Substituent Effects on Intramolecular Energy ... - ACS Publications

therefore, in regard to an iron wire at 1000° seem to be entirely reasonable. ... This phenomenon is the result of an intramolecular energy transfer ...
0 downloads 0 Views 2MB Size
N. FILIPESCU, W. F. SAGER, AKD F. A. SERAFIN

3324

decreased to a steady-state value. The number of N atoms in the burst was estimated to be from 10 to 20 times the number of ?;H radicals. ,4s shown in Fig. 11, no burst of S H zradicals was observed. It seems very probable that the burst of N atoms may be interpreted as resulting from the formation of some FeeX a t low temperature as the temperature is being raised from room temperature to 1000°. Such a nitride is known to form rapidly at low temperature and is also known to decompose at temperatures above 500'. The evolution of r\"in such large quantities initially is a little puzzling. If the Fe4N contained from 5-10% of a compound Fe,NH, as an impurity, the burst of NH radicals would be acounted for.8 Such N H groups have never been reported in connection with the formation of iron nitrides. It is quite possible, however, that the presence of this small amount of NH could have been overlooked. It is interesting to note that a t steady state, the evolution of N, XH, and XHZ reach small but finite

steady-state levels. It would appear on the basis of approximate calibrations that the number of nitrogen atoms is five- to tenfold greater at steady state than the number of NH groups though this ratio must still be taken as an approximation. The presence of a high concentration of nitrogen atoms on the surface of an iron catalyst is consistent with the current theories of ammonia synthesis and decomposition. The presence of NH groups on well-reduced catalysts is perhaps less well established than the presence of nitrogen atoms although, when certain promoters such as alumina are present, considerable evidence exists that NH and perhaps XH2 groups are also to be found as intermediates on the surface of an active iron catalyst. The observations here made, therefore, in regard to an iron wire a t 1000' seem to be entirely reasonable.

(8) T h e referee pointed o u t that F e N H may also be formed.

Substituent Effects on Intramolecular Energy Transfer. 11. Fluorescence Spectra of Europium and Terbium p-Diketone Chelates

by N. Filipescu, W. F. Sager, and F. A. Serafin Department of Chemistry, The George Washington University, Washington 6 , D. C.

(Received J u n e ll?, 1964)

The fluorescence spectra of europium and terbium p-diketone chelates are modified significantly on changing substituents in the organic ligand. The relative intensity, spectral distribution, shifting, and splitting of the fluorescence lines are discussed in relation to the nature of substituents, their position, molecular configuration, and the over-all intramolecular energy transfer.

Introduction Sharp-line emission and absorption characteristic to the rare earth ions imbedded in crystalline solids consists primarily of electric-djpole transitions between the levels of certain multiplets that can be built from the inner f-electrons that these ions possess. The The Journal of Phusical Chemistry

clearly expressed discrete structure is closely related to the immediate environment surrounding the ion, and, therefore, the ions of the rare earth elements can serve as sensitive probes introduced in systems being investigated. The spectra of rare earth ions incorporated in inorganic single crystals have been success-

SUBSTITUENT EFFECTS ON INTRAMOLECULAR ENERGY TRANSFER

fully interpreted by considering the crystalline field as a perturbation on the free ion Hami1tonian.l The Zeeman effect, paramagnetic studies, and polarized light have also been used to elucidate the behavior of these ions in different environments.2 Rare earth ions incorporated jn organic complexes by coordination through a donor atom such as oxygen or nitrogen, when excited with light absorbed by the ligand, exhibit narrow-line emission at approxirnately the same frequencies as the inorganic single crystal system. This phenomenon is the result of an intramolecular energy transfer from the electronic states associated with the organic complex to localized intra4f shell energy levels of the ions.3 In a previous paper of this series4 the ultraviolet absorption spectra of a number of substituted rare earth @-diketone chelates and the phosphorescence spectra of the gadolinium complexes have been analyzed. The present work is concerned with the fluorescence spectra of the europium and terbium p-diketonates and the over-all intramolecular energy transfer governing the absorption-emission process.

Experimental The preparation of different p-diketones and the rare earth chelates has 'been described previously. The fluorescence spectra of europium and terbium complexes were recorded on a Cary Model 14 spectrophotometer using the fluorescence attachment. The compounds were dissolved at a concentration of M in EPA ( 5 parts diethyl ether, 5 parts 3-methylpentane, and 2 parts absolute ethanol by volume). The diethyl ether and the methylpentane were distilled from metallic sodium ribbon, and the reagent grade ethanol was distilled from magnlesium ethoxide. The fresh solutions of europium or terbium chelates were introduced into 10-mm. quartz tubes and immersed in liquid nitrogen jn a transparent quartz dewar. The frozen samples were irradiated with two sources: a beam from the internal mercury vapor lamp in the cary14 spectraphotometer the from the front and one from an externd mercury vapor lamp 9W, Model SI, 3660 (Ultraviolet pro,ducts,I ~ ~ . ) to the sampleslit beam. For all the aromatic 8-diketone chelates the exciting light was filtered through an aqueouf CuSOp and a Corning 5840 (3000-4000-A* transmission range), whereas for the aliphatic @-diketonates (acetylacetonates, trifluoroacetylacetonates> and fluoroacetylacetonates) the internal Cary 14 ultraviolet Source Was filtered tjhrough a wiso4 Water solution and a Corning 9863 filter and coupled with an external short wave ultraviolet lamp 9W, Model SL 2537, perpendicular to the sample-beam direction. The log scale was used on -the recorder, and all t,he fluorescence spectra

3325

were extrapolated to an equivalent 0.5-mm. slit for comparison. A semiquantitative evaluation of the relative intensities is possible by comparing the area under the curve of a given peak on a linear R.I. = f(v) scale.

Results and Discussion The ground state Qf both europium and terbium trivalent ions is a 7F state. The lowest excited states inside the 4f-shell are the 5D0, 5D1, and 5Dzlevels for europium and the 5D4 level for terbium. Under the influence of the electric field of the surrounding crystal lattice or ligand, the excited states as well as the ground 7F levels will separate into a number of states with different energy. The number of sublevels resulting from the splitting and the extent of their separatioh are understood rather well for rare earth ions situated in a field of known symmetry in ionic crystals, but the calculations made attempting to explain the experimental intensities can at best be described as semiquantitative.' The theoretical problems encountered are difficult to treat not only because they involve simultaneous interaction among radiation, matter, and phonons but also because crucial specific details of the wave functions are not known.6 For rare earth ions, the difficulty in estimating intensities of electric dipole transitions is that they arise from the admixture into 4fn of transitions of opposite parity. To calculate such admixtures, not only must the energies and eigenfunctions of configurations such as 4fn-15d be known, but also that part of the ligand or crystal field potential responsible for the admixing.lf For an ion in a chelate, similar to an ion in a crystal, the influence of the field of the surrounding atoms on this ion will be great for all (1) (a) J. Van Vleck, J . P h y s . Chem., 41, 57 (1937); (b) H. Bethe, Ann. P h y s i k , 3 , 1 3 3 (1929) ; (0) B. R.. J u d d , Proc. ROU. SOC. (London), A2289 120 (1955) ; (d) G . S. Ofelk J . Chem. Phys.9 37, 511 (1962) ; (e) J. D. Axe, ibid., 39, 1154 (1963); (f) B. R. J u d d , Phys. Rev., 127, 750 (1962); (g) S. Freed, S. I. Weissman, and I?. E. Fortress, J . Am. Chem. Sot., 63, 1079 (1941); (h) S. Freed and S. I. Weissman, J . Chem. Phys., 6 , 297 (1938). (2) (a) P. Selwood, J . Am. Chem. Soc., 56, 2392 (1934); (b) A. Schmillen, Ann. Physik, 39, 502 (1941) : (c) S. Freed, J . Chem. Phys., 8 , 2 9 1 (1940) ; (d) F. Spedding, p h w . Rev:.,38, 2080 (1931) ; (e) E. H. Carlson and G. H. Dieke, J . Chem. Phys., 34, 1602 (1961); (f) A. Friedrich, K. H. Hellwege, and H . Lammermann, 2.P h y s i k , 159, 524 (1960). (3) (a) 5 . I . Weissman, J . Chem. P h y s . , 10, 214 (1942); (b) G. A. Crosby and M .Kasha, Spectrochim. Acta, 10, 377 (1958) ; (c) G . A . Crosby, 1%.E. Whan, and R. M. Alire, J . Chem. Plays., 34, 743 (1961); (d) R. E. Whan and G. A. Crosby, J . Mol. Spectry., 8 , 315 (1962) ; (e) N. Filipescu, M. R. Kagan, N. McAvoy, and F. A, Serafin. Nature, 196, 467 (1962); (f) G. A. Crosby and R. E. Whan, J . Chem. Phys., 36, 863 (1962); (9) H . Samelson, A. Lempicki, V. A. Brophy, and C. Brecher, ibid., 40, 2547, 2553 (1963); (h) N. McAvoy, N. Filipescu, M. R. Kagan, and F. A. Serafin, J . Phys. Chem. Solids, 25, 461 (1964). (4) W . F. Sager, N. Filipescu, and F. A. Serafin, paper I of this series, J , phys. them,. (in press), (5) D. L. Dexter, J . Chem. Phys., 2 1 , 8 3 6 (1963).

