this configuration was identified as having the best overall performance. Additionally, the inclusion of the side lenses increased the focal length by 52% and improved the spatial resolution of the non-uniform microlens array to the subwavelength scale compared to that of the single microlens (without side lenses).
A further analysis was conducted by varying the number of side lens groups integrated into the non-uniform microlens array. Configurations with zero, one, two, three and, four groups of side lenses were compared. The results are presented in Figure 3a–d, with additional data on the configuration with four side lens groups provided in Figure S3a. The focal length data are plotted onto the Y-axis in Figure 3e, while the FWHM is depicted in Figure 3f; additional focal length data can be found in Figure S3b. These findings demonstrate that incorporating one group of side lenses achieves the optimal balance between the spatial resolution and focal length. Specifically, the FWHM reached 780 nm, which was smaller than the incident wavelength. Compared to that in the configuration with zero side lenses, the spatial resolution improved by 2%, and the focal length reached 2.22 μm, representing a 52% enhancement over that with the single microlens without side lenses. These results highlight the effectiveness of the non-uniform microlens array design. By optimizing both the size and number of side lenses, the non-uniform microlens array achieves a significantly improved spatial resolution and focal length compared to these values with a single microlens. This study demonstrates that the inclusion of side lenses provides a simple yet effective strategy for enhancing the performance of microlenses, making them suitable for high-precision optical applications.
3.2. Universality of the Non-Uniform Microlens Array Design for Rectangular Lenses
To evaluate the universality of the non-uniform microlens array design, we applied this structure to rectangular lenses, modeling them using the same simulation parameters as those employed for cylindrical microlenses. As a preliminary step, we optimized the side length of a single microlens (without side lenses), with the simulation results illustrated in Figure S4a–i. Based on this analysis, a rectangular microlens size of 5 μm was selected for further study. Side lenses with a size of 2.5 μm were then added to both sides of the central lens for subsequent calculations.
The simulation results for the single rectangular microlens (without side lenses) and the non-uniform rectangular microlens array (with side lenses) are presented in Figure 4a and Figure 4b, respectively. The findings revealed that incorporating side lenses into the non-uniform design significantly enhanced the performance of the rectangular microlenses. Specifically, the focal length of the non-uniform rectangular microlens array increased by 26% compared to that of the single microlens.
These results demonstrate the adaptability and versatility of the non-uniform microlens array structure. While originally developed for cylindrical microlenses, the design is equally effective when applied to rectangular microlenses. This adaptability highlights the potential of the non-uniform microlens array design for various lens geometries, significantly improving both the focal length and spatial resolution across different applications.
3.3. Raman Signal Enhancement Using a Non-Uniform Rectangular Microlens Array
The microlens designs were created using three-dimensional modeling software, and both the single microlens (with a side length of 60 μm and no side lenses) and the non-uniform rectangular microlens array (with side lengths of 30 μm) were fabricated using advanced 3D printing technology. These two microlens configurations were subsequently employed to collect the Raman spectra of 10 mM methylene blue aqueous solution using a confocal Raman microscope (Xiamen SHINS Technology Co., Ltd.). For reference, the assignment of the Raman peaks for methylene blue is provided in Table S1.
The original spectra measured using the non-uniform microlens array and the single microlens are shown in Figure 5a,b. The Gaussian fitting of the spectra shows peaks centered at 1623 and 1633 cm−1. The integrated peak areas at 1633 cm−1, attributed to the stretching vibration of the C=C bonds of photosensitive resin (Figure S5) [34], as shown as Figure 5c, are almost the same under both microlenses, as the photosensitive resin is the material of the microlens itself. Meanwhile, the integrated peak areas centered at 1623 cm−1, attributed to the C-C ring stretching of MB [35], increased by about 85% after using the non-uniform microlens array, demonstrating a slight enhancement in the Raman signals.
This substantial improvement in the Raman intensity highlights the effectiveness of the non-uniform microlens array structure in amplifying the Raman signals. These findings demonstrate the practical advantages of the non-uniform microlens array design for enhancing the signal sensitivity and spatial resolution in Raman spectroscopy. Furthermore, these results underscore its potential for applications in subsurface remote Raman detection, where a precise and non-invasive molecular analysis is required.
