Holographic optical elements have emerged as an important technology in modern photonics, significantly advancing various fields through their versatile applications [1,2]. Holographic optical element (HOE) waveguides are integral components in the development of augmented and virtual reality devices, facilitating the redirection of light to produce immersive, high-quality visual experiences [3–5]. Notably, products such as the Microsoft HoloLens [6] and Magic Leap One [7] utilize HOE waveguides to enhance user experiences. Additionally, HOE waveguides are under development for solar energy harvesting [8], aiming to improve the performance of photovoltaic cells through efficient sunlight guidance. Furthermore, in biomedical imaging, HOE waveguides can enhance the precision and clarity of diagnostic techniques by directing light within compact [9,10] intricate devices.

These applications illustrate the versatility of HOE waveguides in managing optical signals. Leveraging holographic principles, the design and fabrication of HOE waveguides allow for precise and diverse redirection of light, facilitating their integration into complex systems.

The capability of HOE waveguides to guide light effectively is well-documented [11–15], underscoring their importance. This proficient management of optical signals highlights their critical role in optimizing signal handling and system design across a wide range of advanced technological applications.

Research into HOE waveguides is driven not only by their diverse functionalities but also by their effective modeling and efficient fabrication [13,14,16]. Holography involves two main processes: recording and reconstruction. The recording process employs a photosensitive material with properties that can be photochemically altered under light exposure. Reconstruction occurs when the recorded HOE structure is illuminated with a light beam having a wavefront similar to the one used during recording. In the case of HOE waveguides, the reconstructed beam is produced at an angle that ensures it meets the material-air interface at or above the critical angle, causing the diffracted beam to undergo total internal reflection (TIR). By carefully selecting the reconstruction wavelength and considering the desired input and output angles, the characteristic spacing and orientation of the fringes for a waveguide can be determined.

These parameters, once established, allow for the determination of the recording conditions for an HOE waveguide, as detailed in Section 2.

Reports detailing a flexible recording process demonstrate how a grating can be recorded using one wavelength and operate as a coupler at a different wavelength, eliminating the need for complex and bulky optical alignments. This flexible recording process has been demonstrated using a 532 nm wavelength to identify the Bragg angles necessary for fabricating volume phase transmission holographic gratings. The gratings operate as waveguides in either the in-coupler or out-coupler regime when probed with a 632.8 nm wavelength [16]. Additionally, a detailed model for determining the recording angles required for the fabrication of HOE waveguides across a range of recording wavelengths was reported. This model has highlighted the potential of transmission gratings recorded at 532 nm for waveguiding at 632.8 nm, and reflection gratings recorded at 532 nm for waveguiding at 405 nm, with experimental verifications [13,17,18].

A similar approach to HOE waveguide design has been utilized in Refs. [19–21], where signal delivery and extraction from devices such as fluorescence-based sensing systems were investigated. In such systems, the delivery of an excitation signal and the extraction of subsequent fluorescence are critical. For instance, luminescence-based sensors use specific dyes that emit light when excited by a particular wavelength [19,20]. The intensity and wavelength of the emitted light can provide information about the environment or analyte being studied.

This study presents a design concept aligned with the demonstration in Fig. 1, representing a configuration that progresses toward wearable applications. The light from the LED source is directed by the HOE-guided signal to a photoluminescence sensor embedded in a culture chamber, positioned in direct contact with the sensing region, such as a wound site. The photoluminescent signal is then guided out of the system for characterization by a photon detector, facilitating the monitoring of quenched fluorescence, indicative of relevant parameters, such as oxygen levels.