Volume 68, Number 11

November, 2964

N. FILIPESCU, W. F.SAGER, AND F. A. SERAFIN

3326

the outer 6s- and 5d-electrons and considerably smaller for the internal electrons of the 4f-shell. The spectra of europium and terbium ions in chelates, as in crystals, possess a pronounced discrete structure which is naturally explained as due to transitions between levels of the f-shell. The prohibition for electric dipole transitions between levels of one parity is related to the presence in the free ion of a center of symmetry which coincides with the nucleus. But if, upon chelation of the ion, its nucleus ceases to be a center of symmetry, then “forced” dipole transitions between levels of one parity appear. These emission lines characteristic of intra-4f parity-forbidden transitions are observed with high intensity for both europium and terium in 0diketone chelates. It has been shown that noncentrosymmetry due to oscillatory motion1 is minor compared to field distortion introduced by the p-diketone structure and substituents attached to these rare earth chelates. The departure from regular octahedral symmetry in the @-diketonechelates investigated is due to a number of reasons: (1) The oxygen atom orbitals involved in bonding to the metallic ion do not have axial symmetry along the bond because of resonance with the enolate ring, a feature which annihilates the molecular center of symmetry; ( 2 ) different radicals attached to the chelating ring affect both the metal-oxygen distance and the electron density associated with the oxygen atoms bound to the ion; (3) solvent molecules of high polarity. may tend to associate with the ion introducing new (but probably much weaker) perturbing effects and possibly slight modification of the regular arrangement of the nearest neighbors, As a rule, the less symmetrical the geometric arrangement surrounding the ion, the more numerous the lines appearing in the spectrum due to internal Stark splitting of the “unperturbed” ionic levels. The distribution of the nearest neighbors around a trivalent ion in 0-diketone chelates is very close to a configuration of 01,symmetry, whereas the whole molecule of a symmetrically substituted @diketonate belongs to the CaVgroup symmetry. The number of sublevels in which the individual ionic levels are split under the influence of internal Stark fields of known symmetry is easily predicted as dependent on the total angular momentum quantum number J.6 The maximum number of lines to be observed for individual transitions from 6D0, 3D1 (or 5D4) to the low-lying ’F levels of europium (or terbium) ion in an Oh or Csvfield is given in Table I. Generally, the number of lines in the fluorescence spectra of europium and terbium p-diketonates agrees with a distorted octahedral field about the ion approaching a Csvsymmetry depending on substituents. The Journal of Physical Chemistry

Table I Eu3+ max. number of lines Eu3” max. nutnber of lines Tb3+ Max. number of lines

.

6Dot o Oh c3v

‘D, to Oh c3v

6D4t o Oh c3v

7Fo 7FF1 7F2 1 1 2 1 2 3 7Fo ?F1 7Fz 1 1 2 4 6 2 7Fo 7F1 ?F2 4 4 8 6 12 18

7Fs

?F4

3 5 7Fa 3 10

4 6

?F4

4 12 ?Fa 7Fa 12 16 30 36

7F~ 4 7

6 9

7F5

F2

4 14

6 18 ?F6 ’Fe 16 24 42 54

Appreciable variation is observed in relative intensity and number of individual lines in the fluorescence spectra of differently substituted europium and terbium 0-diketone chelates. The various substituents change both the symmetry and the strength of the molecular field surrounding the ion and modify the interaction of the inner 4f-shell of the ion with its environment. The differences in the emission spectra of europium or terbium chelates grouped in series A (meta- and para-substituted dibenzoylmethides) indicate a significant interaction through the aromatic benzene ring between distantly appended substituents and the enolate chelate ring. The successive steps of the over-all intramolecular energy transfer, the result of which is the ionic fluorescence, are affected by the substituents as given in the following paragraphs. Absorption in the Near-Ultraviolet “(x + XI). Substituents directly attached to the 0-diketone chelate ring (series B) or more distantly (series A) shift A,, of the main molecular absorption peak. The absorption intensity does not vary significantly on changing substituents. The energy of the excited singlet (which changes on replacing substituents) does not affect directly the energy transfer from the organic ligand to the chelated ion. Intersystem Crossing (8‘ -+ T ) . This radiationless process depends markedly on the substituents; different efficiencies were observed for the various chelates investigated. The efficiency was evaluated from the phosphorescence spectra (T 4 5 ) of the gadolinium chelates at 77’K. in a solid matrix, when quenching of the triplet state is negligible. Competing processes are the organic fluorescence (S’ -+ S) and the radiationless deactivation of the excited singlet as thermal energy. The organic fluorescence has a low yield so that the amount of energy available at the triplet level depends on the efficiencyof S’-+T intersystem crossing (6) (a) H. Bethe, Ann. Physik, 3, 133 (1929); (b) H. Gobrecht, ibid., 28, 673 (1937).

SUBSTITUENT EFFECTS ON INTRAMOLECULAR ENERGY TRANSFER

which competes with an intramolecular radiationless deactivation. Triplet State. The energy of the triplet state is directly dependent on the substituents. The requirement for an efficient migration of energy to the ion is that the triplet stat e be equal or above the resonance level of the chelated ion. Apparently shifting, splitting, or line width of individual transitions are not dependent on the energetic location' of the triplet, Energy Transfer p o r n the Organic Triplet to the Rare Earth Ion. This is a balance-like process in that the organic part goes by a radiationless transition to the ground-state configuration, as the ionic 4f-electron is raised to an excited state. The mechanism of energy transfer to theion involves a strong interaction between the triplet state or a vibrational level of the chelate in the triplet configuration with the resonance level of the ion. Consequently, the closer the triplet state to the emitting level of the rare earth ion, the more efficient is the coupling between the two states and the migration of energy to the ion. Attached substituents affect this step through their influence on the energetic location of the triplet and the nature of the metaloxygen bond since the donor capability of the oxygen atoms depends on the substituents through resonance or inductive effects in both singlet and triplet states. Competing processes are the organic phosphorescence (T --t S) and the thermal dissipation of the triplet state energy through vibrational coupling to the surroundings (solvent) For chelates having the triplet state near (equal or above) the emitting level of the chelated ion, the organic phosphorescence is negligible which shows efficient transfer to the ion. Solvent quenching of the triplet state is also insignificant in tJhe rigid matrix a t 77°K.3 Figures in the previous paper of this series4 illustrate the relative positions of the triplet states of different chelates (as determined from the phosphorescence of the gadolinium p-diketonates) and the resonance levels of europium (6Doand 5D1) and terbium PD.4). The fluorescence emission of the excited europium ion depends on the perturbing field of the ligand which varies from chelate to chelate. On increasing interaction between the 4f shell and the nearest neighbors, i.e., an increase in the internal electric field, both the lower and the upper combining levels are displaced downward. wI

c F

1'-

-

The displacement of the upper levels, the less shielded levels, is usually greater, and, hence, the lines are dis,

3327

placed with an increase in the molecular field, as a rule, in the direction of longer wave lengths (v' < v). Of particular interest is the zero line 5Do-+ in europium ion which is not split in electric fields. This transition from an upper level with J = 0 to the lower 'F level corresponding also to J = 0 which is strictly forbidden for the free ion becomes possible for the ion in the chelate, since the' quantum number J only approximately preserves its meaning. Displacement of the zero line gives an indication of the magnitude of the internal electric field. According to the observed structure of the band (splitting), it is possible to estimate the symmetry of the field, which is directly related to the number of lines corresponding to purely electronic transitions. The magnitude of the splitting is related to the magnitude of the electric field. Superposition of oscillatory transitions upon the purely electronic transitions in the fluorescence spectra of Eu3+,Gda+,and Tb3+,has been found to be negligible.' In p-diketonates, europium ion transitions are observed between unsplit levels, between levels of which only one is split, and between upper and lower levels where both are split by the internal electric field. The fluorescence spectrum of terbium consists of transitions originating at a level with J = 4, which is highly split in both O h and C3"field symmetry, and terminating at highly split levels with J = 6, 5 , 4, 3. The line width is temperature dependent, but the fluorescence spectra were recorded at the same temperature (77'K.) for all chelates. The dependence of the line width and line contour of surroundings is obvious on comparing spectra of the rare earth ions in different crystals, glasses, and solutions at the same temperature. Line broadening is associated with irregular distribution of ions in static electric fields of different magnitude and symmetries as in glasses, or superposition of oscillatory transitions on purely electronic transitions. In rare earth chelates the line width may also depend on the location of the triplet state of the ligand vs. the emitting level of the ion and the existence of stereoisomers. Some general considerations on the fluorescence spectra of europium chelates are summarized below.

General Considerations on Fluorescence Spectra of Europium Chelates (1) The requirement for energy migration to the ion is that the triplet state be close to or above the resonance level of the europium ion. The shifting, splitting, and line width of individual transitions are not dependent on the energetic location of the triplet. (2) The relative intensity of the fluorescence lines is not directly dependent only on the amount of energy Volume 68,Number 11

ATovember, 1964

3328

available a t the triplet state and transferred to the ion but primarily on the probability of radiative deactivation compared to nonradiative dissipation of energy from the excited ion. This probability is very sensitive to changes in substituents in the ligand. (3) The molecular electric field surrounding the ion is affected strongly by substituents, determining important variation in line intensity, line width, shifting, and splitting of the individual lines. The intensity of this molecular field is changed to some extent on replacing substituents, but the symmetry of the field is modified significantly and seems to determine the spectroscopic behavior of the chelated ion. (4) The ratio of intensity of two different transitions originating from the same resonance level varies greatly for different chelates, indicating that once energy is transmitted to the ionic resonance level (5D1 or 6Do), transitions to different low-lying 'F levels are not in the same statistical ratio for europium ions in any chelates, but that very different transition probabilities for transitions from the same level are determined by the surroundings (substituents). ( 5 ) The parity-forbidden intra-4f shell transitions in europium ion are strong in the investigated chelates; this indicates the existence of a noncentrosymmetrical field about the ion. The 5Do+ 7Foline which would be strictly forbidden in a regular octahedral field is more intense for asymmetrically substituted p-diketonates than for those with symmetrical substituents in both series studied, showing that the departure from o h symmetry is substituent dependent. (Similar behavior has been observed for other transitions.) The shifting of the unsplit 5D0+ 7Foline shows minor changes in the molecular field intensity, so that the major kause for differences in fluorescence spectra of europium chelates is the change in symmetry of the internal Stark field. (6) Electron-donor methoxy groups substituted on dibenzoylniethide chelate increase the amount of energy transferred from the triplet state of the ligand to the ion, probably by an increased orbital overlap in the metal-oxygen bond. An opposite effect, decrease in total ionic emission, was observed for the electron-withdrawing nitro substituents, attached in both meta and para positions. (7) An increase in the conjugated system by attaching aromatic substituents (p-phenyldibenzoylmethide, dinaphthoylmethides) resulted in enhanced ionic emission. (8) The aliphatic p-diketonates (acetylacetonate, trifluoroacetylacetonate, and hexafluoroacetylacetonates) showed a weak ionic fluorescence, and this was attributed to the existence of a large gap between the resonance levels of the europium ion and the triplet The Jaurnal of Physical Chemistry