3.4. Morphological Characterization and Focal Length Analysis of the Single Lens and the Non-Uniform Microlens Array
To evaluate the morphological features of the fabricated microlenses, we employed the SEM to measure the microlenses. SEM images of the single microlens (without side lenses) and the non-uniform microlens array are presented in Figure 6a and Figure 6b, respectively. Due to inherent limitations in the 3D printing process, both microlenses exhibited stepped edges. The single microlens, shown in Figure 6a, has a rounded rectangular shape with a side length of 60 μm, while the non-uniform microlens array, depicted in Figure 6b, incorporates side lenses measuring 30 μm. Meanwhile, due to the small size of the microlenses, to avoid failure in 3D printing fabrication and to increase the fabrication success rate, multiple microlens arrays were fabricated on the same platform, with a spacing of 800 μm between each microlens array. These structural features were further simulated using COMSOL Multiphysics, with the excitation wavelength (λ) set to 1550 nm for consistency between the experimental and simulated analyses.
To validate the experimental results, we compared the measured focal lengths of the microlenses to their corresponding simulations. As shown in Figure 6c, the experimental focal length of the single microlens without side lenses at 1550 nm was measured to be 96 μm, closely aligning with the simulated value of 105 μm. Similarly, the focusing characteristics of the non-uniform microlens array are illustrated in Figure 6d, where the experimental focal length was determined to be 158 μm, compared to the simulation result of 145 μm.
To reduce the variability and ensure reliable measurements, the focal length experiments were repeated four times for both microlens designs. The results, summarized in Figure 6e, show an average experimental focal length of 96 μm for the single microlens, with a deviation of less than 8.6% from the simulation result of 105 μm. In comparison, the average experimental focal length of the non-uniform microlens array was 158 μm, while the simulated focal length was 145 μm, resulting in a discrepancy of less than 9%. This represents a 64.6% increase in the focal length (from 96 μm to 158 μm) for the non-uniform design compared to that with the single microlens configuration in the experiment.
Additionally, comparisons between the experimentally measured spatial resolution and the simulated spatial resolution are provided in Figure S6a,b. These results further confirm the strong correlation between the experimental data and simulations, highlighting the accuracy and robustness of the microlens design and the fabrication process. The significant improvements in the focal length and spatial resolution achieved using the non-uniform microlens array design underscore its potential for high-precision optical applications.
4. Discussion
This study has demonstrated the exceptional potential of photonic nanojet (PNJ)-based non-uniform microlenses in advancing subsurface remote Raman sensing. The augmented performance of our design is likely attributed to the synergistic interplay between the central and side lenses [22,32,36]. In traditional single microlenses, it is highly probable that the lateral divergence of incident light diminishes the focusing efficiency. This phenomenon leads to the majority of the PNJs congregating predominantly on the lens’s surface, thereby precluding their departure from the surface and subsequent focusing at a distal location. Nevertheless, the integration of the side lenses, which have a radius half that of the central lens, potentially redirects and refocuses the scattered light via the constructive interference of the overlapping PNJs. In light of these findings, we were able to generate a focused irradiance spot. Its full width at the FWHM is less than the wavelength, and its focal length extends beyond several wavelengths. This focused spot not only intensifies the electric field with respect to the incident field but also enables high-resolution and long-range focusing. This mechanism achieves a 52% focal length extension (2.22 μm) and a subwavelength resolution (780 nm FWHM at 785 nm excitation), effectively balancing depth penetration and spatial resolution, a critical advancement over uniform designs.
Compared with previous studies, our work can ensure both the resolution and focal length of a lens simultaneously. For example, when compared with the research by Ling et al. [31], they investigated various microparticle lens structures, like microsphere segments, hollow microspheres, and ellipsoids. By optimizing the parameters of a hollow microsphere lens, they managed to achieve an ultra-high resolution with a full width at FWHM of 140 nm for near-surface focusing, which was accomplished by modulating parameters such as the inner diameter of the hollow microsphere. However, the focus of the lens adheres to the lens’s surface. In our study, to control the variables and ensure consistency with Ling et al.’s parameters, we chose to work with the same wavelength of 550 nm. Our non-uniform microlens array, different from their design, which was mainly for near-surface high-resolution focusing, focused on the practical trade-offs. We used finite element simulations to optimize the size and number of side lenses in our non-uniform microlens array. As shown in Figure S7, at 550 nm excitation, our non-uniform microlens array could reach a resolution with a FWHM of 301 nm and an extended focal length of 1.2 μm (In Ling’s research, when the focal length reached 1.14 μm, the full width at FWHM was 350 nm.). Compared with Ling et al.’s design, our non-uniform microlens array shows a better performance in balancing the resolution and focal length for subsurface applications. This balance between the resolution and focal length is crucial for subsurface remote Raman sensing, allowing us to break through the limitations of traditional single microlens and uniform microlens array designs and offering a more efficient approach to subsurface molecular analysis.
