J. Phys. Chem. 1981, 85, 751-753
The behavior in Figure 1 is consistent with this picture. Though our arguments are rather qualitative, they do appear to be consistent with our observations, and those of Brawer and Weber. Phase separation is known to occur in alkali-borate glassesz4and may affect the spectral properties at high 5Do-7FoEu3+transition energies.I3 The smooth undistorted inhomogeneous line shapes observed for the glasses used in this study suggest that the Eu3+ ion resides in a continuum of sites. Furthermore, our smooth linear frequency dependency of the homogeneous line width seems to continue in the energy region previously proposed to have spectral properties from the new phase. This might suggest that the homogeneous width is similar in this energy range for the two phases.
Conclusions We have performed fluorescence line-narrowing experiments on a variety of Eu3+doped glasses, and on borate glasses with controlled changes in Na20 modifier concen(24) R. R. Shaw and D. R. Uhlmann, J. Am. Ceram. SOC.,51, 377 (1968).
75 1
tration. The importance of the tetrahedral unit structure of the glass network in determining the homogeneous width has been shown by the monotonic increase in the homogeneous width of a given 5Do-7Fopacket with increasing Na20 concentration in the borate glasses. The variations in the inhomogeneous width in the borate glasses with the NazO concentration has been discussed in terms of the variation in the distribution of trigonal and tetrahedral units around the Eu3+site. The linear frequency dependence of the homogeneous line width for different packets within the 5Do-7F,, inhomogeneously broadened line can be explained in terms of the variation of the TLS coupling constant with the crystal field of the different sites. These explanations are consistent with experiment and exhibit the close relationship between the static structural and dynamical properties of glasses, as seen through the properties of optical impurity transitions. Acknowledgment. The authors thank Drs. S. A. Brawer and M. J. Weber for many stimulating discussions during the course of this work. The support of the Office of Naval Research (ContractNo. Nw14-75-0245 for E.C. and R.O., and Contract No. N00014-75-C-0602for J.N., W.H., and M.A.E.) is gratefully acknowledged.
Laser Excitation of SF6 in a Transparent Nozzle Beam: Collision-Assisted Absorption and Up-Pumping M. I. Lester, D. R. Coulter,+ L. M. Casson, G. W. Flynn, and R. B. Bernstein’ Department of Chemisfty, Columbia University, New York, New York 10027 (Received: December 9, 1980; In Final Form: January 28, 198 1)
Irradiation of SF6 in a transparent capillary-nozzlesupersonic beam source by a low-power cw cozlaser has provided molecular beams with a laser-induced increase in total energy of -0.6 eV molecule-’ (corresponding to the absorption of approximately five laser photons). Some 80% of this energy increase is retained as internal excitation of SF6.
Previous experiments have shown’ that internal energy can be deposited in a beam of SF6by irradiation with a low-power cw COz laser in the “collisional region” at the nozzle exitS2v3Greatly enhanced excitation has now been achieved by irradiating inside the nozzle. Both arrangements take advantage of molecular collisions to overcome inherent weak absorption in polyatomic molecule beams. Collisions assist by broadening absorption lines, promoting rotational hole filling, and providing up-pumping through (V-V) energy transfer. Following irradiation, excited beam molecules move rapidly into the collision-free region, freezing much of the absorbed energy in vibrational modes. The previous apparatus’ has been provided with an infrared-transparent, capillary-nozzlebeam s o u r ~ e .The ~ nozzle is a triangular cross section, 10-mm long capillary formed by clamping a flat ZnSe plate on a polished block of stainless steel containing a V-shaped channel (0.2 mm deep X 0.3 mm wide). The flowing gas within the capillary is irradiated (“double pass”) by a cw C02laser at 10.6 pm J e t Propulsion Laboratory, Caltech, Pasadena, CA 91125. 0022-3654/81/2085-0751$01.25/0