To achieve this, a prototype sensing system was fabricated to evaluate the feasibility of the operational design. It comprises photopolymerizable hybrid sol-gel (PHSG)-based HOE slanted gratings serving as couplers and a sensing chamber embedded in a polydimethylsiloxane (PDMS) layer. PDMS was selected for its compatibility with microfluidic system integration and its suitability for photoluminescence-based sensing within the embedded chamber. Additionally, its widespread use in microfluidic devices [22,23] and prominence in wearable applications made it a practical and accessible choice for this study, combining functional advantages with ease of implementation. The HOE couplers were recorded using a 476.5 nm laser, with the inter-beam angle set to achieve a spatial frequency of 1720 lines/mm. The sensor chamber was filled with a 630–660 nm absorbing methylene blue solution that is characterized by a peak fluorescence signal at 700 nm. An HOE, functioning as an in-coupler at 632.8 nm, guided the excitation signal into the sample chamber, while another HOE, operating as an out-coupler at 700 nm, extracted the fluorescence signal.

This is the first instance, to the authors’ knowledge, of fabricating couplers in the water- and thermally resistant, holographically capable PHSG material for photoluminescence-based sensing [21,24,25]. Compared to other holographic materials, PHSG provides a unique balance of optical performance and robustness. Most photopolymers offer flexibility and ease of processing but are often susceptible to moisture-induced degradation and lower long-term stability. Dichromated gelatin provides high diffraction efficiency but lacks environmental resilience. Silver halides exhibit excellent diffraction efficiency and long-term stability but require complex chemical processing. To ensure efficient performance of the HOE waveguiding system, several strategies were employed. Firstly, various recording energies ($50-750\mathrm{mJ}/{\mathrm{cm}}^2$) were used to find the optimal fabricating PHSG-based coupler conditions. Secondly, different recording geometries were explored to examine waveguide versatility for signal management. Thirdly, angular and peristrophic multiplexing was utilized to couple one probe signal in four different directions. This advancement offers a robust holographic framework for developing systems that enable simultaneous delivery and extraction of optical signals using PHSG-based HOE waveguides.

HOE coupler operation is defined by the gratings fringe orientation and fringe spacing. The fringe orientation should enable beam reconstruction in a direction at an angle equal to or greater than the angle required for TIR between two material boundaries, such as air and the photosensitive material. This angle, ${\theta }_{\mathrm{TIR}}$, is determined by the refractive index of the materials. A software tool was developed to calculate the recording parameters of the holographic waveguide based on the operation wavelength, recording wavelength, and layer refractive index. The first step in designing a holographic coupler involves considering the device’s operational wavelength ($\lambda$) and input angle (${\theta }_1$) within the medium. The Bragg law is used to model the characteristic fringe spacing ($\mathrm{\Lambda }$) of a slanted volume transmission mode grating with these operational parameters: $$\mathrm{\Lambda }=\frac{m\lambda }{2n\sin ({\theta }_B)},$$(1)where $m$ is the order of diffraction, $n$ is the refractive index of the material, and ${\theta }_B$ defined as $${\theta }_B=\frac{{\theta }_1-{\theta }_{\mathrm{TIR}}}{2}$$(2)is the Bragg angle. At normal incidence (${\theta }_1=0°$), ${\theta }_B$ becomes half of ${\theta }_{\mathrm{TIR}}$ as illustrated in Fig. 2(a). Similarly, under these conditions, the angular position of the fringes, determined by the angle between the fringe and the normal to the layer characterized by the slant angle (${\theta }_s$), is equal to ${\theta }_B$.

These parameters are used to calculate the spatial frequency [Eq. (1)], defining the precise grating configuration necessary for the HOE to function optimally at the designated wavelength, input angle, and TIR angle, suited for the targeted application.

Recording a transmission mode volume coupler at the operational wavelength of 632.8 nm is not feasible because interference cannot occur within the material’s volume. This is because a beam undergoing total internal reflection, when incident at an angle of 42° or greater, cannot penetrate the material boundary, as shown by the dotted ${\theta }_{\mathrm{TIR}}$ beam in Fig. 2. Therefore, using a shorter wavelength, which results in a smaller ${\theta }_B$ and a reduced incidence angle, is necessary for holographic coupler development. This approach eliminates the need for bulky optics, such as prisms, to help minimize this angle [26]. By utilizing the operational fringe spacing value obtained from Eq. (1), the recording ${\theta }_B^{\prime }$ can be determined for an appropriate recording wavelength through a rearrangement of the equation [13,16,27].

