Photoabsorption and photoionization of the valence and inner (P 2p, 2s) shells of PF3: Absolute oscillator strengths and dipole-induced breakdown pathways

Chemical Physics ELSEVIER

Chemical Physics 215 (1997) 397-418

Photoabsorption and photoionization of the valence and inner (P 2p, 2s) shells of PF3: absolute oscillator strengths and dipole-induced breakdown pathways Jennifer W. Au, Glyn Cooper, C.E. Brion * Department of Chemistry, The University of British Columbia, 2036 Main Mall, Vancouver, B.C. V6T IZI, Canada

Received 26 September 1996; in final form 6 November 1996

Abstract Absolute photoabsorption oscillator strengths (cross sections) for the valence and inner (P 2p, 2s) shells of phosphorus

trifluoride (PF3) have been measured using dipole (e, e) spectroscopy in the 5-300 eV equivalent photon energy region (1 eV FWHM). The discrete spectral structures in the valence region and in the vicinity of the phosphorus 2p inner shell have also been studied at higher resolution (0.05 and 0.1 eV FWHM, respectively). Using these data and the S( - 2) sum rule, the electric-dipole polarizability for PF 3 has been determined for the first time. In addition, dipole (e, e + ion) coincidence spectroscopy has been used to obtain the photoionization efficiencies, photoion branching ratios, and absolute partial oscillator strengths for the molecular and dissociative photoionization channels of PF 3 from the first ionization threshold up to and above the phosphorus L shell edges to 175 eV. A consideration of the photoabsorption and photoionization measurements together with thermochemical data and results from previously published photoelectron branching ratios and photoelectron-photoion coincidence (PEPICO) studies provides information on the dipole-induced breakdown pathways of PF3 in the photon region below 100 eV. © 1997 Elsevier Science B.V. All rights reserved.

1. Introduction Absolute photoabsorption and photoionization oscillator strengths (cross sections) in the vacuum UV and soft X-ray regions provide useful information for the quantitative study of radiation damage and also for astrophysics, aeronomy, and electron microscopy [1,2]. Such information is also required for the quantitative evaluation of quantum mechanical calculations. However, the valence shell spectral data available in the literature for electronic excitation of the third row molecule phosphorus trifluoride (PF 3) is " Corresponding author.

very limited, with few measurements o f absolute oscillator strengths. Following the relative photographic work of Humphries et al. [3] in the limited valence-shell region, 6 . 9 - 1 1 . 8 eV, M c A d a m s and Russell [4] reported absolute optical density measurements from 5 to 10 eV. Inner-shell excitation in PF 3 has been examined by synchrotron radiation [5-7], but only the work of Ishiguro et al. [5] reported absolute photoabsorption spectra for the P 2p and 2s inner-shell regions. In other related work Sodhi and Brion [8,9] obtained relative intensities for valence and inner shell (P 2p, 2s and F ls) excitation of PF 3 using electron-energy-loss spectroscopy (EELS) at 2.5 keV impact energy.

0301-0104/97/$17.00 Copyright © 1997 Elsevier Science B.V. All rights reserved. PII S 0 3 0 1 - 0 1 0 4 ( 9 6 ) 0 0 3 7 0 - 9

398

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418

The outer valence ionization of PF 3 has been studied extensively by photoelectron spectroscopy (PES) [6,10-14] and photoelectron-photoioncoincidence (PEPICO) spectroscopy [15-17]. Both the PES studies by Green et al. [14] and Viiyrynen et al. [6], obtained using synchrotron radiation, reported photoelectron branching ratio data for the seven outer valence bands of PF 3. In contrast, no data are available for the inner valence ion states, 3e- I and 5a~- t, other than the theoretical ionization energies determined using SCF-Xa-SW [18] and continuum MS-Xc~ [19] calculations. Photoionization mass spectroscopy [20,21], electron impact [22-24], and threshold-photoelectron-photoion-coincidence (TPEPICO) [25] measurements have contributed additional information on the appearance potentials and relative intensities of molecular and dissociative ions observed in the breakdown of phosphorus trifluoride. In the inner-shell region of PF 3, the phosphorus L and fluorine K photoelectron core binding energies have been recorded using XPS [26-30], and synchrotron radiation has been used to determine the spin-orbit splitting between the 2p~-]2 and 2p392 components of the P 2p-~ ionization [6]. In other work, the core photoionization of PF 3 has also been studied by L2.3VV Auger spectroscopy [6,31], synchrotron radiation [32], and photoion-photoion coincidence (PIPICO) spectroscopy [33]. Fast electron scattering techniques at negligible momentum transfer have continued to provide a reliable alternative to direct optical absorption experiments for the measurement of absolute photoabsorption oscillator strength (cross-section) data of atoms and molecules [2,34-37]. The results obtainable using dipole (e, e) and dipole (e, e + ion) coincidence spectroscopies have been shown to be quantitatively equivalent to measurements obtained from photoabsorption and photoionization mass spectroscopy, respectively [2]. In addition to being easily tunable over a wide energy range (e.g., 5-300 eV in the case of PF3), the technique of dipole (e, e) spectroscopy can be used to obtain absolute measurements (via the VTRK or S ( - 2) sum rules) and does not require the logarithmic transformation involved in the BeerLambert law for the determination of absolute oscillator strengths. Use of the Beer-Lambert law technique in direct optical measurements can result in large errors in absolute cross sections caused by

"line-saturation" (bandwidth/linewidth interaction) effects [34-36,38]. Furthermore, use of the dipole (e, e) method does not involve the corrections for higher orders and stray light which are usually required when using monochromated synchrotron radiation for photoabsorption measurements. In this work, we now report the absolute photoabsorption spectrum of phosphorus trifluoride measured using dipole (e, e) spectroscopy from 5 to 300 eV at an energy resolution of 1 eV FWHM. The discrete features and absolute intensities in the valence and inner (P 2p, 2s) shell regions have been studied in greater detail at higher resolution (0.05-0.1 eV FWHM). In addition, the photoionization efficiencies, photoion branching ratios, and absolute partial photoionization differential oscillator strengths (PPOS) for the molecular and dissociative photoionization of PF 3 have been obtained using dipole (e, e + ion) coincidence spectroscopy from the first ionization threshold up to the P 2s edge. These data have been used together with photoelectron branching ratios [6,14] and PEPICO [15,16] data to obtain quantitative information concerning the dipole-induced breakdown scheme for PF 3 under vacuum ultraviolet and soft X-ray radiation.

2. Experimental methods The dipole (e, e) and dipole (e, e + ion) coincidence spectrometers used to obtain absolute photoabsorption oscillator strengths and time-of-flight (TOF) mass spectra for PF 3 have been described in detail in earlier publications [34,39-44]; therefore, only brief accounts of the instruments and experimental procedures will be presented here. In the high (0.05-0.1 eV FWHM) and low (1 eV FWHM) resolution dipole (e, e) spectrometers, electrons with high incident energies, E 0, (3000 and 8000 eV, respectively) were employed to excite the gaseous PF 3 sample. The high-energy, inelastically scattered electrons were energy-loss (E) analyzed about a mean scattering angle of zero degrees and were detected by a channeltron electron multiplier. Under such conditions of high impact energies with E << E 0 and very small scattering angles (3.0 × 10 -5 [45] and 1.4 × 10 - 4 s r [41], respectively), the momentum transfer is negligi-

J.W. Au et al./ Chemical Physics 215 (1997) 397-418

ble and electric-dipole transitions dominate the electron-energy-loss spectra [34]. The resulting background-subtracted, relative electron-energy-loss spectra were converted to relative photoabsorption spectra using the energy-loss dependent Bethe-Born conversion factors previously determined for the high [34,45] and low [40,41] resolution dipole (e, e) spectrometers [2,34]. In these experiments the energy-loss scale is equivalent to photon energy. The absolute differential oscillator strength scale for the long-range (5-300 eV), low resolution photoabsorption data was established using the valenceshell Thomas-Reiche-Kuhn (VTRK) sum-rule [46] (see Section 3.2.1 for details). The high resolution, relative photoabsorption spectra of the valence (5-50 eV) and P 2p (132-150 eV) shells were put onto an absolute scale by normalization to the low resolution data in the smooth continuum regions where no sharp structures exist. The absolute photoabsorption oscillator strengths of the present measurements are estimated to have an accuracy of better than _ 5%. The reliability and accuracy of the dipole (e, e) method [2] has been confirmed by the excellent agreement of the experimental oscillator strengths for helium [34] and molecular hydrogen [35] with highly accurate ab initio calculations. The energy scale for the high resolution data was calibrated by simultaneous admission of PF 3 and helium and referencing the PF 3 spectral features to the 2 J P ~ 1 tS transition of helium at 21.218 eV [47]. The low resolution data (1 eV FWHM) were energy-calibrated by adjusting the electron analyzer pass energy in the spectrometer so that the energy positions of the spectral profiles observed matched those obtained in the high resolution spectrum after it had been convoluted with a 1 eV FWHM Gaussian function (in order to mathematically degrade the resolution of the high resolution spectrum to the 1 eV FWHM of the low resolution spectrum). TOF mass spectra were obtained for PF 3 using P Is F IS P 2s P 2p inner shells

(5al)2(3e) 4

inner valence

(7e)°(9al)°(10al)°(1 lal)°(8e)°(9e) 0 virtual valence

399

dipole (e, e + ion) coincidence spectroscopy by detection of fast, forward-scattering electrons in coincidence with positive ions extracted at 90 ° to the incident beam at selected values of the energy loss (equivalent photon energy). The ion extraction voltages and TOF analyzer lens system ensure uniform collection of ions with up to -- 20 eV excess kinetic energy of fragmentation [40-42]. This is essential if accurate branching ratios and partial oscillator strengths are to be obtained for molecular and dissociative photoionization processes. A single stop timeto-amplitude converter (using electron start, ion stop) was used to determine the TOF of the ions. Mass sensitivity correction factors determined for the microchannel-plate ion detector used in this work [44] were applied to the TOF mass spectra to obtain the ion fragmentation branching ratios. The detection efficiency of the ion extraction and transportation system of the spectrometer is -~ 10% using the microchannel-plate detector [44]; thus, both ions formed in Coulomb explosions in dissociative double ionization processes will be detected in approximately equal proportions (i.e. 10% for the fastest ion and 9% (i.e. 10% of 90%) for the slower ion). The sample of PF 3 with 99% minimum stated purity was obtained from Ozark-Mahoning and was used without further purification. No impurities could be detected in the TOF mass spectra of PF 3.

3. Results and discussion 3.1. E l e c t r o n i c structure

The phosphorus trifluoride molecule, which has trigonal pyramidal geometry and belongs to the C3v symmetry group, has a ground state (~A~) electronic configuration in the independent particle model given by (6al)2(4e)4(7al)2(5e)4(la2)2(6e)4(8al)2 outer valence

400

J.W. Au et al. / Chemical Physics 215 (1997) 397-418

where the l la~, 8e, and 9e virtual valence orbitals involve P 3d character [19,48]. The vertical ionization potentials (VIP's) for the seven outer-valence ion states have been determined to be 12.27 (8a~-~), 15.88 ( 6 e - I ) , 16.30 (lab-t), 17.46 (5e-I), 18.60 (7a~- 1), 19.50 (4e- ]), and 22.55 eV (6a~-I ) using He I and He II PES [13]. No experimental data or

, ,,ta; AE= > 0

60

many-body Green's function calculations are available for the two inner-valence ionization processes (3e-t and 5a~-~) which are each expected to be split into a range of final ion states (poles) by electron correlation effects [49]. Therefore, in the present work the independent particle orbital energies of 35.48 and 36.47 eV, respectively, from the contin-

eVfwhm

-. • "

1

60

. . . . . . . , a ' n , , J ~ ' , , " ~ ' - " " = - ~ -; - ~ -'-- -"" - - T - --". . . . . . . . . . . . . .