N. FILIPESCU, W. F. SAGER, AND F. A. SERAFIN

state of the ligand, causing inefficient transfer to the ion. (9) Combinations of aromatic-aliphatic substituents on p-diketones (benzoylacetonate, benzoyItrifluoroacetonate, theonyltrifluoroacetonate) showed a strong ionic emission. This was attributed to efficient energy migration to the trivalent ion coupled with high radiative transition probability between the excited state and the low-lying 7F levels due to an increased asymmetry about the ion. (10) The distribution of the energy transferred to the ion among the individual possible transitions originating at the two 5D1and 5Dolevels and terminating a t different 7F levels cannot be correlated with the location of the triplet state of the ligand. Thus, arylalkyl substituted p-diketonates (BA, BTA, TTA) fluoresce stronger from the 5Do level (the 5Do4 7F transition represents more than 60% of total emission) although the triplet state lies a t wave lengths shorter than the corresponding diary1 p-diketonates (DBM, DTM) which in turn emit more strongly from the 5D1 level (5D1--t 7F4 > 65%). (11) The splitting of the fluorescence lines corresponds to molecular electric fields having a distorted O h symmetry, but not exceeding in number of lines the Csv symmetry. The 5Do + 7F1 line is split into a doublet in asymmetrical ligands in both series A and B and is singlet in symmetrical chelates, showing a greater departure from octahe(dra1 symmetry in unsymmetrically substituted diketonates (the line should be a singlet in a field of 01, symmetry and doublet in a field of CsVsymmetry), (12) The line width of the fluorescence transitions is substituent dependent. A regular variation of the line width of the weak 6Do+ ?Foand 5D1 + 7F3transitions is observed in series A where monophenyl substitution on dibenzoylinethide sharpened these lines and para disubstitution of either CHIO, NOz, or F had a broadening effect. If stereoisomers are formed in the case of asymmetrically substituted chelates, they have essentially equivalent electric fields surrounding the ion. Otherwise, line broadening would result from emission a t very close different frequencies. This is not observed. (13) The strong EDo + 7Fz and 6D1 -+ IF4 lines are broadened significantly on para substitution of electron-withdrawing groups (nitro, fluoro) on the dibenzoylmethide chelate in series A, whereas not much change is observed on meta substitution of either methoxy or nitro groups or para methoxy substitution. Following is a more detailed discussion of the individual lines in the fluorescence spectra of europium pdiketonates.

SUBY'I'ITUEhY' EFFECTS O N

INTIZAMOLECU14ARENERGY TRANSFER

The total relative intensities of ionic fluorescerice for europium chelates (rc.gardlcss of how this emission is distributed aniong individual transitions) varied with the chelate as given in the two series in Table I1 (the total intensity of wropiurn diberizoylriictharie was taken as unity). Table I1 Series A p-Phen yldibenzoylmethide Ili-m-methoxydibenzoylinethide m-Methoxydibenzoylmethide 1)ibenzoylmethide p-Methoxydibenzoylmethide p-N itrodibenzoylmethide Di-p-methosydibenzoylmethide m-Nitrodibeneoylmethide IX-m-ni trodiberizoylmethide Di-p-nitrodibenzoy lmethide Di-p-fluorodibenxoj lmethide Series B Di-2-naphthoylmethide Theonyltrifluoroacetonate Di-1-naphthoylmethide Benaoylacetonate Difuroy lmethide Benzoyltrifluoroacetonnte Dibenxoylmethide Ilitheonylmethide IXisonicotylmethide Hexafluoroacety lacetonate Trifluoroacetylacetonate Acetylacetonate

4.6 1.81 1.13 1

0.95 0.75 0.62 0.23 0.23 0.12 0.04 14.01 I). 12 9.12 4.38 3.24 1.32 1

0.37 0.15 0.10 0.01 0. 01

The over-all fluorescerice intensity characteristic to the ion depends on the amount of energy available at the organic triplct and on the efficiency of energy transfer to the ion. These two factors vary for different appended substituents. Thus, mefhozy groups, substituted in meta position on dibenzoylniethide chelates, tend to enhance the ionic emission, whereas paramethoxy substitution decreases the europium fluorescence. The effcct is more pronounced for the di- than for monomethoxy-substituted dibenzoylniethides. The phosphorescence-relative intensities of the Gd chelates indicate a more efficient intersystem crossing for paraniethoxy-subst ituted dibenzoyltnethide than for the meta isomers. The reverse order in ionic fluorescence (m-CH30 > p-CH30) can be explained in part by the exclusive inductive effects of methoxy groups in meta posit ion, whereas in para-substituted dibenzoylmethides resonance effects tend to decrease the efficiency of energy from the organic triplet to the ionic. The inductive effect of the electron-donor inethoxy group at-

3329

tached to a dibcrizoylmcthide system increases the electron density in the conjugated system, and possibly in the metal-oxygen bond, with the result of a higher orbital overlap and consequently an easier transfer of energy to the ion. An opposite effect is expected arid observed for the nitro-substituted dibcnzoylniethides of europiurn. The electron-attracting nitro groups appended para or meta decrease the total ionic emission of europium. The effect is more powerful for di- than for nionosubstituted dibenzoylniethidcs. para-AIonosubst itution of a CH30or S O 2 group in europium dibenzoy1mt:thide decreases slightly the over-all ionic emission intensity but on para-disubstitution of either CHsO or S O 2 (or F) the intensity is significantly decreased. This indicates that resonance effects of para substituents opposing energy transfer to the ion slightly overconie the asymnietry-effects introduced by monosubstitution which favor ionic emission. When asymmetry eff ccts are removed by para-disubstitution, only the (opposing) resonance effects remain, decreasing the ionic emission significantly. Europiutn p-phcnyldibenzoylnxthide has a special position in this series. Its strong ionic fluorescence indicates that an increase in the aromatic system enhances the amount of criergy transferred to the europium ion (the organic triplet of the chelate is not significantly shifted compared to dibenzoylmethide). This fact is also corifirrncd by the two naphthyl-substituted diketones in series B which have substantially higher ionic emission than the dibenzoylnicthide chelate (14.01 and 9.12). In series B the location of the triplet state varies over a wider range. The weak ionic fluoresccrice from aliphatic 0-diketone chelates (AcA, TFAcA, and HFAcA) can be explained by the existence of a large gap between the resonance levels of europium ion (5D1, 5Do) and the triplet level of the chelates (more than 3500, 2000, and 1000 cm.-l, respectively). Combinations of aromatic-aliphatic substituents on p-diketones (TTA, BTA, BA) result in an increased Eu3+fluorescence, compared to the parent symmetrical aromatic (DTJI, DBRI) or aliphatic (HE'AcA, AcA) diketone chelates. On replacing one substituent iri a diaromatic @-diketone (DRhI, DTM) with a CH3 or CF3 group (BA, BTA, TTA), the energy of the triplct state is displaced toward short)er wave lcngt hs, further above the 5Doarid 5D1 levels of Eu3+ ion. This displacement should not increase the efficiency of the energy transfer to the ion. The strong ionic emission in asymmetrical aryl-alkyl-substituted 0-diketone chrlatcs (TTA, BA, BTA) results from an efficient encrgy migration to the t rivalerit europium ion coupled w i t h a Volume 68, Number I 1

November, 1964

K. FILIPESCU, W. F. SAGER,AND F. A. SERAFIN

3330

high transition probability between the two excited levels 5Doand 5D1 and the low-lying 'F multiplet. The amount of energy transferred to the ion depends on substituent influence in the S' w+ T interrsystem crossing (amount of energy available at T), in the energetic location of the triplet vs. ionic resonance levels, and in the efficiency of transfer through the metaloxygen bond to the ion. But the relative fluorescence intensity of the ion is the energy dissipated radiatively from the excited level, a process which competes with a nonradiative degradation process. The conipetition between the radiative process and quenching, probably via vibronic coupling to the ligands and thence to the surroundings, depends upon other factors such as temperature, solvent, energy gap, etc., which are inore or less equivalent for all chelates and on the transition probabilities for the radiative process. The fluorescent lines from Eu3+ ion correspond to intra-4f shell "forbidden, , transitions, which become more allowed, the more the internal electric field due to the particles surrounding the ion is distorted from a symmetry configuration where the ion occupies the center of symmetry. This departure from centrosynimetry (in this case O h ) is dependent on substituents and expected to be greater for unsymmetrical )-diketones. This would explain the higher relative intensity of BA, BTA, and TTA compared to DBM, DTM, AcA, and TFAcA. There is no apparent reason to consider

1

I

I

I

6200

6100

6000

Wave length,

I

1

A.

5900

5800

Figure 2 . Fluorescence spectrum of europium tris( p-methoxydibenzoylm$thane j (EupM); M in EPA, 77"K., 5 A./sec.

0.25 MM SLIT

I

I

I

niuu

I

I

6300

6200

6100

6000

Wave length,

5900

5800

5700

b.

Figure 3. Fluorescence spectrum of europium tris( di-m-methoxydihenzoylmethane j ( EuDmM) ; ili' in EPA, 77"K., 5 d./sec.

I 6300

6200

I 6000 Wave length,

6100

b.

I

1

I

5900

5800

5700

Figure 1. Fluorescence spectrum of europium tris(dihenaoy1methane) (EuDBMj; in EPA, 77"K., 5 d./sec.

The Journal of Physical Chemistry

that the aliphatic substituent increases the amount of energy transferred to the ion. Instead, it creates a more unsymnietrical field in conibination with an aromatic substituent, thus increasing the transition probability of the radiative process in the Eu ion which competes more favorably with the nonradiative energy dissipation. Another distinction is made here between the different europium @-diketonechelates. The

SVBSTITUENT EFFECTS ON INTRAMOLECULAR ENERGY TRANSFER

6300

6200

6100 6000 Wave lenrrth k.

5900

5800

Figure 4. Fluorescence spectrum of europium tris(di-2-theonylmethane) (EuDTM); 10-6 Jf in EPA, 77”K., 5 A./sec.