This approach enables the design of an HOE structure with the same fringe orientation and spacing as the one needed for the operational wavelength, while recording at a different wavelength, as demonstrated in Fig. 2(b). Finally, to adapt the model for practical use in an optical setup, Snell’s law was applied to determine the recording angles in air.

In applications where the incoming signal (1) experiences TIR at normal incidence, (2) propagates efficiently through the system, and ultimately (3) strikes the center of a light-interrogated region, the HOE design is important to overall system functionality. Understanding the optical path of the coupled signal within the material is important for optimizing the system operation. The number of reflections within the medium is directly influenced by the angle at which the light enters the layer and the mutual position of the HOE and the interrogated region.

This research focuses on a waveguide-assisted approach to photoluminescence-based sensing. For waveguide integration into a microfluidic chamber embedded into PDMS, device dimensions and grating parameters are critical for system operation. Compact dimensions are essential for device portability, packaging, and broader applicability. In this model, where the HOE substrate is a microscopic glass slide (1 mm thick) positioned atop the PDMS system (5 mm thick), a total height of 6 mm is considered. Both the HOE coupler and sensing chamber have a diameter of 10 mm. To evaluate the required system dimensions for accurate central illumination of the interrogated chamber, with minimal signal losses, a range of ${\theta }_{\mathrm{TIR}}$ angles are considered.

In one version of the device, the culture chamber is placed in direct contact with the sensing site (wound), while the HOE is positioned on the opposite plane, enabling direct illumination without the need for any reflections. This configuration, however, limits the possibility of parallel interrogation by other techniques (e.g., phase contrast or fluorescence microscopy), as the HOE would need to be in close proximity to the interrogated region, thereby obstructing direct access to it. To avoid this limitation, we propose the HOE to be located at some distance from the sensing chamber. This distance will depend on whether the HOE and the sensing chamber are positioned on opposite sides of the PDMS layer (Fig. 1) or on the same side of the PDMS layer [Fig. 3(a)]. In the following analysis, both the HOE and the sensing chamber are placed on the same surface side. This arrangement was selected as proof-of-concept and for the ease of preparation in our lab. Consequently, only an odd number of total internal reflection (TIR) events occur before the excitation beam interacts with the sensor chamber.

Even though PDMS is transparent in the visible range and losses from TIR are expected to be minimal, with increased propagation path of the beam an increased optical loss can be expected. This is mainly due to Rayleigh scattering, caused by any fluctuations of the density and the refractive index in the polymer material. Other sources that can be responsible for optical loss can be traced back to contaminating dust and pores from the fabrication process. Minimizing the number of TIR bounces reduces the propagation length and thus the optical losses and supports a more compact sensor design, which is particularly beneficial for integration into a patch-based framework. To balance these considerations, the modeling focused on configurations with either one or three TIR events, ensuring sufficient chamber excitation. The following equation was used to determine the required distance ($D$) between the HOE in-coupler and sensing chamber: $$D=(N+1)h\tan {\theta }_{\mathrm{TIR}},$$(3)where $N$ is the number of TIR events and $h$ is the thickness of the PDMS layer (assuming the HOE thickness is very small in comparison to the PDMS layer). The relative positions of the HOE and the sensor chamber are presented in Fig. 3(a).

As shown in Fig. 3(b), angles of incidence within the substrate where ${\theta }_{\mathrm{incident}}>{\theta }_{\mathrm{TIR}}$, require a longer propagation path to reach the center of the sensing chamber.

The distance between the HOE and the sensing region center can range from 1.18 to 3.27 cm for ${\theta }_{\mathrm{TIR}}$ values between 42° and 80° at the selected $N$ values. Miniaturization of these systems is desirable, as it allows for the development of compact, portable devices suitable for a wide range of applications. Miniature sensing systems offer significant advantages, including ease of integration into various environments, improved patient comfort in medical applications, and enhanced portability for field use. Therefore, employing smaller ${\theta }_{\mathrm{TIR}}$ values and consequently shorter devices is advantageous to our study. This system demonstrates versatility as it can be adapted for various applications requiring different sizes of the sensing element.

