Vibrational assignment of phenazine and phenazine-d8: Crystal spectra in polarized light and force constants calculations

spcctrochimica Acta, 1904,Vol. 20, pp. 1503to 1516. Pergamon Press Lbd. Printedin NorthernIreland

Vibrational assignment of phenazine and phenazine-d,: crystal spectra iu polarized light and force constants calculations N. NETO, F. AMBROSINO and S. CALIFANO Istituto Chimico dell’Universit8 Via Mezzocannone (Received

4, Naples,

13 January

di Napoli, Italy

1964)

Abstracts-The

infra-red crystal spectra of phenazine and phenazine-ds have been measured between 4000 and 400 cm-l. Spectra in polarized light have been obtained for two types of polycryst,alline highly oriented films, one showing the (201) crystal plane in the film plane, the other having a random axial orientation around the 6 crystal axis. A zero order force con&ant calculation was made for the infrared active planar blocks B,, and B,, using a modified Urey-Bradley field, transferred from anthracene and pyrazine. The vibrational assignment of the infrared active fundamentals is discussed on the basis of the dichroic behaviour of the observed bands and of the predict)ion of the force field. The crystal field splitt,ing of several bands has been studied in further det,ails.

RECENTLY the infra-red spectra of several substituted phenazines, including phenazine itself and phenazine-d,, have been reported by STAMMER and TAURINS [ 11. From the analysis of these spectra they have been able to furnish a list of characteristic phenazine ring bands. This has prompted us to report on a more complete and detailed vibrational analysis of the infrared active modes of phenazine and phenazine-d, done as part of a research project on the vibrational spectra of cyclic molecules. The evidence for this assignment has been found through a force constant calculation based on a modified Urey-Bradley field and through the analysis of the crystal spectra in polarized light of the two isotopic species. Both methods have been extensively used in our laboratory and can be confidently trusted as the most powerful techniques for the interpretation of the spectra of large cyclic molecules. EXPERIMENTAL

The infra-red spectra were recorded by means of a Beckman IR-9 spectrograph bet,ween 400 and 4000 cm-l. Polarization measurements were made on polycrystalline oriented films, obtained by slow crystallization of a melted film between KBr windows on a metallic bar with a temperature gradient. A region of full orientation of 3 mm radius was selected in the film and the spectrum was measured on it with the aid of the Beckman beam condenser-polarizer unit. Two different types of film orientation were observed. The first one, called A in the text, gives rise to the spectra of Figs. 3 and 5 and is the one most commonly obtained. 111 C. STAMMER and A. TAURINS, Spectrochim.

Acta lg,

1503

1625 (1963).

1.504

N.

