electroosmotic flow and dispersion in microfluidics
play

Electroosmotic flow and dispersion in microfluidics Sandip Ghosal - PowerPoint PPT Presentation

IMA Tutorial: Mathematics of Microfluidic Transport Phenomena December 5-6, 2009 Electroosmotic flow and dispersion in microfluidics Sandip Ghosal Associate Professor Mechanical Engineering Department Northwestern University, Evanston, IL, USA


  1. IMA Tutorial: Mathematics of Microfluidic Transport Phenomena December 5-6, 2009 Electroosmotic flow and dispersion in microfluidics Sandip Ghosal Associate Professor Mechanical Engineering Department Northwestern University, Evanston, IL, USA E-mail: s-ghosal@northwestern.edu http://www.mech.northwestern.edu/fac/ghosal

  2. A kitchen sink (literally!) experiment that shows the effect of electrostatic forces on hydrodynamics Courtesy: Prof. J. Santiago’s kitchen

  3. On small scales things are different! 2 R body forces ~ R 3 interfacial forces ~ R 2 interfacial charge dominates at small R 3

  4. Electroosmosis through porous media FLOW Charged Debye Layers E Reuss, F.F. (1809) Proc. Imperial Soc. Naturalists of Moscow

  5. Electroosmosis v Debye E Layer ~10 nm Substrate = electric potential here v = m eo E Electroosmotic mobility

  6. Electrophoresis Debye Layer of counter ions + + + + + v - Ze + + + + + + E v = m ep E Electrophoretic mobility

  7. Equilibrium Debye Layers (Neutral) φ ( z ) is the mean field (l << L ) c i = c i 0 exp( - z i e f / k B T ) λ 2 f ε z 2 = - r e (Poisson) ¥ ρ e = c i z i e i + + + + + + + Gouy-Chapman Model Co-ion (+) Counter-ion (-)

  8. λ D ~ 0.3 nm z If in GC model, x = z i e f / k B T << 1 exp( x ) 1 + x then Debye-Huckel Model 2 f z 2 + f 2 = 0 l D f = z exp( - z / l D ) λ D ¥ z i 2 = e k B T 2 e 2 ) - 1 l D 4 p ( c i 0 φ + + + + + + + ζ i Co-ion (+) (zeta potential) For 1M KCl Counter-ion (-)

  9. Thin Debye Layer (TDL) Limit 2 u 2 f ε µ z z 2 = - 4 pr e z 2 + r e E = 0 � � 2 z 2 u - e E f � � = 0 u ( z ) � � 4 pm φ (0) = z & u (0) = 0 u ( z ) = e E ( f - z ) 4 pm Debye Layer u ( ) = - e E z φ (0) = z 4 pm (Helmholtz-Smoluchowski slip BC)

  10. Electroosmotic Speed E 100 micron 10 nm 2 u µ z 2 = 0 u e η / 2) = υ ε = − εζ Ε u ( z = 4 πµ 10 nm u ( z ) = u e

  11. Slab Gel Electrophoresis (SGE)

  12. Light from UV source Sample Injection Port Sample (Analyte) UV detector Buffer (fixed pH) + -- CAPILLARY ZONE ELECTROPHORESIS

  13. Capillary Zone Electrophoresis (CZE) Fundamentals c ( x , t ) u eo x u = u eo + u ep 2 c c t + u c x = D x 2 σ 2 ~ 2 Dt = 2 DL / u = 2 DL /[( m eo + m ep ) E ] N ~ L 2 / s 2 = ( m eo + m ep ) V /(2 D ) dm ~ m / N ~ 30 kV ) N ~ 10 6 (for V Ideal capillary

  14. “Anomalous dispersion” mechanisms In practice, N is always LESS than this “ideal” (diffusion limited) value. Why? • Joule heating • Curved channels • Wall adsorption of analytes • Sample over loading • ……….

  15. Non uniform zeta-potentials u e z is reduced Pressure Gradient + = Corrected Flow Continuity requirement induces a pressure gradient which distorts the flow profile

  16. What is “Taylor Dispersion” ? D eff = D + a 2 u max 2 192 D G.I. Taylor, 1953, Proc. Royal Soc. A, 219 , 186 Aka “Taylor-Aris dispersion” or “Shear-induced dispersion”

  17. Zone Broadening by Taylor Dispersion A B Clean CE Time Delay “Dirty” CE Resolution Degraded Signal Weakened

  18. Parabolic profile due to induced pressure Experiment using Caged Fluorescence Technique - Sandia Labs E Detection Caged Dye EOF suppressed Laser sheet (activation)

  19. Mathematical Modeling (I) The Flow Problem : what does the flow profile look like in a micro capillary with non-uniformly charged walls? (II) The Transport Problem : what is the time evolution of a sample zone in such a non-uniform but steady EOF? (III) The Coupled Problem : same as (II) but the EOF is unsteady; it is altered continuously as the sample coats the capillary. 19

