Far-infrared intensity and normal coordinate studies on pyridine-halogen complexes

Journal of Molecular

Structure

ElsevierPublishing Company,Amsterdam.Printedin the Netherlands

FAR-INFRARED INTENSITY AND NORMAL ON PYRIDINE-HALOGEN COMPLEXES

G.

W.

BROWNSON

Chemistry

AND

Department,

COORDINATE

,147

STUDIES

J. YARWOOD

University

of Durham,

Durham

City (England)

(ReceivedJuly13th, 1971)

ABSTRACT

Absolute integrated intensity data have been determined for the two low frequency bands due to v(D-I) and v&X) in pyridine-IX (X = I,CI,Br) complexes. Along with the normal coordinates, calculated using a linear triatomic model, these data have been used to calculate dipole derivatives aF/aRj values. Pyridine-&-IBr spectra have been used to estimate possible values of the interaction force constant k,,. The dipole moment change $/a&,_, calculated for the pyridine-I2 complex using a simple model is considerably lower than that observed, implying that “charge-transfer” effects contribute significantly to the band intensities. Numerous attempts have been made strengths of the halogens and interhalogens

in the past

to compare

the acceptor

with strong donors such as pyridine. From thermodynamic (K,) measurements ‘s2 , for example, the acceptor strength decreases in the order ICI > IBr > I,. The same conclusion was drawn from studies on the frequencies of the bands3-5 arising from v(I-X) and v(D-I) stretching modes of the molecule pyridine-I-X, which occur in the far-infrared region. Complications arise, as already pointed out6-‘, using K= measurements because the equilibrium constant includes entropy contributions, introduced, for example, by steric effects (reliable AW values being unavailable to date). The dangers of using relative frequency shifts have also been discussed6-’ and it is clear that changes inforce constant must be compared if one is to take account of vibrational mixing, especially of the two low frequency stretching modes (of species A,, if the complex molecule has C,, sym metry). Such mixing may cause both frequency and intensity changes of differing amounts in going from one compIex to another and it is desirable to compare transition moments, iJji/aR if possible. Hitherto a linear triatomic model has been used6-8 to estimate the normal coordinates and transition moments relating to the two low frequency modes v(I-X) and v(D-I). This approximation should be a reasonable one in view of the J. Mol. Structure. 10 (1971) 147-153

148 TABLE

G. W. BROWNSON,

J. YARWOOD

1

FREQUENCY,

INTENSITY,

Complex

Pyridine-ICI (benzene)

NORMAL

COORDINATFi

AND

TRANSITION

Y(I-X) (cm-‘)

v(D-I)” (cm-‘)

B (I-X)b

292

140

11900*200

MOhtENT

DATA

FOR PYRIDINE--IX

k (rz!y/tA-1)

B (D-I)b

3700*200

COMPLEXEZ

(z$&‘]

0.0 0.1 0.2 0.3 0.4

1.36 1.41 1.45 1.48 1.51

OS

1.52

0.6

1.52

0.1 0.6 Pyridine-IBr (benzene)

206 (205)’

133-35 (127)’

6500+100

2800*100

0.0 0.1 0.2 0.3 0.4

0.5 0.6

1.00 1.18 1.28 1.36 1.40 1.43 1.43

0.1 0.6 Pyridine-I= (cyclohexane)

183

93

25OOf

Pyridine (cyclohexane)

1830

94’

29300

= All values are f 1 cm-‘. b Intensities in “darks” (cm- r cm* mmole-I). Beer-Lambert law plots. E Units are lo-‘= gt. d UnitsareDA-‘. c “Ccu-rected” values (see text)_ r Values for CsD=N-IBr. = Data of Lake and Thompson (ref. 10).

