Hydrogel Micropillar Array for Temperature Sensing in Fluid

Abstract

Localized temperature sensing and control on a micron-scale have diverse applications in biological systems. We present a micron-sized hydrogel pillar array as potential temperature probes and actuators by exploiting sensitive temperature dependence of their volume change. Soft lithography-based molding processes were presented to fabricate poly N-isopropyl acrylamide (p-NIPAAm)-based hydrogel pillar array on a glass substrate. Au nanorods as light-induced heating elements were embedded within the hydrogel pillars, and near-infrared (NIR) light was used to modulate temperature in a local area. First, static responses of the micro-pillar array were characterized as a function of its temperature. It was shown that the hydrogel had a sensitive volume transition near its low critical solution temperature (LCST). Furthermore, we showed that LCST could be readily adjusted by utilizing copolymerizing with acrylamide (AAM). To demonstrate the feasibility of spatiotemporal temperature mapping and modulation using the presented pillar array, pulsed NIR light was illuminated on a local area of the hydrogel pillar array, and its responses were recorded. Dynamic temperature change in water was mapped based on the abrupt volume change characteristics of the hydrogel pillar, and its potential actuation using NIR light was successfully demonstrated. Considering that the structure can be arrayed in a two-dimensional pixel format with high spatial resolution and high sensitive temperature characteristics, the presented method and the device structure can have diverse applications to change and sense local temperatures in liquid. This is particularly useful in biological systems, where their physiological temperature can be modulated and mapped with high spatial resolution.

Graphical Abstract

I. Introduction

Localized temperature sensing and actuation have diverse applications in understanding fundamental biological processes and developing biomedical devices. Temperature indicates the physiological conditions of cells, and their biochemical activities are strongly correlated with temperature fluctuations [1, 2]. Temperature variation affects structural forming parameters in cells, leading to changes in their mechanical properties [3, 4]. Abnormal metabolisms in cells induce temperature variation [5], and monitoring temperature has been used as a biomarker to identify functional disorders in cells and tissues [6]. Abrupt temperature modulation can alter neuronal functions by inducing excitation and inhibition of neural cell activities [7]. Manipulation of temperature-sensitive ion channels within cells triggers a new possibility of controlling the signaling pathways of neuronal cells [8, 9].

There are growing interests in engineered interfaces to sense and control the temperature in a localized area on a micron-scale within physiologically relevant environments. Nevertheless, Monitoring temperature in a micron-scale in a liquid environment has been particularly difficult. Different types of temperature sensors [10–12] based on the temperature dependence of their electrical properties have been employed to measure local temperature variations. While these sensors can provide long-term temperature measurements with high sensitivities, they are typically useful to measure at a local point and have limitations to capture the spatial distribution of temperature with high resolution. Furthermore, they need to be close to a target location without disturbing measurement environments. Non-contact optical detection is useful for monitoring dynamic temperature profiles. Based on the optical approach, they provide temperature profiles with high spatial resolution. Infrared camera technologies [13, 14] based on measuring infrared flux from a target surface have been greatly advanced and implemented to measure the temperature profile in fluidic channels. However, their applications are not suitable for micron-scale temperature monitoring in liquid due to the high absorption loss of infrared signals in a liquid environment. Many current methods to monitor dynamic temperature profiles rely on temperature-dependent spectral responses from dyes or nanoparticles [15–25]. In their typical implementations, measurements are done in a microscope setting with a high-resolution objective lens, and a separate calibration curve was measured to obtain the temperature dependence of their fluorescent emission intensity [17, 20, 24], lifetime [16, 21, 22], or spectral shift [19]. Based on the calibration curve, temperature profiles were mapped on a target object. However, there are still challenges remaining in obtaining reliable results. For example, fluorescent signals can be affected by laser intensity fluctuation and strongly depend on environmental changes due to particle concentration and chemical surrounding variation, local scattering, and absorption. Furthermore, photobleaching can induce fluorescent signal changes with time.

In this paper, we present an approach using temperature sensitive hydrogel pillar array to map two-dimensional temperature profiles in a water environment. Poly N-isopropyl acrylamide (p-NIPAAm) hydrogel, which has been extensively used as a biocompatible material [26, 27], was devised to be formed as an array on a glass substrate. While p-NIPAAm hydrogel has been used in bulk temperature sensors [28–30] based on a unique temperature response near its low critical solution temperature (LCST), to our knowledge, there is no demonstration using p-NIPAAM hydrogel in a micro-pillar array as temperature probes to map temperature profile in a liquid environment. We also show that LCST can be readily adjusted by copolymerizing with acrylamide (AAM). To demonstrate temperature mapping using p-NIPAAm hydrogel pillar array, Au nanorods are incorporated into the hydrogel during the fabrication. Localized heating on the hydrogel pillar array was generated by NIR laser beam illumination, and its temperature profile in liquid was mapped based on the calibrated mechanical response of the hydrogel array. In addition to the temperature sensing, this will open a photothermal actuation capability using the micro-pillar array. While the demonstration of the proposed method was achieved with a micron-sized pillar array, their dimension, in conjunction with a microscope resolution, can be further reduced by standard microfabrication technology. The hydrogel pillar array will have the potential to measure and control localized temperature variation in a liquid environment with high spatiotemporal and temperature resolutions.