  20. (I) The Flow Problem 20

  21. Formulation (Thin Debye Layer) y a x z L

  22. Slowly Varying Channels (Lubrication Limit) y x a z L Asymptotic Expansion in

  23. Lubrication Solution From solvability conditions on the next higher order equations: F is a constant (Electric Flux) Q is a constant (Volume Flux)

  24. Lubrication Theory in cylindrical capillary Boundary conditions Solution distance: a 0 velocity: u e = − ες 0 Ε 0 4 πµ Ghosal, S., J. Fluid Mech., 2002, 459 , 103-128 Anderson, J.L. & Idol, W.K. Chem. Eng. Commun., 1985, 38 , 93-106

  25. The Experiments of Towns & Regnier Towns J. & Regnier F. Anal. Chem. 64 , 2473 (1992) Experiment 1 Protein + EOF Mesityl Oxide 100 cm Detector 3 Detector 2 Detector 1 (85 cm) (50 cm) (20 cm)

  26. Understanding elution time delays t ζ = 0 ζ = 1 x L x L z = ( L - x ).1 + 0. x dx = 1 - x dt = u = £ L L x = L (1 - e - t / L ) t + L (at small times)

  27. Application: Elution Time Delays - + x = X ( t ) ζ = ζ 0 + e - a x ( z 1 - z 0 ) ζ = ζ 0 dX dt = u = - e < z > E = F ( X ) 4 pm

  28. Best fit of theory to TR data Ghosal, Anal. Chem., 2002, 74, 771-775

  29. Electroosmotic flow with variations in zeta Anderson & Idol Ajdari Ghosal ( Lubrication Theory ) Geometry Cylindrical Plane symmetry Parallel Amplitude Small Wavelength Long Variable zeta zeta,gap zeta,gap Chem. Eng. Comm. Phys. Rev. Lett. Vol. J. Fluid Mech. Vol. 459 2002 Reference Vol. 38 1985 75 1995 Phys. Rev. E Vol. 53 1996

  30. (II) The Transport Problem 30

  31. The Experiments of Towns & Regnier Towns J. & Regnier F. Anal. Chem. 64 , 2473 (1992) Experiment 2 300 V/cm (fixed) 15 cm M.O. _ + PEI 200 100 cm Detector remove zeta potential

  32. Taylor Dispersion in Experiment 2 EOF X

  33. Experiment 2: determining the parameters ζ 1 = - 0.326 z 0

  34. Diffusivity of Mesityl Oxide WILKE-CHANG FORMULA

  35. Theory vs. Experiment Ghosal, S., Anal. Chem., 2002, 74 , 4198-4203

  36. (II) The Coupled Problem 36

  37. CZE with wall interactions in round capillary 2 a 0 ζ = ζ 0 s ( x , t ) c ( r , x , t ) (on wall) (in solution) u e = - ez 0 E 4 pm Length = a 0 Re = u e a 0 / n (less than 1) Time = a 0 / u e (greater than 10) Pe = u e a 0 / D

  38. Flow+Transport Equations

  39. Method of strained co-ordinates T = e t , X = e x φ = φ ( r , X , T )

  40. ( a 0 x << 1) Asymptotic Solution Dynamics controlled by slow variables S.Ghosal JFM 2003 491 285 S.Datta & S.Ghosal Phy. of Fluids (2008) 20 012103

  41. DNS vs. Theory

  42. DNS vs. Theory Shariff, K. & Ghosal S. (2004) Analytica Chimica Acta, 507 , 87-93

  43. Eluted peaks in CE signals Reproduced from: Towns, J.K. & Regnier, F.E. “Impact of Polycation Adsorption on Efficiency and Electroosmotically Driven Transport in Capillary Electrophoresis” Anal. Chem. 1992, 64 , pg.2473-2478.

  44. QuickTime and a YUV420 codec decompressor are needed to see this picture. QuickTime and a YUV420 codec decompressor are needed to see this picture. 44

  45. QuickTime and a YUV420 codec decompressor are needed to see this picture. QuickTime and a YUV420 codec decompressor are needed to see this picture. 45

  46. QuickTime and a YUV420 codec decompressor are needed to see this picture. QuickTime and a YUV420 codec decompressor are needed to see this picture. 46

  47. Summary • Problem of EOF in a channel of general geometry was discussed in the lubrication approximation. • Full analytical solution requires only a knowledge of the Green’s function for the cross-sectional shape. • In the case of circular capillaries, the lubrication theory approach can explain experimental data on dispersion in CE. • The coupled “hydro-chemical” equations were solved using asymptotic methods for an analyte that adsorbs to channel. walls and alters its zeta potential. http://www.mendeley.com/profiles/sandip-ghosal/

Recommend


More recommend