Optical Trapping with Focused Surface Waves

Abstract

Near-field optical trapping can be realized with focused evanescent waves that are excited at the water-glass interface due to the total internal reflection, or with focused plasmonic waves excited on the water-gold interface. Herein, the performance of these two kinds of near-field optical trapping techniques is compared using the same optical microscope configuration. Experimental results show that only a single-micron polystyrene bead can be trapped by the focused evanescent waves, whereas many beads are simultaneously attracted to the center of the excited region by focused plasmonic waves. This difference in trapping behavior is analyzed from the electric field intensity distributions of these two kinds of focused surface waves and the difference in trapping behavior is attributed to photothermal effects due to the light absorption by the gold film.

1. Introduction

Optical tweezers can be realized using a high numerical aperture (NA) objective that focuses a laser beam to a diffraction-limited spot.^[1–4] In the past decades, optical tweezers have attracted great interest, because of their potential in biological applications, particularly in the research of single molecule dynamics and manipulation of DNA. However, the focal spot of a high NA objective cannot be smaller than the diffraction limit and is elongated along the longitudinal axis. The large focal spot and the elongated trapping volume lead to a significant background emission, which makes it difficult to observe the dynamics of single-labeled biomolecule. To solve this problem, a near-field optical trapping technique based on focused evanescent waves illumination has been proposed and experimentally demonstrated by Min et al.^[5–9] The focused evanescent field is produced using an annular incident light and a high NA objective. The evanescent waves induced by the total internal reflection (TIR) at a water-glass interface constructively interfere at the center of the excited region, which can form a sharp optical virtual probe at the center and reduce the lateral trapping size. The same optical configuration can be used to induce surface plasmons (SPs) on a gold film. The focused plasmonic fields can also work as optical tweezers, which have been used for trapping metallic micron-sized particles^10–11 However, to the best of our knowledge, there are no reports which compare the focused evanescent waves and focused plasmonic waves for optical trapping of single or multiple micron-sized dielectric beads when there are many beads in the field-of-view (FoV) of a high NA objective. This comparison will clarify the roles of optical forces and photothermal effects in trapping, which will guide the future applications of these near-field optical trapping techniques.

2. Results and Discussion

2.1. Trapping PS Beads with Focused Plasmonic Waves and Focused Evanescent Waves

Figure 1 shows the experimental configuration. A 671 nm laser beam is expanded by two lenses and then passed through a pair of axicons, resulting in a nearly collimated annular beam.^[12] This annular beam is focused by a high NA objective (100 X, NA = 1.49) onto the substrate. A light-emitting diode (LED) is placed above the samples to illuminate and observe the beads. Two collection lenses are used to image the front focal plane (FFP) and back focal plane (BFP) of the objectives onto the cameras (detectors 1 and 2).^[13] Filter 1 is used to partially reject the laser beam and allow the light from the LED to reach the detector 1 to monitor the motions of the beads. Filter 2 is used to allow only 671 nm laser beam to reach detector 2 and to verify the excitation of evanescent waves at the water-glass interface and plasmonic waves at the water-gold interface. Two substrates are used to compare the evanescent field and plasmonic field for optical trapping. The evanescent field is formed using a bare glass substrate. The plasmonic field is formed using a glass substrate coated with a thin gold film about 45 nm thick. A drop of the water solution containing polystyrene (PS) beads with diameter at about 1.6 pm is placed on the two substrates. More details on the experimental setup and samples preparation are described in the Experimental Section.

Figure 1.

Schematic diagram of the experimental setup. A laser beam at 671 nm is first expanded by two lenses and then passes through a pair of axicons, resulting in a nearly collimated annular beam. This annular beam is focused by an oil immersed objective (100 ×, NA 1.49) onto the sample to trap the beads on the substrate. Filter 1 is used to partially reject the excitation laser beam and allow the light from the illuminator to reach detector 1 (for front focal plane (FFP) imaging). Filter 2 is used to allow only the laser beam to reach detector 2 (for back focal plane (BFP) imaging).

