This looks like what we need, but it's behind a paywall.
Turns out my lab library has a subscription that gets me behind the paywall.
Being Belgian brewers, the authors are actually looking to
optimize the production of ferulic acid (FA) as a phenol precursor. They investigate the impact of mash temperature, pH, duration, thickness, grist coarseness and composition, and stirring regime.
Temperature: Free FA production peaks at 40C/104F before dropping 70% to a flat minimum above 60C/140F.
pH: Free FA production peaks at pH 5.8, dropping by 25% in the pH 5.2-5.6 range of most interest to us.
Duration: A 30 minute mash reduced the free FA production by 25%; a 2+ hour mash increased it by 50%.
Thickness: Free FA production was maximized at medium density (quoted as 10 degrees Plato) dropping by up to 10% at higher or lower densities.
Grist coarseness: Free FA production was reduced by 25% when changing the grinding disk separation from 0.1mm to 2mm.
Composition: Baseline results were from pure pilsner malt; replacing 50% of the grain bill with adjuncts typically halved free FA production, except with wheat and rye, and Vienna and Munich malts.
Stirring: Continuous stirring increased free FA production by 50% compared with an unstirred mash.
Their conclusions were that free FA could be maximized by including a low temperature rest (40C) during the mash, keeping the pH relatively high (5.8), and stirring a lot.
They then set up a full-scale test brew and measured free FA concentrations throughout. Note that in the figure below the scale starts at 1.5ppm, somewhat disguising the fact that most of the production still comes from step (a) - the initial low-T rest - with the remainder coming from the full duration of the mash - to step (d). Sparging reduces the concentration but not the total content, while boiling increases the concentration but reduces the content due to thermal decarboxylation. Finally hop additions increase free FA by ~10%.
If you're interested in the full paper, pm me.