180” around the diketone ring-substituent bond, an identical configuration results. Group (a) includes DBM, DpRlI, DpN, DpF, DPyM, AcA, and HFAcA. The second group includes all the symmetrical diketones where 180” rotation of the substituent brings about a different steric configuration. Group (b) consists of DFuM, DTRI, DmRI, D m S , DlNAI, and D2KM. These chelates are the “less symmetrical” of the symmetrical P-diketones. Resonance in a p-diketone chelate involves only one substituent a t a time and the chelating ring. One substituent in this group may be rotated out of the chelate ring plane and “frozen” in that position in the solid matrix a t 77°K. A substituent of this “less symmetrical” type would need a much larger energy to rotate back in the chelating plane compared to a phenyl, p-methoxyphenyl, methyl. etc. (group a). The resulting “fixed” configuration is rather unsynimetrical for the europium ion. The remaining chelates are the asymmetrical ones having two chemically different substituents attached t o the

Figure 5 . Fluorescence spectlrum of terbium tris(dibenxoy1methane) (TbDBM);

complexes (DFuM, DBRI, DTM, DpM, DmRI, DmX, DpS, DpF, DliY11, D2NA1, AcA, HFAcA, DPyM) are divided into two groups which behave differently spectroscopically. I n the first group are considered those chelates, where on rotating a substituent by

3331

M‘ in EPA, 70”K., 5 Lk./sec.

chelating ring. Group (e) includes pB1, p X , mM, m X , pPh, TTA, BA, BTA, and TFAcA. An increased ionic fluorescence is observed on replacing CH, groups with CF, in the AcA, TFAcA, HFAcA series. This trend niay be easily explained Volume 68, Number 11

November, 1964

3332

N. FILIPESCU, W. F.'SAGER, AND F. A. SERAFIN

h

4

3 E! e,

f* bi

il

6300

6200

5900

5800

6500

Wave length,

5400

5000

4900

4800

1.

Figure6. F'luorescence spectrum of terbium tris(di-p-fluorodibenzoylmethane)(TbDpF); 1 0 - b M in EPA, 77'K., 5 A./sec.

by the lowering of the triplet state energy of the ligand, thus decreasing the gap to the 5D1ionic level from more than 3500 to approximately 1000 cm.-l. KO such enhanced fluorescence is observed on replacing a methyl with a trifluoromethyl group in BA (to BTA), the triplet level being lowered by only 300 cm. - I . The total fluorescence intensity in symmetrical aromatic p-diketones decreased in the following order: 1-naphthyl > 2-naphthyl > fury1 > phenyl > thienyl > 4-pyridyl. From these findings one may predict an expected relative ionic fluorescence intensity of Eua+ in 0diketone chelates with different attached substituents. An alkyl substituent combined with a large aromatic The Journal of Physical Chemistry

radical (naphthyl, biphenyl, phenanthryl, etc.) attached on a R-CO-CH2-CO-R' system would probably offer the combination of efficient transfer of energy to the ion and high transition probability for the radiative process, provided the triplet state has the proper location us. the ionic resonance level. Both alkyl (straight chain, branched chain, haloalkyl, substituted alkyl, etc.) and aryl (aromatic mono- or polynuclear hydrocarbons, linearly linked or condensed, heterocyclic radicals, etc.) may be selected to provide the most adequate S'--+T intersystem crossing, location of triplet state, and transfer to the ion. Chelates having from poor to very bright ionic fluorescence can be prepared from ligands with various substituents. The fluores-

SUBSTITUENT EFFECTSON INTRAMOLECULAR ENERGY TRANSFER

I

5900

5800

5500

Wave length,

b.

5400

3333

I

I

5900

5800

0.1

M M SLIT

1

I

I

4900

Figure 7. Fluorescence spectrum of terbium tris( benzoyltrifluoroacetonate) (TbBTA) ; 10-6 M in EPA, 77”K., 5 b./sec.

cence spect.ra of europium and terbium chelates are illustrated in Fig. 1-0;‘ and the data are listed in Tables I11 and IV.

5500

Wave length,

b.

5400

4900

Figure 8. Fluorescence spectrum of terbium tris( trifluoroacetylacetonat?) (TbTFAcA) ; 10-6 M in EPA, 77”K., 5 A./sec.

and doublet for the unsymmetrical ones, even if differences in ligands are marginal, as in series A where only distant substituents are changed. This is a very significant indication that the departure from O h synimetry is substituent dependent. Some other information can be drawn from close examination of the 5Do+ 7F1lines of the chelates. The amount of splitting ( A v ) could be taken as a qualitative measure of the distortion from centrosymmetrical o h configuration. para- and meta-monosubstituted (CHBO or KO,) dibenzoylmethides have a greater splitting than asymmetrical p-diketones in series B (BA, TTA, BTA, etc.). Substituents which are not symmetrical on free rotation around the bond with the chelate ring

Individual Lines in En Chelates : Splitting of Electronic Levels The only transitions observed in the fluorescence spectra were from the two resonance levels 5D1 and 5D0 to the lower four levels 7Fo....4of the ground multiplet. Transitions to 7F5and 7F6 ( A J 3 4) are very weak and therefore were unobserved by the phototube detector which had a limited sensitivity a t wave lengths greater than 7000 A. The 5Do-+ 7Foline is singlet for all europium chelates investigated, as expected for a J = 0 J = 0 transition, unsplit by fields of any symmetry. The spectral purity of this transition is remarkable. The noncentrosymmetrical configuration in chelates allows this strictly “forbidden” transition with a high probability. In general, any change iii the dibensoylmethide structure 0 S resulted in increased relative intensity for the 0 + 0 line (expressed as percentage of total fluorescence). The variation of the relative intensity of the 5Do4- generate a small but distinct doublet for the 5Do + ‘F1 transition. (This type of substituent belongs to 7Fo transition in europium chelates is expressed in Table V (in % of total emission) for series A and W. (7) Figures 1-4 (europium chelates) and 5-8 (terbium chelates) are The 5D0 + 7F1transition terminates on a level with illustrated in t h e paper. T h e entire collection of fluorescence J = 1 which is unsplit in a field of o h symmetry but spectra (44 diagrams) has been deposited as Document No. 8150 with’ t h e AD1 Auxiliary Publication Project, Photoduplication Sersplits into two levels under the perturbing field of ligands vice, Library of Congress, Washington 25, D . C . A copy m a y be having a less symmetrical arrangernent such as Csv. secured by citing t h e Document Number and by remitting $6.25 for photoprints or $2.50 for microfilm in advance payment by check This line, corresponding to the EDo + 7F1 transition, or money order payable t o : Chief, Photoduplication Service, is a singlet for all essentially symmetrical fi-diketones Library of Congress. -+

~

~

~

Volume 68, Number I i

November, 196.4

?

3334

T\i.

FILIPESCU, w.F. SAGER,

AXD

F. A.

SERAFIN

Table I11 : Individual Fluorescence Linea of Europium Chelates Europium ion transition

A d at b S X 9

ern.-'

'/&nax,

cm.-1

AV,Q

hmm?

cm.-1

R.U.

0

Eu( \

Europium Dibenzoylmethide 'DI

--

7F4

+

'Do 'DO 6D1 'Do

'Fz TI

-

?Fs

-+

16,077 16,176 16,308 16,773 17,065 17,256

48 40 27 81 39 33

64 149 129 170 96 85

6Di -L 'F4

--

?Fz

'DI

+

7F3

'Do

-+

'Fo

5Do 'Do

16,090 16,173 16,305 16,750 16,903 17,073 17,277

39 37 27 95 95 50 42

0

Eu(CHs0 \

'Di

DO

-

-+

'F4

7F~

'Do ---t T i 5D1- 'Fa

eDo

16,090 16,292 16,835 17,036 17,301

26 16 108 139 97

100.4 101.3 101.2

10 102.7 100.3 100.0' 100

'Do 'Do ED1 'Do

'F4

-

+

-+ -+

'Fz 'FI 'F3

?Fo

16,064 16,181 16,313 16,736 16,863 17,079 17,241

32 53 81 81 81 39 33

0

-+

16,181 16,287 16,793 17,010 17,241

87 73 135 120 55

100.8' 101.9 102.2 100.7 100.8 1003 100.6

173 96 208 135 64

6Do 5D1 'Do

+

-+

T 1

'F3

TO

T h e Journal of Physical Chemistry

53 48 32 86 51 17

128 102 69 21 1 107 74

99

30.1 4.3 2.3 0.2

... ...

... ...

72.7

83

18.5 4.1

153

...

...

...

(EuDpM) 1.7 96.0 1.5 0.3. 0.6

... ... ...

... ...

(EupN) 27.4

117

68.2 3.1

127

0.8 0.5

...

... .

.

I

(EuDpN) 57.8 32.0 5.4 3.3 1.5

... ...

... ...

I_\

)3

(EuDmM)

OCHs

102 2 102.4 102 6 101.4

10'

1,713 959 161 98 44

COCHCOQ

CHsO 16,181 16,273 16,311 16,835 17,001 17,241

1NOz)S

101.8 101.2 100.13 100.02 100.1 I-\

-

0

COCHCO \

Eu(

)3

178 4,806 12,402 312 255 138 85

0

Eu(02N \

Europium Di-m-methoxydibenzoylmethide

'D1 + 'F4 5Do T z

0

1COCHCO \ /

Eu(0zN \

49 136 151 87 94 61 41

Europium Di-p-nitrodibenzoylmethide ~ D --F I 'F4 SDO ?Fz 'Do -.+ "F1 ~ D I 'Fs &Do

253 14,535 220 45 86

63.0

4.6 3.3

E u ( C H 3 0 0C O C H C O a O C H &

49 84 225 174 150

om. -1

(EupW

)3

2,540 14,193 4,253 767 192 1,047 769

100.9

Europium p-Nitrodibenzoylmethide

'DI

0

COCHCO \

101.8 102.4 102.1

J-Level splitting,

(EuDBM) 798 14,381 7,251 1,046 548 60

102.4 102.2 10' 1 101.1 100.2

83 152 81 196 115 110 110

Europium Di-p-methoxydibenzoylmethide

0

/ COCHCO \ 1) 3 101.3

Europium p-Methoxydibenzoylmethide

Percentage total emission. %

Relative intensity, A.U.c

6

102.2

9,272 12,434 13,240 2,405 2,080 4,280

21.2 58.7

...

5.5 4.8 9.8

...

38 ...

...

SUBSTITUE~VT EFFECTS

ON

INTRAMOLECULAR ENERGY TRANSFER

3335

Table I11 (Continued) Europium

AY'

ion

VUl*li

transition

cm.-l

at

1/dha=,

A",O

hmcm

em.-'

om. -1

R.U.