II. Design and fabrication

Micro-molding processes have been widely used and reviewed to create three-dimensional hydrogel arrayed patterns [31, 32]. Particularly, soft lithography-based molding processes were presented in this paper to fabricate poly N-isopropyl acrylamide (p-NIPAAm)-based hydrogel pillar array on a glass substrate. The fabrication process steps for the hydrogel micro-pillar array are shown in Figure 1.

Fig. 1.

The fabrication process for p-NIPAAm hydrogel micro-pillar array. (a) A silicon substrate is prepared. (b) Photolithography to prepare a master mold of a patterned micro-pillar array using photoresist and silane-treatment for easy detachment. (c) PDMS is poured on the master mold and capped with a glass slide. Then, PDMS is cured. (d) PDMS mold is prepared by detaching it from the master silicon mold. (e) NIPAAm hydrogel solution mixed with Au-NDs is poured, covered with a glass slide, and cured on the PDMS mold. (f) p-NIPAAm hydrogel micro-pillar array with embedded Au-NDs is prepared by detaching it from the PDMS mold

First, a silicon substrate was spray-cleaned with solvents (acetone, methanol, and isopropanol), and then dried with nitrogen, followed by baking at 200 °C for 10 min. Additional cleaning using oxygen plasma (20 sccm O2 flow, 100 W RF power, 180 mTorr pressure, and 5 min) was performed on the substrate (Figure 1(a)). Standard photolithography using AZ 4620 resist was performed on the silicon substrate. To promote the adhesion between the substrate and AZ 4620 resist, AZ BARLi-II 90 was first spin-coated on a silicon substrate and baked at 200 °C for 10 min. AZ 4620 resist was spin-coated at 2500 rpm and baked at 110 °C for 4 min. Resist edge beads were removed using a Q-tip soaked in acetone. After UV pattern exposure and develop, the measured thickness of the resist pattern was around 14 μm. This completes silicon mold fabrication for the next polydimethyl siloxane (PDMS) process (Figure 1(b)). To begin a PDMS mold fabrication, the silicon mold was briefly treated using oxygen plasma (20 sccm O2 flow, 100 W RF power, 180 mTorr pressure, and 30 sec) and treated using silane (trichloro (1H,1H,2H,2H-perfluorooctyl) silane) under vacuum in a desiccator for overnight. 10:1 weight ratio of PDMS base and curing agent (Dow Corning Sylgard 184) was mixed and degassed without air bubbles. A pre-cleaned microscope glass slide was also prepared and surface-treated using oxygen plasma (20 sccm O2 flow, 100 W RF power, 180 mTorr pressure, and 5 min) to promote the adhesion between the glass substrate and PDMS layer. The degassed PDMS mixture was carefully poured on the silicon mold without creating air bubbles, and the glass slide was placed over the PDMS mixture (Figure 1(c)). After the PDMS was cured at 60 °C for 4 hours, a PDMS mold was prepared by separating it from the silicon mold (Figure 1(d)). Hydrogel micro-pillar array was fabricated by replicating the structure from the PDMS mold. Ultraviolet (UV)-curable Au nanorod/p-NIPAAM hydrogel solution was prepared. The detailed preparation of Au nanorod/p-NIPAAm hydrogel can be found in our previous publication [33]. In short, Au nanorods were synthesized by seed-mediated synthesis method. Their optical absorption peak was tuned to 815 nm of wavelength by adjusting silver nitrate amount during their synthesis processes. A pellet of Au nanorod precipitate was obtained by centrifugation from 20 mL of freshly synthesized Au nanorod suspension. Then, it was functionalized with PVP surfactant to avoid their aggregation in the p-NIPAAm hydrogel solution. The functionalized Au nanorods were rinsed with deionized water and collected by centrifugation. After the water was evaporated, PVP functionalized Au nanorods were suspended in photocurable NiPAAm hydrogel solution, which contains 0.2 g of NIPAAm monomer, 0.01 g of MBAAm crosslinker, and 20 μL Darocur 1173 UV initiator in 1mL DMSO solution. Separately, a glass substrate was cleaned with solvents, followed by oxygen plasma (20 sccm O2 flow, 100 W RF power, 180 mTorr pressure, and 5 min). To enhance the adhesion between hydrogel and the glass substrate, the glass substrate was then treated with a few drops of aminopropyl triethoxy silane (APTES) in a vacuum desiccator overnight. To begin hydrogel micro-pillar array fabrication, the PDMS mold was placed in a vacuum desiccator, and the hydrogel solution was poured on top of the PDMS mold. Brief vacuum was applied to ensure that the hydrogel solution was effectively introduced into voids in the PDMS mold. After that, the silane-treated glass substrate was placed on top of the hydrogel solution and then exposed to UV light (360 nm, 9 W UV bulbs) for 1 min to cure the hydrogel (Figure 1(e)). To release the cured hydrogel from the PDMS mold, it was soaked in water overnight at room temperature and separated from the PDMS mold (Figure 1(f)). Figure 2 shows an example of a freestanding hydrogel micro-pillar array layer in water without a glass substrate. With APTES treatment on the glass substrate, the adhesion between the glass substrate and hydrogel pillar layer was strong enough for repeated temperature sensing and actuation during the following measurements.

