~oMICHIGAN STATE UNIVERSITY

Contrt r N0C014-68-A-0109-0003

-R 014 - 203

Technical Report No. 8 (33)

THE RAMAN SPECTRUM OF CRYSTALLINE LITHIUM NITRATE

by

R. E. Miller, R. R. Getty, K. L. Treuil, and G. E. Leroi

(Subff.tted for publication in the Journal of Chemical Physics)

OT

Reproduction in whnle oi- in part is permitted for any -.urpose of the United 3tatesO

uovernment. Distribution of this document is unlimited.

October, 1968

Reproduced by tfh*CLEAR N, OU SE

for Fcderal $o entf, & Technicalsnorgficn5'geld Va ;22|5"

1DO

THIS DOCUMENT IS BESTQUALITY AVAILABLE. THE COPY

FURNISHED TO DTIC CONTAINED

A SIGNIFICANT NUMBER OF

PAGES WHICH DO NOT

REPRODUCE LEGIBLY.

THE RAMAN SPECTRUM OF CRYSTALLNE LITHIUM NITRATE

by

R. E. Miller, R. R. Getty, K. L. Treuil and G. E. Leroi

Department of Chemistry, Michigan State University, East Lansing, Michigan 48823

(Received October 1968)

ABSTRACT

Polarized Raman spectri of an anhydrous single crystal of LiNO3 have been

obtained under optimum experimental conditions. The internal vibrations ofthe nitrate ion are observed as follows: v (Eg) = 735.4, V1(A19) = 1071. 0 and

V3(E ) = 1384.6 cm .these frequencies are essentially independent of the

cation in the host lattice. The two raman-active (E ) lattice modes have beeng

located and assigned to primarily translational (124. cm - 1) and librational

(237.2 cm -1) motions. No indication of coupling between these vibrations is

observed. The frequencies of the external vibrations relative to those of

NaIO 3 demonstrate that they are dependent on the volume of the crystalline

unit cell. Except for small intensity contrbutions in some orientations

which are probably due to depolarization effects in the birefringen" crystal,

the polarizability activities of both the internal and external modes follow

the expected symmetry selection rules.

INTRODUCTION

10rman scattering has been observed from LINO in the solid, 1 melt, 2

and solution, 3 and tne results discussed by several authors. - 5 However, there

have been no reported attempts to conduct a polarization study of an anhydrous

single crstal, or to locate the Raman-active external lattice frequercies of

the solid. Lithium nitrate crystallizes in the calcite structure (D space

33group) with -the nitae ions occupying L3 sites 3nd the lithium ions S 6 sites.6

Thus five Raman-active fundamental vibraticns arme expected. 7 Three are attributable

to internal vibrations of the nitrate ion: v I(A19), 3(E ), and v4(Eg); the

previous results for these vibrations in the various phases cf LUNO 3 have been

summarized by Bues. 2 The remaining Raman fundamentals are E external (lattice)g

oscillations. The normal modes of tie lattice vibrations were depicted in our8

previous work on NaNO3 ; for convenience they are reproduced in Fig. 1. Although

one of the vibrations is translational in character and the other librational,

since they have the same symmetry mixing is allowed. It would be interesting

to know whether the nature of the cation affects the degree of mixing, which8

was found to be undetectable for sodium nitrate.

The Raman frequencies for NO3 internal modes in the solid, melt,

and solution for a seri-s of alkali nitrates have been compared by Bues,' who

interpreted the small shifts in the symmetric stretching frequency v, (less

than 4%) according to a "free volume" ,model. Even smaller frequency differences

were noted for v3 and v4, and it was argued that these motions should be less

sensitive to the packing of the envirorment. However, it was also shown that

the internal frequencies in LINO3 are quite dependent on tha atount of water

present and the physical state of the saple. 2,3 If the solid samnles used in

-2-

'he early studies were not of the hi"hest quality, for example because of the

extr el - hy ':-pi - .: ture of lithium nitrate, ircorrect frequencies would have

been observe, fr the crystal. This would significantly affect the hypothesis

formulated to P:lain the small frzquency shifts in the NO3- vibrations with

change in cation.

Anomalous polarizat. )n intensity ratios for the Raman bands of crystals

with the calcite struct-re hve been Zenerally observed, and interpreted for9 8

CaCO3 and HaNO3 as being pri EL;Uy due to depolarization of the incident and/or

scattered light in the optically anisotropic crystals. Since the ratio of the

molecular refractivity, a measure of birefrigence, is almost the same for- sodium

and lithium nitrates, a careful polarization study of crystalline LiNO3 sho*ild

Provide a test for this explanation.