Relative

*

intensity, A.U.G

-+

?F4

16,090 16,155 16,202 16,265 16,324 16,722 16,892 16,926 17,065 17,241

68 65 32 16 30 38 67 54 30 55

-

16,064 16,132 16,171 16,231 16 329 16,734 16 858 116,987 17.241

5D1

~

~

39 39 42 48 32 59 73 67 33

2,669 5,100 3,100 9,059 3,780 196 832 526 391 1,668

77 74 60 59 60 48 76 59 64 101

Europium m-Nitrodibenzoylmethide

0

0

Eu(\ / COCHCO \ OzN

161 1 028 1,374 460 1 217 204 343 539 290 ~

~

100 7

100 9 100

0

Eu(

COCRCO

+

'F4

16,069 116,129 16,173 16,223 16,329 16,722 36,852 16,992 17,235

53 48 48 53 34 73 65 65 33

6D1 + 'F4

-

6Do -+ 'Fz ?Do 7 F ~ 6Di

6Do

-

+

'F3

'Fo

bDo 'F3 5D1 -+ 'Fi +

~~

16,108 16,168 16,369 16,689 16,912 16,992 1'7,071 17)280 15,420 113 500 1~a,675 113,834 ~

34 53 48 51 51 48 33 27 230 138 97 111

400 948 1,186 798 986 171 244 529 290

io

115 136 80

ELI( 82 81 121

78 152 80 79 105 260 160 140 150

COCHCO 103 102.9 102.3 101.4 102.2 101.7 102.3 100.4 100

101.' 101,3 101.2

112

47.0

59

5.6

204

1.43 6.1

... ...

53.9

167

21.67 9.7

124

...

..

9.6 '5.2

NO2

54 55 54 54 89

Eurojpiurn p-Phenyldibenzoylmethide

39.8

(EuDmN)

OzN 'Dl

cm. -1

(EumN)

/ ) 3

100 7 10' 5 101 6 101 101 6 100 6

50 52 54 50 90 87 127 137 80

Eun-opium Di-m-nitrodibenzoylmethide

J-Level splitting,

(EumM)

Europium m-Methoxydibenzoylmethide

5D1

Percentage total emission, %

Q-/> )s

60.1

154

17.8 7.5

130

9.5 5.2

...

(EupPh)

37,500 37,138 10,823 1,130 10,064 2,205 7 , 232

67. 3

60

9.8 10.1

223

8.5

79

100

0.1 0.2

...

4.0

334

180 I , 373 1,666 1,474

...

...

