Negative ion states of the halogens - American Chemical Society

Sep 18, 1984 - Department of Chemistry, University of Houston UP, Houston, Texas 77004 .... In the late 1970's Ayala, Wentworth, and Chen6 measured th...
1 downloads 0 Views 801KB Size
J . Phys. Chem. 1985, 89, 4099-4105

4099

Negative Ion States of the Halogens E. C. M. Chen School of Sciences and Technology, University of Houston CL, Houston, Texas 77058

and W. E. Wentworth* Department of Chemistry, University of Houston UP, Houston, Texas 77004 (Received: September 18, 1984; I n Final Form: May 20, 1985)

The negative ion states of the halogens, including the spin orbital substates, are characterized by calculating Morse potential energy curves from experimental data. Most of the curves are uniquely determined or are overdetermined. The properties of the negative ion curves have been related to those of the neutral by using three dimensionless parameters to modify the Morse curves of the neutral. Where extra data exist, the calculated properties agree well with the experimental values. For the two highest energy states of iodine, there are insufficient data so that reasonable parameters have been estimated. These curves are the only complete set of curves in the literature obtained from experimental data. The parameters are very consistent for the different halogens and suggest reasonable ranges for other molecules. The bond orders calculated from simple molecular orbital theory and the vibrational properties calculated from Badger’s rule are consistent with the values calculated from the parameters. The ionic radii can be calculated from the internuclear distances for the ground states of the neutral and the negative ion, assuming additivity of the ionic and atomic radii. Thus from this analysis, the ionic radii are F = 1.21 A, C1- = 1.63 A, Br- = 1.71 A, and I- = 1.90 A, in excellent agreement with ionic crystal radii.

Introduction

The halogen atoms have the highest electron affinities of all of the elements. As a result of this strategic position in the periodic table, the bond dissociation energies of the homonuclear diatomic anions are low, making it difficult to even characterize the ground states of the negative ions, much less, the excited states. At present, no single experiment has been used to characterize these negative ion states. During the past decade new data for these ions have been obtained from electron thermal electron attachment,6 electron swarm and alkali metal and photodestr~ction~~ studies. Two recent reviews have summarized the spectroscopic datal6 and alkali metal beam data.” Combining this new data with the earlier electron impact data,18-z0spectroscopic information,z1s2zand S C F calculation^,^^^^^ a good de( I ) Tam, w. C.; Wong, S . F. J. Chem. Phys. 1978,68, 5626. (2) Kurepa, M. V.; Belic, D. S. J. Phys. B 1978, 1 1 , 3719. (3) Kurepa, M. V.; Babic, D. S.; Belic, D. S. J. Phys. B 1981, 14, 375. (4) Azria, R.; Abouaf, R.; Teillet-Billy, D. J . Phys. B 1982, 16, L569. (5) Chantry, P. J. Westinghouse R & D Document No. 78-926-ATTACH-RI-1978. (6) Ayala, J. A.; Wentworth, W. E.; Chen, E. C. M. J. Phys. Chem. 1981, 85, 768. (7) Brooks, H. L.; Hunter, S. R.; Nygaard, K. J. J. Chem. Phys. 1979, 71, 1870. (8) Sides, G. D.; Tiernan, T.0.;Hanrahan, R. J. J . Chem. Phys. 1976, 65, 1966. (9) Schultes, E.; Christoloudides, A. A,; Schindler, R. N. Chem. Phys. 1975, 8, 354. (10) Baede, A. P. M. Adu. Chem. Phys. 1975, 30, 463. (11) Lacmann, K. Adu. Chem. Phys. 1980, 42, 513. (12) Anderson, R. W.; Herschbach, D. R. J. Chem. Phys. 1975,62,2666. (13) Tang, S. Y.; Leffert, C. M.: Rothe, E. W. J. Chem. Phys. 1975, 61, 2592. (14) Hubers, M. M.; Klevn, A. W.: Los. J. Chem. Phvs. 1976, 17. 303. (15) Lee, L. C.; Smith, G. P.; Moseley, J. T.; Cosby, P. C.; Guest, J. A. J . Chem. Phys. 1979, 70, 3237. (16) Andrews, L. Annu. Rev. Phys. Chem. 1979, 70, 79. (17) Kleyn, A. W.; Los, J.; Gislason, E. A. Phys. Rep. 1982, 90, 1. (18) Truby, F. K. Phys. Rev. A 1971, 4, 617. (19) Frost, D. C.; McDowell, C. A. Can. J. Chem. 1960, 38, 407. (20) DeCorp, J. J.; Steiger, R. P.; Franklin, J. L.; Margrave, J. L. J. Chem. Phys. 1970, 53, 936. (21) Delbecq, C. J.; Hayes, W.; Yuster, P. H. Phys. Reu. 1961,121, 1043.

