Question 2 of 6: A2 — Spin-coating film-thinning law
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
Paper format. EGBC 04-CHEM-B1 Transport Phenomena, May 2016, 3 hours, open-book. Six problems in three sections (A Fluid Mechanics, B Heat Transfer, C Mass Transfer); candidates attempt one from each section plus a fourth (four of six at 25 marks each). All six problems are solved below as a complete study resource. Appendix A of the paper supplies the equations of change (continuity, Navier–Stokes, energy and species tables) in rectangular, cylindrical and spherical coordinates — these are quoted rather than re-derived.
Reference texts: R. B. Bird, W. E. Stewart & E. N. Lightfoot, Transport Phenomena (2nd ed., Wiley) — the equations of change and the differential momentum / energy / species balances; J. R. Welty, C. E. Wicks, R. E. Wilson & G. L. Rorrer, Fundamentals of Momentum, Heat and Mass Transfer (Wiley) — pipe friction, internal-flow heat transfer and boundary-layer mass transfer; F. P. Incropera & D. P. DeWitt, Fundamentals of Heat and Mass Transfer (Wiley) — the Dittus–Boelter correlation and constant-heat-flux internal flow; R. H. Perry & D. W. Green, Perry’s Chemical Engineers’ Handbook (9th ed.) — transport properties; T. B. Reddy & standard metallurgical mass-transfer literature (Eisenberg, Tobias & Wilke, J. Electrochem. Soc. 1954) — the rotating-cylinder correlation.
Question 2: A2 — Spin-coating film-thinning law (25 marks)
Given. A liquid film of thickness $h(r,t)$ on a wafer spinning at constant angular velocity $\omega$; liquid density $\rho$, viscosity $\mu$, kinematic viscosity $\nu=\mu/\rho$; the film is thin so lubrication (thin-film) scaling applies — radial pressure gradients are negligible and the free surface is shear-free.
Find. (a) $h(t)$ from an initial thickness $h_0$; (b) the long-time (very-thin-film) limiting form.
Fig. A2: Spin coating. Centrifugal body force $\rho\omega^2r$ drives a radial outflow whose depth-integrated flux thins the film; a radial momentum balance plus a thin-film mass balance gives the thinning law $dh/dt\propto-h^3$.
Approach. Balance the centrifugal body force against viscous shear to get the radial velocity profile, integrate it for the outward flux per unit circumference, then close a thin-film continuity equation to obtain an ODE for $h(t)$; integrate and take the long-time limit.
Radial momentum (lubrication) balance. With the free surface shear-free and pressure gradients negligible, the centrifugal body force is balanced by viscous shear: $\mu\dfrac{\partial^2 u_r}{\partial z^2}=-\rho\omega^2 r,$ where $z$ is measured up from the wafer.
Velocity profile. Integrating with no-slip at the wafer $u_r(0)=0$ and zero shear at the free surface $\partial u_r/\partial z|_{z=h}=0$: $$u_r(r,z)=\frac{\rho\omega^2 r}{\mu}\left(hz-\frac{z^2}{2}\right).$$
Outward flux per unit circumference. $q=\displaystyle\int_0^h u_r\,dz=\frac{\rho\omega^2 r}{\mu}\left(\frac{h^3}{2}-\frac{h^3}{6}\right)=\frac{\rho\omega^2 r\,h^3}{3\mu}=\frac{\omega^2 r\,h^3}{3\nu}.$
Thin-film continuity. Conservation of liquid on the axisymmetric film gives $\dfrac{\partial h}{\partial t}+\dfrac1r\dfrac{\partial (r\,q)}{\partial r}=0.$ Substituting $q$ and noting that a spatially-uniform film has $\partial h/\partial r=0$ so $\tfrac1r\partial_r(r\,q)=\tfrac{2\omega^2h^3}{3\nu}$: $$\boxed{\dfrac{\partial h}{\partial t}=-\frac{2\omega^2}{3\nu}\,h^3}.$$
Integrate the thinning ODE. Separating $-h^{-3}dh=\tfrac{2\omega^2}{3\nu}dt$ from $h_0$ at $t=0$: $\dfrac{1}{2h^2}-\dfrac{1}{2h_0^2}=\dfrac{2\omega^2}{3\nu}t$, i.e. $$\boxed{h(t)=\frac{h_0}{\sqrt{\,1+\dfrac{4\omega^2 h_0^2}{3\nu}\,t\,}}}\qquad\text{(part a)}.$$
Very-thin-film (long-time) limit. When $\dfrac{4\omega^2 h_0^2}{3\nu}t\gg1$ the "1" and hence $h_0$ drop out: $$\boxed{h(t)\to\sqrt{\frac{3\nu}{4\omega^2 t}}=\frac12\sqrt{\frac{3\nu}{\omega^2 t}}}\qquad\text{(part b)}.$$ The thickness becomes independent of the initial amount deposited.
Quantity
Result
Thinning ODE
$dh/dt=-(2\omega^2/3\nu)\,h^3$
(a) Thickness vs. time
$h(t)=h_0\big/\sqrt{1+(4\omega^2h_0^2/3\nu)\,t}$
(b) Very-thin-film limit
$h(t)\to\tfrac12\sqrt{3\nu/(\omega^2 t)}$ (independent of $h_0$)