40

40

tO) 20

0

E

•

-'

•

•

0

valence Llence ~

-

m

.

.

.

.

i

50

-

-

-

20

.Q

100 v

U)

0

O •~

O ~

6o

40 20

M,,,,,M,,,,,o ~

£3

50

0

100

150

200

250

0 300

•° " ~•" ~ " ~

LR dipole (e,e), Present w o r k Ph Abs [5] Ig Atomic sum (P+3F), T h e o r y [55] z~ Atomic sum (P+3F), Exp't [56] ....... Valence extrapolation

j~~ k

•

~

40

60

80

100

120

P 2P3/2'1/2

10

6 4

. D

lh..

O o o c-Q_

oc'o

20 0

20

O

¢.O

£

40

i

c-"

09 60

o ~

8

tO

o m

O

b)

E

~

-. ........:...... ;........:....... vr___----'-Ti

0

o o~ o o t--13..

6 --.~..

2 (c)

P2p,2s

~

..... ~:"iiiiil;i

4

................................................................... 2

0

0 150

200

Photon

250

300

Energy (eV)

Fig. I. Absolute photoabsorption differential oscillator strengths for PF 3 obtained at a resolution of 1 eV FWHM: (a) Long-range spectrum measured from 5 to 300 eV; (b) the valence region (5-130 eV); and (c) the P 2p and 2s inner-shell regions (125-300 eV) shown on an expanded scale. The dashed lines represent the valence-shell extrapolation curve obtained by fitting equation (1) to the high-energy portion of the valence spectrum from 85 to 130 eV (see text for details). The inset in (a) shows the photoionization efficiencies (rti) of PF 3 measured from I1 to 130 eV.

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418

uum M S - X a calculations of Powis [19] have been used as average estimates of the IP's for these ionization processes. The ordering of the virtual valence orbitals is that determined from ab initio [5] and M S - X a [19,48] calculations. The mean binding energy for the P 2p edge, as measured by XPS [28], is 142.07 eV, and values of 0.86 eV [6] and 0.90 eV [9] have been reported for the spin-orbit splitting between the 2p3/' 2 and 2p~-92 components. Based on a statistical weighting of 2:1, the 2p3/12 and 2p~-/12 edges have been estimated to be at 141.77 and 142.67 eV [9]. The value for the P 2s VIP, as measured by XPS [28], is 199.49 eV.

3.2. Photoabsorption measurements 3.2.1. Long-range photoabsorption oscillator strength spectra The low resolution (1 eV FWHM) absolute photoabsorption spectrum of PF 3 spanning the valence, P 2p, and P 2s regions (5-300 eV) is shown in Fig. l(a). Panels (b) and (c) show the valence (5-130 eV) and inner (P 2p, 2s) shell regions, respectively, in greater detail on expanded scales, and the corresponding data are given numerically in Tables 1 and 2. Also shown for comparison in Fig. l(c) is the synchrotron radiation data above 142 eV reported by Ishiguro et al. [5], which are in excellent agreement with the present work except in the vicinity of the second continuum shape resonance (159 eV) where they are lower than the present measurements by --~ 6%. The dashed curves in (a) and (c) represent the estimated contribution from the valence-shell spectrum (determined from the VTRK sum-rule normalization and extrapolation procedure discussed below) to the total photoabsorption above 130 eV. The TRK sum-rule in the valence-shell modification [46] has been used to normalize the long-range, relative photoabsorption spectrum (1 eV FWHM) obtained in the present work (see Section 2) to absolute values. In this procedure the contribution of PF 3 to the valence shell above 130 eV (the onset of the P 2p spectrum) was estimated by fitting a polynomial of the form

df dE

= A E - 2 + B E -3 + C E -4,

(1)

401

to the relative valence-shell photoabsorption data (d f / d E ) over the energy range, 85-130 eV, and extrapolating to infinite energy. The values determined for the best fit parameters, A, B, and C, are - 2 . 3 2 7 4 × 104 eV, 2.2561 X l07 eV 2, and - 1.1931 X 109 eV 3, respectively. This type of polynomial function has been found to be effective in previous studies [37,50-52] for fitting the higher energy portions of measured relative oscillator strength distributions of polyatomic molecules. The relative valence-shell photoabsorption spectrum was then integrated from the first excitation threshold to infinite energy, and the total area was set to an oscillator strength value of 27.00. This value corresponds to 26 valence electrons in PF 3 plus a small estimated contribution (1.00) for Pauli excluded transitions from the inner shells to the occupied valence orbitals [53,54]. Theoretical [55] and experimental [56] atomic differential oscillator strength sums (or atomic mixture rules (AMR) [57]) for the constituent atoms (P + 3F) are plotted together with the present photoabsorption results in Fig. l(b) and (c). The theoretical [55] valence-shell atomic data have also been used to model the underlying valence contribution (Fig. l(c)) in the P 2p, 2s photoabsorption region above 130 eV, and are in good agreement with the extrapolation based on the curve fit to 300 eV. It can also be seen in Fig. l(b) and (c) that such atomic sums provide good estimates of the oscillator strengths in the valence continuum above 40 eV and in the core continuum above -- 170 eV (i.e. in regions, away from the ionization edges, where any molecular effects become negligible). The static electric-dipole polarizability ( a N) of a molecule can be determined from the S(-2) sum rule, ,~1

s(-2)

= aN = f£,

df

d--/(e) dE,

(2)

if a sufficiently accurate absolute photoabsorption spectrum is available up to at least = 100 eV [58-60] (data above 100 eV in general contribute less than 1% to the summation). Alternatively, equation (2) may be used to establish the absolute oscillator strength scale if a sufficiently accurate value of the static dipole polarizability is available [60]. Since the low energy region of the oscillator strength is heavily weighted by the 1 / E 2 factor in S ( - 2 ) , the

402

J.W. Au et al. / Chemical Physics 215 (1997) 397-418

Table 1 Absolute differential oscillator strengths for the total photoabsorption and the dissociative photoionization of PF 3 (1 eV FWHM) up to 130 eV Photon

Differential oscillator strength (10 -2 eV- t) a

energy (eV)

Photo-absorption

5,0 5,5 6,0 6,5 7,0 7.5 8.0 8,5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 23.5 24.0 24.5 25.0 25.5 26.0 26.5 27.0 27.5 28.0 28.5

0.50 0.59 0.16 0.13 4.00 24.96 40.27 15.14 19.46 22.02

PF~

PF~

0.69 2.54 4.71 6.41 9.61 11.53 12.88 13.03 10.57 9.03 8.09 8.20 8.64 8.36 8.18 7.43 6.77 6.60 6.09 5.95 5,43 5,44 5,02 4.66 4442 3,89 3,79 3,24 3,18 3,04 2,81 2,65 2.52 2.42 2.34 2.19

0.03 0.04 0.06 0.06 0.14 0.19 0.53 1.41 3.76 8.21 18.72 23.51 30.10 35.48 41.23 44.46 47.84 50.77 53.61 55.84 57.85 58.46 57.48 56.63 54.85 53.89 51.34 49.50 46.20 43.67 41.81 40.42 37.40 35.58 34.58 33.63

PF +

Photoionization PF 3 +

PF~ +

P+

PF 2+

F+

p2 +

efficiency r~i

11.03

11.21 16.21 17.22 17.48 22.56 30.75 34.46 43.23 47.92 47.50 43.00 39.29 38.60 40.44 43.84 49.41 51.89 54.61 57.37 59.69 61.97 63.54 64.67 63.85 63.40 61.76 60.56 58.90 56.72 53.58 51.09 49.13 47.82 44.62 42.61 41.71 40.36

0.04 0.15 0.27 0.29 0.32 0.34 0.31 0.30 0.30 0.40 0.68 0.82 0.96 1.00 b

0.18 0.27 0.77 1.34

2.12 2.49 2.77 3.77 3.98 4.20 4.27 4.38 4.62 4.54 4.40 4.42 4.28

0.11 0.13 0.13 0.16 0.20 0.37 0.25

J.W. A u et al.// Chemical Physics 215 (1997) 397-418

403

Table 1 (continued) Photon

Differential oscillator strength (10-2 eV- ~) a

energy (eV)

Photo-absorption

PF3~

PF~

PF +

29.0 29.5 30.0 32.0 34.0 36.0 38.0 40.0 42.0 44.0 46.0 48.0 50.0 55.0 60.0 65.0 70.0 75.0 80.0 90.0 100.0 110.0 120.0 130.0

38.29 37.67 36.97 32.77 32.15 30.39 28.86 27.82 25.60 23.73 22.62 21.47 21.10 19.88 18.14 16.53 14.63 13.20 11.77 10.03 8.27 6.84 5.79 4.73

2.09 2.07 1.88 1.60 1.46 1.38 1.26 1.30 1.16 1.15 1.05 1.05 1.02 1.01 0.86 0.84 0.72 0.61 0.58 0.46 0.38 0.26 0.22 0.18

31.73 31.08 30.48 26.43 24.89 22.96 20.97 19.40 17.74 16.28 15.46 14.40 14.12 12.90 11.66 10.39 8.96 7.75 6.68 5.49 4.38 3.69 2.91 2.33

4.01 3.84 3.80 3.15 3.03 2.76 2.52 2.42 2.15 1.99 1.81 1.80 1.63 1.56 1.43 1.22 1.08 0.99 0.86 0.73 0.58 0.46 0.38 0.29

PF~ +

0.08 0.09 0.15 0.13 0.15 0.14 0.16 0.14 0.12 0.10 0.07 0.09 0.07 0.08 0.06 0.05 0.02 0.04 0.03

PF~ +

P+

0.19 0.19 0.21 0.16 0.20 0.17 0.19 0.17 0.17 0.13 0.13 0.12 0.13 0.09 0.07 0.06 0.04

0.46 0.67 0.82 1.59 2.47 2.60 2.81 2.75 2.41 2.14 2.09 1.99 1.97 1.90 1.70 1.62 1.49 1.44 1.35 1.19 0.95 0.84 0.75 0.63

PF 2+

F+

p2 +

0.11 0.23 0.30 0.37 0.42 0.46 0.48 0.49 0.42 0.41 0.40 0.40 0.35 0.30 0.25 0.20 0.17

0.31 0.63 1.19 1.49 1.58 1.51 1.55 1.45 1.51 1.56 1.51 1.56 1.51 1.56 1.48 1.41 1.32 1.09 1.05 0.90

0.08 0.16 0.21 0.23 0.24 0.23 0.23 0.20 0.24 0.16 0.17 0.15

Photoionization efficiency "r/i

a O" (Mb) = 109.75 d f / d E (eV-~). b The photoionization efficiency is normalized to unity above 17.5 eV (see text for details).

