An integrated actuating and sensing system for light-addressable potentiometric sensor (LAPS) and light-actuated AC electroosmosis (LACE) operation

Abstract

To develop a lab on a chip (LOC) integrated with both sensor and actuator functions, a novel two-in-one system based on optical-driven manipulation and sensing in a microfluidics setup based on a hydrogenated amorphous silicon (a-Si:H) layer on an indium tin oxide/glass is first realized. A high-intensity discharge xenon lamp functioned as the light source, a chopper functioned as the modulated illumination for a certain frequency, and a self-designed optical path projected on the digital micromirror device controlled by the digital light processing module was established as the illumination input signal with the ability of dynamic movement of projected patterns. For light-addressable potentiometric sensor (LAPS) operation, alternating current (AC)-modulated illumination with a frequency of 800 Hz can be generated by the rotation speed of the chopper for photocurrent vs bias voltage characterization. The pH sensitivity, drift coefficient, and hysteresis width of the Si₃N₄ LAPS are 52.8 mV/pH, −3.2 mV/h, and 10.5 mV, respectively, which are comparable to the results from the conventional setup. With an identical two-in-one system, direct current illumination without chopper rotation and an AC bias voltage can be provided to an a-Si:H chip with a manipulation speed of 20 μm/s for magnetic beads with a diameter of 1 μm. The collection of magnetic beads by this light-actuated AC electroosmosis (LACE) operation at a frequency of 10 kHz can be easily realized. A fully customized design of an illumination path with less decay can be suggested to obtain a high efficiency of manipulation and a high signal-to-noise ratio of sensing. With this proposed setup, a potential LOC system based on LACE and LAPS is verified with the integration of a sensor and an actuator in a microfluidics setup for future point-of-care testing applications.

I. INTRODUCTION

In recent years, point-of-care testing (POCT) has received great attention for improving patient health. To meet these kinds of demands, integrated microfluidics with lab on a chip (LOC) or micro total analysis systems (μTAS) have been considered potential candidates with the advantages of a small sample and reagent volumes, a fast response, and less technician work and cost over conventional laboratory operations.¹ Because of the complicated composition of biomedical analytes, highly reliable and efficient bioanalytical techniques are required with trustworthy separations before measurement.² Various actuated magnetic-bead immunoassays have been proposed by means of a stand-alone LOC with an on-chip sensor to implement a practical and disposable device for acceptable, sensitive POCT.^3–5 In general, magnetic beads have been well proven to immobilize probe species to capture the target analyte with the advantages of a large surface-to-volume ratio and easy manipulation by an external magnetic field.^5–8 The smaller the diameter of the magnetic beads is, the larger the surface area will be in a fixed total volume. The surface of magnetic beads was successfully functionalized with different probe species including capture antibodies.^7,8 Even commercial magnetic beads are available with standardized kits for enzyme-linked immunosorbent assay (ELISA) detection.⁷ These well-constructed magnetic beads with functional groups can be used for reactions in a microfluidics system such as binding assays for better sensitivity or a lower limit of detection (LOD) with small amounts of target analyte for analytical applications. With the help of micromagnetic structures integrated in LOC and variable magnetic field operation,^7,9,10 magnetic beads labeled with biological analytes such as proteins,^4,7,11 nucleic acids,^12,13 viruses,^14–16 and cells^17,18 can be easily controlled to transport, separate, and mix fluids in microchannels with high-efficiency separation for subsequent analysis. However, the controllable magnetic force can be generated by extra magnetic electrodes or an external magnetic field, which is not easy to achieve at an affordable cost with POCT. This kind of technological development is similar to the application of alternating current (AC) electrokinetic mechanisms including dielectrophoresis (DEP)¹⁹ and AC electroosmosis (ACEO)²⁰ for particle manipulation in microfluidics. Generally, both systems are necessary to create a specific arrangement of conductive electrodes in microfluidics with the burdens of a costly, time-consuming, and technically loaded fabrication process and limited usage. The kind of constraints of DEP and ACEO has been dramatically improved by an emerging optically induced electrokinetic (OEK)-based technique by means of specific illumination and a corresponding adsorbed semiconductor layer (e.g., hydrogenated amorphous silicon, a-Si:H) to generate virtual electrodes in an efficient and flexible way.^21–23 Recently, an OEK method was successfully demonstrated for the manipulation of magnetic microbeads by commercial digital projectors, which can be considered a potential replacement of magnetic fields in the conventional limitation of system design.²⁴ In the setup of OEK operation, the AC frequency of the applied voltage between top and bottom indium tin oxide (ITO) electrodes can be used to provide polarization of the electrolyte and magnetic beads in microfluidics. Once the illumination pattern is driven by a direct current (DC) into the a-Si:H layer, a uniform electric field can be generated between illuminated and unilluminated parts as a function of virtual electrodes. Based on the requirement of microparticle manipulation, virtual electrodes with desired movements can be easily and quickly obtained by customer-designed patterns in the software.^24–26 Then, polarized particles can be attracted or repulsed by the combination of their polarity and the electric field gradient. OEK-based operations can be generally classified into two major parts according to the AC frequency range: optically induced dielectrophoresis (ODEP) (e.g., 30–150 kHz) and light-actuated AC electroosmosis (LACE) (e.g., 8–30 kHz). In our experience, cell- and magnetic bead-based manipulation can be suggested to operate in the ODEP^25,26 and LACE operation ranges,²⁴ respectively. The maximum manipulation speed of magnetic beads with a diameter of 1 μm can be 1441.1 μm/s with the proper setup of the thickness of the a-Si:H layer, operation parameters, and conductivity of the electrolyte. In particular, LACE operation can be simply used to concentrate, mix, group, transport, and pattern massive amounts of micro- or nano-particles (e.g., magnetic microbeads,²⁴ polystyrene bead,²⁷ and carbon nanoparticles²⁸) or biomolecules (e.g., nucleic acid²⁷ and antibodies²⁹), which is suitable for sample processing in the biosensing-based microfluidic system. Even with this superior manipulation working as an actuator, direct label-free sensing of functionalized magnetic beads is not available in the literature, which can be a chance for novel investigations.

