JOURNAL OF CHEMICAL PHYSICS VOLUME 113, NUMBER 8 22 AUGUST 2000 The pure rotational spectrum of CaNH2 and CaND2 „ X̃ 2 A 1 …: Additional proof of planarity M. A. Brewster and L. M. Ziurysa) Department of Chemistry, Department of Astronomy, and Steward Observatory, 933 North Cherry Avenue, Tucson, Arizona 85721 共Received 8 February 2000; accepted 22 May 2000兲 The pure rotational spectrum of CaNH2 in its X̃ 2 A 1 ground electronic state has been recorded using millimeter/submillimeter direct absorption methods in the frequency range 320–537 GHz as well as that of CaND2. The species were created by Broida-oven techniques. Eleven rotational transitions were observed arising from the v ⫽0 and v 6 ⫽1 states of CaNH2, and eight transitions were recorded for the v 3 ⫽1 and v 4 ⫽1 levels. For CaND2, eight transitions ( v ⫽0) were also measured. For the majority of these transitions, K a doublets corresponding to K a ⫽0 – 5 were observed and fine structure splittings were resolved in every component. These spectra were analyzed using an S-reduced Hamiltonian; rotational, centrifugal distortion, and spin–rotation parameters were determined for CaNH2, CaND2, and the three observed vibrationally excited states. An r 0 structure has also been calculated. The data are consistent with calcium amide being a planar molecule with C 2 v symmetry and having predominately ionic bonding, as indicated by previous optical studies. © 2000 American Institute of Physics. 关S0021-9606共00兲00432-3兴 I. INTRODUCTION tional transitions of CaNH2, SrNH2, and BaNH2 using laserinduced fluorescence 共LIF兲.17,18 These studies were subsequently followed by additional LIF measurements of CaNH2 and SrNH2 by Bernath and co-workers,19 who observed the Ã→X̃ and B̃→X̃ transitions. Higher resolution electronic spectroscopy of CaNH2 was then carried out by Whitham and collaborators,20,21 Marr et al.,22 and, more recently, Morbi and co-workers.23,24 These studies produced high resolution analyses of the C̃ – X̃, B̃ – X̃, and à – X̃ systems of CaNH2, including the determination of rotational and fine structure constants of the ground state. Such measurements indicated that the alkaline earth amides were planar with C 2 v symmetry. Up to the present time, however, pure rotational spectroscopy had not been conducted for any metal amide species. In this paper we present the measurement of the pure rotational spectrum of CaNH2 in its 2 A 1 ground electronic state, using submillimeter-wave direct absorption techniques. Eleven rotational transitions of this free radical were recorded in its v ⫽0 state, including K a components as high as K a ⫽7, as well as numerous lines arising from the v 4 ⫽1 state, the out-of-plane puckering motion, and the v 3 and v 6 modes. The pure rotational spectrum of CaND2 was also obtained in its ground vibrational and electronic state. No evidence of inversion was found for either molecule. These data were analyzed using the S-reduced asymmetric top Hamiltonian and rotational, centrifugal distortion, and fine structure constants were accurately determined for CaNH2 and CaND2. An r 0 structure has been obtained as well. Here we present our data and its analysis, as well as an interpretation of our findings in terms of the bonding in metal amide species. Small, metal-containing molecules offer a unique opportunity to examine the competition between ionic and covalent bonding. Because of the generally low ionization potentials of most metals, their bonds to other elements are likely to be ionic to a first approximation. Indeed, the properties of many metal halide species such as CaF can be explained in the context of ionic models, such as a ligand field approach,1 or using electrostatic polarization effects.2 However, for some of the metals in the earlier rows of the periodic table, strictly ionic bonding may not be able to account for molecular properties. This situation is certainly observed in the metal monohydroxide and monocyanide molecules. The heavier monohydroxide species such as RbOH, CsOH, BaOH, and SrOH have strongly bound, linear structures,3–6 while large amplitude motions and quasilinear behavior are found for KOH, NaOH, and MgOH.7–10 CuOH and AgOH are bent with angles near 108°–110°,11 similar to covalent H2O. An analogous trend is found in the cyanides. Here the ionically favored structure is a T shape, as found in NaCN12 and KCN,13 where the M⫹ cation is thought to ‘‘orbit’’ the CN⫺ anion. As covalent forces become comparable, the cyanides take on the linear, isocyanide form, observed in MgNC14 and LiNC.15 Another interesting class of molecules where the competition between bonding types is reflected in their geometries are metal amide species of the general formula MNH2. The covalently bonded amide is likely to have a structure similar to NH3, i.e., a pyramidal shape, while the ionic compounds 16 M⫹NH⫺ 2 , will be planar with C 2 v symmetry. Such species were first discovered by the Harris group, who detected opa兲 Electronic mail: [email protected] 0021-9606/2000/113(8)/3141/9/$17.00 3141 © 2000 American Institute of Physics 3142 J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 II. EXPERIMENT The data were recorded using one of the millimeter/ submillimeter direct absorption spectrometers of the Ziurys group. The details of the construction of this spectrometer are given elsewhere25 and will only be briefly outlined here. The