2916
OPTICS LETTERS / Vol. 39, No. 10 / May 15, 2014
Digital holographic microscopy with coupled optical fiber trap for cell measurement and manipulation Samira Ebrahimi,1 Ali-Reza Moradi,1,2 Arun Anand,3 and Bahram Javidi4,* 1
2 3
Department of Physics, University of Zanjan, P.O. Box 45195-313, Zanjan, Iran Optics Research Center, Institute for Advanced Studies in Basic Sciences, P.O. Box 45137-6673, Zanjan, Iran
Optics Laboratory, Applied Physics Department, Faculty of Technology and Engineering, MS University of Baroda, Vadodara 390001, India 4
Department of Electrical and Computer Engineering, University of Connecticut, Storrs, Connecticut 06269-4157, USA *Corresponding author: [email protected] Received February 10, 2014; revised April 7, 2014; accepted April 8, 2014; posted April 8, 2014 (Doc. ID 205951); published May 8, 2014 We present an integrated optical system for three-dimensional (3D) imaging of micrometer-sized samples, while immobilizing and manipulating the samples by means of an optical fiber trap. Optical traps allow us to apply and measure pico-Newton-sized forces, and perform detailed measurements of micrometer-sized dielectric systems in the field of biology. The integrated 3D system can be used as a major tool in the field of biophysics. The trap is built using a tapered optical fiber to enhance the effective numerical aperture of the fiber. The trapping system is mounted on a conventional microscope, in which the two eyepieces’ output ports are used as the paths of an off-axis self-referencing digital holographic microscopy (DHM) setup. The trap is calibrated using a high-speed camera, and trap stiffness is determined through the power spectrum method. The compact setup provides an elegant apparatus for temporally stable DHM for 3D imaging of optically controlled samples. Three-dimensional information and quantitative phase contrast images of the trapped samples are obtained by postprocessing the recorded digital holograms. Experiments were performed on lipids and red blood cells. Quantitative phase contrast images and temporal evolution of optical thickness of trapped samples are presented. © 2014 Optical Society of America OCIS codes: (170.6900) Three-dimensional microscopy; (120.3890) Medical optics instrumentation; (120.5050) Phase measurement; (350.4855) Optical tweezers or optical manipulation; (060.2370) Fiber optics sensors; (280.1415) Biological sensing and sensors. http://dx.doi.org/10.1364/OL.39.002916
Optical trap (OT) is a method to immobilize and manipulate micron-sized samples in suspension [1–4]. The high NA microscope objectives that are used in most conventional optical trapping setups suffer from short working distances. OTs based on optical fiber enable manipulation at much larger depths, which has application in biophysics and nanoscience research [5]. In the first demonstration of optical fiber trapping (OFT), two counterpropagating laser beams were introduced directly into the sample from the side by two aligned optical fibers [6]. Since then several implementations of OFT have been reported including 3D single-beam OFT [7]. Optical fibers can be tapered by heating and stretching in order to change their light coupling or light propagation properties [8]. The use of such fibers improves the stability of the fiber trap by increasing the beam divergence angle and providing greater accessibility to a large trapping volume. The increased effective numerical aperture of a tapered fiber may be calculated as NAi R × NA0 ;
(1)
where NAi and NA0 are the input and output numerical apertures, respectively, and R represents the ratio of taper input diameter to output diameter [9]. Among the methods for imaging phase objects by measuring phase variation, digital holographic microscopy (DHM) is an effective and nondestructive technique providing quantitative phase images of biological samples and living organelles [10–13]. The reconstruction of holograms recorded by digital sensor is carried out numerically providing whole field information about the object [14]. In off-axis holography a small angle between 0146-9592/14/102916-04$15.00/0
the object and reference wave is introduced. This separates the undiffracted reference beam from the virtual and real objects at the reconstruction plane. However, the introduction of a separate reference beam requires the use of additional optical elements, which makes the setup highly sensitive to environmental and mechanical vibrations. This, in turn, results in uncorrelated optical path length changes in the two arms of the interferometer, leading to higher noise. Common-path geometry for interferometric techniques greatly reduces these uncorrelated phase changes. Common path geometry employing self-referencing can be ideal for low noise interferometric microscopy. This was recently demonstrated using a Lloyd’s mirror configuration [15]. Here we utilize a compact and simple common-path off-axis DHM setup based on a conventional binocular microscope. Further, the ability to combine the techniques of DHM and OFT is demonstrated. Earlier work has shown it is possible to