|Home | About | Journals | Submit | Contact Us | Français|
The Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, Georgia 30332 (USA), School of Chemistry and Biochemistry, School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332 (USA), younan.xia/at/bme.gatech.edu
An optofluidic system was constructed from a diode laser as the energy source, an aqueous suspension of plasmonic nanostructures as the photothermal transducer, and a glass capillary for measuring the volumetric expansion of the suspension. The suspension could be driven to move up the capillary by more than 30 mm and be used to control the operation of an electrical switch.
In localized surface plasmon resonance (LSPR), the conduction electrons of a plasmonic nanostructure are driven by the alternating electric field of an incident light to collectively oscillate at a resonant frequency.[1,2] As a result, the incident light will be absorbed by the plasmonic nanostructure. Some of the absorbed energy will be emitted as light through scattering while the rest will be dissipated by Landau damping, resulting in light-to-heat conversion. The photothermal conversion has been used to induce localized heating of the host medium, which is of great interest for applications in a number of fields, including photonics, electronics, catalysis, phase transformation, and biomedicine.
The photothermal heating based on plasmonic nanostructures has also be been explored to induce and/or control fluid motion. In the first demonstration, Lee and co-workers were able to drive liquid to move along microfluidic channels over a distance as far as 500 µm by using plasmonic heating with Au colloids. In this case, the liquid was driven to move through a complex mechanism that involved a combination of localized evaporation, condensation, and surface wetting. In another demonstration, Wu and co-workers were able to control liquid flow through microfluidic channels by combining the temperature-dependent sol-gel transformation with a composite of Au nanorod and poly(dimethyl siloxane) to generate localized obstructions within specific paths. Most recently, Halas and co-workers exploited the localized heating of SiO2/Au nanoshells within a solution to vaporize water and ethanol with sunlight. The nanoshells were thought to heat up rapidly upon irradiation and their surfaces were surrounded by thin layers of high-temperature vapor that carried the nanoshells to the top of the suspension before the vapor was released into the atmosphere above.
Despite these reported demonstrations, we noticed that the absorption cross sections of the plasmonic nanostructures, a fundamental property that determines their photothermal conversion efficiencies, were seldom measured experimentally in these studies. A quantitative understanding of this parameter is essential to the engineering of plasmonic nanostructures for a variety of applications like fluid transport, optical imaging contrast enhancement, and cancer treatment. In a prior study, we demonstrated the use of photoacoustic imaging for experimental measurements of the absorption cross sections of Au nanostructures, but this approach requires expensive equipment and complex procedures.
Herein, we report a plasmon-assisted optofluidic (PAOF) system that is consisted of a diode laser as the energy source, an aqueous suspension of Au nanostructures as the photo-responsive fluidic medium, and a capillary as the means to monitor the change in temperature caused by volumetric expansion of the medium. Three different types of Au nanostructures were compared in this study: nanocages, nanorods, and hexapods; all of them can be tuned to absorb light in the near-infrared region. To achieve controllable fluid motion, we took advantage of the well-known thermal expansion phenomena of a liquid. With this system, we could drive a fluid to rise in a capillary by as much as 30 mm, depending on the type of nanostructures suspended in the medium and/or the concentration. We could also experimentally measure the absorption cross sections and photothermal conversion efficiencies of various types of plasmonic nanostructures. Furthermore, we demonstrated the feasibility of utilizing the fluid motion caused by plasmonic heating to construct a photo-sensitive electrical switch.
Figure 1a shows a schematic of the PAOF system. It contained a diode laser with a central output wavelength at 808 nm, a photo-responsive medium, and a microfluidic device illustrated in Figure 1b. The microfluidic device was assembled from a poly(dimethyl siloxane) (PDMS) block (1) with flat surface, a PDMS block patterned with a reservoir (2) on the surface, a plastic tube with an inner diameter of 0.5 mm (3) that serves as a channel for pumping in the liquid medium, and a glass capillary with an inner diameter of 0.6 mm (4) serving as a “ruler” to measure volumetric expansion of the liquid. All joints were sealed with PDMS. The exact dimensions of these components are summarized in Figure S1. In a typical experiment, 0.7 mL of a medium containing the plasmonic nanostructures was loaded, through the plastic tube, into the reservoir between blocks 1 and 2 until the medium reached a height of 1 cm in the capillary. The opening on tube 3 was then plugged with a short glass rod. Since PDMS is transparent in the visible and near-infrared regions, the diode laser was able to penetrate through the top PDMS block to reach the liquid medium containing the plasmonic nanostructures. As such, the medium was quickly heated to a temperature as high as ca. 84 °C, depending on the type of nanostructures contained in the medium and their concentration. The medium then thermally expanded into capillary 4 due to the increase in temperature and the height of the liquid column was recorded during irradiation with a stationary camera. Figure 1C shows a photograph of the actual microfluidic device.