0022-3654/85/2089-4099$01.50/0

scription of the negative ion states of the molecular halogens is possible. In the late 1970’s Ayala, Wentworth, and Chen6 measured the rate constants and the activation energies for thermal electron attachment to chlorine, bromine, and iodine using the electron capture detector. Potential energy curves were calculated for the two lowest negative ion states by using a single parameter to modify the Morse potential of the neutral molecule. At the time, the higher states could not be defined. Now, there are sufficient data to define almost all of the curves. The negative ion curves have been related to the Morse potentials of the neutral by using three dimensionless parameters. The values of the parameters obtained for the halogens can serve as guidelines for cases where there are insufficient experimental data. The work has been divided into three phases: (1) the development of the relationship between the properties of the negative ion and those of the neutral; (2) the discussion of the general calculation procedure; and (3) the discussion and comparison of the results for the halogens with those in the literature. There are sufficient data to define the Morse potentials for all of the negative ion states of the halogens, including the spin substates, except for the two highest energy states of iodine. These calculations rely on the assignment of the experimental data to specific electronic states. No attempt will be made to consider the many possible applications of these results to other studies. Theoretical

The Morse potential referenced to zero energy at infinite separation of the ground state atoms is given by

~ ( X Z=)-2Dx, exp(-p(r

- re))

+ Dx, exp(-2p(r

- re))

(1)

where Dx, is the spectroscopic bond dissociation energy, r is the X-X separation, and re = r at the minimum of U ( X 2 ) . The constant, @, is defined by

p = ve(2*Zp/D)’~2 where we is the fundamental vibrational frequency and is the reduced mass. The parameterized Morse potential for the negative ions is given by (22) Person, W. B. J. Chem. Phys. 1963, 38, 109. (23) Gilbert, T. L.; Wahl, A. C. J. Chem. Phys. 1971, 55, 5247. (24) Rescigno, T. N.; Bender, C. F. J . Phys. B 1976, 9, L329.

0 1985 American Chemical Society

4100 The Journal of Physical Chemistry, Vol. 89, No. 19, 1985

L‘(x,-)= - 2 k ~ D x ,exp(-kB8(r - re)) + kRDx, exp(-2k~(?(r - re))- EAx

+ E*x

(2)

where the parameters kA, kB, and kR are constants, Ex* is the energy of an excited state of X relative to the ground state, and EAx is the electron affinity of X. This is also a Morse potential and the relationships between the negative ion properties and the neutral properties can be derived following the procedures given in ref 6. Dx; = DX,(kA2/W -

re.X2

=

(3)

In ( k ~ / k ~ ) kB8X2

@X;

=

kBPX2

+ Te.X2

(4)

(5)

Chen and Wentworth the discrepancy between the calculated distributions using eq 11, and the experimental data is accounted for in terms of the survival factor which varies with energy. This could be done for bromine and the survival factor determined as a function of energy. It should be emphasized that the majority of the present curves have been obtained from experimental data. That is, the dimensionless parameters kA, kR, and kB have been used only as an expedience for obtaining values of the Morse parameters. Alternatively, the curves could have been determined by using the frequency, dissociation energy, and internuclear distance for each of the negative ion states, independent of the properties of the neutral molecule. In fact, this was essentially the procedure used by Person22earlier. However, in cases where insufficient data exist, reasonable curves can be obtained from the neutral curves by using estimates for the dimensionless parameters as was done for the two highest states of iodine.