combine off-axis DHM and OT [16], but it requires special optics for beam splitting and recombination. The combination of DHM and OT is of particular advantage for quantitative visualization of 3D structures that are trapped by laser beam. The integration of DHM and OT is straightforward if a single microscope objective is used for both DHM and OT. However, the microscope objective lens can be optimized either for acquisition of digital holograms or trapping of the objects. Complete decoupling of OT and DHM paths is an elegant approach to overcome the limitations. An additional side view objective may be used for independent access to the hologram recording components as well as the trapping configuration. However, for this setup, long working distance microscope objectives have to be used, and © 2014 Optical Society of America
May 15, 2014 / Vol. 39, No. 10 / OPTICS LETTERS
the conventional sample chamber has to be replaced by a square glass capillary. OFT can be used to overcome these practical limitations. The experimental setup that we used is mounted on an upright microscope and is shown in Fig. 1. The DHM part of the setup is based on the combination of the methods presented in [15] and Mach–Zehnder geometry. Both the object and the reference waves are obtained from the light transmitted through the sample (object wavefront). Light from a He–Ne laser (MEOS, 632.8 nm, 5 mW) is used for the generation of the holograms. This beam after reflection from a dichroic mirror (DM) and the microscope condenser (C, NA 1.25, Olympus) is transmitted through the sample (S). The transmitted light carries the complete object information. After it passes through the microscope objective (MO2, 100×, NA 1.25, WD 0.17 mm, Olympus), it is separated into two waves by the binocular module of the microscope (BM). One of the waves is reflected from a beam splitter (BS) onto the imaging sensor (DCC1545M, Thorlabs, 8 bit dynamic range, 5.2 μm pixel pitch) and interferes with the second wavefront that is reflected from the mirror (M). The common path setup is highly stable and provides easy control of interference fringe frequency. Holograms are formed from the interference of the object wavefront with a portion of the same wavefront that does not contain object information. The M together with the BS creates an off-axis geometry. The main advantage of the presented setup over the one based on the use of Lloyd’s mirror configuration is the capability to mount the system on a conventional microscope. Another laser (DPSS, 85- GSS- 309, Melles Griot, 532 nm, maximum output power of 3 W) beam is used as the source in the OFT setup. The laser beam is focused on the entrance of an optical fiber by a microscope objective (MO1). Experiments were conducted in the sample chambers with an open surface (not covered by a slip). The chambers were constructed of coverslips, secured by tape, and
Fig. 1. Integrated DHM and OFT setup. MO, microscope objective; DHM laser, laser for digital holographic microscopy; DM, dichroic mirror; S, sample; C, condenser; BM, binocular module of microscope; M, mirror; BS, beam splitter; 2θ, angle of tapered fiber.
2917
filled with a small volume of the sample. The tapered end of the fiber was mounted at an angle of 15° on a x-y-z micropositioner and inserted into the chamber from above and was capable of changing the trap position. The inset of Fig. 1 shows the image of the fiber tip. The NA of the fiber was calculated as 1.04, and the tip’s tapered angle (2θ) was 25.8°. The insertion angle of the fiber end is a critical parameter for OFT [17]. A large insertion angle produces a weak trapping force on the object. According to the linear relation between the trapping force exerted on a particle and the displacement of the particle with respect to the trap center, the trap stiffness k can be computed and used for force measurements. We first calibrated the trap by using the video microscopy method [18] and power spectrum analysis [19,20]. Position detection and tracking of the trapped particle can be achieved by high-speed camera imaging. The movement of the particle in the OT was observed via video position detection. In our setup the image sensor at a full resolution of 1280 × 1024 pixels had a speed of 25 fps. However, with a reduced region of interest (65 × 65 pixels) speeds up to 1000 fps can be reached. Polystyrene microspheres of 1.65 μm diameter were trapped and used for OFT calibration. The power spectrum of the Brownian motion for a trapped particle in each direction is given by Lorentzian function [19]. Acquiring power spectrum data and fitting it with a Lorentzian function in each direction, yields the corresponding corner frequencies. The corner frequency f c is related to the trap stiffness as f c k∕2πγ, where γ 6πηa is the drag coefficient, η is the fluid viscosity, and a is the radius of the trapped particle. Figure 2 shows the power spectrums along the x and y axes for an OFT
Fig. 2. Power spectrum of a trapped polystyrene particle at 200 mW laser output power in (a) x and (b) y directions. f c in the inset is the corner frequency.