A sensor structure similar to that of the a-Si:H layer has been widely investigated in the application of light-addressable potentiometric sensor (LAPS).³⁰ In the operation mechanism of LAPS, the DC bias voltage is given between a reference electrode and a bottom electrode to provide the depletion region of the semiconductor layer first. Then, illumination with a modulated AC frequency is applied to stimulate electron–hole pair generation in the depletion region and follows an AC photocurrent by carrier separation from the DC bias voltage.^31,32 With changes in the DC bias voltage, the operation mode of the semiconductor layer (e.g., the a-Si:H layer) can be modified from the accumulation, depletion, and inversion regions. The induced photocurrent of LAPS can be increased from the accumulation to the depletion region and saturated in the inversion region. On the other hand, the surface binding potential changes as the ion concentration of electrolyte changes (e.g., the hydrogen ion concentration, pH) and follows a parallel shift of the original characteristics of the photocurrent vs DC bias voltage curve. Then, the sensitivity of the ion concentration with a linear response range can be calculated by fitting the representative DC bias voltage of the ion concentration of the electrolyte. For example, the pH sensitivity of the conventional sensing membrane of LAPS ranges from 45 to 58 mV/pH in a pH range from 2 to 12.^30–35 With the movement of the illumination spot, different locations of sensors and their responses can be measured and collected. Afterward, all the responses of each point can be used to generate a two-dimensional (2D) chemical image for volume data and a clear picture in a complex sensing environment.³¹ With various surface modifications of LAPS, the quantitative signals of biological analytes such as the pH,^30,33,34 other ions,^35,36 proteins,³⁷ nucleic acids,^38,39 bacteria,⁴⁰ and cells^41–43 can be obtained with the advantages of an easy fabrication process, integration of microfluidics, and chemical imaging ability. The possibility of LAPS measurement for the gradient drug effect to monitor the acidification of cells by endotoxin-assembled magnetic beads with a magnetic field can also be found.⁴³

Based on the previous discussion, the similarity of photoelectric-driven mechanism and device structures (e.g., the a-Si:H layer) could be suggested to integrate OEK and LAPS for actuating and sensing functions in a single LOC. In contrast, the clear difference in operation parameters between OEK and LAPS is the key topic to overcome a novel LOC system. For example, DC and AC signals are applied in OEK and LAPS in the illumination part, respectively. In the bias voltage part, AC and DC signals are applied in OEK and LAPS, respectively. Because of the major requirement of two systems, the input signal of illumination with dynamic patterns and a bias voltage are rarely provided and switch smoothly between DC and AC in all currently available setups. To integrate LOCs with sensor and actuator in a high operational efficiency and low setup expansion, a system that can perform LAPS and LACE operations is first proposed and evaluated in this study.

