I am having trouble with my glass coverslips for primary neuronal cultures. After troubleshooting, I've narrowed it down to the coating. which keeps detaching at a specific step. Here is my prep process:
Coat coverslips in poly-l-lysine, 1 mg/mL overnight
It's a completely different approach, but you could give "Matrigel (Growth Factor Reduced)" (Corning), or "Geltrex" (Gibco) a try. I am using Matrigel for iPSC derived neurons, because I always had problems with other coatings (PLL, Ornithine, Laminin) and wasted months of my time, trying to optimize my coating procedure, before testing this stuff. Matrigel is quite expensive, but you can get at least 500 ml extracellular matrix out of 1 bottle of Matrigel (1:50 dilution), which would be sufficient for 2.000 12mm coverslips (250 µl coating volume).
A colleague is using Laminin/Poly-D-Lysine precoated coverslips (Corning) for primary neuronal cultures, but these are even more expensive...
Some manufacturers leave a very thin film of "grease" on glass coverslips. Whether this is by design (e.g. to keep coverslips apart and easier to separate) is not entirely clear. Acid washing your glass coverslips can help to improve downstream processes, and a colleague had good results with acid washing coverslips before coating with poly ornithine/laminin for her neuronal cultures. Might be worth a try!
I don't know offhand which acid wash protocol she used, but a quick search suggested this one from Cold Spring Harbour, which should be a good starting point. http://cshprotocols.cshlp.org/content/2008/5/pdb.prot4988.long