Bill,

Gold particles (a.k.a. gold sol) will adsorb antibodies without the need for
any fancy coupling. Such antibody coated sols are used in many diagnostic
tests (e.g. the ept pregnancy test sold in the USA).  Planar surfaces, like
your gold coated cantilevers will also strongly adsorb antibodies. When I
link antibodies to sol I simply mix sol (typically 30nm in diameter at about
10e11 sol particles per mL with an antibody at about 20 micrograms/mL in a
pH 8.5 buffer of low ionic strength (so as not to cause the sol to
aggregate) - 10mM sodium tetraborate works fine. Leave for 10 minutes then
block any bare areas of gold with a cheap protein (bovine serum albumin
(BSA)at 1mg/ml works fine). One coated the particles are colloidally
stable - in fact the Science Museum in London still has some of Michael
Faraday's protein coated sols in glass vial from the the 1890s I think.

For your cantilevers I would do likewise, contact with a 10 - 100
microgram/ml antibody solution in buffer then contact with a protein
blocking solution (1mg/ml BSA)

Notes:
Only about 1-5% of your antibody will be functional, but don't worry about
this because you have to get really clever with the linking chemistry to get
higher specific activities and even then you will probably not get the 20
fold plus improvement you might naively expect. (The Biacore SPR instrument
uses some nice chemistry - dextran hydrogel linked via a thio-C16-epoxy to
gold, but if you are trying to generate a force when the antibody binds its
target then this may not be so good as the hydrogel will probably expand
normal to the gold surface rather than parallel to the gold surface)

Your antibody should be free of other adsorbing species (i.e. other proteins
and detergents)

Protein adsorption (i.e. mass of adsorbed protein per unit area) will
increase with protein concentration and reach a plateau adsorption above
which increasing the protein concentration will not increase the surface
concentration of the adsorbed protein.

You should a higher affinity of the protein for the surface around the
isoelectric point of the protein, at this pH the protein has its lowest
solubility and so has a greater tendency to adsorb to the surface.

Variables I would explore in coating the cantilevers are
Cleaning - you do not want to adsorb protein to any loosely attached dirt
'because it may come off  and will lead to irreproducible antibody coatings
Antibody concentration (10-100 micrograms per ml is a good range)
pH of buffer - work within 2 pH units of the isolectric point (ask supplier
if they know this - use a monoclonal antibody as the i.e.p. will be better
defined)
Ionic strength - vary the NaCl concentration
Blocking protein - BSA is commonly used, beta casein (re-crystallised is
often better but more expensive).

Hope the above helps. There is loads of literature on this area in the
scientific domain.

My background is in attaching proteins to surfaces for use in biosensors, I
currently work for a diagnostic company in the UK and if you have an
interesting application for your cantilevers and you wish maybe we could
collaborate.

Best of luck with your endeavours.

Bob Davies.