J. Chem. SOC.,Faraday Trans. 2, 1981,77, 1281-1291 Luminescence of Porphyrins and Metalloporphyrins Part 3.-Heavy-atom Effects BY ANTHONYHARRIMAN Davy Faraday Research Laboratory, The Royal Institution, 21Albemarle Street, London W1X 4BS Received 18th December, 1980 The fluorescence quantum yields and lifetimes for a series of metal tetraphenylporphyrins were found to decrease with increasing spin-orbital coupling constant of the central metal ion. The triplet states gave similar effects at both 77 and 293 K, as shown by luminescence and flash photolysis studies. In particular, the triplet-state lifetimes decreased in the order Mg> Zn >Pd =Cd >Hg and it was possible to relate the rate constants for non-radiative decay processes to the spin-orbital coupling constant of the metal ion although different metal ions interacted with the porphyrin .rr-system to a different extent.Previous papers'72 in this series have compared the luminescence properties of several first -row transi tion-me ta1 tetrap henylporp h yrins ;the nature of the central metal ion was found to influence both the ground-state absorption spectra and the lifetimes of the porphyrin excited singlet and triplet states. The most pronounced effects were induced by paramagnetic metal ions that interacted strongly with the porphyrin .rr-system (e.g. Cu") since this situation introduced some degree of relaxation into the spin-forbidden singlet-triplet transitions. Thus, the paramagnetic metalloporphyrins can be regarded2 as intramolecular examples of the well-known external paramagnetic effect where the introduction of paramagnetic species, such as 02,into the solvent results in enhancement of non-radiative tran~itions.~" In this paper, we investigate the influence of heavy-atom metal ions on the luminescence properties of rneso-tetraphenylporphyrinand, in addition, we have measured the corresponding room-temperature triplet-state absorption spectra and lifetimes and compared the observed trends with the luminescence data.Our interest in this field is twofold. First, the nature of metalloporphyrins allows intensive study of intramolecular effects on photophysical properties but, so far, such systems have received little investigation. Although Gouterman etal.' have made a detailed study of the influence of the central metal ion on the luminescence properties of Group IV porphyrins, this work was complicated by axial ligation. The present work involves a simplified system where the axial ligands can be ignored.Secondly, we are aware8-lo that there is a growing interest in the use of metallopor- phyrins as photosensitisers for the dissociation of water into H2 and 02.These systems are aimed at the storage of sunlight in the form of H2 fuel and it is an inherent necessity for such systems that the photosensitiser absorbs strongly throughout the visible region and possesses a relatively long excited-state lifetime. The diamagnetic metalloporphyrins studied here possess such properties and may have application for solar-energy devices.l1 EXPERIMENTAL MATERIALS Compounds were prepared and purified as described in previous papers. 1*2 Special attention was paid to the synthesis of HgTPP'2-'4 and the final product was recrystallised from 1281 1282 LUMINESCENCE OF PORPHYRINS AND METALLOPORPHYRINS dichloromethane +hexane. Methylcyclohexane (B.D.H.)was dried with molecular sieves and redistilled. METHODS Absorption spectra and luminescence spectra, lifetimes and yields were recorded as described previously, Singlet excited-state lifetimes were measured in our laboratory by Dr. V.Fidler using the picosecond time correlated single-photon counting technique. For all yield and lifetime measurements, outgassed dilute solutions were used and the reported values are the average of several independent measurements.The spectral responses of the instrument were corrected using the methods suggested by Argauer and White." Room-temperature triplet absorption spectra were recorded with conventional ,us(pulse