S IR C H A N D R A S E K H A R A V.

RA M A N

The molecular scattering of light

Nobel Lecture, December 11, 1930

The colour of the sea

In the history of science, we often find that the study of some natural phe-
nomenon has been the starting-point in the development of a new branch of
knowledge. We have an instance of this in the colour of skylight, which has
inspired numerous optical investigations, and the explanation of which, pro-
posed by the late Lord Rayleigh, and subsequently verified by observation,
forms the beginning of our knowledge of the subject of this lecture. Even
more striking, though not so familiar to all, is the colour exhibited by oce-
anic waters. A voyage to Europe in the summer of 1921 gave me the first
opportunity of observing the wonderful blue opalescence of the Mediter-
ranean Sea. It seemed not unlikely that the phenomenon owed its origin to
the scattering of sunlight by the molecules of the water. To test this explana-
tion, it appeared desirable to ascertain the laws governing the diffusion of
light in liquids, and experiments with this object were started immediately
on my return to Calcutta in September, 1921. It soon became evident, how-
ever, that the subject possessed a significance extending far beyond the special
purpose for which the work was undertaken, and that it offered unlimited
scope for research. It seemed indeed that the study of light-scattering might
carry one into the deepest problems of physics and chemistry, and it was this
belief which led to the subject becoming the main theme of our activities at
Calcutta from that time onwards.

The theory of fluctuations

From the work of the first few months, it became clear that the molecular
scattering of light was a very general phenomenon which could be studied
not only in gases and vapours but also in liquids and in crystalline and
amorphous solids, and that it was primarily an effect arising from molecular
disarray in the medium and consequent local fluctuations in its optical den-
sity. Except in amorphous solids, such molecular disarray could presumably
268 1930 C.V.RAMAN

be ascribed to thermal agitation, and the experimental results appeared to

support this view. The fact that molecules are optically anisotropic and can
orientate freely in liquids was found to give rise to an additional type of
scattering. This could be distinguished from the scattering due to fluctua-
tions in density by reason of its being practically unpolarized, whereas the
latter was completely polarized in the transverse direction. The whole sub-
ject was critically reviewed and the results till then obtained were set out in
an essay published by the Calcutta University Press in February 1922.
The various problems requiring solution indicated in this essay were in-
vestigated with the aid of a succession of able collaborators. It is possible to
mention briefly only a few of the numerous investigations which were car-
ried out at Calcutta during the six years 1922 to 1927. The scattering of light
in fluids was studied by Ramanathan over a wide range of pressures and
temperatures with results which appeared to support the "fluctuation" the-
ory of its origin. His work also disclosed the remarkable changes in the state
of polarization which accompany the variations of intensity with tempera-
ture in vapours and in liquids. Liquid mixtures were investigated by Kames-
wara Rao, and furnished optical proof of the existence in such systems, of
simultaneous fluctuations of density, composition, and molecular orientation.
Srivastava studied the scattering of light in crystals in relation to the thermal
fluctuations of density and their increase with temperature. Ramdas inves-
tigated the scattering of light by liquid surfaces due to thermal agitation, and
established a relation between surface-tension and surface-opalescence. He
also traced the transition from surface-opalescence to volume-opalescence
which occurs at the critical temperature. Sogani investigated X-ray diffrac-
tion in liquids, in order to connect it with their optical behaviour, and test
the application of fluctuation theory to X-ray scattering.

