FLUORESCENCE SPECTRAAND LIFETIMES OF Sm3+ (1 16) on glass. The tests of the theoretical curves of Figures 4 and 8 were effected a t X 445 and 495 nm, these being the absorption maxima of the dimer and monomer, respectively, of the T M C dye. Figure 12 deals with LB layers of pure barium arachidate which is a nonabsorbing substance a t 632.8 nm. The full line (curve b) almost fits the experimental data. The dashed curves a and e show that A depends strongly on the difference n, - n,. This demonstrates nicely the sensitivity of the method. Since the variations of $ did not exceed I”(compare Figure 2), we have not made a plot of the experimental $’s. The thickness of one layer of barium arachidate prepared under the same experimental conditions was determined very accurately a t 2.660 nm by (reflection) ellipsometry.1° The experimental resuN s of transmission ellipsometry of an absorbing anisotropic film are depicted in Figure 13. Since this film of TNC-barium arachidate (1 :16) was anisotropic in both the index of refraction and the index of absorption, A as well as $ were strongly
3397 affected by an increase in the film thickness. The thickness of one layer of this mixed film was assumed t o be 2.74 nm.9 The “best fit” of the optical constants with the experimental data (inserted circles) is n, = 1.56, n, = 1.50)K , = 0, and K~ = 0.052. I n Figure 14 the dichroic ratio D,/D, of the same film a t 445 nm has been plotted us. the film thickness. The theoretical curves a and c show the sensitiveity of D,/DB to small changes of thc optical constants of the film. Curve b almost represents the “best fit”. Figure 14 alone does not give sufficient information to calculate both K, and K~ and to estimate n, and n,, Since D, and D,are determined separately, K , can be calculated directly from the experimental values of D,. The dichroic ratio a t another wavelength (495 nm) is qualitatively the same as that shown in Figure 14.
Acknowledgments. I am indebted to Dr. E. P. Honig for many discussions and for his critical reading of the manuscript. The experiments were partly performed by ICIr. E. Davelaar.
Fluorescence S ectra and Lifetimes of Sm3+ in POC13-SnC14
by P. Tokousbalides and J. Chrysochoos” Department of Chemistry, University of Toledo, Toledo, Ohw
48006
(Received J u l y 6 , 19’72)
PuSlicatwn costs assisted by Owens-Illinois (Corporate Research Laboratories) Toledo, Ohio
The fluorescence efficiency of Sm3+in POCls-SnC14 is markedly enhanced relative to the efficiency in aqueous solutions. Both the fluorescence efficiency and fluorescence lifetime depend very. strongly om the ratio of POC13and SnCL. Optimum results are obtained a t a molar ratio of about 10: 1. ,4t both low concentrations of Sm3+ and low [P0Cl~]/[SnClr]ratios the fluorescence lifetimes approach a high value of 3.3 msec. At higher concentrations of Sm3+ considerable self-quenching is observed with a self-quenching rata constant, kol of about 8 X 102 1M-l sec-l. A discussion of the above data is presented.