NETO. F. AMBROSINO and S. CALIFANO

The identification of the crystal axes in the film plane has been made through an X-ray analysis of the film. The result was that the (2Oi) crystal plane lies in the film plane, i.e. it contains the axes b and a + 2~. The (2Oi) crystal plane is one of the planes of highest electron density and occurs naturally in a-phenazine crystals [S]. The b axis is parallel to the direction of crystal growth. The second type of film, called B in the text, gives rise to the spectra of Figs. 4 and 6 and is obtained in very few cases, generally through a quick cooling of the sample. B films are not well oriented as A films ; the X-ray analysis shows that the 6 axes lies in the film plane, but that there is a random axial orientation in the plane perpendicular to b. Phenazine-d, was prepared following the same procedure used for anthracene-d,, [2] and pyrene-d,, [3], namely by means of several exchanges between phenazine and calcium deuteroxide in a Carius tube at about 300°C. The temperature was set at this relatively low value in order to realize a compromise between an appreciable rate of exchange and a small decomposition of the substance. The final product was purified through several sublimations under vacuum. Isotopic purity was controlled at each exchange from the intensity of the CH and CD stretching bands and in the final test was better than 98 per cent. Happily, our spectrum is coincident with that reported by Stammer and Taurins on a sample made by the Merck Sharp and Dohme of Canada. FORCE CONSTANT CALCULATION The power of frequency prediction using the force constants found for related molecules has been emphasized by several authors. For instance, the force constants obtained for benzene have been used to predict the fundamental frequencies of several ring molecules. Both the valence force field [4] and the Urey-Bradley field [5-71 have been successfully used in this respect. Since in our laboratory we have recently obtained the modified Urey-Bradley force constants for anthracene [6] and pyrazine [7], we decided to use the same type of force field, for the calculation of the in-plane vibrations of phenazine. Thus we have transferred from the anthracene field the force constants connected with the two side rings and from pyrazine those connected with the central ring. The external NCC bending and repulsion force constants have been assumed equal to the corresponding force constants of anthracene. The Kekule force constant [5] has been averaged between the two values found for anthracene and pyrazine. The numerical values of the force constants are shown in Table 1, together with the calculated frequencies for the B,, and B,, species. The results for the infrared inactive planar species are not reported due to the lack of Raman data on phenazine. The calculations were performed using a program written for a Bendix G-26 [2] S. CALIFANO,J. Chem.Phys. 36, 903 (1962). [3] 8. CALIFANOand G. ABBONDANZA, J. Chem. Phys. 39, 1016 (1963). D. B. SCULLY [4] D. E. FREEMANand I. G. Ross, Spectrochim. Acta 16, 1393 (1960). D. H. WIFFEN,Spectrochinz. Acta 16,1409 (1960). [5] J. SHEERER, J. Chem. Phys. 36, 3308 (1962). [SJ S. CALIFANO,Ricerca Sci. 3, 461 (1963). 171 M. SCROCCO, C. DI LAVRO and S. CALIFAYO,Spectrochim. Acta (in press).

and

Vibrational

assignment

Fig. Table

1. Calculated

of phenazine

1. Internal

coordinates.