absolute p h o t o a b s o r p t i o n data ( 5 - 3 0 0 e V ) obtained in the present w o r k are m o r e than sufficient for deriving the electric-dipole polarizability for P F 3. This is o f s o m e interest since no e x p e r i m e n t a l values o f ot s for P F 3 h a v e b e e n published thus far. In o t h e r w o r k from this laboratory e q u a t i o n (2) has b e e n used to obtain a N v a l u e s f r o m the V T R K sum-rule normalized spectra for other p h o s p h o r u s halides, PCi 3 [61] and P F 5 [62], w h i c h are in g o o d a g r e e m e n t with experimental and theoretical estimates reported in the literature [ 6 3 - 6 5 ] for these molecules. In the present w o r k the electric-dipole polarizability o f P F 3 has been d e t e r m i n e d to be 44.32 × 10 -25 c m 3 using the photoabsorption data in Table 1 and Fig. 1 in equation (2). This value differs significantly f r o m the theoretical estimates p e r f o r m e d by Lippincott et al. (29.02 X 10 -25 c m 3) [66] and P a n d e y et al. (29.154 × 10 -25 c m 3) [67] obtained using the delta-function potential model. H o w e v e r it should be noted that

difficulties in c o m p u t i n g the dipole polarizabilities for group 15 trifluorides h a v e been reported by these authors [66,67] since the experimental data upon which the necessary polarity corrections can be based are not available. T h e r e f o r e at best the theoretical values o f a N reported for P F 3 [66,67] are questionable. F u r t h e r m o r e , the reliability o f the presently reported oscillator strength scale is supported by the g o o d a g r e e m e n t with the data of Ishiguro et al. [5] in the P 2p, 2s r e g i o n and with the atomic s u m s [55,56] o v e r a w i d e range in both the v a l e n c e and inner shell c o n t i n u u m regions. 3.2.2. V a l e n c e - s h e l l p h o t o a b s o r p t i o n s p e c t r u m

T h e v a l e n c e - s h e l l absolute photoabsorption spectrum o f PF 3 f r o m 5 to 50 e V obtained at 0.05 e V F W H M resolution in the present w o r k is presented in Fig. 2(a) together with the positions o f the vertical ionization potentials o f the seven o u t e r - v a l e n c e ion

404

J.W. Au et al. / Chemical Physics 215 (1997) 397-418

states as given in Section 3.1. Fig. 3 shows an expanded view of the discrete region (7-23 eV) of the spectrum; the assignments given on the manifolds have been taken from Ref. [8] with the exception of feature 3 which was attributed [8] to both the 4s Rydberg and 9a~ virtual valence transitions. In reference [8] transferability of both virtual valence

and Rydberg term values was assumed in assigning the valence shell spectrum. Therefore, the positions shown for the virtual valence transitions will be, at best, very approximate. The Jahn-Teller split band (see features 1 and 2 in Fig. 3) centered at a mean energy of 8.06 eV were previously assigned to 7e(tr*) ~ 8a~-1 [4,8]. The

Table 2 Absolute differential oscillator strengths a for the total photoabsorption b of PF 3 in the P 2p, 2s inner-shell region from 125 to 300 eV Photon energy (eV)

Differential oscillator strength (10 -z eV- l)

Photon energy (eV)

Differential oscillator strength (10- 2 eV- t)

Photon energy (eV)

Differential oscillator strength (10- 2 eV- i)

Photon energy (eV)

Differential oscillator strength (10 2 eV- i)

125.0 125.5 126.0 126.5 127.0 127.5 128.0 128.5 129.0 129.5 130.0 130.5 131.0 131.5 132.0 132.5 133.0 133.5 134.0 134.5 135.0 135.5 136.0 136.5 137.0 137.5 138.0 138.5 139.0 139.5 140.0 140.5 141.0 141.5 142.0 142.5 143.0

5.02 5.11 5.04 5.02 4.98 4.94 4.90 4.83 4.75 4.69 4.73 4.65 4.66 4.59 4.55 4.52 4.52 4.47 4.70 5.06 5.28 5.48 5.73 5.63 5.11 5.13 5.57 5.83 5.72 5.54 5.32 5.23 5.21 5.20 5.28 5.36 5.53

143.5 144.0 144.5 145.0 145.5 146.0 147.0 148.0 149.0 150.0 151.0 152.0 153.0 154.0 155.0 156.0 157.0 158.0 159.0 160.0 161.0 162.0 163.0 164.0 165.0 166.0 167.0 168.0 169.0 170.0 171.0 172.0 173.0 174.0 175.0 176.0 177.0

5.91 6.26 6.60 7.03 7.30 7.51 7.48 7.24 6.88 6.52 6.43 6.36 6.46 6.64 6.85 7.16 7.38 7.47 7.51 7.47 7.38 7.19 7.05 6.92 6.80 6.68 6.54 6.41 6.36 6.20 6.12 6.08 6.01 5.91 5.85 5.82 5.80

178.0 179.0 180.0 181.0 182.0 183.0 184.0 185.0 186.0 187.0 188.0 189.0 190.0 191.0 192.0 193.0 194.0 195.0 196.0 197.0 198.0 199.0 200.0 202.0 204.0 206.0 208.0 210.0 212.0 214.0 216.0 218.0 220.0 222.0 224.0 226.0 228.0

5.68 5.66 5.66 5.56 5.49 5.46 5.40 5.36 5.32 5.30 5.31 5.29 5.41 5.63 5.99 5.43 5.22 5.20 5.09 5.09 5.11 5.02 4.91 4.79 4.81 4.87 4.80 4.74 4.77 4.69 4.72 4.59 4,57 4.53 4.48 4.44 4.39

230.0 232.0 234.0 236.0 238.0 240.0 242.0 244.0 246.0 248.0 250.0 252.0 254.0 256.0 258.0 260.0 262.0 264.0 266.0 268.0 270.0 272.0 274.0 276.0 278.0 280.0 282.0 284.0 286.0 288.0 290.0 292.0 294.0 296.0 298.0 300.0

4.33 4.24 4.16 4.12 4.01 3.96 3.92 3.84 3,79 3,75 3,66 3,61 3.57 3.52 3,49 3.45 3.35 3.35 3.29 3.27 3.24 3.21 3.16 3.11 3.06 3.05 2.98 3.00 2.92 2.93 2.89 2.85 2.81 2.79 2.75 2.72

a o- (Mb) = 109.75 d f / d E ( e V - I ) . b The partial photoabsorption oscillator strengths for the P 2p, 2s shells can be obtained by substituting the A, B, and C parameters given in Section 3.2.1 into Eq. (1) and then subtracting the resulting valence-shell contributions from the corresponding total photoabsorption values in this table.

J.W. Au et al. / Chemical Physics 215 (1997) 397-418 100

...~100 "7

>

o

O u t e r v a l e n c e VIP's

80

~

o

~

Valence shell

'T > • 0

t"

+

80

....

60

I

~

t

I "

7e*

!

+

20 AE = 0.05 eV fwhm ¢n 0

e-

£

20

30

40

50

30

--

0

5

10

15

0

4s < - - 8a-11~ i . ~

o

£3

¢-

o o9 J:l

.

10

5s

;,.

.

7

8

9

I

;

i

,/'N,

v

/ I

I

I

I

,

.4

5

6.

7

.I

10

11

12

13

14

3

(b)

.

.

.

.

.

11

.

.

o o tQ.

.

I

e

~

40

I

7e*

20

integrated oscillator strength value of 0.4604 (Table 3) presently determined for this transition is very close to the value of 0.48 reported in the optical density study by McAdams and Russell [4]. Two vibronic progressions centered at 9.52 and 11.08 eV (Fig. 2(b)) are clearly observed in the presently reported spectrum. The relative [3] and absolute [4] optical absorption spectra previously measured in the very limited energy regions, 6.911.8 and 5-10 eV, respectively, also showed these vibrational structures. In the present work, 23 vibronic bands have been deconvoluted from the first progression and 29 have been identified from the second• The energy positions for these vibronic bands are in good agreement with those reported by Humphries et al. [3] within experimental uncertainty. The average vibrational separation, 460 cm- J (0.058

_1!

4s ~

5S

I

,~

5S i

~ 7al 1

la~

5s 0

+

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

I

l

I

8

9

10

15

J..............

16

I

: ..............................................................................

11

17

I

I

12

13

18

19

20

21

22

23

Photon Energy (eV)

Photon Energy (eV) Fig. 2. (a) Absolute high resolution (0.05 eV FWHM) photoabsorption spectra of the valence-shell region of PF 3 from 5 to 50 eV. The hatched lines indicate positions of the outer valence VIP's as determined by high resolution PES [13]. Panel (b) shows the two vibronic progressions in the bands at 9.52 and 11.08 eV. Note that the photoabsorption differential oscillator strengths can be converted to absolute photoabsorption cross sections using the relation, o- (Mb) = 109.75 X ( d f / d E ) ( e V - I ).

/

.l~,,J

' .......5S II............. ~~ .... ~ 4eI

~,"1~

4S

I

14

12

7~j'"

('~'~

£=

9

'.

II 12

60

20

i,.~~~,,~

.~ 4s

.

.£

oe13- 0

I 4s -1

t'-

[-Fr~ IIIIIIIFIIIIIIjl t 1

vl=O

Band separation = +460 cm q

~

20

(/'3

10

40.

~ 0

I

4s

~]

;/e" ~

40 t-

I

7e*

,,

E~ r- 40

c- 60

0

405

Fig. 3. The features observed in the discrete high resolution (0.05 eV FWHM) valence-shell absolute photoabsorption spectrum from: (a) 7 to 14 eV and (b) 14 to 23 eV, The assignments are taken from Ref. [8]. The integrated oscillator strengths for the eight energy regions separated by vertical dotted lines are given in Table 3.

eV), for both progressions corresponds closely to the u 2 symmetric bending mode of PF 3 (487 cm -~) in the IR spectrum [68]. Table 3 Absolute photoabsorption oscillator strengths for regions of the valence-shell spectrum of PF 3 Integrated energy region a

(eV) 7.39-8.68 8.68-10.26 10.26-12.16 12.16-13.62 13.62-16.24 16.24-17.34 17.34-18.54 18.54-20.88

Oscillator strength ( 10- 2 ) Present work

Ref. [4]

46.04 25.42 28.42 41.56 117.43 42.78 58.23 136.80

48 28

a See Fig. 3 for the respective spectral regions.