Procedures for Calculations Equations 1-11 have been programmed on an Apple 11+ computer and the graphs have been plotted on a C Itoh dot matrix printer with a scale of 1/144 in. = 0.01 eV. The neutral curves are defined by the quantities De, re, and Y,. To calculate the curves for the negative ions, the parameters kA, kR, and kBand the values of EAx and Ex*must be input. Thus three or more experimental EV = Dx,(l - 2 k ~ kR) - EAx + Ex* -‘/hv,,x, data points uniquely define or overdetermine the curves. If there (7) are fewer than three data points, then assumed values of one or more parameters must be used. Even if there are three data points, the calculation is not simply a matter of substituting these into the modified Morse potential since some of the data is implicit, e.g., the various cross sections. Interestingly enough, these two equations define kA and kR since Since an iterative procedure is used the calculated data may not there is only a slight dependence of these quantities on kB. The exactly reproduce the experimental values. Sometimes, two other constant, kB, can be determined from the full width at different experiments give the same data. For example, the width half-maximum (fwhm) of the dissociative electron attachment of the electron attachment cross section at zero energy is autocross section or the activation energy for thermal electron attachment, or it can be estimated from Badger’s r ~ l e ~ ~ s ~ ~matically small and cannot be used as independent data. Examples of experimental data are the adiabatic and vertical (9) /3 = Cij/((ri - di,j)’/2Dx21/2) electron affinities, cross section, and activation energy for dissociative electron attachment; and the negative ion absorption The terms Cijand dijare constants for a given row in the periodic spectra, vibrational frequency, internuclear separation, and table, which leads to the following relationship: photodestruction or photodetachment cross sections. If the vertical energy and the adiabatic electron affinity are available, then eq 7 and 8 yield the parameters k, and kR, and kBcan be determined from the fwhm for the dissociative electron attachment cross section, the internuclear distance, the activation energy for thermal Based on eq 3, a purely repulsive curve can only be obtained electron attachment, or the vibrational frequency. If these are if kA = 0, thus reducing the number of parameters to two and not available then the value of kBcan be estimated from Badger’s giving a curve which is not strictly of the Morse form. From eq rule. 3, the bond order is given by BO = kA2/kR. From simple moFor all of the halogen ground states and first excited states, lecular orbital theory, the bond order is one-half of the net number there are three or more data points. For the second excited states, of bonding electrons which is 0.50 for the halogens. it was assumed that the curves are totally repulsive so that only The cross sections for dissociative electron attachment can be two other data points were required. For the highest state this calculated from the following expression which assumes unit or assumption was also made, but for all except iodine the electronic constant probability of dissociation. The general temperature absorption maximum is an extra experimental point. In the case range of mass spectrometric experiments is T < 500 K so that of the 211u,1i2 and 22g+ states for iodine there are insufficient data only the first two vibrational levels have been used in the calcuso that values of the parameters were assumed. lations. The details are given in ref 6 and the other references cited therein. Results and Discussion The potential energy curves (eq 1 and 2) and the calculated distributions for dissociative electron attachment (eq 11) are the results of this study. These are shown in Figures 1-4, with the experimental distributions given as dashed lines. The relative n is the vibrational quantum number, u is the velocity of the magnitudes of the cross sections are not significant since the electron, and IT,,({)are the Hermite polynomial solutions to the equation for the cross section does not contain a survival factor. harmonic oscillator. In Figure 5, the calculated and experimental distributions for each As will be seen in the Results and Discussion section, eq 11 state are normalized individually for better comparison. The gives a good fit between the calculated distributions and the parameters used to generate these curves are given in Table I along experimental data in all cases except for bromine. On this basis, with the number of experimental data points for each state. The we feel justified in assuming a constant probability of dissociation assumption of a purely dissociative curve for the higher states has (survival factor), independent of energy. In the more exact theory, been included as a data point. There are at least three points for each curve, except for the two highest states of iodine. Nine curves are overdetermined as can be seen in Table 11. The calculated (25) Badger, R. M. J . Chem. Phys. 1934, 2, 128. values used to define the curves are marked with an asterisk. In (26) Badger, R. M. J . Chem. Phys. 1935, 3, 710. Two two most frequently measured quantities for negative ions are the vertical energy, EV, and the adiabatic electron affinity of the molecule, EA. These can be related to the parameters as follows:

+

The Journal of Physical Chemistry, Vol. 89, No. 19, 1985 4101

Negative Ion States of the Halogens

Br $. Br

F.F

F

bI

I

1

I

I

2

I

I

3

I

I

4

I

I

5

I

+

f -

I

6

r[A]

Figure 1. The Morse potentials for F2

from eq 1 and and the cross sections for dissociative electron attachment calculated from eq 11, T = 25 OC. (---) experimental data taken from ref 1, 5 , and 20 normalized to the calculated cross section. The relative magnitudes are not significant.

Figure 3. The Morse potentials for Br2 calculated from eq 1 and 2 and the cross sections for dissociative electron attachment calculatedfrom 11, = 25 OC. (---I Experimentaldata taken from ref 4 normalized to the calculated cross section, The relative magnitudes are not signif-

icant.

2-

u

(ev>

-

-

1+1

0-

-2

-2 - %",

I'+ I-

-

I +I-

-

-4

7

P

FI

Figure 2. The Morse potentials for CI2 calculated from eq 1 and 2 and the cross sections for dissociative electron attachment calculated from eq 11, T = 25 OC. (---) Experimental data taken from ref 3 and 4 normalized to the calculated cross section. The relative magnitudes are not significant.

Table 111, the properties and parameters for the ground state are compared with estimates from other sources such as the SCF calculations of Gilbert and Wah123and of Rescigno and the curves of Hubers, Kleyn, and Losl4 and Person,22 and our earlier curves.6 N o attempt has been made to consider all of the theoretical curves in the literature. Two items listed in Table I are not experimental data points. These are kB obtained from Badger's tule and the bond order estimated from the net number of bonding electrons. The value of kB is an especially valuable comparison for the ground state while the bond order is more important for the excited states. Badger's rule does not apply for molecules with very low bond dissociation energies. However, Keyes" has assumed a constant value of kB = 1 in using a single parameter Morse potential to (27) Keyes, R. W. J . Chem. Phys. 1958, 29, 523.

:"., "0.

'

1

"

'

1

n

.n

1.