frequencies of infra-red of phenazine

act,ive in plane vibrations

B,,‘

~~~ CYAN,%

1505

and phenazine-rl,

%I CAi,D,

3058 3053 1624 1478 1345 1297 1119 928 665 246

C,,N,H,

C,,S,D,

3056 3048 1514 1435 1421 1260 1150 1005 865 625

2253 2243 1484 1430 1319 1221 874 827 760 609

2261 2251 1602 1421 1333 1051 879 815 636 230

Kzc = 3.95

K oH = 4.7

Klcc = 5.60

KV.Z cc = 4.36

K CN = 5.7

H&

= 0.65

H&,

= 0.35

Hxcc = 0.70

H (.NC = o.i.5

H[,,

= 0.68

H Hcc = 0.35

F& = 0.63

F NC = 0.65

F Cc = 0.70

Ftc = 0.53

F& = 0.55 F CH = 0.32

F,,,,

= -0.05

bending force constants in lo-”

high speed computer. already described [6].

F,,,,

= -0.05

p = 0.24

erg/r&d2others in mdyn/A.

The program, coded in SNAP machine language, has been The nomenclature of internal coordinates is shown in Fig.1. CRYSTAL

SPECTRA

a-phenazine crystallizes in the monoclinic system, space group P2,,, (CZh5)with two molecules in the unit cell [S], each molecule being located on a Ci site symmetry. The selection rules for the isolated molecule, as well as for the unit cell are given in the correlation diagram of Table 2. Table 2 shows that each normal, mode of the free molecule is split in the crystal into two unit cell modes and that the only gas-inactive modes which gain infra-red activity in the crystal are those of A,, species. From the atomic coordinates, given

1506

N. NETO,

F. AMBROSINO and

S. CALIFANO

by HERBSTEIN and M. J. SCHMIDT [S], it is easily predicted, in the oriented gas approximation, the polarization character of the infra-red active modes, for light incident on any crystal plane. Of course, these predictions have a direct meaning for monoclinic cells only if the b crystal axis is contained in the plane, since the b axis is the only symmetry axis and coincides with one of the principal axes of the dielectric tensor. Happily this occurs in our case for both types of films, thus ensuring a constant orientation of the electric vector when the polarized light travels in the crystal. Table 3 shows the proportionality factors for band intensity when linearly polarized light hits the (2Oi) crystal plane, with the electric vector oriented parallel Table ~____

Molecular

2. Select’ion rules for a-phenazine

C’i

D,ll

No. 11 10 4 5 5 r III 10

IR

R

A,

ia

%,

ia

BPS H 38 Au B lu B 2u B 3u

ia ia ia

pdpdp--,-SO dpia-ia---30 iaia-

Species

Factor

Site group

group

MZ MY MX Table

Film

A

No.

c

Species

IR

R

NO.

*

ia

M,x,,,,,

IR

R

A,

ia

P

B,

A,

ia Mb

dP ia

B,,

M,,

ia

P’-30 -30 dp‘-30

3. Proportionality factors for band inknsity the (201) crvstal alanc Plane

B 1U

B 21L

(201)

0.47 0.51

0.04 0.05

2h

Species

30 -4,

group

B 3u 0.49 0.44

in Axis h n 4 2c

or perpendicular to the direction of the b axis. These values are simply the squares of the projection of unit vectors oriented along the molecular reference system axes, on the b crystal axis or on a direction perpendicular to b in the (20i) crystal plane. In what follows band polarization shall be indicated by the symbol () ( ) parallel or (1) perpendicular, which are always referred to the b crystal axis. The most remarkable difference between A and B film spectra is due to the occurrence of several strong polarized bands in B spectra, which are almost absent in A spectra. Actually the molecular orientation in the crystals of a-phenazine is such that the short molecular axis is oriented almost perpendicular to the (2Oi) crystal plane. This is shown schematically in Fig. 2. Thus B,, vibrations, which originate a dipole variation in this direction are almost absent in A spectra obtained with polarized light incident on the (20i) plane. In B films, due to the random axial orientation in the plane perpendicular to the 6 axis, the B,, transitory moment will have a non zero component in the direction perpendicular to b, thus giving rise to light absorption when the electric vector [X] F. H. HERBSTEIN

and M. J. SCHMIDT,

ActaCryst. 8, 399 (1955);

8, 406 (1955).

Vibrational

assignment

of phenazine

and phenazine-ds

1507

oscillates in this direction. In using polarization measurements for the assignment of the observed bands to the proper symmetry species, it is necessary to realize the limitations imposed by the oriented gas model. In principle this model is valid only for the ideal case of non-interacting molecules at large intermolecular distances. In a real crystal, we expect therefore two types of deviations from the predictions of the oriented gas model: (a) band splitting due to the coupling of modes in the unit cell (factor group splitting) ; (b) changes in band intensity due to the crystal field and to intermode mixing.

