Raman Spectroscopy of Alkali Metal-Ammonia Solutions
1753
Raman Spectroscopy of Alkali Metal-Ammonia Solutions Willie L. Smith and William H. Koehler” Department of Chemistry, Texas Christian University, Fort Worth, Texas 76729 (Received February 5, 1973) Publication costs assisted by The Robert A. Welch Foundation
A laser Raman spectrometer has been constructed which readily accepts a specially designed dewar assembly. After the resolution and polarization performance characteristics of the spectrometer were determined, polarization studies on liquid ammonia were undertaken. Depolarization ratios for the fundamental modes of ammonia are reported, and a doublet has been observed for the v2 band. Alkali metal-amM were studied. No scattering center attributed monia solutions in the concentration range 0-50 X to the solvated electron was found, nor were the positions of the bands altered by the presence of the solute in this concentration region. The intensity of the scattered radiation was found to decrease with increasing metal concentration and has been attributed to the absorbance of these solutions. The Raman effect in F centers is discussed and compared to alkali metal-ammonia solutions.
Introduction The nature of the species present in a very dilute metal-ammonia solution continues to be a matter of speculation. Numerous models have been proposed, but as yet no single theory is clearly superior. A discussion of the various models will not be attempted in this work because numerous review articles are avai1able.l Copeland, Kestner, and Jortner* have proposed a model for localized excess electron states in ammonia, and this model has a particular bearing on the present work. According to Copeland, et al., very dilute solutions may be described in terms of unassociated solvated cations and electrons. The electron is envisioned as existing in a cavity created by some number of preferentially oriented ammonia molecules. The proposed model predicts a totally symmetric vibration in the ground state with a frequency given by v = (1/2n)[K/p31’2 where K is 1/2(a2Et/aR2)R = R ~ and , p = Nm”, Et is the total energy of the ground state, R is the cavity radius, N is the number of solvent molecules in the primary solvation shell, and in”3 is the reduced mass of the ammonia molecule. Depending on the value of N and VO,the electronic energy of the quasifree electron state, the totally symmetric vibration is predicted to be between 25 and 60 cm -I. R ~ s c happlying ,~ the treatment of Klick and Schulman* to metal-ammonia solutions, has suggested that the symmetric “breathing” mode may be in the 400- to 700-cm-1 region. These results were obtained using the formulation for the frequency of the “breathing” mode which is given by
v g = (2hT/h )[W,,,(LT)/W,,,(HT>12 where vg is the frequency of the ground-state vibration, Wh/2 is the width at half-height of the absorption band, and LT and H T refer to the low- and high-temperature limits, respectively. Regardless of the model used to explain the observed properties of these solutions, the effect of the metal on the solvent structure is of primary importance. Several investigators have addressed themselves to this problem, but at
present the results are incomplete and appear contradictory. Beckman and P i t ~ e r using ,~ external reflection techniques, studied the intermediate and concentrated solutions in the infrared. Burow and Lagowski,6 using internal reflection techniques, addressed themselves to much the same problem. Rusch,? using transmission techniqu’es, studied the effect of metal concentration on the 3300cm-1 envelope of ammonia. The results of the aforementioned investigations are summarized in Table I. In addition to the effect of metal on the solvent sQucture, the assignment of the Raman bands of liquid ammonia is contradictory and still speculative. It is not the purpose of this work to review all the vibrational data reported on ammonia; however, the Raman data available on liquid ammonia are summarized in Table 11.8-16
Experimental Section A laser Raman spectrometer was constructed from three basic components: (1) Control Laser Corporation Ar+ laser, (2) Spex 1401 double monochromator, and (3) Solid State Research photon-counting detection system. The spectrometer utilized the conventional 90” geometry and is shown schematically in Figure 1. Radiation of appropriate frequency and intensity was focused on the sample cell (Figure 2) and the scattered radiation collected at right angles. Both the 4880- (50-350 mW) and the 4579-81. lines (maximum power 100 mW) were used in this work. The beam was focused to a diameter of approximately 0.2 mm at the sample cell. The scattered radiation was collected and collimated with a conventional camera lens (55 mm f / 2 ) , passed through an analyzer, and focused on the entrance slit with a lens compatible with the f number of the monochromator. Prior to the slit assembly, a dove prism was used to rotate the image 90” (this rendered the image compatible with the vertical entrance slit), and a calcite wedge scrambler was inserted to remove the polarization dependence of the monochromator. An SSRI photocounting detection system was employed with the Spex 1401 monochromator. This detection system incorporates a Bendix Channeltron photomultiplier tube which has a dark count of 10 counts/sec at room temperature. The Journal of Physical Chemistry, Voi. 77, No. 74, 1973
Billie L. Smith and William H. Koehler
1754 CELL DEWAR SUPPORT LENS (65 MY 1/21
CONTROL LASERCORP. A: LASER
ANALYZER FIELD LENS DOVE PRISM SCRAMBLER MANUAL SLIT ASSEMBLY
SSRI DETECTION SYSTEM CHANWELTROW PM TUBE
Figure 1. Laser Raman spectrometer.