~

~~~~

Volume 68, Number 11 Nouernber, 1964

N.FILIPESCU, W. F. SAGER, AND F. A. SERAFIE

3336

Table I11 (Continued)

Europium ion transition

k&x*

om.-'

'/hme.x,

A",O

om.-'

om.-'

hmsx,b

R.U.

Eu(F

Europium Di-p-fluorodibenaoylmethide 16,129 16,308 16,906 17,053 17,271

105 92 139 92 58

186 141 175 91 109

16,113 16,166 16 279 16,324 16,773 16 835 16,992 17,241 15,372 18,587 18,744

32 27 19 27 22 40 27 22 194 116 111

16 13 40 41 32

T h e Journal

of" Physical

Chemistry

73 29 57 81 53 11 208 81 55 55

COCHCO

100.8 100.3 100.' 100.1 100.8

133 69 65 54 55

3,400 25 024 15,460 9,011 1,115 1,513 6 234 8,415 172 6,026 1,855

134 162 84 139 95 90 260 120 100 126

Eu( O C O C H C O D )s

55.4 14.4 12.7 7.6 9.9

... ... ... ...

36.3

53

31.3

45

3.3

62

...

102.2 101.9 100.7 100.8 100.9

6,538 1,887 182 135 237

8 8 \

8.0 10.8 0.2 10.1

... ... 157

(EuDTM)

S

Eu( / \ COCHCO

cm.-i

(EuDFuM)

0

S

J-Level splitting,

(EuDpF)

625 162 143 86 112

72 72 60 60 68 72 71 90 300 150 150

Europium Di-1-naphthoylmethide

16,207 16,335 16,793 16,883 16,978 17,241 15,347 18,553 18,657 18,691

0

0

0

Europium Ditheonylmethide 16,142 16,313 16,736 16,807 17,036

A.U.C

Eu( O C O C H C O O ) 3

Europium Difuroylmethide

Percentage total emission, %

Relative intensity,

AY' a t

)s

72.8 21.0 3.5

... ... 171

2.6

(EuDlNM)

35,084 109,725 4,950 23,794 15,889 28,000 169 445 662 935

16.0 50.0 13.1

...

7.2 12.7 -. 0 . 1

...

0.9

138

... 90

... ...

S U B S T I T U E N T EFFECTS ON INTRAMOLECULAR

3337

ENERGY TRANSFER

Table I11 (Continued) Europium ion transition

AV,a

hmsx?

om. -1

R.U.

-

-+

?Fz 7F1

6Do

'Di

'F4

-+

'F3

7F~

6Do 'Do

+

6Di

-+

'Fa 'Fz

16,194 16,273 16,300 16 ,335 16,749 16,821 16,886 16,972 17,065 17 ,241 15,337 17,841 18,067

81 48 48 46 46 55 57 73 33 36 200 139 166

155 50 50 50 84 80 90 100 83 112 360 200 200

-

-+

DO 6Do 6Di

.-)

-+

6Do

'F4

7F~ 7F1 F 'a ?Fo

16,163 16,308 16,889 16,978 17,256

84 79 122 108 44

99,855 45,990 45,990 44 ,730 13,913 11,932 16,126 9,748 11,790 36,542 190 379 422

196 117 213 159 110

Europium Benzoylacetonate

ED1 'Do

-

+

'IF4

-

'Fz

6Do

'F1

6Di

'Fa

16,213 116,281 16,329 116,764 16,863 116,978 17,050 17,241 15,347 17,841 18,057 118,553 18,692

54 43 43 59 81 54 50 29 160 138 128 1.08 108

92 72 72 99 120 110 81 94 160 150 150 130 140

~

43 27 57 43 50 57 100

102 89 189 104 91 140 160

1,815 395 458 471 495

101.3

10 100.6 100.7 10

, 0

Eu( \

COCHCOCHs)s

102.6 102.9 102.9 101.8 102 101~ 101.3 108.4 100 100 100 101.1 100,7

]Europium Beneoyltrifluoroacetonate 16,181 16,281 16,779 16,998 17,224 18,308 181 553

em.-1

29.6

...

40.4

62

12.4

37

6.4

93

10.8 0.1 0.2

226

...

$ I ~ ( N > O C H C O ~ ) ~ (EuDPyM)

Europium Diisonicotylmethide

~DI

J-Level splitting.

(EuD2NM)

Ewropium Di-2-naphthoylmethide 6Di 6Do

Percentage total emission, %

Relative intensity, A.U.C

101.7

10',6 100.3 101.4 10

...

15.0 59.4

...

9.9

99

5.0

72

9.0 0.1 0.2

216

1.4

139

29.4 44.6 11.9 5.9 0.3 5.0 2.8

...

... ...

...

...

(EuBA)

15,810 31,379 31,379 3,423 7,050 4,329 902 9,546 120 106 139 1,090 447

C0CHCOCFs)s 10a.3 102.8

49.9 10.9 12.6 13.0 13.6

48

(EuBTA)

9,377 14,236 3,797 1,891 95 1,595 900

Volume 68, Number 11

November, 1964

N. FILIPESCU, W. F. SAGER, AND F. A. SERAFI?;

3338

Table I11 (Continued) Europium ion transition

hrnex?

R.U.

Percentage total emission, %

Relative intensity, A.U.O

J-Level splitting, cm. -1

Europium Theonyltrifluoroacetonate 6D1-+

7F4

6Do -.+ 'Fz

-

6D0

7F1

+

"F3 7Fo

-+

7Fa

6D1 6Do 6Do 6Dl

16,126 16,176 16,287 16,313 16,770 16,827 16,869 17,007 17,241 15,267 18,560 18,741

21 52 30 22 38 38 38 24 22 179 72 54

60 71 102 70 65 65 65 94 54 260 160 100

12,781 43,750 101,736 17 984 8,835 7,032 4,442 17,793 1,946 155 3,026 261

50

54.5

26

9.2

99

8.1 0.9 0.1 1.5

... 181

~

Europium Acetylacetonate 'D1+ 7 F a and EDo 7F2

25.7

...

( EuAcA)

16,292

197

221

100.04

61

13.0

...

16,906

240

262

100.01

186

39.7

.. ... .

18,315

167

370

100.1

222

47.3

r i .

-+

6Do-+

--

7F1

and 6Di 7Fs 6D1 'Fz

-

Europium Trifluoroacetylacetonate

6D1 7 F 4 and 6Do 'Fz 'Do + 'Fi and ~ D I 7F3 6Dl 7Fz

-

Eu(CFaCOCHCOCH3)a

( EuTFAcA)

16,179

147

194

100.1

171

46.1

16,949

102

120

100.'

102

27.5

18,298

122

356

100.2

98

26.4

.,,.

.

--f

Europium Hexafluoroacetylacetonate 6Di .-c 7F4 and 6Do + 7F2

EDo

-

Eu( CFsC0CHCOCFs)s

...

(EuHFAcA)

16,228

55

158

101.2

1,062

40.7

...

16,878

102

183

100.9

769

29.5

...

17,986 18,298 18,632

166 138 250

360 358 380

100.05 100.2 100.18

194 254 330

17.1

312

12.6

...

?FI

and 6D1-+

'Fs

6Di 4

7F9

6D1 +. 7F1

a The width of the spectral range covered by the entire band. units. c Arbitrary units (recorder units X cm.-').

the "less symmetrical'' group b). The amount of splitting (Av) of the 6Do--.t 'F1 transition varies as a function of substituents as: pPh > mRI > p;\I > D2KS1 > DmN > pN > m?i > BA > TTA > D l Y M > DTM > DFull; unsplit: DBM, DpllI, DpN, DmM, DpF; no transition: AcA, TFAcA, HFAc14. T h e Journal

5j Physical

Chemistry

* The height

of the peak on a linear scale, in (arbitrary) recorder

The line splitting is not due to stereoisomers. The magnitude of splitting is around 100 tin.-', and the syinrnetrical DFuAI, DTM, DlKAI, D2NA1, and DmN europium chelates show the same range of 6 D -c ~ 7F1line-splitting. Expressing the relative intensity of the 5Do 7F1

-

SUIMTITUENT

3339

EFFECTS OK INTRAMOLECULAR ENERGY 'TRANSFER

Table IV : Individual Fluorescence Lines of Terbium Chelatee

Terbium ion transition

YIIlZlXl

hmsnl

om.-'

R.U.Q

A", cm.-'

Relative intensity, A.U.b

Percentage total emission, %

Terbium ion transition

6D4

-+

4

6D4-.+

'F6

7Fb

7F4

--L

20,539 20,470 20,458 (sh)c 20,442 20,367 (sh) 20,346 20 ,243 20,202 20,161 18,430 18,382 18,349 18,298 18,265 17 227 17,167 17,138 17,065 17 ,992 16,920 16,878 16,849 16,234 (sh) 16,181 (sh) 16 , 129 16,077 (sh) 10,051(sh) 16,026 (sh) ~

'D4

102.6 108.4 103.6 103.6 103 7 102.6 102.66 102.6 104.3 103.66 lo3.@ 104.4 103.2 100.6 100.8 100.8 10 100.7 100.65 100.8 100.R 100 100 100.01 100 100 100

378

238,140

659,352

6D4 -.+

-

-

6D4

7F6

7F6

bD4 + 7F4

ED4 -.+ 'Fa

20,534(sh) 20,483 20 ,429 20,354 20,194 20,020 19,928 18,433 18,399 18,352 18,298 17,212 17 , 167 17,065 17,001 16,92O(sh) 16,892 (sh) 16,858 16,116

108 103.6 103.6 lo3,@ 103.2

378

1,140

0.1

'D4 208

104

0.01

102 102.1 102 102 101.7 lo'.' 103 lo2.? lo2,' 103.4 100.5 101.3 10°.02

-.+

'Fs

20,305 10 (organic) 18,433(sh) 100.' (organic) 18,302 101.8

Relative intensity, A.U.b

473

18,880

24.9

178

56,159

73.3

372

1,488

1.9

264

133

0.2

(TbpN)

1,000

5,000

63.4

100

50

0.6

90

2,835

36.0

Terbium Di-p-nitrodibenzoylmethide

20 ,305 101.3 (organic) 18,433 100.7 (organic) 18,298 101.7

550

5,500

67.7

250

625

7.7

80

2,000

24.6

Terbium rn-Methoxydibenzoylmethide 606

381,780

12.5

135 2,667,940

87.4

106.1 10 101.6 101.6 354

3,540

0.1

101.3 101.4

101.6 100.02

... ...

104.3

101.4

cm.3

Tb(02NaCOCHCOQ).:

102.3

102.3 104.8 lo4,'

R.U.a

Au,

Terbium p-Nitrodibenzoylmethide

Terbium p-lVlethoxydibenzoylmethide

6D4

-.+

-

73.3

20,513 20,450 20; 346 20,284 20,202(sh) 20,040 (sh) 7F6 18,443 18,416(sh) 18,355 (sh) 18,298 17,138 7F4 16,891 7Fs 16,276

+ 'F6

26.4 'D4

165

cm.?

hart,

Terbium Di-p-methoxydibenzoylmethide

Terbium Dibenzoylmethide

'D4

%ax,

Percentage total emission, %

190

96

0

20,504 (sh) 20 ,450 20,367 20,284 20,202 20,040 (sh) 18,433(sh) 18,382 18,332 15,298 18,132 17,135 16,978 16.114

102.7 103.1

lo8,' 103 1023 102.5

464

146,392

39.7

110

219,010

59.4

372

2,976

0.8

174

87

103.5

103.7 103,8 104.6 102.5 101.4 100.8 100

0.02

Volume 68, Number 11 iVovember, 1964

N. FILIPESCU, W. F. SAGER, AND F. A. SERAFIN

3340

Table IV (Continued)

Terbium ion transition

%*XI

em.-'

hmax, R.U.a

A",

om.-'

Relative intensity, A.U.b

Percentage total emission, %

Terbium ion transition

b a x ,

h,,,

om.-'

R.u.~

Av, cm.-1

Terbium Di-p-fluorodibenzoylmethide

Terbium Di-m-methoxydibenzoylmethide

Tb(F QCOCHCOQF)a

CH30 20,420(sh) 20,395 20,346 20,305(sh) 20,202 (sh) 20,120 (sh) 18,433 18,365(sh) 18,305 17,138 16,978 16,103

OCH? 10Ze 102 7 102 7 1026

38,808

19 7

l o 21 10' 8

~

103 7

lo30

168

167,580

85 1

loo6

255

384

0 2

100 4 100

179

89

0 04

103 4

Terbium m-Nitrodibenaoylmethide

OpN 20,304 20,060 (organic) 18,977 (organic) 18,416 18,305

100.7 100.7

190 500

475 1,250

4.1 10.7

10°.s

350

350

2.9

193

9,650

82.3

.. . 6D4

7F6

20,060 (organic) 18,832 (organir ) 18,423 18,298

+

102.6

6D4-P 'Fe

102

500

25,000

58.8

300

2,400

5.6

190

15,105

35.5

20,325 (sh) 20,284 (sh) 20,243 20,202 20,120 (sh) 20,040 (sh) 20,000 (sh) 'D4 + 'Fs 18,443 18,392 18,355 18,298

102 6

6Dq-+

'F5

104.8

106.1 104,3

564 3,553,200

11.0

228 28,636,800

88.9

103.6

10"4 103.3 106.1 106.6

10"b

lo4 101.9 192,3 102.4 102.1 101.8 101.8 102.2 100.3 100.01

372

20,916

181

144

0.06

0

loo.' 100.4 392

588

0.2

150

375,900

99.8

100.~

108.8

108.4 103.7 104 2

Tb((-)COCHCO(-)), S

102

600

30,000

66.5

101.8 102 7

190

15,105

33.5

The Journal of Physical Chemistry

(TbDFuW

Terbium Ditheonylmethide

...

19,900 (organic) 18,443 18,298

104.7

T b ( 0 COCHCO(-)), 0 0

Terbium p-Phenyldibenzoylmethide

...

(TbDpF)

Terbium Difuroylmethide

Terbium Di-m-nitrodibenzoylmethide

. ..

20,429 7Fe 20,442 20,346 20,202 20,026 20 000 19,940 18,426 6 D 4 7F6 18,399 (sh) 18,349 18,298 18,198(sh) 6D4 + 7F4 17,241 17,167 17,057 17,004 16,949 (sh) 16,915(sh) 16,869 16,116 6 D 4-, 7F' 16,008 6D4

309

Relative intensity, A.u.~

Percentage total emission, %

19,493 102.3 (organic) 6D4+ 7F6 18,308 101.8 ... 18,083 101.6 (organic) 'Dq -+ 'F4 17,406 100.4