2918
OPTICS LETTERS / Vol. 39, No. 10 / May 15, 2014
trapped polystyrene with the laser output power of 200 mW. Several validation experiments were performed to determine the setup’s capabilities. The experiments were conducted for three different objects: polystyrene microspheres, 1-palmitoyl-2-oleoyl- sn-glycero-3-phosphocholine (POPC) lipids, and red blood cells (RBCs). After calibration using polystyrene microspheres, the tip of OFT fiber was brought closer and used to apply pressure on lipids. A stack of POPC lipid, one of the most important bilayer lipids in biological experiments, was used as a sample. Lipids are a major class of biomolecules that includes fatty acids, waxes, glycerol, triacylglycerols, phospholipids, and cholesterols. The structure of lipids varies tremendously and they are involved in a wide array of processes from compartmentalization of the cell with membranes to energy storage and cell signaling [21]. The OFT tip was used to stretch the lipid membrane and induce morphological change by the force of the fiber trap. Figure 3 shows conventional microscopy images of a stack of POPC lipid that is stretched in different directions. The direction of the stage movement with respect to the tip position in each row is shown by the arrow in the first image of the row. The amount of membrane stretching is shown in the inset by a binary image of the region of interest. The reduction of the gray-level images to binary images is performed by applying a thresholding algorithm on the selected stack of lipid. The threshold value is adapted according to the local image characteristics. In order to confirm that the stretch of the lipid is due to the presence of optical force and not due to the adhering to the end of the tip, a control experiment is conducted in which after stretching the lipid the laser was switched off and it was found that the lipid returned to its normal state. The OFT was then coupled with DHM for quantitative phase measurement to directly observe the morphology changes happening to a trapped RBC. The samples were prepared at 37°C, and the experiments were conducted at room temperature, i.e., 25°C. Freshly collected human RBCs were diluted with a buffer containing 150 mM NaCl to obtain a 0.1% hematocrit value, making it suitable for
Fig. 3. Stretching of POPC lipid membrane in different directions (shown by arrows) by OFT tip. Scale bar in image (a) is 10 μm. Images are obtained by conventional microscopy.
single-cell trapping experiments. A single RBC was trapped by the tip of OFT to examine the performance of the DHM OFT system. Figure 4 shows the results of RBC trapping with DHM. Objects shaped like RBC usually try to align themselves with the beam axis in order to find the maximum volume of trapping region [22]. We trapped a single RBC stably at a laser output power of 500 mW and acquired digital holograms of it simultaneously through the DHM system. Figures 4(a)–4(c) show the hologram of the cell while the stage is moved laterally. The cells remained trapped for movement velocities of up to 35 μm∕s. We utilized the angular spectrum propagation approach in scalar diffraction theory for numerical reconstruction of the holograms [23]. The phase and intensity of the reconstructed wavefront can be computed from the propagated complex wave field of the hologram at the reconstruction plane. The reconstructed phase patterns are unwrapped by Goldstein’s branch-cut method to convert them into continuous-phase distributions. Figures 4(d) and 4(e) are the associated intensity and phase patterns obtained from the reconstructed hologram. In order to remove the background contaminations from the sample container and the fluid, we recorded a reference hologram in which no RBC was trapped. Figure 4(f) is the phase contrast image of the RBC obtained after subtracting the background phase information obtained from the reference hologram. Figures 4(g)–4(i) are different illustrations created using the obtained phase information. As can be seen, the phase of the RBC and the rest of the field of view have different value ranges. Assuming negligible changes for refractive index, the thickness at an arbitrary point on the trapped RBC can also be calculated and hence the volume of the immobilized cell can be obtained. Figure 5 shows the temporal variations of a trapped RBC thickness, at the points given in Fig. 4(h).
Fig. 4. Trapped RBC: (a), (b), (c) lateral movement of the stage while the RBC remained trapped, (d) the intensity, and (e) the phase pattern of the hologram, (f) the filtered phase difference distribution, (g) the reconstructed 2D thickness profile of the field of view, and (h) the 2D and (i) 3D thickness profiles of the selected region of interest of (g).
May 15, 2014 / Vol. 39, No. 10 / OPTICS LETTERS
Fig. 5. Temporal variations of a trapped RBC thickness, at the points A, B, and C shown in Fig. 4(h).