The anisotropy of molecules

As stated above, the state of polarization of the light scattered in fluids is

connected with the optical anisotropy of the molecules. Much of the work
done at Calcutta during the years 1922 to 1927 was intended to obtain data
concerning this property and to establish its relations with various optical
phenomena. Krishnan examined a great many liquids, and by his work
showed very clearly the dependence of the optical anisotropy of the mole-
cule on its chemical constitution. Ramakrishna Rao studied the depolariza-
 MOLECULAR SCATTERING OF LIGHT 269
tion of scattered light in a very large number of gases and vapours, and ob-
tained information of high importance for the progress of the subject. Ven-
kateswaran studied the scattering of light in aqueous solutions to find the in-
fluence on it of electrolytic dissociation. Ramachandra Rao investigated liq-
uids having highly elongated molecules and also highly polar substances
over a wide range of temperatures, and discovered the influence of molecular
shape and molecular association on the depolarization of scattered light in
liquids.
The interpretation of the observations with liquids involved the develop-
ment of a molecular theory of light-scattering in dense media which was
undertaken by Ramanathan, myself, and Krishnan. A revised opalescence
formula was derived which differed from that of Einstein and yielded results
in better agreement with observation. Krishnan and myself also published a
series of investigations showing how the optical anisotropy of the molecules
deduced from light-scattering could be utilized to interpret the optical and
dielectric behaviour of fluids, and also the electric, magnetic, and mechanical
birefringence exhibited by them. The conclusions derived from these studies
enabled a connection to be established between the molecular anisotropy
observed in fluids and the optical, electric, and magnetic aeolotropy ex-
hibited by solids in the crystalline state.

The investigations referred to above were in the main guided by the classical
electromagnetic theory of light, the application of which to the problems of
light-scattering is chiefly associated with the names of Rayleigh and of Ein-
stein. Nevertheless, the possibility that the corpuscular nature of light might
come into evidence in scattering was not overlooked and was in fact elab-
orately discussed in the essay of February 1922 which was published at least
a year before the well-known discoveries of Compton on X-ray scattering.
While our experiments in the main appeared to support the electromagnetic
theory of light, evidence came to hand at a very early stage of the investiga-
tions of the existence of a phenomenon which seemed to stand outside the
classical scheme of thought. The scattering of light in transparent fluids is
extremely feeble, much weaker in fact than the Tyndall effect usually ob-
served in turbid media. It was experimentally discovered that associated with
the Rayleigh-Einstein type of molecular scattering, was another and still
270 1930 C.V.RAMAN

feebler type of secondary radiation, the intensity of which was of the order
of magnitude of a few hundredths of the classical scattering, and differed
from it in not having the same wavelength as the primary or incident radia-
tion. The first observation of this phenomenon was made at Calcutta in
April 1923 by Ramanathan who was led to it in attempting to explain why
in certain liquids (water, ether, methyl and ethyl alcohols), the depolariza-
tion of scattered light varied with the wavelength of the incident radiation.
Ramanathan found that after exhaustive chemical purification and repeated
slow distillation of the liquid in vacuum, the new radiation persisted un-
diminished in intensity, showing that it was a characteristic property of the
substance studied and not due to any fluorescent impurity. Krishnan ob-
served a similar effect in many other liquids in 1924, and a somewhat more
conspicuous phenomenon was observed by me in ice and in optical glasses.

The optical analogue of the Compton effect

The origin of this puzzling phenomenon naturally interested us, and in the
summer of 1925, Venkateswaran attempted to investigate it by photograph-
ing the spectrum of the scattered light from liquids, using sunlight filtered
through colour screens, but was unable to report any decisive results. Ra-
makrishna Rao in his studies on the depolarization of scattering during 1926
and 1927 looked carefully for a similar phenomenon in gases and vapours,
but without success. This problem was taken up again by Krishnan towards
the end of 1927. While his work was in progress, the first indication of the
true nature of the phenomenon came to hand from a different quarter. One
of the problems interesting us at this time was the behaviour in light-scatter-
ing of highly viscous organic liquids which were capable of passing over into
the glassy state. Venkateswaran undertook to study this question, and re-
ported the highly interesting result that the colour of sunlight scattered in a
highly purified sample of glycerine was a brilliant green instead of the usual
blue. The phenomenon appeared to be similar to that discovered by Rama-
nathan in water and the alcohols, but of much greater intensity, and, there-
fore, more easily studied. No time was lost in following up the matter. Tests
were made with a series of filters transmitting narrow regions of the solar
spectrum and placed in the path of the incident beam, which showed that in
every case the colour of the scattered light was different from that of the in-
cident light, and was displaced from it towards the red. The radiations were
 MOLECULAR SCATTERING OF LIGHT 271
also strongly polarized. These facts indicated a clear analogy between the
empirical characters of the phenomenon and the Compton effect. The work
of Compton had made familiar the idea that the wavelength of radiation
could be degraded in the process of scattering, and the observations with
glycerine suggested to me that the phenomenon which had puzzled us ever
since 1923 was in fact the optical analogue of the Compton effect. This idea
naturally stimulated further investigation with other substances.
The chief difficulty which had hitherto oppressed us in the study of the
new phenomenon was its extreme feebleness in general. This was overcome
by using a 7-inch refracting telescope in combination with a short-focus lens
to condense sunlight into a pencil of very great intensity. With these ar-
rangements and using complementary light-filters in the path of the inci-
dent and scattered beams, as was done by Ramanathan in 1923, to isolate the
modified radiations, it was found that they could be readily observed in a
great many liquids, and that in many cases they were strongly polarized.
Krishnan, who very materially assisted me in these investigations, found at
the same time that the phenomenon could be observed in several organic
vapours, and even succeeded in visually determining the state of polarization
of the modified radiations from them. Compressed gases such as CO, and
N2O, crystalline ice, and optical glasses also were found to exhibit the mod-
ified radiations. These observations left little doubt that the phenomenon
was really a species of light-scattering analogous to the Compton effect.