5. Conclusions
This study demonstrates the potential of photonic nanojet (PNJ)-based non-uniform microlens array to significantly enhance subsurface remote Raman sensing. By systematically optimizing the lens parameters, we showed that integrating side lenses into the microlens design improves both the focal length and spatial resolution, outperforming traditional single microlens and uniform microlens array designs. Specifically, a central lens radius of 5 μm, paired with the side lenses having a radius half that of the central lens, yielded a 52% increase in the focal length (achieving 2.22 μm) and a subwavelength spatial resolution compared to these values in the single microlens. Additionally, the non-uniform microlens array enhanced the Raman intensity by 85%, further validating its effectiveness in amplifying Raman sensing and enabling the precise detection of subsurface analytes.
The adaptability of the non-uniform microlens was confirmed by successfully applying the design to rectangular lenses. Despite differences in the geometry, the non-uniform structure consistently improved the focal length (64.6%, from 96 μm to 158 μm) and Raman signal intensity, showcasing its versatility for various lens configurations. The experimental measurements of the focal length at an excitation wavelength of 1550 nm closely aligned with simulations, with discrepancies of less than 9%, highlighting the precision of the fabrication process and the reliability of the design. The implications of this work extend to diverse fields, including biomedicine, environmental monitoring, and materials science, where precise subsurface detection and high-resolution chemical analysis are essential. The significant improvements in the focal length, spatial resolution, and Raman intensity achieved using the non-uniform microlens array provide an effective, non-destructive solution for acquiring information on subsurface analytes. Furthermore, the demonstrated compatibility of this design with 3D printing technology suggests that it can be readily fabricated and scaled to practical applications.
Looking ahead, future research will be dedicated to surmounting the existing limitations via synergistic innovations. Dynamically tunable materials, such as lead lanthanum zirconate titanate (PLZT), will be incorporated to achieve real-time refractive index adjustments in complex heterogeneous environments like biological tissues [37,38]. Precision-engineered platforms equipped with real-time feedback control mechanisms will be developed to ensure the optimal alignment between the microlenses and samples. Hybrid systems that combine non-uniform microlenses with plasmonic nanoparticles could enhance the Raman signals by more than 103 [30,39]. Notably, the subwavelength focusing capability and extended focal length of non-uniform microlens arrays are expected to enable their integration with nonlinear Raman techniques such as stimulated Raman scattering (SRS) and hyper-Raman scattering (HRS). Additionally, integration with advanced imaging modalities like hyperspectral imaging and adaptive optics will be explored. These endeavors are set to revolutionize subsurface sensing technologies, broadening their applications across clinical diagnostics, environmental monitoring, and materials science.
Conceptualization: E.-M.Y. and X.-Y.L. Methodology: E.-M.Y. and X.-Y.L. Validation: H.-Y.L. and W.-D.Y. Investigation: E.-M.Y. and X.-Y.L. Data curation: H.-Y.L. and X.-Y.L. Writing—original draft preparation: E.-M.Y. and X.-Y.L. Writing—review and editing: E.-M.Y., J.L. and X.-Y.L. Visualization: E.-M.Y. and X.-Y.L. Supervision: E.-M.Y. and J.L. Funding acquisition: J.L. and E.-M.Y. All authors have read and agreed to the published version of the manuscript.
Not applicable.
Not applicable.
The original contributions presented in this study are included in the article/
The authors declare no conflicts of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Enlarge this image.Figure 1. (a) Schematic of the non-uniform microlens array. The gray structures placed on the surface are the microlens array. The star-like structures in the subsurface area are the analytes to be detected. A 785 nm laser with a Gaussian beam shape illuminates the structure from the top. (b) The simulated distribution of the electric field intensity of the non-uniform microlens array under the illumination of a Gaussian beam with a 785 nm wavelength. Eloc and E0 denote the local and incident electric fields, respectively.