In systems with an optical path of 1.18 cm (one TIR event), losses per bounce can be considered negligible. For longer optical paths and larger devices, modeling of these losses will include considering the sum of the absorption coefficient of the material and the scattering, which will increase with the number of TIR events occurring in the PDMS layer.

It has been demonstrated that HOE waveguides can be used for both in-coupling [13–17] and out-coupling signals [14,15] in an optical device as shown in Fig. 2(a). HOE couplers, like all volume phase transmission holographic structures, can reconstruct either of the light wavefronts used during the recording process. Generally, for in-coupling, the light beam with a predetermined wavelength is incident on the coupler at the operational input angle, ${\theta }_1$. Consequently, a beam propagating at the ${\theta }_{\mathrm{TIR}}$ angle is reconstructed, and thus the beam propagates through the layer by TIR. In contrast, for out-coupling with the same HOE properties, the expected input angle is ${\theta }_{\mathrm{TIR}}$, and the reconstruction will occur at ${\theta }_1$. For example, a signal propagating within a material at a known ${\theta }_{\mathrm{TIR}}$ that meets the Bragg condition of an HOE along its path will be out-coupled from the material at a known angle ${\theta }_1$ as demonstrated in Fig. 4(a). For application in a sensor based on photoluminescence the properties of the second HOE will have to be tuned for operation at a different wavelength. This is utilized in our prototype system, where the dye solution filling the photoluminescence chamber is excited at one wavelength, in-coupled by the first HOE, and emits a fluorescent signal at a different wavelength, which is out-coupled by the second HOE. A theoretical model is developed to calculate the properties of the two HOEs operating at the two different wavelengths simultaneously. The model can be adapted for a range of ${\theta }_{\mathrm{TIR}}$, ${\theta }_1$, wavelengths, and optical materials.

This section evaluates the feasibility of integrating an HOE for out-coupling to characterize a fluorescent signal in a PDMS system. Fluorescence emitted from the sensing chamber can be modeled as a point source signal. Our model shows that a point source, which emits light uniformly in all directions, naturally directs 27% of the light horizontally (left or right) based on TIR conditions and 23% vertically (upward or downward), as seen in Table 1. Assuming an intensity of $1\mathrm{mW}/{\mathrm{cm}}^2$ from the source and a chamber diameter of 10 mm, the total power directed via both TIR and upward/downward illumination from the point source is similar.Table 1.

Area, Area Percentage, and Power for Light Traveling in Two Directions

Therefore, selecting the direction for light collection, whether from the top or side of the point source, does not differ significantly in efficiency. However, to enable future diagnostics of the chamber area, such as microscopic imaging, the top central region of the device is not preferred for out-coupling. To facilitate such signal collection, this study designs a waveguide to out-couple the incoming fluorescent wavelength at a known ${\theta }_{\mathrm{TIR}}$ for diagnostics and integrating it into the PDMS system in the location shown in Fig. 1.

The out-coupling waveguide can be fabricated by employing the method described in Section 2. For minimum disturbance of the recording setup the out-coupler has the same spatial frequency but a different slant angle. This is to ensure that the fluorescence signal at peak wavelength is incident to the out-coupler at the Bragg angle.

To study the versatility of the HOEs as in- or out-couplers, waveguide multiplexing was employed, using both angular and peristrophic multiplexing to enable simultaneous signal guidance and detection. Multiplexing can provide HOE waveguide fabrication operating at the same ${\theta }_{\mathrm{TIR}}$ and reconstruction wavelength, allowing coupling in different directions. Angular multiplexing was achieved by maintaining the fringe spacing of the HOE waveguide, as described in Section 2.B, while changing the fringe orientation of the grating [Fig. 4(b)]. Recording on both sides of the surface normal (positive and negative) results in the reconstruction of ${\theta }_{\mathrm{TIR}}$ in the opposite direction, as visualized in Fig. 4(b).

