4892
J. Phys. Chem. 1984, 88, 4892-4895
Relativistically Parameterized Extended Huckel Calculations. 8. Double-( Parameters for the Actinoids Th, Pa, U, Np, Pu, and Am and an Application on Uranyl Pekka Pyykko* Department of Chemistry, University of Helsinki, 001 00 Helsinki, Finland
and Leif Laaksonen Department of Physical Chemistry, Abo Akademi, 20500 Abo (Turku), Finland (Received: January 12, 1984)
Double-{ Slater type fits to the radial, relativistic and nonrelativistic Hartree-Fock functions are given for the neutral atoms Th, Pa, U, Np, Pu, and Am. An application on uranyl indicates increased U ( 5 f ~ ) - O ( 2 p ~bonding, ) compared to earlier relativistic s i n g l e t or nonrelativistic multiple-t calculations. The current understanding of bonding of uranyl is discussed. A possible CI mechanism for obtaining an eventual tlg HOMO for UF,, instead of a t,, one, is pointed out.
1. Introduction Radial wave functions of a t least double-!: (DZ) quality a r e needed to adequately describe the d and f shells of transition metals in various semiempirical calculations. No such wave functions have been available for the actinoids. The applications of the "relativistically parameterized extended Huckel" (REX) method1,* on various uranium compounds3were carried out using the single-!: (SZ) default functions in the program.2 The purpose of the present paper (considered as part 8 of the REX series; part 7: ref 4) is to provide such parameters for the actinoids T h to A m for use elsewhere. Secondly, we discuss in section 3 certain aspects of the uranyl ion UOZ2+using these parameters (for summaries of recent literature on uranyl, see ref 5 and 6).
2. Parameters The atomic orbital energies in Table I and the radial wave functions were obtained from average-of-configuration Dirac-Fock calculations7 on the neutral atoms using Desclaux's program.* In the radial fits in Table 11, a ncdeless, normalized, two-component function P i ( r ) was fitted to the radial four-component density p(r) = Pi2 Q? a t ro and m with ro outside the last node, minimizing the quantity
+
The uranium fits are shown in Figure 1. The others were of comparable quality. For the elements Pa, U, and N p the chosen 7s26d15Pconfiguration is the experimental ground state. It was also chosen for Pu and A m to facilitate comparisons along the series. For the quadrivalent Th, both the 7s17p16d2parameters and the 7s26d15f1 parameters are given. The oxygen parameters are reproduced in Table 111. W e note that the estimated uranium exponents and coefficients 6d:
2.581 (0.7608), 1.207 (0.4126)
5f
4.943 (0.7844), 2.106 (0.3908)
TABLE I: Orbital Energies from Average-of-Configuration Dirac-Fock or Hartree-Fock Calculations for 7s26d15f" Configurations of Neutral Atoms -6, eV atom A0 L*b Le NR Th 7s 5.31 4.43 33.23 25.01 26.63 6P 5.09 4.87 6d 7.02 6.26 5.80 5f 13.71 7s" 7.07 5.69 37.10 28.50 29.88 6Pa 7.23 6.84 6d" 8.83 3.59 3.09 3.23 7P" Pa 7s 5.41 4.49 34.93 25.95 27.44 6P 6d 5.19 4.95 7.15 7.93 7.33 5f 15.52 5.51 U 7s 4.53 36.51 26.76 28.16 6P 5.23 4.97 6d 1.25 9.42 8.67 5f 17.25 NP 7s 5.61 4.58 38.03 27.48 28.83 6P 5.23 4.96 6d 7.32 10.80 9.90 5f 18.91 5.71 Pu 7s 4.62 39.51 28.15 29.46 6P 5.20 4.92 6d 7.37 12.11 11.06 5f 20.53 5.80 Am 7s 4.67 40.97 28.76 30.05 6P 5.15 4.87 6d 7.40 5f 13.36 12.14 22.11
"7s'7p16d2. bL*:j = l - ' J 2 . cL: j = 1 + ' J 2 , of ref 9 are in qualitative agreement with ours.