J. Mol. Structure, 10 (1971) 147-153

40

195Oi

2790s

60

0.0 0.1 0.2 0.3 0.4 0.5 0.6

1.13 1.27 1.37 1.45 1.50 1.54 1.55

0.1 0.2 0.5 0.6

1.27 1.37 1.54 I .55

Errors are standard errors from a least squares fit of the

NORMAL

0.63 0.62 0.62 0.64 0.68 0.73

6-57 6.74 6.87 6.97 7.04 7.08

0.80

7.10

0.73 0.62 0.60 0.62 0.63 0.68 0.76

6-75 8.16 8.87 9.31 9.61 9.76 9.75

COORDINATES

-0.88 -0.22 3-0.41 1.02 1.62 2.21 2.82

-3.92 -1.84 -0.39 +0.79 2.03 3.17 4.35

OF PYRIDINE-HALOGEN

2.68 2.22 1.76 1.31 0.85 0.40 -0.08

9.36

9.40 9.39 9.34 9.25 9.13 8.96

7.08

8.93

5.40 4.13 3.03 1.80 0.62 -0.66

9.57 9.74 9.72 9.53 9.21 8.72

6.61 7.27 7.67 8.12 8.62 8.92 9.30 4.80” (4.7)C 6.62E t (6.8)” 2.02 4.44 5.92 7.04 8.12 9.05 9.80 ( ::i:: 9.17= (9.10)

0.34 0.31 0.31 0.33 0.36 0.41 0.48 0.32 0.32 0.42 0.48

I

149

COMPLEXES

-7.93

-6.93 -6.28 - 5.45 -4.64 -3.75 -2.73 -6.80” -3.97” -9.40

-7.80 -6.52 -5.37 -4.07 -2.76 -1.31 -7.60 -1.65

9.12 9.99 10.51 1090 Ii.19 11.37 11.41

-0.61 +o.so 1.52 2.51 3.45 4.43

6.87 5.62 4.49 3.38 2.26 1.15 -0.07

9.82 9.94 9.99 9.89 9.69 9.39 8.98

1.83 3.00 3.91 4.72 5.49 6.18 6.84

-6.31 -5.60 -4.97 -4.31 -3.60 -2.89 -2.09

9.92 10.49 il.34 11.42

-0.66 +0.47 3.46 4.43

5.67 4.52 1.14 -0.07

9.98

10.00 9.41 8.96

2.84 3.93 6.63 7.41

-6.70 -6.00 -3.71 -2.79

- 1.9I

relatively high frequencies of the pyridine, A,, vibrations which are changed very little6 on going to the complex. There has, however, been a major problem in the past in that ~~~z~~e~va~u~sof the interaction constant k, 36--8 have had to be used. This problem has now been avoided by the use of pyridine-d, to obtain extra frequency data without changing the donor properties. We have calculated the force constants fi2 and f2s for a range of interaction constants and plotted the usual ellipses’ to find the mathematical solution for these two constants. Fig. 1 shows these elhpses for pyridine-TBr and pyridine-d,-IBr in the region where they J. Mol. Sfrucfure,

10 (1971)

147-153

150

G. W.

BROWNSON,

J. YARWOOD

0.6 -

o-7 0.6 -

0.5 %3 0.4-

0.3 020-1 -

0.0 0.9

1.0

1.1

k,Fig. 1. Force

constant

ellipses

12 er

for pyridine-IBr

1.3

1.4

(a) and pyridine-d,-IBr

1.5 (b)

using a linear tri-

atomic model.

cross. The two crossing points correspond to k,3 = 0.1-0.2 mdyn A-‘* and k, 3= OS-O.6 mdyn A-‘, so it is necessary to choose the correct solution before proceeding to calculate the transition moment

this interaction

constant

data. In view of previous estimates of

(i.e. about 0.4 mdyn A-1)8*g

we wouId expect the 0.6

mdyn A- ’ value to be more acceptable but we have obtained the normal coordinate and the transition moments for the whole range of k,3 values to see how they vary. The results are shown for the three complexes in Table 1. It should

matrix (L-l)

be noted that we have not observed sufficient frequency shift on going to the pyridine-d,

complexes

of iodine and iodine monochloride

for a unique solution to be

obtained so we have assumed the same values of k13 for these complexes. There is, of course, another set of solutions offi andf2, not shown in Table 1. We have chosen the solutions for which the force constants ellipses cross. It should be emphasisedthat the choice of force constants made (a large f,_x with a small&,, for example) depends on the assignment of the original frequencies to the normal vibrations of the pyridine-I-X system. It has been usual in the past3-” to assign the band at higher wave number to the v&X) vibration and the other band to the intermolecular stretching mode, v(D-I). Since it is the lower frequency band which shows a frequency shift on isotopic substitution it seems likely that this band

corresponds to the mode which is mostly D-I stretching. This is the opposite from complexes. The two the situation found by Gayles12 for trimethylamine-halogen l

1

mdyn A-*

= 10sf N m-l in S.I. units.