Introduction The fluorescence efficiency of the rare earth ions in aqueous solutions and in various organic solvents depends on the energy difference between the lowest excited state and the highest ground state of the rare earth ion.lb2 Furthermore, the higher the energy of the vibrational modes associated with the solvent, the lower %hefluoresccnce efficiency of the A decrease in the iiuorescence efficiency of Eu3+ in organic solvents was attributed to the radiationless deexcitation of Eu3+via the overtones of the - 4 - E t and -0-H groups of the solvent molecules both in the primary and secondary mlvation sphere of the rare earth i ~ n . ~ , ~ The ratios of the fluorescence intensities of the rare
earth ions in DtO and in DzO-H~O,i e . , II)/ZD+H, was found to increase in proportion to the concentration of the hydroxyl group. 3,6 Similar results were observed with rare earth ions in alcohols.’ These results indicate a nonradiative transfer from the lowest excited state of (1) A . Heller, J . Mol. Spectrosc., 2 8 , 208 (1988). (2) V. L. Ermolaev and E. B. Sveshnikova, Opt, Spektrosk., 2 8 , 98 (1970). (3) A. Heller, J. A m e r . Chem. SOC.,88, 2058 (1966). (4) J. Chrysochoos, Spectrosc. Lett., 5 , 57 (1972). ( 5 ) J. Chrysochoos, Chem. Phys. Lett., 14, 270 (1972). (6) J. L. Kropp and M. W. Windsor, J. Chem. Phys., 45, 761 (1966). (7) N. A. Kazanskaya and E. B. Sveshnikova, Opt. Xpektrosk., 2 8 , 376 (1970). T h e Journal of Physical Chemistrg, V o l . 76, *Vo. BS, 1972
P. TOKOUSBALIDES AND J. CHRYSOCHOOS
3398 the rare earth ion to the overtones of certain vibrational modes of the solvent. Such a mechanism, however, has been accepted on qualitative rather than quantitative evidence. The energy difference between the 4G6/2 and 4F~1/2 states of Sm3+ is about 7400 cm-l.* This energy difference matches some of the low overtones of the stretching vibrational modes of such groups as -tC-H, -0-ET, -tC-D, - O-D, etc. I n solvents containing such groups, Sm3+should exhibit a very weak, if any, fluorescence efficiency. This is the case. On the other hand, Sm3+should exhibit a high fluorescence yield in solvents which do not contain groups with high energy vibrational modes. Such solvents include POCL in combination with SnC14, ZrCL, SbC15,etc. The usefulness of such solvents has been confirmed in the case of Nd3+.g*10
"O0T. I50
Experimental $&io Fluorescence spectra were measured with the Aminco Bowman spectrophotofluorimeter using an RCA-1P21 photomultiplier and a 200-W xenon-mercury lamp. Absorption spectra were taken using the Cary-14 spectrophotometer. Fluorescence lifetimes were obtained by exciting the samples with a xenon flash lamp whose flash duration was about, 20 Msec. The excitation light was filtered with a UG-5, ultraviolet transmitting, glass filter. The emitted light was also filtered with a filter whose peak transmission was a t 6000 with a halfwidth of about 125 8 (Optics Technology, Inc., set No. 1018; filter KO.800). Therefore, the fluorescence lifetimes measured are associated with the emission band of Sm3+a t 595 mpj Le., 4G6,, 0H7,2. The rare earth ion was used in the form of Smz03 (99.9-99.99%). Ultrapure POCL and SnCll were used with no further purification. Samarium oxide mixed with appropriate amounts of POCls and SnCh dissolved slowly at 70-80" under reflux conditions and in the absence of water vapor and Con. 'The reaction was completed within 24 to 48 hr, giving rise to transparent solutions containing up to 3 X 10-I M Sm34-. All samples were kept in a dry box. --f
Results and Disciissisn Typical fluorescence spectra are shown in Figure 1A. This figure a1so contains the fluorescence spectrum of 6m3+ in aqueous solutions for comparison. An enhancement in the fluorescence efficiency of Sm3+ in POC13-SnC14 relative to that in aqueous solutions is apparent. The ratio of the fluorescence intensities in the two Bystems is of the order of lo3. A further enhancement, by about a factor of 3, is observed by lowering the temperature from +25 to - 2 " . Fluorescence spectra a t lower temperatures were unattainable because the freezing point of the samples under consideration was about - 2 " . Variation in the composition of the POCldhCPd system has a considerable effect upon the fluorescence spectra of Sm3+. Such an effect is The Journal of PhysicaJ Ciiemistry, Vol. 76,No. 93, 1978
L I (
Figure 1. (A) Fluorescence spectra of 2 X 10-1 M Sm3+, uncorrected; A,, 402 mp: (a) Sm3+in POC13-SnCln ([P0CI3]/[SnCl4]= 10.34) a t -2'; (b) Sm3+in POC13-SnC14 a t 25" ; (6) Sm3+in aqueous solution. (B) Fluorescence spectra of 4 X M Sm3+in POC13-SnCL; A,, 402 m p : (a) [P0Cl~]/[SnCl4]= 10.34; (b) [POCla]/[SnCla] = 13.24; (c) [POCl,]/[SnClaJ = 27.76.