3. Application on Uranyl and Discussion General. The closed-shell uranyl ion UOZ2+has, like COz, no proton affinity in solution5 and shares with it a
(2)
(1) L. L. Lohr Jr. and P. Pyykko, Chem. Phys. Lett., 62, 333 (1979). (2) L.L. Lohr Jr., M. Hotokka, and P. Pyykko, QCPE, 12, 387 (1980). (3) P. Pyykko and L. L. Lohr Jr., Inorg. Chem., 20, 1950 (1981). (4) A. Viste, M. Hotokka, L.Laaksonen, and P. Pyykko, Chem. Phys., 72, 225 (1982). (5) C. K. J~rgensenand R. Reisfeld, J . Electrochem. SOC.,130, 681 (1983); Srrucr. Bonding (Berlin), 50, 121 (1982). (6) V. A. Glebov, "Electronic Structure and Properties of Uranyl Compounds", Energoatomizdat, Moscow, 1983 (in Russian). (7) J. P. Desclaux, At. Data. Nucl. Data Tables, 12, 311 (1973). (8) J. P. Desclaux, Comput. Phys. Commun., 9, 31 (1975).
0022-3654/84/2088-4892$01.50/0
(3)
MO structure" for the 12 (U(7s26d15P)+ 20(2p4) - 2) valence
electrons (levels 6-1 1 in Figure 2). Above them one has unfilled U 5f levels and below them the semicore levels 1-5 for the 10 (U(6p6) + 20(2s2)) electrons. Overlap Integrals. Newman" noted in 1965 that the relativistic (9) K. Tatsumi and R. Hoffmann, Inorg. Chem., 19, 2656 (1980). (10) S. P. McGlynn and J. K. Smith, J . Mol. Spectrosc., 6, 164 (1961). (11) J. B. Newman, J . Chem. Phys., 43, 1691 (1965). 0 1984 American Chemical Society
Double-{ Parameters for Various Actinoids TABLE I 1 Relativistic (R) and Nonrelativistic (NR) Double-!: Radial Fits for Neutral Atoms Th to Am' atom case A 0 cI I:! c2 r2 R 7s 0.605102 2.094 127 0.511 158 1.241 327 Th 6p* 0.923008 3.821 006 0.165 580 1.777290 6p 0.899349 3.437 886 0.190674 1.695662 6d* 0.578890 2.755 778 0.571 682 1.440736 6d 0.573946 2.653 959 0.575 122 1.392806 5f* 0.567917 5.154 089 0.593 592 2.467 971 5f 0.570 646 5.063 873 0.594 164 2.403 133 NR 7s 0.588065 1.836539 0.524 353 1,100631 6p 0.928278 3.371 935 0.160 550 1.539 589 6d 0.623803 2.811 171 0.503 291 1.553 788 5f 0.582 128 5.475 435 0.547 607 2.863 285 Thb R 7s 0.658421 2.146527 0.445 180 1.301 313 6p* 0.172496 2.321 681 0.882 102 3.964 545 6p 0.856562 3.588 045 0.200 991 2.150226 6d* 0.619009 2.862 939 0.509411 1.578 158 6d 0.623800 2.744 81 1 0.504651 1.511 848 7p* 0.546624 1.775592 0.580404 1.027 171 7p 0.535245 1.570547 0.588 272 0.915935 NR 7s 0.621012 1.896 152 0.480 324 1.164 375 6d 0.679002 2.862312 0.436 507 1.609 406 7p 0.537549 1.574 134 0.580 773 0.930 108 Pa R 7s 