duration lO~s,maximum energy 200 J) and frequency doubled Nd3+ ns (pulse duration 20 ns, maximum energy 100mJ) flash photolysis equipment. Again, all measurements were made with outgassed dilute solutions and, with the microsecond flash photolysis experiments, the photolysis lamps were filtered with aqueous K2Cr207 to remove light of A C 500 nm. Triplet- state absorption spectra were measured point-by-point and extinction coefficients were determined16 by the complete bleaching method and have an expected accuracy of *t15'/0.Spin-orbital coupling constants were obtained from literature sources. 17318Redox poten- tials refer to the first oxidation potential (us. SCE) of the metalloporphyrin measured in non-aqueous solution and were obtained from the literature. With the exception of HgTPP, the values used were those compiled by Kampas et al.19 The redox potential for HgTPP in CH2C12 solution was measured by cyclic voltammetry but should be regarded as a fairly crude estimate. RESULTS AND DISCUSSION ABSORPTION SPECTRA The absorption spectra of metalloporphyrins are dominated by intense (rr*) transitions which occur in the visible region. The most intense band is the B band which is situated between 400-470 nm and is the origin of the second (TT*)excited singlet state.The origin of the first (TT*)transition occurs between 550-650 nm and is termed Q(0,O). Both energy (E)and the oscillator strFFgth (f)of the Q(0,O)band depend upon the central metal ion and they can be used as a simple measure of the degree of interaction between the metal ion and the porphyrin n-system. Thus, with increased interaction the energy is raised and there is a reduction in the oscillator strength.20 Fig. 1 shows the relationship between energy and oscillator strength for the Q(0,O)transition of the metalloporphyrins used in this study. Fig. 1infers that there is very little interaction between metal ion and porphyrin r-system for HgTPP whilst for PdTPP the interaction is quite strong.The extent of interaction is controlled by a combination of several factors, including the size of the metal ion, the geometry of the metalloporphyrin and electrostatic and inductive effects but, even so, fig. 1 provides a useful fingerprint for comparison of the relative efficiency with which orbitals on the metal ion interact with the porphyrin n-system. FLUORESCENCE SPECTRA At room temperature, typical metalloporphyrin (~r*)fluorescence' was obser- ved for MgTPP, ZnTPP, CdTPP and PdTPP. The fluorescence spectra showed very small Stoke's shifts (ca. 150cm-') and good mirror symmetry with the lowest-energy absorption bands whilst there was a good match between the excitation spectrum and A. HARRIMAN 1283 15 16 17 18 19 E[Q(O,0)]/103 cm-' FIG.1.-Relationship between the energy (E)and the oscillator strength (f)of the Q(0,O) transition. the ground-state absorption spectrum. The fluorescence quantum yields (&) and excited singlet-state lifetimes (7s) recorded in outgassed methylcyclohexane solu- tion, exhibited the expected heavy-atom effect in that both yields and lifetimes decreased in the order Mg >Zn >Cd == Pd (table 1). In fact, the 7s and C$F values found for CdTPP and PdTPP were close to the limits of detection. Considerable difficulty was encountered in attempting to obtain a reproducible & for HgTPP. In methylcyclohexane solution, fluorescence was observed, with a maximum at 655 nm, but C$F depended upon the history of the sample. Quite different c$~values were found for different samples of HgTPP and, for a given sample, C$F increased with the age of the solution.The excitation spectrum did not match the ground-state absorption spectrum of HgTPP but showed a better agreement with the absorption spectrum of metal-free TPP. Thus, the observed fluorescence was probably due to TPP present as an impurity and we note that HgTPP is extremely sensitive towards acid catalysed demetallation. On the basis of an excitation spectrum, no fluorescence was observed that could be assigned to HgTPP and, from these studies, the maximum dF for HgTPP is That fluorescence from HgTPP should be weak is not too surprising in view of the TABLE 1.