shown in Figure 1B. The fluorescence efficiency of Sm34 increases as the [POCl3]/ [SnCLj decreases. However, the concentration of SnCl4 has to be within certain limits to attain clear solutions. The rare earth salts are insoluble in both pure Poci3 and SnCh but they are fairly soluble in certain combinations of these two solvents. The conductivities of both pure POC13 and 1 X lom8 and SnC14are very low," i.e., 1.6 X 52-' cm-', respectively. However, POCL with 6% (in moles) SnCh exhibits conductivity of about 6 X W1em-', indicating the presence of the ionized forms of both POC1, and SnCla, possibly via 2P0CI3 f SnClr +2POC1zSSnC~e2-
A complex, 2POC13 SnCla, precipitates a t the beginning, but it is dissolved in an excess of POCI,. Therefore, it appears as if the P0Cl2+ion may react, with the oxide (8) G. Ha Dieke and H. M. Grosswhite, A p p l . Opt., 2 , 675 (1963). (9) A. Lempicki and A. Heller, AppZ. Phys. Lett., 9, 108 (1966). (10) N. Blumenthal, C . B. Ellis, and D. Grafstein, 1. Chem. Phys., 48, 5726 (1968). (11) F. Collier, H. Dubost, R. Kohlmuller, and C. Raoult, C. R. Acad. Sci., Ser. C , 267, 1065 (1968).
339
Bm&
+ BPOC&
1-
-+%m3+
+ 4POCla + PzOs
e 5m3+ formed is then associated with the SnC1628. .fit 1c1w ~ o ~ ~ e i ~ t r a t iofo nSnCle s there is not sufficient ~ ~ o ~ ~ of ROC&+ c ~ ~to react ~ t with r ~SmzO3, t ~ ~ ~ ~ ~ wheresas a t high eo ntraiions of SnC14 the complex * snc'14 preoi tes. Therefore, the variation in [SnC141 has to kse confined within these two limits. l'he effecl of &he jb 6=Bs]/ [SnCL] ratio upon the fluoreicence lrfetirneii i dmwn in Figure 2A. The flixare:soencc lil'etiares increase 8s the ratio decreases. This change in tAe IP5)@13]/[SnCL] ratio reflects a cons ~ d e change r ~ ~ in~ Lhe ~ ~ o ~ ~ c e n t r of a t SnCL ~ o ~ but a rat hcr ~~~~~~~~~~1~ charge in the concentration of POCL. C the Biiorcseenee lifetimes upon 1 j~ shown in Figure 2B. Since it is sather unlikelj, that :s IO?& xnorease in [VOCl,] can be a s ~ o c ~ a wit t e ~n the ~ ~ ~ r ~ ofa ~ thei ofluorescence n lifetinses okrscnwB, i: iieenis more reasonable that such a arariatioir may ireiwte t o [SnCL]. If we visualize Sm3+ ions -,urrormder4hy SiaCl2- ions, the quenching effect of such a solvntio.;l sphere will be insignificant due to the low energy ~ ~ ~modcs~ involved. , ~ At lower ~ tcon- ~ ~ ~ ~ ~ ~ of ~ L ~solvation ~ osphere n may ~ 8 n CY,,~ Lhc~primary contain alw I'OCls motecufes. Such molecules could ~~~~~~
00
~
a
~
I'0
Figure 3. (A) Ultraviolet absorption spectra of 4 X M Sm3+ in POC13-SnC14 us. air; L; = 1 cm: (a) [POC13]/[SnC14 = 13.24; (b) [POCl3]/[SnCl*]= 56.81. (E) Fluorescence intensity (normalized) of Sm3+ in POCl3-SnC14 ([POC13]/[SnC14] = 10.34); A,, 402 mp: 0, Xi1 595 mp; e, A f l 560 mp.
;I
/
/ /' / o
IO.?