In the initial study, we compared the performance of three different types of Au plasmonic nanostructures: nanocages, nanorods, and hexapods. All of them can be engineered to generate heat through absorption of near-infrared light. The detailed procedures for the syntheses of these nanostructures can be found in the Supporting Information. Figure 2, a–c, shows transmission electron microscopy (TEM) images of these nanostructures. The nanocages in Figure 2a exhibited a porous morphology (with a hollow interior and pores at the corner sites) with an edge length of 45 nm and a wall thickness of ca. 5 nm. The nanorods in Figure 2b had an average diameter of 17 nm and an aspect ratio of ca. 3.3. The hexapods in Figure 2c showed a highly branched morphology with an octahedral core of ca. 25 nm in edge length and six arms on all the vertices. The distance between the ends of two opposite vertices was ca. 60 nm. In principle, the LSPR peaks of all these nanostructures can be precisely tuned to any wavelength in the near-infrared region by adjusting their wall thickness, aspect ratio, and arm length, respectively. In this study, all of their major LSPR peaks were tuned to the region of 805–810 nm to match the central output wavelength (808 nm) of the diode laser. Figure 2d shows UV-vis-NIR extinction spectra of these three different types of Au nanostructures.
We used an organic dye, indocyanine green (ICG), as a reference of calibration. To establish a relationship between the temperature and the height of liquid in the capillary, we calibrated the PAOF system by adding 0.7 mL of 6.5 µM ICG solution into the device, irradiating the medium with the diode laser, monitoring the temperature (T) of the medium in real time with an infrared camera, and recording the height (h) of the medium in capillary 4 whenever the temperature reached the specific values of 24, 37, 50, and 70 °C. The power density of the laser was set to 0.4 W·cm−2 and the illumination area was 1.13 cm2 for all the experiments. As shown in Figure S2, a linear relationship was found between h and T. The slope (Δh/ΔT) was calculated to be 0.52 mm °C−1. Since all the media we used were based on aqueous solutions and they should have more or less the same volumetric expansion coefficient, it is not unreasonable to assume that the calibration value for Δh/ΔT can also be applied to media containing different types of Au nanostructures. The linear relationship between h and T suggests that one can determine the temperature of a medium by simply measuring the height (h) of liquid in the capillary, just like a thermometer.
We then evaluated the plasmon-assisted heating process over a period of 40 min: the diode laser was turned on for 10 min to heat the medium, and then turned off for 30 min to let the medium cool down. This process was repeated three times and the height was recorded at an interval of one minute. For nanocages, we made measurements for four different concentrations at 1.0×109, 2.5×109, 5.0×109, and 1.0×1010 particles/mL (Figure 3a) while we only examined one concentration of 1.0×1010 particles/mL for both nanorods and hexapods (Figure S3). The curves clearly show the periodicity of change in height (Δh). As expected, the Δh value increased with the concentration of the nanostructures due to the increased power of absorption for the solution. For nanocages of 1.0×1010 particles/mL in concentration, we were able to drive the fluid up to a height of ca. 30.5 mm in the capillary. We then calculated the change in temperature (ΔT) based on the calibration value for Δh/ΔT and the results are summarized in Table 1. For nanocages with different concentrations, the ΔT varied from ca. 11.2 to ca. 58.7 °C, which corresponded to a maximum temperature of 82.7 °C when we started from room temperature (24 °C). For nanorods and hexapods, the ΔT values were less than half of that for nanocages at the same particle concentration.