Enlarge this image.Figure 2. (a–c) The simulated electric field distribution of the non-uniform microlens array with different sizes of the side lenses, which are 1 μm, 2.5 μm, and 5 μm for (a–c), respectively. The insets are enlarged views of the PNJs. The vertical dashed white line represents the Z-axis, while the horizontal line denotes the X-axis. (d) The line profiles of the electric field along the horizontal lines shown in the insets in (a–c) for the non-uniform microlens array with various sizes of side lenses. (e) The relationship between the FWHM in (d) and the size of the side lenses. The dotted line represents the FWHM equals to 785 nm, and the asterisk represents the FWHM of the lens with different radius. The FWHM is narrowest (780 nm) when the side lens size is 2.5 μm. (f) The relationship between the focal length and the sizes of the side lenses, revealing that the focal length reaches the maximum when the side lens is 2 μm in diameter. The asterisk represents the focal length of the lens with different radius. The dotted line denotes the radius of side lens equals to 2.5 μm.
Enlarge this image.Figure 3. (a–d) The simulated distribution of the electric field corresponding to 0, 2, 4, and 6 side lenses, respectively. The horizontal and vertical white dashed lines are the central lines of the PNJ along the X-axis and the Z-axis. The insets are enlarged views of the electric field around the PNJs. (e) The line profile of the electric field along the horizontal white dashed lines shown in (a–d). (f) The relationship between the extracted FWHM in (e). When there are two side lenses, the FWHM reaches the narrowest value, which is 780 nm less than the incident wavelength of 785 nm, thus achieving a subwavelength resolution. The asterisk represents the FWHM of the lens with different number of side lens. The dotted line represents the FWHM equals to 785 nm.
Enlarge this image.Figure 4. (a,b) The simulated electric field distribution of a (a) single rectangular microlens and a (b) non-uniform rectangular microlens array with side lenses. The white dotted line in (a,b) indicates the center of the microlens along Z-axis. (c) The line profile of the electric field along the white dashed line in (a,b). The purple dotted line in (c) represents the focal plane of the non-uniform rectangular microlens array, and the brown dotted line indicates the focal plane of a single rectangular microlens.
Enlarge this image.Figure 5. Comparison of the Raman spectra of 10 mM MB solution recorded using microlenses with (a) and without (b) side lenses. The Gaussian fitting of the spectra shows peaks centered at 1623 (purple) and 1633 cm−1 (orange). (c) represents the Raman peak areas at 1623 cm−1 and 1633 cm−1 in the Raman spectra recorded using the microlenses with (solid circle) and without (hollow circle) side lenses.
Enlarge this image.Figure 6. (a) SEM top view of the single microlens without side lenses. The scale bar is 30 μm. (b) SEM top view of the non-uniform microlens array. The scale bar is 30 μm. (c,d) Experimental electric field distribution (left) alongside the simulated electric field distribution (right) of (c) the single microlens and (d) the non-uniform microlens array, respectively. The white dotted line indicates the position of the focal plane based on the experimental results. (e) Comparison between the experimental focal length and the simulated values based on four repeated experiments.