In our experiments, the incident annular beam was focused by the high NA objective. The incident angle of the annular beam onto the substrates was controlled by the diameter of the annular beam, which could be adjusted by the position of the two axicons. To realize the optical trapping, the incident angle onto the substrate should be larger than the TIR angle, therefore only evanescent waves can be excited on the substrate. Before trapping the beads, the BFP images of the laser beam reflected from the two substrates are recorded by detector 2 as shown in Figure 2a,b, on which a narrow bright ring appears. The evanescent waves at the water-glass interface can be generated in the case of glass working as the substrate (Figure 2a). Based on the known NA of the objective and diameter and thickness of the bright ring, the range of the incident angles can be derived as from 63° to 79°, which are larger than the TIR angle at the water-glass interface (61.39°). Figure 2b shows that a pair of dark arcs appear inside the bright ring, which is the typical feature of the SPs excited by a linearly polarized beam at a water-gold interface. From the diameter of this dark arc and NA of the objective, the surface plasmon resonance (SPR) angle at the water-gold interface can be derived as about 72°,^[12,14] which is consistent with the numerically calculated SPR angle (about 70°) at the water-gold interface.

Figure 2.

Back focal plane (BFP) images of the reflected laser beam and illustrations on the formation of optical virtual probes. The polarization of the incident beam is along x-axis (linear polarization). a) BFP image of the reflected laser beam from the glass substrate. b) BFP image of the reflected laser beam from the gold film. The dark arcs indicate surface plasmons (SPs) excited by a linearly polarized beam. c) Schematic of trapping beads by an optical virtual probe in the evanescent field on the glass substrate. d) Schematic of trapping beads by an optical virtual probe in plasmonic filed on the gold film.

The focusing of the plasmonic waves at the water-gold interface and evanescent waves at the water-glass interface are illustrated in Figure 2c,d, respectively. In this experiment, the water-glass interface and the water-gold interface are slightly below the FFP of the objective (the focus of the laser beam is inside the water solution), and then the excited evanescent waves and plasmonic waves propagate toward the center of the excited region and form optical virtual probes, as illustrated in Figure 2c,d.^[15,16] For optical forces acting on PS beads, the gradient force can be enhanced by a stronger and sharper electric field and always points the peak of the field, whereas the scattering force is determined by the light propagation as well as the power flux.^[17] The gradient force induced by the optical virtual probe points to the focused center and the scattering force derives from the pushing as the waves propagate inward to the central peak. The direction of the push is hence in the same direction as the attractive gradient force and both of them contribute to optical trapping. In these experiments, when the laser beam is focused onto the gold film for excitation of SPs, the power of the incident annular beam is about 10 mW (the detector of the power meter is placed at the rear aperture of the objective). The power of the incident laser beam is increased to 40 mW for the excitation of the evanescent waves on a glass substrate. This trapping process is shown with the successive images in Figure 3. For the PS beads in water on the gold film, the optical virtual probe works as a near-field tweezer. From Figure 3a-e, three beads (labeled with the red circle) are drawn into the center of the excited region and then are trapped. The dynamic trapping process is shown in Movie S1, Supporting Information. These experimental results clearly show that several PS beads can be trapped together in the center of the excited region on a gold film.

Figure 3.

Trapping polystyrene (PS) beads with focused plasmonic waves and focused evanescent waves. a-e) Successive images of three beads (diameter of 1.6 pm) trapped by the focused plasmonic waves on the gold film. The image sequence is from (a) to (e) with a time interval between images of 2.5 s. f-j) Bead 1 is trapped by the focused evanescent waves on the glass substrate. As the stage is moved from right to left, the untrapped bead (bead 2) moves from right to left. Successive images of this process are recorded in (f)-(j). During this movement, bead 1 does not move and bead 2 cannot be trapped even when it passes through the center of the excited region.

For the PS beads on the bare glass substrate, where the evanescent waves (water-glass interface) induced by TIR form the optical virtual probe, only one bead can be trapped, as shown in Movie S2, Supporting Information. Although there are several beads in the field-of-view of the objective on glass substrate, these beads do not flow toward the center, which is different from that shown in Figure 3a-e) and Movie S1, Supporting Information, for the gold film. To further verify this observation, the stage is moved from right to left and untrapped bead moves to the center of the excited region (the position of the optical virtual probe), as shown in Figure 3f-j. During this stage movement, the position of bead 1 (marked with the red circle) is always at the center of the excited region, showing that bead 1 is trapped by the focused evanescent waves. However, bead 2 moves from right to left. Bead 2 is not trapped or attracted even when very close to the center of the excited region. This phenomenon demonstrates that only one PS bead can be trapped in the focused evanescent field even when there are other beads around the trapping center. However, as shown in Figure 3a-e, for the focused plasmonic field on a gold film, several beads can be trapped simultaneously.