To quantitatively evaluate the capability of converting light into heat for different types of plasmonic nanostructures, we calculated the energy conversion efficiency (η) using the following equation:
where Q is the total heat generated in the solution which depends on ΔT for a given particle concentration and E is the total energy output of the laser over the 10-min period (both of them are defined and derived in the Supporting Information). Table 1 summarizes Δh, ΔT, and η for nanocages, nanorods, and hexapods at various concentrations. As the concentration of nanocages increased from 1.0×109 to 2.5×109, 5.0×109, and 1.0×1010 particles/mL, η increased from 12.1±0.8% to 27.9±1.7%, 46.9±2.9%, and 63.6±4.2%, respectively (see Table 1 and Figure 3b). For nanorods and hexapods at a concentration of 1.0×1010 particles/mL, their η values were 22.1±1.7% and 29.6±1.9%, respectively, which were less than half of that for nanocages at the same concentration. The disparity in η for the three types of nanostructures is believed to be caused by their differences in absorption coefficient (μ) associated with absorption cross section (σ), an intrinsic property of the plasmonic nanostructure that determines its ability to absorb light at a particular wavelength.
One of the major applications of the PAOF system is to experimentally measure the μ and thus calculate the σ of a plasmonic nanostructure. To demonstrate this capability, we used ICG as a reference again because its absorption coefficient (μICG, which equals to 140 m−1) is well documented. After adding 0.7 mL of 6.5 µM ICG solution into the device, we irradiated the system for 10 min and a change in height (ΔhICG) of 33.2 mm was measured. We then derived the μ associated with the σ of a plasmonic nanostructure by firstly calculating the heat (Q) from eq. (S1). The irradiation energy being absorbed (Eabs) can be expressed as eq. (S4). Since the absorbed energy (Eabs) is completely converted to heat (Q) through Landau damping, we obtain eq. (S6). When the same device is used for both ICG and a suspension of Au nanostructures, the following equation can be used to calculate the μ of the nanostructure (μNS):
where M is the molar concentration of the nanostructures and NA is the Avogadro’s constant.
The experimentally obtained μ and σ are listed in Table 1 for each type of Au nanostructure at different concentrations. The magnitude of σ for the nanocages was on the same order as what was derived from theoretical calculations (16.3×10−15 m2), suggesting that the PAOF system can be adopted for quantitative measurements of μ and σ. When comparing the σ values of the three different types of Au nanostructures, we found that the nanocages had the highest magnitude while the nanorods had the lowest and the values of σ for both nanorods and hexapods were less than half of that of nanocages. This finding agrees with the observed trend for η, in which nanocages exhibited the highest efficiency in converting light to heat. Although prior studies demonstrated that the σ of plasmonic nanostructures could be measured by either photoacoustic imaging or optical coherence tomography, the technique reported in this study provides a much simpler and more convenient tool as it does not require any specific equipment or training.
By taking advantage of the light-driven motion of liquid in a PAOF system, we demonstrated a photo-sensitive electrical switch. In this case, we saturated an aqueous suspension of Au nanocages (1.0×1010 particles/mL) with NaCl to make it conductive and then used as the fluidic medium. As shown in Figures 4a and S4, we constructed the switch from a small vial connected to a capillary containing two standard copper wires with a gap of 5 mm between their ends (as marked by * and ** symbols in Figure 4a). When the solution was irradiated by the diode laser, it heated up and expanded along the capillary, bridging the gap between the two wires and completing an electrical circuit. The connection turned on a small light-emitting diode (LED). After turning off the laser irradiation, the solution level dropped and the LED was switched off. As shown in Figure 4b, the electrical switch could be operated through four cycles of on and off irradiations without losing its performance, suggesting that it could be used to control electrical devices without requiring mechanical movement or large solid-state transistors. To our knowledge, this was the first example demonstrating the potential use of fluid motion controlled by plasmonic heating to construct a functional electrical device.
In summary, we have demonstrated a plasmon-assisted optofluidic system by combining the unique features of microfluidics and plasmonic heating of noble-metal nanostructures. By taking advantage of the well-known thermal expansion of a liquid, we could move a fluid column via plasmonic heating. Using this simple system, we were able to quantitatively measure the photothermal conversion efficiencies of various plasmonic nanostructures and their absorption cross sections. We also demonstrated the feasibility of constructing a photo-sensitive, electrical switch by controlling the liquid level with plasmonic heating. Our plasmon-assisted optofluidic system provides a simple and effective platform for remotely controlling the motion of a fluid and is expected to find use in various areas involving molecular transport, biochemical analysis, catalysis, photonics, and electronics.
This work was supported in part by a research grant from NCI (1R01 CA138527), an NIH Director’s Pioneer Award (DP1, OD00798), and startup funds from Washington University in St. Louis and Georgia Institute of Technology. Y.X. was also partially supported by the World Class University (WCU) program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (R32-20031).
Supporting information for this article is available on the WWW under http://dx.doi.org/.