Photoinitiators are indispensable components in the formulation of inks used for printing packaging materials. Residual photoinitiators on the surface of packaging materials may migrate and contaminate the contents， posing potential risks to human health. Toxicological experiments have shown that photoinitiators pose various risks， including carcinogenicity， reproductive toxicity， and dermal toxicity. However， current quality standards for pharmaceutical packaging materials， both domestically and internationally， do not impose limits on photoinitiators. This study focuses on 12 photoinitiators， which are either restricted or prohibited in certain fields such as food packaging materials. The selected medicinal composite membranes， widely employed in pharmaceutical packaging and requiring significant ink volumes， serve as representative samples. This research holds significant implications for improving the quality standards of pharmaceutical packaging materials and ensuring medication safety. While gas chromatography-mass spectrometry and ultra performance liquid chromatography-tandem mass spectrometry have been reported for photoinitiators detection， UPC²-MS/MS has not yet been reported for detecting these compounds. The principle of UPC² is based on supercritical fluid chromatography， where the mobile phase primarily consists of supercritical CO₂ with a minimal use of organic solvents， aligning with the trend of green chemistry. When coupled with MS， this method also enhances the sensitivity of detection. Therefore， a novel method based on UPC²-MS/MS was established for the determination of the 12 photoinitiator residues in medicinal composite membranes. MS conditions， makeup solution conditions （solvent， additive type， additive ratio， flow rate）， and sample pretreatments were systematically optimized. Photoinitiator residues were extracted from the medicinal composite membranes using acetonitrile by ultrasonication. The chromatographic separation was achieved on a UPC² CSH^TM Fluoro-Phenyl column （100 mm × 3.0 mm， 1.7 μm） under gradient elution using a mobile phase composed of supercritical carbon dioxide （A） and methanol （B）. Gradient elution was performed as follows： 0-1.5 min， 100%A-95%A； 1.5-2.0 min， 95%A-80%A； 2.0-3.0 min， 80%A-70%A； 3.0-4.0 min， 70%A-100%A； 4.0-5.0 min， 100%A. A methanol-water （99∶1， v/v） mixture was employed as the makeup solution to enhance MS response. The mobile phase flow rate， makeup solution flow rate， column temperature， and injection volume were respectively set at 1.5 mL/min， 0.2 mL/min， 50 ℃ and 1 μL. The automatic back pressure regulator maintained 13.79 MPa. The 12 photoinitiators were monitored in multiple reaction monitoring （MRM） mode with electrospray ionization （ESI） under positive conditions. Quantification was performed using the external standard method. The results showed that all 12 photoinitiators exhibited good linear relationships in the range of 0.1-2.0 μg/mL， with correlation coefficients （r）>0.995. The limits of detection （LODs） and quantification （LOQs） for the 12 photoinitiators were 0.03 μg/mL and 0.1 μg/mL， respectively. At 1LOQ， 2LOQ， and 10LOQ levels， the average recoveries of the 12 photoinitiators ranged from 80.7% to 119.7%， with relative standard deviations （RSDs） between 1.0% and 5.6%. Twelve batches of medicinal composite membranes were tested using this method， and six batches were found positive. The detected photoinitiators were 4-methylbenzophenone， methyl-2-benzoylbenzoate， benzophenone， and 2-isopropylthioxanthone. Among them， methyl-2-benzoylbenzoate showed the highest concentration， but it did not exceed the proposed limit. The established method is accurate， sensitive， rapid， environmentally friendly， and simple pretreatment， which is suitable for monitoring residual photoinitiators in medicinal composite membranes.