2.2. Numerical Simulation of the Optical Virtual Probes

To understand this difference, we performed numerical simulations on the electric field distributions of the two near-field optical virtual probes with the finite-difference time-domain (FDTD) method.^[18] In these simulations, the refractive indexes of glass and water are 1.515 and 1.333, respectively. The refractive index of gold at wavelength of 671 nm is set as 0.14 + i*3.80 based on experimental values.^[19] The incident annular beam has linear polarization (along X-axis) and the range of the incident angles is from 63° to 79°, which is consistent with the experimental conditions. Figure 4a,c show the normalized electric field distributions on the X-Z plane that is perpendicular to the surface of the substrate. For the optical virtual probes induced by either the focused plasmonic waves or focused evanescent waves, their strength decays rapidly with the distance from the surface of the substrate, which highly reduces the axial size of the trapping volume. Therefore, optical trapping of microobjects can be limited in the near-field region of the substrate.^[5]

Figure4.

Numerical simulations of the optical virtual probes. |E₀| is the maximum amplitude of the incident beam. a,b) Normalized electric-field (|E|/|E₀|) distributions of the focused plasmonic waves (water/gold interface) in the X-Z plane and X-Y plane, respectively. c,d) Normalized electric-field (|E|/|E₀|) distributions of the focused evanescent waves (water-glass interface) in the X-Z plane and X-Y plane, respectively. The Z-axis is normal to the surface of the substrate. The white double-headed arrows indicate the polarized direction of the incident beam. The white dashed lines on (a) and (c) represent the water-gold interface and water-glass interface, respectively.

In Figure 4b,d, the normalized electric field distributions on the surface of the substrates (X-Y plane) show the optical virtual probes located in the center, which can induce the optical gradient forces to attract the beads. Due to different polarization properties of the plasmonic waves (water-gold interface) and evanescent waves (water-glass interface), the focused patterns on Figure 4a,b are different from that on Figure 4c,d, respectively. For example, the center of the focused plasmonic field is dark (Figure 4b), whereas the center of the focused evanescent field is bright (Figure 4d).^[15] But the lateral size (along X-axis or Y-axis) of the two virtual probes are both much smaller than the diameter of the PS beads used in the experiment. Additionally, in our experiment, the incident laser beam with radially polarization was also used to excite the plasmonic waves at the water-gold interface that could form a solid virtual probe with a bright center and was the same as that formed by the focused evanescent waves (Figure 4c).^[10,11,15] In this case, the experimental result is shown in Movie S3, Supporting Information and many beads are trapped by the focused plasmonic waves under a radially polarized incident beam. The difference in the trapping of PS beads shown in Figure 3 cannot be explained from the electric field distribution of the two near-field optical virtual probes.

2.3. Thermal Effect on Gold Film

As reported elsewhere,^[20,22] whenever plasmonic tweezer are used for trapping, the problem of heating due to the light absorption by metal must be taken account. In our experiment, the heat could be efficiently generated because of SPs absorption by the gold film. To test this assertion, numerical simulations of the temperature distribution and thermal convection were performed in the focused plasmonic field on the gold film using the finite element method (FEM).^[23] The power of the incident beam is 10 mW and the environment temperature is set to 293.15 K. Figure 5a,c show the temperature distributions in the vertical X- Z plane and horizontal X-Y plane for the focused plasmonic field, respectively. The maximum temperature can reach 55 °C in the focused center due to the strong electric field located in the center, and the temperature gradient will lead to the thermal convection in the water solution. Figure 5b,d show the thermal convections in the X-Z plane and X-Y plane, respectively. The calculated results demonstrate that the convective fluids move toward the center while near the substrate and then direct upward just above the center, which will create a circulating zone around the focused center in the water. Because of the convection, many beads suspended in water will be drawn into the center from adjacent regions and then be trapped together, which is in good agreement with the experimental results.

