Ultrasonic investigation of Mesoporous Silicon   Julien Bustillo1, Jérôme Fortineau1, Gaël Gautier1, Marc Lethiecq1 julien.bustillo@univ-tours.fr 1Université François Rabelais de Tours, CNRS, CEA, INSA CVL, GREMAN UMR 7347, Tours, France Abstract: Ultrasonic characterization of mesoporous silicon is proposed to estimate the porosity and the thickness of porous silicon layer. Two sets of samples are formed in order to retrieve independently each parameter: one with similar porosity and another with similar thickness. Secondly, an immersion ultrasonic measurement technique is developed and a modeling based on material homogenization is used to simulate wave propagation through the porous layer. An optimization between simulated and experimental spectra, based on a genetic algorithm, allows porosity and thickness to be estimated. They are compared to destructive measurements and good estimation of porosity and thickness is observed, even when layer thickness is small compared to wavelength of the ultrasonic wave. Sample presentation Sample Current density (mA/cm²) Etching duration (min) PoSi thickness (µm) PoSi Porosity rate (%) A 42 30 115 39 B 20 81 37 C 10 41 36 D 80 97 E 45 114 <100> p-type silicon HF based solution Pulsed current electrochemical etching Pore diameter of 20 nm 2 sets of porous silicon Porosity measured by gravimetic method Thickness measured using optical microscope SEM Slice view Ultrasonic measurement method d Transducer Transducer Input Output tSi cSi ρSi tPoSi cPoSi ρPoSi Water d Sample Experimental and theoretical transmission spectra through sample 2 Modeling (a) and experimental setup (b) used for ultrasonic characterization. Data processing Wave propagation model Frequency domain analysis Optimisation based on Genetic Algorithm Non monotone mutation Good convergence behavior Porosity deduced from acoustical impedance One dimensional modeling Biot’s theory modeling Transfer matrix method Transmission coefficient calculation and measurement Parameter retrieval processing scheme Porosity and thickness estimations PoSi thickness (µm) PoSi Porosity rate (%) Sample A B C D E Destructive Method 115 81 41 39 42 36 Acoustical Method 121 37 44 Relative error (%) 6 10 5 3 Retrieved values compared to values measured by destructive methods Good agreement even for thin layers Maximum discrepancy of 10% for thickness estimation Maximum discrepancy of 6% for porosity estimation Simultaneous and non destructive estimation of porosity and thickness Conclusion: An ultrasonic method has been developed and applied to porous silicon characterization using an inverse problem solving. A model has been implemented to calculate theoretical transmission spectrum and an optimization has been performed using a genetic algorithm. A good agreement is observed between retrieved thicknesses and porosity rates and values obtained by destructive methods. Discrepancies are respectively of 10% and 6% for thickness and porosity. Therefore, this method seems promising to characterize mesoporous silicon layer made using electrochemical processes. PSST 2014 – March 9th-14th –Alicante