IN

Fig.

2. Schematic

drawing

of the molecular

orientat,ion in tho (201) crystal plane.

The occurrence of factor group splittings is not a serious limitation to the use of the oriented gas model, inasmuch as the vibration assignment is concerned. In practical cases of molecular crystals, the factor group splitting is appreciable only for few bands, if the highest resolution of a grating spectrograph is not used. Even in such cases, the use of the mixed crystal technique often permits to localize the components of the splitting and to obtain the unperturbed frequency value. The second type of perturbation: i.e. the intensity variation due to the crystal field is instead a severe drawback to the oriented gas model. In fact the occurrence of speci$c intermolecular interactions will cause a different intensity variation for each component of the splitting. In the language of the oriented gas model this corresponds to say that the crystal field induced transitory dipole is not in the direction of the transitory dipole associated with each normal coordinate. The greater is the angle between the induced and original transitory moment, the more the experimental dichroism will deviate from the prediction of Table 3. In the case of a-phenazine band splitting has been detected for several cases which have been studied in details with the mixed crystal technique. The mixed crystal spectra are shown in Fig. 7 and will be discussed in the next section. The observed dichroism fulfills qualitatively the predictions of Table 3 for almost

1508

N. NETO, I?. AMBROSINOand S. CALIFANO

all bands. Only in three cases, namely for the bands ‘at 958, 858 and 478 cm-l a strong parallel polarization, unpredicted from Table 3, has been observed. The last two bands are however absent in the solution spectrum and therefore are reasonably assigned to the A, species. These bands and the corresponding bands of phenazine-d, which occur at 753, 628 and 438 cm-l, will be discussed in the next section. Thus, since only one band over about thirty shows a type of polarization which is incoherent with the prediction of the oriented gas model, we conclude that, the predictions of the oriented gas model can be confidently used in this case for the purpose of establishing a vibrational assignment, specially if supported by a detailed analysis of the band splittings. VIBRATIONAL ASSIGNMENT

BzTcSpecies As stated in the preceding section fundamentals of this species should be very weak or even unobservable in A spectra while strong and completely (1) polarized in B spectra. The (1) bands at 3043 and 3015 cm-l of phenazine-d, and those at 2286 and 2265 cm-l of phenazine-d, are clearly the two CH stretching modes of this species. Three ring modes are predicted from Table 1 at 1624, 1478 and 1345 cm-l for &N,H, and at 1602, 1421 and 1333 cm-l for C,,N,D,. The assignment of the strong bands at 1631, 1476 and 1598, 1418 cm-l respectively is therefore unquestionable. In the 1300-1400 cm-i region only a weak (I) band occurs at 1307 cm-’ in the spectrum of phenazine and has been taken as fundamental. The corresponding band of phenazine-d, has not been detected. Three other fundamentals, associated with ring deformation are predicted at 928, 665 and 246 cm-l and at 815, 636 and 230 cm-l respectively. The two strong (1) bands of phenazine at 904 and 658 cm-l and the band of phenazine-d, at 820 cm-l are thus easily assigned. The last mode of phenazine-d, has been assigned to a band at 626 cm-i on the following ground. A medium intensity doublet with a (II) component at 628 and a (I) component at 626 cm-l occurs in the B spectrum (Fig. 5), while in the A spectrum (Fig. 3) the (1) component almost disappears. Therefore we assume that a BzlLfundamental at 626 is overlapped by another band at 628 cm-l of a different symmetry species. A doublet occurs in B spectra of phenazine-d, with a strong (1) component at 1149 and a weak (II) component at 1140 cm-l. In A spectra the (I) component is strongly reduced in intensity as expected for B,,, bands. Figure 7 shows the spectrum of a mixed crystal with a molecular ratio of about one molecule of C,,N,H, to The doublet structure collapses, giving rise to a single ten molecules of C,,N,D,. band at 1146 cm-l. This clearly shows the occurrence of a factor group splitting of a B2,&fundamental. Owing to the prediction of Table 1 of a CH deformation mode at 1119 cm-l, there are few doubts in the assignment of this band. The same behaviour is observed in the case of the 895-889 cm-l doublet of phenazine-d,, close to the The second CH deformation mode is predicted at calculated value of 879 cm-i. 1297 and 1050 cm-l respectively for C,,N,H, and &N2D8. Thus the strong (1)