-FLUORESCENCE PROPERTIES OF SOME METAL TETRAPHENYLPOR-PHYRINS IN OUTGASSED METHYLCYCLOHEXANE AT ROOM TEMPERATURE compound 4F TJns k,/107 s-' kisc/108s-' MgTPP ZnTPP 0.15 0.03 9.2 2.7 1.6 1.7 0.9 3.5 CdTPP PdTPP 4x1Op4 2~ 10-~ 0.065 0.020 1.8 1.7 154 500 HgTPP 40-3 - 2.0 - 1284 LUMINESCENCE OF PORPHYRINS AND METALLOPORPHYRINS expected heavy-atom effect but it is rather disappointing to have to place the maximum & at such a high level.Interestingly, the rate constant for fluorescence (kF),calculated from the Strick- ler-Berg equation, remains constant for all the metalloporphyrins (table 1)and, if it is assumed that internal conversion from the first excited singlet state to the ground state is unimportant*' for these compounds, then the observed decreases in &and T~ can be attributed to enhanced intersystem-crossing.Based on this assumption, values for the rate constant for intersystem-crossing (kist) have been calculated (table 1). PHOSPHORESCENCE SPECTRA Low-temperature phosphorescence from the first excited (TT*)triplet state of MgTPP and ZnTPP has been described in a previous paper.' The phosphorescence quantum yields (t$p) and lifetimes (Q), measured in methylcyclohexane at 77 K, are collected in table 2. Also shown in the table are #p and T~ values for PdTPP, CdTPP TABLE2.-PHOSPHORESCENCE PROPERTIES OF SOME METAL TETRAPHENYLPOR- PHYRINS IN METHYLCYCLOHEXANE AT 77 K MgTPP ZnTPP 0.015 0.012 45 26 0.3 0.5 0.22 0.38 CdTPP 0.04 2.4 17 4.0 PdTPP 0.17 2.8 61 2.9 HgTPP 0.01 0.2 50 50.0 and HgTPP; the emission observed from these porphyrins (fig.2) was assigned to normal (TV*) phosphorescence since the metal ions are diamagnetic and the singlet-triplet energy gaps were in good agreement with those expected' for porphyrin (VV*) excited states. For all five compounds, there was a good match 650 700 750 800 850 900 950 h/nm FIG. 2.-Phosphorescence spectra of CdTPP (-), PdTPP (----) and HgTPP (....) recorded in methylcyclohexane at 77 K. A. HARRIMAN 1285 between the excitation spectrum and the ground-state absorption spectrum recorded in methylcyclohexane at 77 K and there was no spectral change upon further cooling. [The & observed' for TPP was extremely low (ca.lo-') and the phosphorescence maximum occurred at 865 nm so that the phosphorescence reported here for HgTPP seems unlikely to arise from a TPP impurity.] Table 2 also contains values for the rate constants for phosphorescence (kp)and non-radiative deactivation of the triplet state (kT) and both sets of values showed a pronounced heavy-atom effect.Unlike the other compounds, the triplet excited state of PdTPP exhibited luminescence in outgassed fluid solution at room temperature which was almost identical to that observed at 77 K and, as such, it was attributed to normal (.rr.rr*) phosphorescence.22 The 293 K spectrum was slightly broader than that found at 77 K and there was a small red shift (ca. 100 cm-') whilst T~at 293 K (T~385 ps)= was lower than the 77 K value but these effects are quite normal.TRIPLET-STATE ABSORPTION SPECTRA The triplet-triplet absorption spectra of the metalloporphyrins recorded in outgassed methylcyclohexane solution at room temperature are shown in fig. 3 whilst the absorption maxima and extinction coefficients (E), derived from these spectra, are collected in table 3. The spectra observed for the different metalloporphyrins are remarkably similar and show two intense bands in the visible region together with a series of relatively weak bands in the near-infrared. Of the two intense visible bands, the lowest-energy band was always the most intense, having a typical extinction coefficient of ca. 10sdm3rnol-'cm-', and was always broader than the higher- energy band. The ratio of extinction coefficients of these two visible bands was found to be quite sensitive towards the central metal ion.The absorption in the