406

J.W. Au et a l . / Chemi,'al Physics 215 (1997) 397-418

Table 4 Absolute oscillator strengths for the 4s ,---8a~ I vibronic bands of the valence-shell photoabsorption spectrum of PF3 Vibrational quantum number (v') 0 1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Photon energy (eV) 8.82 8.9l 8.98 9.04 9.10 9.16 9.22 9.28 9.33 9.39 9.45 9.50 9.56 9.62 9.68 9.73 9.79 9.85 9.91

!9

9.97

20 21 22

10.04 10.11 10.19

Oscillator strength (10 "~) 0.09 0.15 0.32 0.49 0.70 0.92 1.20 1.45 1.70 1.87 1.98 2.04 2.02 1.96 1.73 1.52 1.27 1.02 0.82 0.61 0.45 0.31 0.18

The H O M O (8a~) of PF 3 is primarily the phosphorus lone-pair orbital [69], and ionization of an electron from the H O M O results in the manifestation of vibronic structures in the 8a~-l ion ground state with an interval o f 470 c m - t (averaged from v' = 5 to v' = 20) in the high resolution photoelectron spectrum reported by Maier and Turner [13]. The intensity distribution and width o f the F r a n c k - C o n d o n vibrational envelope of the 9.52 eV band of the dipole (e, e) excitation spectrum as shown in Fig. 2(b) corresponds very closely with that observed in the 8a~-t photoelectron spectrum [13]. This similarity with ionization would be expected for a Rydberg transition since Rydberg orbitals are generally diffuse and atomic-like with the Rydberg electron viewing the positively charged ion core as a point charge. These observations therefore strongly support the assignment of the first vibronic progression centered at 9.52 eV solely to the 4s ~ 8a~-~ Rydberg transition [4] with no underlying contribution from a 9at(it ~ ) ~ 8a~- t virtual valence transition. Sodhi and

Brion [8], whose data are not sufficiently resolved to show any vibronic structures, had earlier suggested that the 9al virtual valence transition might also be contributing to the band at 9.52 eV. Table 4 gives the absolute oscillator strength values for each of the 23 vibronic peaks shown in Fig. 2(b). The oscillator strength for the entire band from 8.68 to 10.26 eV, 0.2542 (Table 3), is in good agreement with the value of 0.28 reported earlier [4]. This good agreement and also that for the band from 7.39 to 8.68 eV (see above) further supports the accuracy of the presently determined absolute oscillator strength scale and the resulting value of the static dipole polarizability (see Section 3.2.1). Based on a quantum defect of 1.78 determined for the 4s ~ 8a~-t transition at 9.52 eV, the Rydberg formula was used in Ref. [8] to determine the position of the 5s Rydberg. This is shown on the 8a~-t manifold in Fig. 3(a), and it can be seen that, as expected, the occurrence of a second vibronic envelope (feature 4) matches very well with the predicted position of the 5s ~ 8a~- ~ transition. Therefore, while the broad band at 11.08 eV also possibly has contributions from 7 e ( c r * ) ~ 6 e - l and la~-I, the vibrational structures must presumably arise from the transition to the 5s Rydberg state. An oscillator strength value of 0.2842 has been presently obtained for this band (integrated over the 10.26-12.16 eV region). The oscillator strengths for five more energy regions shown by the dotted vertical lines in Fig. 3, are listed in Table 3. This represents an extension of the very limited absolute data ( 5 - 1 0 eV) [4] for the valence-shell photoabsorption of PF 3 that is currently available in the literature. It should be noted that while the spectral features in Fig. 3 eorrespond well to the 7 e ( t r * ) and ns Rydberg assignments [8], transitions to the n p and nd orbitals are also formally-allowed for PF 3 and may be expected to contribute [70]. Recent theoretical investigations using continuum MS-X,a calculations [19,71] suggested the presence of two shape resonances in each of the a~ and e continuum channels in the valence-shell ionization continuum. The partial photoionization cross sections were calculated over the electron kinetic energy range 0 - 3 0 eV for each of the three outermost valence orbitals, 8a I, 6e, and l a 2 [19]. The lower energy resonance (9al(cr * ) ~ 8a~-~), which was found to be

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418 P 2p pre-edge _

8 ¸

3.2.3. Phosphorus 2p and 2s inner-shell photoabsorption spectra

. -8

"7

> 6 0 ~'¢... ¢:~

N

4

4

~9

O

em

D eO O if) O "6 en

I

II 9a1"4s

~ 2

I 3

t 11 I I 4 5678

7e" 7e"

2 (a)

o

1

134

co

I

a

136

138

~ 564<:1 t I 2p~t; 2 I 9

140

I 18

0 142

(b) f , ~ . . . . . . . .

8

6

6

I

......~......................... 2p~, ,12 ~I! d~ub~e~ lOa"1 ex~ati

2 0 140

145

.[

o

4

F- I I

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

0 155

160

180

10a;

7e" 4p 261

2 ~ ..............I . . I

0-/(c)

.........

.14 ".16 190

165

170

P 2s pre-edge & post-edge [_ 6 I

200

t'~ v cO .¢_, (3

Or)

Of) Or)

2

(3 tO

2

i

3d (al+e+e)

01 .......... o oo.................................... 4 /

4

P 2p post-Nge

150

~E =, eV ~,~ I

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

I

0 ¢n o~

o 0 eIX.

1-4

l [-2

,.......... i0........................................ . . ,t 0 210

220

230

240

407

250

Photon Energy (eV) Fig. 4. (a) The absolute photoabsorption differential oscillator strengths (0.1 eV FWHM) measured in the P 2p discrete region (solid circles) compared with the optical spectrum (line) previously obtained using synchrotron radiation [5]. The integrated oscillator strengths for the six energy regions separated by dotted vertical lines are given in Table 5. Panels (b) and (c) show the P 2p post-edge and P 2s regions respectively: The solid circles represent the high resolution (0.1 eV FWHM) dipole (e, e) data, while the open circles show the lower resolution (1 eV FWHM), long-range data. The dotted lines at the base of the spectra in panels (a)-(c) represent the contributions to the absolute oscillator strengths for inner-shell photoabsorption from the underlying valence-shell estimated from the fitting and extrapolation procedures (see text for details).

characterized by sp type wavefunctions from the P atom and p type wavefunctions from the F ligand, was predicted [19,71] to be located ~ 3 eV above the ionization threshold. In contrast, the remaining a~ and e shape resonances were calculated to be dominated by d type continuum functions. This suggests that similar to the P 2p spectrum (see Section 3.2.3 below), the broad continuum peak observed at = 35 eV in Fig. l(b) can be attributed to valence-shell excitations to the 10a~ virtual valence orbital.

In addition to the low resolution data shown in Fig. l(a) and (c), the absolute differential oscillator strengths in the discrete P 2p inner-shell region of PF 3 have been further investigated in the present work at higher resolution (0.1 eV FWHM), and the data are shown as solid circles in Fig. 4(a) together with the optical work (line) of Ishiguro et al., [5] which matches closely with the dipole (e, e) data in both shape and magnitude. The bands situated from 134 to 137 eV in Fig. 3(a) have been previously assigned [5,9] to the virtual valence transitions, 7e((r * ) ~ 2p - t . The phosphorus 2p orbitals of PF 3 transform as a I and 2e under C3v symmetry. The direct product of the initial 2p(e) and final 7e(cr*) orbitals indicates the existence of two dipole-allowed excited states, tA~ and IE, which are spin-orbit split into four transitions (see Fig. 4(a)) [5,9]. Only three peaks are apparent in this energy region. The assignments in Refs. [5,9] indicated that the center band represents the sum of 7e((r*): rE*-- 2p3/~(e): IA i and 7e((r * ): IAj *2p~-/Iz(e): IA l [5,9]. It can be seen that the main peak energies of features 1-3 (134.93, 135.70, and 136.48 eV) also reflect the 2p3/2 and 2p~/2 spin-orbit splitting (0.86 eV [6]) for transitions to both the IE and ~A~ excited states. Recently, a high order ab initio calculation and synchrotron radiation study [7] predicts that feature 3 is in fact an LS coupled 7e((r*): IA I ~2p3/~(e): IA I state which has no spin-orbit partner. The series of absorption bands from 137 eV up to the P 2p ionization thresholds arise from transitions of the 2p core photoelectrons to Rydberg orbitais [9]. Features 5 and 7 at 138.60 and 139.40 eV, respec-I tively, have been assigned to 4s ~ 2p3/2,~/2, [9] based on their splitting and also because of the sharpness of their profiles which is typical of Rydberg transitions [70]. The positions of the 4s Rydbergs correspond to larger term values than those expected for an ns quantum defect of 1.75 eV determined from the valence shell (see Section 3.2.2 above). This can be attributed to Rydberg-valence mixing with the underlying virtual valence transition to the 9a~((r* ) orbital (see below). Ishiguro et al. [5] later interpreted these two peaks as being due to transitions to the 4p Rydberg orbitals based on an

408

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418

Table 5 Absolute photoabsorption oscillator strengths for regions of the P 2p inner-shell spectra of PF 3 Integrated

Oscillator strength (10-2)

energy region a this work

Ref. [5]

(eV)

133.60-136.10 136.10-137.07 137.07-139.19 139.19-139.64 139.64-140.06 140.06-141.01

valence + P 2p shell P 2p shells

valence + P 2p shells

13.29 5.81 11.69 2.66 2.20 4.79

13.52 5.84 12.20 3.22 2.27 4.88

2.53 1.72 3.09 0.86 0.54 1.12

a See Fig. 4 for the respective spectral regions.

average term value of 2.5 eV and a quantum defect of 1.67. However, with the P 2p3/2 and P 2pl/2 VIP's occurring at 141.77 and 142.67 eV [9], the term value (T = VIP - E) for each of the two sharp structures clearly cannot be 2.5 eV. Since VIP's of 141.04 and 141.97 eV were used in assigning the optical spectrum [5], it appears that the authors had misinterpreted the VIP's reported in Ref. [26]. Minimal basis set ab initio calculations performed by Ishiguro et al. [5] predicted that np Rydbergs occur in the photoabsorption spectrum of PF 3 because of the C3v symmetry of the molecule. However, it has been found in previous studies that atomic-like selection rules are a useful guide to relative intensities in the interpretation of molecular core excitation spectra to Rydberg states, since the initial and final states are both essentially atomic in character. On this basis, 4p ~ 2p-I transitions would not be expected to make a larger contribution than 4s *--2p -~ , and thus the latter assignments as in Ref. [9] are preferred for peaks 5 and 7 on Fig. 4(a). The shapes of the two bands in the 137-139 eV region suggest that the shoulders on the lower energy sides of the 4s Rydberg transitions (features 4 and 6) arise from the --I underlying 9 a j ( o -* ) ~ 2p3/2.1/2, transitions [9]. However, both the minimal basis set ab initio [5] and the MS-Xc~ continuum [19] calculations in the literature predicted the 9a~ virtual valence orbital to be situated near the continuum, but this is not experimentally observed. Nevertheless, it is possible that the a~ anti-bonding orbital theoretically determined [5,19] is in fact the 10a~.

The tentative assignments of the higher Rydberg features converging to the ionization continua in Fig. 4(a) have been guided by those in Ref. [9]. Since the calculated term values and quantum defects used in Ref. [9] were not given, the positions of the 5s and 4d Rydbergs indicated on the manifolds have been determined in the present work using the Rydberg formula with estimated quantum defects of 1.75 and 0.2, respectively. The value of 1.75 in the former case has been taken from the 4s ~ 8a;- ~ valence-shell transition, since there is no evidence of Rydberg-valence mixing in the valence spectrum in contrast to the P 2p region. The nd quantum defect of 0.2 is as suggested by Robin [70] for transitions involving third row elements. Table 5 compares the absolute oscillator strengths for the P 2p discrete photoabsorption of PF 3 (plus the underlying valence continuum) for the six regions separated by the vertical lines in Fig. 4(a) determined from the present measurements, with those of the optical work of Ishiguro et al. [5]. It can be seen that there is generally good agreement between the two sets of data in all regions within experimental error. Also given in Table 5 are the partial oscillator strengths for the P 2p shell component of the spectrum obtained by subtracting the valence shell contributions estimated from the fitting and extrapolation procedures in Section 3.2.1. Two strong shape resonances are observed in the P 2p ionization continuum at 146 and 159 eV (Fig. 4(b)). Recent continuum MS-Xc~ calculations [19] have predicted that the first resonance structure (feature 12 at 146 eV) results from the a t channel associated with the l = 1 component of the fluorine ligands. This is consistent with earlier ab initio calculations which suggested an assignment of 10at(tr*) 2p(al) -I [5]. However, the small shoulder (feature 11 at 144.5 eV) situated at the leading edge of this resonance has not been modeled by any of the theoretical methods [5,19,48]. This feature, which is also evident in the optical absorption [5] and electron-energy-loss [9] spectra, has been ascribed [9] to a double excitation process (i.e. simultaneous excitation of a 2p and a valence electron). In contrast, the broader resonance located = 16 eV above the P 2p edge (feature 13 at 159 eV) is attributed to excitation from the 2p core to the phosphorus continuum 3d orbitals according to the

J.W. Au et al. / Chemical Physics 215 (1997) 397-418

MS-X(r calculations [19]. The ordering of the atomic-like 3d *-- 2p -1 transitions, 3dz:(1 la l) < 3dxz ' yz(8e) < 3dx2_y2 xy(9e) [48], is reflected by the relative energy positions of the a~ and e channel partial photoionization cross sections theoretically determined in Ref. [19]. The two pre-edge structures in the P 2s spectrum (Fig. 4(c)) have been previously observed in both the electron-energy-loss [9] and optical absorption [5] spectra. They were assigned to the 7e(g * ) * - - 2 s (192.0 eV) [5,9] and 4p *-- 2s-i (197.5 eV) [9] excitations. Above the P 2s threshold, the very broad resonance at = 219 eV arises from the a~ continuum channel and is analogous to the 10al(cr*)*--1 2P3/2.1/2 transition in the P 2p spectrum [19]; therefore, it has been tentatively assigned to 10a~(cr * ) *-2s-~ in the present work.