Peristrophic multiplexing was implemented by rotating the PHSG layer 90° around its normal. The already angularly multiplexed PHSG layer was rotated and again angularly multiplexed, resulting in four 632.8 nm waveguides stored in the material. Illuminating the multiplexed structure with a 632.8 nm beam at 90° incidence would achieve ${\theta }_{\mathrm{TIR}}$ for four different directions: left, right, up, and down. HOE waveguide multiplexing would allow for simultaneous probing and site excitation, enabling the characterization of more than one sensing region.

The development process of the PHSG, as reported in previous research [15,28,29], is employed in this study. The four-stage process involves synthesizing the sol base and then adding modifiers to promote its transformation into the PHSG gel. The process initiates with (1) the complexation of zirconium (IV) propoxide (ZPO, 0.727 g) to zirconium complex (ZCO) in the presence of methacrylic acid (MAA, 0.135 g), and (2) the generation of a hybrid silicate matrix via the hydrolysis of 3-trimethoxysilylpropyl methacrylate (MAPTMS, 15 g) with 0.1 mol/L nitric acid (${\mathrm{HNO}}_3$, 1.6 mL).

In the subsequent step, this matrix is combined with the ZCO, which forms the sol base. The third step involves the dropwise addition of deionized water (1.088 g), facilitating further hydrolysis. In the final stage, gelation proceeds following the addition of photosensitizer bis(eta.5-2,4-cylcopentadien-1-yl)-bis(2,6-difluoro-3-(1H-pyrrol-1-yl)-phenyl) titanium (Omnirad 784, 0.05 g), a modifier (3-aminopropyl)triethoxysilane (APTES, 0.450 g), and solvent isopropanol (4 mL) into 10 g of the sol base. Following 10 h of stirring the photosensitive hybrid sol-gel synthesis is completed.

Fresh (0-day) PHSG sol was drop-casted onto a leveled glass slide ($37.5\mathrm{mm}\times 25\mathrm{mm}$, 1 mm thick) in a dark room at ambient conditions. The layers were then dried in the oven (Binder, model ED56) at 125°C for a period of 45 min.

PDMS was prepared by combining a silicone elastomer base with a silicone elastomer curing agent in a ratio of 10:1. The mixture was then transferred into two 3D-printed molds and allowed to degas for 30 min. The PDMS filled molds, consisting of a lid with an input channel and a base with a chamber, were cured in the oven at 60°C for 1 h. After cooling, the PDMS lid and base were bonded together using liquid PDMS and returned to the oven for an additional 10 min. Following this step, the PDMS-embedded chamber was fully enclosed, with access available only through the channel [see Fig. 9(a) shown later]. The PDMS system dimensions are $25\mathrm{mm}\times 10\mathrm{mm}\times 25\mathrm{mm}$, with a chamber diameter of 10 mm.

0.013 mmol/L solution of methylene blue in deionized water was used. This solution filled the PDMS chamber. The optical density and relative fluorescence of the solution were characterized using the SpectraMax M3 multi-mode microplate reader. Finally, a 3D-printed attachment was used to clamp the filled chamber to the glass slide.

Development of the 632.8 nm waveguide also employed angular multiplexing, which involved two subsequent recordings of HOEs with $+21.5°$ and $-21.5°$ slant angles in the same region of the PHSG layer as demonstrated in Fig. 4(a). For peristrophic waveguide multiplexing, the sample was first used to record two angularly multiplexed gratings with $+/-21.5°$ slant angles as in Fig. 4(b). The recorded sample was then rotated by 90° and two more angularly multiplexed gratings were recorded.

The optimum recording conditions for HOE coupler fabrication in PHSG were evaluated by recording at intensities of 2.5, 3, 5, 10, and $12.5\mathrm{mW}/{\mathrm{cm}}^2$ and exposure times 20 and 60 s. The development of 632.8 nm couplers at different recording conditions was characterized by monitoring real-time diffraction efficiency growth curves and Bragg selectivity curves to determine the optimal exposure.

