Colloid and adsorption physics are a very complicated business and often have a lot of surprises. In 1991 Kramarcy and Sealock published a paper in JHC Vol. 39, No. 1,pp. 37-39, 1991: “Commercial Preparations of Colloidal Gold-Antibody Complexes Frequently Contain Free Active Antibody”. Their data indicate that proteins adsorbed onto colloidal particles of 5nm and larger can dissociate from the particle surface with time and that, at times even shortly after manufacturing, colloidal gold reagents may contain free binding molecules. This is not necessarily the result of bad manufacturing practice, as adsorption and desorption are in equilibrium at all times. Some proteins (there are even variations between antibodies from different animal species) are more liable to become dissociated than others, and the conditions of coupling play a role as well.