The points A, B, and C are chosen out of the cell, on the side of the cell, and in a central part of the trapped cell, respectively. In spite of the Brownian fluctuation of the cell in the trap, the mean membrane thickness and small deviations of the measured thickness show the stability of the trap site. Analysis of cell membrane fluctuations through reconstruction of digital holograms can be used to study the effect of important organic compounds or external stimuli on structural properties of cells [24–26]. In conclusion, we have demonstrated an integrated digital holographic microscopy and optical micromanipulation system based on the use of a binocular conventional microscope. A tapered optical fiber tip was used to create a controllable and simple optical trap, while the trapped object was monitored live by recording digital holograms. Validation experiments were performed to confirm the capabilities of the setup. The developed method being common path in nature is very stable during mechanical vibrations and can be used as a portable system to attain detailed dynamic information of trapped single cells. A. R. Moradi acknowledges support of the Center for International Scientific Studies and Collaboration within the ICRP program. References 1. A. Ashkin, Phys. Rev. Lett. 24, 156 (1970). 2. K. C. Neuman and S. M. Block, Rev. Sci. Instrum. 75, 2787 (2004). 3. P. J. Rodrigo, L. Kelemen, D. Palima, C. A. Alonzo, P. Ormos, and J. Glückstad, Opt. Express 17, 6578 (2009).
2919
4. S. Zwick, T. Haist, Y. Miyamoto, L. He, M. Warber, A. Hermerschmidt, and W. Osten, J. Opt. A 11, 034011 (2009). 5. S. K. Mohanty and M. W. Berns, J. Biomed. Opt. 13, 054049 (2008). 6. A. Constable, J. Kim, J. Mervis, F. Zarinetchi, and M. Prentiss, Opt. Lett. 18, 1867 (1993). 7. S. I. Eom, Y. Takaya, and T. Hayashi, Precis. Eng. 33, 3 (2009). 8. E. R. Lyons and G. J. Sonek, Appl. Phys. Lett. 66, 1584 (1995). 9. J. Hecht, Understanding Fiber Optics, 5th ed. (Prentice Hall, 2005). 10. P. Marquet, B. Rappaz, P. Magistretti, E. Cuche, Y. Emery, T. Colomb, and C. Depeursinge, Opt. Lett. 30, 468 (2005). 11. C. J. Mann, L. Yu, C. Lo, and M. K. Kim, Opt. Express 13, 8693 (2005). 12. P. Ferraro, D. Alferi, S. De Nicola, L. De Petrocellis, A. Finizio, and G. Pierattini, Opt. Lett. 31, 1405 (2006). 13. L. Miccio, P. Memmolo, F. Merola, S. Fusco, V. Embrione, A. Paciello, M. Ventre, P. A. Netti, and P. Ferraro, Lab Chip 14, 1129 (2014). 14. U. Schnars and W. Jueptner, Digital Holography: Digital Hologram Recording, Numerical Reconstruction, and Related Techniques (Springer, 2005). 15. V. Chhaniwal, A. S. Singh, R. A. Leitgeb, B. Javidi, and A. Anand, Opt. Lett. 37, 5127 (2012). 16. M. Daneshpanah, S. Zwick, F. Schaal, M. Warber, B. Javidi, and W. Osten, J. Display Technol. 6, 490 (2010). 17. K. Taguchi, M. Tanaka, and M. Ikeda, Opt. Rev. 8, 156 (2001). 18. O. Otto, C. Gutsche, F. Kremer, and U. F. Keyser, Rev. Sci. Instrum. 79, 023710 (2008). 19. F. Gittes and C. F. Schmidt, Methods Cell Biol. 55, 129 (1998). 20. P. M. Hansen, I. M. Tolic-Norrelykke, H. Flyvbjerg, and K. Berg-Sorensen, Comput. Phys. Commun. 174, 518 (2006). 21. D. L. Nelson and M. M. Cox, Lehninger Principles of Biochemistry, 3rd ed. (Worth, 2000). 22. S. C. Grover, R. C. Gauthier, and A. G. Skirtach, Opt. Express 7, 533 (2000). 23. A. Anand, V. K. Chhaniwal, and B. Javidi, J. Display Technol. 6, 500 (2010). 24. Y. Kim, K. Kim, and Y. Park, in Measurement Techniques for Red Blood Cell Deformability: Recent Advances, Blood Cell—An Overview of Studies in Hematology (INTECH, 2012), pp. 167–194. 25. K. Lee, K. Kim, J. Jung, J. Heo, S. Cho, S. Lee, G. Chang, Y. Jo, H. Park, and Y. Park, Sensors 13, 4170 (2013). 26. F. Merola, L. Miccio, P. Memmolo, G. Di Caprio, A. Galli, R. Puglisi, D. Balduzzi, G. Coppola, P. Netti, and P. Ferraro, Lab Chip 13, 4512 (2013).