0.597556 2.150872 0.520 179 1.271 281 6p* 0.936349 3.896519 0.166856 1.636565 6p 0.892264 3.540 184 0.194 341 1.794491 6d* 0.581351 2.833 506 0.571 387 1.473274 6d 0.575363 2.726 236 0.576 003 1.422 463 5f* 0.577 112 5.411 229 0.574 390 2.663 886 5f 0.582 836 5.293208 0.572023 2.581 728 N R 7s 0.583631 1.867 875 0.530053 1.116211 6p 0.921770 3.459 452 0.163 156 1.637667 6d 0.617030 2.893 644 0.511 148 1.595219 5f 0.594795 5.659 275 0.530 901 2.990 790 U R 7s 0.605004 2.181 612 0.514928 1.281 648 6p* 0.214212 2.494 079 0.847 518 4.187 597 6p 0.817813 3.752975 0.249 321 2.256 677 6d* 0.579182 2.910766 0.576032 1.504097 6d 0.573101 2.795 308 0.580877 1.448 904 5f* 0.588 131 5.564286 0.558 589 2.774 723 5f 0.594 113 5.422 060 0.556 806 2.672061 N R 7s 0.580067 1.896604 0.534789 1.130308 6p 0.846729 3.660923 0.212294 2.205 998 6d 0.622495 2.943 859 0.508 568 1.611032 5f 0.592 494 5.884 970 0.530473 3.134846 R 7s 0.598932 2.231 561 0.522 178 1.308015 NP 6p* 0.207 126 2.533 786 0.854 469 4.287 085 6p 0.811 996 3.843 554 0.255 327 2.321 246 6d* 0.576677 2.979 870 0.581 374 1.528 879 6d 0.572180 2.850 944 0.584838 1.466449 5f* 0.589787 5.814 158 0.551 561 2.942313 5f 0.594 408 5.676 028 0.551 076 2.839082 N R 7s 0.577130 1.923 402 0.538 818 1.143 336 6p 0.839826 3.740 972 0.219 607 2.267 424 6d 0.615597 3.016538 0.517086 1.645600 5f 0.519727 3.254 057 0.600 608 6.067 8 12 Pu R 7s 0.593943 2.279 518 0.528 274 1.333083 6p* 0.224044 2.650022 0.838 346 4.416 323 6p 0.257596 2.349 538 0.811 794 3.918518 6d* 0.575610 3.036 789 0.585 653 1.545 561 6d 0.568674 2.907 518 0.591 408 1.483913 5f* 0.542993 3.083 943 0.594 341 6.028 517 5f 0.597183 5.897 228 0.544 240 2.982 238 NR 7s 0.574664 1.948 700 0.542310 1.155551 6p 0.838 318 3.808 396 0.223017 2.294 688 6d 0.609291 3.084 859 0.525 136 1.676706 5f 0.509 587 3.364 059 0.608 626 6.240 975 Am R 7s 0.589793 2.326 154 0.533 462 1.357 332 6p* 0.854179 4.504 232 0.205 988 2.678 344 6p 0.820963 3.976 807 0.251 263 2.341 563 6d* 0.573102 3.091 944 0.591 451 1.560618 6d 0.566665 2.952 869 0.596 747 1.494345 5f* 0.505424 3.132894 0.628 254 6.088 680 5f 0.625 501 5.990 569 0.512053 3.051 148 N R 7s 0.572563 1.972 792 0.545 381 1.167 124 6p 0.229955 2.319 035 0.834 204 3.874 331 6d 0.596224 3.169 440 0.539448 1.719 333 5f 0.496 122 3.456 964 0.620 302 6.391 532 'The 7s26d'5f" electron configuration is 5tssumed. unless otherwise specifieid. b7s17p16d2.