instrument consists of a phase-locked Gunn oscillator/ Schottky diode multiplier as the source of radiation, an evacuated reaction chamber that incorporates a Broida-type oven as a source of metal vapor, and a liquid helium-cooled InSb bolometer as the detector. The radiation is modulated at 25 kHz to achieve phase-sensitive detection and is propagated quasioptically through the cell using a scalar feedhorn, a polarizing grid, and a series of Teflon™ lenses. After transmission through this double-pass system the resultant radiation is recorded at 2 f , thus obtaining second derivative spectra. The CaNH2 radical was produced by reaction of the metal vapor with a mixture of NH3 and argon in a gas glow discharge. The calcium vapor was generated in a Broida-type oven, and entrained in a mixture of NH3 and argon introduced through the bottom of the oven in such a fashion as to achieve a partial pressure of 12–15 mTorr of NH3 and 25–30 mTorr argon in the gas cell. The dc discharge was then applied from a cathode placed above the grounded oven. Depending upon the type of electrode used and the details of the vaporizing source in the oven, a voltage range of 20–50 V was found to result in optimum discharge conditions at a corresponding current of 400–800 mA. It was not critical that any particular voltage be used to produce the radical, but rather that a rich lavender aura of atomic emission emanated from the calcium within the crucible. CaND2 was produced by an identical method using ND3 instead of NH3. Transition frequencies were measured for 165 lines of CaNH2 ( v ⫽0), 98 lines of its v 4 ⫽1 state, 114 lines of its v 6 ⫽1 state, 58 lines of its v 3 ⫽1 state, and 142 lines of CaND2 in the frequency range of 322–535 GHz. All measurements were made from an average of an even number of 5 MHz scans, half of which are taken in increasing frequency and the other half in decreasing frequency. Most of these observations required only two such scans, but weak signals, especially the v 3 ⫽1 vibrationally excited state, required more extensive signal averaging. Typical line widths were 800–1500 kHz over the entire frequency range. III. RESULTS Table I contains a subset of the data obtained for calcium amide and its deuterium isotopomer. CaNH2 is most likely planar with C 2 v symmetry, as will be discussed later on, and therefore should have a 2 A 1 ground electronic state. In this case the species is a near prolate asymmetric rotor, and the quantum numbers which label each transition are N, K a , K c , and J. N designates the total angular momentum, and J indicates the spin–rotation interaction, where J⫽N⫹S. Unlike the case of the symmetric top, however, the projection of N onto the symmetry axis, K, is no longer a constant of the motion. The ⫾K degeneracy present in symmetric rotors is lifted and indices corresponding to the limiting prolate (K a ) and oblate (K c ) cases are used to label these energy levels. M. A. Brewster and L. M. Ziurys Moreover, for a planar, C 2 v species, the dipole moment lies only along the symmetry axis 共in this case the â axis兲, which means that all electric-dipole allowed transitions are a type. Hence, as shown in Table I, only a-type transitions were measured, meaning ⌬K a ⫽0 and ⌬K c ⫽⫾1 for ⌬N⫽⫾1 and ⌬J⫽⫾1. 共This fact is additional evidence for planarity.兲 Moreover, a general property of this notation is that K a ⫹K c ⫽N or N⫹1 and considering the above selection rules, the K a components with K a ⫹K c ⫽N lie higher in frequency than those with K a ⫹K c ⫽N⫹1 for a particular N. A total of 11 rotational transitions were measured for CaNH2 in both its v ⫽0 state and v 6 ⫽1 mode, the asymmetric (B 2 ) stretch. Individual spin–rotation doublets per K a component were measured in each transition usually for the K a ⫽0, 1, 2, and 3 and sometimes the K a ⫽4, 5, 6, and 7 lines, as illustrated in Table I. 共The K a ⫽1, 2, and 3 components are split as well due to asymmetry doubling, but usually these doublets are more widely separated兲. Eight transitions were additionally measured for the v 3 ⫽1 state, a symmetric (A 1 ) stretch, and the same number for the v 4 ⫽1 out-of-plane puckering motion. Again, in each transition, components for K a ⫽0 – 5 were usually observed for these two vibrationally excited states, all which were split into spin–rotation doublets. Finally, eight transitions of CaND2 were recorded, for K a ⫽0 up to as high as K a ⫽7. The average spin–rotation splitting was found to be on the order of 32–39 MHz. The transition frequencies of both molecules and all excited vibrational states were assigned using the same algorithm. For CaNH2, a range of 460–490 GHz was initially scanned in 100 MHz increments and the frequency of all absorptions were recorded. For CaND2 this starting range consisted of two blocks: 467–474 and 451–458 GHz. Harmonically related spin–rotation doublets were then fit to an effective B, D, and ␥ and additional transitions searched for, which were incorporated into the fit. Rotational quantum numbers were readily assigned on the basis of the optically derived constants,22,23 as well as the magnitude of the spin– rotation splittings. Each set of lines fit to an effective B, D, and ␥ corresponded to a K a component. Finally, considering differences in line intensities arising from ortho/para spin statistics, tentative K a , K c assignments were made and a ⫾K relationship was