Figure 2. Sample cell.
TABLE I: Infrared Spectroscopy of Metal-Ammonia Solutionsa Ref
Assignment
Solvent
Concentration
> 0.5) 1030-1050 3190 3370
5
(MPM v2
v1 v3
2 X
6
214 v1
v3
3220 3280 341 0
M Na
1.7 X
3220 3280 3410
3220 3280 341 0 M Li
1.0 X
lo-*M Na 1.3 X l o - ' M Na
1.0 X 1O-'MK
1.4 M Na
10.3 M Na
3240
3220 3300 341 0
3230 3230 341 0
341 0
1.0 X l o - ' M Li
lo-'
1.0 X
MK
31 70 3280 3380
3230 3280 3420
3220 3280 341 0
3240 3280 3400
5 X 10-3[Li]
5 X 10-3[K]
5 X 10-2[Li]
5 X 10-2[K]
31 58 3265 3429
31 56 3268 3426
3153 3250 3409
31 52 3252 341 3
-
~
7
2V4 Vi
v3 a Ail
31 55 3286 3453
-
~~~~
__-_
values in cm-'
TABLE I t : Raman Spectroscopy of Liquid Ammonia Date
Observed bands and assiqnments
1929 1930 1930 1936 1953 1954 1968 1970 1970 1972
1594(v4) 1580( ) 1624(V4)
1070 ( V 2 ) 1061 ( V Z ) 1060 f 5 1060 f 5
1641 ( ~ 4 ) 1 645(V4) 1634 f 5
3210( 3216( 3208( 3210( 3212(
) ) ) ) )
3218(2~4) 321 5 ( 2v4) 3206 ( v i ) 3218( ) 3215 f 2 3214 f 2
The Raman cell itself was little more than a U-tube. The extrusion permitted the closest approach of the scattering center to the collection lens within the confines of the dewar. This design also enabled the focal point of the collection lens to be set at the scattering center which resulted in good collimation. The dewar design permitted two passes of the laser when a mirror was placed at the exit window. The Journal of Physical Chemistry, Vol. 77, No. 14, 1973
3310( ) 3304 ( V i ) 3296(V I ) 3300( ) 3303 (Vi ) 33OO(V i ) 3301 ( V I ) 3296(2~4) 3303( ) 3301 f 2 3300 f 2
t, "C
Ref
-40 4-25
8 9 10 11 12
-40 4-25 3373 ( v 3 ) 3 3 8 4~ 3 ) 3363(~3) 3386( ~ 3 ) 3380 f 2 3377 f 2
-34 to -70 4-25 -55 4-40 to -80 -67 4-25
13 14 15 16
Present work
The solutions were prepared and studied in a modified version of the dewar assembly described by Quinn.17 Prior to solution preparation, the assembly was subjected to a cleaning procedure previously described.l* To prepare a solution, a quantity of ammonia was distilled off sodium and condensed in the insert. The volume was determined from the calibrated pipet and was known to within 2~1%. A known weight of high-purity metal (Alfa Inorganics,
1755
Raman Spectroscopy of Alkali Metal-Ammonia Solutions Inc.) was added using a winch assembly previously described.18 Temperature of the solutions was monitored with a thermistor which extended into the optical cell. Room temperature spectra were recorded with the ammonia in sealed tubes. In general, spectra were recorded without the analyzer in the fore optics and then with the analyzer set first parallel and then perpendicular. The spectra were recorded a t constant slit width (usually 300 p ) which corresponded to spectral band widths in the order of 4-6 cm-l. The spectrum was generally scanned at 10 cm-l/min for the first several hundred wave numbers then increased to 50 cm-l/min.