Supplementary Materials
The following supporting information can be downloaded at
References
1. Liu, X.; Xu, Z.; Fang, G.; Li, N.; Hasi, W. SERS combined with self-optimizing machine learning algorithm for quantitative detection of norfloxacin and ciprofloxacin in milk. Spectrochim. Acta Part A Mol. Biomol. Spectrosc.; 2025; 329, 125641. [DOI: https://dx.doi.org/10.1016/j.saa.2024.125641]
2. Lombardi, J.R. Enhanced by organic surfaces. Nat. Mater.; 2017; 16, pp. 878-880. [DOI: https://dx.doi.org/10.1038/nmat4958]
3. Han, X.X.; Rodriguez, R.S.; Haynes, C.L.; Ozaki, Y.; Zhao, B. Surface-enhanced Raman spectroscopy. Nat. Rev. Methods Primers; 2022; 1, 87. [DOI: https://dx.doi.org/10.1038/s43586-021-00083-6]
4. Li, L.-F.; Zhou, R.; Cui, J.-Q.; Yan, H.-P.; Wang, Z.-Z. Improved uniformity of deposited metallic layer on hybrid micro/nano-structured Si substrates fabricated by two-step laser ablation for SERS application. J. Cent. South Univ.; 2022; 29, pp. 3312-3322. [DOI: https://dx.doi.org/10.1007/s11771-022-5154-y]
5. Yi, J.; You, E.-M.; Hu, R.; Wu, D.-Y.; Liu, G.-K.; Yang, Z.-L.; Zhang, H.; Gu, Y.; Wang, Y.-H.; Wang, X. et al. Surface-enhanced Raman spectroscopy: A half-century historical perspective. Chem. Soc. Rev.; 2025; 54, pp. 1453-1551. [DOI: https://dx.doi.org/10.1039/D4CS00883A] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39715320]
6. Kneipp, J.; Seifert, S.; Gärber, F. SERS microscopy as a tool for comprehensive biochemical characterization in complex samples. Chem. Soc. Rev.; 2024; 53, pp. 7641-7656. [DOI: https://dx.doi.org/10.1039/D4CS00460D] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38934892]
7. Xu, Y.; Aljuhani, W.; Zhang, Y.; Ye, Z.; Li, C.; Bell, S.E.J. A practical approach to quantitative analytical surface-enhanced Raman spectroscopy. Chem. Soc. Rev.; 2025; 54, pp. 62-84. [DOI: https://dx.doi.org/10.1039/D4CS00861H] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39569575]
8. Yi, J.; You, E.-M.; Liu, G.-K.; Tian, Z.-Q. AI–nano-driven surface-enhanced Raman spectroscopy for marketable technologies. Nat. Nanotechnol.; 2024; 19, pp. 1758-1762. [DOI: https://dx.doi.org/10.1038/s41565-024-01825-9]
9. Wang, H.-L.; You, E.-M.; Panneerselvam, R.; Ding, S.-Y.; Tian, Z.-Q. Advances of surface-enhanced Raman and IR spectroscopies: From nano/microstructures to macro-optical design. Light Sci. Appl.; 2021; 10, 161. [DOI: https://dx.doi.org/10.1038/s41377-021-00599-2]
10. Yang, Y.; Creedon, N.; O’Riordan, A.; Lovera, P. Surface Enhanced Raman Spectroscopy: Applications in Agriculture and Food Safety. Photonics; 2021; 8, 568. [DOI: https://dx.doi.org/10.3390/photonics8120568]
11. Vatnik, I.D.; Gorbunov, O.A.; Churkin, D.V. Generation of Narrow Modes in Random Raman Fiber Laser Based on Multimode Fiber. Photonics; 2024; 11, 2. [DOI: https://dx.doi.org/10.3390/photonics11010002]
12. Yamamoto, Y.S.; Oshima, Y.; Shinzawa, H.; Katagiri, T.; Matsuura, Y.; Ozaki, Y.; Sato, H. Subsurface sensing of biomedical tissues using a miniaturized Raman probe: Study of thin-layered model samples. Anal. Chim. Acta; 2008; 619, pp. 8-13. [DOI: https://dx.doi.org/10.1016/j.aca.2008.02.027] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18539166]
13. Zhou, Y.; Zhang, Y.; Xie, H.; Wu, Z.; Shi, B.; Lin, L.L.; Ye, J. In Vivo Surface-Enhanced Transmission Raman Spectroscopy and Impact of Frozen Biological Tissues on Lesion Depth Prediction. ACS Nano; 2024; 18, pp. 35393-35404. [DOI: https://dx.doi.org/10.1021/acsnano.4c12469]
14. Valentino, G.M.; Shetty, P.P.; Chauhan, A.; Krogstad, J.A.; Weihs, T.P.; Hemker, K.J. Nanotwin formation in Ni–Mo–W alloys deposited by dc magnetron sputtering. Scr. Mater.; 2020; 186, pp. 247-252. [DOI: https://dx.doi.org/10.1016/j.scriptamat.2020.05.031]