looked for among the B eff’s. Defining B eff a as the effective B for the upper (⫹)K a or lower (⫺)K a components, the following approximate relationship was 0 ⫹1 ⫺1 ⫺(Beff ⫹Beff )/2 then used for the assignments, namely B eff ⫹1 ⫺1 ⫹2 ⫺2 ⫹2 ⫺2 ⫹3 ⫽(B eff ⫹Beff )/6⫺(B eff ⫹Beff )/6⫽(B eff ⫹Beff )/10⫺(B eff ⫺3 ⫹Beff )/10, and so on. This relationship is the same 1:3:5 ratio exhibited by the K structure of a prolate symmetric rotor. In this case, we have compensated for the lifting of the K degeneracy by averaging the effective rotational constants of the K a doublets. In addition to the fine structure, magnetic hyperfine and electric quadrupole interactions are also possible as both the 14N(I⫽1) and 1H(I⫽1/2) atoms have nuclear spins. Due to the large values of N in the spectra, such interactions were small enough to not be resolved. Other tools used in the assignments were overall intensities and spin statistics. The lines that were most intense were assigned to the ground ( v ⫽0) state, which has spin J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 Rotational spectrum of CaNH2 3143 TABLE I. Selected transition frequencies for CaNH2 and CaND2 (X̃ 2 A 1 ). a CaNH2 v ⫽0 Transition N ⬙ (K a⬙ ,K c⬙ )J ⬙ N ⬘ (K a⬘ ,K c⬘ )J ⬘ 23共1, 23共1, 23共5, 23共5, 23共5, 23共5, 23共4, 23共4, 23共4, 23共4, 23共3, 23共3, 23共3, 23共3, 23共2, 23共2, 23共0, 23共0, 23共2, 23共2, 23共1, 23共1, 27共7, 27共7, 27共7, 27共7, 27共1, 27共1, 27共5, 27共5, 27共5, 27共5, 27共4, 27共4, 27共4, 27共4, 27共3, 27共3, 27共3, 27共3, 27共2, 27共2, 27共0, 27共0, 27共2, 27共2, 27共1, 27共1, 29共1, 29共1, 29共5, 29共5, 29共5, 29共5, 29共3, 29共3, 29共3, 29共3, 29共0, 29共0, 29共1, 24共1, 24共1, 24共5, 24共5, 24共5, 24共5, 24共4, 24共4, 24共4, 24共4, 24共3, 24共3, 24共3, 24共3, 24共2, 24共2, 24共0, 24共0, 24共2, 24共2, 24共1, 24共1, 28共7, 28共7, 28共7, 28共7, 28共1, 28共1, 28共5, 28共5, 28共5, 28共5, 28共4, 28共4, 28共4, 28共4, 28共3, 28共3, 28共3, 28共3, 28共2, 28共2, 28共0, 28共0, 28共2, 28共2, 28共1, 28共1, 30共1, 30共1, 30共5, 30共5, 30共5, 30共5, 30共3, 30共3, 30共3, 30共3, 30共0, 30共0, 30共1, 23兲 23兲 18兲 19兲 18兲 19兲 19兲 20兲 19兲 20兲 21兲 20兲 21兲 20兲 22兲 22兲 23兲 23兲 21兲 21兲 22兲 22兲 20兲 21兲 20兲 21兲 27兲 27兲 22兲 23兲 22兲 23兲 23兲 24兲 23兲 24兲 25兲 24兲 25兲 24兲 26兲 26兲 27兲 27兲 25兲 25兲 26兲 26兲 29兲 29兲 24兲 25兲 24兲 25兲 27兲 26兲 27兲 26兲 29兲 29兲 28兲 22.5 23.5 22.5 22.5 23.5 23.5 22.5 22.5 23.5 23.5 22.5 22.5 23.5 23.5 22.5 23.5 22.5 23.5 22.5 23.5 22.5 23.5 26.5 26.5 27.5 27.5 26.5 27.5 26.5 26.5 27.5 27.5 26.5 26.5 27.5 27.5 26.5 26.5 27.5 27.5 26.5 27.5 26.5 27.5 26.5 27.5 26.5 27.5 28.5 29.5 28.5 28.5 29.5 29.5 28.5 28.5 29.5 29.5 28.5 29.5 28.5 24兲 24兲 19兲 20兲 19兲 20兲 20兲 21兲 20兲 21兲 22兲 21兲 22兲 21兲 23兲 23兲 24兲 24兲 22兲 22兲 23兲 23兲 21兲 22兲 21兲 22兲 28兲 28兲 23兲 24兲 23兲 24兲 24兲 25兲 24兲 25兲 26兲 25兲 26兲 25兲 27兲 27兲 28兲 28兲 26兲 26兲 27兲 27兲 30兲 30兲 25兲 26兲 25兲 26兲 28兲 27兲 28兲 27兲 30兲 30兲 29兲 23.5 24.5 23.5 23.5 24.5 24.5 23.5 23.5 24.5 24.5 23.5 23.5 24.5 24.5 23.5 24.5 23.5 24.5 23.5 24.5 23.5 24.5 27.5 27.5 28.5 28.5 27.5 28.5 27.5 27.5 28.5 28.5 27.5 27.5 28.5 28.5 27.5 27.5 28.5 28.5 27.5 28.5 27.5 28.5 27.5 28.5 27.5 28.5 29.5 30.5 29.5 29.5 30.5 30.5 29.5 29.5 30.5 30.5 29.5 30.5 29.5 CaND2 v 3 ⫽1 v 4 ⫽1 v 6 ⫽1 v obs v o⫺c v obs v o⫺c v obs v o⫺c 423 568.487 423 607.455 423 967.803 423 967.803 424 003.796 424 003.796 424 927.080 424 927.080 424 963.201 424 963.201 425 621.273 425 622.957 425 657.665 425 659.217 426 023.097 426 059.693 426 182.100 426 219.306 426 240.169 426 276.136 428 963.680 428 998.121 490 671.053 490 671.053 490 706.401 490 706.401 493 910.017 493 949.044 494 421.031 494 421.031 494 457.036 494 457.036 495 531.136 495 531.136 495 567.299 495 567.299 496 338.803 496 342.014 496 375.061 496 378.290 496 778.103 496 814.694 496 878.375 496 915.736 497 121.575 497 157.562 500 186.995 500 221.507 529 037.563 529 076.632 529 611.725 529 611.725 529 647.703 529 647.703 531 659.643 531 664.276 531 695.995 531 700.563 532 169.870 532 207.312 535 752.388 0.009 0.013 0.007 0.007 0.056 0.056 ⫺0.074 ⫺0.070 ⫺0.105 ⫺0.101 ⫺0.100 0.075 ⫺0.003 0.047 0.020 0.033 ⫺0.061 ⫺0.016 0.066 0.024 0.003 ⫺0.003 0.021 0.021 ⫺0.040 ⫺0.040 0.021 0.034 ⫺0.036 ⫺0.036 0.008 0.008 0.093 0.106 0.102 0.114 0.030 ⫺0.023 0.007 ⫺0.015 ⫺0.020 ⫺0.051 0.039 0.035 ⫺0.172 ⫺0.028 0.014 0.024 ⫺0.012 0.017 ⫺0.060 ⫺0.060 ⫺0.031 ⫺0.031 ⫺0.071 ⫺0.048 0.015 ⫺0.010 0.029 ⫺0.007 0.017 427 835.328 427 872.632 0.023 ⫺0.016 423 299.049 423 338.169 ⫺0.058 0.018 424 849.606 424 849.606 424 886.182 424 886.182 430 776.420 430 781.401 430 811.688 430 816.632 498 875.253 498 912.635 502 373.395 502 384.160 502 408.661 502 419.467 507 214.122 507 247.826 v obs v o⫺c ⫺0.038 ⫺0.034 0.029 0.032 428 216.097 428 216.097 428 251.811 428 251.811 ⫺0.014 ⫺0.006 0.013 0.021 425 441.575 425 478.301 425 604.249 425 641.519 425 596.643 425 632.951 428 007.746 428 042.632 0.022 ⫺0.040 ⫺0.044 0.047 0.089 0.001 ⫺0.164 0.029 429 230.464 429 266.374 429 330.311 429 366.800 429 564.266 429 599.495 ⫺0.033 ⫺0.043 0.008 ⫺0.006 ⫺0.015 ⫺0.036 493 600.479 493 639.608 ⫺0.034 0.014 495 416.685 495 416.685 495 453.226 495 453.226 0.014 0.024 ⫺0.011 ⫺0.001 496 109.688 496 146.428 496 232.308 496 269.644 496 354.948 496 390.988 499 080.370 499 115.279 528 708.775 528 747.973 ⫺0.071 ⫺0.160 ⫺0.018 ⫺0.001 0.062 ⫺0.194 ⫺0.038 0.137 ⫺0.054 0.043 499 341.663 499 341.663 499 377.441 499 377.441 500 096.467 500 102.862 500 132.063 500 138.428 500 499.010 500 534.938 500 491.271 500 527.993 501 026.536 501 601.625 ⫺0.057 ⫺0.033 0.023 0.047 0.059 0.077 ⫺0.032 ⫺0.024 0.021 ⫺0.018 0.007 ⫺0.011 ⫺0.021 0.007 531 494.257 531 531.806 534 571.750 ⫺0.081 0.071 ⫺0.117 0.021 0.006 0.007 ⫺0.032 ⫺0.027 ⫺0.044 ⫺0.015 ⫺0.050 ⫺0.002 0.030 0.030 ⫺0.023 v ⫽0 536 005.608 536 042.479 ⫺0.037 ⫺0.036 v obs v o⫺c 427 042.946 427 074.865 426 070.079 426 103.489 428 489.442 428 519.664 ⫺0.037 ⫺0.088 ⫺0.052 ⫺0.062 ⫺0.063 ⫺0.024 452 682.928 452 716.958 ⫺0.016 0.026 456 186.229 456 219.952 462 226.370 ⫺0.026 0.046 ⫺0.011 3144 J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 M. A. Brewster and L. M. Ziurys TABLE I. 