Finally, in an interpretation of the infrared combination spectra of

calcite and the alkali nitrates, Schroeder, !'eir, and Lippincott 11 suggested

that a torsional lattice vibration about the optic axis should exist with a-l

fu-.damental frequency of 20-33 cm . Although such 4 vibration should be Raman-

inactive by symmetry, it was reported in early studies of calcite; 1 2' 1 3 it was

not observed in later investigations of NaITO83 and CaCO3 . 9 Since measuremeits

in the low frequency region were difficult due to scattered light and grating

ghosts, it seemed worthwhile to confirm the absence of the torsional motion in

the case of LiNO3 .

> In order to determine the external lattice vibration frequencies of LiNO

and to unambiguously establish the internal nitrate ion frequencies, as well as to

study the effect of the alkali cation on these vibrations, the Raman spectrinu of a

large, anhydrous single crystal of lithium nitrate has been obtained. Using modernRaman instrumentation a complete polarization study h been made$ and the dvaiiable

spectra region extonded to within lO'cm- I of t.. ex--ting radiation.

--3"

EXPERIMENTAL

The crystal used in this investigation was obtained from the Harshaw

Cherical Company. Unfortunately, unlike NaNO 3 , a clear, homogeneous single

crystal of LiNO3 can be grown only with some difficulty, and cannot be control-

lably cleaved. The 3ample received in our laboratory was a cube approximately

5mm on a side, polished to size from a larger ingot. The faces were mutually

perpendicular within 10, and the z-axis (optic axis) was normal to one of them.

in order to facilitate orientation, the crystal was mounted in a W and W

eucentric gonianeterwhich had been placed in a small Plexiglass drybox.' This

cnsured that the sar!ple remained anhydrous; at no time was the crystal exposed

t_- the room atmosphere.

The oriented single crystal was first irradiated using a Spectra-

0Physics Model 125 He-!le laser emitting about 70 mW at 6328A. The 900 scattered

light was collected by appropriate optics, dispersed by a Spex Nodel 1400 double

rnoncchromatnr, and detected using a ther-mo-electrically cooled ITT FW-130

photorultiplier (S-20 response), The photomultiplier output signal was -znplified

with a Victoreen VTE-l micro-microammeter, and displayed on a Moseley 7100B

strip chart recorder.1 4 The laser power was adequate to discern the general

features and intensity patterns of the Paman spectrum. More detailed studies

347 each band were made using a S;ectra-Pbysics Model 140 argon ion laser emitting

0 0about 600 mW at 480A and 700 mW at 5145A as the excitation source.

In general, the spectra were obtained with the incident beam focused

within the sample using a solid angle of less than 20, and with the scattered

light gathered using f/2 optics. Depending on the band being studied and the

-1polarization conditions, spectral slit widths i'n the range 0.2 - 2 cm were

-4-l

employed. The accuracy of the repcrted frequencies is within 0.5 cm-. All

measurements were made at room temnerature.

RESULTS AID DISCUSSION

As indicated in the third column of Table I, for a rhombohedral crystal

with D3d symmetry only four of the polarizability tensor derivatives are unique.

According to the group theoretical analysis a and a are always related duexx yy

to symmetry, as are aXz and ayz Representative Raman spectra of crystalline

LiNO3 in the xx-, zz-, xy-, and xz-or'lentations are shown in Fig. 2. The polar-

ization is specified by the commonly accepted notation! 5 of tite t)e x(yz)y,

where the initial -. d final letters denote the direction of the incoming and

oDserved light in the crystal coordinate rystem, and the letters in parenthesis

indicate the polarization axes of the incident and scattered radiation. Thus in

this example a measure of the a -component of the polarizability tenp.oryz

derivative would be obtained.