R curve A B
c
D
pure BH3
2.3 x 5.2 2.0 x
-67’
M
i! jj
Slits
C
300/300/300~
scan 50 cm-l/min
350 mW, 4880 A
Results afid Discussion Prior to initiating the metal-ammonia studies, the resolution and polarization performance of the spectrometer were determined. The 459-cm-l line of carbon tetrachlorii cm-1 ide was investigated and the three most prominent bands Figure 3. Raman spectra of sodium-ammonia solutions. occurred a t 456.1, 458.8, and 460.6 cm-l. These three bands arise from the isotopic distribution of chlorine and have been reported by HerzberglS at 455.1, 458.4, and 461.5 cm-l. 60 350 scanmW, slits 50 300/300/300 4880 A cm-l/min k No attempt was made in the present study to eliminate all the errors generally associated with quantitative polarization measurements. The design of the spectrometer 50 and cell assembly was predicted on obtaining Raman spectra of metal-ammonia solutions rather than absolute depolarization ratios. Consideration was given to such factors as monochromator polarization efficiency, image magnification, and cell design, but when necessary these parameters were compromised. For a detailed discussion concerning the effect of these parameters on polarization measurements, the reader is referred to several excellent papers on the subject.20,21 Carbon tetrachloride was used to determine the accuracy of the polarization measurements reported in this work. 1 saw 3300 l4UU Depolarization ratios of 0.79, 0.78, and 0.01 were deterii cm-1 mined for the 225-, 312-, and 459-cm-l bands, respectiveFigure 4. ( V I , u 2 , 2 u 4 ) bands of liquid ammonia. ly. Comparison of these results reveals that the depolarization ratios are higher than the predicted and experiratios for v 2 and v4 are probably the most quantitative to mental values previously r e p ~ r t e d . Although ~ ~ - ~ ~ not abdate. solute, the measurements were generally in error by no Typical Raman spectra of metal-liquid ammonia solumore than 10%. tions are shown in Figures 3, 6, and 7. The broad band ( i ~ The Raman spectrum of liquid ammonia was deter= 200 cm-1) observed in pure ammonia was not observed mined and salient features of a typical spectrum are in the metal solutions. Careful attention was given to the shown in Figures 3 and 4. With reference to Figure 3, a region 0-1000 cm-I because of the predicted “breathing” broad structureless band centered around 200 cm-1 is evimode vibration; however, no new scattering center was dent and is attributed to solvent-solvent interaction. Figobserved. In fact, the spectrum of pure ammonia was inure 4 shows bands a t 3380, 3301, and 3215 cm-1 which M solution in distinguishable from that of a 2.3 X have been assigned to v3, V I , and 2v4, respectively. Althe region 0-100 cm-l. The only essential difference bethough the v 3 assignment is well accepted, the assigntween the spectrum of pure ammonia and any solution in ments of the 2V4 and V I are not as certain. The bands obM ) was in intensithe concentration region (2-50 x served at 1060 and 1634 cm-I have been assigned as vz ty. As is evident in Figures 3, 6, and 7, the scattered inand u4, respectively. A relatively high resolution spectrum of the 1 0 6 0 - ~ m - ~ tensity decreases with increasing metal concentration. Figure 8 is a plot of band maxima as a function of metal ( V Z ) band reveals an unresolved doublet which is shown in concentration and reveals that in the concentration range Figure 5 . This doublet has been previously reported12 and 0-5 x 10-3 M there are no shifts in band positions. Alarises from inversion. Based on numerous runs, the best though the data are not shown, the 1060- and 1640-cm-1 estimate of the splitting is 10-15 cm-l. Attempts were bands were also unaffected by metal up to 5 x M. At made to resolve the remaining bands in the spectrum, but higher concentrations, the intensity of the v p and v4 bands no other doublet structure was found. Depolarization ratios are presented in Table 111, and inwas too low to be observed. It is perhaps important to spection reveals that the results are in qualitative agreeM) note that three concentrations (2.3, 5.2, and 20 X represent sequential addition of metal, whereas the other ment with those previously reported. The depolarization
>
The Journal of Physical Chemistry, Vol. 77, No. 14, 1973
1756
Billie L. Smith and William I-I. Koehler -67' C Slits 300/300/300 scan 50 crn-l/:in
350 mW. 4880 A
l6O0I
Figure 7.
solutions.
