Robotic in situ stiffness cartography of InP membranes by dynamic force sensing Jean-Ochin Abrahamians 1,2 , Bruno Sauvet 1 , J´ erˆ ome Polesel-Maris 2 , R´ emy Braive 3,4 and St´ ephane R´ egnier 1 Abstract—Typical methods of measuring mechanical proper- ties at the micro-scale are destructive, and do not allow proper characterisation on resonant MEMS/NEMS. In this paper, a cartography of local stiffness variations on a suspended micro- membrane is established for the first time, by a tuning-fork- based dynamic force sensor inside a SEM. Experiments are conducted on InP membranes 200nm thin, using a 9-DoF nano- manipulation system, complemented with virtual reality and automation tools. Results provide stiffness values ranging from 0.6 to 3 N/m on a single sample. I. I NTRODUCTION As of now, the behaviour of micro- and nano-scale res- onators is not fully understood due to nonlinearities in their dynamics [1]. Research on the subject primarily consists in modelling and analysis. Reliable values from actual measure- ments of their mechanical properties are desired in order to provide the parameters to be used in simulations, confirm current suppositions concerning these small-scale dynamics, and predict their resonant behaviour. Typical measurement methods involve inherently destructive protocols which either rely on indenting samples, or applying enough pressure to deflect MEMS sensors. These methods can make the measurement itself unreliable [2]. Furthermore, they do not allow precise mapping of several points on a single sample, as the mechanical properties of the sample are potentially modified after each measurement. Suspended micro-membranes (Fig. 1) are especially frag- ile (Fig. 2). Hence, non-destructive measurement on these samples first requires non-destructive mechanical positioning with accurate nanometer-range resolution. Since accurate manipulation has to be conducted by nanometric steps, it is excessively time-consuming if entirely handled by unas- sisted human operators. Automating some of the repetitive positioning operations is useful in alleviating the workload. This semi-automation brings the duration of experiments down to achievable levels for research purposes, and is an important first step before full automation. Virtual reality is another practical tool which, in addition to facilitating manual operations and providing a representation of confined 1 Institut des Syst` emes Intelligents et de Robotique, Universit´ e Pierre et Marie Curie, CNRS UMR 7222, 4 Place Jussieu, 75005 Paris, France. {abrahamians, regnier}@isir.upmc.fr 2 CEA, IRAMIS, Service de Physique et Chimie des Surfaces et Inter- faces, F-91191 Gif-sur-Yvette, France. 3 LPN-CNRS, Laboratoire de Photonique et de Nanostructures, 91460 Marcoussis, France 4 Universit´ e Denis Diderot, 75205 Paris, France setups, can run simulations concurrent to the experiment. Developing these tools is part of the process towards enabling meticulous manipulation. Although the state of the art in micro-robotics now provides actuators which are accurate and dexterous enough to perform the required operations in terms of positioning and control, invasive measurement methods are unsuitable for use on micro-membranes. Therefore, the aim of this work is to demonstrate a proof of concept for a micro- scale stiffness cartography measurement method on fragile structures. Herein, the stiffness of a suspended InP membrane is locally measured by contact at several points of its surface, using a self-sensing quartz tuning fork probe controlled in frequency modulation. Experiments are conducted in situ through a robotic nanomanipulation system implemented in a scanning electron microscope. Section II summarises the state of the art in micro-robotics, and mechanical properties measuring, relevant to this study. Section III details the equipment and manipulation setup. Section IV describes the stiffness measurement method, and the experimental results thereby obtained. Fig. 1. Suspended membrane. Dimensions 10x20 μm, thickness 200 nm. II. STATE OF THE ART A. Micro-robotic manipulation Nowadays, micro-manipulation experiments are often per- formed under SEMs (scanning electron microscopes) rather than optical microscopes, partly because classical optical microsopy reaches its limits in resolution, but mainly because 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) November 3-7, 2013. Tokyo, Japan 978-1-4673-6358-7/13/$31.00 ©2013 IEEE 1016