The observed frequencies, assignments, and relative intensities as a

function of polarization are summarized in Table I. Theoretical consideratioc

of the relative intensities have been developed by Bhagavantum1 6 and Saksena. 17

Like the other crystals in its class, lithium nitrate is highly birefringent,

and therefe anvmlous depolarization effects, such as the presence of a give7%

vnBration in a polarized spectrum where it is theoretically forbidden, are to

b expected. Fig. 2 shows that such small residual intensity is sometimes

observed. This subject has been considered in the discussions of the Raman

rpect:a of calrite and scidiio nitrate. 8 Furthermore, in the course of checkir.r.

the polarized spectra usin alternate incident and scattered light axes, it -4as

-.. ",hi , . -. -.. ... te ai. .g.b -. .. dEii -- *J,-i i-C'-

hi --.r*. ..., . • i - - -- -':-

obzc-ved that the relitive intensities of the bands sometimes !:hanged although

the name tensor compnent was being observed. This effect is most striking for

v in the a polarization, wich is shown in Fig. 3. Again the optical

inhaogeneity of the VI O3 can account for this observation, as discussed in

detail for the cast of calcite by Porto, et al -

A. internal Modes

The symmetric stretching vibration (v1 ) of the No3- ion in crystalline

LiO 3 is observed at 1071. 0 cm"1 . This frequency is close to that observed8 for

NaNO3 (v I = 1068 cm -1), but appreciably different from the value of 1086 cm- 1

previously reported for solid Li0 3 ,1 which was used in the subsequent discussion

of frequency trends among various alkali nitrates. 2- 5 Detailed scans of the v1

region in the n,-polarization using the approximately 0.7 watt of power available0 0

in each of the argon ion laser lines at 4880 A and 5145 A showed the presence of-l

a sharp band at 1050 ± 1 cm with an intensity about 1/200 of the strong peak

at 1071 cm- 1 . The relative intensity sugests that the weak band at 1050 cm- I

is due to the vI vibration of a nitrate ion containing one oxygen -18 atom. 18

Calculation of the isotopic frequency assuming a valence force field and a planar

geometry for the XYZ 2 molecule 19 gives excellent agreement with the observed

value, thus confi'.ming this assignment.

The antisymmetric stretching vibration (v3) is located at !384.6 -

a value slightly lower than the previously reported frequency of 1391 cm- 1 ,

but essentially identical to v3 of the nitrate ion in HaNO,, which is observed8

at 1385 cm- 1. Careful examination of this spectral region using argon ion-1

excitation revealed a weak band at 1354 cm . Isotope shift calculations suggest

the assigrzent of this band to v3 of sNO 3" ions, present in 0.37% natural

abun4ance.

-6-

The remaining internal fundamental, the in -pP. ne bending motion, v4 is-1

observed at 735.4 cm . This frequency is slightly hl',her than that listed by

Sues (728 cm- ), and also hOigher tan the value emrved 8 for NaNO3 (724 cm-)

The frequency shift of 1-1/2% in t 14; vibratlon uprn. change in cation, although

the largest for the internal modes, confirms that 'cie bonding in the nitrate ion

is essentially unaffected by tbe packing in the crystal. None of the fundamentals

show frequency shifts of sufficiently significant magnitude to be correlated with

the available "free vo_ -ane" in the various respective crystal lattices.

The depola -izition behavior exh-.bited by the internal vibrations, shown

in Fig. 2, closely follcws exuectati.i. For the EE modes, v, and V4, the grout

theoretical selection rules predict that the allowed polarizaility components

are axx = - Iyy = CL and X, = . However, it is expected that the intensity

in the xz- and yz-spectra will be relatively weak, 1 7 since in the limit of anisolated nitrate ion (D3h symmetry) the out-of-plane polarizability derivative

components approach zero during the in-plane vibrations. The absence of v3 and

V in the x(zy)z-spectrum of Fig. 2 demonstrates that there is little interionic

coupling in the LiNO3 lattice. It is satisfying to note that the intensities of

these bands in the yy- and xy-spectra are indeed quite similar.

For the A symmetric stretching vibration, v1 , the only nonvanishingpolarizability derivatives should be a x and a. I addition, symmetry

considerations suggest that a > a The intensity ratio (a /a ) isYY zz "y Z

approximately 10 for the 1071 cm" band of LiNO3 the corresponding ratio "l also

8 910 for NaNO 3 , and is about 5 for CaCO 3.

--

The v band also appears weakly in the x(yx y-spectr-= of Fig. 2,

and is barely discernible in the x(sy)z orientation; in both cases xz shotld

be absent. Furthermore, as shown in Fig. 3, the relative intensity of the

1071 m- i band in the xy-polarization is much higher when the -crystal is rotated

so that light either enters or is gathered along the crstallographic z -axis.