3.3. Molecular and dissociative photoionization 3.3.1. Valence-shell region Valence-shell TOF mass spectra of phosphorus trifluoride have been recorded at low resolution (1 eV FWHM) as a function of energy loss (photon energy) from the first ionization threshold up to the onset of excitation from the P 2p shell (130 eV). The positive ions formed from the molecular and dissociative photoionization of PF 3 are PF~+, PE~ +, and F + where x equals 0 to 3. Fig. 5 shows a typical TOF spectrum obtained at 70 eV. The broadness of the peaks due to the smaller, singly charged ions,

1O0

I ~

!

PF2

PF~+

80 ~

__

h~ = 70 eV

pF~+

60

!

1x25

20

PF3 p

!,

•

0

-

2

3 4 Time of Flight (#sec)

'

'

5

Fig. 5. Time-of-flight mass spectrum of PF 3 recorded at an equivalent photon energy of 70 eV.

409

PF +, P+, and F +, indicates that they are formed with appreciable kinetic energies of fragmentation either from direct dissociative photoionization or from Coulomb explosion of multiply charged ions. In particular, the P+ peak is quite broad when first formed at 26 eV, while the F ÷ peak further broadens above 55 eV. In the inner-shell region these ion peaks become increasingly broad with increase in photon energy. In the dipole (e, e + ion) experiment, the relative photoionization efficiency is defined to be the ratio of the total coincidence ion signal to the number of forward scattered electrons as a function of energy loss (i.e. equivalent photon energy). These values for PF 3 at 1 eV FWHM become effectively constant within experimental uncertainty above 17.5 eV. Making the reasonable assumption that the photoionization efficiency (r h) is unity at high energies, the r/i values at 17.5 eV and above have been normalized to one. From the photoionization efficiencies shown in the inset of Fig. l(a) (and given numerically in Table 1), it can be seen that r/i rises to a local maximum at = 13 eV before increasing quickly to unity at 17.5 eV. The leveling off of the r/i at --- 13 eV corresponds to a large peak in the low resolution total photoabsorption (Fig. l(b)) and indicates that fluorescence a n d / o r dissociation to neutral products are competing with autoionization back down to the 8a~-~ continuum from the super excited states converging on the higher valence VIP's in this region (see Fig. 3 for details of the excited states at higher resolution). The photoion branching ratios presented in Fig. 6 were determined by integrating the mass peaks in the background-subtracted TOF spectra at the photon energies shown. The data were corrected for the mass (re~e) sensitivity of the microchannel plate ion detector using the response function reported in Ref. [44]. As the molecular ion decreases in abundance above 16 eV, PF2+ becomes the dominant cation throughout the valence-shell photoionization of phosphorus trifluoride. Above 36 eV small amounts of stable, doubly charged ions, pF,2+ (x = 0-3), are produced. It is clear that major changes occur in the ionic photoffagmentation of PF 3 in going from the valence to the inner-shell region. The data shown in Fig. 6 for the phosphorus 2p and 2s inner-shell regions above 130 eV are discussed in detail in

410

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418

Section 3.3.3 below. The numerical values of the branching ratios can be generated by dividing the ion oscillator strengths in Table 1 by the product of the photoabsorption oscillator strength and the photoionization efficiency at each photon energy. Absolute partial photoionization differential oscillator strengths (PPOS) for the production of molecular and fragment ions from PF 3 were obtained from the triple product of the absolute total photoabsorption differential oscillator strength (see Section 3.2.1), the photoionization efficiency, and the photoion branching ratio for each ion, as a function of photon energy. The PPOS for molecular and dissociative photoionization of PF 3 are shown in Fig. 7 and are listed numerically in Table 1. The hatched lines represent the vertical ionization potentials (VIP's) of

'o

o

t--

lO,

.A

~® 50

'g

Total photoabsorption

_ ~,p.~'~. P 2n

~""---

x_x7 2S

-

O

60

~

ion states

P 2P3~.1/=

~

PF3

P 2s

xm .~

~'Vertical IP's

'

' '

Z~'

~: -CI

0.0 0.2

E o •=

0.0

N

2

"E: ._o

0.4

o "13_

0.0 1,5

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

P1~22~"

~,. . . . . . 1 - . . . . .....,

"

.

~"""

''-

2

. PF

. . .0..-. o o . • o

"

....

......

;

_

_

L .................

H,

P

" " . . . . . . "-- ~ - . - - - , - _ _ _

LIILI~"

0

o.o 0.2. o.o

.

pF +

°-~--.. . . . . . . . . . .

" ° ° ° ° °

. pF~+~

13_ ~& Valence

o l 4,

/ 16'

-

0

: i o

i'Z___ .....

I J l %,..

ul

~

:

I I~I~;~ 5 0 .Ix l s o ~

9

=--

A~'\ Valence Photoabsorption

50,

P

.

F~

p~

I1~

! ll~. ..

.

. •

20

i

.

i

i

.'-... . . . . . . . .

40

.

.

.

i

~._~ ..... ~

.

-

i

.......

.

,m•

60

80

100

120

140

Photon Energy (eV)

V•°

f~"--.._..

60-

° , , ,

,

,

.

°

PF]

,

•

o c-

~Oo•

0

to 0 2

pF ÷ p " - ~ . . . . . . . . . . . PF3

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

e-

PF~ +

m

p÷

,,..~ L,

.

,

•

o

•

,

•

•~',*

°

°

°

°

°

•

•

•

.

° ~

•

. ,,,,,"

°

•,~.m

• o • ° • •

~

'.%.,.

e.-

o ~ , , • • ° • o •

o o eQ.

"~.~...

!

2+

..o

0

~ 2 + I"F

~ . _.,,~°••oo•

• •

~ • • °

30 ¸ F + •

0 10

,,,.,.,,•..••

Fig. 7. Absolute partial photoionization differential oscillator strengths for the valence-shell molecular and dissociative photoionization of PF 3 (11 -130 eV). The vertical arrows represent the calculated thermodynamic potentials [72} of the singly-charged cations (see Table 6). The positions of the seven outer-valence vertical ionization potentials (VIP's) measured by PES [13] and the average energies for the two inner-valence ionization processes determined using MS-Xa calculations [19] are denoted by the hatched lines in the PF~- panel. The lower energy regions of the partial photoionization data for PF~- and P+ have also been plotted at 150 and five times the absolute intensities, respectively, in order to show more clearly the onsets of the multiple thresholds. The dotted curves represent valence-shell extrapolations used to estimate the differential oscillator strengths in the inner-shell region (see Fig. 13 below). The absolute total photoabsorption spectrum from Fig. 1 is also shown in the top panel.

°

p2+

,m.'' ,

40

.=m

oo°°•

•

80

•

•

120

160

200

P h o t o n Energy (eV)

Fig. 6. Branching ratios for the molecular and dissociative photoionization of PF 3. The positions of the seven outer-valence vertical ionization potentials (VIP's) measured by PES [13], the average energies for the two inner-valence ionization processes determined using MS-Xa calculations [19], and the P 2p and 2s VIP's derived from XPS [9] are denoted by the hatched lines in the PF~- panel. The absolute total photoabsorption spectrum from Fig. 1 is also shown in the top panel.

the seven outer [13] and two inner [19] valence orbitals obtained as discussed in Section 3.1. The appearance potentials (AP's) of the cations measured in the present work are shown in Table 6 in comparison with values from previously published work. The present results for the singly charged ions are reasonably consistent with previously reported values obtained from photoionization mass spectrometry (PIMS) [20] and electron impact [22,23,25] except in the case of the electron impact value [22]

J.W. Au et al./ Chemical Physics 215 (1997) 397-418

Table 6 Calculated and measured appearance potentials for the production of positive ions from PF3 Process Appearancepotential (eV) calculated a experimental [20] [22] [23] [25] 11.4 10.8 14.7 13.1 16.9 19.1 20.8 14.6 18.5 20.7 22.3 24.5 26.2 23.2 28.5 31.5 33.1

11.5 11.0 15.0

11.43 11.65 11.6 11.38 14.55 15.5 13.5 14.27

21.0

20.5

26.0 28.5

32.3

33.0 35.0 39.0 39.0 49.0

correspond to extensive fragmentation of the molecule to the respective ions and dominantly atomic neutral fragments. 3.3.2. Dipole-induced breakdown o f PFs

this work b PIMS electron impact PF~ PF~ +FPFf +F PF+ +2FPF+ + F + F PF + +F 2 PF + +2F P+ +3FP++F+2FP++F 2+FP÷+2F+FP+ + F + F 2 P÷ + 3F F ÷ + PF2 F + +PF+F F+ + P + F 2 F÷ + P + 2 F PF~÷ PF~ ÷ PF 2+ p2+

411

36.0

a Calculated using thermochemical data from Ref. [72] assuming zero kinetic energy of fragmentation. b +leV.

for P+. No appearance potentials for the doubly charged ions are presently available in the literature. Table 6 also contains thermodynamic appearance potentials calculated for all possible processes leading to the production of singly charged ions from PF 3 using heats of formation of neutral species and singly charged ions [72], and assuming zero kinetic energy of fragmentation. These values for the singly charged molecular and fragment ions are denoted by vertical arrows in Fig. 7. Note that while negative ions cannot be detected from the dipole (e, e + ion) experiment, the appearance of P F f at 11 eV (see Table 6 and branching ratios in Fig. 6) corresponds to the resonant ion-pair formation process, P F / + F [22]. The main onset for P F f production by dissociative photoionization is at 15.0 eV, which is in good agreement with the thermodynamic prediction. The observed onsets for PF ÷, P+, and F ÷ formation