OPTICS LETTERS / Vol. 39, No. 10 / May 15, 2014
Digital holographic microscopy with coupled optical fiber trap for cell measurement and manipulation Samira Ebrahimi,1 Ali-Reza Moradi,1,2 Arun Anand,3 and Bahram Javidi4,* 1
2 3
Department of Physics, University of Zanjan, P.O. Box 45195-313, Zanjan, Iran Optics Research Center, Institute for Advanced Studies in Basic Sciences, P.O. Box 45137-6673, Zanjan, Iran
Optics Laboratory, Applied Physics Department, Faculty of Technology and Engineering, MS University of Baroda, Vadodara 390001, India 4
Department of Electrical and Computer Engineering, University of Connecticut, Storrs, Connecticut 06269-4157, USA *Corresponding author: [email protected] Received February 10, 2014; revised April 7, 2014; accepted April 8, 2014; posted April 8, 2014 (Doc. ID 205951); published May 8, 2014 We present an integrated optical system for three-dimensional (3D) imaging of micrometer-sized samples, while immobilizing and manipulating the samples by means of an optical fiber trap. Optical traps allow us to apply and measure pico-Newton-sized forces, and perform detailed measurements of micrometer-sized dielectric systems in the field of biology. The integrated 3D system can be used as a major tool in the field of biophysics. The trap is built using a tapered optical fiber to enhance the effective numerical aperture of the fiber. The trapping system is mounted on a conventional microscope, in which the two eyepieces’ output ports are used as the paths of an off-axis self-referencing digital holographic microscopy (DHM) setup. The trap is calibrated using a high-speed camera, and trap stiffness is determined through the power spectrum method. The compact setup provides an elegant apparatus for temporally stable DHM for 3D imaging of optically controlled samples. Three-dimensional information and quantitative phase contrast images of the trapped samples are obtained by postprocessing the recorded digital holograms. Experiments were performed on lipids and red blood cells. Quantitative phase contrast images and temporal evolution of optical thickness of trapped samples are presented. © 2014 Optical Society of America OCIS codes: (170.6900) Three-dimensional microscopy; (120.3890) Medical optics instrumentation; (120.5050) Phase measurement; (350.4855) Optical tweezers or optical manipulation; (060.2370) Fiber optics sensors; (280.1415) Biological sensing and sensors. http://dx.doi.org/10.1364/OL.39.002916
Optical trap (OT) is a method to immobilize and manipulate micron-sized samples in suspension [1–4]. The high NA microscope objectives that are used in most conventional optical trapping setups suffer from short working distances. OTs based on optical fiber enable manipulation at much larger depths, which has application in biophysics and nanoscience research [5]. In the first demonstration of optical fiber trapping (OFT), two counterpropagating laser beams were introduced directly into the sample from the side by two aligned optical fibers [6]. Since then several implementations of OFT have been reported including 3D single-beam OFT [7]. Optical fibers can be tapered by heating and stretching in order to change their light coupling or light propagation properties [8]. The use of such fibers improves the stability of the fiber trap by increasing the beam divergence angle and providing greater accessibility to a large trapping volume. The increased effective numerical aperture of a tapered fiber may be calculated as NAi R × NA0 ;
(1)
where NAi and NA0 are the input and output numerical apertures, respectively, and R represents the ratio of taper input diameter to output diameter [9]. Among the methods for imaging phase objects by measuring phase variation, digital holographic microscopy (DHM) is an effective and nondestructive technique providing quantitative phase images of biological samples and living organelles [10–13]. The reconstruction of holograms recorded by digital sensor is carried out numerically providing whole field information about the object [14]. In off-axis holography a small angle between 0146-9592/14/102916-04$15.00/0
the object and reference wave is introduced. This separates the undiffracted reference beam from the virtual and real objects at the reconstruction plane. However, the introduction of a separate reference beam requires the use of additional optical elements, which makes the setup highly sensitive to environmental and mechanical vibrations. This, in turn, results in uncorrelated optical path length changes in the two arms of the interferometer, leading to higher noise. Common-path geometry for interferometric techniques greatly reduces these uncorrelated phase changes. Common path geometry employing self-referencing can be ideal for low noise interferometric microscopy. This was recently demonstrated using a Lloyd’s mirror configuration [15]. Here we utilize a compact and simple common-path off-axis DHM setup based on a conventional binocular microscope. Further, the ability to combine the techniques of DHM and OFT is demonstrated. Earlier work has shown it is possible to combine