II. MATERIALS AND METHODS

A. The two-in-one LOC system setup

To integrate the operations of basic chemical sensing characterization and magnetic-bead manipulation in a microfluidic system, a two-in-one (2-in-1) LOC system composed of LAPS and LACE functions is first designed and established in this study. First, a high output current of the LAPS and the manipulation force of LACE are the most important LOC performances, which is better for obtaining a light source with a high illumination intensity. Therefore, a high-intensity discharge (HID) xenon lamp (MR16, Prosper Optoelectronic, Taiwan) with a power of 32 W from an electronic ballast was selected. This HID xenon lamp was filled with xenon gas in a quartz tube, and the xenon electrons were released by an electronic ballast. A high-intensity illumination source can be obtained with the advantages of a broadband capability, stable brightness, and long lifetime.⁴⁴ The response of irradiance vs wavelength of the HID xenon lamp is shown in Fig. S1 in the supplementary material. As mentioned in Sec. I, the major difference in illumination between LAPS and LACE is the time-dependent modulation of the input signal such as alternating current (AC) and direct current (DC). To achieve modulated illumination with a certain frequency in LAPS measurements (e.g., AC mode) and constant-intensity illumination in LACE manipulations (e.g., DC mode) integrated with the same system, the optical chopper with its controller (MC1F10, Thorlabs, USA) was installed in the illumination path to obtain the possibility of two kinds of light modulations according to the rotation speed of the chopper. For example, DC and AC mode illuminations can be obtained by fixing the position and rotating the chopper. In addition, illumination patterns on the back side of an LOC chip must be controlled with desired procedures in the LAPS and LACE operation in this developed system. For example, a specific procedure including the movement of illumination patterns in the microfluidics setup functioned as LACE manipulation first, followed by a fixed illumination pattern functioning as LAPS measurement after manipulation, which could be easily performed. Therefore, a digital light processing (DLP) module (Light craft 6500, Texas Instrument, USA) was selected to reflect illumination and define patterns of illumination on the back side of LOC for the sensing area of LAPS and manipulation of LACE, respectively. The whole DLP module is composed of a DLPC900 controller and a digital micromirror device (DMD) module with an array of 1920 × 1080 pixels in an area of 16.51 × 9.14 mm².

After two key components were determined, the illumination path had to be well arranged to meet the requirement of a high uniformity of illumination (e.g., input signal to LOC chip) and less interference of undesired illumination (e.g., noise to LOC signal). First, the turned-on xenon lamp was worked as constant-intensity illumination in LACE through the open area of the fixed chopper. On the other hand, the rotating chopper with the open and block areas in the illumination path could be used to obtain the on and off illumination with a regular frequency, which can be applied as the input signal for LAPS measurements. The frequency of the on and off illumination can be decided by the spacing of open and block areas in the chopper and the rotating speed defined by the controller. With the high rotating speed of this chopper, the modulated frequency of the on–off ratio can be increased. The frequency of the chopper can be modified from 20 to 1000 Hz in this setup. Because of the spreading illumination of the HID xenon lamp and the dimensional limitation of the open area on the chopper, two focus lenses were placed between the HID xenon lamp and chopper to have a small illumination spot (e.g., <9 mm in diameter) as the projected area on the chopper, which is smaller than the dimension of the open area (e.g., 10.3 mm). The collimation lens next to the HID xenon lamp was first used to have a collimated beam as the parallel illumination path. Then, a focus lens was used to reduce the illumination area for passing through the open area of the chopper without obstruction. After illumination passed the chopper, a focus lens and a collimation lens were used to obtain a collimated beam, and a reflective lens was placed in the illumination path to reflect the illumination directly on the DMD. A circle with a diameter of approximately 18 mm was used for the projection area with a uniform illumination intensity to cover the whole surface of the DMD module. Moreover, the path and angle of reflection illumination from the DMD module was optimized according to the 12° tilt angle of the DMD for the projected illumination patterns on the back side of the LOC chip. To avoid the inference light from the reflection of the remaining area out of the DMD module, black tape covered the other areas of the DMD module to minimize undesired illumination. Two possible control modes of this DLP module including the pattern on the fly mode by the universal serial bus (USB) port and the video mode by the high-definition multimedia interface (HDMI) port were investigated. In the pattern on the fly mode, the built-in software can be used to control output patterns displayed on the DMD by a binary image file with a resolution of 1920 × 1080 pixels. On the other hand, dynamic patterns can be displayed on the DMD in the video mode by means of a video file designed by presentation slices. The desired patterns of LAPS or LACE can be easily achieved by the DLP controller and its software provided by Texas Instruments Inc. (TI). Two lenses with focal lengths of 50 and 25 mm with an inserted aperture were used to reduce the projected pattern to half of the DMD area (e.g., 8.25 × 4.57 mm²) on the back side of the LOC chip.