[P(16), 5.5-8.5 W; area -1 cm2]focused with a 0.4-m focal length lens at any given point along the capillary upstream or downstream of the exit. The electron bombardment ionizer-quadrupole mass filter monitors the number density n of SF6in the beam and provides time-of-flight (TOF) measurements of velocity distributions. The energy flux (power, W) carried by the beam is measured with the pumped liquid He-cooled bolometer.’ For many different irradiation positions, mass spectrometer and bolometer measurements of the mechanically chopped (25 Hz) molecular beam were recorded (via lock-in detection) with and without cw laser excitation. TOF measurements were (1)D. R. Coulter, F.R. Grabiner, L. M. Casson, G. W. Flynn, and R. B. Bernstein, J. Chem. Phys., 73, 281 (1980). (2) Confirmation of this result comes from the electron diffraction observations of COP laser-irradiated jets of SF6 by L. S. Bartell, S. R. Goates, and M. S. Kacner, Chem. Phys. Lett., 76, 245 (1980). (3) Somewhat related experiments on energy deposition of SF6 jets by COz laser irradiation have been reported by A. A. Vostrikov, S. G. Mironov, A. K. Rebov, and B. E. Semyakin, Sou. J. T e c h . Phys., 49,2680 (1979). (4) B. J. C. Wu and G. A. Laguna, J. Chem. Phys., 71, 2991 (1979).
0 1981 American Chemical Society
752
The Journal of Physical Chemistry, Vol. 85,No. 7, 1981
6
3
-Con
LASER PO61 10.6pm
A a.5w
4-
\
-I
C
1.0 I
I
I
I
I
I
I
I
I
I
I
I
1
I
I/,\
E
v
400
7 Y
P (3
a W
z w X
c
t
.-
'il
cy
tation, and the TOF results (uB,aJ. (All of the results shown are for a pressure of 205 torr of SF6in the oven.) Irradiating the beam downstream of the exit yielded small effects similar to those observed with the original n0zzle.l1~ Irradiation inside the nozzle induced a large enhancement (ca. sixfold) in bolometer output accompanied by an increase in velocity of the beam molecules and an increase in their number density. These effects are seen to be maximized 1mm upstream of the exit. Laser irradiation of pure Ar or of SF6at wavelengths far from resonance yielded no enhancement of bolometer signal. Under all irradiation conditions with or without SF6 in the capillary the temperature rise of the stainless steel nozzle was merely a few degrees, even over the course of many minutes of continuous irradiation. The total,translational, and internal energy per molecule in the beam, calculated from the data as in ref 1, are also shown in Figure 1. Irradiation inside the nozzle yields an increase in both the average kinetic energy (K) and the average internal energy (I). At the optimal irradiation position the increment in average total energy (T) due to laser excitation is -9.5 X J molecule-l (4760 cm-l, 0.6 eV), equivalent to approximately five 10.6-pm photons. Of this absorbed energy -80% is retained as internal excitation. The results are interpreted via a collision-assisted absorption and up-pumping mechanism. During ezrpansion the pressure and thus collision rate decreases along the capillary. As the point of irradiation is moved "upstream", the collision frequency in a given irradiation volume increases, resulting in increased absorption.' The high internal excitation achieved by irradiating inside the nozzle also implies that many vibrational states are collisionally coupled to the v3 pumped state of SF,. However, when the collision frequency is further increased, V-T relaxation and energy transfer to the walls limit the energy retained in SF6 vibration. The increase in kinetic energy with laser excitation indicates that at least some V-T relaxation has occurred before SF6molecules reach the collision-free region of the molecular beam. This observation is consistent with expectations based upon the time scale for relaxation in the bulk gas phase5 at a pressure in the range of 50-75 torr, comparable to that estimated for the SF6in the capillary immediately upstream of the nozzle exit. Irradiating far upstream (in the high-pressure region) should simulate bulk heating, yielding equilibration among all degrees of freedom. The extent to which laser excitation near the nozzle exit produces a nonequilibrium distribution which differs from that for bulk heating is not yet known. Experiments using a heated nozzle source are underway to study this question. The present excitation technique provides a highly energetic molecular beam whose energy content can be rapidly modulated, synchronously with the laser. Problems such as surface reactions and thermal decomposition associated with conventional heated-oven sources are avoided. This excitation method should be of general applicability to many molecules and is of potential use in reactive scattering studies of vibrationally excited species6 and multiphoton excitation diagnostics.'
