Small unilamellar phospholipid vesicles (SUVs) are widely used as a model system in which to study the molecular interactions which occur in biological membranes. Israelachvili et al. and Mitchell and Ninham argue that provided the phospholipid concentration is sufficiently low, lipid aggregates will spontaneously form into SUVs which represent the equilibrium state of the lipid in water. Helfrich, Fromherz and Lasic adopt an alternative model and picture the phospholipid vesicle as a distortion of a planar membrane which requires energy for its formation and is inherently unstable.

In the present work we examine these two approaches and their consequences in explaining the physical properties of SUVs. A quantitative basis for our discussion has been devised by adapting a recently published data set of Parsegian et al. to permit an approximate estimate of the curvature energy associated with vesicle formation. We conclude that the spontaneous vesicle formation model only applies to a limited class of molecules where the electrostatic term is the dominant free-energy contribution which drives the bilayer to a more curved geometry. In the case of phosphatidylcholine we propose that the flexibility of the zwitterion eliminates this term causing the planar geometry to be the preferred state. Based on our calculation of curvature energy we are able to predict the diameter of SUVs of phosphatidylcholines to within 10% of the experimental value and to account for the effects of lysolecithin and cholesterol on the size of such vesicles.