off-axis DHM and OT [16], but it requires special optics for beam splitting and recombination. The combination of DHM and OT is of particular advantage for quantitative visualization of 3D structures that are trapped by laser beam. The integration of DHM and OT is straightforward if a single microscope objective is used for both DHM and OT. However, the microscope objective lens can be optimized either for acquisition of digital holograms or trapping of the objects. Complete decoupling of OT and DHM paths is an elegant approach to overcome the limitations. An additional side view objective may be used for independent access to the hologram recording components as well as the trapping configuration. However, for this setup, long working distance microscope objectives have to be used, and © 2014 Optical Society of America
May 15, 2014 / Vol. 39, No. 10 / OPTICS LETTERS
the conventional sample chamber has to be replaced by a square glass capillary. OFT can be used to overcome these practical limitations. The experimental setup that we used is mounted on an upright microscope and is shown in Fig. 1. The DHM part of the setup is based on the combination of the methods presented in [15] and Mach–Zehnder geometry. Both the object and the reference waves are obtained from the light transmitted through the sample (object wavefront). Light from a He–Ne laser (MEOS, 632.8 nm, 5 mW) is used for the generation of the holograms. This beam after reflection from a dichroic mirror (DM) and the microscope condenser (C, NA 1.25, Olympus) is transmitted through the sample (S). The transmitted light carries the complete object information. After it passes through the microscope objective (MO2, 100×, NA 1.25, WD 0.17 mm, Olympus), it is separated into two waves by the binocular module of the microscope (BM). One of the waves is reflected from a beam splitter (BS) onto the imaging sensor (DCC1545M, Thorlabs, 8 bit dynamic range, 5.2 μm pixel pitch) and interferes with the second wavefront that is reflected from the mirror (M). The common path setup is highly stable and provides easy control of interference fringe frequency. Holograms are formed from the interference of the object wavefront with a portion of the same wavefront that does not contain object information. The M together with the BS creates an off-axis geometry. The main advantage of the presented setup over the one based on the use of Lloyd’s mirror configuration is the capability to mount the system on a conventional microscope. Another laser (DPSS, 85- GSS- 309, Melles Griot, 532 nm, maximum output power of 3 W) beam is used as the source in the OFT setup. The laser beam is focused on the entrance of an optical fiber by a microscope objective (MO1). Experiments were conducted in the sample chambers with an open surface (not covered by a slip). The chambers were constructed of coverslips, secured by tape, and
Fig. 1. Integrated DHM and OFT setup. MO, microscope objective; DHM laser, laser for digital holographic microscopy; DM, dichroic mirror; S, sample; C, condenser; BM, binocular module of microscope; M, mirror; BS, beam splitter; 2θ, angle of tapered fiber.
2917
filled with a small volume of the sample. The tapered end of the fiber was mounted at an angle of 15° on a x-y-z micropositioner and inserted into the chamber from above and was capable of changing the trap position. The inset of Fig. 1 shows the image of the fiber tip. The NA of the fiber was calculated as 1.04, and the tip’s tapered angle (2θ) was 25.8°. The insertion angle of the fiber end is a critical parameter for OFT [17]. A large insertion angle produces a weak trapping force on the object. According to the linear relation between the trapping force exerted on a particle and the displacement of the particle with respect to the trap center, the trap stiffness k can be computed and used for force measurements. We first calibrated the trap by using the video microscopy method [18] and power spectrum analysis [19,20]. Position detection and tracking of the trapped particle can be achieved by high-speed camera imaging. The movement of the particle in the OT was observed via video position detection. In our setup the image sensor at a full resolution of 1280 × 1024 pixels had a speed of 25 fps. However, with a reduced region of interest (65 × 65 pixels) speeds up to 1000 fps can be reached. Polystyrene microspheres of 1.65 μm diameter were trapped and used for OFT calibration. The power spectrum of the Brownian motion for a trapped particle in each direction is given by Lorentzian function [19]. Acquiring power spectrum data and fitting it with a Lorentzian function in each direction, yields the corresponding corner frequencies. The corner frequency f c is related to the trap stiffness as f c k∕2πγ, where γ 6πηa is the drag coefficient, η is the fluid viscosity, and a is the radius of the trapped particle. Figure 2 shows the power spectrums along the x and y axes for an OFT
Fig. 2. Power spectrum of a trapped polystyrene particle at 200 mW laser output power in (a) x and (b) y directions. f c in the inset is the corner frequency.
