What is the surface roughness of your substrate?

The smoothness of a surface may reflect that of the substrate or the surface tension of the film–air interface. Poly(acrylic acid) is polar and able to hydrogen bond so it is likely to have a high surface tension or surface free energy. If the film was annealed (heat treated above glass transition) in a high humidity atmosphere, hence lowering the relative surface tension between the film–air and moisture plasticising the poly(acrylic acid), then it may become smoother. This assumes that you do not want to add a plasticiser or surface additive such as a partially-miscible semi-polar polysiloxane, to permanently modify the  poly(acrylic acid), which should give a smoother surface. Many products are available to make polymer films (such as paint) smoother, for example Du Pont additives.

I make this suggestion because I used to work developing coatings and some coatings were smooth and glossy while others would develop wrinkles, called orange peel. Coatings that had longer alkyl chains or that were oil based tended to be smoother than highly polar polymer based coatings. Polyethylenes with oxidised functional groups (after corona treatment) decrease polarity in dry air over about a week as the polar groups migrate into the surface, giving a lower and more stable surface free energy. If then exposed to moisture the polar groups migrate again to the surface.

I dont know your polymers but you can use UV-curable polymerization method. Because evaporation causes surface roughness as a result you cannot get smooth surface. In UV-curable method all ingredients are reactive, no evaporation and you obtain smooth surface.

I hope this will halp you

Dear Jensen,  substrate I am using has a roughness of around 4-5 angstrom maximum.

Two thoughts: 4-5 nm is extremely thin and some believe as a matter of principle that a film this thin may not be continuous over a Si wafer. The "hills" on the wafer would not be covered in PAA, even though the ellipsometer says "4.975 nm" or whatever. phase contrast afm would be useful in this case. Also, I assume your wafers are clean? If you open and close the boat often they will become dirtied by ambient greases quickly and won't be as easy to spin on.

Second thought: you do not mention solvent, although this is the most important choice when trying to spin coat smooth films. For PAA I would suggest 1-butanol off the top of my head, although references may have better choices. One time I obtained a thin PAA film by using it in amine salt form. A colleague dissolved it in trimethylamine/water (50/50) solution, precipitated it in THF and vacuumed dry. This form could be spin coated from dilute methanol solution into very thin and smooth films.

I do not anticipate thermal or solvent annealing applied after spin coating to make your film smoother. Another thin you could try is spin coating a very thick film of PAA, and baking it on a hot plate long enough to drive a grafting reaction between PAA and wafer surface. Then you could rinse unbound polymer off in good solvent and be left with a nice monolayer. Since PAA could form anhydrides at high temperatures this may not actually be feasible.

What is the molecular diameter of PA? Ploymers typically have larger than 20nm coil diameters. If this is the case it may not be possible to have a roughness measured in angstroms.