Figure 4. The Morse potentials for I2 calculated from eq 1 and 2 and

the cross sections for dissociative electron attachment calculated from eq 11, T = 25 "C. (---) Experimental data taken from ref 1 normalized to the calculated cross section. The relative values of the magnitudes are not significant. describe excited states of diatomic molecules. Thus the value of kB should fall between that of the ground state of the negative ion and 1.O. There is a very good comparison between the values calculated from Badger's rule and the values used for the ground-state curves. In the case of the 2rIu,1,2 state for bromine, a constant value of kB was assumed to prevent the spin substates from crossing in the Franck-Condon region. The bond order for the negative ion ground state is predicted to be 0.5 and for the first excited state 0-0.2 from the net number of bonding electrons. The corresponding calculated values range from 0.54 to 0.74 and from 0.01 to 0.09. The higher excited states have a zero bond order so kA = 0. We have selected from the reported experimental values of the adiabatic electron affinities those which agree to fO.l eV. These are given in Table I1 along with the average value and the values

4102 The Journal of Physical Chemistry, Vol. 89, No. 19, 1985

Chen and Wentworth

TABLE I: Parameters for the Neutral and the Negative Ions of tbe Halogens molecule state N k, kR ks EA

S, A-'

D., eV

1.oo 4.16 1.62 2.90 4.85

1.oo 0.60 0.52 0.40 0.43

0.00 3.40 3.40 3.40 3.40

2.87 1.72 1.48 1.15 1.24

1.63 1.26 0.02 0.00 0.00

1.42 1.92 3.02

892 462 54

1.oo 1.15 0.45 0.45 0.00 0.00 0.00

1.oo 2.45 2.30 2.42 2.40 2.80 4.40

1.oo 0.60 0.60 0.60 0.55 0.60 0.70

0.00 3.61 3.61 3.61 3.61 3.50 3.50

2.01 1.21 1.21 1.21 1.11 1.21 1.41

2.52 1.38 0.22 0.21 0.00 0.00 0.00

1.99 2.62 3.34 3.39

565 249 101 98

4 3 3 3 4

1.oo 1.21 0.35 0.20 0.00 0.00 0.00

1.oo 2.42 1.63 1.86 2.70 3.20 4.88

1.oo 0.63 0.60 0.70 0.83 0.83 0.83

0.00 3.36 3.36 3.36 3.36 2.91 2.91

1.95 1.23 1.17 1.37 1.62 1.62 1.62

2.00 1.22 0.15 0.04 0.00 0.00 0.00

2.28 2.85 3.60 3.92

323 158 53 33

1.oo 1.32 0.30 0.10 0.00 0.00 0.00

1.oo 2.52 1.30 1.82 2.80 3.30 4.45

1.oo

3 4 4 4 1 2

0.63 0.73 0.73 0.69 0.90 0.90

0.00 3.06 3.06 3.06 3.06 2.12 2.12

1.84 1.16 1.34 1.34 1.34 1.66 1.66

1.58 1.10 0.11 0.01 0.00 0.00 0.00

2.67 3.23 3.76 4.83

215 113 41 12

3 4 3 4 5 3

3 3

3 4 3

4

I

P

.5 2

4

2 4

+ EX*, eV

1.oo 1.76 0.15 0.00 0.00

10 12

6

8

6

8 10

E[eVI

E'leVI

Figure 5. The cross sections for dissociative electron attachment for the halogen molecules calculated from eq 11, T = 25 OC. ( - 0 - 0 ) Experimental data taken from the following references: F2 1, 5, 20; C12 3, 4; Br22; and I2 1. The distributions are normalized to the calculated cross

section for each state. The relative magnitudes are not significant. used in the calculations. The vertical electron affinities are less well defined but there are two sources: the interpretation of alkali metal beam reaction cross sections with the molecule^'^ and the correlation of the charge transfer complex maximum absorption with the halogens as acceptors. There is no experimental value for the vertical electron affinity of FZ.The charge transfer values are relative to the electron affinity of the I atom and have been corrected to the currently best available value, 3.06 eV. Interestingly, the Morse potentials presented by PersonZZwere calculated to explain the charge transfer complex absorption. The error in the values obtained from alkali metal beam studies is quoted as *0.05 eV, but the values obtained from the two different techniques differ by more than this (see Table 11). The vertical electron affinity of C1, is an independent calculation from these potential energy curves but the calculated value does not agree

r,,

A

w,,

cm-'

with the experimental value within the error quoted above. The vertical electron affinities of iodine and bromine have been used to define the ground-state curves. The absorption spectra of all of the halogen negative ions have been measured in the solid state by Delbecq, Hayes, and Yuster.21 The ions were generated by X-ray irradiation of alkali halide crystals with added impurities at liquid nitrogen temperatures. Andrews16also measured the electronic absorption spectra of the negative ions of the halogen molecules formed by reaction with alkali metals in an argon matrix at 17 K. The photodestruction spectra of Clz- was measured by Lee, Smith, Mosely, Cosby, and GuestIs in the gas phase. These latter results are about 0.1-0.2 eV higher than the alkali metal crystal values. For consistency, the values of Delbecq et a1.*' have been used. The curves can be made to fit the other data by changing k B about 10%. The vibrational frequencies of the ground state of the negative ions have been measured by Raman spectroscopy of the matrix isolated negative ions. These values have been used to define kB for the ground states. The preexponential terms and the activation energies for the rate constants for thermal electron attachment to all of the halogens have been measured.6 The rate constants were measured at more than two temperatures for all of the halogens except for fluorine.&9 The preexponential terms are larger for chlorine and fluorine than for bromine and iodine, implying that there is a smaller contribution to thermal electron attachment by the ground states of the negative ions of the latter two molecules. The activation energies have been used to define the ground states for fluorine and chlorine. For bromine and iodine, the activation energy corresponds to the first excited state but has not been specifically used to define these curves so that these data are a point of comparison. The experimental activation energy is a lower limit for the crossing points of the other curves, including the ground state in agreement with the data shown in Table 11. The dissociative electron attachment cross sections for all of the halogens have been measured by Tam and Wong' who found three peaks at energies less than 6.0 eV for all of the halogens except for fluorine, where only a single peak at zero was observed. However, Chantry5 and DeCorpo, Steiger, Franklin, and MargraveZoobserved two other peaks at 3.0 and 6.0 eV for fluorine. Very accurate measurements of the dissociative electron attachment cross sections for chlorine and bromine have been reported by Kurepa and B e l i ~and ~ , ~for chlorine by Azria, Abouaf, and