2918
OPTICS LETTERS / Vol. 39, No. 10 / May 15, 2014
trapped polystyrene with the laser output power of 200 mW. Several validation experiments were performed to determine the setup’s capabilities. The experiments were conducted for three different objects: polystyrene microspheres, 1-palmitoyl-2-oleoyl- sn-glycero-3-phosphocholine (POPC) lipids, and red blood cells (RBCs). After calibration using polystyrene microspheres, the tip of OFT fiber was brought closer and used to apply pressure on lipids. A stack of POPC lipid, one of the most important bilayer lipids in biological experiments, was used as a sample. Lipids are a major class of biomolecules that includes fatty acids, waxes, glycerol, triacylglycerols, phospholipids, and cholesterols. The structure of lipids varies tremendously and they are involved in a wide array of processes from compartmentalization of the cell with membranes to energy storage and cell signaling [21]. The OFT tip was used to stretch the lipid membrane and induce morphological change by the force of the fiber trap. Figure 3 shows conventional microscopy images of a stack of POPC lipid that is stretched in different directions. The direction of the stage movement with respect to the tip position in each row is shown by the arrow in the first image of the row. The amount of membrane stretching is shown in the inset by a binary image of the region of interest. The reduction of the gray-level images to binary images is performed by applying a thresholding algorithm on the selected stack of lipid. The threshold value is adapted according to the local image characteristics. In order to confirm that the stretch of the lipid is due to the presence of optical force and not due to the adhering to the end of the tip, a control experiment is conducted in which after stretching the lipid the laser was switched off and it was found that the lipid returned to its normal state. The OFT was then coupled with DHM for quantitative phase measurement to directly observe the morphology changes happening to a trapped RBC. The samples were prepared at 37°C, and the experiments were conducted at room temperature, i.e., 25°C. Freshly collected human RBCs were diluted with a buffer containing 150 mM NaCl to obtain a 0.1% hematocrit value, making it suitable for
Fig. 3. Stretching of POPC lipid membrane in different directions (shown by arrows) by OFT tip. Scale bar in image (a) is 10 μm. Images are obtained by conventional microscopy.
single-cell trapping experiments. A single RBC was trapped by the tip of OFT to examine the performance of the DHM OFT system. Figure 4 shows the results of RBC trapping with DHM. Objects shaped like RBC usually try to align themselves with the beam axis in order to find the maximum volume of trapping region [22]. We trapped a single RBC stably at a laser output power of 500 mW and acquired digital holograms of it simultaneously through the DHM system. Figures 4(a)–4(c) show the hologram of the cell while the stage is moved laterally. The cells remained trapped for movement velocities of up to 35 μm∕s. We utilized the angular spectrum propagation approach in scalar diffraction theory for numerical reconstruction of the holograms [23]. The phase and intensity of the reconstructed wavefront can be computed from the propagated complex wave field of the hologram at the reconstruction plane. The reconstructed phase patterns are unwrapped by Goldstein’s branch-cut method to convert them into continuous-phase distributions. Figures 4(d) and 4(e) are the associated intensity and phase patterns obtained from the reconstructed hologram. In order to remove the background contaminations from the sample container and the fluid, we recorded a reference hologram in which no RBC was trapped. Figure 4(f) is the phase contrast image of the RBC obtained after subtracting the background phase information obtained from the reference hologram. Figures 4(g)–4(i) are different illustrations created using the obtained phase information. As can be seen, the phase of the RBC and the rest of the field of view have different value ranges. Assuming negligible changes for refractive index, the thickness at an arbitrary point on the trapped RBC can also be calculated and hence the volume of the immobilized cell can be obtained. Figure 5 shows the temporal variations of a trapped RBC thickness, at the points given in Fig. 4(h).
Fig. 4. Trapped RBC: (a), (b), (c) lateral movement of the stage while the RBC remained trapped, (d) the intensity, and (e) the phase pattern of the hologram, (f) the filtered phase difference distribution, (g) the reconstructed 2D thickness profile of the field of view, and (h) the 2D and (i) 3D thickness profiles of the selected region of interest of (g).
May 15, 2014 / Vol. 39, No. 10 / OPTICS LETTERS
Fig. 5. Temporal variations of a trapped RBC thickness, at the points A, B, and C shown in Fig. 4(h).