J. Mol. Srructuri,

10 (1971) 147-153

NORMAL

COORDINATES

OF

PYRIDINE-HALOGEN

151

COMPLEXES

vibrations are, of course, expected to be “mixed” and the L-’ matrix elements in Table 1 show that this is the case. These normal coordinates also show that, as expected, the closer the two frequencies the more severe is such mixing. The fact that a significant frequency shift is observed only for the iodine monobromide complex can be rationalised in this way since the vibrational coupling is greatest for this complex (the two frequencies are closest in this case). Gayles12 concluded that, for the trimethylamine complexes, k, 3 negative. This is not the case for pyridine complexes and, since as the D-I bond is stretched the I-X bond is strengthened (the halogen becoming closer to the “free” molecule), we would expect k,, to be positive13. Intensity data for both vibrations have now been obtained for all three complexes. Since there was a discrepancy between our intensity data’ for pyridine-I, and those of Lake and ThompsonlO we have remeasured the intensities of the two far-infrared bands in cyclohexane over a range of pyridine concentrations. The intensity of the low frequency band is again significantly different from the previous value’ ‘. No major variations in band intensity are apparent although we are further investigating the effect” of solvent polarity. The data shown in Table 1 are average values using K, = 140 1 mole-’ (ref. 2). It should be noted that, although the indeterminant signs of the dipole derivatives (i3i$aQi), lead to two values of az/i3Ri in each case, we have chosen the sign combination of the (a~laQi), parameters which leads to the two transition moments having opposite signs’. In order to compare directly the values of a~/dRr-x for the three complexes it is necessary to subtract the contributions made by the “free” halogen in an “inert” solvent. We have recently remeasured the intensities of the v&Cl) and v(I-Br) bands for these interhalogens in carbon tetrachloride and heptane. The “&atomic” contributions* a;/& are 2.5 D A-l and 0.7 D A-’ respectively*. On subtracting these values from the computed a,u/aR,-x values we get the “corrected” values shown in Table 1. The alternative is to subtract the values of (a,$@,),

for

the l‘free” halogen before computing 8~/aR,_X and dj?/dRD_1 for the complex. The two calculations should be roughly equivalent and do indeed lead to similar values corrected of a%/i3RI_x for both complexes. It may be seen from Table 1 that in all three cases the relative absolute value of the two transition moments is reversed on going from k, 3 = 0.1 to 0.6 mdyn A-‘. At k,, = 0.6 the a~/aR,_, value is the greatest but at k,, = 0.1 the aG/aRD_, value is now the larger. Although, at first sight, we may except i3ji/dR,_x > &/aRD_, it is clear that if we accept the values for k13 = 0.6 then the a~/aR,_, data are inthe order 1Br > I2 > ICI while the aji/aRD+ data are in the order ICI > IBr > 12. There will, of course, be errors in these parameters, due to the use of a simplified model, but the data show that the ICI complex is the “strongest” complex using one transition moment and the “weakest” using the other. On the other hand, if * 1 D A-’

= 3.335x 1O-2o Coulombin S.I. units. J. Mol. Structure,

10 (1971) 147-153

152

G. W. BROWNSON,

J. YARWOOD

we examine the data for k,, = 0.1we find that the. order of transition moments is now ICI > IBr > I2 using aji/aR,-,_, and IBr > ICI > I2 using -aj$aR,-x. Since thermodynamically ‘s2 the iodine complex is very much weaker th& the other two this would appear to be the most reasonable choice of the two. Further, a value of kl3 in the region of 0.1-0.2 mdyn A- ’ fits better with the value of 0.22 mdyn A-’ obtained using Badgers rule14. Assuming this choice to be correct the question arises whether it is reasonable

to have $/6X,_,

> L$/aR,_x_ Since the charge redistribution due to complexation

and during vibration is not known, predicting dipole moment changes is very difficult. It seems clear from the recent NQR data 15.16that the halogen molecules in these complexes are considerably polarised and that the-extent of charge-transfer may be smaller (about 20-30 OA) than at one time thought. It seems clear that polarisation forces are also involved. Table 2 shows dipole moments induced in

the iodine molecule by pyridine, using the simplest possible fiodellg, as a function of the average distance between the molecules. Pyridine is treated as a point dipole (of 2.2 Debye) and the polarisability of the iodine is taken as 17.5 x 1O-24 cm3 (ref. 17). On this basis the observed dipole moment of the complex (4.5 D)18 corresponds to a distance of - 3.2 a and the dipole moment change, aji/aR,_,, is 2.3 D A-l_ For k13 = 0.1mdyn A- ’ this is considerably lower than the observed TABLE 2 INDUCED