This striking orientation dependence is clearly due to the particular properties

of the LiMO3 crystal, and not to an inherent oreakdwn of syometry selection

ritles. Indeed, such behavior is expected for highly birefringent crystals,

if either the incident or scattered light is convergent. This effect has been-

discussed in detail by Porto, Gicrdmaine, and Damen, 9 and is also observed in

calcite20 and sodium nitrate. If the incident polarized light is not extremely

well collimated and- colinear w-th the intended crystalline axis interference

between the ordinary and extraordinary components of other polarizations will

contribute to the scattering. For example, in our z(yx)y spectrm depolarized

z(xx)y light, where v, is strongly allowed, would also be detected. Similarly,

since the scattered radiation is collected over a finite solid angle, contributions

from unwanted polarizations will also be measured. Thus depolarized scattering

from x(yy)z will be contained in our x(yx.z-spectru. The dependence of these

effect3 on solid angle has been discussed by Porto, et al; 9 depolarization is

important for divergence angles as small as 10. In our experiments the half-

angle of the focused incident light was about 0.50, whereas the scattered

radiation was collected over a cone with -rng!e of approximately lo. Thus the

anomalous a - intensity for v would be expected to be stronger in the x(yx)z-xy

spectrum than in z(yx y; this is borne out in Fig. 3. These interference effects

are vi~tziaazaa axa t&* Xs Seuc 510-'C 'Oat&r

or scattered ligh't FgLives r~s to zx- cr -Dlztco, .atere ;-az A-

vil-rat:bon iqs aLso forbidden. The neak 4Z~eararce vr-. i&Z z(yx)y setr-

must therefore be attributed to cmystat' itt i berfczoaS sa 5iigrsent f t

5e20e -4' the laser- bea-m. Snethle ayrSv' oretro eas *vsted to !o%-tL a-e

the iteony ec-ectaftci, it ikWely th-at thbe unavctddalble, maurmal 2s tr

dieas in t:-v- b2IXO sarn-le are repfl- c- most Cdc S-%jFt_

m. azan spectre sS nFigs. 2; ~43wt-e ctvtaiWd tN the He-

laser. -;The higher- pnr- Art laser ues alouse-d acm - Spw--A-!r -az

of exceflent qual t t.Y. T:e hgr Scanterfrint 'rensifty Dt-ited aruvslits

to b-a used, but no aw fine ttrtoir wa rc -U~cm~, ~~ t-e

ccupar-able to* thr se I - the !Be-KieCrezct ~eec~cotv sc-arrna

due to the interc.J funmiamewta's f 'r above- the ttmai inztMsixv Scaie r. vsa

sape tral featsures were ots-erved -In adM the iscojc bandi mo

abov~,sernaloctujotion L~#e wee chstred - - eklee t17

assigned1 to 2'-; - :%-e ie~~ .- ch is i=7rce-&t72

o-Wy ere2-_ -a occur at bc' u a-, 16 a cm icrfL- theats -.emd ~--ai

ob-served at 146 t, 2115 t 213±2a are ass. £ne4 v-

and v. I ca e & ofa=2 ~~~~:-

sa l. Mthef- vter %8k band- t-. fjg at '2159 2r 34 r=- * ni cano =-M De&

a-, ied. -I-t sho uld be noted thaz even unr- the inteCs arfon :lOtno tat:

a evidence of R=an scar-terirg due tco~ialt=~s---

tode andI tte external (atc)vAia osc~l eow~rd

-9-

B. Lattice Modes

The low frequency region of the Raman spectrum of c--ystalline LINO3

was investigated to within about 10 cm"I of the exciting line. No evidence of

the postulated low-frequency torsion around the optic axis was noted, even with

the higher power exciting lines. Two lattice vibrations are observed (rigs. 2, 3)

at 124. cm and 237. 2 cm , and are attributed to the expected translational and

librational normal modes, in analogy to sodium nitrate. 8 These assignments are

supported by the depolarization ratios of the two bands, which are summarized in

Table I. Both vibrations are of E symmetry, where the allowed polarizabilityg

components are a - a and a However, as Saksena 17 has shown,

the in-plane translation should have axz = ayz 0 and the out-of-plane libration

should have a - a = a 2 0, if the coupling between the vibrations is small.xx yy xy

This is precisely the behavior shown in Figure 2, where the 124 c- 1 band

(translation) is absent in the x(zy)z-spectrum, and the intensity of the 237 cm-1

haA ' (libration) is smallest in the x(yy)z-spectrum. The latter intensity is

even slightly lower than in the x(zz)y-spectrum where E bands are not allowed

by symmetry. The residual zz-scattering probably arises from the birefringence

effects discussed above. In each polarization the LiNO3 spectra are superior to

those Pre-iously reported for NaNO3 ; since the crystals have s;milar birefringence,

it is likely that crystal imperfections contribute most to the additional anomalous

intensity in sodium nitrate. In neither crystal was evidence of mixing between

the lattice modes apparent.