The points A, B, and C are chosen out of the cell, on the side of the cell, and in a central part of the trapped cell, respectively. In spite of the Brownian fluctuation of the cell in the trap, the mean membrane thickness and small deviations of the measured thickness show the stability of the trap site. Analysis of cell membrane fluctuations through reconstruction of digital holograms can be used to study the effect of important organic compounds or external stimuli on structural properties of cells [24–26]. In conclusion, we have demonstrated an integrated digital holographic microscopy and optical micromanipulation system based on the use of a binocular conventional microscope. A tapered optical fiber tip was used to create a controllable and simple optical trap, while the trapped object was monitored live by recording digital holograms. Validation experiments were performed to confirm the capabilities of the setup. The developed method being common path in nature is very stable during mechanical vibrations and can be used as a portable system to attain detailed dynamic information of trapped single cells. A. R. Moradi acknowledges support of the Center for International Scientific Studies and Collaboration within the ICRP program. References 1. A. Ashkin, Phys. Rev. Lett. 24, 156 (1970). 2. K. C. Neuman and S. M. Block, Rev. Sci. Instrum. 75, 2787 (2004). 3. P. J. Rodrigo, L. Kelemen, D. Palima, C. A. Alonzo, P. Ormos, and J. Glückstad, Opt. Express 17, 6578 (2009).
2919
4. S. Zwick, T. Haist, Y. Miyamoto, L. He, M. Warber, A. Hermerschmidt, and W. Osten, J. Opt. A 11, 034011 (2009). 5. S. K. Mohanty and M. W. Berns, J. Biomed. Opt. 13, 054049 (2008). 6. A. Constable, J. Kim, J. Mervis, F. Zarinetchi, and M. Prentiss, Opt. Lett. 18, 1867 (1993). 7. S. I. Eom, Y. Takaya, and T. Hayashi, Precis. Eng. 33, 3 (2009). 8. E. R. Lyons and G. J. Sonek, Appl. Phys. Lett. 66, 1584 (1995). 9. J. Hecht, Understanding Fiber Optics, 5th ed. (Prentice Hall, 2005). 10. P. Marquet, B. Rappaz, P. Magistretti, E. Cuche, Y. Emery, T. Colomb, and C. Depeursinge, Opt. Lett. 30, 468 (2005). 11. C. J. Mann, L. Yu, C. Lo, and M. K. Kim, Opt. Express 13, 8693 (2005). 12. P. Ferraro, D. Alferi, S. De Nicola, L. De Petrocellis, A. Finizio, and G. Pierattini, Opt. Lett. 31, 1405 (2006). 13. L. Miccio, P. Memmolo, F. Merola, S. Fusco, V. Embrione, A. Paciello, M. Ventre, P. A. Netti, and P. Ferraro, Lab Chip 14, 1129 (2014). 14. U. Schnars and W. Jueptner, Digital Holography: Digital Hologram Recording, Numerical Reconstruction, and Related Techniques (Springer, 2005). 15. V. Chhaniwal, A. S. Singh, R. A. Leitgeb, B. Javidi, and A. Anand, Opt. Lett. 37, 5127 (2012). 16. M. Daneshpanah, S. Zwick, F. Schaal, M. Warber, B. Javidi, and W. Osten, J. Display Technol. 6, 490 (2010). 17. K. Taguchi, M. Tanaka, and M. Ikeda, Opt. Rev. 8, 156 (2001). 18. O. Otto, C. Gutsche, F. Kremer, and U. F. Keyser, Rev. Sci. Instrum. 79, 023710 (2008). 19. F. Gittes and C. F. Schmidt, Methods Cell Biol. 55, 129 (1998). 20. P. M. Hansen, I. M. Tolic-Norrelykke, H. Flyvbjerg, and K. Berg-Sorensen, Comput. Phys. Commun. 174, 518 (2006). 21. D. L. Nelson and M. M. Cox, Lehninger Principles of Biochemistry, 3rd ed. (Worth, 2000). 22. S. C. Grover, R. C. Gauthier, and A. G. Skirtach, Opt. Express 7, 533 (2000). 23. A. Anand, V. K. Chhaniwal, and B. Javidi, J. Display Technol. 6, 500 (2010). 24. Y. Kim, K. Kim, and Y. Park, in Measurement Techniques for Red Blood Cell Deformability: Recent Advances, Blood Cell—An Overview of Studies in Hematology (INTECH, 2012), pp. 167–194. 25. K. Lee, K. Kim, J. Jung, J. Heo, S. Cho, S. Lee, G. Chang, Y. Jo, H. Park, and Y. Park, Sensors 13, 4170 (2013). 26. F. Merola, L. Miccio, P. Memmolo, G. Di Caprio, A. Galli, R. Puglisi, D. Balduzzi, G. Coppola, P. Netti, and P. Ferraro, Lab Chip 13, 4512 (2013).