The Journal of Physical Chemistry, Vol. 88, No. 21, 1984 4893 TABLE III: Double4 Parameters for Oxveen
case A 0 R 2s 2p* 2p NR 2s 2p
CI
1 0.726 16345 0.725 57229 1 0.725 398 01
I:! c2 (2 2.225 4389 1.6302094b 0.351 051 45 3.582 1220 1.627 3602 0.351 630 12 3.574 1087 2.222 9082 1.6280663 0.351 732 52 3.573 8795
'Erroneously reproduced in ref 4 as 1.3602094 TABLE IV: Overlap Integrals between Real Atomic Orbitals of Uranium and Oxygen at R = 171 pm@ uranium A 0 oxygen A 0 case 7s 6P 6d 5f 2su a 0.274 0.207 0.487 0.088 b 0.232 0.301 0.443 0.099 c 0.274 0.239 0.447 0.139 2PU a 0.003 0.281 0.109 0.114 b 0.001 0.300 0.115 0.089 c 0.046 0.274 0.080 0.101 2PT a 0.080 0.333 0.082 b 0.137 0.322 0.089 C 0.109 0.327 0.126
"The I* (j = 1 coefficients are used for p, d, and f AOs. bCases a-c refer to the following: (a) single-!:, relativistic (as in ref 3); (b) multiple-!:, nonrelativistic (as in ref 12 and 13); (c) multiple-!:, relativistic. TABLE V: Uranium Populations (in Percent) and the Total Overlap Populations for thi Uranyl Ion at R = 171 pm from Single-{ (SZ) and Double-!: (DZ) REX Calculations' percentage on U MO A0 sz DZ 1.2 1.3 U" (11) 6P 5f 86 79 0.7 1.1 7 d / 2 ) (9) 6P 5f 8.6 19 ug (8) 6d 0.4 0.5 T g ( ' / 2 ) (6) 6d 8.7 8.3 totalb 6d-2p 0.26 0.25 5f-2p 0.17 0.31 0.59 0.69 total overlapb u-0
"The MO numbers refer to Figure 2. The T ( ~ / populations ~) are comparable with the given ones. bOverlap population per oxygen.
SCF expansion of uranium 5f orbitals considerably increases their involvement in bonding. Glebov and N e f e d o ~ ' * , recently '~ formulated a new relativistic, self-consistent, valence-shell MO method in which they, however, used available nonrelativistic orbitals. On the other hand, our earlier REX calculations3 used S Z orbitals. In table IV we compare the overlap integrals for these three cases. Firstly, we note that the case b ones agree with those of ref 13, as they should. Secondly, going from case a to c increases the 5f-2sa and 5f-2pa integrals by over 50%, while the 6d integ?als actually decrease. The 5f contribution to the a, MO (orbitals 9 and 10; see Figure 2) rises from 8.6% to 19% when going from SZ to DZ. The total 5f-2p population increases from 0.17 to 0.31 and now exceeds the 6d-2p one of 0.25. Why Is Uranyl Linear? A plot of the D Z orbital energies as a function of the bending angle closely resembles the SZ one (Figure 2 of ref 3). The sum of the five lowest orbital energies remains constant while the penultimate, aJ3/J MO 10, is raised by 0.9 eV by bending from 180 to 120' (SZ: 1.1 eV). The increase of the 5f character from S Z to D Z has thus not increased the bending force constant. This provides a new suggestion that the linearity of uranyl could be ascribed to increased 6pa-2pa antibonding, while bending, in this 5fa-2pa bonding MO. The 6~312coefficients of this MO are 0.30 and 0.26 for SZ and DZ, respectively. Earlier, Tatsumi and Hoffmann9 ascribed the linearity to the u HOMO ("6p activated oxygen 2p AOs" as a u (12) V. A. Glebov and V. S. Nefedov, Koord. Khim., 7, 1664 (1981). (13) V. A. Glebov and V. S. Nefedov, Koord. Khim., 7 , 586 (1981).
4894
+
The Journal of Physical Chemistry, Vol. 88, No. 21, 1984
1.a0t 0. s0
f
I 1,
CRANIUM DENSITY
!5F)3
( 6 0 ) : !?S)2
Pyykko and Laaksonen
-OR I GI NIL FITTED
...SINGLE-2
7l e
91 0.
5fl
a@-
0. 6 0 +
0.5a
f
0. 5C-
11
0. 30:
60*
2.00
4.00
6.00
C I I
I
\
EL.
8.00
R(h.U.)
