As an upgrade to conventional photoacoustic tomography (PAT), pump-probe PAT (PP-PAT) exploits the transient excited-state absorption of photosensitizers to effectively eliminate interference from strong endogenous absorbers such as blood, enabling the accurate detection of target molecules. However, current PP-PAT techniques predominantly rely on photosensitizers with short triplet-state lifetimes, such as methylene blue (MB), restricting PP-PAT systems to relatively short pump-probe delay windows. Due to the Grüneisen relaxation effect, strongly absorbing background tissues can also generate weak, transient thermally induced photoacoustic signals that overlap within this narrow window, introducing artifacts and limiting molecular imaging specificity. To address this issue, we propose a physical strategy using a solid-state polyvinyl alcohol (PVA) matrix with high degrees of alcoholysis. PVA's dense 3D hydrogen-bonding network spatially confines MB molecules, suppressing non-radiative relaxation and extending their triplet-state lifetime for background-suppressed PP-PAT. Transient triplet differential (TTD) characterization and imaging were performed on free MB, PVA-MB composite films (~200 μm thick), pure hemoglobin (Hb), and Hb-MB mixtures using UV-Vis spectroscopy, fluorescence spectroscopy, and a PP-PAT system. Results show that the PVA matrix significantly prolonged the triplet lifetime of MB while preserving its ground-state absorption and excited-state transition absorption characteristics. As PVA alcoholysis increased (74.5%, 88%, and 99%), the triplet lifetime rose from 13.73 μs to 17.70 μs and 63.26 μs, respectively—over 40-fold longer than in pure water (1.57 μs). Compared to free MB, this solid-state microenvironment increased the optimal pump-probe delay by a factor of 4.17, achieving a 2.99-fold improvement in PP-PAT specificity under equivalent absorption conditions. This study elucidates how microenvironmental rigidity regulates photosensitizer behavior in PP-PAT, offering valuable guidance for developing highly specific PP-PAT approaches.