-10-

Unlike the internal fundamentals, the lattice frequencies are highly

dependent upon the cation. The librational frequency is 1.9 times that of the

tr'-slation in each case, but the LiNO 3 vibrations are a factor of 1.27 higher

in7 .1-quency than the corresponding NaN0 3 bands. As Figure 1 shows, this cannot

be attributed to the lighter mass of the lithium ion, since the cation is not

involved in either Raman-active normal mode. However, it is quite reasonable

that these relatively large-amplitude low-frequency nit.2&.t ion vibrations will

be dependent on the unit cell volume. Indeed, the volume of the NaNO 3 unit

cell is approximately 30% larger than that of LiNO3;6 thus the external NO3-

vibrations will be more highly hindered in the LiNO3 lattice, and would be expected

to exhibit higher frequencies.

SUHMARY

The polarized Raman spectrum of an anhydrous single crystal of LiNO3

has beev obtained under optimum experimental conditions. The Raman-active fundamerta:

vibrations G-1 the nitrate ion in the lattice have been established " follows:

V (E) 735. 1071. and v (E) 1384.6 cm- . Several bands due to4 g 41 0 1 g

isotopic molecules and ctArbinatit-: . o I2MM.rentals were observed. The frequencies

of the nitrate ion internal vibrati-,na L.- -. sentially independent of the cation

in the host lattice; thus there is no tvidence to support the "free volume" theory2

for the frequency shifts advanced by Bues, For some of the bands minor depolar-

ization anomalies occur in the form of residual intensity in polarization

orientations where Raman scattering is not expected according to symnetry. Although

much of this intensity can be accounted for by the highly birefringent nature of

the LiNO crystal, it is clear that crystalline imperfections may also contribute.

I

This is demonstrated by the fact that NaNO3 , which has about the same birefringence,

exhibited stronger anomalous depolarization than LiNO The predictedU

sensitivity of the interference effects due to birefringence to the orientation

of the crystalline sample is quite striking In the pnecelA ei:.!'te.

The Raman-active lattice modes of LiNO have been obtained, ..M ass[gned

to primarily translational (12u (.-) ard librationi (237.2 cm i ) vibratLons3

respectively. No indication o ccupli:g between t!ese 12 Vibratiors i3 observed,

There is no evidence of an additional lowtfr' q,-ncv torsional lattice !,,de.

The frequencies of the external vibratics of the nitrate ion dJo depend strongly

on the host lattices. Althou-,b the ratio rf the librat-on/traansation frequency

is about the same in U!NO3 as In NaH0 3 , vhe valuiea of the frequencies themselves

are about 30% higher. A shift in this d.xettion is cults reasonable in light,

of the significantly smaller unit cell voluae of LO.

Partial support 1rc the U. S. Off izs of Naval Research is gratefully

acknowledged.

References -12-

,. Vz He Thatte and A. S. Oanevan* Indian . rhysic.8, 3- ).

2.1,. BUft, Z, Fjyj,41k Chom., Nze Folge. l0, 1(1957).

3. J. 2. M Sthieu and H. Lc"iburvp, Discuse. Faradlay Soc. 9, 196(1950).

5. K. W. F. K ira~a1 e~r en-cl 1 &*tnfliaad-uni jaixtbuch dcrPhys i d 9V1, Liep '-Le3).

6. 1. W. G. Wyckoff, The__.mruetur* af OCdtz, ,Ctici Catalog 1o .ay, Inc.Zew York, 13-)l 2nd Bd.

7. S. Bbagavatnam ani T. yent .itayudu, Proc. Intl. Acad. Sci.A9, 224(1939).

r. l. L. RoUsseau, R. E. K4ller, aad 6, E, Lerol, J. Chem. Phys.4a, 3409(1968).