15. Park, J.E.; Yonet-Tanyeri, N.; Vander Ende, E.; Henry, A.-I.; Perez White, B.E.; Mrksich, M.; Van Duyne, R.P. Plasmonic Microneedle Arrays for in Situ Sensing with Surface-Enhanced Raman Spectroscopy (SERS). Nano Lett.; 2019; 19, pp. 6862-6868. [DOI: https://dx.doi.org/10.1021/acs.nanolett.9b02070]
16. Simas, M.V.; Olaniyan, P.O.; Hati, S.; Davis, G.A., Jr.; Anspach, G.; Goodpaster, J.V.; Manicke, N.E.; Sardar, R. Superhydrophobic Surface Modification of Polymer Microneedles Enables Fabrication of Multimodal Surface-Enhanced Raman Spectroscopy and Mass Spectrometry Substrates for Synthetic Drug Detection in Blood Plasma. ACS Appl. Mater. Interfaces; 2023; 15, pp. 46681-46696. [DOI: https://dx.doi.org/10.1021/acsami.3c10174] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37769194]
17. Peng, C.; Jin, L.; Wang, F.; Yang, H.; He, H. Laser transparent multiplexed SERS microneedles for in situ and real-time detection of inflammation. Biosens. Bioelectron.; 2023; 225, 115079. [DOI: https://dx.doi.org/10.1016/j.bios.2023.115079] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36738731]
18. Fardian-Melamed, N.; Skripka, A.; Ursprung, B.; Lee, C.; Darlington, T.P.; Teitelboim, A.; Qi, X.; Wang, M.; Gerton, J.M.; Cohen, B.E. et al. Infrared nanosensors of piconewton to micronewton forces. Nature; 2025; 637, pp. 70-75. [DOI: https://dx.doi.org/10.1038/s41586-024-08221-2]
19. Deng, Y.; Liu, Z.; Song, C.; Wu, J.; Liu, Y.; Wu, Y. Topology Optimization-Based Computational Design Methodology for Surface Plasmon Polaritons. Plasmonics; 2015; 10, pp. 569-583. [DOI: https://dx.doi.org/10.1007/s11468-014-9842-9]
20. Veluthandath, A.V.; Murugan, G.S. Photonic Nanojet Generation Using Integrated Silicon Photonic Chip with Hemispherical Structures. Photonics; 2021; 8, 586. [DOI: https://dx.doi.org/10.3390/photonics8120586]
21. Allen, A.; Waldron, A.; Ottaway, J.; Carter, J.; Angel, S. EXPRESS: Hyperspectral Raman Imaging Using a Spatial Heterodyne Raman Spectrometer with a Microlens Array. Appl. Spectrosc.; 2020; 74, 000370282090622. [DOI: https://dx.doi.org/10.1177/0003702820906222] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32031013]
22. Xie, C.-Z.; Li, C.-H.; Chang, Y.-C.; Chen, Y.-F. Optofluidic Accumulation of Silica Beads on Gel-Based Three-Dimensional SERS Substrate To Enhance Sensitivity Using Multiple Photonic Nanojets. ACS Appl. Mater. Interfaces; 2023; 15, pp. 31703-31710. [DOI: https://dx.doi.org/10.1021/acsami.3c04569] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37343114]
23. Fales, A.M.; Strobbia, P.; Vo-Dinh, T.; Ilev, I.K.; Pfefer, T.J. 3D-printed phantoms for characterizing SERS nanoparticle detectability in turbid media. Analyst; 2020; 145, pp. 6045-6053. [DOI: https://dx.doi.org/10.1039/D0AN01295E] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32766656]
24. Ren, Y.-X.; Frueh, J.; Zhang, Z.; Rutkowski, S.; Zhou, Y.; Mao, H.; Kong, C.; Tverdokhlebov, S.I.; Liu, W.; Wong, K.K.Y. et al. Topologically protected optical pulling force on synthetic particles through photonic nanojet. Nanophotonics; 2024; 13, pp. 239-249. [DOI: https://dx.doi.org/10.1515/nanoph-2023-0740] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39635297]
25. Uma, S.; Jayesh, G.; Gokul, N.; Basudev, R. Optical trapping with photonic nanojets using various sized polystyrene spheres yields two different regimes. Proceedings of the SPIE Photonics Europe; Strasbourg, France, 7–12 April 2024; 129910U.
26. Ma, L.; Gu, G.; Zhang, Y.; Yang, H. A Simulation Study of Photonic Nanojet Generated by a Combination of a Dielectric Microsphere and a Hemisphere Lens. Proceedings of the 2022 Asia Communications and Photonics Conference (ACP); Shenzhen, China, 5–8 November 2022; pp. 1995-1997.