Fragmentation ratios for dissociative photoionization from each electronic state of the molecular ion (PF~) should be constant when the photon energy exceeds the upper limit of the Franck-Condon region, except in local regions where autoionization effects are significant or where multiple photoionization is appreciable [42]. Therefore, the absolute PPOS for the production of each singly charged ion from PF 3 can be expected to be a linear combination of PPOS of the contributing electronic ion states. This is the basis that permits the deduction of the dipoleinduced breakdown pathways of phosphorus trifluoride by careful consideration of the experimental appearance potentials (Table 6), valence-shell branching ratios (Fig. 6), and absolute partial differential oscillator strengths for molecular and dissociative photoionization (Fig. 7) measured in the present work, together with VIP's [14], Franck-Condon widths of the electronic ion states [ 14], photoelectron branching ratios [6,14], and He I PEPICO data [15] available in the literature. The photoelectron branching ratios for production of the seven outer valence states for P F f reported by Green et al. [14] and V~iyrynen et al. [6] in the energy ranges 3 0 - 9 5 eV and 35-141 eV, respectively, have been used in conjunction with the presently obtained absolute total photoionization differential oscillator strengths (total photoabsorption differential oscillator strengths × photoionization efficiencies, as given in Table 1) to derive electronic ion state PPOS for PF 3 as follows: The contribution from the two inner valence ion states (3e-~ and 5a~-a), which were not reported in the photoelectron studies [6,14], have been approximated by using the sum of the presently measured P+ and F + PPOS (see discussion below), and this sum was then subtracted from the total photoionization differential oscillator strength. This estimation using the sum of the P+ and F + PPOS is based on the appearance potentials of these ions (Table 6) which are above the upper limit of the Franck-Condon region for production of the 6a~-~ ion state [14] (the last outer valence VIP). Since doubly charged ions are not in

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418

412

general expected to arise from the decomposition of singly charged ions (except in the case of VVV Auger processes [73]), the (relatively low intensity) PPOS for production of the PF~2+ ( x = 0 - 3 ) ions were also subtracted from the total photoionization differential oscillator strength. The resulting modified "total" photoionization oscillator strengths were then multiplied by the photoelectron branching ratios [6,14] at each photon energy to yield PPOS for the production of each of the seven outer valence electronic states of PF3+. These PPOS are shown as open circles [14] and stars [6] on Fig. 8; the electronic state PPOS as given by ion sums (solid circles) have been derived from the presently obtained PPOS data using the dipole-induced breakdown scheme discussed below.

(a) :

'> 1o

PF;, This wo~ o *

• ~,~

- -

.c:

5

¢-.

m

• •

+

..~ I " PF~,This work ~+7a1+O.88(4e),

PES[14]

~ + T a l + O . 8 8 ( 4 e ) ,

PES [6]

40

\

o -.~

"%*,0

20

"°*G' oa' •

t"-

•

=-

121 r"

o

~

4

O

10

•

5

8a 1 , PES [6] 8a I , MS Xct [19] P F ; , This work

*

% %~

•"

~ *

0

0.12(4e)+6a~, PES [141

•

,

,

e

~

lo.O.ol~oe

o

.

0 (d)

t-

'>

OO0ooooo

c•

* 8aI , PESt14]

•

0

N o ~6

(a)."

8 a 1 , M S Xo. [19]

\

o

(b) • '

t~ ~

8a r , PES[14] 8a I , PES {6]

~

5

~

o

...

= • =¢1¢10 - n-

20

40

- F ÷ . P ÷ (30+5a 1), This work ~ F ~ [55]

L '-

= = o a=uouo===

60

80

•

n

.

1 O0

Photon Energy (eV)

~ 40 _~ 20

~

F

•

%o

;

;

0

~°9~'V~

(c)

?

eo

o000009

~p .

. 4e, P_ES[,1_.4}

.,'% ." k

I * 40, PES [6] +

O.21(PF*)

.~

0

~ ._N

3 ~

•

(d~

•

.'.. /,,,.

I

This work

•

eogoeoeo

.

6a 1 , _PE$ [L4]

~.~;1,: p ~ i61" ":0%~

,

o .

•

i--

5

°

_- o ~ ( ~ 3 , Th+. . . . k •

O

~°0°eoeo0

~ '-"

o

.

" ~+P_+ [3_e+f.a~) ,-T_his ~ork [ = 3 x F2s, Ph Ion [55]

**. ,•

.

"*''

~-

•

•

5

eo

~-

Fig. 9, Differential oscillator strength sums determined from the proposed dipole-induced breakdown scheme for PF 3 below 130 eV. The presently measured absolute partial photoionization differential oscillator strength sums for molecular and dissociative photoionization are compared with electronic ion state partial oscillator strength sums derived from PES branching ratio measurements [6,14] and MS-Xet continuum calculations [19].

Thiswork

•*'**

~

~

, k

;

O

~

•

.

.

.

•

•

.

I

0 20

40

60

80

1 O0

Photon E n e r g y (eV) Fig. 8. Absolute electronic ion state partial differential oscillator strengths (open circles and stars) derived from photoelectron branching ratios (30-95 eV) from Refs. [6,14] and the presently measured total pbotoabsorption and photoionization efficiency data (see Table 1). The solid circles represent estimates of electronic ion states PPOS derived from the absolute PPOS of the molecular and fragment ions obtained in the present work (see Fig. 7) as discussed in Section 3.3.2.

The appearance potential of the PF~- molecular ion (11.5 eV) is at the adiabatic ionization potential of the ground electronic state, 8a~-~ (VIP = 12.3 eV) [14], and the He I PEPICO data of Reynolds et al. [15] shows that PF~- is formed essentially only from the 8a~-s state. The PPOS of PF~- are compared in Fig. 9(a) with the PPOS of the 8a~-~ state derived from the photoelectron branching ratios [6,14], and also with a calculation of the 8a~-t ionization performed using the M S - X a method [16]. It can be seen that the PPOS of PF~- are in particularly good agreement with the 8a~-~ electronic state PPOS derived from the PES work of Green et al. [14] over the entire energy range of the latter measurements. The calculation [19], which shows two resonances centered at 14.5 and 22.0 eV, does not reproduce very well the PES [14] or the PF~- experimental results on Fig. 8(a).

J.W. Au et al. / Chemical Physics 215 (1997) 397-418

Two appearance potentials were measured in the present work for PF~- (see Fig. 7). The first (low) onset at 11.0 eV corresponds to the thermodynamic AP calculated in Table 6 for the ion pair formation, PF 3 --* PF~- + F - . The second PF~- onset at 15.0 eV is in the region of the ionization energies of the (6e-~ + la 2 l) states [14] and corresponds to dissociative photoionization to P F ~ + F (Table 6). The PPOS curve for PF~- (Fig. 7) continues to increase up to 21 eV, and all other ions appear at or above 21 eV. This, together with consideration of VIP's and Franck-Condon widths, suggests that further contributions must occur to PF~- from the 5 e - ~, 7a~- ~, and 4 e - i ion states. It can be seen from Fig. 9(b) that the electronic ion state PPOS sum (6e-~ + la~ ~ + 5 e +7a~ -I + 0 . 8 8 ( 4 e - i ) ) using both PES data sets [6,14] is in excellent agreement with the shape of the PF~- partial oscillator strength curve. The PEPICO measurements [15] indicate that these five ion states exclusively dissociate to PF~- at the He I photon energy used. The AP of the PF ÷ ion (21.0 eV) lies at the high energy limit of the 4 e - ~ Franck-Condon width [ 14] and below the adiabatic ionization potential of the 6a~-~ state. This suggests a small contribution ( = 12%) from the 4e -~ and a dominant contribution from the 6a~- t ion states. PF ÷ could not be observed to dissociate from the 4e-~ in the PEPICO experiment [15], and this is to be expected since the upper photon energy limit of the He I PEPICO data (21 eV) [15] is situated right at the PF ÷ onset of 21 eV. Fig. 9(c) shows that the agreement between the electronic ion state PPOS sum 0 . 1 2 ( 4 e ) + 6al derived from the PES data of Green et al. [14] is generally in very good agreement with the PF ÷ PPOS. The P+ and F ÷ ions have A P ' s of 26 and 33 eV, respectively, which are well above the upper limit Franck-Condon width of the 6a~-~ state; therefore these two ions must arise from inner valence (3eand 5a~-~) photoionization [19] a n d / o r dissociative multiple photoionization processes. Thus, the ( P + + F ÷) PPOS sum should provide a reasonable estimate of the summed ( 3 e - i + 5a~-l) PPOS if the contributions to these two ions from dissociative multiple photoionization processes are relatively small. A crude estimate of the contribution from the inner-valence states may also be obtained from the calculated

413

Dipole Induced Breakdown Pathways for PF3 ~

Ion States VIP (eV)

Mctecule in Ground State

+ 5at

.... I

PrearrnC~rtn|y~//'O0°/° ett~.~/'~tO/O"10001` ~OO°l

Products (eV)

AP

P* + F+ 26.0 33.0

I

Fig. 10. The dipole-induced breakdown pathways following valence-shell photoionization of PF3 below 130 eV. The vertical

ionization potentials (VIP's) for the seven outer-valenceelectronic ion states have been taken from PES [14].

atomic F 2 s - l photoionization differential oscillator strengths, since the 3e and 5a~ orbitals have predominantly F 2s character [14]. As shown in Fig. 9(d), three times the oscillator strengths of F 2s[55] (from the three fluorine atoms in PF 3) adds up to significantly less than the ( P + + F ÷) PPOS sum. It can also be noted that 80% of (3 X F 2s) is similar in shape and magnitude to the F ÷ PPOS. The differences are likely due to the molecular character in the inner-valence orbitals a n d / o r contributions from dissociative multiple photoionization processes. The peak observed at 38 eV in the P+ PPOS coincides with the shape resonance seen in the valence-shell photoabsorption (Fig. 1). This molecular phenomenon is not present in the (3 x F 2s) sum as expected. The dipole-induced breakdown scheme for production of singly charged cations from PF 3 in the valence shell region according to the above deductions is summarized in Fig. 10. Fig. 8 illustrates the absolute electronic state PPOS derived from PES measurements plus the total differential photoabsorption oscillator strengths from the present work, [6,14] as described above, together with those estimated from sums of their corresponding dissociative ion products. It should be noted that the ion sums provide estimates of the electronic state PPOS at energies below the lower limit of the PES measurements [6,14] as well as at higher energies. Turning to a consideration of the doubly charged ions, it can be seen that the shapes of the PPOS for

414

J.W. Au et al. / Chemical Physics 215 (1997) 397-418

production of PF32+ and PF~ + (Fig. 7) are quite similar, although their onsets are different. Similarly, the PF 2+ and p2+ ions have different appearance potentials but similar partial differential oscillator strength shapes. Therefore, it is possible that these four dications arise at least in part from different VVV Auger processes involving groups of energy poles arising from many-body effects in the two inner-valence (3e -1 and 5a~-I) ionization processes. More complete information on the dipole-induced breakdown of the singly and doubly charged cations must await measurements of adiabatic and vertical double ionization potentials as well as further photoelectron-photoion coincidence (PEPICO) and phot o i o n - p h o t o i o n coincidence (PIPICO) studies. Many-body G r e e n ' s function calculations as well as further photoelectron and electron momentum spectroscopy experiments are required to elucidate the nature of the binding energy spectrum of PF 3 in the inner-valence region and the role of many body (electron correlation) effects. 3.3.3. Phosphorus 2p and 2s inner-shell regions In order to investigate the changes in the dipoleinduced breakdown of PF 3 in going from valence-

shell photoionization to phosphorus 2p and 2s excitation and ionization, T O F mass spectra have been recorded at a series of energies from 130 to 193 eV at a resolution of 1 eV FWHM. Typical spectra are shown in Fig. 11 at (a) 130 eV (pre-edge, valenceshell region) and in the inner-shell region at photon energies of (b) 136.5 eV corresponding to the virtual valence transitions ( 7 e ( t r * ) ~ 2 p ) , (c) 138.5 eV (Rydbergs ~ 2p), (d) 147 and (f) 157 eV (the continuum shape resonances (9al(cr * ) ~ 2p and 3d ~ 2p, respectively), (e) 152 eV (the trough between the two shape resonances), (g) 169.5 eV (the 2p ionization continuum), and (h) 193 eV (7e(tr * ) ~ 2s excitation). Similar to the valence-shell case at 70 eV (see Fig. 5), five singly charged ions ( P F f , P F f , PF +, P+, and F +) and four doubly charged ions (PF32+, PF22÷, PF 2+, and p2+) are detected at all energies, although the relative intensities are quite different in the P 2p, 2s inner-shell regions. Recent PIPICO measurements for PF 3 [33] in the P 2p region have reported time-correlated pairs of singly and doubly charged fragment species resulting from Coulomb explosion of dissociative doubly and triply charged ions. However, no stable triply charged ions were observed in the T O F mass spectra [33] in

