Non-contact Mesoscale Manipulation Using Laser Induced Convection Flows Emir Vela, C´ ecile Pacoret, Sylvain Bouchigny, St´ ephane R´ egnier, Klaus Rink and Arvid Bergander Abstract— Laser induced convection flows is a new and promising method to achieve better manipulation of mesoscale objects (above 1 μm and below 500 μm) in a liquid medium. The temperature gradient created by laser absorption generates natural and thermocapillary (or Marangoni) convection flows. These flows are used to perform the manipulation itself. In this paper, we demonstrate for the first time that large and heavy particles can be dragged using the Marangoni convection flows. Experiments based on these phenomena show that fast and accurate underwater micromanipulation of particles up to 280 μm is possible using only a convergent 1 480 nm laser beam. I. INTRODUCTION Today the challenge for micromanipulation are to design highly flexible and effective systems which could control particles of large variety of sizes, shapes, materials and weights. Many devices exist and are dedicated to one type of object and applications either with direct contact like grippers [1] and cantilevers [2], or without contact like electrophoresis [3], magnetic [4], optical tweezers [5], [6] and microfluidic systems [7]. Many different tools are needed because of scale effects: objects from a few nm to a few mm undergo different interactions. Inertia and gravity effects become negligible, while adhesive, capillary and surface forces arise at the μm- scale [8]. Under the frame of the GOLEM European project, the need for a system capable of driving mesoscale particles (1 μm to 1 mm) was critical. The aim is to study the assembly of bio-functionalised components for which, the platform should be able to safely manipulate objects of many sizes and kinds e.g. MEMS, μ-tools and biological parts. For such an application, non-contact techniques provide the most interesting characteristics. Especially, it has been shown [9], [10] that individual and group movements can be easily achieved with temperature control and a large workspace. A common approach in non-contact micromanipulation, that we first considered for the GOLEM project, is to use laser light radiation pressure to trap small particles of E. Vela is with the sensory interfaces laboratory (LIS) at the French Atomic Energy Commission (CEA), Fontenay-aux-Roses, FRANCE (www- list.cea.fr) emir.vela-saavedra@cea.fr C. Pacoret is with the sensory interfaces laboratory (CEA), Fontenay- aux-Roses, FRANCE and with the Institute of Intelligent and Robotic systems (ISIR), Universit´ e Pierre et Marie Curie, FRANCE (www.isir.fr) cecile.pacoret@cea.fr S. Bouchigny is with the sensory interfaces laboratory (CEA), Fontenay- aux-Roses, FRANCE sylvain.bouchigny@cea.fr S. R´ egnier is a professor at ISIR, Universit´ e Pierre et Marie Curie, FRANCE stephane.regnier@upmc.fr Klaus Rink and Arvid Bergander are with OCTAX Micro- science GmbH, GERMANY klaus.rink@octax.de, a.bergander@ieee.org good refractive index in the focal point of a convergent laser beam. This setup is called “optical tweezers” and it was first demonstrated by Ashkin [11]. Since then, optical tweezers have routinely achieved piconewton forces and nanometric precision in the fields of natural sciences. Also, holographic systems allow manipulation of hundreds of particles at the same time [12]. And lately, optical vortices have also generated group movements, and present original rotating capacities [13]. However, this technology has some limitations regarding the size (dimensions less than 20 μm), shape (spherical) and refractive index of the particles. Speci- fications of the GOLEM project do not fit in this technology. The energy required to displace the particles is much higher than radiation pressure can achieve (a few pN). We propose in the following section a novel laser micro- manipulation approach closer to microfluidic methods and we will prove the feasibility of the strategy adopted. With convection flows, large and heavy particles can be dragged and we can exploit the fact that forces rise with the particle sizes. In section III, simulations are presented. Setup and experiments will be detailed in section IV. A brief summary of our first exploration results will be proposed in section V, followed by our conclusions on the potential of this technique and the outlook for new developments for micromanipulation in section VI. II. CONVECTION FLOW MANIPULATION PRINCIPLE The basic principle of convection flow micromanipulation is shown in Fig. 1. A convergent laser beam is shot on the bottom of a Petri dish, increasing the temperature locally. A convection column rises and drags particles toward the laser beam at the bottom of the Petri dish. An opposite effect appears at the free surface: when the convection column reaches the surface, floating particles are dragged away from the laser beam. Two phenomena actually drive this current: natural con- vection flow and Marangoni convection flow. The first one results from the Archimedes lift force as water density decreases with temperature. The second occurs when a temperature gradient is created on the free surface. This gradient may be generated by the flow of hot liquid coming from the bottom by natural convection or directly by the laser radiation if it is not fully absorbed in the liquid before reaching the surface. The flows at the microscale are highly laminar, as 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems Acropolis Convention Center Nice, France, Sept, 22-26, 2008 978-1-4244-2058-2/08/$25.00 ©2008 IEEE. 913