The Journal of Physical Chemistry, Vol. 89, No. 19, 1985 4103

Negative Ion States of the Halogens

TABLE II: Experimental and Calculated Properties of the Halogen Negative Ion States 2XU+

EA(v), eV

EA(ad), eV

E*,' (eV)

ye:

cm-I

BO

kB

calcd -0.69

lit.

calcd

lit.

calcd

BR'

calcd

MO

calcd

0.02

0.00'

452 to 475

462'

0.53

0.60

0.50

0.74

14 22 12

-0.71

0.04

0.05.

225 to 264

249'

0.60

0.60

0.50

0.54

-1.47 -1.2 -1.6

14 22 12

-1.36'

0.50

162

158'

0.63

0.63

0.50

0.61

-1.72 -1.7 -1.7

14 22 12

-1.67'

1.44

114 to 116

113'

0.63

0.63

0.50

0.69

lit.

ref

calcd

F2

3.08 2.94 3.01

31 32 av

3.00;

Cl2

2.52 2.38 2.40 2.50 2.50 2.45 2.46 2.46

33 31 34 35 36 31 38 av

2.46;

-1.02 -1 .o -1.2

Br2

2.51 2.62 2.55 2.55 2.64 2.50 2.56

31 39 40 37 34 35 av

2.57"

I2

2.60 2.58 2.42 2.60 2.55 2.52 2.50 2.55

33 31 39 41 31 40 35 av

2.57;

lit.

ref

2 n .i

E W , eV

E*, eV

fwhm, eV

E(abs), eV

BO

lit.

ref

calcd

lit.

ref

calcd

lit."

calcd

lit.d

calcd

MO

calcd

win i

F2

0.00

1

0.00

0.5

1, 5

0.6;

0.02

0.00'

1.65

1.66;

0-0.2

0.01

both

CI2

2.50 2.60 2.40

1 2 4

2.50' 2.60'

1.3 1.2 1.3

1 2 4

1.3; 1.3;

0.30

1.58 1.65

1.58; 1.65'

0-0.2 0-0.2

0.09 0.08

312 1/2 112

Br2

0.60 1.60 1.40

3 3 1

0.50; 1.60'

0.8 0.8 1.2

3 3 1

0.8' 0.9;

0.01

0.01 0.15

1.38 1.65

1.37, 1.65'

0-0.2 0-0.2

0.08 0.02

312 1/2 112

12

0.03 0.90

1 1

0.03' 0.90'

0.1 1.0

1 1

0.1. 1.0'

0.03 0.01

0.03 0.08

1.08 1.55

1.08; 1.55'

0-0.2 0-0.2

0.07 0.07

312 1/2

0.34

2nu,i E(v), eV

fwhm, eV

E*, eV

calcd

spin i

0.56

2.80

0.00

0.00'

both

1.5;

0.97

2.90

0.00

0.00'

1.6'

1.05

3.05

0.00

0.00;

3/2 both 112

1.8;

0.46

2.10

0.00

0.00'

2.4

0.65

2.22

0.00

0.00;

1.5 2.0

0.35 0.53

2.12' 2.81

0.00 0.00

0.00;

calcd

lit.

ref

calcd

F2

3.00

5, 20

3.00"

1.3

5, 20

1.3;

C12

5.40 5.50 6.00

4 1, 2 4

5.40;

1.4 1.6 1.6

4 1 4

3.80 3.70 5.40

3 1 3

3.80* 5.40;

1.8 1.7 1.6

3 1 3

2.50

1

2.50;

1.5

1

Br2

I2

BO

MO

ref

6.00,

E(abs), eV calcd

lit.

lit.

calcd

lkd

2.12

o.oo*

312 312 112 3/2 112

2z*+

E(vh eV

F,

ci, Br2 I2

lit. 6.00 9.60 8.60

ref 5, 20 2 3

fwhm, eV calcd

6.00" 9.60' 8.60; 8.20

lit. 2.6 3.6 2.0

ref 5, 20 2 3

E*, eV calcd 2.6 3.5 3.3 3.3

lit.

calcd 0.85 1.19 0.83 0.63

BO

E(abs), eV lit.d 3.65 3.40 3.22 3.10

calcd 3.65; 3.40' 3.22' 3.10"

MO

calcd

0.00 0.00 0.00 0.00

O.OO* O.OO* 0.00;

0.00,

"Reference 6. *Reference 16. CBR denotes kB calculated from Badger's rule eq 10. dReference 21.