Figure 5.

Numerical simulations of the temperature distribution and thermal convection in the focused plasmonic field. The power of the incident laser is 10 mW. a,c) Temperature distribution on the X-Z plane and X-Y plane, where the X-Y plane is 5 pm above the water-gold interface. b,d) Distribution of thermal convective velocity on the X-Z plane and X-Y plane, where the X-Y plane is 5 pm above the water-gold interface. The background shows the velocity distributions of convection in water, and the black arrows indicate the direction of convection in water.

In order to verify the heat generated in the focused plasmonic field, the temperature change was detected with a temperature-dependent fluorescent dye, as shown in Figure 6a. The fluorescence intensity of rhodamine B (RhB) solution is strongly temperature dependent. As the temperature increases, the fluorescence intensity of RhB decreases.^[24,25] The RhB molecules were dissolved in water and placed on the gold film and glass substrate, respectively. The plot of normalized fluorescence intensity versus incident laser power on different substrate is shown in Figure 6b. The normalized results clearly show that the fluorescence intensity on gold film decreases rapidly with increasing power while the fluorescence intensity on glass substrate does not show evident change when the incident power increases, which shows the heat generated on the gold film is much more than that on glass substrate. The difference in the trapping behavior is attributed to the strong photothermal effects on the gold film.

Figure 6.

Verification of heat generation via measurement of the temperature-dependent fluorescence intensity. a) Experimental setup for measurement of the temperature-dependent fluorescence intensity. A 532 nm laser beam is used to excite the dye molecules (RhB in a water solution) with a constant power. A 671 nm laser beam is used to excite the plasmonic waves on the gold film or evanescent waves on the glass substrate. Two filters are used to reject the laser beams at 532 and 671 nm to ensure that only the fluorescence can reach the detector. b) Normalized fluorescence intensity versus incident power on different substrates. The red hollow dots indicate the fluorescence intensity on the glass substrate. The blue hollow dots indicate the fluorescence intensity on the gold film.

2.4. Trapping Beads During the Moving of Stage Along Z-Axis

To further verify this point, the stage on which the substrates are placed was moved along Z-axis to change direction of the optical force acting on PS beads. When the incident focal point is inside the water solution above the glass substrate, the surface waves propagate inward to the focused center and form the virtual probe, as illustrated in Figure 7m. In this case, the optical force will attract PS beads to the center of the excitation region. When the focus of the incident annular beam is immersed in the glass substrate below the water solution, the surface waves propagate outward and the virtual probe cannot be formed, as illustrated in Figure 7n.^[16] In this case, the optical force will push the beads away from the center. In Figure 7a, a high concentration bead solution is placed on the gold film and many beads (more than thirty) are drawn to the area where the laser beam is focused (marked with red circle). When the stage is moved upward along the Z-axis (away from the stationary objective), then the focal point of the incident annular beam is shifted from the water solution to the glass substrate (the gold film is deposited on the glass). During this movement, the trapped beads are still accumulated at the center of the excited region, as shown in Figure 7a-h. This phenomenon clearly verifies that the trapping of multiple beads on the gold film is not due to the optical forces, but due to the thermal convections.

Figure 7.

Trapping beads during the moving of the stage along Z-axis away from the stationary objective. a-h) Many beads are simultaneously trapped by the focused plasmonic waves on a gold film. From (a) to (h), the stage is moved upward along the Z-axis with a step of 2 pm. During this movement, the beads are still accumulated in the excited region. i-l) One bead is trapped by the focused evanescent waves on the glass substrate. From (i) to (l), the stage is moved upward along the Z-axis with a step of 1 pm. During this movement, this bead escapes from the excited region. m) When the focal point of the incident annular beam is inside the water solution above the glass substrate, the surface waves propagate to the center and form the virtual probe. n) When the stage is moved upward along the Z-axis away from the stationary objective and the focal point is inside the glass substrate below the water solution, the surface waves propagate outward and the virtual probe disappears.