Vibrational

assignment of phenazine and phenazine-ds

1509

Table 4. Infra-red spectra of C,,N,H, Solution

Intensity

Y

A

E(M) to b

3085 3062 3047 3016

26 36 30 8

1955 1933 1914 1847 1837 1817 1775 1712 1730 1709 1628

10

1559 1522 1516 1478 1473 1444 1433 1366 1362 1328 1279 1268 1239 1232 1211 1153 1138 1113

6 45 73 24 3 3 36 42 41 10 5 2 3 2 12 3 60 75

1073 1066 1027

10 7 1

996

32

957 904

821

750 655 596

10 8 7 2 8 10 5 3 6

17 13

84

100 35 65

B film

film WI)

to b

E(rl) to b

3119 3090 3061 3043 3015 2983 2896 2831 2761 1970 1942 1918 1849

0

0

0

21 41 5 3 0 0 0 2 G 2 4 1

9 23 19 24 0 0 0 3 20 3 10 .5

15 31 !

1822 1778 1758 1733 li12 1636 1631 1554 1522 1516 1476 1471

4 0 A 7 5 7 5 50 85 0 0

1431 1362 1325 1307 1285 1247 1234 1214 1210 1148 1140 1116 1110 1077 1069 10.39 1004 998 992 958 904 863 858 820 754 753 750 744 658 595 478

3

4 6 0 9 10 13 15 12 50 93 6 0

0 0 0 2 6 0 6 8

10 10 58 44 48 10 5 3 3 36 8 7 3

0

20

5 0 9 9

6 5 50 65 0

n

50

73

60

52

70

11 0 0 0 4 15

li 0 0 0 8

76 16 0 0 a 5 “8

40 22

8 1” 0

78 10 15 0 50

0 0

11 5 70

42 78 .1

49

0

;

36

0

66 81

90

100

90 100

100

100 100 100 0 78 15

76 81 24 6 7 4 9 50 76

47 23 65 0 0 50 81

‘>

2 10 13 45 48 12 50 70 70 64

28

33 78 5 10 0 30

WI)tob

12 90 0

100 100 0 78 15

60 90 0

1510

N.

2400 Fig.

3. Infra-red vector

NETO,

2000

AMBROSINO

1600

spectrum parallel

F.

Table

.

.

A film in polarized

. electric

5. Infra-red

vector

spectra

Solution

A

2930

“858

1317 2305 z95 1284 2657

1625

1.; 8

~ -Pmelectric

to b.

of C,,N,D,

film

B film E(l)

to b

E(Il) to h

E(i)

to 6

2960

6

6

6

6

2930

15

15

15

16

2860

7

7

7

7

2790

5

5

5

5

2773

3

3

3

3

2664

8

8

8

8

2575

6

6

6

6

2558

4

4

6

6

3

6

8

13

11

i

2340

3

2317

5

2707 .

14

;

30

28

20

2293

10

0

“3

33

17

2286

10

0

23

33

5

2265

0

30

3

64

2193

2

2

3

3

2144

2

2

4

4

4

4

1:

5

2082

1

1950

0

0

1860

2

2

72

3

3 5

1790

1

P

4

1670

2

4

4

IO

1626

12

14

18

30

1.595

4

3

10

“0

1569 1490

light,.

perpendicular

Crystal

E(ll) to b 1960

S. CALIFANO

1.

1GOO

of phenazine-d, to b;

and

5

10

1556

2

5

1491

SO

84

1472

5

17

30

‘1

18

80

84

5

10

1417

10

1418

0

0

0

40

1407

8

1405

0

0

0

36

1396

3

1387 1381

94

1385

90

93

90

93

64

1361

2

:!

3

3

1342

3

1342

5

7

10

10

1312

4

1314

3

3

7

7

10

Vibrational assignment of phenazine and phenazine-d8

1511

Table 5 (continued) Crystal

Solution

Intensity Intensity

W to b

B film

<4 film

J-VI) to b

E(L) to b

1290

5

1310

6

6

10

10

1280

6

1283

10

12

21

21

1258

28

1205

2

1174 1164

1277

1%

14

25

24

1255

79

80

79

80

12.36

3

5

3

3

1201

7

7

10

10

4

1166

77

79

80

80

70

1150

3

4

10

10

1139

2

4

7

9

1129

2

2

4

i

1098

0

0

0

4

2

1130 1070

2

1072

3

5

6

8

1047

16

1047

0

0

3

40 17

1038

0

0

3

1024

0

6

2

15

12

15

18

21

963

5

952

925

2

935

4

5

7

6

910

5

922

13

5

34

10

904

3

3

15

892

2

828

6

895

5

6 30

23

890

15

867

2

2

4

6

854

3

30

6

41

830

14

15

30

48

820

30

52

817

30

817

15

771

25

773

81

0

78

769

75

0

76

5

10

33 10 4

757

6

759 757

7

10

27

743

70

744

83

85

78

80

715

0

0

4

669

2

2

0 r

8

660

643

3

628

18

662

10

7

642

2

2

628

70

609

30

602

30

2; 5

3 15 2

70 5

626

50

610

33

35

35

100

603

100

97

100

97

30

574

78

80

78

80

530

1

530

0

0

5

3

451

3

451

0

10

0

10

438

6

0

6

0

576

band of C,,N,D, at 1047 cm-l is easily assigned. have chosen the weak (I) band at 1285 cm-l B,,

E(IJ)to b

In the spectrum

30