共Continued.兲 CaNH2 v ⫽0 Transition N ⬙ (K a⬙ ,K c⬙ )J ⬙ N ⬘ (K a⬘ ,K c⬘ )J ⬘ 29共1, 29共2, 29共2, 29共4, 29共4, 29共2, 29共2, 29共4, 29共4, 29共6, 29共6, 29共6, 29共6, 29共7, 29共7, 29共7, 29共7, 30共1, 30共2, 30共2, 30共4, 30共4, 30共2, 30共2, 30共4, 30共4, 30共6, 30共6, 30共6, 30共6, 30共7, 30共7, 30共7, 30共7, a 28兲 27兲 28兲 25兲 26兲 27兲 28兲 25兲 26兲 23兲 24兲 24兲 23兲 22兲 23兲 22兲 23兲 29.5 28.5 28.5 28.5 28.5 29.5 29.5 29.5 29.5 28.5 28.5 29.5 29.5 28.5 28.5 29.5 29.5 29兲 28兲 29兲 26兲 27兲 28兲 29兲 26兲 27兲 24兲 25兲 25兲 24兲 23兲 24兲 23兲 24兲 30.5 29.5 29.5 29.5 29.5 30.5 30.5 30.5 30.5 29.5 29.5 30.5 30.5 29.5 29.5 30.5 30.5 v 3 ⫽1 v obs v o⫺c 535 786.948 0.044 v obs v o⫺c CaND2 v 4 ⫽1 v 6 ⫽1 v ⫽0 v obs v o⫺c v obs v o⫺c 534 606.696 531 704.131 531 403.465 530 659.461 530 659.461 531 740.213 531 440.165 530 695.844 530 695.844 0.073 0.178 0.260 0.034 0.050 0.024 0.113 ⫺0.167 ⫺0.150 536 734.172 536 087.199 534 862.203 534 862.203 536 769.150 536 123.197 534 897.984 534 897.984 0.023 0.012 ⫺0.088 ⫺0.048 0.045 0.018 ⫺0.005 0.035 v obs v o⫺c 462 256.737 459 178.142 457 413.133 ⫺0.013 0.000 0.012 459 208.113 457 445.121 ⫺0.019 ⫺0.007 455 279.575 455 279.575 455 311.183 455 311.183 454 164.624 454 164.624 454 196.418 454 196.418 0.053 0.053 ⫺0.057 ⫺0.057 0.028 0.028 0.033 0.033 In megahertz. For the complete data set see Ref. 44. statistics characteristic of A 1 symmetry. For molecules with C 2 v geometry, particle exchange of fermions 共which applies to protons in this case兲, results in the odd K a components being favored statistically over even K a lines by a factor of 3:1, provided the vibrational wave function is symmetric. The spectra assigned to the ground state had in fact stronger odd K a lines relative to nearby even ones by about a factor of 3, as expected. The three sets of weaker lines recorded were then attributed to vibrationally excited states. According to optical experiments,16 the three modes lying lowest in energy are the v 6 antisymmetric N–H bend, the v 4 puckering motion, and the v 3 Ca–N symmetric stretch, which lie approximately 320, 347, and 520 cm⫺1 above the v ⫽0 level. The weakest lines of the three sets had K a intensities identical to the ground state, and hence could be confidently assigned to the v 3 symmetric stretch. The other two progressions had spin statistics opposite to that of the ground state, namely, the even K a lines were a factor of 3 stronger than odd K a lines. Therefore, they had to arise from asymmetric vibrational states such as the v 4 (B 1 ) and v 6 (B 2 ) modes. The intensity of one pattern was slightly greater than the other, and it was identified as arising from the v 6 state, which lies somewhat lower in energy than v 4 . However, there is some ambiguity in this assignment, as the intensity difference between the spectra was small. On the other hand, the constants for the lines assigned to the v 4 state are closest in value to those of the ground state, which might be expected of the puckering motion rather than the asymmetric stretch. Moreover, our v 4 and v 6 assignments follow the pattern found in H2CO. Another possible interpretation of the CaNH2 spectrum is for the ground state to be undergoing inversion. A molecule that possesses an average structure that is orthorhombic, but has achieved such a configuration by inversion, also will exhibit the same statistical weighting as observed in the v ⫽0 state. The spin statistics observed in optical spectra could be interpreted as planarity or possible inversion.20,22 共In the lower resolution spectra the two inversion states could be blended兲. Moreover, the proton statistics in the upper inversion state are identical to that in the v 4 mode. Hence, the spectra identified as arising from the v 4 state could be actually coming from the upper inversion level. Indeed, the rotational constants for the v ⫽0 and the proposed v 4 states are quite similar, as expected for two inversion levels.26,27 For example, B varies by 28 MHz and C by 1 MHz between the two sets of spectra. On the other hand, inversion levels are usually close in energy and hence one would expect quite similar line intensities. In this case, the ground state spectra were about a factor of 2.5 stronger than those of the other state. Furthermore, molecules undergoing inversion develop a small but non-negligible c-type dipole moment and hence exhibit c-type transitions. Such transitions have been readily detected for such inverting molecules as NH2CN and NH2NC. 26,27 In contrast, no obvious c-type lines were observed in the spectra of CaNH2. Overall, the evidence supports a planar, C 2 v geometry for calcium amide. For CaND2, the proton spin statistics change again. Substituting bosons (D:I⫽1) for fermions (H:I⫽1/2) alters the weighting scheme to 3:6, odd K a to even K a . This pattern was observed in the spectra of CaND2. Also, deuterium substitution created a heavier