Real-time growth curves and Bragg selectivity curves were measured with an in-house-built computer supported setup and software (Pantera). The setup consisted of a rotational stage and an Arduino board (Arduino Nano Every) linked photodiode (SLCD-61N5) measuring the intensity of a 632.8 nm transmitted beam. For Bragg selectivity curves the transmitted intensity was measured as the grating rotated through the Bragg angle. The Pantera software displays the transmitted intensity, which is then used to calculate the DE using the following equation: $$\mathrm{DE}=1-\left(\frac{I-I_0}{I_{\mathrm{in}}-I_0}\right),$$(4)where $I$ is the measured transmitted beam intensity, $I_0$ is the zero-bias intensity measured with no input light to account for the system’s baseline noise, and $I_{\mathrm{in}}$ is the incident probe beam intensity. This method for determining diffraction efficiency is justified by the fact that the recorded grating is a volume holographic grating, producing a single order of diffraction. Consequently, any light removed from the transmitted beam is assumed to be redirected into the diffracted beam.

The spectral characterization of the waveguides was performed by probing with a halogen light (AvaLight-HAL-S) and fiber-coupling the signal into the FLAMT-T-VIS-NIS- ES USB powered spectrometer for real-time visualization with the OceanView software.

Firstly, the impact of recording exposure energy on the DE of slanted volume transmission gratings for use as holographic couplers was examined. Figure 6 demonstrates the average DE achieved at each recording condition for gratings recorded using two different exposure times (20 and 60 s) and varying intensities (2.5, 3, 5, 10, and 12.5 mW/cm). Under the studied recording conditions ($50-750\mathrm{mJ}/{\mathrm{cm}}^2$), couplers were fabricated with a DE of 0.6, with a variation of $\pm 0.15$. This variation, along with the observed lack of trend, suggests that within the studied range, the recording process does not strongly depend on exposure energy, and the differences in DE may be attributed to variations in film thickness rather than inherent sensitivity to exposure energy. However, some degree of dependence may be revealed beyond the studied range and for different layer thicknesses. In the subsequent sections, an exposure of $3\mathrm{mW}/{\mathrm{cm}}^2$ for 20 s will be considered efficient for fabricating PHSG waveguides with a 0.6 DE.

The absorbance and fluorescence of the 0.013 mmol/L methylene blue water-based solution have been characterized as described in Section 3.B. Figure 7(a) reveals that the dye solution exhibits two absorption peaks at 633 nm and 660 nm. Upon excitation at 633 nm, the subsequent fluorescence observed at 700 nm is depicted in Fig. 7(b).

This data was used to design the 632.8 nm in-coupler and the 700 nm out-coupler as detailed in Table 2. Consequently, the in-coupling HOE was designed to operate at 632.8 nm, at normal incidence and ${\theta }_{\mathrm{TIR}}$, reconstruction angle of 43°. The out-coupler was designed to operate at 700 nm, with an input angle ${\theta }_{\mathrm{TIR}}^{\prime }$, of $-46°$ and reconstruction angle of 90°. Fluorescence is assumed to emit isotropically from a point source, covering the entire 360° domain. For angles relative to the surface normal, TIR occurs when light propagates at angles greater than the critical angle. In this case, for a critical angle of approximately $\pm 42°$, the portion of the fluorescence signal undergoing TIR corresponds to angles outside this range. Thus, fluorescence at 700 nm, with an input angle of $-46°$, will satisfy the TIR condition.

The conditions employed allow the recording of a 632.8 nm waveguide and a 700 nm waveguide, both satisfying approximately 56° angle difference between interfering beams. This arrangement allows to produce gratings of the same spatial frequency. Fringe orientation as shown in Fig. 4(a) and an increased slant angle (Table 2) will result in the structure operating as a 700 nm out-coupler.