near- infrared region seemed to consist of at least three absor tion maxima, typical maximum extinction coefficients being ca. 5 x lo3dm3 mol- Pcm-', and were rela- tively unaffected by the nature of the central metal ion. With the exception of HgTPP, the transient absorption decayed by first-order kinetics, reforming the ground-state metalloporphyrin, and the decay rate constants (kD) are collected in table 3. At higher concentrations of porphyrin, the decay kinetics deviated from first-order owing to triplet-triplet interactions and this was particularly so for the longer-lived triplets such as MgTPP.For HgTPP, the decay TABLE 3.-TRIPLET ABSORPTION DATA FOR SOME METAL TETRAPHENYLPOR-PHYRINS IN OUTGASSED METHYLCYCLOHEXANE AT ROOM TEMPERATURE compound A/nm €/lo4dm3mol-' cm-' k,/103 s-' MgTPP 415 485 2.7 7.2 0.74 ZnTPP 400 3.8 0.83 470 7.1 CdTPP 415 2.2 3.77 490 5.7 PdTPP 385 3.0 2.63 450 4.7 HgTPP 420 495 2.4 8.6 30.9 1286 LUMINESCENCE OF PORPHYRINS AND METALLOPORPHYRINS x Y.-v, Q)-0 cd .IU a Q) .I c-z A Y .I -0-cd .I Ya 0) ._U-Q) I I 1 x U.I z 71 0 .I Y a Q) .-Y-2 x Y.-c -0-cd .I* a Q) .+* cd -2 x w .d E:v, -0 c 0 .IY a 0 .I Y- z I I I I I I LOO 500 600 7 00 800 900 A/nm FIG.3.-Triplet absorption spectra of the metalloporphyrins recorded in outgassed methylcyclohexane at room temperature.(a)MgTPP; (6) ZnTPP; (c)CdTPP; (d)PdTPP; (e)HgTPP. A. HARRIMAN 1287 kinetics was best analysed in the form of two exponentials. The minor species was quite long lived (kD =r lo3s-’) and was attributed to the triplet state of metal-free TPP whilst the major component was much shorter lived (kD= lo5s-l) and was assigned to the triplet state of HgTPP. Upon aeration of the solutions, similar spectra were observed but the triplet lifetimes were reduced to <200 ns. The spectra and lifetimes observed for ZnTPP and PdTPP were in good agreement with those rep~rted~~-~~ by previous workers but the room-temperature triplet-state properties of the remaining metallopor- phyrins have not been subjected to a previous systematic study.Despite an absence of detailed experimental results, there have been several attempts to describe a theory for triplet-triplet absorption spectra of p~rphyrins.~’-’~ Gouterman2’ has developed a theory that considers the upper state of the triplet spectra as arising from either a singly excited configuration or a doubly excited one. Detailed calculations on the doubly excited configuration have predicted2’ two allowed groups of states in the region of 400-500nm. The two groups were predicted to have equal intensities whilst the total intensity was expected to be approximately half that of the singlet-singlet spectrum.The calculations also predicted a weakly forbidden band further to the red. In general, these-predictions have been borne out by the observed triplet spectra, at least for the visible regione2’ Although no detailed calculations have been reported for the singly excited configuration, it is assumed that these states give rise to the absorption maxima observed in the near-infrared. The triplet spectra reported here show some interesting aspects regarding the influence of the central metal ion upon the intensity and position of the two visible bands that should enable some refinement of the above theory. Thus, as the degree of interaction between metal ion and porphyrin 7.r-system is increased (as shown by fig.l),the energy of both visible transitions increases also. In addition, there is a general relationship between the relative intensities of these two visible bands and the extent of interaction between metal ion and porphyrin n-system. As the interaction increases, the extinction coefficient of the higher-energy band increases relative to that of the lower-energy visible band. These effects are very similar to those noted for the ground-state absorption spectrum, although not so marked, and suggest that the same type of metal-porphyrin interactions prevail in both states. The absorption bands observed in the near-infrared region of the triplet spectra do not appear to be influenced to any real extent by the nature of the central metal ion.This finding is consistent with their assignment as singly excited configurations. DEACTIVATION OF THE EXCITED STATES It is well-established that, for reasons of symmetry, spin-orbital coupling is exceptionally weak in planar n-electron systems built-up exclusively from 2p, atomic orbitals.28 However, upon insertion of a transition-metal ion into the centre of the porphyrin ring a new spin-orbital coupling pathway is opened, as first discussed by Ake and G~uterman.’