(TbDTM)

5 250

25,000

77.9

90 200

2,835 4,000

8.8 12.5

200

252

0.8

SUBSTITUENT EFFECTS

ON INTRAMOLECULAR

3341

ENERGY 'TRANSFER

Table 1V (Continued)

Terbium ion transition

Yrnaxl

hmaxl

Au,

cm.-1

R.U.a

om.-'

Relative intensity, A.U.b

Percentage total emission. %

Terbium ion transition

ED4 -+ 100.6 19,212 (organic) 6D4 +. 7 F ~18,308 100.2 6D4 -* 7 F 4 17,322 100.0'

400

800

88.3

70 100

56 50

6.2 5.5 6D4 -P

Terbium IDi-2-naphthoylmethide

.. .

102 19,286 (organic) 6D4 -P 7F6 18,298 101.7 10 , . , 18,050 (organic)

400

20,000

75 180

1,850 900

87.9

ED4 -+

7Fe

7F.5

7F4

8.1 4.0

'D4

Terbium Diisonicotylmethide T~(N''COCHCOO)S

h,,,,

cm. -1

R.U.'

Av,

crn.-l

Relative intensity, A.U.6

%

Terbium Benzoylacetonate

Terhium Di-1-naphthoylmethide

.,.

kl*XI

Pelcentage total emission,

-+

'Fa

20,513 20 ,471 20,396 (sh) 20,367 20,255 20,202 20,080 (sh) 18,433 18,416 18,392 18,342 18,275 17,164 17,123 17,079 (sh) 17,036 16 ,892 16,810 16,103 (sh) 16,064

104 lo3.' lo3.' 1W.a 103 103 lo2 104.2

433

272,160

2.9

168 1,331,400

81.6

104.2

103.8 104.2 104.6 103

102.6 102.' 102.2 101.7

364

28,938

181

91

1.8

101.4

100 100.0'

0

('I'bDPyM) Terbium Benzoyltrifluoroacetonate

20,513 20,408 20,202 (sh) 20,141 18,433 18,305 17,167 17,036 (sh) 16,949 16,915 (sh) 16,883 (sh) 16,155

102 102.2

528

16,632

20.6

128

64,000

79.1

100.05

347

174

100.0' 100 100

0.2

180

90

0.1

COCHCOCFa)3 (TbBTA)

101.8

101.6 103,' 102.9 100.1 100.01

5D4 -+ 7Fa

5D4

7F5

6D4 -.L

7F4

20,475 20,442 20,367 20,325 18,420 18,365 18,305 17,188 17,106 16,915

102.6 lo2,' 102.9 102.6 1043 103.7 103.6 101.3 101.4

145

36,144

14.6

70

220,850

85.6

117

928

0.4

10

Terbium Theonyltrifluoroacetonate Terbium Acetylacetonate Tb( CHICOCHCOCH~)3(TbAcA)

T b ( ( O C o C H c o c Fs)a (TbTTA) S 20,500 20,442 20 , 362 18,416 18,359 18,298 18,275 17,182 17,100 16,929

102,e 102.7 102.9 10"3 104 103,9 103.6 100.9 100.9 100.6

7Fe 20,513 20,367 20,300 (sh) 20,177 (sh) 20,121 (sh) 'D4 'F6 18,440 18,345 (sh) 18,315 18,265 (sh) 18,172 6D4-+ 'F4 17,182 16,935

'D4

187

46,686

7 7

141

556,080

92, 1

.-+

-+

345

1,038

0 .2

102.6 102 6 102.3 101,~ 10',8 103.3 102.9 los,' 102.9 102.3

413

32,960

23.6

268

106,396

76.1

320

489

0.4

100.3

Volume 68, Number 2 1

November, 1364

N. FILIPESCU, W. F. SAGER, AND F. A. SERAFIX

3342

Table IV (Continued)

Terbium ion transition

Table V

%ax5

hmax,

Av,

Relative intensity,

om. -1

R.U.a

cm.-‘

A.U.b

Series A

Percentage total emission, %

DPyM ( 1 3 , 6 ) D1NM (12.7) D2NM (10.8) DFuM (10.8)

Terbium Trifluoroacetylacetonate Tb( CH3COCHCOCF3)S (TbTFAcA) ‘D4 -+ 7Fe 20,534 20,408 20,284 20,202 (sh) 6D4+ 7F6 18,433 18,399 18,349 18,282 18 , 165 18,067 5D4 -+ ’F, 17,188 17,138 17,021 16,935 Terbium -+

BA ( 9 . 0 ) 332

831,992

16.5

366 4,220,160

83.5

mM ( 6 . 1 ) mN ( 5 . 2 ) DmN ( 5 . 2 ) pPh ( 4 . 0 ) p M (3.3) D p S (1.5)

103.5

TTA ( 0 . 9 ) DpM ( 0 . 6 ) p x (0.5) 320

960

0.01

DBM ( 0 . 2 )

100.6

Hexafluoroacetylacetonate (TbHFAcA)

7F, 20 ,471 20,429 20,367 (sh) ’D4 + ’F6 18,467 18,423 18,355 18,298 18,268 (sh) ’D4 + 7F4 17,100 ‘D4

DpF (9.9) DmM ( 9 . 8 )

103,s

103,a 104.6 104.6 lo4.* 1053 104 103,~ 101.2 10 100.4

104 104,1

Series B

229

Tb(CF3COCHCOCF3)z

1,145 000 ~

BTA ( 0 . 3 ) DBM ( 0 . 2 ) DTM ( 0 . 0 ) HFAcA ( 0 0) TFAcA ( 0 . 0 ) AcA ( 0 . 0 )

30.8

Table VI 104,4 104.7 104.6 104.3

Series A

205 2,562,240

69.0

D l N M (13 1) DpF (12.7)

104.1

102

125

6,250

DPyM (12 6) D2KM (12 4) BTA (11 9 )

0.2

a The height of the peak on a linear scale in (arbitrary) recorder units. Relative intensity taken as ‘/pAvhmitx in arbitrary units (recorder units X cm.-l). The average height of a band L,, is taken as lOZh,/n, where h, is the height of the individual peak on the log I = f(v) scale and ni the number of peaks in a band. (Only the relative intensity of the strongest band is tabulated.) (ah) designates shoulder.

transition as percentage of total emission, the decreasing order in the two series (A and B) is given in Table VI. The KDo--t 7Fzline is the most intense line originating from the 5Do level for all chelates. The terminating level having J = 2 can be split into two sublevels in a field of Oh symmetry or into three sublevels in a Csv field. The recorded fluorescence spectra show singlets for the chelates investigated,s except for mhl, DmRI, D F u l I , BA, TTA (doublet), and D2N!\I (triplet), all belonging t o the asymmetrical c group or “less symmetrical” b group. I n series A, the 5Do transition is generally sharpened and intensified by methoxy substitution in -.f

The Journal of Physical Chemistry

Series B

pPh (10 1) BA (9 9 )

m F (9 7) TTA ( 9 . 2 ) DmK (7 5 ) mM ( 5 6 ) DmAI ( 5 5) DpN ( 5 4) DBM (4 3) P b T ( 4 1)

DBM ( 4 3) D T M ( 3 5) DFuM ( 3 3)

PN ( 3 . 1 ) DpM ( 1 . 5 ) HFAcA (0 0) TFAcA (0 0 ) AcA ( 0 0 )

both para or meta on dibenzoylmethane. h reverse effect (broadening and decrease in intensity) is observed on p a m or meta nitro substitution. If the rela~~

(8) It is possible, and in some cases very probable, t h a t further resolution of t h e fluorescent bands (i e . , by lowering the temperature) would reveal a band structure finer than t h e one observed.

is clxprcsscd as pcrcctntt'ivt: iiiteiisity of thc 5110 -+ agt: of tot a1 cniissiori, the europiuin chclat,os arrange as shown in Table VII.

Table VI1 Scries A

Srrirs I3

11pLI (!IO 0 ) pN (68 2 )

1312 (59 4)

solving power of the spcctrophotometcr to be distinguished or becausc. they may overlap. This transition was rclcorded as singlet for DpRI, DpN, ml1, DpV, I l T l I , IllNAI, D2KA1, Dl'ylI, EA, RTA, AcA, TP'AcA, and HVAcA; doublet for DBAI, pl1, pN, pl'h, IlI~uAI,and TTA; triplet for D m l I ; and quadruplet for 772N and DmX. The intensity of the 5Dl + 71", transition being expressed in percentage of total intensity, the order in the two series is shown in Table VIII.

T>m.\I (58.7) TTA (54 5 ) 111 N L I (50 0)

Table VI11

m h I (47 0)

I 3 T A (44 6 ) 1)2NLI (40 4 ) I)pS (32 0 )

~ ; L(30I 1) m N (21 7 )

I)FUM (31 3) 1)13A1 ( 8 0 1) i ) n r (21 0)

PA1 (18 5 ) I h S (17 8 ) I ) p V (14 4 )

Series A

Series I3

DTiM (72.8) pM ( 7 2 . 7 ) pPh ( 6 7 . 3 ) nnbr ( 6 3 . 0 ) J>mK ( 6 0 .1 ) IIpS ( 5 7 . 8 ) 1)pI' ( 6 5 . 4 ) m x (53.9)

DRM (63.0)

DPyM (49.9) TFAcA (46.1) HFAcA (40.7)

11PySI (10 9) pl'h ( 9 . 8 )

mM (39.8)

N o pure 5D0+ 7 1 2 transition was observed for AcA, ?'l;iicA, arid HE'AcA, but a broad band with A,,

locatcd sornrwherc between the expected wave lengths of "I, + 7V4 and 5D0 + 71"2. The band may be a resultant of the two, but the resolution of the band structure does riot pcrniit confirmation or rejcction of this possibility. Icluorescent bands corresponding to the 5D0 + 7Fj transition were recorded for only six chelates (pI'h, DFuAl, D l N h I , D2N;\IJ RA, and TTA), all of low relative intensity, and no bands were observed corresponding to the 6Do + 7F4, 5Do + 7175, or 6Do + 7F6transitions. From the fluorescent lincs corresponding to transitions originating a t the 5D1level, only those terminating a t 'IC4 and 7F'3 lower lcvels were consistently observed. In a few cases (pl'h, DE'uRI, DlNI4, BTA, TTA, and HE'AcA) one, two, or three bands were recorded i i i t hc SDl + 7F1frequency range. The s J l l + 'IF4line is the most intense line originating level, for all chelates. The terminating froin the 6D1 level having J = 4 can be split into four sublevels in a field of oh syrnnlc%ry or into six sublevcls in a Cav field, whereas the 5Dl lcvcl is unsplit i n oh and doubly split in Csvsymmetry. Actually riot all the possiblc lines are 0bscrvt.d either tmause t he diffcrcnce in energy bctwccm the sublcvrls is too small for the re-

IIFuhI (36 3 ) D2NM (29.6) BTA (29 4 ) pN (27.4) DmM (21.2)

TTA DlNM RA AcA

(25 (16 (1R (13

7) 0) 0) 0)

DpM ( 1 . 7 )

The 6D1 + 7173transition is weaker than 5D1+ 7F'1, but has been observed for all chelates investigated except the aliphatic diketones (AcA, TFAcA, HFAcA). The recorded line was a doublet (pPh, D2Nl1, and RA) or singlet (all the rest). The relative intensity (percentage of total emission) is given in Table IX.

Line Splitting in Fluorescence Spectra on Europium Chelates The different transitions in the europium ion arc affected in different ways by changing substituents in the surrounding ligand. The number of lines resulting froni the splitting of an individual transit ion and t he magnitude of this splitting are given in Table X. In series A , nionosubstitution in para results i n splitting of the 5D,, + 7F1transition into a doublet with a separation of 153 c m - * for pAI, 127 cxn-l for pN,

N. FILIPESCU, W. F. SAGER,AND F. A. SERAFIN

3 344

and 223 cm.-I for pPh, whereas the same transition for DBM and the two para-disubstituted DpM is a singlet. Similarly the 6D1 7F4transition is a doublet for p M and pN, whereas for the disubstituted DpR4 and DpN it is a singlet. The 5D0 4 'Fz line is split into a doublet only for the two meta-methoxy-substituted dibenzoylmethides (ml4 and DmM). pPh is quite different (as number of lines and magnitude of splitting) compared to the other substituted dibenzoylmethides.

Line Width of Fluorescence Transitions in Europium Chelates

+ .