Bragg selectivity curves of the couplers were obtained using a 405 nm laser probe to verify the recorded structures. As shown in Fig. 8, the 632.8 nm coupler, when probed with a 632.8 nm beam at normal incidence, exhibits coupling with a single peak in the transmitted order. When probed with a 405 nm laser, the grating reconstructs two diffraction orders at $-12°$ and $-59°$. Similarly, the 700 nm coupler, with a different slant, demonstrates two diffraction orders at 15° and 61° when probed with a 405 nm laser. This characterization confirmed the accuracy of the model.

Figure 9(a) shows the integration of the PDMS fluorescence chamber onto a glass substrate. Due to the refractive index mismatch between glass ($n=1.5$) and PDMS ($n=1.43$), at the sample-PDMS boundary, refraction will occur. The incident beam angle at the boundary is 43° and the refracted beam propagates in the PDMS layer at 46° from the normal to the layer. Once entered the PDMS layer, light remains guided through TIR (critical angle at the PDMS-air interface is 44°). While there is a refractive index mismatch between the sample and PDMS, its impact on optical losses is minimal. At an incidence angle of 43°, Fresnel losses at this interface are on the order of 0.2%, ensuring efficient transmission into the PDMS layer. To further minimize scattering and unintended losses, liquid PDMS was applied during assembly to solidify the material interfaces, eliminating air pockets and reducing refractive inconsistencies. For the development of a flexible wearable sensor device, the HOE will need to be fabricated in a sol-gel layer directly coated onto the PDMS layer. It is important to note that the PDMS chamber is intended to assess the feasibility of exciting a solution and extracting the fluorescence with HOE couplers before integration into a functional wearable sensing system.

Figure 9(b) demonstrates the performance of the system developed. The 632.8 nm laser beam is incident on the in-coupler, which excites the methylene blue in the chamber, and its fluorescence is guided out of the system by the out-coupler.

By situating an optical fiber attached to a spectrometer perpendicular (90°) to the 700 nm out-coupler, the signal was detected and characterized. The successful guidance of the 632.8 nm signal into the chamber and the extraction of the fluorescent signal are evidenced by the intensity versus wavelength graph shown in Fig. 9(c). This system showcases the capability of the PHSG HOE in- and out-couplers to guide a sufficient signal for the device operation and its characterization.

To evaluate the feasibility of using a flexible substrate for in-contact wearable sensing applications, such as a patch-based framework, a polycarbonate film (Bayer Makrofol DE 1-1 CC, 500 μm) was employed. Similar to the glass substrate samples, 0.50 mL of the PHSG solution was drop-cast onto the polycarbonate film and then dried in the oven. The preliminary results, as shown in Fig. 10(a), demonstrate the samples’ functionality in combination with a PDMS film, highlighting the PHSG coupler’s ability to reliably operate even when slightly deformed or curved, thus showcasing its potential for flexible and wearable sensor applications. Future work will explore the direct deposition of sol-gel onto PDMS, eliminating the use of an intermediate glass substrate and streamlining the fabrication process.

For simultaneous signal delivery and extraction, such as measuring analyte concentration at multiple sites, PHSG waveguide multiplexing can be employed. This sensing system would be capable of sensing a larger area, enabling more comprehensive and accurate monitoring of a region. Additionally, it could enable real-time mapping of analyte concentrations at specific sites, thereby enhancing the precision and effectiveness of a sensing system. The possibility of waveguide PHSG multiplexing for the same operational wavelength and input angle was investigated. Angular multiplexing was employed to change the fringe orientation while maintaining the same spatial frequency and slant angle.

Two couplers were multiplexed, with one HOE recorded at a ${\theta }_{\mathrm{TIR}}$ of $+43°$ and the other at –43°, using the optical arrangement described in Section 3.C. The Bragg selectivity curves of the two consecutively recorded couplers, probed with a 632.8 nm beam at normal incidence, are shown in Fig. 10(b). The two 70% DE gratings confirm the multiplexing capabilities of the PHSG layer and validate the model’s operation under the same Bragg conditions.