~ This mechanism derives from the conjugation of the metal d,,, d,, T-type atomic orbitals with the first pair of antibonding p* orbitals on the porphyrin nucleus to which the electron is promoted upon e~citation.~’ For heavy- metal porphyrins, this mechanism should be efficient and the spin-orbital coupling should be dominated by the one-centre contribution that comes from the central metal ion.The effects of substitution of a heavy-metal ion or a paramagnetic ion into an aromatic nucleus have been well-documented but the increase in the rate constant 1288 LUMINESCENCE OF PORPHYRINS AND METALLOPORPHYRINS for non-radiative decay from the excited state (kNR)is the most characteristic28 and can be used as a measure of the relative effectiveness of different metal ions as perturbers of spin-orbital coupling. The interpretation of such effects has usually followed an empirical approach in which the magnitude of changes in kNR, obtained from luminescence studies, have been correlated with the atomic spin-orbital coupling constant (l)of the substituted atom.This empirical approach has been most successful28 and has led to formulation of a relationship between kNR and J. Thus, if the triplet state (TI)mixes only with the perturbing singlet (Sp) and if it is known that the substituent does not affect the nature nor energy of the excited states, then kNR for a given homologous series of aromatic molecules may be expressed where B is the same constant for all members of the series. If l for one particular centre outweighs all the others and if only two atomic orbitals on that centre have much significance then we may write (Spl~lTI) CAcBf (2) where CAand CBare atomic-orbital coefficients. It follows that kNR B(c*c,o~- (3) Normally, as with a series of compounds such as the halogenonaphthalenes, it is assumed that both B and the orbital coefficients CAand CBremain constant throughout the series3' and hence kNR can be related to J, but this is not the case with metalloporphyrins. It is quite clear from the data presented in this paper that the central metal ion affects the nature and the energy of the excited states of the metalloporphyrin. As such, we would expect that for a series of metal TPP complexes both B and CAand CBwill change.Furthermore, spectroscopic studies have shown2' that the different sublevels of the lowest-energy triplet state couple to a different extent with the singlet manifold and, therefore, kNRshould be more correctly written as kgk where i refers to a particular sublevel T,, T,, or T,.For Mg and many metal-free porphyrin~,~~ it has been found that the T, and T,,sublevels are much more active for non-radiative transitions than the T, sublevel. This is not the case when the central metal ion is heavy32 and for Zn porphyrins the T, sublevels are more coupled to the singlets and are the dominant sublevels as regards non-radiative decay. Thus, for a series of metalloporphyrins eqn (3) is too simple to express the influence of the central metal ion upon the photophysical properties. Instead we can write where Z=[XEoxCACB/AE (5) and E O is the redox potential of the metalloporphyrin, which is a measure of the extent of charge transfer between metal and porphyrin, and AE is the difference in energy between the initial and final states.In order to use this equation, it is necessary to select one of the metalloporphyrins as a standard and compare kNR for the others against this reference. Although MgTPP exhibits the least amount of spin-orbital coupling, as shown by the luminescence properties, the most appro- priate standard is ZnTPP since for all the metalloporphyrins studied here, with the exception of MgTPP, we can assume that non-radiative transitions involve only the T, sublevel of the lowest-energy triplet state. In addition, we have shown previously A. HARRIMAN 1289 that it is possible to correct kNRfor variations in the energy gap between initial and final states so that we may rewrite eqn (4) as kNR = kkRZz (6) and ZO= 6 X E X CACB.