Figure 1. Dirac-Fock radial electron densities for uranium. The full curves are the original Dirac-Fock densities, the dotted curves the present double-!: fits, and the dashed curves the single-!: fits of ref 2 and 3.
donor to U 5fa), while W a d P suggested straightforward 5 f r - 2 ~ ~ bonding as the reason. Our 6 p 3 , 2 r 2 p a overlap is 0.136 while Tatsumi and Hoffmann had only 0.049 and dismissed this contribution. On the other hand, our €(HOMO) is almost independent of the bond angle. How Much 5f in the a,,? The present D Z value in Table V is 79%. Tatsumi and Hoffmannq obtained 29% and Glebov and Nefedovtz 85% while values of 59%, 57%, 40%, and 53% were obtained from quasi-relativistic M S Xa,14pseudopotential Hartree-Fock,I5 DiraeSlater DVM,16 and quasi-relativistic M S Xa” calculations, respectively. J ~ r g e n s e nhas ~ , ~argued ~ that this might be an artifact of the M O model. Which HOMO? The highest occupied molecular orbital ( H O M O ) is of special importance as the lower state in luminescence. Experimentally, Denning et al.Iga argue that the H O M O of uranyl is the a,, in agreement with the REX ordering (both SZ and DZ) A,
= a, c A, c
a,
(4)
DeKock et al.lqbfind, using a perturbative HFS method, that the a, is H O M O partly because of “pushing from below” by the 6p and partially because of the relativistic destabilization of the large 5f contribution. Both the a,, H O M O of uranyl and the analogous, theoretical tl, H O M O of UF6 may be due to pushing from below by the 6p shell. It then is most interesting that a recent intensity analysis of the XPS spectraZoof UF6 suggests a tl, HOMO, overlapping with the tl,(ysu) and above the tlu(y6,). This reopens the theoretical issue and might suggest important correlation effects, which are missing from all the present calculations. Role of 6p Orbitals. Experimentally, both the XPSz’ and the 235Uinternal conversion spectra22of uranyl compounds clearly show strong U(6p)-O(2s) hybridization. This hybridization is reproduced by all calculations (see levels 1-5 in Figure 2). As (14) (15) (16) (17) (1981). (18) (19)
M. Boring and J. H. Wood, J . Chem. Phys, 71, 392 (1979). W. R. Wadt. J . Am. Chem. SOC.,103, 6053 (1981). P. F. Walch and D. E. Ellis, J . Chem. Phys., 65, 2387 (1976). J. H. Wood, M. Boring, and S. B. Woodruff, J . Chem. Phys., 74,5225
C. K. Jprgqnsen, Chem. Phys. Lett., 89, 455 (1982). (a) R. G. Denning, T. R. Snellgrove, and D. R. Woodwark, Mol. Phys., 37, 1109 (1979); (b) R. L. DeKock, E. J. Baerends, P. M. Boerrigter, and J. G. Snijders, Chem. Phys. Lett., 105, 308 (1984). (20) (a) N. Martensson, P:A, Malmqvist, and S . Svensson, Chem, Phys. Lett., 100, 375 (1983); (b) N. Martensson, P,-& Malmqvist, S . Svensson, and B. Johansson, to be submitted for publication. (21) B. W. Veal, D. J. Lam, W. T. Carnall, and H. R. Hoekstra, Phys. Reu. B: Solid State, 12, 5651 (1975). (22) D. P. Grechukhin, V. I. Zhudov, A. G . Zelenkov, V. M. Kulakov, B. V. Odinov, A. A. Soldatov, and Yu. A. Teterin, Pis’ma Zh. Eksp. Teor. Fiz., 31, 627 (1980).
3o
42
t
I
U
uo;‘
0
Figure 2. Single-!: (SZ) and double-{ (DZ) orbital energies for uranyl at R = 171 pm. The order of the orbitals is the same in the two cases. The symmetry labels are shown above the SZ case (lu: rnj = parity odd; 3g: mj = 3/2, parity even; etc.), and the orbitals are numbered above the DZ case.
both the bonding and antibonding levels are occupied, we consider the eventual contribution of the 6p orbitals to bonding as entirely unsettled. Glebov (ref 6, Chapter 2.4), from EHT-level arguments, suggests such contributions.