Figure 10(c) visualizes the simultaneous coupling of a 632.8 nm beam incident at 90° by the two multiplexed waveguides. Similarly, optical filtering capabilities of the HOE waveguide system for operation with a 632.8 nm beam at normal incidence are demonstrated in Fig. 10(d) where the incident beam is a broad-spectrum white light.

Additionally, the HOE structure was tested using two approaches. In Fig. 11(a), white light was coupled along the sample axis, with the 632.8 nm wavelength out-coupled at 90°. In Figs. 11(b) and 11(c), the white light illuminated the HOE externally, demonstrating its function as a classical transmissive element. Adjusting the input angle allowed different wavelengths to satisfy the Bragg condition.

The spectrum of the out-coupled signal, detailing the specific wavelengths achieved at different incidence angles, was obtained, and is presented in Fig. 11(d). It is worth noting that a very narrow part of the spectrum is selected, with an average full width at half maximum of 1.46 nm. If a highly divergent light source was employed, the need to satisfy the Bragg condition and a range of ${\theta }_{\mathrm{TIR}}$ for multiple wavelengths needs to be considered. This spectral analysis underscores the HOE’s capability to selectively filter and guide different wavelengths by tuning the incident angle, highlighting its potential to be used in versatile optical arrangements.

The study also explored peristrophic multiplexing within the PHSG layer. Initially, angular multiplexing was employed to record two 632.8 nm waveguides, functioning as both in- and out-couplers within a single plane. The layer was then rotated by 90°, and the same angular separation was used to record an additional set of two 632.8 nm waveguides (in- and out-couplers). This process effectively resulted in the development of four operational 632.8 nm waveguides within the same PHSG layer, currently achieving diffraction efficiency of 20% per grating.

Upon illumination, for the same incidence angle, ${\theta }_{\mathrm{TIR}}$ reconstruction was achieved in four distinct directions. This outcome indicates the successful implementation of peristrophic multiplexing to increase the density of functional optical paths within a single layer. The experimental results are illustrated in Fig. 12, which visually demonstrates the ${\theta }_{\mathrm{TIR}}$ reconstruction and the directional diversity achieved through multiplexing in a PHSG layer.

This study demonstrated for the first time the effective design, fabrication, and integration of HOE waveguides in a photopolymerizable hybrid sol-gel matrix. The successful recording of 632.8 nm and 700 nm in- and out-coupler regime waveguides using a single 476.5 nm optical arrangement further highlights the robustness and versatility of the HOE waveguide model employed.

The integration of these waveguides into a PDMS microfluidic system for the first time effectively guided excitation light at 632.8 nm and extracted fluorescence signal with a peak wavelength at 700 nm, showcasing their potential for application in wearable fluorescence-based sensing and diagnostics. Additionally, the study’s exploration of waveguide multiplexing in PHSG enabled simultaneous management of multiple optical signals operating at the same wavelength and incident angle. These advancements in efficient fabrication, high and reliable DE, and versatile signal management underscore the potential of PHSG-based HOE waveguides to enhance photonic device performance across a spectrum of applications.

Future work focuses on validating this system under application-relevant conditions, such as wound-like environments that simulate analyte concentration gradients and temperature variations at the sensing site for photoluminescence-based sensing. The studies will evaluate the device’s robustness, response time, and accuracy, offering critical insights into its performance for wearable sensing applications and will be published elsewhere. To further enhance functionality, multiplexing strategies will be explored, including the integration of HOE waveguides operating at different wavelengths for a fixed incident angle, as well as the combination of waveguides functioning in both transmission and reflection modes under the same incidence. These advancements aim to improve sensitivity, selectivity, and adaptability, supporting the development of a more versatile sensing platform.

Acknowledgment. The authors gratefully acknowledge the FOCAS Research Institute for providing access to facilities, equipment, and technical assistance. Additionally, they extend their appreciation to colleagues at TU Dublin for their valuable discussions and insights. Javier Arguelles Lopez is acknowledged for the design and development of the data acquisition system and Pantera software used in this study.