We have used a random copolymer of deuterated styrene and p-hydroxystyrene [dPS((1-f))-r-PPHSf], where f is the molar fraction of PHS, to strengthen the weak phase boundary between PS and poly(2-vinylpyridine) (PVP), The fracture toughness (G(c)) of the phase boundary was measured using an asymmetric double cantilever beam method, and the areal chain density (Sigma) of copolymer at the phase boundary was determined by forward recoil spectrometry (FRES). The interfacial strength was extremely sensitive to the composition of the random copolymer with an optimum value found for f approximate to 0.03. The maximum measured fracture toughness was around 250 J/m(2) at f = 0.022, while no significant strengthening was observed at f = 0.01 and f = 0.066. Such a strong composition dependence of G(c) is in marked contrast to the case of dPS((1-f))-r-PVPf copolymers, for which the maximum strengthening is seen at f approximate to 0.5 with significant strengthening still observed at f = 0.4 and f = 0.6. The differences in the compositions for optimum strengthening and in the composition sensitivity of strengthening in these two cases are attributed to the hydrogen bonding between different PHS units and between PHS units and the PVP homopolymer as well as the absence of composition drift in our dPS((1-f))-r-PPHSf random copolymers. Using FRES, we determined the fracture mechanism to be chain pull-out for f = 0.01 and f = 0.066 over the entire range of Sigma probed. For copolymers with f = 0.022 and f= 0.045, a mixture of chain scission and chain pull-out prevailed at low areal chain densities while craze formation followed by craze breakdown was observed at higher areal chain densities.