To collect the output photocurrent in the LAPS measurement, a low noise current preamplifier (SR570, Stanford Research System, USA) and a data acquisition card (USB-6351, National Instrument, USA) were used for bandpass filtering of noise and data acquisition, respectively. For an acceptable signal-to-noise ratio, the longer integration time to collect output photocurrent is necessary. The readout speed is approximately 1 pixel/s according to the balance between HID lamp illumination with the AC frequency modulation by a chopper and frame refresh rate of DMD. A function generator (33210A, Keysight Technologies, USA) was used to provide an AC signal of peak-to-peak voltage (Vpp) to the top and bottom ITO electrodes of the LOC chip in the LACE operation. Finally, a digital microscope (VHX-6000, Keyence, Taiwan) was used to monitor and record the manipulation of magnetic beads and light patterns in the LACE actuation from the top side of the LOC chip. Pictures of this 2-in-1 LOC system and schematic plots of the illumination path are shown in Figs. 1(a) and 1(b), respectively.

FIG. 1.

(a) Picture of the whole setup of the 2-in-1 LOC system including the a-Si:H chip and illustration of key parts and (b) a schematic plot of the illumination path and the corresponding optical components.

B. Sample preparations for LAPS and LACE

To confirm the performance of the LAPS measurement of this developed system, a conventional Si₃N₄ LAPS chip was fabricated using a double-side polished p-type silicon substrate (Summit-Tech, Taiwan) with a thickness of 500 μm, which is identical to that of our literature.⁴⁵ After a standard RCA clean, silicon dioxide (SiO₂) and silicon nitride (Si₃N₄) layers with thicknesses of 50 nm were grown in a high-temperature oxidation furnace tube by low-pressure chemical vapor deposition (LPCVD) on both sides. This Si₃N₄ layer can work as an etch stop layer on the back side during the wet etching process and as a sensing film in pH sensing characterization during LAPS measurement. An opening window with dimensions of 900 × 900 μm² on the back side was defined by a photolithography process, reactive ion etching (RIE) by the CF₄ gas plasma environment, and wet etching by buffered oxide etching (BOE). Then, the whole substrate was placed in a special wafer holder for a 4-in. wafer (Single4 LAA, AMMT GmbH, Germany). With the front-side protection of high thermal stability and the chemical resistance from this holder provided by poly-ether-ether-ketone (PEEK) and a special O-ring, the back side of the substrate covered by a Si₃N₄ layer with open areas was immersed in 20% potassium hydroxide (KOH) solution at 80 °C for wet etching of the Si substrate. The Si substrate was thinned down to a thickness of 100 μm, which was approximately controlled by the total etching time. The etched thickness of the Si substrate was measured with an alpha stepper (Dektak XT, BRUKER, USA), and side-view images were obtained with a digital microscope (VHX-6000, Keyence, Taiwan). After KOH etching, all remaining Si₃N₄ and SiO₂ layers were removed on the back side by RIE and BOE, respectively. A back-side metal contact was made by thermal evaporation, and an open metal hard mask of 1 × 1 cm² was used for an aluminum layer with a thickness of 300 nm and an illuminated area. With this ohmic contact layer, the induced photocurrent could be measured and collected. This substrate was cut into an area of 1.4 × 1.4 cm² and then attached to a printed circuit board (PCB) with a back-side open area of 1 × 1 cm² with silver glue. A polydimethylsiloxane (PDMS) tank with an open area of 1 × 1 cm² made from a designed mold was attached to the Si₃N₄ surface of this thin Si substrate for pH sensing characterization. A cross-sectional picture of this fabricated Si₃N₄ LAPS using a VHX-6000 digital microscope is shown in Fig. S2 in the supplementary material. For the LACE operation, a hydrogenated amorphous silicon (a-Si:H) chip with microfluidics was fabricated.⁴⁶ A schematic plot of this chip with a clear definition of each layer is shown in Fig. S3 in the supplementary material. Indium tin oxide on glass (ITO glass, Ruilong, Taiwan) was used as the bottom substrate and top substrate. An a-Si:H layer fabricated by high-density plasma chemical vapor deposition (HDPCVD) was deposited on the bottom ITO/glass substrate. Using laser-sculptured microfluidics on polyimide tape (10 mm, S-turbo D.I.Y. & Hardware Co., Ltd., Taiwan), the top and bottom substrates were attached together. Via holes of microfluidics were placed on the top ITO/glass substrate and could be used to inject a specific solution for the experiment and drain out the solution after the experiments.