27. Gašparić, V.; Ristić, D.; Gebavi, H.; Ivanda, M. Resolution and signal enhancement of Raman mapping by photonic nanojet of a microsphere. Appl. Surf. Sci.; 2021; 545, 149036. [DOI: https://dx.doi.org/10.1016/j.apsusc.2021.149036]
28. Dantham, V.R.; Bisht, P.B.; Namboodiri, C.K.R. Enhancement of Raman scattering by two orders of magnitude using photonic nanojet of a microsphere. J. Appl. Phys.; 2011; 109, 103103. [DOI: https://dx.doi.org/10.1063/1.3590156]
29. Das, G.M.; Laha, R.; Dantham, V.R. Photonic nanojet-mediated SERS technique for enhancing the Raman scattering of a few molecules. J. Raman Spectrosc.; 2016; 47, pp. 895-900. [DOI: https://dx.doi.org/10.1002/jrs.4928]
30. Das, G.M.; Ringne, A.B.; Dantham, V.R.; Easwaran, R.K.; Laha, R. Numerical investigations on photonic nanojet mediated surface enhanced Raman scattering and fluorescence techniques. Opt. Express; 2017; 25, pp. 19822-19831. [DOI: https://dx.doi.org/10.1364/OE.25.019822] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29041669]
31. Ling, J.; Wang, Y.; Guo, J.; Liu, X.; Wang, X. Near-Field Nano-Focusing and Nano-Imaging of Dielectric Microparticle Lenses. Nanomaterials; 2024; 14, 1974. [DOI: https://dx.doi.org/10.3390/nano14231974]
32. Heifetz, A.; Simpson, J.J.; Kong, S.-C.; Taflove, A.; Backman, V. Subdiffraction optical resolution of a gold nanosphere located within the nanojet of a Mie-resonant dielectric microsphere. Opt. Express; 2007; 15, pp. 17334-17342. [DOI: https://dx.doi.org/10.1364/OE.15.017334] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19551027]
33. Maslov, A.V.; Astratov, V.N. Resolution and Reciprocity in Microspherical Nanoscopy: Point-Spread Function Versus Photonic Nanojets. Phys. Rev. Appl.; 2019; 11, 064004. [DOI: https://dx.doi.org/10.1103/PhysRevApplied.11.064004]
34. Chai, K.; Dzulqurnain, D.D.B.; Su’ait, M.S.; Ahmad, A.; Ataollahi, N.; Lee, T.K. Real-Time Raman Spectroscopy of Photopolymerization Dynamics in Ethylene Glycol Methyl Ether Acrylate-g-Epoxidized Natural Rubber. ACS Appl. Polym. Mater.; 2025; 7, pp. 1771-1783. [DOI: https://dx.doi.org/10.1021/acsapm.4c03593]
35. Li, C.; Huang, Y.; Lai, K.; Rasco, B.A.; Fan, Y. Analysis of trace methylene blue in fish muscles using ultra-sensitive surface-enhanced Raman spectroscopy. Food Control; 2016; 65, pp. 99-105. [DOI: https://dx.doi.org/10.1016/j.foodcont.2016.01.017]
36. Li, R.; Jiao, Y.; Tang, H.; Shi, Z.; Zhang, Y.; Xiao, L. Photonic Nanojet Generated by a Coated Hemisphere with Tunable Core. Proceedings of the 2024 International Applied Computational Electromagnetics Society Symposium (ACES-China); Xi’an, China, 16–19 August 2024; pp. 1-3.
37. Huang, C.; Li, D.; He, T.; Peng, Y.; Zhou, W.; Yang, Z.; Xu, J.; Wang, Q. Large Quadratic Electro-Optic Effect of the PLZT Thin Films for Optical Communication Integrated Devices. ACS Photonics; 2020; 7, pp. 3166-3176. [DOI: https://dx.doi.org/10.1021/acsphotonics.0c01234]
38. Keiichi, N. Epitaxial PLZT waveguide technologies for integrated photonics. Proceedings of the Integrated Optoelectronic Devices 2005; San Jose, CA, USA, 22–27 January 2005; pp. 34-43.
39. Wang, M.; Yan, Y.; Zhao, C.; Jiang, Y. Laser Parallel Nanofabrication of Dual-Au-Nanoholes on Microsphere-Lens-Array for Polarization-Dependent SERS. Adv. Mater. Technol.; 2024; 9, 2302244. [DOI: https://dx.doi.org/10.1002/admt.202302244]
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.