molecule. K a components up to K a ⫽8 were observed in this species, although only lines through K a ⫽7 were included in the fit. For CaNH2, only components with K a as large as 5 were recorded, with the exception of a few K a ⫽7 lines. The K a ⫽6 features were usually too weak to be measured. The complexity of the CaNH2 spectrum is illustrated in Fig. 1, which is a stick diagram of the N⫽25→26 rotational transitions covering 458–469 GHz. The heights of the stick figures show the approximate relative intensities of the various spectral features. Spin–rotation splittings are neglected. The lines assigned to the v ⫽0 state clearly dominate the J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 FIG. 1. A stick spectrum of the N⫽25→26 rotational transition of CaNH2 (X̃ 2 A 1 ) showing the lines observed over a 10 GHz region and their approximate relative intensities, as a function of vibrational and K a quantum number. The transitions originating in the ground vibrational state ( v ⫽0) are clearly the strongest, with spin statistics favoring the odd K a lines over the even K a components. The v 3 ⫽1 satellite lines, which arise from the Ca–N stretch, are the weakest, with identical spin statistics as the ground state. Features with intermediate intensities are attributed to the v 4 ⫽1 out-ofplane puckering motion and the v 6 ⫽1 asymmetric bend. Both these states exhibit opposite spin statistics from the ground state. spectrum, especially for odd K a transitions. Features assigned to the v 4 and v 6 modes are clearly weaker, with even K a transitions being stronger in this case. Finally, the v 3 spectral lines have the lowest intensity. Representative spectra of CaNH2 are shown in Figs. 2 and 3. Figure 2 presents spectra of the N⫽26→27 rotational Rotational spectrum of CaNH2 3145 FIG. 3. Vibrational satellite spectra arising from the N⫽26→27 transition of the v 6 ⫽1 state of CaNH2 near 483 GHz. Again the K a ⫽0, 2 ⫹ , 2 ⫺ , 3 ⫹ and 3 ⫺ components are visible in the spectrum, as well as one of the K a ⫽1 components of the v ⫽0 state. The spin–rotation doublets are clearly resolved for each K a component, and the even K a lines are stronger than the odd K a features. The spectrum covers over 1 GHz in frequency and is a composite of ten scans, each 100 MHz in coverage and about 1 min in duration. transition of the ground vibrational state near 479 GHz. Only the K a ⫽0, 2 ⫹ , 2 ⫺ , 3 ⫹ , and 3 ⫺ components are present in this spectrum, which covers about 800 MHz. The spin– rotation doublets are resolved in every K a component. While the two K a ⫽2 transitions 关N(K a ,K c )⫽26(2,25) – 27(2,26) and 26共2, 24兲–27共2, 25兲兴 are substantially separated in frequency, the K a ⫽3 transitions 关26共3, 24兲–27共3, 25兲 and 26共3, 23兲–27共3, 24兲兴 are virtually collapsed. The K a ⫽3 components are also clearly the strongest. Additionally present in this spectrum is a very weak line corresponding to the 26共1, 26兲–27共1, 27兲 transition of the v 6 ⫽1 state of CaNH2. In Fig. 3, several lines arising from the N⫽26→27 transition of CaNH2 in its v 6 ⫽1 state are presented near 482 GHz. In contrast to the v ⫽0 spectrum, here the even K a lines 共K a ⫽0, and 2兲 are stronger than the odd K a features (K a ⫽3). Again, the spin–rotation splittings are resolved in every component. A doublet arising from the 26共1, 25兲– 27共1, 26兲 transition in the ground state is also visible in this spectrum, which covers 1 GHz in frequency. The intensity of this line relative to the v 6 features is considerably larger. IV. ANALYSIS FIG. 2. Spectra arising from the N⫽26→27 transition in the ground vibrational state of CaNH2 near 479 GHz. The quantum number labeling is N (K a ,K c ). The K a ⫽0, 2 ⫹ , 2 ⫺ , 3 ⫹ and 3 ⫺ components of this transition are shown, as well as one K a ⫽1 component of the v 6 ⫽1 state. Each component is split into doublets as a result of fine structure interactions, and the K a ⫽3 lines are about a factor of 3 stronger in intensity than the other components, as predicted by spin statistics for C 2 v symmetry species. This spectrum is a composite created from eight separate 100 MHz scans, each ⬃1 min in duration. Since CaNH2 is a near prolate asymmetric rotor, an appropriately modified S-reduced Hamiltonian of Watson28 in the I r basis was used to model the spectra. This extended Hamiltonian accounts for rotation, its centrifugal distortion correction, electron spin–rotation interaction, and in one case, centrifugal distortion correction to the spin–rotation coupling, and can be expressed as H eff⫽H rot⫹H cd⫹H sr⫹H srcd . 共1兲 Watson’s S-reduced Hamiltonian plus several additional higher order terms was used for the H rot⫹H cd contribution, which has the general form 3146 J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 M. A. Brewster and L. M. Ziurys TABLE II. Rotational constants of CaNH2 and CaND2. a CaND2 CaNH2 a Constant v ⫽0 v 6 ⫽1 v 3 ⫽1 v 4 ⫽1 v ⫽0 A B C ⑀ aa ⑀ bb ⑀ cc S ⌬ NK DN D NK d1 d2 h1 h2 h3 HN H NK H KN L NNK L NK L KKN P NNK P NKK P KN 392 127共89兲 9009.0649共47兲 8782.7559共49兲 45.7共1.8兲 32.063共99兲 41.110共96兲 ⫺5.1(1.4)⫻10⫺6 0.010 383 3共14兲 1.874 04共78兲 ⫺0.000 335 0共15兲 ⫺0.000 112 1共23兲 ¯ 3.3(2.1)⫻10⫺9 7.8(2.9)⫻10⫺10 ¯ 3.917(48)⫻10⫺5 ⫺0.006 701共75兲 ¯ ⫺9.4(3.5)⫻10⫺8 ⫺5.28(26)⫻10⫺5 ⫺6.27(73)⫻10⫺11 2.012(45)⫻10⫺8 ⫺1.637(31)⫻10⫺6 360 362共204兲 9098.572共10兲 8830.5110共94兲 39.5共1.9兲 31.73共13兲 39.94共13兲 ¯ 0.010 871 6共84兲 2.1677共17兲 ⫺0.000 543 1共32兲 ⫺0.000 144 1共27兲 ¯ 6.6(1.8)⫻10⫺9 ¯ 9.3(4.6)⫻10⫺9 1.97(16)⫻10⫺5 0.020 19共42兲 ¯ ⫺1.14(12)⫻10⫺6 ⫺5.06(28)⫻10⫺4 ⫺3.42(36)⫻10⫺10 ¯ 1.010(45)⫻10 ⫺5 354 053共1431兲 9154.777共50兲 8854.055共49兲 34.1共2.8兲 31.68共13兲 39.01共12兲 ¯ 0.009 872共29兲 2.410共44兲 ⫺0.000 304共24兲 0.000 088共17兲 ⫺8.5(1.2)⫻10⫺8 ¯ ⫺7.77(57)⫻10⫺9 ⫺1.54(14)⫻10⫺7 9.63(81)⫻10⫺5 0.0755共96兲 ⫺1.30(38)⫻10⫺8 ¯ ⫺0.006 11共60兲 ¯ ¯ ¯ 479 141共1509兲 8981.1407共92兲 8783.7724共86兲 45.7共2.1兲 32.43共13兲 41.14共13兲 ¯ 0.010 727 3共71兲 1.4242共71兲 ⫺0.000 263 3共26兲 0.000 122 8共49兲 ¯ ⫺6.3(2.2)⫻10⫺9 ¯ ¯ 2.839(35)⫻10⫺4 ⫺0.0906共20兲 4.51(53)⫻10⫺9 ⫺0.000 018 07共18兲 6.618(99)⫻10⫺3 ¯ ¯ ¯ 195 668.8共6.6兲 7807.6329共74兲 7484.6850共64兲 22.5共2.3兲 27.886共88兲 35.376共76兲 ⫺2.24(80)⫻10⫺6 0.007 053 6共12兲 1.251 64共43兲 ⫺0.000 396 2共15兲 ⫺0.000 192 6共18兲 ¯ 5.8(1.1)⫻10⫺9 1.76(15)⫻10⫺9 ¯ 2.520(15)⫻10⫺5 ⫺0.002 130共29兲 ¯ 1.93(12)⫻10⫺7 ⫺2.80(83)⫻10⫺6 ⫺2.82(37)⫻10⫺11 ¯ ⫺1.015共82兲⫻10⫺7 rms of fit 0.069 0.037 0.036 0.088 0.038 In megahertz; errors quoted are 3 and apply to the last quoted decimal place. H rot⫹H cd⫽AN2x ⫹BN2y ⫹CNz2 ⫺D N 共 N2 兲 2 ⫺D NK N2 Nz2 2 2 4 4 ⫹d 1 N2 共 N⫹ ⫹N⫺ ⫹N⫺ 兲 ⫹d 2 共 N⫹ 兲 ⫹H N 共 N2 兲 3 ⫹H NK 共 N2 兲 2 Nz2 ⫹H KN N2 Nz4 ⫹L NK 共 N2 兲 2 Nz4 ⫹L KKN N2 Nz6 ⫹L NNK 共 N2 兲 3 Nz2 ⫹ P NNK 共 N2 兲 3 Nz4 ⫹ P NKK 共 N2 兲 2 Nz6 ⫹ P KN N2 Nz8 , 共2兲 where N⫾ ⫽Nx ⫾iNy . Watson’s D K and H K parameters were not employed because no transitions were recorded that depended on K a , K c alone. The one unpaired electron in CaNH2 gives rise to fine structure doublets that are accounted for by the third term in Eq. 共1兲. Owing to the molecules’ orthorhombic symmetry, the only components of the spin–rotation matrix that are nonzero are the diagonal terms and are included in H eff as H sr⫽1/2 兺␣ ⑀ ␣␣共 N␣ S␣ ⫹S␣ N␣ 兲 . 共3兲 In many cases it was found that inclusion of a centrifugal distortion correction to the spin–rotation improved the fit. The most significant form of this correction was found to be S H srcd⫽⌬ NK 共 N"S兲 N2 Nz2 terms including a minimal set of centrifugal distortion parameters, while fixing A to previously reported values of Marr et al.22 Typically, D N and D NK were sufficient initially, but d 1 and d 2 were used in some cases. After this crude fit had been refined, subsequent K a components and higher order centrifugal distortion constants were included until the total rms of the fit was below 100 kHz, the estimated experimental accuracy. During the addition of the higher order terms, A was allowed to float in the fit. Higher order centrifugal distortion corrections outside of the set in Eq. 共2兲 were necessary to model the highest K a doublets (h 1 ,h 2 ,h 3 ). During the final stage of this process, the centrifugal distortion to the spin–rotation was included in the Hamiltonian and refined to produce the constants presented in Table II. It should be noted that collapsed K a doublets 共in particular, K a ⫽4兲 were given as separate lines of input to the fitting routine. 共4兲 and was the only correction retained in the final analysis. The data were analyzed using the least-squares fitting routine, SPFIT, developed by Pickett and co-workers at JPL. To achieve a minimum set of parameters that accurately reproduced the data, it was found best to fit the K a ⫽0, 1, and 2 components only with the pure rotation and spin–rotation V. DISCUSSION One of the most interesting questions concerning metal amide species is their geometry, which is a reflection of their bonding. One test for a planar geometry concerns the inertial defect. To a first approximation, a small, positive value for this quantity, defined as ⌬⫽I c ⫺I a ⫺I b , indicates a planar molecule. However, as discussed in a series of papers by Oka and Morino,29,30 the interpretation of the inertial defect is not always this simple. The inertial defect is composed of three contributions: vibrational, electronic, and centrifugal distortion components, i.e., J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 Rotational spectrum of CaNH2 TABLE III. Inertial defects of CaNH2 and related species.a CaNH2 CaND2 NaNH2 SrNH2 H2CO NH2CN ND2CN NH2NC NH2CHO ND2CHO 3147 TABLE IV. Structural parameters of CaNH2. ⌬0 ⌬6 ⌬4 ⌬3 ⌬ elec Reference 0.157 0.210 0.079 0.177 0.0577 ⫺0.285 ⫺0.746 ⫺0.756 0.008 ⫺0.015 0.127 ¯ ¯ 0.053 ¯ ¯ 0.291 ¯ ¯ ⫺0.000 15 ⫺0.000 21 ¯ This work This work 34 33 31 26 26 27 31 31 r0 共millimeter wave兲 r CaN (Å) 2.126 共3兲 r NH (Å) 1.018共3兲 HNH 105.8° 共5兲 r 0 共optical兲a Ab initiob 2.122 共6兲 2.13 ¯ 1.02 105.5–106.0 共4.6兲 105.6° NH2c NH⫺d 2 ¯ ¯ 1.025 1.041共15兲 103.1° 102.1° 共3.1兲 a Reference 22. Reference 16. c Reference 36. d Reference 37. b a兲 In amu Å2. ⌬ 0 ⫽⌬ vib⫹⌬ elec⫹⌬ cent . 