(7) Although 6 can be obtained from atomic spectroscopy and E O can be measured by cyclic voltammetry, before we can make a quantitative evaluation of eqn (7) we require a numerical estimate for CACB and we have made the assumption that the CACBterm is equal to the energy of the absorption transition, either the ground-state Q(0,O) transition or the lowest-energy visible band in the triplet spectrum, so that CACB =EQ(O,O) or Em. (8) Now, using the above assumptions, we can relate Zofor a particular metallopor- phyrin to Zofor ZnTPP and compare calculated and observed Zo values. First, we have applied eqn (6) to the observed fluorescence properties of the metalloporphyrins (table 1) where kNRrefers to the rate constant for intersystem- crossing to the triplet manifold (kist).The energy gap between excited singlet and triplet states remains constant (ca. 4000 cm-') so that k& refers to kist for ZnTPP and, as such, the observed Zovalues are collected in table 4. Also shown in table 4 TABLE 4.-HEAVY-ATOM ENHANCEMENT OF INTERSYSTEM-CROSSING Z$etal)/ZfW compound 1/103 cm-' E"/V obs. calc. ZnTPP 1.0 0.71 1.0 1.o CdTPP 1.7 0.63 6.6 1.49 PdTPP 1.46 1.03 11.9 2.32 HgTPP 5.0 0.55 - 3.62 are the calculated Zo terms and, whilst the data are extremely limited, there is a reasonable correlation between the observed and calculated 20values. On the basis of this correlation, we have estimated that kiscfor HgTPP should be in the region of ca. 1x 10" s-'.Secondly, we have applied eqn (6) to the deactivation of the excited triplet states of the metalloporphyrins at 293 and at 77 K. At 293 K kNR refers to kD,as measured by flash photolysis, and at 77 K kNRrefers to kT, obtained from the luminescence measurements. Since the energy of the triplet state depends upon the nature of the central metal ion, it was necessary to correct k& for variations in the energy gap. This correction was achieved using the procedure described earlier' and involved measurement of the triplet lifetimes for a series of Mg porphyrins throughout which the triplet energy varied. A linear relationship was found' between kNR and the triplet energy and we have assumed that Zn porphyrins follow identical behaviour.Thus, extrapolation of the data allowed estimationpf kNRfor a Zn porphyrin at any particular triplet energy and so values for kD and kT for CdTPP, PdTPP and HgTPP were obtained (table 5). Also shown in table 5 are the observed and calculated Zo values and, again, there is a reasonable correlation between experimental and calculated data at both temperatures. 1290 LUMINESCENCE OF PORPHYRINS AND METALLOPORPHYRINS TABLE5.-HEAVY-ATOM ENHANCEMENT OF DEACTIVATION OF THE TRIPLET STATE kTO kDO Zbmetal)/Zfh) compound /S-l /lo3s-l obs. (77 K) obs. (293 K) calc. (293 K) ZnTPP 38 0.83 1.o 1.o 1.o CdTPP 79 1.74 2.25 1.47 1.49 PdTPP 11 0.24 5.19 ,3.31 2.32 HgTPP 72 1.59 8.33 4.41 3.62 Despite the many assumptions and simplifications that have been made during this treatment, it appears that the non-radiative deactivation processes for this series of diamagnetic metalloporphyrins can be expressed in the form of eqn (6).As yet, the data are restricted to only a few compounds and it is hoped that more experimental results will be available in the near future and it should then be possible to carry out a more elaborate verification of eqn (6). However, from the data that are available (tables 4 and 5)it is clear that