1 1000
1050
(VI,
w
v z , 2u4) bands of ammonia in sodium-ammonia
1150
1100
n cm-1
Figure 5. up band of liquid ammonia.
Slits 300/300/300 p scan 10 crn-l/mm
curve A B
8001
Pure N% 2.3
X
M
6; 60Ql ,
conesntra,ioa
01
metal
(.oles/llfer)
Figure 8. Band positions as a function of metal concentration.
Ii t-
m d
P
400(
c d
20
40
60
80
A cm-l
Figure 6. Raman spectrum of sodium-ammonia solutions.
three concentrations represent three different solution preparations. These results are in disagreement with the infrared work of Rusch7 who reported a shift in the positions of V I and v3 with increasing metal concentration (Table I). Some caution must be exercised in comparing the two sets of data because the band positions reported by Rusch were obtained from computer-resolved infrared spectra and the band positions reported in this work are based on unresolved Raman data; however, the spectra presented in Figure 7 are of sufficiently high resolution that very little uncertainty exists in band positions. The Journal of Physical Chemistry, Vol. 77,No. 74, 7973
The decrease in scattered intensity with increasing metal concentration is readily explained by considering the absorption spectrum of the solutions; the spectrum is asymmetric and tails into the visible region of spectrum. Using the optical data of Gold and a plot was made of ( A l l ) as a function of concentration at 5830 A. Absorbance/pathlength, A / l , was used rather than absorbance to eliminate the pathlength dependency; 5830 A corresponds to Au = 3300 cm-1. Figure 9 shows at least a qualitative relationship between absorbance and Raman scattering intensity. Using 4880-A excitation, Au =_. 3300 cm-1 corresponds to 5830 and the absorbance a t this wavelength increases with concentration. There is also appreciable absorption at 4880 A so that two effects are operative which tend to diminish the scattered intensity. First the excitation line energy is reduced due to absorption and secondly the scattered radiation is subject to absorption. At concentrations greater than M , the 4880-A line was completely absorbed and severe heating resulted. The spectrum of the 5.5 x 10-3 M solution was obtained using 4580-A excitation. Although the power of this line was less than the 4880-A line, the decreased absorption at 4580 A, the u4 dependence, and the decreased absorption at 5390 (Au = 3300 cm-l) resulted in a meaningful spectrum. Attempts to study concentrations greater than 5.5 x M with the 4580-A line were unsuccessful
1757
Raman Spectroscopy of Alkali Metal-Ammonia Solutions TABLE I l l :
Polarization Studies of Liquid Ammonia Ref
1060 c m - ’
1640 cm-’
8 9 14 15
Present work
50% polarization p = 0.15
3215 c m - ‘
3300 c m
-’
3380 cm-’
Polarized Polarized
Polarized Polarized
Depolarized Depolarized
p = 0.08
p = 0.08
p = 0.4
>95%
>95%
0% p
=:
0.5
p
= 0.01
p
-
0
= 0.01
p
= 0.5
due to absorption of both the excitation and emission frequencies. Unfortunately it is impossible to state absolutely whether the predicted scattering center is or is not present, Arguments can be raised that the concentration of these centers was too low to be detected. Increasing the concentration is of little value because of the corresponding loss in intensity and the certainty of association in the solutions. Since the analogy is often made between F centers and metal-ammonia solutions and since much of the theoretical treatment of these solutions has as its origin F center theory, it is interesting to extend the treatment of the Raman effect in F centers to metal-ammonia solutions. Kleinrnan26 has calculated the Raman polarizability tensor using the expression
where aR is the polarizability tensor, f the oscillator strength, wGF the angular frequency of the electronic absorption, wo the angular frequency of the excitation energy, M the reduced mass, oG the angular frequency of the lattice vibration, and X G and XF the positions of the energy minimum of the ground and excited states, respectively. p is the “absorption-emission discrepancy” as shown in the relationship
( + w ~ m ( x ) ’ + a=~P ~ ( + F ~ M G F ( X ) ~ P F J when and $Fl represent the lowest vibrational states associated with the ground electronic state and the first excited electronic state, and MGF(X) is the matrix element of the dipole moment operator between electronic states G and F. Kleinman calculated LYR = 9 X cm3 for an F center using He-Ne excitation. This value is very close to the value of aR =- 2 X lO-z5 cm3 reported27 for the strong lines of carbon tetrachloride. Worlock and PortoZ8 reported observing Raman scattering from F centers in KC1 and NaC1. The Raman line was broad (-200 cm-l) and centered at AV = 200 cm-1. In these experiments the concentration of scattering centers was approximately 1016-1017 centers/cm3. Radhakrishna and Sehgal29 reported similar spectra of F centers in KC1 and KBr at densities of l0l8 centers/cm3. Kleinman’s equation was applied to metal-ammonia solutions using the following data: f = 0.77; wGF = 1.26 x (1.5 b); coo = 3.68 X 1Ol5 (4880 hi); fi varies from 1 to 10; M = 1.7 >( g (six NH3 molecules); wG = 6.4 x 10l2 (35 cm-1); X c = 2.2 hi; XF = 2.6 8.The following result was obtained If p = 1, then CXR = 0, but as 0 becomes large 0 1 ap~ proaches 6 X as a limit. Unfortunately, little is known about the excited state and consequently fi is
waVe180gth
ti)
Figure 9. Absorbance and Raman excitation/emission.