共5兲 These three contributions can add together for a given molecule in a way that the inertial defect is small, yet the species is not planar. This effect is seen in formamide, for example.31 The only certain way to access the meaning of the inertial defect is to examine each of the three terms, and establish whether there is some unusual cancellation in their summation. The inertial defect for CaNH2 was found to be ⌬ 0 ⫽0.157, as shown in Table III. This value is relatively small and positive. For CaND2, the inertial defect is ⌬ 0 ⫽0.210, and hence there is no sign change with deuterium substitution. Several of the terms contributing to ⌬ 0 can be evaluated for CaNH2. From the spin–rotation constants, the g tensor can be calculated,29 which results in ⌬ elec⫽⫺0.000 15. Hence, this contribution is not large and has not significantly reduced the value of ⌬ 0 . Part of ⌬ vib can also be derived, because rotational constants for the v 3 , v 4 , and v 6 modes have been measured. These three vibrational states make a contribution of 1/2(⌬ 3 ⫹⌬ 4 ⫹⌬ 6 )⫽0.236 共see Table III兲 to ⌬ vib . Because ⌬ cent is always small and positive,30 its contribution to ⌬ 0 is negligible; therefore, the major remaining contribution is from the inertial defects of the v 1 , v 2 , and v 5 modes. We did not observe these states. Considering the other numbers, 1/2(⌬ 1 ⫹⌬ 2 ⫹⌬ 5 ) must equal approximately ⫺0.079, which does not appear particularly unusual. The inertial defect determined for CaNH2 is also comparable to what has been observed for other planar metal amides, such as SrNH2, 32,33 and NaNH2, 34 which have ⌬ 0 ⫽0.177 and ⌬ 0 ⫽0.079. In contrast, the inertial defects observed for similar pyramidal species that undergo inversion are much larger and negative. As shown in Table III, ⌬ 0 for cyanamide and its deuterium isotopomer are ⌬ 0 ⫽⫺0.285 and ⌬ 0 ⫽⫺0.746; the value for isocyanamide is ⌬ 0 ⫽⫺0.756. Moreover, although the inertial defect is small and positive for formamide 共0.008兲, it changes sign 共⫺0.015兲 on deuterium substitution of the NH2 group. The inertial defects for CaNH2 and CaND2 are thus consistent with a planar geometry. An r 0 structure for CaNH2 was calculated on the basis of the rotational data and the results presented in Table IV. The bonds lengths and H–N–H bond angle were established from a nonlinear least-squares fit to the moment of inertia equations,35 using the A, B, and C rotational constants of CaNH2 and CaND2. The calculation results in r Ca–N ⫽2.126 Å and r N–H⫽1.018 Å, with H–N–H⫽105.8°. These values are in excellent agreement with ab initio calculations, which suggest r Ca–N⫽2.13 Å, r N–H⫽1.02 Å, and H–N–H ⫽105.6°, as well as estimates from the optical study of Marr et al.22 At face value the N–H bond length and H–N–H bond angle of CaNH2 are closer to those of the NH2 radical, which has r N–H⫽1.025 Å and H–N–H⫽103.1°, 36 than those of NH⫺ 2 . This ion has a N–H bond distance of 1.041 共15兲 Å and an H–N–H angle near 102.1° 共3.1兲.37 Such geometric similarities suggest that CaNH2 has a large degree of covalent character, as opposed to only a Ca⫹ NH⫺ 2 structure. Considering the uncertainties, however, the geometries of the NH2 radical and NH⫺ 2 are very similar, so a detailed comparison cannot be done. The Ca–NH2 bond length can be compared with other calcium–ligand bond distances. In Table V, these distances are presented for CaCH3, 38 CaNH2, CaOH,39 and CaF.40 The Ca–F and Ca–OH bond distances are quite similar 共1.955 and 1.985 Å兲, and scale approximately as the atomic radii. The Ca–CH3 and Ca–NH2 bond lengths, on the other hand, are appreciably longer 共2.326 and 2.126 Å兲. The lengthening of the calcium–ligand bond in these cases does not scale as atomic radii, and likely arises in both molecules from steric hindrance induced by the hydrogen atoms, and/or a change in bonding character. Another insight into the bonding in CaNH2 can be gained from examining the spin–rotation constants. Because the main contribution to this parameter is second-order spin– orbit coupling, the magnitude of the spin–rotation constant is a reflection of the p, d, f,... character of the orbital of the unpaired electron, when normalized by the rotational constant.41 The second-order contribution to ␥, the spin– rotation constant, can be expressed mathematically as TABLE V. Calcium–ligand bond lengths.a Species CaCH3 CaNH2 CaOH CaF a r Ca–L (Å) b 2.326 2.126 1.985 1.955 共1.952兲c From an r 0 structure unless stated otherwise. Assumes r C–H⫽1.1 Å. c R e structure in parentheses. b Reference 38 This work 39 40 3148 J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 M. A. Brewster and L. M. Ziurys TABLE VI. Normalized spin–rotation constants for calcium radicals. Species Ground state ␥ /B Reference CaOH CaF CaNH2 CaCCH CaCH3 CaH ⌺ ⌺ 2 A1 2 ⌺ 2 A 2 ⌺ 0.0035 0.0038 0.0041a 0.0064 0.0073b 0.010 39 40 This work 42 38 43 2 2 scheme. Comparison of the spin–rotation parameters determined in this work with other small calcium-bearing species additionally suggests primarily ionic bonding. Other studies are now in progress for metal amides to establish when, and if, more covalent species of this type exist. ACKNOWLEDGMENTS This research is supported by NASA Grant No. NAGW 5-3785 and NSF Grant Nos. CHE-98-17707 and AST-9820576. 1 Actually 2 关 ⑀ bb /B⫹ ⑀ cc /C 兴 . b Actually ( ⑀ bb ⫹ ⑀ cc )/2B. a ␥ /B⯝⫺2 ⫻ 2 兺 ␣ 具 ␣ 兩 Lx 兩 ␣ ⬘ 典具 ␣ ⬘ 兩 aLx 兩 ␣ 典 ⫹ 具 ␣ 兩 aLx 兩 ␣ ⬘ 典具 ␣ ⬘ 兩 Lx 兩 ␣ 典 E ␣ ⫺E ␣ ⬘ ⬘ , 共6兲 where ␣ represents the ground electronic state and ␣⬘ indicates near-by excited states; a is the spin–orbit constant of the excited state ␣⬘, and Lx is the x component of the angular momentum. For a symmetric top, ␥ /B is replaced by 1/2( ⑀ bb ⫹ ⑀ cc )/B, and for an asymmetric top, 冉 S. F. Rice, H. Martin, and R. W. Field, J. Chem. Phys. 82, 5023 共1985兲. T. Törring, W. E. Ernst, and J. Kändler, J. Chem. Phys. 90, 4927 共1989兲. 3 D. R. Lide and R. L. Kuczkowski, J. Chem. Phys. 46, 4768 共1967兲. 4 D. R. Lide and C. Matsumura, J. Chem. Phys. 50, 3080 共1969兲. 5 M. A. Anderson, W. L. Barclay, Jr., and L. M. Ziurys, Chem. Phys. Lett. 196, 166 共1992兲. 6 M. A. Anderson, M. D. Allen, W. L. Barclay, Jr., and L. M. Ziurys, Chem. Phys. Lett. 