Table 7 Absolute differentialoscillatorstrengths for the total photoabsorption and the molecularand dissociativephotoionizationa in the valenceand the P 2p, 2s inner shell regions of PF3 from 130 to 192.5 eV Photon Partial oscillatorstrength ( 1 0 - 2 e V - ~) b energy (eV)

Photoabsorption

PF~-

PFf

PF"

PF3 +

PF~+

P+

PF2+

F+

p2 +

130.0 135.0 136.5 138.0 138.5 139.5 141.0 144.0 145.0 146.0 147.0 148.0 152.0 157.0 162.0 175.0 192.5

4.73 5.28 5.63 5.57 5.83 5.54 5.21 6.26 7.03 7.51 7.48 7.24 6.36 7.38 7.19 5.85 5.71

0.19 0.18 0.27 0.20 0.16 0.18 0.19 0.13 0.13 0.11 0.13 0.15 0.09 0.11 0.06 0.07 0.08

2.33 1.93 2.26 1.77 1.88 1.81 1.74 1.48 1.41 1.42 1.54 1.57 1.35 1.23 1.15 1.00 0.76

0.29 0.48 0.47 0.48 0.47 0.39 0.28 0.33 0.40 0.40 0.37 0.33 0.31 0.34 0.32 0.25 0.21

0.03 0.03 0.05 0.02 0.04 0.03 0.05 0.08 0.11 0.i0 0.11 0.09 0.10 0.13 0.11 0.07 0.05

0.05 0.06 0.06 0.09 0.08 0.10 0.06 0.07 0.10 0.08 0.12 0.09 0. I0 0.08 0.11 0.09 0.08

0.63 0.97 0.89 1.15 1.15 1.11 1.00 0.97 1.11 1.20 1.22 1.16 1.05 1.20 1.14 0.91 0.91

0.17 0.20 0.22 0.21 0.24 0.21 0.20 0.57 0.71 0.75 0.76 0.68 0.64 0.72 0.74 0.54 0.45

0.90 1.22 1.23 1.41 1.58 1.48 1.49 2. | 3 2.51 2.85 2.67 2.61 2.20 2.85 2.90 2.35 2.54

0.15 0,21 0.19 0.23 0.23 0.22 0.20 0.51 0.56 0.61 0.57 0.56 0.53 0.70 0.68 0.56 0.63

a Assuminga photoionizationefficiencyof unity. Note that these values include the contributionfrom the underlyingvalence shell. b O" (Mb)= 109.75 d f /d E (eV-I).

J.W. Au et a l . / Chemical Physics 215 (1997)397-418

agreement with the present work. It can be seen from Fig. 11 that the most significant changes in the T O F mass spectra in going from the valence to the P 2p,2s region are the large increases in the relative abundances of F +, p2+, p F ÷, and PF~ + at energies above 138.5 eV in the P 2p photoionization continuum region. In the P 2p virtual valence (136.5 eV, Fig. 1 l(b)) and Rydberg (138.5 eV, Fig. I l(c)) excitation regions, the most noticeable changes from the valence-shell spectra are in the relative abundances of PF +, P+, and F ÷. At higher photon energies in the P 2p continuum, F + becomes the most abundant ion. The P÷ and F ÷ peaks are very broad, particularly in the P 2p ionization continuum, indicating considerable kinetic energy of fragmentation. In assessing the

8

>

(1)

o vr-dE:

P2 p ~ .~'

415 .............. "'""

P 2S E """--,..o.°..,...°...,l°.,..,.......

4 0 0.3

Total Photoabsorption ..................................................................................

.

......

,

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

o.0 t--o 2 ....., ~÷~,

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

PF;

PF 2

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

m ............................ •

o.O

O ~

0.0 0.1 o.o

c-

0.1

......

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

:__:

.

÷

0.0

C3 o

....

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

0

._N 0.6 ¢-

.o_ 0.o o h~ = 130 eV, 100

(a)

Valence

h~ : 152 ev, Trough

(e)

100

0 tn

2

--

0

"2

0.6

13_

0.0

..~-= 130

0 100 -

ht) = 136.5 eV, 7e* < - - 2p "t

(b)

+i

PF;

/PF~ +

0

.~

loo

rr

'11)= 138.5 eV, 1 Rydbergs < - - 2p-

(C)

=LL,.L,_

100

hl) : 147 eV, Continuum resonance

(d)

h~ = 169.5 eV, P 2p4 continuum

F+I

2

3

4

5

(g) 100

0 h~ = 193 eV, 7e* < - - 2s "1

(h)

I O0

P~*p~"

3

4

150

160

170

180

190

F+

200

P h o t o n E n e r g y (eV)

(f) 100

pF2÷ pF~2"

c-

h~) = 157 eV, Continuum resonance

140

-----"

PF;

5

Time of Flight (~sec)

Fig. 11. Time-of-flightmass spectra recorded at selected energies in the P 2p and P 2s inner-shell region: (a) the valence-shell continuum at 130 eV, (b) 136.5 eV correspondingto the 7e(tr* ) - I ]/2' virtual valence transition, (c) 138.5 eV correspond*--- 2P3/2, ing to Rydbergs , - 2p3/J2,1/2, (d) 147 eV and (f) 157 eV corresponding to the two continuum shape resonances, (e) 152 eV

corresponding to the trough in between the two continuumshape resonances, (g) 169.5 eV correspondingto the P 2p continuum; and (h) 193 eV correspondingto the 7e(tr *)*-- 2s-~ excitation.

Fig. 12. Absolute partial photoionization differential oscillator strengths for the molecularand dissociativephotoionizationof PF3 in the P 2p and P 2s regions from 130 to 193 eV. The dotted curves represent the estimated PPOS contributionsfrom the underlying valence-shellcontinuum in the inner-shell region. The top panel presents the total (valence plus P 2p, 2s) absolute photoabsorption spectrum where the thresholds of the P 2p and 2s edges are shown as hatched lines. measured data, it should be remembered that the T O F mass spectra show the abundances of the ions relative to the highest peak in the spectrum, whereas the photoion branching ratios (see below) indicate the percentages of the total photoionization contributed by each ion detected taking into account the peak area (i.e., considering both the width and height of each peak). The photoion branching ratios for production of the singly and doubly charged cations obtained in the present work in the P 2p, 2s regions (including the contribution from the underlying valence shell) are presented in the higher energy region of Fig. 6. The error bars are derived from the square roots of the total and background coincidence counts and represent statistical error only. Note that numerical values of the branching ratios can be generated from the

416

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418

ratio of the individual ion PPOS to the total photoabsorption (Table 7) at each photon energy. In moving from the valence-shell to the 2p inner-shell continuum region, it can be seen from the photoion branching ratio curves that significant changes occur in the ionic photofragmentation of PF3: The relative amount of PF~- decreases markedly, whereas the relative yield of PFf shows little change. In contrast, the relative yields of PF ÷ and P+ show localized increases in the pre-edge excitation region, while F + and all doubly charged ions generally show significant increases throughout the 2p inner-shell region. The largest changes are in the branching ratios of PF~-, P+, PF 2÷, and p2+. However, it should be noted that since branching ratios are quantities relative to the total ionization, the branching ratio data of Fig. 6 do not show in a direct manner how the individual ionic photofragmentation channels are affected by the P 2p discrete excitation and ionization continuum regions. Such detailed quantitative information is however provided by comparison of the individual PPOS ( d f / d E) curves for molecular and dissociative photoionization with the P 2p photoabsorption as shown below in Fig. 12. For example, even though the branching ratio for PF~- markedly decreases (Fig. 6), the PPOS (Fig. 12 below) shows no significant change in P F f absolute yield in the P 2p region. The absolute partial photoionization differential oscillator strengths (Fig. 12, Table 7) for the production of molecular and fragment ions in the P 2p and 2s inner-shell regions of phosphorus trifluoride were determined from the product of the total photoabsorption differential oscillator strength and photoion branching ratio at each photon energy, and assuming photoionization efficiencies of unity (this assumes effects due to double or higher multiple photoionization, including dissociative processes such as Coulomb explosions, are small compared with direct single photoionization processes). The valence-shell PPOS contributions in the inner-shell region have been estimated by fitting polynomial curves (see dotted lines in Figs. 7 and 12) to the valence-shell PPOS data and then extrapolating them into the inner-shell region. The branching ratios (Fig. 6) and absolute partial photoionization oscillator strengths (Fig. 12) of the fragment ions can be discussed with reference to the total photoabsorption in the valence

continuum, the P 2p pre-edge excitation region, and the P 2p post-edge continuum region (see top panels of Figs. 6 and 12). It can be seen that the largest cations, PF~- and PF~-, are produced almost exclusively from the underlying valence-shell continuum. Although some of the molecular ion, PF~-, and possibly a very small amount of PF~ is produced from the 7e(a * ) ~- 2p- ~ core excited virtual valence and Rydberg states, there appears to be no appreciable increase in these yields above the 2P3/2" ~/2 ionization edges. In contrast, the other three singly charged ions (PF +, P+, and F +) are produced with increasing probability from P 2p and 2s excitation and ionization processes with decreasing size of the fragment ion. The singly charged ions, PF +, P+, and F +, and the doubly charged species, PF~ +, PF22÷, PF 2÷, and p2+, all show contributions from both the P 2p neutral excited states and the ionization continuum. Above the P 2p excitation threshold the PPOS of all four doubly charged ions (PF 2+, PF22÷, PF 2+, and p2+) arise primarily from inner-shell rather than valence shell processes. This observation can be attributed to autoionization decay (i.e., resonance Auger processes) being favored below the ionization threshold where the excited 2p electron remains as a spectator or participates in the autoionizing decay of the 2p-~ hole state to give singly charged ions, whereas above the 2p ionization threshold Auger processes become the principle decay channels, giving rise to doubly charged species [31]. In general, with the exceptions of PF2+, P÷, and F ÷, the singly charged cations are the predominant species in the P 2p pre-edge region, while all the dications make increased contributions in the post-edge continuum. In the P 2p continuum, doubly charged ions produced by Auger processes are expected in some cases to fragment by Coulomb explosion (i.e. dissociative double photoionization processes) to give two singly charged species. This will lead to greater yield of the lighter ion products, and as shown in Fig. 12, P+ and F + make the largest contributions from singly charged ions in the P 2p ionization continuum. These fragments also possess considerable kinetic energy of fragmentation, as can be seen by the increasingly broad peaks of the P+ and F ÷ ions in the TOF mass spectra above 138.5 eV (Fig. 11). This interpretation is supported by the PIPICO studies of Hitchcock et al. [33] which reported increasing inten-

J.W. Au et a l . / Chemical Physics 215 (1997) 397-418

sities of the (F +, P+ ), (F ÷, PF + ), and (F +, PFf ) ion pairs with increasing photon energy. At 145.6 eV small amounts of the (F +, p2+) and (F +, PF 2+) ion pairs were also observed [33]. However, since the pZ+ and PF 2+ ion peaks observed in the present work are relatively sharp and narrow (Fig. 11) even at the higher photon energies, it is unlikely that they are produced from triply dissociative photoionization to any large extent. The shape of the PPOS for PFf in Fig. 12 suggests that the (F +, P F f ) correlated pairs observed in the PIPICO experiments [33] arise principally from double ionization processes involving valence shell electrons even at 145.6 eV. In this regard, it should also be noted that the 10% detection efficiency of the presently used TOF mass spectrometer (see Section 2) causes both ions formed in Coulomb explosions to contribute to the TOF mass spectra in approximately equal proportions. Finally, note that there is no way to relate the PIPICO intensities [33] quantitatively to those of the present TOF mass spectra measurements, and only qualitative comparisons can be made. In particular, it is not possible to deduce the relative contribution to given fragment ions from dissociative single ionization and dissociative double (or higher) multiple photoionization processes. In both experiments the data also involve contributions from the underlying valence shell.