Teillet-Bill~.~ These are shown in Figure 5 . Azria et al! reported measurements as a function of angle of observation for chlorine. However, they do not report a peak at 9.6 eV as do Kurepa et

aL3 The data for fluorine and iodine were read from graphs presented by DeCorpo et al.,*O C h a n t r ~and , ~ Tam and Wong,' and are less reliable. The scale on the graph given by Tam and

4104

The Journal of Physical Chemistry, Vol. 89. No. 19, 1985

Chen and Wentworth

TABLE III: Comparison of the Ground States for the Negative Ions of the Halogens B, A-’ ye, cm-’ molecule kA a,, eV re, A kB kR 4.16 0.60 1.26 1.92 1.72 462 F, 1.76

ClZ

Br2

12

EA(ad), eV

EA(v), eV

E*, eV

BO

ref

-0.69 -0.81 -0.8 1 -3.05

0.00 0.00 0.00 6.28

0.74 0.99 0.80 0.80

a

2.23 1.84 1.00

5.03 4.25 1.24

0.58 0.56 1.00

1.67 1.34 1.36

1.91 1.94 1.49

1.65 1.61 2.87

511 448 801

3 .OO 3.40 3.08 3.08

1.15 1.20 1.10 1.04 1.00

2.45 2.83 2.16 2.11 2.03

0.60 0.65 0.54 0.54 1.00

1.38 1.30 1.43 1.31 1.26

2.62 2.65 2.61 2.64 2.34

1.21 1.31 1.09 1.09 2.01

249 262 228 218 396

2.46 2.38 2.50 2.39 2.34

-0.71 0.00 -1.19 -1.02 -1.02

0.05 0.00 1.71 0.52 0.04

0.54 0.51 0.54 0.51 0.49

a

1.21 1.31 1.27 1.11 1.00

2.42 2.80 2.63 2.18 1.88

0.63 0.63 0.63 0.62 1.00

1.23 1.24 1.24 1.14 1.08

2.85 2.90 2.88 2.85 2.61

1.23 1.23 1.23 1.20 1.95

159 160 160 150 236

2.57 2.59 2.59 2.49 2.61

-1.38 -1 .oo -1.18 -1.44 -1.60

0.53 0.13 0.25 1.26 0.27

0.61 0.61 0.61 0.57 0.53

a

1.32 1.09 1.40 1.20 1.00

2.52 2.04 2.83 2.24 1.86

0.63 0.70 0.63 0.67 1.00

1.10 0.93 1.10 1.02 0.86

3.23 3.15 3.27 3.17 3.00

1.16 1.29 1.16 1.23 1.84

113 115 113 115 157

2.57 2.40 2.58 2.50 2.33

-1.67 -1.70 -1.43 -1.73 -1.70

1.44 2.07 0.48 2.15 0.38

0.69 0.58 0.69 0.64 0.54

a

23 24 6 23 22 14 6 3 22 14 6 22 7 14 6

“This work. Wong’ was shifted to agree with the tabulated data for the onset and maxima. The assignments of the peaks in chlorine are based on the angular measurements by Azria et aL4 For fluorine, the low-energy peak is due to two states. In bromine and iodine, the ground state crosses the neutral at energies greater than 0.5 eV so that the contribution of the ground state to the zero energy peak is negligible. In the case of iodine, we have followed Person22 and have assigned the two lower peaks to the 211&,j2and the 211u,3/2 spin orbital states. The ground-state curves are uniquely defined except for chlorine. The choice of the activation energy over the vertical electron affinity to define the curve for chlorine is supported by the sharp peak in the dissociative electron attachment cross section at zero energy. The agreement between the calculated and experimental cross sections can be seen in Figures 2 and 5 . This represents an additional data point so that the ground-state curve is overdetermined by two points. The properties of the current ground-state curves, listed in the first rows of Table 111, can be compared with our earlier ones listed in the last row. The major differences are in the internuclear distance and the frequency, which are primarily determined by the factor kB which was assumed to be one. In the case of fluorine, the current values of the activation energy and the vertical electron affinity are quite different because the activation energy was assigned to the crossing of the excited state. Person22defined the ground-state curves by estimating values of D, re,and v, so that, for these curves, the vertical electron affinity and the activation energy are the points of comparison. The theoretical curves are generally independent but Rescigno and Bender24 shifted the theoretical curve to fit the experimental adiabatic electron affinity. The agreement between the curves for bromine and iodine is artificial since much of the same data (for example v, and EA (adiabatic)) have been used to define the curves. The major differences rest in the activation energies and the vertical electron affinities, but all of the curves show a “back side” crossing. The good agreement (average deviation f0.03A) between the values of the internuclear distance calculated from the current curves and the values given in the literature is very important. Since we have not specifically used these data to define the curves, it is an independent point of comparison. In addition, there is very good agreement between the internuclear distances for the current curves and the sum of the covalent radius of the halogen atom and the ionic radius of the anion. The greatest deviation is 0.14 A for Iodine. The traditional set of ionic radii summarized by Pauling2*were obtained by division of the minimum internuclear separation, ro, (28) Pauling, L.“The Nature of the Chemical Bond”; Cornell University Press: Ithaca, NY, 1960; 3rd ed, pp 537-540.