205, 415 共1993兲. 7 E. F. Pearson, B. P. Winnewisser, and M. B. Trueblood, Z. Naturforsch. A 31, 1259 共1976兲. 8 P. Kuijpers, T. Törring, and A. Dymanus, Chem. Phys. 15, 457 共1976兲. 9 D. A. Fletcher, M. A. Anderson, W. L. Barclay, Jr., and L. M. Ziurys, J. Chem. Phys. 102, 4334 共1995兲. 10 A. J. Apponi, M. A. Anderson, and L. M. Ziurys, J. Chem. Phys. 111, 10919 共1999兲. 11 C. J. Whitham, H. Ozeki, and S. Saito, J. Chem. Phys. 110, 11109 共1999兲. 12 J. J. van Vaals, W. L. Meerts, and D. Dymanus, Chem. Phys. 86, 147 共1984兲. 13 T. Törring, J. P. BeKooy, W. L. Meerts, J. Hoeft, E. Tiemann, and A. Dymanus, J. Chem. Phys. 73, 4875 共1980兲. 14 E. Kagi, K. Kawaguchi, S. Takano, and T. Hirano, J. Chem. Phys. 104, 1263 共1996兲. 15 J. J. Van Vaals, W. L. Meerts, and A. Dymanus, Chem. Phys. 82, 385 共1983兲. 16 P. F. Bernath, in Advances in Photochemistry, edited by D. C. Neckers, D. H. Volman, and G. Von Bunau 共Wiley, New York, 1997兲, Vol. 23, Chap. 1. 17 R. F. Wormsbecher, M. Trkula, C. Martner, R. E. Penn, and D. O. Harris, J. Mol. Spectrosc. 97, 29 共1983兲. 18 R. F. Wormsbecher, R. E. Penn, and D. O. Harris, J. Mol. Spectrosc. 97, 65 共1983兲. 19 A. M. Bopegedera, C. R. Brazier, and P. F. Bernath, J. Phys. Chem. 91, 2779 共1987兲. 20 C. J. Whitham, B. Soep, J. P. Visticot, and A. Keller, J. Chem. Phys. 93, 991 共1990兲. 21 C. J. Whitham and C. Jungen, J. Chem. Phys. 93, 1001 共1990兲. 22 A. J. Marr, M. Tanimoto, D. Goodridge, and T. C. Steimle, J. Chem. Phys. 103, 4466 共1995兲. 23 Z. Morbi, C. Zhao, and P. F. Bernath, J. Chem. Phys. 106, 4860 共1997兲. 24 Z. Morbi, C. Zhao, J. W. Hepburn, and P. F. Bernath, J. Chem. Phys. 108, 8891 共1998兲. 25 L. M. Ziurys, W. L. Barclay, Jr., M. A. Anderson, D. A. Fletcher, and J. W. Lamb, Rev. Sci. Instrum. 65, 1517 共1994兲. 26 W. G. Read, E. A. Cohen, and H. M. Pickett, J. Mol. Spectrosc. 115, 316 共1986兲. 27 E. Schäfer, M. Winnewisser, and J. J. Christiansen, Chem. Phys. Lett. 81, 380 共1981兲. 28 J. K. G. Watson, in Vibrational Spectra and Structure, edited by J. R. Durig 共Elsevier, Amsterdam, 1997兲, p. 1. 29 T. Oka and Y. Morino, J. Mol. Spectrosc. 6, 472 共1961兲. 30 T. Oka and Y. Morino, J. Mol. Spectrosc. 11, 349 共1963兲. 31 C. C. Costain and J. M. Dowling, J. Chem. Phys. 32, 158 共1960兲. 32 C. R. Brazier and P. F. Bernath 共unpublished兲. 33 J. M. Thompsen, P. M. Sheridan, and L. M. Ziurys 共unpublished兲. 34 J. Xin, M. A. Brewster, and L. M. Ziurys, Astrophys. J. 530, 323 共2000兲. 35 W. Gordy and R. L. Cook, Microwave Molecular Spectra 共Wiley, New York, 1984兲. 36 P. B. Davies, D. K. Russell, B. A. Thrush, and H. E. Radford, Proc. R. Soc. London, Ser. A 353, 299 共1977兲. 1 冊 1 ⑀ bb ⑀ cc . ⫹ 2 B C Clearly the value of the spin–rotation constant increases proportionally as the coupling between the ground and excited states via the Lx operator. Normalized spin–rotation constants for calcium-bearing radicals in doublet electronic ground states are shown in Table VI. The trend from ionic to covalent bonding is illustrated in the spin–rotation constants. The smallest values are for the most ionic compounds, CaF and CaOH, with ␥ /B ⫽0.0038 and 0.0035. The species with the more covalent bonds such as CaH have higher values near 0.01. The normalized spin–rotation parameter for CaNH2 is 0.0041— closer to those determined for ionic molecules than covalent species. 共An identical number is found for CaND2, as expected兲. Therefore, the spin–rotation constants suggest primarily ionic bonding in CaNH2, similar in degree to that found in halides and hydroxides. This property is consistent with the planar geometry, as well. In contrast, molecules like CaCCH and CaCH3 appear to have larger normalized constants and greater covalent character. VI. CONCLUSION Analysis of the pure rotational spectrum of CaNH2 in its ground vibrational and v 3 ⫽1, v 4 ⫽1, and v 6 ⫽1 excited states indicates that this metal amide is planar with C 2 v symmetry, as suggested by previous optical studies. There is no evidence in the spectra of an inverting molecule. Although the inertial defect for CaNH2 of ⌬ 0 ⫽0.157 is not as small as other planar species like H2CO, the value is consistent with other planar metal amides, and does not change sign on deuterium substitution. Measurements of the deuterium isotopomer, CaND2, have allowed for an r 0 structure determination that is consistent with an Ca⫹NH⫺ 2 -type bonding J. Chem. Phys., Vol. 113, No. 8, 22 August 2000 37 L. M. Tack, N. H. Rosenbaum, J. C. Owrutsky, and R. J. Saykally, J. Chem. Phys. 84, 7056 共1986兲. 38 M. A. Anderson and L. M. Ziurys, Astrophys. J. Lett. 460, L77 共1996兲. 39 L. M. Ziurys, W. L. Barclay, Jr., and M. A. Anderson, Astrophys. J. Lett. 384, L63 共1992兲. 40 M. A. Anderson, M. D. Allen, and L. M. Ziurys, Astrophys. J. 424, 503 共1994兲. 41 X. Liu, S. C. Foster, J. M. Williamson, L. Yu, and T. A. Miller, Mol. Phys. 69, 357 共1990兲. Rotational spectrum of CaNH2 3149 M. A. Anderson and L. M. Ziurys, Astrophys. J. Lett. 444, L57 共1995兲. W. L. Barclay, Jr., M. A. Anderson, and L. M. Ziurys, Astrophys. J. Lett. 408, L65 共1993兲. 44 See EPAPS Document No. E-JCPSA6-113-004032 for a complete list of measured transition frequencies for CaNH2 ( v ⫽0, v 3 ⫽1, v 4 ⫽1, v 6 ⫽1) and CaND2 ( v ⫽0). This document may be retrieved via the EPAPS homepage 共http://www.aip.org/pubservs/epaps.html兲 or from ftp.aip.org in the directory /epaps/. See the EPAPS homepage for more information. 42 43
© Copyright 2025 Paperzz