Acknowledgements Financial support for this work was provided by the Natural Sciences and Engineering Research Council of Canada and the Canadian National Networks of Centres of Excellence (Centres of Excellence in Molecular and Interfacial Dynamics). One o f us ( J W A ) gratefully a c k n o w l e d g e s an F.J. N i c h o l son s c h o l a r s h i p . T h e a u t h o r s w o u l d also like to t h a n k P r o f e s s o r A.P. H i t c h c o c k f o r useful d i s c u s s i o n s and for p r o v i d i n g us with his i n n e r - s h e l l total and partial ion yield data as well as P I P I C O m e a s u r e m e n t s p r i o r to publication.

References [1] M. Inokuti, ed., Proceedings of workshop on electronic and ionic collision cross sections needed in the modeling of

417

radiation interaction with matter, Argonne, IL, 1983 (Argonne National Laboratory, Argonne, IL, 1984) Rep. No. ANL-8429. [2] J.W. Gallagher, C.E. Brion, J.A.R. Samson, and P.W. Langhoff, J. Phys. Chem. Ref. Data 17 (1988) 9, and references therein. [3] C.M. Humphries, A.D. Walsh, and P.A. Warsop, Discuss. Faraday Soc. 35 (1963) 148. [4] M.J. McAdams and B.R. Russell, Chem. Phys. Lett. 18 (1973) 402. [5] E. lshiguro, S. lwata, A. Mikuni, Y. Suzuki, H. Kanamori, and T. Sasaki, J. Phys. B 20 (1987) 4725. [6] I.J. V~iyrynen, T.A. Kaurila, R.G. Cavell, and K.H. Tan, J. Electron Spectrosc. Relat. Phenom. 61 (1992) 55. [7] N. Kosugi, R.G. Cavell, and A.P. Hitchcock, Chem. Phys. Lett., in press. [8] R.N.S. Sodhi and C.E. Brion, J. Electron Spectrosc. Relat. Phenom. 37 (1985) 145. [9] R.N.S. Sodhi and C.E. Brion, J. Electron Spectrosc. Relat. Phenom. 37 (1985) 97. [10] A.W. PoUs, HA. Lempka, D.G. Streets, and W.C. Price, Phil. Trans. R. Soc. London A 268 (1970) 59. [11] P.A. Cox, S. Evans, A.F. Orchard, N.V. Richardson, and P.J. Roberts, Discuss. Faraday Soc. 54 (1972) 26. [12] PA. Bassett and D.R. Lloyd, J. Chem. Soc. Dalton Trans. (1972) 248. [13] J.P. Maier and D.W. Turner, J. Chem. Soc. Faraday Trans. 2 68 (1972) 711. [14] J.C. Green, N. Kaltsoyannis, K.H. Sze, and M.A. MacDonald, J. Chem. Soc. Dalton Trans. (1991) 2371. [15] D.J. Reynolds, E.H. van Kleef, and I. Powis, J. Chem. Phys. 95 (1991) 8895. [16] I. Powis, J. Chem. Phys. 99 (1993) 3436. [17] I. Powis, Chem. Phys. Lett. 198 (1992) 473. [18] M. Xin, L.C. Chiu, D. Li, and Y. Pan, J. Electron Spectrosc. Relat. Phenom. 33 (1984) 93. [19] I. Powis, Chem. Phys. Lett. 215 (1993) 269. [20] J. Berkowitz, J.P. Greene, J. Foropoulos Jr., and O.M. Neskovic, J. Chem. Phys. 81 (1984) 6166. [21] J. Berkowitz and J.P. Greene, J. Chem. Phys. 81 (1984) 4328. [22] D.F. Torgerson and J.B. Westmore, Can. J. Chem. 53 (1975) 933. [23] C.R.S. Dean, A. Finch, P.J. Gardner, and D.W. Payling, J. Chem. Soc. Faraday Trans. 1 70 (1974) 1921. [24] W.H. Beattie, Appl. Spectrosc. 29 (1975) 334. [25] R. Riiede, H. Troxler, Ch. Beglinger, and M. Jungen, Chem. Phys. Lett. 203 (1993) 477. [26] R.G. Cavell, Inorg. Chem. 14 (1975) 2828. [27] R.G. Cavell and R. Sodhi, J. Electron Spectrosc. Relat. Phenom. 15 (1979) 145. [28] R.N. Sodhi and R.G. Cavell, J. Electron Spectrosc. Relat. Phenom. 32 (1983) 283. [29] B.E. Mills, R.L. Martin, D.A. Shirley, J. Am. Chem. Soc. 98 (1976) 2380. [30] W.L. Jolly and W.B., Perry, Inorg. Chem. 13 (1974) 2686. [31] F.P. Larkins, E.Z. Chelkowska, Y. Sato, K. Ueda, E. Shigemasa, and A. Yagishita, J. Phys. B 26 (1993) 1479.

418

J.W. Au et aL / Chemical Physics 215 (1997) 397-418

[32] A.P. Hitchcock, N. Kosugi, and R.G. Cavell, in preparation. [33] A.P. Hitchcock, J. Bozek, F. Schlachter, T. Tyliszczak, D. Kilcoyne, and R. Cavell, unpublished results. [34] W.F. Chan, G. Cooper, and C.E. Brion, Phys. Rev. A 44 (1991) 186. [35] W.F. Chan, G. Cooper, and C.E. Brion, Chem. Phys. 168 (1992) 375. [36] W.F. Chan, G. Cooper, R.N.S. Sodhi, and C.E. Brion, Chem. Phys. 170 (1993) 81. [37] J.W. Au, G. Cooper, G.R. Burton, T.N. Olney, and C.E. Brion, Chem. Phys. 173 (1993) 139. [38] R.D. Hudson, Rev. Geophys. Space Phys. 6 (1971) 305. [39] S. Daviel, C.E. Brion, and A.P. Hitchcock, Rev. Sci. Instrum. 55 (1984) 182. [40] C. Backx and M.J. Van der Wiel, J. Phys. B 8 (1975) 3020. [41] C. Backx, G.R. Wight, R.R. Tol, and M.J. Van der Wiel, J. Phys. B 8 (1975) 2050. [42] F. Carnovale and C.E. Brion, Chem. Phys. 74 (1983) 253. [43] X. Guo, G. Cooper, W.F. Chan, G.R. Burton, and C.E. Brion, Chem. Phys. 161 (1992) 453. [44] G. Cooper, Y. Zheng, G.R. Burton, and C.E. Brion, Rev. Sci. lnstrum. 64 (1993) 1140. [45] W.F. Chan, G. Cooper, X. Guo, and C.E. Brion, Phys. Rev A 45 (1992) 1420. [46] G.R. Burton, W.F. Chan, G. Cooper, and C.E. Brion, Chem. Phys. 177 (1993) 217. [47] W.L. Wiese, M.W. Smith, and B.M. Glenon, Atomic transition probabilities, Vol. 1. Hydrogen through neon NSRDSNBS. Circ. No. 4 (US GPO, Washington, DC, 1966). [48] Z.F. Liu, J.N. Cutler, G.M. Bancroft, and K.H. Tan, Chem. Phys. 168 (1992) 133. [49] W. von Niessen, J. Schirmer, and L.S. Cederbaum, Comp. Phys. Rep. 1 (1984) 57. [50] G. Cooper, G.R. Burton, and C.E. Brion, J. Elec. Spectrosc. Relat. Phenom. 73 (1995) 139.

[51] W.F. Chan, G. Cooper, and C.E. Brion, Chem. Phys. 178 (1993) 40 I. [52] W.F. Chan, G. Cooper, and C.E. Brion, Chem. Phys. 180 (1994) 77. [53] J.A. Wheeler and J.A. Bearden, Phys. Rev. 46 (1934) 755. [54] M. Inokuti, personal communication (1985). [55] J.J. Yeh and I. Lindau, At. Data Nucl. Data Tables 32 (1985) 1. [56] B.L. Henke, P. Lee, T.J. Tanaka, R.L. Shimabukuro, and B.K. Fujikawa, At. Data Nucl. Data Tables 27 (1982) 1. [57] J.W. Au, G. Cooper, G.R. Burton, and C.E. Brion, Chem. Phys. 187 (1994) 305. [58] U. Fano and J.W. Cooper, Rev. Mod. Phys. 40 (1968) 441. [59] M. Inokuti, Rev. Mod. Phys. 43 (1971) 297. [60] T.N. Olney, N.M. Cann, G. Cooper, and C.E. Brion, Chem. Phys. submitted. [61] J.W. Au and C.E. Brion, Chem. Phys., in press. [62] J.W. Au and C.E. Brion, Chem. Phys., submitted. [63] N. Hacket and R.J.W. Le Fbvre, J. Chem. Soc. (1961) 2612. [64] G. Nagarajan, Z. Naturforsch. A 21 (1966) 238. [65] S.S. Batsanov, Refractometry and chemical structure (Consultants Bureau, New York, 1961). [66] E.R. Lippincott, G. Nagarajan, and J.M. Stutmann, J. Phys. Chem. 70 (1966) 78. [67] A.N. Pandey, A. Bigotto, and R.K. Gulati, Acta Phys. Pol. A 80 (1991) 503. [68] M.K. Wilson and R. Polo, J. Chem. Phys. 20 (1952) 1716. [69] J. Rolke and C.E. Brion, Chem. Phys. 207 (1996) 173. [70] M.B. Robin, Higher excited states of polyatomic molecules, Vol. I (Academic Press, New York, 1974). [71] I. Powis, J. Chem. Phys. 103 (1995) 5570. [72] J.G. Lias, J.E. Bartmess, J.F. Liebman, J.L. Holmes, R.D. Levin, and W.G. Mallard, J. Phys. Chem. Ref. Data 17 (1988) Suppl. No. 1. [73] G.B Armen and F.P. Larkins, J. Phys. B 24 (1991) 741.