TABLE I V Atomic and Anionic Radii for the Halogen Atoms rx-, A

atom F

(covalent), gas electron rx, 8, phase“ density*

CI Br I

0.71 0.99 1.14 1.33

1.21 1.63 1.71 1.90

1.16 1.64 1.73 2.04

empirical c

b

Pauling“

1.19 1.67 1.82 2.06

1.19 1.65 1.80 2.01

1.33 1.81 1.96 2.16

“This work. bReference30. CReference29. “Reference 28. in proportion to the inverse ratio of the effective nuclear charges of the constituent ions. Shannon29and Morris30 have summarized improved values of empirical “radii” based on multiple correlations. Another approach to measuring the ionic size is to examine the X-ray diffraction pattern of crystals to determine the loci of points in space between the anion and cation where the electron density falls to a minimum to define the ionic radius. These results, which are considered to be the “best” current values, have been summarized by Morris.30 A comparison of the various ionic radii from the literature is given in Table IV along with the covalent radii of the halogens and the gas-phase ionic radii of the anions determined by substracting the covalent radii from the internuclear distance of the ground state. Since the internuclear distance was not used to define the curves, this procedure gives “gas phase” ionic radii which can be directly compared with the solid-state data. For the lower spin state of the first excited states, the activation energy is an extra point, except for chlorine. The definitive data are the absorption spectra from the ground state of the negative ion and the dissociative electron attachment cross sections. The fwhm values for the distribution are given in Table I1 but the total fit to the distributions can be seen in Figures 1-5. The spin orbital splitting is observed for all the halogens except fluorine and in(29) Prewitt, C. T. Acta Crystallogr., Sect. A 1976, A32, 75 1. (30) Morris, D. F. C. Struct. Bonding 1968, 4, 63. (31) Chupka, W. A,; Berkowitz, J.; Gutman, D.J . Chem. Phys. 1971, 5 5 , 2124. (32) Harland, P. W.; Franklin, J. L. J . Chem. Phys. 1974, 61, 1621. (33) DeCorpo, J. J.; Franklin, J. L. J . Chem. Phys. 1971, 54, 1885. (34) Dispert, H.; Lacmann, K. Chem. Phys. Lett. 1977, 47, 533. (35) Tang, S.Y . ;Leffert, C. B.; Rothe, E. W.; Reck, G.P. J . Chem. Phys. 1975. -, 62. - - , 132. --~

(36) Lacmann, K.; Herschbach, D.R. Chem. Phys. Lett. 1970, 6, 106. (37) Baede, A. P. M. Physica 1972, 59, 541. (38) Dunkin, D.B.; Fehsenfeld, F. C.; Ferguson, E. E. Chem. Phys. Lett. 1972, 15, 257. (39) Hughes, B. M.; Lifshitz, C . ;Tiernan, T. 0.J . Chem. Phys. 1973, 59, 3162. (40) Baede, A. P. M.; Auerbach, D. J.; Los, J. Physica 1973, 134. (41) Moutinho, A. M. C.; Aten, J. A,; Los, J. Physica 1971, 53, 471.

J . Phys. Chem. 1985, 89, 4105-4112 creases from chlorine to iodine. Lee et al.I5 give a bond dissociation energy of 0.16 eV for the first excited state of chlorine, as compared to a value of 0.22 eV given in Table I. Kurepa et alS3and Personzz report purely dissociative curves for bromine and iodine, respectively, while dissociation energies of 0.16 and 0.04 eV for bromine and 0.1 1 and 0.01 eV for iodine are shown in Table I. The two spin states for iodine cross on different sides of the neutral. Our earlier excited-state curves were purely dissociative, except for bromine. The second excited state curves are uniquely determined, assuming dissociative curves, except for iodine. For chlorine and bromine the two spin states are apparent. The fit of the data to the two separate peaks is shown in Figures 2, 3, and 5. However, rather than allow the spin orbital states of bromine to cross in the Franck-Condon region, we have chosen to keep the kBvalues constant and sacrifice the fit of the distribution. There are four data points for the highest excited states for all the halogens except iodine. The extra data are the dissociative attachment distribution given in Figures 1, 2, 3, and 5. The greatest deviation is for bromine.

