That's a common problem when using NaOH. Generally, I've found using dioxane/water instead of THF/water can go some way to stopping the phases from separating.
Even better, I just found an experimental detail on Beilstein which uses H2O/ethylene glycol as the solvent. They actually make the NaOH themselves since it's a radiolabelled version, but other than that I would try replicating this exactly as it says.
Sodium metal (25 mg, 1.09 mmol) was added to H218O (0.5 mL, 25 mmol) in ethylene glycol (2 mL, twice distilled under vacuum from sodium). Diphenylacetonitrile (193 mg, 1.00 mmol) was added, and the resulting mixture was heated at 150 °C for 17 h. To the cooled mixture was added aqueous HCl (0.2 M, 30 mL) and the solution was extracted with ether. The combined organic layers were dried with MgSO4 and the solvent was removed, leaving the acid as white solid (205.7 mg, 0.96 mmol, 95%).
Grotjahn et al, J. Am. Chem. Soc, 2001, 123, 8260-8270