4. A Possible Configuration Interaction Effect on the HOMO of UF, The tl, of UF6 is a nearly pure ligand lone-pair orbital while the highest occupied t,, has a U 5f character, N(5f7, estimated in the various models19b*23*24 as 5% to 27%. In view of, the large value of the 5f intrashell Slater integrals (Fk( k = 0, 2, 4, 6) of 17.1, 8.4, 5.4, and 3.8 eV, respectively, in the present atomic Dirac-Fock calculations), it is interesting to propose a symmetry-dependent correlation effect, involving the highest occupied, tlu, and the empty 5f shells:
9 = Cll(tlu)61 + C21(tlu)4(5f721
(5)
For the tl, the lack of metal character should make this mixture much smaller. As an order-of-magnitude estimate for this increase of the tl, ionization energy, we take W t l , ) = I ( ( t l J 6 l ~ 1 2 - ~ l ( t l u ) 4 ( ~ f ) 2 ) 1 2 / ( t t l , - f5f) ( N ( 5 f ) ~ ( 5 f , 5 f ) ) ’ / ( ~ ~-, 4 ” N (0.1 X 17)2/3
N
1 eV (6)
Thus, this mechanism, driven by intra-atomic exchange integrals on uranium, is in principle capable of changing the tl, H O M O of the various S C F theories to the t,, one, proposed by MArtensson et aLZ0 We note that, at the C I level, without spin-orbit splitting, Hay et a1.25indeed obtain a tl, slightly above tl,. For uranyl, analogous but less clear-cut mechanisms, due to the lower symmetry, also operate. (23) C. Y. Yang in “Relativistic Effects in Atoms, Molecules and Solids”, G. L. Malli, Ed., Plenum Press, New York, 1983, p 346. (24) S. Larsson, private communication. (25) P. J. Hay, W. R. Wadt, L. R. Kahn, R. C. Raffenetti, and D. H. Phillips, J . Chem. Phys., 71, 1767 (1979).
J . Phys. Chem. 1984, 88, 4895-4897
5. conclusions (1) A doub1e-c S T O fit describes adequately both the 6d and the 5f functions of the actinoids. The 6p functions are well described a t the single-{ level. The diffuse 7s functions are also well described but unimportant in bonding. (2) Going from single-{to doublecincreases the (5fa) bonding. (3) It is suggested; as in ref 3 and 15, that the linearity of the uranyl ion should be attributed to the rU MO. The energy increases by bending is however now ascribed to increased 0-
4895
(2pa)-U( 6pa) antibonding. -(4) As first pointed out-by Newman in 1965, relathistic 5f functions are necessary for describing the bonding of uranium. The use of the present functions should considerably improve a recent method by Glebov and Nefedov,12 originally based on nonrelativistic functions. Registry No. UOz2+,16637-16-4;Th, 7440-29-1; Pa, 7440-13-3; U, 7440-61-1; Np, 7439-99-8; PU, 7440-07-5; Am, 7440-35-9; UF6, 778381-5; oxygen, 7782-44-7.
LETTERS Infrared-Laser-Induced Gas-Phase Isomerization of Olefins in the Presence of Fe(CO), Peter A. Teng, Frederick D. Lewis,* and Eric Weitz*t Department of Chemistry, Northwestern University, Evanston, Illinois 60201 (Received: May 4, 1984; In Final Form: August 20, 1984)
Irradiation of a mixture of 1-pentene or trans-2-pentene and Fe(CO)Swith a pulsed C 0 2 TEA laser at low pressures in the vapor phase results in the formation of isomeric mixtures of 1- and 2-pectenes and the consumption of Fe(CO),. Isomerization occurs due to heating of the Fe(CO)5 caused by collisional deactivation of the excited pentene. This results in decomposition of Fe(CO)5 to yield chemically reactive coordinatively unsaturated iron carbonyls or iron atoms which can effect the cis, trans and positional isomerization of olefins. Both the initial and final pentene isomer ratios differ significantly from those observed in solution phase photochemically or in thermally initiated isomerization of the pentenes by iron carbonyls.