To check the basic sensing performance in the LAPS measurement of this developed system, standard pH buffer solutions of pH 2, 4, 6, 7, 8, 10, and 12 were purchased from Merck Inc. to check the pH-dependent photocurrent vs bias characteristics. To have a benchmark to our previous study with high stability and reproducibility,²⁴ the streptavidin-coated magnetic microbeads with diameters of 1.0 μm (65001; Invitrogen, USA) were prepared in diluted Tris/borate/EDT (TBE) buffer solutions (J885-5 L; PanReac AppliChem, USA) with a concentration of 2% as the standard solution in LACE manipulation. The conductivity of the 2% TBE solution was measured as 40 μS cm⁻¹. The concentration of the magnetic beads was controlled at 10 μg ml⁻¹.

III. RESULTS AND DISCUSSION

First, the control mode of illumination patterns is the most important step for LAPS and LACE operations. As mentioned in Sec. II, two modes of the DLP module can be selected to control the illumination patterns. In Fig. 2(a), a picture of a circle with a diameter of 2 mm as the illumination pattern is shown according to the video and pattern modes of the different inputs of the DLP module. The circle patterns in both modes show slight distortion, which results from the projection angle of 12° by the DLP module. However, a clear rectangular pattern appears in the video mode, which is the result of the reflection of the DMD induced by the dynamic switching of the digital micromirror with a specific refresh rate of 120 Hz. Based on the operation mechanism, the illumination intensity and its gradient are key factors in LAPS and LACE operations. The basic photocurrent vs bias voltage curves of the Si₃N₄ LAPS chip measured with a modulated frequency of 800 Hz for various diameters of pattern and illumination modes are shown in Fig. 2(b). The standard photocurrent vs bias voltage clearly appears with the response of the accumulation, depletion, and inversion regions to the increase in the bias voltage from 0.5 to 2.5 V. High and low photocurrents appear in the inversion and accumulation regions (e.g., at 2.5 and 0.5 V), which matches the behavior of p-type Si. A clear trend of the photocurrent decreases as the diameter of the illumination spot size [e.g., D (μm) in the figure] decreases. To better compare various diameters and illumination modes, the photocurrents measured at 0.5 and 2.5 V for the accumulation and inversion regions are collected for redrawing, as shown in Figs. 2(c) and 2(d). In Fig. 2(c), the photocurrent measured at 0.5 V can be referred to as the background photocurrent or noise level. The photocurrent measured by the video mode has a higher level than that measured by the pattern mode for all groups for the diameter of the spot size, which can be concluded from the undesired rectangular pattern of illumination. With a small spot size diameter, the background response can be slightly decreased because the photocurrent can still be contributed by an undesired rectangular pattern. However, this behavior of background noise can be further reduced with an approximately 53% reduction in the group with a spot size diameter of 500 μm. The lower the background photocurrent is, the lower the noise impact will be. In Fig. 2(d), the photocurrent measured at 2.5 V can be referred to as the responsive photocurrent or signal level. With a large diameter, the photocurrent reaches a high level in video and pattern modes. In the pattern mode, the photocurrent dramatically decreases with the diameter of the spot size, which is a linear response of the illumination area and the corresponding intensity. However, the photocurrent can be maintained at a certain level when the diameter of the spot size is decreased from 2000 to 500 μm in the video mode, which can be attributed to the undesired rectangular pattern of illumination. The highly responsive photocurrent has a high signal level, which could result in a better signal-to-noise ratio. For a clear comparison, the SNR for the LAPS operation of this developed system is defined as the photocurrent measured at 2.5 V divided by the value measured at 0.5 V in this study. The SNR of different diameters of spot size and operation mode is shown in Fig. 2(e). All SNRs in the pattern mode with diameters larger than 500 μm are higher than those in the video mode. This dimension of the illumination pattern (e.g., 1000 μm) can be a future limitation of this developed system for the spatial resolution of chemical images in LAPS operation. The highest SNR of 3.16 is found in the group with a spot size diameter of 2000 μm in the pattern mode, which will be used as the standard parameters for the LAPS measurement in pH sensing characteristics. To have a high spatial resolution, the optimization of the illumination path for the background noise minimization can be suggested in this DMD-based system.