indeterminate, but unless is very small LYR is of the same magnitude as that obtained for F centers. It should also be pointed out that concentrations in the range 5 x 10-4-5 x M correspond to 3 X 10f7-3 X 10l8 electrons/cma (assuming no association). In view of these results it is perhaps surprising that no Raman effect was observed in the metal-ammonia solutions; however, caution must be exercised in carrying the analogy too far. Several factors could be operative which could reduce the scattering below detectable limits. The interactions in this system may be much less localized than in F centers which would produce broad and continuous Raman shifts. Certainly the difference in lifetimes of the centers must be considered because F centers are very long lived compared to the lifetime of a given cavity (as deduced from nuclear magnetic resonance relaxation times).30 The relative ease of reforming cavities coupled with the low cavity concentration may reduce the concentration of cavities a t any instant below detectable limits. The high radiation density is another factor whose effect is difficult to predict. Localized heating might possibly reduce the concentration of cavities through density effects or even destroy cavities. In this context it should be mentioned that spectra were determined with and without focusing the laser and a t very low laser power outputs. There was no discernible difference (other than intensity) between these spectra and those observed a t high power outputs. There is also the possibility that a cavity as has The Journal of Physicttl Chemistry, Vol. 77, No. 14, 1973
I 758
H. Lutz, E. Breheret, and L. Lindqvist
been described does not exist. Although no scattering center was found, the conclusion must be reached that alkali metals do not perturb the solvent structure in a detectable manner in the concentration region studied.
Acknowledgment. We gratefully acknowledge the financial support of The Robert A. Welch Foundation. We are indebted to the University of Dallas for the loan of the Spex 1401. References and Notes (1) Metal-Ammonia Solutions, Physicochem. Prop., Coiioq. Weyl / I ,
1970,l(1971). (2) D. A. Copeland, N. R. Kestner. and J. Jortner, J. Chem. Phys., 53, 1189 (1970). (3) P. F. Rusch, Universite Cathoiique de Lilie. France, private cornmunication.
(4)G. C, Klick and J. H.Schulrnan, "Solid State Physics," Vol. 5, Academic Press, New York, N. Y., 195'7,p 97. (5)T. A. Beckman and K. S. Pitzer, J. Phys. Chem., 65, 1527 (1961). (6) D.F. Burowand J. J. Lagowski, J. Phys. Chem., 72,.169 (1968). (7) P. F. Rusch, Ph.D. Dissertation, The University of Texas at Austin, 1971. (8) P. Daure, Ann. Phys., 12,375 (1929). (9)S.Bhagavantam, lndian J. Phys., 5, 54 (1930). (IO) A.* Dadieu and K. W . Kohlrausch. Naturwissenschaften, 18, 154
(1930). (11) M. G.Costeanu, C. R. Acad. Sci., 207, 285 (1938).