Different trapping behavior was found on the glass substrate. When a high concentration bead solution is placed on the glass substrate, only one bead can be trapped as shown in Figure 7i. When the glass substrate is moved upward along the Z-axis (away from the stationary objective), the focal point inside the water solution (Figure 7m) is shifted to the inside of the glass substrate (Figure 7n). During this movement, the trapped bead (marked with red circle) escapes from the center of the excited region as shown in Figure 7i-l, which is consistent with the change of the optical force acting on PS bead. This phenomenon demonstrates that there is no obvious optothermal effect when the laser beam is focused onto the bare glass substrate and the trapping of single bead is due to the optical force in the evanescent field on glass substrate. Based on the comparisons shown in Figure 7, the trapping differences of these two kinds of focused surface waves can be explained from the different photothermal effects on the gold film and the glass substrate.

2.5. Effect of the Incident Power on the Trapping

As the incident power increases, the heat generated by the focused plasmonic field on the gold film increases, which is expected to result in stronger thermal convection. A series of images were captured as the power was increased to 40 mW (Figure 8). The bead marked with red circle is first drawn into the excitation region (Figure 8a-e), and then it is pushed away along Z-axis from the water-gold interface when it approaches to the center of the excited region. Beads cannot be trapped in the central region because the convective force directed upward exceeds the gravity of the beads. This process can be more clearly observed in Movie S4, Supporting Information, which shows that some beads even cannot approach the center and are pushed away at the boundary of the excited region due to the strong thermal convective force. The 40 mW laser beam was also used to excite the evanescent waves on the glass substrate, but we did not observe the phenomenon that the beads are pushed away in focused evanescent field, which further verifies that the thermal effect on a glass substrate is much weaker than that on a gold film at the same incident power.

Figure 8.

Effect of the incident power on the trapping. The incident power is increased to 40 mW. a-j) Successive images of beads trapped by the focused plasmonic waves. The image sequence is from (a) to (j) with a time interval between images of 1 s. a-e) The bead marked with red circle is first drawn into the excitation region. f-j) When the bead approaches to the center of the excited region, it is pushed away along Z-axis from the water-gold interface.

3. Conclusion

In conclusion, two kinds of surface waves, the evanescent waves at a water-glass interface and plasmonic waves at a metal-water interface, are excited and focused by a high NA objective, which can be used to trap dielectric micro-/nanobeads in liquid. For trapping using the focused evanescent waves on the glass substrate, only one bead can be trapped. In contrast, for trapping based on the focused plasmonic waves on the gold film, many beads can be simultaneously attracted to the center of the excitation region. The difference in trapping behavior is attributed to the photothermal effects.^[26,27] Heat is generated in the focused plasmonic field due to the light absorption by the gold film, which can lead to thermal convection in the water solution. The beads suspended in water will be drawn into the center of the excited region by the thermal convective fluids and accumulated at the excited region. Numerical simulations show that convective fluids near the substrate move toward the center of the excited region, which agree well with the experimental results. Our work presented here is the first exploration on the photothermal effects of the focused plasmonic waves by a high NA objective and their function in trapping, which may find broad applications in trapping of biomolecules.^[28]

4. Experimental Section

Experimental Setup and Samples Preparation

The schematic of the experiment setup is shown in Figure 1. The laser beam was from a solid state laser with a wavelength of 671 nm. The cone angle of the axicon was 10.0° (AX2510-A, Thorlab). The samples were placed on a high-resolution reference-class piezo stage system (PInano Cap XYZ Piezo System, Physik Instrumente (PI) GmbH & Co. KG, Germany) that could precisely control the distance between the substrate and the objective (Figure 7) and adjust the relative distance between the untrapped beads and the excited region (Figure 3). Two detectors (Retiga 6000, Qlmaging, Canada and Neo-sCMOS, Andor, UK) were used to record the FFP and BFP images. The gold films were prepared by depositing Au on glass coverslips (0.17 mm thickness) with E-beam evaporator (K. J. Lesker, LAB 18, base pressure: 10⁻⁶, deposited velocity rate: 3 A s⁻¹) for 150 s. The original solution of PS beads (4016A) was purchased from Thermo Fisher Scientific, USA, whose diameter was 1.587 pm ± 0.018 pm. The original solution was diluted to an appropriate concentration and a drop of this solution was placed on a bare glass substrate or a gold film.