-
r'
2
Letters
l(f)
I
o , 8 8 1 LASER
O6
IRRADIATION POSITION ( m m )
OFF
Flgure 1. Plotted as a function of laser irradiation position along the infrared-transparent nozzle are (a) the relative change in bolometer ( W,/ W,) and (b) mass filter (n,/n,) signals (open circles: bolometer data corrected for "laser scatter"'); (c) the stream velocity v, and (d) the width parameter a, from TOF velocity analysis; (e) the calculated average total (T), internal (I), and kinetic (K) energy per beam molecule (open points indicate use of interpolated velocity distributions); and (f) the ratio of the increment in I (i.e., A€'"'')to that in T (A€''') due to laser excitation.
made at fewer positions of laser irradiation; thus velocity distributions were interpolated at intermediate positions. Separate bolometer experiments have shown that laser scatter along the capillary axis is undetectable except when the laser irradiates the exit surface of the nozzle. Figure 1 shows the following data plotted vs. laser irradiation position: the relative change in bolometer ( WL/ W,) and mass filter (nL/no) signals upon laser exci-
(5) (a) J. I. Steinfield, I. Burak, D. G. Sutton, and A. V. Nowak, J. Chem. Phys., 52, 5421 (1970); (b) R. D. Bates, J. T. Knudtson, G. W. Flynn, and A. M. Ronn, Chem. Phys. Lett., 8 , 103 (1971); (c) W. D. Breshears and L. S. Blair, J. Chem. Phys., 59, 5824 (1973). (6) U. Agam, M. Eyal, and F. R. Grabiner, Chem. Phys. Lett., 68,35 (1979). ( 7 ) P. A. Schulz, A. S. S u d b ~E. , R. Grant, Y. R. Shen, and Y. T. Lee, J. Chem. Phys., 72, 4985 (1980).
J. Phys. Chem. 1981, 85, 753-754
Acknowledgment. The authors thank Dr. K. T. Wu for help with the experiments and Dr. F. R. Grabiner and Professor L. S. Bartell for valuable discussions. This research received financial support from NSF Grant CHE-
753
77-11384 (to R.B.B.) and D.O.E. Contract DE-AS02-ER7843-02-4940 (to G.W.F.). Partial equipment support was provided by NSF (CHE 77-24343) and JSEP (DAAG2979-C-0079) (to G.W.F.).
o-Carboranylcarbene. A Largely Localized Divalent Carbon Intermediate Richard S. Hutton, Helnz
D. Roth,”
Bell Laboratories, Murray Hill, New Jersey 07974
and Sarangan Chari Department of Chemistry, Princeton University, Princeton, New Jersey 08544 (Received: December 1 1, 1980)
Triplet EPR spectra of o-carboranylcarbeneand its 2-methyl derivative have been observed at 5 K in frozen solutions. The zero-field splitting parameters indicate little delocalization of the ?r spin density onto the carboranyl moiety, in contrast to the results observed for benzenoid aromatic moieties.
