Einstein (1905) derived an expression for the diffusion coefficient of
an isolated spherical colloid. Since that work, there have been two g
eneral methods for analyzing Brownian diffusion in colloidal suspensio
ns at finite concentrations. One follows Einstein's thermodynamic argu
ment postulating a gradient in chemical potential as being the driving
force behind diffusion together with a thermodynamic analysis of sedi
mentation-diffusion equilibrium (Batchelor, 1976). The other approache
s diffusion in a statistical fashion, deriving the macroscopic diffusi
on coefficient from a microscopic analysis of Brownian motion (Felderh
of, 1978). In principle, both methods are correct and should give iden
tical results, but the distinctly different approaches have produced s
ome controversy. To test the various theories, of Brownian diffusion,
experiments were conducted measuring the sedimentation and Brownian di
ffusion coefficients of uncharged rigid spheres. Sterically stabilized
silica spheres dispersed in cyclohexane were used as a model colloid.
The osmotic compressibility of this system was found to be well descr
ibed by the Carnahan-Starling equation for hard spheres. The sedimenta
tion coefficient of the silica spheres was measured over a wide range
of concentration in a closed bottom container. A light extinction meth
od was used to monitor the fall speed of the interface that develops d
uring gravity sedimentation. The diffusion measurements were made usin
g Taylor's hydrodynamic stability method. A laser optical-fiber system
capable of direct monitoring of the penetration depth and concentrati
on profile of the diffusing species along the diffusion column was dev
eloped. The measurements were found to be in fair agreement with Batch
elor's theoretical results for sedimentation and Brownian diffusion of
hard spheres.