the heavy-atom effect is most pronounced for intersystem-crossing from the excited singlet to triplet state. Non-radiative deac- tivation of the triplet state in fluid solution at room temperature is relatively insensitive to the heavy-atom effect. Finally, we can comment on the use of the above compounds as photosensitisers for solar-energy storage devices.Previous work','' has shown that water-soluble ZnTPP derivatives are very efficient sensitisers for the photoreduction of water to hydrogen but, before such systems can be used in practical devices, it is necessary that the quantum yield for formation of hydrogen is increased. Since the active photo- sensitiser in these systems is the triplet state of the Zn porphyrin, it is a reasonable assumption that improvements in the yield of hydrogen can be made by increasing the yield of the triplet state. This can be achieved by replacing Zn with a heavier metal ion and if Cd or Pd ions are used the triplet lifetimes at room temperature are not shortened too drastically.Of the two metal ions, Cd seems to offer the most promise since the absorption spectrum of a Cd porphyrin is always red-shifted relative to the corresponding Pd porphyrin and this allows collection of a higher fraction of the solar spectrum." Consequently, work is now in progress to evaluate water-soluble Cd porphyrins as photosensitisers for the production of hydrogen. I thank the S.R.C.,the E.E.C. and G.E. (Schenectady) for financial support and I am greatly indebted to Prof. Sir George Porter, F.R.S. for many helpful discussions. A. Harriman, J. Chem. SOC., Faraday Trans. 1, 1980,76, 1978. * A. Harriman, J. Chem. SOC., Faraday Trans. 1, 1981, 77, 369. D. F. Evans, Nature (London), 1956,178, 534. G. Porter and M. R. Wright, Discuss. Faraday SOC.,1959, 27.G. J. Hoytink, Mol. Phys., 1960,7, 67. J. N. Murrell, Mol. Phys., 1960, 7, 319. M. Gouterman, F. P. Schwarz, P. D. Smith and D. Dolphin, J. Chem. Phys., 1973,59, 676. K. Kalyanasundaram and M. Gratzel, Helv. Chim. Acra, 1980, 63, 478. A. Harriman and M. C. Richoux, J. Photochem., 1980,14, 253. 10 A. Harriman, G. Porter and M-C. Richoux, J. Chem. SOC., Faraday Trans. 2, 1981, 77, 1175. 11 A. Harriman, G. Porter and M-C. Richoux, J. Chem. SOC., Faraday Trans. 2, 1981,77, 833. 12 G. D. Dorough, J. R. Miller and F. M. Huennekens, J. Am. Chem. SOC., 1951,73,4315. A. HARRIMAN 1291 13 M. F. Hudson and K. M. Smith, Tetrahedron, 1976, 32, 597. 14 M. F. Hudson and K. M. Smith, Tetrahedron Lett., 1974, 2223. 15 R. J.Argauer and C. E. White, Anal. Chem., 1964,36, 368. 16 G. Porter and M. W. Windsor, Proc. R. SOC. London, 1958,245,238.17 C. E. Moore, Circ. U.S. Natl. Bur. Stand., 1949-1958, 1-111. 18 Quelques Probltmes de Chimie Mine'rale, ed. R. Stoops (Dixibme Conseil de Chimie, University of Brussels, 1956).19 F. J. Kampas, K. Yamashita and J. Fajer, Nature (London), 1980, 284, 40. 2o M. Gouterman, The Porphyrins, ed. D. Dolphin (Academic Press, New York, 1978),vol. 3, p. 1. 21 P. G. Bowers and G. Porter, Proc. R. Soc. London, 1967,296,435.22 M. P. Tsvirko, V. V. Sapunov and K. N. Solovev, Opt. Spectrosc., 1973, 34, 1094. 23 L. Pekkarinen and H. Linschitz, J. Am. Chem. Soc., 1960,82,2407.24 A. Harriman, N. Searle and G. Porter, J. Chem. Soc., Faraday Trans. 2, 1979, 75, 1515. 25 M. Gouterman, J. Chem. Phys., 1960, 33, 1523. 26 M. Gouterman, J. Mol. Spectrosc., 1961, 6, 138. 27 D. Madge, M. W. Windsor, D. Holten and M. Gouterman, Chem. Phys. Lett., 1974, 29, 183. 28 S. P. McGlynn, T. Azumi and M. Kinoshita, The Molecular Spectroscopy of the Triplet State (Prentice-Hall, New York, 1969).29 R. L. Ake and M. Gouterman, Theor. Chim. Acta, 1969, 15, 20. 30 S. P. McGlynn, R. Sunseri and N. Christodouleas, J. Chem. Phys., 1962, 37, 1818. 31 R. H. Clark, R. E. Connors, T. J. Schaafsma, J. F. Kleibeuker and R. J. Platenkamp, J.Am, Chem. Soc., 1976,98, 3674. 32 G. W. Canters and J. H. Van der Waals, The Porphyrins, ed. D. Dolphin (Academic Press, New York, 1968), vol. 3, p. 531. (PAPER 0/1952)