I
I
G I
P
li
Figure 1. Block diagram of Instrumental system. A = 10-15 V, 7-21.5 A, d.c. Power Supply (Varian Illuminator Power Supply PS 300-1, Varian,
San Carlos, Calif.). B = Eimac 150-W Xenon Arc Lamp (Varian-Eimac Division, San Carlos, Calif.). C = Quartz lens. D = Mechanical Chopper (Model 382B, Ithaco, Ithaca, N.Y.). E = Optional Monochromator (Model H10, J. Y. Optical Systems, Metuchen, N.J.). F, G = Excitation and emission monochromators (Aminco-BowmanSPF, American Instruments Co., Silver Spring, Md.). H = Photomultiplier tube and housing (1P21, Hamamatsu Corp. Middlesex, N.J.). I = High Voltage Power Supply (240 Regulated High Vottage Power Supply, Keithley Instruments, Cleveland, Ohio). J = Preamplifier (Model 164 Preamplifier, Ithaco, Ithaca, N.Y.). K = Lock-in amplifier (Dynatrac 391A, Ithaco, Ithaca, N.Y.). L = X-Y Recorder (Plotamatic 715, MFE Corp., Salem, N.H.). S = Sample compartment (same as reference 1) matrix made necessary spinning of the capillary cell which introduced additional noise and random phase shifts and prevented correct phase measurement which limited the accuracy and precision of lifetime and quantitative measurements. Similar measurements (in clear glasses and snowy matrices) were attempted for morphine and codeine, but the results were not nearly as encouraging. A solution of morphine (400 ppm) gave no appreciable phosphorescence signal, while a solution of codeine (750 ppm) gave a weak signal. In these cases, where emission was very weak, increased amplification led to an enhanced phosphorescence signal as well as noise level. In addition, spectral interferences from stray and scatter light, which could not be phased out, contributed relatively much more to the noise than in the benzophenone and 4-bromobiphenyl cases above. A second monochromator in series was placed between the light source and sample to reduce stray light in the system. In this case, benzophenone and 4-bromobiphenyl were still easily detectable with the limit of detection of benzophenone being raised to =1 ppb. No significant increase in the signal-to-noise ratio of morphine or codeine was seen so that the reduction in stray light was insufficient to allow suitable detection of these drugs.
CONCLUSIONS These observations do not dispute those of Mousa and Winefordner ( I ) because the phase resolution technique apparently works in the case of strong phosphors or concentrated solutions of poorer phosphors if no fluorescence or stray light interference is present. I t also appears to be successful in resolving binary mixtures ( I ) of highly phosphorescent species if the difference between the phase shifts Ba and OB of molecules A and B is of a certain magnitude or greater. In the theoretical treatments ( I ) , stray light and fluorescence were assumed to be negligible. In practice, however, these two factors are usually present unless only strongly phosphorescent species are involved or unless one uses more sophisticated instrumentation to minimize these interferences which would limit the analytical use of this technique to rather restricted conditions. Also, in phase resolved phosphorimetry, molecules with short phosphorescent lifetimes have enhanced response over those with long phosphorescence lifetimes; this situation is desirable when the background phosphorescence from the solvent is of a long-lived nature. However, an inherent disadvantage results because only phosphorescent molecules with lifetimes between 1-50 ms can be easily measured using phase resolved phosphorimetry (7). If the lifetime is too long (>50 ms), measurements must be made at a low frequency where bubbling noise of the liquid nitrogen and other low frequency noises make it difficult, if not impossible, to take accurate and precise data. If the lifetime is too short (