of C1,N,H,

we

species

Table 3 shows that fundamental of this species should be slightly (I) polarized in A films. In the CH stretching region no bands are found with the correct type of polarization. On the basis of the band moment hypothesis, at least one of them is expected to be very weak. To the second we have tentatively assigned the bands at 3061 and 2303 cm-l

respectively,

although

no conclusive

experimental

evidence

has been found.

1.512

N.

NETO, F. AMBROSINO and S. CALIFANO

On the basis of their polarization and intensity the bands of phenazine-d, at 1516, 143i,1362,1210,820 and 595 cm-l and those of phenazine-d, at 1491, 1385, 1255, 1166, 744 and 574 cm-1 are undoubtedly B,, fundamentals. The agreement with the calculated values of Table 1 is very satisfactory and, as in the B,, species, leads to an univocal assignment. The CH bending modes are expected from Table 1 at 1150 and 1005 cm-i for phenazine-d, and at 8'74and 827 cm-r for phenazine-d,. The crystal spectrum of phenazine-d, shows at about 1000cm-l a triplet structure with two (II) components

i 3200 Fig.

’

I 2600

i

I 2400 I I I 2000

I

I,

1600

I / 1400 I,

tctoo

;

izoo

/

?8

I 3

800

I

1

EOO

/

cm4

IO

4. I&m-red spectrum of phenazino-d- B film in polarized light. ---dectric vector parallel to b; . . . . . electric vector perpendicular to b.

at 1004and 992 cm-l and a (1) component at 997 cm-r. Again the spectrum of the mixed crystal of phenazine and phenazine-~~ shown in Fig. 7 clearly demonstrates that the 1004 and 997 cm-l bands are the components of the factor group splitting of a B,, fundamental. In the mixed crystal spectrum both bands disappear, giving rise to a single band at 1001 cm-i, which has been taken as the unperturbed value of this fundamental. As counterpart in phenazine-d, we have chosen the weak band at 829 cm-l. The polarization character of this band is clearly seen only in the A spe~trunl (see Fig. 3), being in the B spectrum strongly overlapped with the (i) band at 820 cm-l. The second CH deformation mode has not been identified in the spectrum, since no bands with B,, polarization occur in the 1200-1000cm-l region. B,, and A, species Fundamentals of this species should be slightly (II) polarized in A spectra, as shown in Table 3. The only band which can be identified without difficulty from the observed polarization is the very strong CH out of plane deformation mode at 751 cm-l in This band, which appears as single phenazine-d, and at 600 cm-l in phenazine-d,.

Vibrational in the

solution

spectrum,

assignment is split

1513

of phenazine and phenazine-cl,

in the

753 and 744 cm--l and two perpendicular

crystal

into

four

components,

two

(II) at

at 755 and 750 cm-l.

The mixed crystal spectrum is of no help in this region due to the overlap with It is therefore difficult to specify the nature of the complex strong bands of C,,N,D,.

2600 Fig.

3200

2400

2000

1600

lGO0

1400

:’

1

1200

1000

600

Ii

GO0

cm-'400

5. Infra-red spectrum of phenazine-d, A film in polarized light. --electric vector parallel to b; . . . . .electric vector perpendicular to b.

2800

2400

2000

1600

1GOO

1400

1200

1000

600

cm-1 600

4

Fig. 6. Infra-red spectrum of phenazine-d, B film in polarized light. --electric vector parallel to b; . . . . . electric vector perpendicular to b. pattern expected

Since only two components of the factor group splitting are observed. (see Table 2), the most plausible assignment is to select the bands at 744

and 750 cm-l as the two components of the splitting of another band, probably a combination band with a lattice mode made strong in the crystal by intensity “borrowing” from the B,,fundamental.

Three completely (II) polarized bands appear in the crystal spectrum of u-phenazine at 957, 858 and 478 cm-l. At least two of them, namely those a’t 957 ancl

N. NETO, F. AMBROSINO and S. CALIFANO

1514

478 cm-l, would reasonably fit the assignment to B,, fundamentals, by comparison with the assignments of naphthalene [4] and anthracene [2]. However their polarization does not meet the requirements of Table 3, since we expected B,, modes only slightly (II) polarized. The possibility that such a polarization arises from the crystal field effect or from intermode mixing seems rather remote since it is hard to explain

r\ I

\

\ \ \

l&l

111

Fig. 7. Splitt.iny