4105

The bond order for the ground state is 0.5 from simple molecular orbital theory while the calculated values vary from 0.54 to 0.74. For the first excited state the calculated values vary from 0.01 to 0.10 in agreement with the count of bonding and antibonding electrons. This justifies the use of a purely repulsive curve for the other excited states. The values of kB obtained from the experimental data and the values calculated from Badger’s rule are consistent, so that in the cases where there are insufficient data this parameter can be estimated from the other two. This parameter for the ground state is an approximate lower limit for the value in the excited states where Badger’s rule does not apply. The dimensionlessquantities for a given state are approximately constant for all of the different halogens except for the ground state of fluorine. Based on this observation, limits can be placed on reasonable values of the dimensionless parameters as follows: 1.9 L kA 1 0.0 5.0 L kR L 1.0 1.0 L ke 2 0.4

Conclusions

The negative ion states of the halogens, including the spin substates, have been characterized by calculating the Morse potential energy curves from experimental data. In order to do this, specific state assignments were made. The Morse potentials of the negative ions were obtained by using three dimensionless parameters to modify the Morse curves of the neutral. Many of the curves are uniquely determined. Where extra data exist, the calculated properties agree well with the experimental values. For and the zZg+states of iodine, insufficient experimental the zIIu,1/2 data exist so that one or two of the parameters were estimated by analogy to the other halogens. These curves are the only complete set of curves in the literature obtained from experimental data and are an improvement over our earlier ones for the first two states.

These limits can be used to test the validity of parameters used for other molecules or can be used to estimate these parameters in the case of limited experimental data. The ionic radii of the halogen atoms have been determined from the internuclear separation in the ground state to be as follows: F = 1.21 A; C1- = 1.63 %.;Br- = 1.71 A; I- = 1.90 A. These values are in remarkable agreement (average deviation of 0.06 A) with ionic radii obtained from X-ray diffraction studies.

Acknowledgment. We thank the Robert A. Welch Foundation, Grant E-095, and the University of Houston CL, Research grant, for support of this work. Registry No. F2‘,12184-85-9; CI,, 12595-89-0;Brc, 12595-70-9;I,, 12190-71-5.

-

The Role of Triplet States in the Trans Cis Photoisomerization of Quaternary Salts of 4-Nitro-4’-azastilbene and Their Qulnollnium Analogues. 6‘ Helmut Comer* and Dietrich Schulte-Frohlinde Max-Planck-Institut fur Strahlenchemie, 0-4330 Miilheim a.d. Ruhr, West Germany (Received: January 2, 1985)

-

The quantum yields of fluorescence (4f)and of trans cis photoisomerization (&-J were measured for four quaternary and -quinolinium; X- = salts of 4-nitro-4’-azastilbene derivatives (A,+X-; A,+ = trans-l-alkyl-4-[4-nitrostyryl]pyridinium I-, C104-,and CH3S04-)as a function of solvent polarity, temperature, and concentration of quenchers. Transients (lifetime 5200 ns at 25 OC,,,A 450 nm) observed by laser flash photolysis were identified as the trans triplet configuration of either the cation (3*A,+)or the ion pair (3*A,+-.X-) in highly and less polar solvents, respectively. In polar solvents at room temperature 3*At+is quenched by oxygen, azulene, ferrocene, and I- at rates close to the diffusion-controlled limit. The quenching measurements under pulsed and steady-state conditions suggest that 3*A,+is an intermediate in the trans cis photoisomerizationroute. In contrast to the pyridinium salts, 4- does not follow for the quinolinium salts linear Stem-Volmer dependences on [azulene], [ferrocene], or [I-]. It is concluded that for the pyridinium salts trans cis photoisomerization occurs only via the triplet mechanism, whereas a second route, bypassing the lowest triplet state, contributes to a certain extent (540%) for the quinolinium salts (mixed singlet-triplet mechanism).

-

-

Introduction

The cis trans photoisomerization of neutral azastilbene derivatives has been the subject of several s t ~ d i e s . ~Introduction -~ (1) Reference 9 is part 5 in the series on cis-trans photoisomerization of stilbazolium salts. (2) Whitten, D. G.; Lee, Y . J. J . A m . Chem. SOC.1972, 94, 9142. (3) Bartocci, G.; Mazzucato, U.; Masetti, F.; Galiazzo, G. J. Phys. Chem. 1980, 84, 847. (4) Bortolus, P.;Cauzzo, G. Trans. Faraday SOC.1970, 66, 1161.

0022-3654/85/2089-4105.$01.50/0

-

of a positive charge into the stilbene-like molecule gives rise to new Phenomena.6-” Quaternary salts of 1-ak’l-4-[4-R( 5 ) For a review see: Mazzucato, U. Pure Appl. Chem. 1982, 54, 1705. (6) Glisten, H.; Schulte-Frohlinde, D. Tefrahedron Left. 1970, 3567. Schulte-Frohlinde, D.; Giisten, H. Liebigs Ann. Chem. 1971, 749, 49.

(7) Glisten, H.; Schulte-Frohlinde, D. Z . Nafurforsch.,B Anorg. Chem., Org. Chem. 1979, H E , 1556. (8) Giisten, H.; Schulte-Frohlinde, D. Chem. Eer. 1971, 104, 402. (9) Gorner, H.; Schulte-Frohlinde, D. Chem. Phys. Lett. 1983, 101, 79.

0 1985 American Chemical Society