(12)S. Kinurnaki and K. Aida, Sci. Rep. Res. lnsf. Tokohu Univ., 6, 186 (1954). (13)C. A. Piint, R. M. B. Small, and H. L. Welsh, Can. J. Phys., 32, 653 (1954). (14)G. Seillier. M. Ceccaidi. and J. P. Leicknam. Method. Phvs. Anal. (GAM), 4, 388 (1968). (15)T. Birchall and I . Drummond, J. Chem. Sac. A, 1859 (1970). (16) 8. Bettignies and F. Wailart, C. R. Acad. Sci., 271, 640 (1970). (17) R . Quinn and J. J. Lagowski, J. Phys. Chem., 73, 2326 (1969). (18)D. F. Burow and J. J. Lagowski, Advan. Chem. Ser., No. 50, 1 11965), (19)G. Herzberg, "infrared and Raman Spectra," Van Nostrand, New York, N. Y.. 1966. (20) H. H. Claasen, H. Seiig, and J. Shanier, Appi. Spectrosc., 23, 8 (1969). (21)C. D. Ailernand, Appl. Spectrosc., 24, 348 (1970). (22)W. F. Murphy, M. V. Evans, and P. Bender, J. Chem. Phys., 47, 1836 (1967). (23)A. F. Slomba, C. D. Hinrnan, and E. H. Siegler. Proc. Conf. Anal. Chem. Appl. Spectrosc., (1965). (24)A. E. Douglas and D. H. Hank, J. Opt. SOC.Amer., 38, 281 (1948). (25) M. Gold and W. L. Joiiy, lnorg. Chem., 1, 818 (1962). (26) D. A. Kleinman, Phys. Rev. A , 134, 423 (1964). (27)J. Brandmuiler and H. Moser, "Einfuhrung in die Rarnan-spektroskopie," Dr. Dietrich Steinkopff Verlag, Darrnstadt, 1962 (28)J. M. Worlock and S. P. S. Porto, Phys. Rev. Lett., 15, 697 (1965). (29)S.Radhakrishna and H. K. Sehgai, Phys. Lett. A, 29,286 (1969). (30) R. Catterall, Metal-Ammonia Solutions; Physicochem. Prop., Coiioq. Weyl / I , 1970,1 (1971).
Effects of Solvent and Substituents on the Absorption Spectra of Triplet Acetophenone and the Acetophenone Ketyl Radical Studied by Nanosecond Laser Photolysis Hanspeter Lutz,' Emilienne Breheret, and Lars Lindqvist* Laboratoire de Photophysique Moieculaire du C. N.R.S., Universite de Paris-Sud, 91405 Orsay, France (Received December 8, 1972) Publication costs assisted by the Centre National de la Recherche Scientifique
The triplet absorption spectra of p-trifluoromethylacetophenone, acetophenone, and p-methyl- and pmethoxyacetophenone in nonpolar solvents were measured by nanosecond laser photolysis using the third and fourth harmonics of a Nd glass laser. Acetophenone was also studied in acetonitrile and water. In the cases where the energy of the lowest 3(n,7r*) state is lower than that of the lowest 3(~,7r*)state, the triplet absorption spectra are characterized by two weak absorption bands a t about 410 and 450 nm and by a strong absorption appearing below 330 nm. When the 3(7r,7r*)'state is lowest, the triplet has a structureless, weak absorption in the visible and a strong maximum a t about 350 nm. The spectra of the ketyl radicals of p-trifluoromethylacetophenone, acetophenone, and p-methylacetophenone were determined in cyclohexane. The spectra are all very similar and resemble the triplet absorption spectra of the ketones which have lowest triplets of 3(n,r*) nature. The rate constants of hydrogen abstraction from 2-propanol by the triplet state of these latter ketones in benzene solutions containing 2 M 2-propanol are 6.2 X lo6, 1.2 x 106, and 1.3 x IO5 M-1 sec-1, respectively. The results indicate that measurements of triplet absorption spectra may be useful in determining the 3(n,n*)us. 3(n,7r*) nature of the lowest triplet state of alkyl phenyl ketones.
Introduction It is known that many alkyl phenyl ketones have almost isoenergetic lowest 3 ( n , ~ *and ) 3(7r,r*) states, a property leading to interesting spectroscopic and chemical consequences. Information about the nature of the lowest triplet state of these compounds has been obtained mainly The Journal of Physical Chemistry, Voi. 77,No. 14,1973
from studies of phosphoresence,2-6 singlet-triplet absorption,7.8 and zero-field splitting.QJ0 Triplet absorption spectra are expected also to be of value in characterizing the triplet state; however, only few measurements of triplet absorption spectra of alkyl phenyl ketones have been reported previously.1l~l2