The line widths of the fluorescence lines are primarily dependent on the nature of the ligand, namely on the substituents attached to the p-diketone chelate ring, and not on the location of the triplet state level or the efficiency of transfer to the ion. A similar variation is observed for the two transitions 7Fo(singlet) and 5D1 4 7F3(singlet with a few exceptions). The line width decreases on para-phenyl 7Fo substitution of dibenzoylinethide for the 5Do line, and does not change for the 6D1 + 7F3line. Significant line-broadening takes place on para disubstitution. The broadest 5Do+. 7Foand 6D1 4 7F3lines in series A are those from DpN, DpF, and DpM. Mononitro substitution in para does not change the line width of either 5D0-t 7Foor 5D1 7F transitions, and only a small broadening is observed on paramonomethoxy substitution. Dml4 has the sharpest 5D0 7Fotransition in series A, whereas mM has the sharpest 6Dl 7F3line. Actually, such a variation was not expected. For asymmetrical P-diketone chelates probably a mixture of stereoisomers is formed during synthesis, and due to' slight differences in the molecular field of these isomers, line broadening was expected, resulting from the emission of very close different frequencies. The observed result seems to indicate that either no stereoisomers are formed, and there are no obvious reasons to sustain this supposition, or, more probably, that the internal electric fields in different stereoisomers are essentially equivalent. The 5D1 4 7F8line in the aliphatic diketonates is narrower for the unsymmetrical TFAcA than for the other two symmetrical AcA and HFAcA. The strong 5D1 + 7F4and 6Do+ 'Fz transitions are affected by substituents in a different way than 5Do + 7Foand 5D1 4'Fa. Although their line width is more difficult to analyze due to splitting, some observations can be made. para-Substitution of electron-withdrawing groups (pN, DpN, DpF) on dibenzoylmethide broadens the 6D14 7F4and 5Do+. 7Fztransitions considerably, whereas inethoxy (in both para or meta) or meta-nitro substitution does not change the line width significantly ( p M , DpM, m M , DmM, mN, DmW). A continuous narrowing of this strong transition is observed for the aliphatic P-diketonates on replacing methyl with trifluoromethyl from AcA to TPAcA and HFAcA. This trend is probably due to a decrease of the large gap between the organic triplet and the 5D1 level of the ion froin 5000 to 4000 and 3000 c ~ n - l , respectively. The three five-membered heterocyclic ring-substituted -+

+ .

Table IX Series A

Series B

AcA HFAcA TFAcA DPyM

(39.7) (29.5) (27.5) (13.0)

mN ( 9 . 6 ) DmN ( 9 . 5 ) pPh ( 8 . 5 )

-+

-+

TTA (8.1) DFuM (8.0) DpF (7.6) DlNM (7.2) D2XM ( 6 . 4 ) BTA (5.9) BA ( 5 . 0 )

DmM ( 4 . 8 ) PM ( 4 . 6 ) DpN ( 3 . 3 ) D T M (2.6) DBM mM pls DpM

(2.3) (1.4) (0.8) (0.3)

DBM ( 2 . 3 )

In series B, the variations from one chelate to another are even greater. It does not seem to be a continuous change or obvious correlation with the different substituents, even for ligands having only different distant substituents. The only similar spectra were those of mN and DmN, and the three aliphatic diketones (AcA, TFAcA, HFAcA). Changes in number of lines observed and magnitude of splitting are very significant on going from one chelate to another. S o general rule relating the number of lines or the magnitude of splitting is obvious. On the other hand, it is very possible that some of the individual lines, components of split transitions, are overlapping or very weak compared to those observed. This would make Table X a rather incomplete basis for this discussion. The Jourhal of Phgsical Chemistry

+ .

SUBSTITUENT EFFECTS ON INTRAMOLECULAR ENERGY TRANSFER

Table X: Splitting in Fluorescence Spectra of Europium Chelates. 5D1 -+

Europium chelate

N.1.0.a

7Fc Au,h

cn-1

D

B M 2 9 9 pM 2 83 Dph4 1 .,, pN 2 1117 D p N 1 ... ... m M 1 DmM 3 11% 167 mN 4 DmN 4 154 60 pPh 2 D p F 1 ... DFuM 2 53 D T L M ~ ... ,.. DlXM 1 D2NM 1 ... DPyM 1 , . , TTA 2 60 BA 1 ... BTA 1 ... A c A 1 ... TFAcA 1 ... HFAcA 1 .,. a

Number of lines observed.

6Do

-

N.1.o.

1 1 1 1

7Fz

5Do -+

Av,

ern.-'

...

ED1

Awn

N.1.o.

cm.3

...

l . . . 2 153

. .

1

,

., I . . . ,

2 3 8 2 59 1 .,, 1 , , , 1 l .

'Fi

, , ,

. .

2 45 I . . . 1 3 1 .., 2 26 2 48 1 ... O . . . 0 .., 0 .,.

2 l I 3 2 2 2 I 2 I 2 3 1 3 2 1

,

.

.

. .

..

127 . . . . 204 124 130 223 . . 62 . . 90 37 ,

.. 99 99

O . . .

0 0

.,, ..,

--t

3345

Number of Observed Lines and Magnitude of Splitting

Transition 7F3 6Di AV,

N.1.o. c m . 7

-

N.1.o.

7F1

Av,

ern.-'

1 .. O . . . I. I. 1. O . . . l . . . O . . . I. . . . O . . . 1. 1 1. ... 0 ... 2 79 0 ... l . . . O . . . 1. 0 1 . . . 0 . . . 7 0 ... 2 2 226 1 O . . . 1 0 ... 2 2 216 1 ... 1 , . . l . . . 1 . . . 1 ... 1 .. . 1 ... 2 312

ID1

-

N.1.o.

Av,

.

O 0 O O

.

O

.

. 2 0

3 0 O 2 2 1 0 0 1

&Do-+ 7Fo

Aul

.

O

.

.

N.1.o.

.

Ah

om.-'

l . . .

. . O . . . l . . . . . 0 . . . 1 . . . . . O . . . l . . O . . . l . . . .

.

O

.

.

.

l . . .

...

0 3

O

.

. . .

'Fa

3.1.0. ern.-'

cm.3

O

-

'Do

7F1

1 ... 0 .., 1 ... 1 ... O . . . l . . . 1 ... 1 ,.. . . . 0 . . . 0 . . . 138 1 1 ... ... 1 1 ... . . . O . . . l . . . 181 1 .,. 1 , , . 139 1 , . . 1 .._ 0 1 . . . O . . . O . . . .., 0 ... 0 ... ... 0 ... 0 ...

334 . . 157

* Magnitude of splitting.

diketonates (DFuM, DTiM, and TTA) have slightly sharper lines in general, compared to the others. S o direct correlations between substituents and line broadening due to superposition of oscillatory transitions on pure electronic transitions could be made.

estimated as being approximately equal to 1/2Avhmax ( A v is the spectral range of the observed band in cm.-' and h,, is the average height of the band equal

General Considerations on Fluorescence Spectra of Terbium Chelates Fluorescence emissilon was observed for all terbium

units-of the individual lines of a band). The total intensity of fluorescence decreased in the two series investigated as given in Table XI (total intensity expressed as arbitrary area units). The very strong total emission of terbium ion in DpF and prill chelates (series A) could be interpreted as due to the ideal location of the 0 -, 0 band of the organic triplet us. the 5D4 resonance level, that is almost the same frequency and slightly above. This would imply a very efficient transfer of energy to the metallic ion. The comparatively poor ionic emission of DBM, pN, DpN, mM, DmM, mN, DmN, and pPh chelates is primarily due to disadvantageous location of the organic triplet, below the 504 terbium level. It is observed that nitro substitution decreases significantly the ionic emission in both meta and (more) in para positions. Disubstitution decreases the intensity more than mononitro substitution. This effect has also been observed for europium chelates, although for terbium the triplet states lie below the emitting level of the ion. Monomethoxy substitution in both meta

chelates investigated, although in some cases the triplet state of the ligand lies below the ED4 resonance level of the ion. I n such cases the ionic emission was much weaker compared to the other chelates where the organic triplet was equal to or above the 6D4 level. When the ionic emission was weak, the broad-band organic phosphorescence was intense, approaching the intensity values recorded for Gd chelates. The most intense fluorescence band was that corresponding to the 5D4-+ 7F5transition for all chelates. Bands associated with transitions to low-lying 7F levels with J = 6, 5 , 4 , and 3 were recorded for several complexes, but no transitions to 7F2, 'F1, and 7Fo were observed. The fluorescence bands are highly split in numerous individual lines as expected for transitions between levels with high J number. The intensity was evaluated in arbitrary area units equivalent to the area under the curve recorded, which was

n

h , / N , hi = lon',n, being the height-in

to

recorder

2=1

Volume 68, Number 11

November, 1964

N. FILIPESCU,

3346

Table XI Series A Di-p-fluorodibeiizoylmethide p-3lethoxydibenzoylmethide Dibenzoylmethide m-Nethoxydibenzoylmet hide Di-m-me t hoxydibenzoylmethide Di-p-methoxydibenzoylrne thide p-Phenyldibenzoylme thide Di-m-nitrodibenzoylmethide m-Nitrodibenzoylniethide p-Ni trodibenzoylmethide Di-p-nitrodibenzoylrnethide

32,211,060 3,053 356 888,736 368,465 196,861 76 660 15, 105 15,105 10,123 2,835 2,000

Series B

Trifluoroacetylacetoriute Hexaflnoronce ty lacetoriate Benzoylacetonnte Dibenzoylmethide Theonyltrifluoroaretonate Difuroylmethide Benzoyltrifluoroacetonate Acetylacetonate Iliison icotylmethide Ditheonylmethide IX-2-naphthoylmethide Di-1-naphthoylmethide

5,053,112 3,713,490 1,632,,589 898,736 603,804 376,488 257,922 139,845 80,896 19,000 1,850 56

or para leads to more intense fluorescence than the respective disubstitution, aniong other reasons being the increased asymmetry of the molecular field.

The Journal

of

Phvsical Chemistry

w.li‘. S A G E R , AND li‘. A. SERAFIN

In series 73, the poor ionic emitters DTJI, DlNRI, and D 2 N i chelates have the organic tAplet well below the 6D4 ionic level. It is interesting to observe that cven in such situations sonic energy is transferred to the ion, although most of it is radiated as organic phosphorescence, which approaches the values obtained for Gd chelates. The partial transfer of energy in those cases may take place from higher triplet states formed by T + l” absorption. The asyninictrical TIi’AcA is a niorc powerful ionic emitter than the sytiitnetrical HFAcA, although the latter has the organic triplet closer t o the jDI ionic level. The other aliphatic P-diketonate (AcA) is a comparatively poor w i i t ter justified by its large energy gap between the organic triplet and 5D4 level. Just as for europium chelates, HA has a more intense fluorescence than HTA, probably because the trifluoroinethyl group tends to increase the ionic character of the nietal-oxygen bond through negative inductive effects. The splitting of the fluorescence bands is rather extensive as expected for transitions involving levels of high J value with much overlapping between individual lines, niany of which, however, arc remarkably sharp. Acknmoledgment. Part of this work has been carried out a t Melpar, Inc., in Falls Church, Va.