9. S. P. 3. Porto, J. A. Oiordcaint, tnd T . ijiDen, Phys. Rev. 147, 6iG.1-96 6).

105. The tbtndbooIk of Chmistry end Physics~, (ThA C~hemical Rubber CoOhio, i957) 35th Ed. Reront editon" contain a misria atr c.i"O.-

11, R. A. Schroed-, C, E. Vlr, and E, R. Lippinctt, J. Re. Yat. Bur. Stdso. 66A

12. J. HI. ibben, Raman Effect and its Clh;mlcal Asplications (Reinhold Publ. Co.,

New York, 1939) p. 428.

13. D. Xirihnamurti, Proc. lnd. Acad. Sci. 46A, 103(1957).

14. R. E. Miilr, K. L. Treuil, R. R. Getty, and t. E. Leroi, Technical Report No.4(29),ORedesign and Construction of a Laser Raman SpectrophotometerM ay, 1968,DDC Accession No. 669797,

15. T. C. Daien, :;. P. S. Porto, and S. B. Tell, Phys. Rev. 142, 570(1966).

16. S. Bhagavantau, Proc. Lad. Acad. Sci. A13, 543(1941.'

17. B. D. Saksona, Proc. Ind. Acad. Sci. All, 229(1940).

18. The natural abundance of 180 is 0.204%.

19. E. Herzberg, Infrared and' Ranan Spec ra of Po-'atonic Noecules (D. Van acss:3n0Co., Inc., Princeton, N. J., 1945).

20. See references cited in reference 9..

21. T. M. K. Nedungadi, Proc, Ind. Acad. Sci. A8, 397(1938).

q IMIv f4

IN M

I~ I S it etI

~ it,

4J I~d

410 0

-4ccr o i

rig 1 The R-im-n-ac tve la t ice ---- --f cvyjstal Iine IN%.3 a. Libration.

b~. Transation.

Fig. 2. Raa Sppg~tra of !IUM, in the four unique polarizatiOn orientations.

131 ~. T le R -ma pctrum of LOUO in the xy-polarization as a functilon of

cryz-,ai ;5dentat ion.

10'

.4L

I - - -

0 J-- I6

. II I I

i

- ---

1200 am40

FRIM C I Iip-1

Securi2ty ClassificaionDOCUMENT CONTROL DATA - R&D

(0b.%utY c.sf~ of titt*. bft~r of aboreset Ad irwdind inft*W14 MO~ '. WntOf4 uhon *%V r~ewt port in clessit.0

ORIGATI G ATIVTV (oop~to wawl m. 09WORT SIECufiT~t C LASS.V#CATIONMichigan State University UnclassifiedDepartment of Chemistry 26 GROUP-----

Miller, RI. E., Gett-y, R. R., freuil, K. L., Leroi, G. E.

*CO* AE741 TOTAL NO. OF VACES REFNo.Bp

49L CONtTRACT OR 40AN ft. O. RIGIMATOR'S REIPORT NUWM9ft3)

MOO~'-6-A-oog-003Technical Report No. 8(33)

ThEN REPORT SGOMS (Amw .EWn*-tm ft" -ay be mmofeie

14L AVAIL AILITY1UNW1ATION4 RTICKS

Distribution of this document is unlimited

111. SUPPLIKOUTARY NOTES I&. SPOsmSuag YSUTARY ACTIVITY

Office of Naval Research-~ Physics Branch, Code 4i21

ti. AOSTRACT

Polarized Raman spectUra of an anhydrous single crystal of LiN% bhe- zi obtainedizndez' optimim experimental Conditions. The internal vibrations of the nitrate ion

s*Obst2rved~as follzowis: I4 735.54 -,,,(Aj~)=17 and v(E 1 384i. cm

these frequencies; ame sssentially independent of the cation in the hosr lattice.The two Raman.active (E ) lattice modes have been located and assigned to primar-ily translational (1 £ cm~ and librational (3.cm ntos No indica-io

=of c- Uvling between these vibrations is observed. The fre~quencies of the externalvibrations relative to those of NalF% demonstrate that they are dependent on the

volume of the crystalline unit cell. Except for small intensity contributions Insome orientations which are probably due to depolarization effects in the

birfrngntcrystal, the polarizability activitie of ohteiterna n

extewal modes follow the expected sy-etry slction rls

0 D ORN 473 ncl S eui tye Classificationt

Secz;:ity Ctassi,,.ca.c...