Fig. 2.

A microscope image of a hydrogel micro-pillar array layer in water that was detached from a glass substrate.

III. Result and Discussion

Figure 3 shows a characterization setup to demonstrate the temperature sensing and actuation of a hydrogel micro-pillar array. Nikon E200 microscope was modified to include NIR laser illumination using a dichroic mirror in the optical path. The characterization of the hydrogel micro-pillar array was performed in the transmission configuration with white light illumination from the bottom of the microscope. A custom-built temperature-controllable chamber shown in the insert image was used to mount a hydrogel micro-pillar array on a glass substrate. A temperature controller (Thorlabs TC200) interfaced with a heating element and a temperature sensor mounted in the chamber. After a hydrogel sample was positioned in the chamber, water was filled, and a thin glass slide was placed on the top of the chamber to prevent water evaporation and clear imaging through an objective lens. The water temperature in the chamber can be set and measured during the experiments. The resulting images from the hydrogel micro-pillar array were captured and recorded in a video format from a digital microscope camera.

Fig. 3.

Schematic diagram of a custom-built characterization setup of hydrogel micro-pillar array.

First, the static temperature response of a hydrogel micro-pillar array was characterized. For this measurement, hydrogel micro-pillar images were taken at each water temperature in the chamber controlled by the temperature controller. To have constant water temperature, each temperature was stabilized for at least 10 minutes before taking images. Once the temperature was stabilized, the size of the pillar was not changed at the set temperature.

Figure 4 shows the top view of the hydrogel pillars at various temperatures. Figure 5 shows normalized pillar sizes as a function of temperature for two different hydrogel compositions from the captured images. 100% NIPAAm represents a hydrogel composition without copolymerizing with acrylamide (AAM). 95% NIPAAm represents the hydrogel composition with 5% AAM and 95% NIPAAM monomers. Other ratios with crosslinker and photoinitiator remain the same. It was observed that the hydrogel micro-pillars with AAM were more structurally rigid than the hydrogel micro-pillars with only NiPAAm. Furthermore, the size changes, as shown in Figure 4, reveal that the volume transition temperature was shifted to a higher temperature by copolymerizing NiPAAm with AAM. The LCST shift of pNIPAAM with the addition of AAM agrees with other previous works [34–36]. It was possible to design the temperature response range of the p-NiPAAm-based hydrogel micro-pillar array by adjusting the ratio of two monomers. Particularly, for the temperature sensing and actuation demonstrated in this paper, this provides an opportunity of tuning the temperature range of the hydrogel micro-pillar array for a specific application. Micro-pillar arrays with 100% NIPAAm and 95% NIPAAm have relatively linear responses within temperature ranges of 25 °C to 33 °C and 33 °C to 42 °C, respectively. The temperature sensitivity can be defined as the change in the size (%) per 1 °C, which results in over 11% per °C. Compared to the typical sensitivities of less than 2% per °C demonstrated in other fluorescent-based methods [23–25], our approach potentially provides more sensitive measurements.

Fig. 4.

Static temperature response of hydrogel micro-pillar arrays with different composition ratios of NiPAAm:AAm.

Fig. 5.

Normalized size characterization of hydrogel micro-pillar arrays with different composition ratios of NiPAAm:AAm.

Next, the dynamic temperature sensing and actuation were demonstrated using the hydrogel micro-pillar array. For this demonstration, pulsed NIR light at 815 nm was illuminated locally on a particular area of the hydrogel micro-pillar array to induce thermo-plasmonic heating through embedded Au nanorods. The peak optical power from the objective lens was measured to 82 μW by a calibrated photodetector (Newport 818-SL). Localized temperature modulation was used to control the hydrogel pillar actuation. Similar concept was also used to sense the localized temperature distribution in water by observing the size change of hydrogel pillars. Figure 6 (a) shows a microscope image of a hydrogel micro-pillar array in the water chamber at 27 °C. When NIR light was illuminated on the local area marked in red circle, it was observed that the pillar size was decreased, as shown in Figure 6 (b). The dynamic change of the pillar size by using ON/OFF laser illumination can be viewed in the supporting video (Supplementary material). The demonstration clearly shows the local actuation of hydrogel pillars using NIR light. After the repeated photothermal actuation, the hydrogel pillars do not show any degradation and return to their initial shape, which confirms their stability as the proposed sensor.