Introduction
olefin isomerization and (b) the formation of product olefins with both initial and final product ratios substantially different than those obtained in photochemical reactions of Fe(CO),' or thermal reactions of Fe3(C0)12.3b
The spectroscopy and chemistry of metal carbonyls continues to be an active area of Much of this interest is due to the ability of coordinatively unsaturated metal carbonyl species to catalyze chemical reactions such as olefin isomerization, hydrogenation, and hydrosilation. While the spectroscopy and photochemistry of Fe(CO)S have been extensively ifivestigated; its thermal chemistry has received relatively little attention. Attempts to effect positional isomerization of terminal alkenes by thermal activation of F e ( C 0 ) 5 resulted in decomposition to yield iron powder without extensive olefin is~merization.~ These observations are consistent with the recent report of Engelking and LinebergerZathat the bond dissociation energy for Fe(CO)4-C0 is larger than that for any of the coordinatively unsaturated species Fe(CO),, n = 1-4. Thus thermolysis of Fe(CO), should result in more rapid decomposition of the catalytically active species than the inactive precursor. Pulsed heating methods offer a potential advantage over conventional thermolysis in that the initially generated internally energized, coordinatively unsaturated species produced during the heat pulse could be collisionally deactivated and effect catalytic reactions between heat pulse^.^ Pulsed laser heating also minimizes dissociation and deposition of material a t the cell walls which occur extensively with conventional heating. We report here the preliminary results of our investigation of pulsed laser-induced infrared multiphoton isomerization of 1-pentene and trans-2-pentene in the presence of Fe(CO)s. Among the significant findings of this investigation are (a) laser excitation of an olefin in an ~ l e f i n - F e ( C O ) mixture ~ results in the gradual decomposition of Fe(CO)S and concomitant
1974, 70, 283-301. (b) Schroeder, M. A.; Wrighton, M. S. J Am. Chem. SOC.1976, 98, 551-8. (c) Graff, J. L.; Sanner, R. D.; Wrighton, M. S. Organometallics 1982,1, 837-42. (d) Salomon, R. G. Tetrahedron 1983, 39, 485-575. (e) Whetton, R. L.; Fu, K. J.; Grant, E. R. J. Am. Chem. SOC.1982, 104, 4270-1. (2) (a) Engelking, P. C.; Lineberger, W. C. J . Am. Chem. SOC.1979,101, 5569-73. (b) Yardley, J. T.; Gitlin, B.; Nathanson, G.; Rosan, A. M. J . Chem. Phys. 1981, 74, 370-9. (c) Ouderkirk, A. J.; Weitz, E. J . Chem. Phys. 1983, 79, 1089-97. (3) (a) Manuel, T. A. J . Org. Chem. 1962,27, 3941-5. (b) Bingham, D.; Hudson, B.; Webster, D.E.; Wells, P. B. J . Chem. SOC.,Dalton Trans. 1974,
1983-1984 Visiting Fellow, Joint Institute for Laboratory Astrophysics, University of Colorado and National Bureau of Standards, Boulder, CO
1521-4. (4) Sonocatalysis of metal carbonyls has been recently reported: Suslick, K. S.;Goodale, J. W.; Schubert, P. F.; Wang, H. H. J . Am. Chem. SOC.1983,
80309.
Experimental Section The cis- and trans-2-pentenes were obtained from Chemical Samples Co. and were specified to be of >98% purity. The 1-pentene sample was obtained from Aldrich Chemical Co. and was specified to be at >99% pure. On FID G C analysis, each sample was found to be >99% pure and was not subject to further purification. All reported yields have the initial concentration of other pentene isomers subtracted out. Isomer analysis was carried out with a &fl-oxydipropionitrile column. Infrared spectra were recorded with a Nicolet series 7000 FTIR spectrometer. Pressure measurements were made with an MKS capacitance manometer. Irradiation of the pentenes was carried out with the collimated unfocused output of a Lumonics K203-2 CO, TEA (1) (a) Wrighton, M.; Hammond, G. S.; Grey, H. B. J. Organomet. Chem.
105, 5781-5.
0022-3654/84/2088-4895$01.50/00 1984 American Chemical Society