The spectroscopic characters of the new effect

Thanks to the vastly more powerful illumination made available by the 7-

inch refractor, the spectroscopic examination of the effect, which had been
abandoned in 1925 as indecisive, now came within the reach of direct visual
study. With a Zeiss cobalt-glass filter placed in the path of the incident beam
and one or other of a series of organic liquids as the scattering substance, a
band in the blue-green region was observed by me in the spectrum of the
scattered light, separated by a dark interval from the indigo-violet region
transmitted by the filter. Both of these regions in the spectrum became sharp-
er when the region of transmission was narrowed by the insertion of an ad-
ditional filter in the incident beam. This suggested the employment, instead
of sunlight, of the highly monochromatic radiations given by a mercury arc
in combination with a condenser of large aperture and a cobalt-glass filter.
272 1930 C.V.RAMAN

With these arrangements the spectrum of the scattered light from a variety of
liquids and solids was visually examined, and the startling observation was
made that the spectrum generally included a number of sharp lines or bands
on a diffuse background which were not present in the light of the mercury
arc.

Fig. I. Spectrum of carbon tetrachloride.

The quartz mercury lamp was so powerful and convenient a source of

monochromatic illumination that, at least in the case of liquids and solids,
photographing the spectrum of scattered light was found to present no extra-
ordinary difficulties. The earliest pictures of the phenomenon were in fact
taken with a portable quartz spectrograph of the smallest size made by the
firm of Hilger. With a somewhat larger instrument of the same type, Krish-
nan obtained very satisfactory spectrograms with liquids and with crystals
on which measurements of the desired precision could be made, and on
which the presence of lines displaced towards the violet was first definitely
established. The experimental difficulties were naturally greater in the case of
gases or vapours, though they could be lessened by working with the sub-
stance under pressure. With an improvised instrument of large aperture
(F/1.8), Ramdas obtained the first spectrograms with a gaseous substance
(ether vapour) at atmospheric pressure.

In interpreting the observed phenomena, the analogy with the Compton

effect was adopted as the guiding principle. The work of Compton had
gained general acceptance for the idea that the scattering of radiation is a
unitary process in which the conservation principles hold good. Accepting
this idea it follows at once that, if the scattering particle gains any energy
 MOLECULAR SCATTERING OF LIGHT 273
during the encounter with the quantum, the latter is deprived of energy to
the same extent, and accordingly appears after scattering as a radiation of
diminished frequency. From thermodynamic principles, it follows that the
reverse process should also be possible. Adopting these ideas, the actual ob-
servations could be interpreted, and the agreement of the observed displace-
ments with the infrared frequencies of the molecules made it clear that the
new method opened up an illimitable field of experimental research in the
study of the structure of matter.