of some phenazine bands; (a) phenazin? crystal of phrnazine and phenazine-d8.

cryst)al;

(b)

mixed

how these effects could completely annihilate the intensity of the perpendicular component of the BltL transition dipole. Furthermore the bands at 478 and 858 cm-l are completely absent in the solution spectrum, a fact that strongly suggest their assignment to the A, species. Two bands showing the same dichroic behaviour appear in the spectrum of phenazine-d, at 628 and 438 cm-l, and have been assigned to the corresponding modes. The 957 cn-i band however is, although weak, definitely present in the solution spectrum. Although a rather sophisticated and involved mechanism should be invoked to explain its presence in solution, the assignment to the A,, species seems

Vibrational

assignment of phenazine and phenazine-ds

1515

the only possible. The counterpart for this band in the spectrum of Ci,N,D, occurs as strong (II) band at 753 cm-l. A rather strong band with the type of polarization expected for B,, vibrations occurs instead in the spectrum of C,,N,H, at 1110-1116 cm-l. The factor group splitting of this band has been investigated with the mixed crystal technique, the result being shown in Fig. 7. This band is very unlikely to be a combination band, being one of the most intense in the spectrum, and therefore has been taken as B,, fundamental although its frequency is about 150 cm-r higher than that of the corresponding mode of Table 6. Fundamental Type

of vibratiolr

frequencies of phenazine and phenazine-ds C1,N,H,

C,,N,D,

Sprcio

C,,N,I-I,

C,zN,%

CH

bendirlg

out

of pl.

(1110)

(922)

(957)

(753)

CH

bending

out of pl.

751

600

858

628

478

438

Ring

deformation

Ring

deformation

King

deformation

f$XCiVS

B I?‘

AU

CH

stretchirlg

3043

2286

3056’

CH

stretchirlg

3015

22G5

(3061)

2303

2253*

Ring

stretching

1631

1,598

1516

1491

Rirlg

stretrhing

1476

1418

1431

1385

Ring

&etching

Ring

stretching

B 2,‘

1307 904

820

B 3u

1362

1255

1210

1166

CH

bend&

in pl.

1149

895

CH

bendirlg

in pl.

1285

1047

1001

658

626

820

240*

230*

Ring

deformation

Kin,q deformation

1150*

874* (829) 744 574

anthracene [6] and naphthalene [4]. As counterpart in the spectrum of C,,N,D, have taken the band at 922 cm-l which shows the same dichroic behaviour.

we

CONCLUSION

The assignment discussed in the preceding section is collected in Table 6. Apart from very few cases (frequencies in brackets) definite evidence has been found for the assignment of the fundamentals through the calculations of Table 1 and the application of the oriented gas model. The similarity with the assignment of anthracene is striking ; for both the B,, and B,, species corresponding fundamentals for the two molecules differ of less than fifty wave numbers with the only exception of the 1210 cm-l fundamental of phenazine which lies at a considerably lower frequency than the corresponding mode of anthracene (1346 cm-l). Unfortunately no Raman data are available for phenazine and therefore no attempt has been made to account for the combination bands which appear in the spectrum. We wish to stress again that the oriented gas model gives a general picture which is in good qualitative agreement with the observed dichroic ratios. However the spectrum of phenazine shows definitely the presence of specific intermolecular interactions leading to the factor group splitting of some bands and to slight variation in the’ dichroism of bands of the same species. Two models have been proposed to account for these intermolecular interaction ;

1516

N. NETO, F. AMBRO~INO and S. CALIFANO

dipole-dipole coupling [9] and hydrogen atom repulsion [lo]. We have observed that in the phenazine spectrum the bands which show the highest splittings are those associated essentially with hydrogen motions and this seems to indicate that the atom-atom repulsion mechanism is probably responsible for a major fraction of the observed splittings. Aclcrcowledgement-The authors wish to express their deepest thanks to Dr. P. GANIS and to Dr. G. DIANA for their help in the analysis of the X-ray spectra. This Research was made possible, through the support and sponsorship of the U.S. Department of the Army under contract No. DA-91-591-EUC-2865. [9] R. M. HEXTER, J. Chem. Phys. 33, 1833 (1960). [lo] D. A. Dows, J. Chem. Phys. 32, 1342 (1960).