FIG. 2.

(a) Picture of the projected light pattern of the video mode and the pattern mode. (b) Photocurrent vs bias voltage curve, (c) the photocurrent at 0.5 V, (d) photocurrent at 2.5 V, and (e) the SNR of Si₃N₄ LAPS measured in a pH 7 buffer solution and a modulated frequency of 800 Hz for various diameters of projected light patterns in two modes.

The procedure for checking the basic pH sensing performance in LAPS operation measured by this developed system is to check the pH-dependent photocurrent vs the bias voltage, time-dependent photocurrent vs the bias voltage, and nonideal memory effect for characterization of pH sensitivity, drift, and hysteresis. As shown in Fig. 3(a), photocurrent vs bias voltage curves were measured in various standard pH buffer solutions from the accumulation region to the inversion region. As the pH increases, the curve shifts to a positive bias voltage due to the OH- ions bonded on the Si₃N₄ surface with negative charge-induced potential changes. The output voltage (Vout) of each standard pH buffer solution was calculated as the voltage at a photocurrent of 0.9 μA. The pH-dependent Vout is shown in the inset of Fig. 3(a). The pH sensitivity and linearity are 52.8 mV/pH and 99.7%, respectively, by linear fitting between Vout and its pH value. To confirm the stability of this system, the photocurrent vs bias voltage curve was measured every 20 min for 600 min, and the Vout of each curve was calculated for drawing in Fig. 3(b). The time-dependent Vout shows a stable trend. The drift coefficient is 0.68 mV/h by linear fitting from 300 to 600 min. To check the solution memory effect, hysteresis measurements with a solution loop of pH 7–4–7–10–4 were performed with a photocurrent vs bias voltage measurement every minute five times in each pH step. As shown in Fig. 3(c), the response within 5 min of these 3 pH steps is stable. The hysteresis width is 10.9 mV from the difference between the last point of the first and the third steps for the pH 7 buffer solution, which can be used to obtain the memory effect of one acidic (e.g., pH 4) and one basic (e.g., pH 10) solution. These pH sensing performances including the pH sensitivity, linearity, drift coefficient, and hysteresis width are similar to that in previous studies,^45,47,48 which can serve as proof to validate the LAPS operation in this developed system. For the possibility of other sensing targets, ion concentration,⁴⁹ cell activity,⁵⁰ bacteria,⁵¹ and DNA⁵² had been proven in LAPS platform.

FIG. 3.

Si₃N₄ LAPS illuminated with a diameter of 2000 μm and a modulated frequency of 800 Hz in the pattern mode: (a) photocurrent vs bias voltage curve measured in a standard pH buffer solution from pH 2 to 12. The pH sensing performance measured by this setup is calculated as shown in the inset. (b) Drift behavior of the output voltage vs time and (c) hysteresis behavior of the output voltage vs pH every minute measured in a cycle in a buffer solution at pH 7–4–7–10–7 for 5 min in each pH step.

