Sensors and Actuators B 154 (2011) 129–136 Contents lists available at ScienceDirect Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb A method for determining the mutual diffusion coefficient of molecular solutes based on surface plasmon resonance sensing F.C.C.L. Loureiro a , A.G.S. Barreto Neto a,b , C.S. Moreira a,c , A.M.N. Lima a,∗ , H. Neff a,d a Department of Electrical Engineering, Center for Electrical Engineering and Informatics, Universidade Federal de Campina Grande, Rua Aprígio Veloso 882, 58429-900 Campina Grande, PB, Brazil b IF-PB, Department of Electromechanics, Brazil c IF-AL, Department of Electronics, Brazil d Northeast Center for Strategic Research (CETENE), Systems and Integrated Circuits Laboratory (LINCS), Recife, PE, Brazil article info Article history: Available online 16 February 2010 Keywords: Microfluidic Diffusivity Surface plasmon resonance sensing Finite element simulation Molecular transport abstract An experimental method, combining surface plasmon resonance sensing and microfluidics, to determine the mutual diffusion coefficient of molecular solutes, as ethanol and bovine serum albumin, is presented. Representative refractive index variations of analyte samples over time, and associated dynamic solute concentration profiles, respectively, have been employed to access molecular transport parameters. From both, Fick’s diffusion length and Taylor’s pulse dispersion methods, solute and solvent mutual diffusion coefficients for diluted ethanol and concentrated protein aqueous solutions have been obtained. Addi- tionally, the dynamic behavior and geometry effects of molecular transport have been exploited using finite element simulations for the 3-dimensional case and confirmed experimentally. The numerical simulation also addresses the influence of temperature effects. © 2010 Elsevier B.V. All rights reserved. 1. Introduction The mutual diffusion coefficient or diffusivity is one of the most important materials transport parameters, and of relevance for many physicochemical and physiological processes, as well as for molecular weight determination [1]. Diffusion coefficients D of solutes in aqueous solution vary widely over several orders of magnitude, from 10 -5 cm 2 /s for small molecules, up to 10 -8 cm 2 /s for biochemical high molar weight solutes. Moreover, D is strongly affected by the chemical composition of the solvent, temperature and solute concentrations. For many materials, reliable quantita- tive data are not available, since measurements are technically demanding. The Taylor diffusive dispersion analysis [2] has been widely employed as the bases for experimental exploitation. Alternatively, radio-active tracer, NMR pulse gradient, chromatographic meth- ods, flow field flow fractionation and light scattering [3–6] have been reported to assess materials transport parameters in solu- tions. Recently, a T shaped microfluidic channel (T-sensor), in connection with epifluorescence microscopy as an optical parti- cle detection method, has been employed to determine diffusion coefficients of large and small molecules [7]. The quantitative method displayed high accuracy, combined with short measure- ∗ Corresponding author. Tel.: +55 8321011135; fax: +55 8321011418. E-mail address: amnlima@dee.ufcg.edu.br (A.M.N. Lima). ment time. Quantitative data were obtained by prediction based on a 1-dimensional (1-dim) model, with the diffusion length scal- ing as (2D t ) n , where t is the time variation and n = 0.5 the power relation. However, recent theoretical research work [8] indicated deviations from the ideal 1-dim Einstein approximation for microfluidic configurations, where the aforementioned scal- ing relation varied as 0.33 ≤ n ≤ 0.67 along the micro-channel extension. Experimental investigations of the scaling relations have been reported [9] by means of confocal fluorescence microscopy. Therefore, transverse diffusive broadening of low molar weight solutes in two-phase laminar flows across the interface between two solutions in micro-channels, at high Peclet numbers, have been analyzed, revealing variations 0.33 ≤ n ≤ 0.5. In this work, a parallel plate microfluidic device, where a sec- tion of one plate is an integral part of a surface plasmon resonance (SPR) sensor, has been explored. The present label-free method relies on monitoring the dynamic evolution of the refractive index of an aqueous solution, while transported to, and passing through, a microfluidic cell. Generally, after leaving from the reservoir, and in absence solute adsorption effects, the recorded temporal refrac- tive index variations and associated response time t tot are largely governed by diffusive (t D ) and convective (t CV ) transport to and within the cell, approximately composed as t tot = t D + t CV . Nevertheless, upon applying appropriate experimental boundary conditions, both transport mechanisms can be clearly separated. The presented method comprises two different approaches: Firstly, an initially rectangular solute concentration profile, leaving 0925-4005/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2010.02.023