9.7
10.2 G---
cpocfJ,M
Figure 2. (A) Fluorescence decay of 4 X 10-2 M Sma+ in POCla-SnCla, he,, 300-450 m p ; X f l 590--630 mp: (a) [POCl3I/[SnClrj = 10.34; (b) [POC13]/[SnC14]= 13.24; (c) [P0613]/[SnClB,, = 66 81. (B) Fluorescence lifetimes of 1 X lom3M Sma as a function of [SnClr] and [POC13]. N
greatly enhance the quenching effect of the solvation sphere owing t o the participation of the relatively high energy stretching vibrational mode of +P=O. This will result in a marked decrease in the fluorescence lifetime. If the change in the solvation sphere is very drastic, it may be accompanied by appropriate changes in the transition probabilities. Under these conditions the change in the fluorescence lifetimes will be much more complex. I n the cases examined in the present study, the absorption spectra, which furnish a direct measure of the transition probabilities, were practically identical (Figure 3A) in the spectral region which vias used for the excitation, Le., 390-410 mk. Therefore, it seems very likely that the effect of the [POCla]/ [SnC14] ratio upon the fluorescence lifetimes is associated with a radiationless deexcitation of Sm33 via the vibrational modes of the solvent. If a solvation sphere i s established containing only SnCP-, no further effect should be observed. This can be verified in Figure 2B. The fluorescence lifetimes increase rapidly up to 0.6 M SnCh and then level off. An extrapolated value of -T = 2 . 2 msec a t [SnCL] = 0 represents the fluorescence lifeThe Journal of Phgsical Chemistry, Vol. 76, N o . 28, I972
P. TOKOUSBALIDES AND J. CHRYSOCHOOS
3400
enhancement of about 1000 is shown for the fluorescence efficiency of Sm3+in POCL-SnC14 relative to that in aqueous solutions. I n the presence of self-quenching the spectroscopy of Sm3+in POC13-SnC14can be outlined as
hv
Sm3f(4F7/%)
Sma+(6H5/2)
Sm3+(4F7/2) -+ Sm*f(4Gs/2) Sm3+(4G5/,)-+ Sm3+(6H)4-h m ;
kfl
+
Sm3+(4G5/2) solvent -+
Srn3+(% or 6F>C heat; k,o~vsea Sm3+(4Gs,,)3. Sm3+(6H,/,)-+ Sm3+(6H)
+ Sm3+(6F);
ICQ
The aforementioned mechanism lead to the following expression for the fluorescence efficiency
-
'#Q
U
=
kfl/(kfl
t
+ + kh
ksoiveeC[solvent ]
0 .
'b
-+ kQ [Sm3+])
(1)
or for the fluorescence lifetime 20
_-
I -
7Q-l
1'2
24
I O ~ X C S ~ ~ +M ],
Figure 4. (A)Fluorescence decay of Sm3+ in POCls-SnClI ([PQCS] /[SnCk] = 10.34); A,, 4 0 0 - 4 5 0 mr; Xrl -590-630 mp: (b) 1 X 10-l M Sm3+;(c) 5 x loF2M Sm3+;(d) 2.5 x M Srn3+;(e) 5 x 10-3 M Sm3+; ( a ) 1 x 10-1 M Sm8+ in K O - S i Q d h O glass. (B) Variation of the reciprocal fluorescence lifetime, r - l , of Sm3+ in POCl&hClr ([POCl,]/[SnCl,j = 10.34) as a function of [Elm*+].
time of Sm3+ in 100% POc13 provided the transition probabilities are the same. The effect of the [POC13]/[SnCL] ratio upon the fluorescence spectra and lifetimes were studied at low concentration3 of Sm3+ to avoid self-quenching of excited Sm3+ ions by neighboring Sm3+ ions in their ground state Such an effect is shown in Figure 3B where the fluorescence intensity, I p / ( l - 10-'Ct), is plotted against [Srn3+]. The effect of [Sm3+]upon the decay of fluorescence is similar (Figure 4A). At very low concentrations of Sm3+in POCb-SnCl4, the fluorescence lifetimcs approach the lifetime of Sm3+ in the glass, Le., 3.3 msec for Sm3+ in POC13-SnC14 and 3.7 rnsec for Sm3I- in the glass. The fluorescence lifetimes of Sin3+in CM30€I,CzN50H, C3H7OH, and HzO were found to be 10.5, 110, 9-0, and