4 LINK A LINKa LINK rOLE WT OLE WT NOLK 5T i

INSTRUCTION3OG ICINA7TNG ACTIVITY E nter the na . and &Iress inpose 1 b t Sec ?ity Classifict 1on0, utin stalndard statements

te se cT.-Ity Cr other organization coporate author) ssui ng1 (1) "Q aified requesters may obtain copies of thisIth, report. report frorl DDC."Za. REPORT SECJ'ITdTY CLASSIFICATION; Enter the oer-) "FoetC-. annowtcement and dissemination of thisal. securnty classification of the report. Indicate whether report by DD is not authorized.""Restriced Data" is ircluded Marking is t be in accord-• nce with appropriate security regulations. (3) "U. S. Government agencies may obtain copies of

this report directly from DDC. Other qualified DDC

reC%.-1re 5r'O. 10 arnd Armed Forces lr44stnai Man-'l ser shill request through,he tr -unwnber. Also, when app~cab!e. sh!3w that optioralmarkvigs have been used for Group 3 and Group 4 as ai;:hotr (4) "U. S military ageocies may obtain copies of thisized report directly from DDC. Other qualified users3. REPCRT TT4LE: Ertertl e ccrete rep.r ti te in all shall reqnest throughcaptal l ters. Ttles in a! cases should be unclassified-

sl a treantfuI title carxnat be selected without classifice-tion, shOw tide cloisifica!.on int all cApitals in parenthesis (5) "AnI distribution of this report is controlled. Quid-

eda:y follOWing the btle. ifeid DDC users shall request through

£DE5CRIP7I'.7E NOQTES. If epprcprte. enter the typer of____________________rnfi .g,, iterim. proxres,, sr.--!.ary. ln7!al. or fial, If the report has been furnished to the Offlo" of TechnicalGive th.o Li.smve date% when a s;e:ifc repar.',n ;erad is Services. Department of Commerce. for sale to the public, Indt-CTverV cat* this fact and enter the price, if known5. PL TH3R ,3> Emer the nwre{.s of author's; as shown on IL SUPPLEMENTARY %OTES: Use for additional explsna-or in the report Enter lst ne-., first name. middle initiaL tory note-$it military. show rank a, brarch of service. The nra% ofthe princpsa author to an aolsotute miniLnunt requirerren.. 12. SPOISORING MIUTARY ACTIVITY: Enter the name of

the departmental project office or laboratory sponsoring (par, REPORT DATL- Enter the date of the report as day. in for) the r~searcb sod development. Include address.

month. year or mornh year. If more than one date appearson the repo. use date of pub! 1catiorL 13. ABSTRACT: Enter an abstract giving a brief end factual

summery of dte document indicative of the report, even though7a TITAL NUGBER OF PAGES: The total page count it may also appear elsewhere in the body of the technical re-shcud follow narmal pagination procedures. i.e.. enter the pa If additional space is required. a continuation sheet shallnumber of pages Containing i-fora,*on. be attached.7b. . UMBER OF REFERENCES Enter the total numbor of It is higly desirable that the abstract of classified reportsreferences cited in the report. be unclassified. Each paragraph of the abstract shall end withSi CON4TRACT OR GRA.T NIUMIER: If approprTaTe. enter an indication of the miLitary sec-irity classification of the in-the applicable number of the contract -r grant under which formation in the paragrsph. represented as (TS, (S), (C). er (U)the report was writter. Ther is no Limitation on the length of the abstract. How-Ab. 8r, & "d PROJECT NUM3ER. Enter It" appropriate ever, die suggested length i from 150 to 225 words.silitay deparuto idntficat onr such as project . 14 KEY WORDS: Key words are technically meaningful terms

or short phrases that characterize a report and may bie used as9a ORIGINATOR'S REPORT NL"VMIR S)- Enter th ofil- indes entries for cataloging the report. Key words must becial report number by which the docunent will be identified selected so that no security classification Is required. Idonfti-aid contl!!ed by the originating activity. This number must tiers, suc as equipment model designation, trade name, militarybe unique to this re-om project code name, gieograthic location, may be used s key

9b OTHER REPORT N.'M3ER(S): If the report has been words but will be followed by an irdication of technical con.asstirned any other report nwmbers fei Ter by the art.narlr text. The assirin-ent of links, roles, end weights is optional.or P1/ the spoeosr), at so en:ter thi s numberfs).

If. AVALLABI TY/ .IMITATION NOTICE&- Enter any tin.:ations on further dis e:uninltion of the report, other tha !4hos

UnclassifledSecurity Classification