Interpretation of the effect

It appears desirable to emphasize that though the conservation principle of

Compton is useful in interpreting the effects disclosed by experiment, it is by
itself insufficient to explain the observed phenomena. As is well known from
studies on molecular spectra, a gaseous molecule has four different species of
energy of increasing orders of magnitude, namely those corresponding to
translatory motion, rotation, vibration, and electronic excitation. Each of
these, except the first, is quantized and may be represented by an integer in
an extended sequence of quantum numbers. The aggregate energy of a mole-
cule may, therefore, assume any one out of a very large number of possible
values. If we assume that an exchange of energy occurs in the collision be-
tween the molecule and the quantum, and limit ourself to the cases in which
the final energy of the molecule is less than that of the incident quantum, we
arrive at the result that the spectrum of the scattered light should contain an
immense number of new lines and should in fact rival in its complexity the
band spectrum of the molecule observed in the emission or absorption of
light. Nothing more different from what is actually observed can be imag-
ined than the foregoing picture. The most conspicuous feature revealed by
experiment is the beautiful simplicity of the spectra of even complicated poly-
atomic molecules obtained in light-scattering, a simplicity that is in strik-
ing contrast to the extreme complexity of their emission or absorption
spectra. It is this simplicity that gives to the study of light-scattering its
special significance and value. It is clear that the effect actually observed was
not and could not have been foreseen from an application of the conserva-
tion principles.

The general principle of correspondence between the quantum and clas-

sical theories enunciated by Niels Bohr enables us, on the other hand, to ob-
274 1930 C.V.RAMAN

tain a real insight into the actual phenomena. The classical theory of light-
scattering tells us that if a molecule scatters light while it is moving, rotating
or vibrating, the scattered radiations may include certain frequencies, differ-
ent from those of the incident waves. This classical picture, in many respects,
is surprisingly like what we actually observe in the experiments. It explains
why the frequency shifts observed fall into three classes, translational, rota-
tional and vibrational, of different orders of magnitude. It explains the ob-
served selection rules, as for instance, why the frequencies of vibration de-
duced from scattered light include only the fundamentals and not the over-
tones and combinations which are so conspicuous in emission and absorption
spectra. The classical theory can even go further and give us a rough indica-
tion of the intensity and polarization of the radiations of altered frequency.
Nevertheless, the classical picture has to be modified in essential respects to
give even a qualitative description of the phenomena, and we have, there-
fore, to invoke the aid of quantum principles. The work of Kramers and
Heisenberg, and the newer developments in quantum mechanics which have
their root in Bohr’s correspondence principle seem to offer a promising way
of approach towards an understanding of the experimental results. But un-
til we know much more than we do at present regarding the structure of
molecules, and have sufficient quantitative experimental knowledge of the
effect, it would be rash to suggest that they afford a complete explanation
of it.

The significance of the effect

The universality of the phenomenon, the convenience of the experimental

technique and the simplicity of the spectra obtained enable the effect to be
used as an experimental aid to the solution of a wide range of problems in
physics and chemistry. Indeed, it may be said that it is this fact which con-
stitutes the principal significance of the effect. The frequency differences
determined from the spectra, the width and character of the lines appearing
in them, and the intensity and state of polarization of the scattered radiations
enable us to obtain an insight into the ultimate structure of the scattering
substance. As experimental research has shown, these features in the spectra
are very definitely influenced by physical conditions, such as temperature
and state of aggregation, by physico-chemical conditions, such as mixture,
solution, molecular association and polymerization, and most essentially by
 MOLECULAR SCATTERING OF LIGHT 275
chemical constitution. It follows that the new field of spectroscopy has prac-
tically unrestricted scope in the study of problems relating to the structure of
matter. We may also hope that it will lead us to a fuller understanding of the
nature of light, and of the interactions between matter and light.

Some concluding remarks

From a physical point of view, the quantitative study of the effect with the
simplest molecules holds out the largest hope of fundamental advances. The
beautiful work of McLennan with liquefied gases, and of R. W. Wood and
Rasetti are pioneer investigations in this field which command the highest
admiration. The quantitative study of the effect with crystals of the simplest
possible chemical constitution is naturally of great importance. The case of
the diamond, which has been investigated by Ramaswamy, Robertson, and
Fox, and with especial completeness by Bhagavantam, is of special interest.
Very surprising results have been obtained with this substance, which may
be the pathway to a fuller understanding of the nature of the crystalline state.
I should also like to draw attention to the work of Krishnamurti, who has
traced a remarkable dependence of the intensity of the spectral lines observed
in scattering on the nature of the chemical bond, and followed the transition
from the homopolar to the heteropolar type of chemical combination. Krish-
namurti’s observation that the paramagnetism of crystals apparently influ-
ences the observed intensity of the displaced lines is one of the most remark-
able ever made in this new field of research.