Usually vesicles and liposomes are un-stable. But colloid chemists know a lot of tricks to make them stable for very long times; by sterical stabilization, adding electrolytes, etc. Think to many formulations commonly used. The true question is, thus, if the stability of such adducts is thermodynamic or not.  I  know cases where vesicular phases are put in a phase diagram; more correctly, it should be intended as a dominance diagram. An example is this; mix together two surfactant solutions of oppositely charged species, having the same concentration. The mother solutions are stable. When one surfactant is in excess (say 2/1) you get vesicles and they are stable. When the ratio is 1/1 you get a precipitate. But this can be redissolved by adding one of the two surfactants! The question is intriguing. Depending on mole ratios you have stable monophasic systems (the two micellar solutions), two vesicular areas and a two phase system, consisting of solvent and precipitate. Note that vesicles are quite poly-disperse; in words, such dispersions are not made by a single species.

It's a good question, Titus, and one that highlights differences between biological/toxicological scientists (dealing with concentrations, numbers, and molecules). and engineers (dealing with transport properties, masses, and discrete portions of matter).  To a physicist and this is inherent in the NIST definition of particle (SP 960-3 "Any condensed-phase tridimensional discontinuity in a dispersed system may generally be considered a particle; e.g., droplets in an emulsion or solids dispersed in a liquid. The term is normally used in reference to solid materials.  An aggregate may also be regarded as a particle") then anything different or a closed boundary in a system will define a particle.

Thus to a light scattering person any heterogeneity or closed boundary will lead to a refractive index differential and thus will produce scattering.  To such a person, a molecule in solution or a nitrogen molecule in air is a 'particle' as it's different from the environment around it and will scatter incident radiation as a result.  The debate of 'what is a particle?' continues...

I think it is not correct to think that the absense of the bulk phase generally means stability. The "all-interface" system would rather be meta-stable and tend to create "bulk" volumes at any location where the kinetic stabilization mechanisms fail for any reason, or am I wrong?

yes this is a main point. I was too focused on micro'emulsion' type systems, when I wrote 'all at interface'. In between I realized that the important point is surplus energy compared to situation in bulk, independently whether bulk phase is already present or will inevitably be formed later. This would be the case e.g. for tiny nanodots when they are composed only of atoms at the interface. To obtain a thermodynamically stable system like colloidal solution, boundary at molecular interface has to be resolved either by polymer solvation or by inclusion into association colloids or layers. Thermodynamic instability of  vesicles/liposomes should have its origin in high entropy.

Further points to consider?

Dear Titus,

Very interesting topic. As the number of liposomal products on the market increase, the issue of their stability and shelf-life gains more importance. You may find the following statements from literature useful:

Vesicles are hollow aggregates, composed of bilayers of amphiphilic molecules, dispersed into and filled with a liquid solvent. The thermodynamics and the kinetics aspects of the phenomena of formation of vesicles is discussed in: 

Adv Colloid Interface Sci. 2010 Dec 15;161(1-2):77-88. doi: 10.1016/j.cis.2009.11.004. Epub 2010 Jan 15.

Thermodynamics and kinetics of vesicles formation processes.

Guida V

*******

One of my favorite scientists, D.D. Lasic, wrote on this topic:

The spontancity of liposome formation and their thermodynamic stability are discussed. It is shown that in general liposomes do not form spontaneously and that some energy input is required to prepare them. This clearly illustrates that their free energy is not the minimum the system can achieve even though they can be made stable for prolonged periods of time. In the case of nonzero spontaneous curvature, however, the spontaneous formation of liposomes is possible and was already observed.

In: Journal of Colloid and Interface Science

Volume 140, Issue 1, November 1990, Pages 302-304

On the thermodynamic stability of liposomes

D.D. Lasic. 

*********

In our liposome / nanoliposome publications we mentioned:

Hydrated phospholipid molecules arrange themselves in the form of bilayer structures via Van-der Waals and hydrophilic/hydrophobic interactions. In this process, the hydrophilic head groups of the phospholipid molecules face the water phase while the hydrophobic region of each of the monolayers face each other in the middle of the membrane (Fig. 2). It should be noted that formation of liposomes and nanoliposomes is not a spontaneous process and sufficient energy must be put into the system to overcome an energy barrier (for detailed discussion see ref. (20)).

In other words, lipid vesicles are formed when phospholipids such as lecithin are placed in water and consequently form bilayer structures, once adequate amount of energy is supplied. Input of energy (e.g. in the form of sonication, homogenisation, heating, etc.) results in the arrangement of the lipid molecules, in the form of bilayer vesicles, to achieve a thermodynamic equilibrium in the aqueous phase. ...

In:

Nanoliposomes: preparation and analysis

MR Mozafari - Liposomes: Methods and Protocols, Volume 1: …, 2010 - Springer

'Input of energy (e.g. in the form of sonication, homogenisation, heating, etc.) results in the arrangement of the lipid molecules, in the form of bilayer vesicles, to achieve a thermodynamic equilibrium ? in the aqueous phase.'

Seems to me rather a form of kinetic 'stability'

Liposomes are formed from lamelar micelles by input of energy. In this way curvature of bi-layer spherical film is of higher energy then when being in lamelar form. That is why this association structure is thermodynamically unstable. It possesses kinetic stabilty due to small size na Brownian motion. (Same opinion as of Titus) .