Fig. 6.

Microscopic images of hydrogel micro-pillar arrays without (a) and with (b) NIR illumination (marked in red circle).

To demonstrate the local temperature sensing in water using the hydrogel micro-pillar array, the measured size changes of the hydrogel pillars were converted into their corresponding temperatures. Since the water temperature was stabilized at 27 °C for this experiment, the initial size at 27 °C was used to normalize the size change, and its corresponding temperature was mapped based on the interpolation of the calibration graph shown in Figure 5. p-NIPAAm-based hydrogel pillar has a unique temperature-volume response curve near its LCST. By utilizing the characteristics, the localized temperature can be indirectly measured by its size change. Figure 7(a) and (b) show temperature distributions extracted from the size change of the hydrogel pillar array from Figure 6(a) and (b), respectively. The temperature between 26 °C and 34 °C can be successfully identified with a spatial resolution of 16 μm pitch distance between pillars. Based on the microfabrication processes, the spatial resolution can be further improved by utilizing a smaller pitch distance.

Fig. 7.

Extracted temperature responses of hydrogel micro-pillar arrays (a) and with (b) NIR illumination.

Figure 8 shows the transient response of a hydrogel pillar (100% NIPAAm) when NIR light was modulated with ON (2 sec) and OFF (3 sec) cycles extracted from one of hydrogel pillars shown in Figure 7. When the NIR light was ON, the temperature was rapidly increased and saturated in 1.2 sec of rise time (defined as 10% to 90% value change) within around 3 oC. When the NIR light was OFF, the temperature was abruptly decreased. Since Au-Nds embedded in hydrogel pillars worked as plasmonic heating elements, the temperature on the illuminated area was increased when NIR light was ON. When NIR light was OFF, the illuminated area temperature was decreased by stabilizing from the liquid temperature of the outside area. As the NIR ON/OFF cycles continued with our experimental condition, the background temperature slightly increased due to the accumulated heat by the repeated thermal generation through NIR light illumination, as shown in Figure 8. Even with the repeated NIR light cycles, the pillar size came back to its original shape without any noticeable damages. It is expected that Au-Nd concentration within hydrogel affects the maximum temperature that can be obtained with a given laser power from our previous paper. Benefiting from the significant volume change within a narrow LCST region, this approach can allow highly sensitive temperature mapping in water. For example, 80% of hydrogel size change is measured within 7 °C of temperature change regarding 100% NIPAAm hydrogel micro-pillar array. Furthermore, the sensing temperature range can be adjusted by co-polymerization of hydrogel monomers, as shown in Figure 5. Particularly, the temperature window determined by the NIPAAm-AAm-based hydrogel is interesting in mapping temperature distribution of in-vitro biological systems in water, of which the nominal physiological temperature range is within a few degree variations from 37 °C. Temperature measurement can be one of the vital parameters that can be correlated with physiological events in biological systems. Based on the microfabrication techniques, the demonstrated structure can be incorporated into micro total analysis systems for biological applications. This will provide a new tool to sense the temperature of biological systems where the liquid water environment poses a technological barrier for conventional temperature mapping using an infrared camera.

Fig. 8.

Transient temperature response of a hydrogel micro-pillar when NIR light is illuminated on the pillar with ON and OFF cycles.

IV. Conclusion

We have demonstrated the temperature sensing of a hydrogel micro-pillar array using Au nanorod embedded NIPAAm-based hydrogel composites. Temperature-sensitive NIPAAm-based hydrogel was structured to micro-pillar arrays based on replica molding processes. The hydrogel pillar array showed significant volume change within a narrow temperature window, which can also be adjusted by utilizing co-polymerization of NIPAAm and AAm. Adopting the unique volume change characteristics, it was demonstrated that the hydrogel pillar array could be used as temperature sensing elements in water. Since the abrupt volume change of the hydrogel pillar array occurs within a narrow temperature region, this allowed sensitive temperature mapping in water by monitoring their volume changes. As optically induced thermal heating sources, Au nanorods were embedded in the hydrogel pillar array, and thermal actuation of the hydrogel-pillar array in a local area was demonstrated using remote NIR illumination. Their temperature mapping in water was also proposed based on the size change of the pillar. Considering that the narrow temperature window can be tuned to physiologically relevant temperature ranges in biological systems, the proposed temperature mapping in water has the potential to be implemented into novel micro-total analysis systems that require monitoring the onset of threshold temperature change and gradient temperature mapping of target cells.