After confirming the LAPS operation, the second part of the operation, LACE, had to be verified systematically. For LACE operation, the a-Si:H chip is designed and fabricated because of the thin-film process on glass, which can be easily integrated with microfluidic and transparent illumination patterns that can be observed from the top side of this chip. Moreover, the a-Si:H layer is a well-proven material applied in commercial products such as a thin-film transistor (TFT) of liquid crystal displays in the panel industry⁵³ and solar cells.⁵⁴ First, photocurrent vs bias voltage curves with spot sizes of various diameters, which is the same setup as the first part of the LAPS measurement, were obtained to confirm the basic photoelectron behavior of this fabricated a-Si:H chip. As shown in Fig. 4(a), the behavior of this fabricated a-Si:H chip, including the accumulation, depletion, and inversion regions, is similar to that of the Si₃N₄ LAPS. However, the photocurrent with the same diameter of illumination patterns from the a-Si:H chip is higher than that from the Si₃N₄ LAPS. For example, the photocurrent at 2.5 V of the a-Si:H chip with a diameter of 2000 μm is 2 μA and is higher than the value for Si₃N₄ LAPS (e.g., 1.34 μA). This result can be concluded from the structure of the thin film of a-Si:H on ITO/glass, which generates more electron–hole pairs and has a shorter transportation path for a high photocurrent.³⁰ With this advantage, a better photocurrent vs bias voltage can be obtained for a small illumination pattern with a diameter of 100 μm compared to that of the Si₃N₄ LAPS. The pH sensing performance of this fabricated a-Si:H chip is also investigated, as shown in Fig. 4(b). The output voltage of three pH buffer solutions of pH values 4, 7, and 10 is calculated as the voltage at a photocurrent of 0.9 μA. The pH-dependent Vout is shown in the inset of Fig. 4(b). The pH sensitivity and linearity are 42.5 mV/pH and 98.1%, respectively, based on a linear fitting between Vout and its pH value. This sensing performance can be from the native oxide on the a-Si:H layer, which is in the acceptable range of SiO₂ in the literature.⁵⁵

FIG. 4.

(a) a-Si:H chip illuminated with a diameter of 2000 μm and a modulated frequency of 800 Hz in the pattern mode for various diameters of projected light patterns and (b) photocurrent vs bias voltage curve measured in standard pH buffer solutions of pH 4, 7, and 10. The pH sensing performance measured by this setup is calculated as shown in the inset.

For the operation of LACE, the most important parameters of manipulation force are the illumination intensity, modulated AC frequency, and applied AC voltage. In general, the modulated AC frequency should be lower than 50 kHz for LACE operation.⁵⁶ The higher the applied AC voltage is, the higher the manipulation force will be but the greater the possibility of electrolysis and subsequent bubbles in the microfluidics will be. The illumination intensity was determined via the HID xenon lamp and illumination path including the lens setup and DMD in this developed system. Therefore, the modulated AC frequency is tested from 50 to 5 kHz with a Vpp of 10 V for AC voltage. With illumination patterns with an area of 100 × 500 μm², the magnetic beads dispensed in the microfluidics channel can be collected at various levels within 5 s, as shown in Fig. 5(a). The lower the modulated AC frequency is, the more magnetic beads can be collected, which matches the operation mechanism of LACE. To further evaluate the difference, the pictures in the illumination pattern after collection were used to perform pattern recognition by ImageJ software, which is a well-proven skill.⁵⁷ As shown in Fig. 5(b), the area marked with a red color is defined as the area with magnetic-bead collection by LACE. No clear collection is apparent at 50 kHz. With an AC frequency of 30 kHz, the collection ratio can be evaluated via the calculated area, which is approximately 5.23% but distributed over a wide range. With the AC frequency reduced to 10 kHz, the area is maintained at 5.23%, but most magnetic beads are collected in a concentrated area. Once the frequency is further reduced to 5 kHz, the area is further concentrated within a larger area of 6.05%, which can be attributed to the large manipulation force at 5 kHz. To further evaluate the manipulation force, the illumination pattern was preformed to travel for a certain distance in a defined time period to check the movement of the magnetic beads. If the magnetic beads can follow the movement of the illumination pattern in a shorter time (e.g., at a higher moving speed), the manipulation force is higher. As shown in Fig. 6, the pictures were collected every 5 s for 25 s for a travel speed of 20 μm/s (e.g., 400 μm in 20 s). It is clearly shown that magnetic beads still remain inside the illumination pattern but near the edge. As time increases, the illumination pattern moves to the left side, and magnetic beads follow the movement. After 20 s, the illumination pattern stopped, and the magnetic beads still moved to the left side with an inertia force provided by LACE. Therefore, the magnetic beads stay on the left side of the illumination pattern. Although LACE operation is proved to manipulate magnetic beads in the proposed system, the manipulation force remains lower than that of the specific system in the literature.²⁴ Moreover, considering subsequent applications of this study, the relationship of bead size, bead surface modification, and LACE-based manipulation is suggested to be confirmed. In the LACE operation, the current published results revealed that the magnetic microbead with a smaller size was observed to have a faster microbead manipulating velocity,²⁴ indicating that the enhanced dimension of beads might lead to a decreased velocity in the manipulation. However, the detailed performance regarding LACE manipulation impacted by the surface modification of beads has not been fully explored or found in the literature, which can be further investigated in the future. To perform a detailed comparison of the relative systems for LAPS and LACE, several representative ligatures were selected, as listed in Table I. It is clearly apparent that this developed system is the first to have LAPS and LACE functions in the same setup. Based on current experimental results, in this study, the concept of a 2-in-1 LOC system with both LAPS and LACE functions was successfully demonstrated. The current setup can be improved in terms of illumination intensity and pattern distortion by optimizing the illumination path. It would be more applicable to integrate sensors and actuators into a single chip and a single system with the achievements provided in this work.