We have observed EPR spectra for two icosahedral carboranylcarbenes (2a,b), following photolysis of the
/H
N2\
‘C
I
I
2
a) X = H
a) X = H
b) X=CH3
bl X=CH3
corresponding carboranyldiazomethanes (la,b),at cryogenic temperatures in frozen solutions. The zero-field splitting parameters indicate a relatively high ?r spin density on the divalent carbon atom; apparently, the carboranyl moiety does not delocalize the adjacent ?r spin density very efficiently. Even before the publication of their synthesis,l the icosahedral carboranes were described as superaromatic on the basis of extensive molecular orbital calculations.2 Such a description seemed justified in view of the extensive electron delocalization in these compounds. This prediction is born out by a reactivity pattern not unlike that of benzenoid aromatics. Icosahedral carboranes were found to undergo “aromatic” substitution reactions3 and ocarboranylcarbene showed thermal rearrangement reacof phenylmethylene and derivative^.^ t i o n ~typical ~ However, in many other features, carboranes exhibit substantial differences from benzenoid systems. The IH (1) Heying, T. L.; Ager, J. W.; Clark, S. L.; Mangold, D. J.; Goldstein, H. L.; Hillman, M.; Polak, R. J.; Szymanski, J. W. Inorg. Chem. 1963,2,
1089. (2) Hoffmann, R.; Lipscomb, W. N. J. Chem. Phys. 1962, 36, 3489. (3) Beall, H. In “Boron Hydride Chemistry”; Muetteries, E. L., Ed; Academic Press: New York, 1975; Chapter 9. (4) Chari, S.; Agopian, G. K.; Jones, M., Jr. J. Am. Chem. SOC.1979, 101, 6125. (5) Jones, W. M.; Brinker, U. H. In “Pericyclic Reactions”; Marchand, A. P.; Lehr, R. E., Eds.; Academic Press: New York, 1977; Vol. I, Chapter 3. 0022-3654/81/2085-0753$01.25/0
NMR spectra fail to show the downfield shifts observed for benzenoid aromatics; indeed, the ‘H resonance of ocarborane occurs at higher field (6 3.5 ppm)6 than even vinylic protons. Similarly, the I3C resonance (6 57 ppm)’ is quite close to shifts characteristic of aliphatic carbons. Another characteristic of benzenoid systems is their ability to delocalize the electron spin in a a orbital of a a radical or a carbene. For a a radical, this ability manifests itself in sizeable hyperfine couplings for nuclei of the aromatic moiety; for a carbene, the reduced a spin density a t the divalent carbon results in reduced zero-field splitting parameters. The EPR spectra of o-carboranylcarbene (%a),and of its 2-methyl derivative (2b), were observed after UV irradiation of the corresponding carboranyldiazomethanes (la,b), in frozen solutions of methyltetrahydrofuran or propylene carbonate at 5 K. The spectrum shown in Figure 1 is typical of both carbenes. The zero-field splittings are DBa= 0.6860 and Eaa= 0.0302 cm-l and DBb = 0.6820 and E P b = 0.0293 cm-l. The difference between the values of D of the two compounds is less than 1% and is smaller than the differences usually observed for a particular carbene in different matrices.* The observation of triplet spectra at temperatures as low as 5 K indicates that the triplet is the ground state or lies within a few calories of the ground state. The values of D and E decrease by less than 1%as the temperature is raised to 40 K, a result which is incompatible with a substantial increase in the motion of the C-H fragment. Comparison of the D values of the o-carboranylcarbenes with that of a carbene having a completely localized a electron, methylene (DCH2= 0.76 cm-l)? indicates that the extent of 7c delocalization onto the carboranyl group is small. A ?r spin density, p = 0.9, is estimated for the divalent carbon from the ratio of D , to DcH2. In contrast, (6) Stanko, V. I.; Khrapov, V. V.; Klimova, A. I.; Shoolery, J. N. Zh. Strukt. Khim. 1970,11, 542. (7) Todd, L. J.; Siedle, A. R.; Bodner, G. M.; Kahl, S. B.; Hickey, J. P. J. Mug. Reson. 1976, 23, 301. (8) Trozzolo, A. M.; Wasserman, E.; Yager, W. A. J.Chem. Phys. 1964, 61, 1663. (9) Wasserman, E.; Hutton, R. S.; Kuck, V. J.; Yager, W. A. J.Chem. Phys. 1971,55, 2593.
0 1981 American Chemical Society