DIPOLE

MOMENTS

AND

TRANSITION

MOMENTS

aji/aRD_,

FOR THE PYRIDINE-12

SYSTEM

r

T;fnd=

Total dipote nronrentb

(4

(Debe)

(Debye)

aiiiaRD-I (DA-‘)

2.0 2.5 2.6 2.9 3.0 3.1

9.64 4.92 4.36 3.16

l.l.84 7.12 6.56 5.36

5.6 3.6

2.55

3.2

2.34

3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.0

2.14 1.95 1.80 1.65 1.52 1.40 1.30 1.20

4.77 4.54 4.34 4.15 4.00 3.85 3.72 3.60 3.50 3.40

2.85

5.05

3.1 2.80 2.30 2.00 1.90 1.50 1.50 1.30 1.20 1.00 1.00

n The dipole moment induced in the iodine molecule by pyridine is

where a is the iodine polarisability and r is the average distance between the moleculeslg. b After adding.p$ridine dipole moment.

J. Mol.

Structure,

10 (1971) 147-153

NORMAL

COORDINATES

OF

PYRIDINE-HALOGEN

COMPLEXES.

153

value- However, since the D-I distance in the solid complex is 2.3 AZ0 the relevant distance for calculation of the induced moment using the simple model is probably more like 3.5-3.6 A. Here the dipole moment change is only about 1.5 D A-‘.. It would appear therefore that the effects of electron “delocalisation” during vibration are important at least in determining the infrared intensities; At present it is not, however, possible to be sure which transition moment should be the largest. We are currently working on a more sophisticated model for these complexes in. an attempt to examine the effects of mixing of these two vibrations with skeletal vibrations of the pyridine ring. Thanks are due to S.R.C. for a studentship (to G.W.B.) and for funds to purchase a far-infrared interferometer.

REFERENCES 1 A. I. POPOV AND R. H. RYGG, J. Amer. Chem. Sot., 79 (1957) 4622. 2 H. D. BIST AND W. B. PERSON,J. Phys. Cilenz., 71 (1967) 2750. 3 S. G. W. GINN AND J. L. WOOD, Trans. Faraday Sot., 62 (1966) 777. 4 I. HAQUE AND J. L. WOOD, Spectrochim. Acru, 23A (1967) 959. 5 S. G. W. GINN, I. HAQUE AND J. L. WOOD, Specrrochinz. Actu, 24A (1968) 1531. 6 J. YARWOOD, Spectrochim. Acru, 26A (1970) 2099; Trans. Faruday Sot., 65 (1969) 934. 7 Y. YAGI, W. B. PERSON AND A. I. POPOV, J. Whys. Chem., 71 (1967) 2439. 8 J. YARWOOD AND W. B. PERSON, J. Amer. Cfzem. Sot., 90 (1968) 594, 3930. 9 A. G. MAKI AND R. FORNERIS, Spectrochim. Acra, 23A (1967) 867. 10 R. F. LAKE AND H. W. THOMPSON, Proc. Roy. Sot. (London), A297 (1967) 440. 11 R. F. LAKE AND H. W. THOMPSON, S’ectrochim. Acra, 24A (1968) 1321. 12 J. N. GAYLES, J. Chem. Phys., 49 (1968) 1840. 13 I. M. MILLS, in M. DAVIES (Editor), Infra-Red Spectroscopy and Molecular Srrzzcrzzre, Elsevier, Amsterdam, 14 F. WATARI,

1963,

Chap.

Spectrochim.

5, p. 194.

Acra, 23A (1967) 1917.

15 H. CRES~ELL-FLEMING AND M. W. HANNA, J. Amer. 16 G. A. BOWMAKER AND S. HACOBIAN, Azzst. J. Chem.,

Chenz. Sot., in press. 22 (1969) 2047.

17. M. W. HANNA, J. Amer. Chem. Sot., 90 (1968) 285. 18 K. TOYODA AND W. B. PERSON, J. Amer. Chem. Sot., 88 (1966) 1629. 19 M. DAVIES, Some EIecrricuI and Opricul Aspecrs of MoIeczzkzr Behuviozzr, Pergamon, 20

1965. 0. HA~.~EL, CHR. RBMMING AND T. TUFTE,

Oxford,

Acra Clrem. Scund., 15 (1961) 967. J. Mol. Srrzzcrzzre, 10 (1971) 147-153