FIG. 5.

(a) Picture of the collection of magnetic beads for 5 s at different frequencies provided by the function generator to the a-Si:H chip in the LACE operation. (b) Density of collection of magnetic beads collected at different frequencies calculated by ImageJ software.

FIG. 6.

Time-dependent manipulation of collected magnetic beads with the movement of the illumination pattern with a constant moving speed of 20 μm/s for 25 s at a modulated frequency of 10 kHz provided by the function generator of the a-Si:H chip in the LACE operation.

TABLE I.

The function and performance comparison of LACE and LAPS relative systems.

Light source	Dynamic pattern	OEK-manipulation	Integrated w/ microfluidics	AC/DC dual input	LAPS-pH sensing	System	Reference
Red laser	No	No	No	No	Yes	LAPS	30
DLP projector	Yes	No	No	No	Yes	LAPS	58
LED	No	No	No	No	Yes	LAPS	59
Hg lamp/DMD	Yes	No	No	No	Yes	LAPS	60
LCD	Yes	Yes	Yes	No	No	ODEP	61
SLM projector	Yes	Yes	Yes	No	No	ODEP	62
Hg lamp/DMD	Yes	Yes	Yes	No	No	ODEP	63
DLP projector	Yes	Yes	Yes	No	No	LACE and ODEP	24
HID xenon lamp/DMD	Yes	Yes	Yes	Yes	Yes	LACE and LAPS	This work

IV. CONCLUSION

A novel 2-in-1 LOC was successfully presented with acceptable LAPS and LACE performance. The degradation and cost of chips transferred from two different systems for actuators and sensors are substantially improved by this design. The most challenging parts of this system are the illumination resource and pattern controllability, which have already been solved using the HID xenon lamp and DLP module. With the assistance of a chopper, the AC and DC illuminations are easily controlled for LACE and LAPS, respectively. The rotating speed of the chopper can be used to define the modulated frequency of illumination for the LAPS. The photoresponse and pH sensing performance are verified using a conventional Si₃N₄ LAPS in this proposed system and are comparable to those of published results. The pH sensitivity is 52.8 mV/pH with a linearity of 99.7%. In the LACE validation, the collection and manipulation of magnetic beads with a diameter of 1 μm was demonstrated with a moving speed of 20 μm/s in 2% TBE solution. The current LAPS and LACE performance should be improved by optimizing the illumination path for intensity enhancement and pattern focus. To approach the application of POCT, the bulky dimension of the whole system can be further minimized by means of a mini DLP projector, illumination path optimization, readout circuit realized by a field programmable gate array (FPGA)-based platform or a full-custom designed chip fabricated by the semiconductor process.

SUPPLEMENTARY MATERIAL

See the supplementary material for more information about the light source and the sensing device.

AUTHORS’ CONTRIBUTIONS

H.-Y.P. and C.-M.Y. contributed equally to this work.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author upon reasonable request.