How K.R Ramanathan Could Detect Raman Scattering in Pure Liquids in the Year 1923 ?

1. Introduction

C.V Raman and his research students in Calcutta in India made systematic investigations in light scattering in different media during 1920’s and came out with some remarkable results which often deviated from the Rayleigh or eleastic scattering. Raman could even anticipate the role of quantum theory in molecular phenomena when he tried explain the color of sea water (Raman, 1922). In the year 1923 K,R Ramanthan one of the out standing students of Raman announced the discovery of inelastic scattering of light in pure liquids which at that time described as weak fluorescence.( Ramanathan, 1923). Ramanthan’s results were further confirmed by Krishnan ( 1925). These studies were the basis for the first report of Raman related to the finding of new radiations ( Raman scattering) present in the scattered light of almost all pure liquids ( Raman and Krishnan, 1928a). Subsequently spectroscopic confirmation of the Raman effect was communicated by Raman ( Raman and Krishnan, 1928b, and Raman, 1928) which fetched him the Nobel Prize for the year 1930.

The evolution of Raman scattering instrumentation can be divided in to three stages:-

i) Sunlight source, complimentary color filters and polarization detectors ( 1923 April -1928 January 16th ).

ii) Mercury quartz lamps, optical filters, spectroscopes, photographic or photoelectric detectors ( 1928 February 28th-1965)

iiii) Laser Raman speectrometers and allied advanced instrumentation ( 1966 onwards; HORIBA, 2025)

The time line of the discovery of Raman effect through the light scattering experiments conducted in Calcutta covering important mile stones during the years 1921-1928 is clearly communicated by CV Raman ( Raman , 1928; Raman, 1930). According to Pinzaru and Keifer ( 2018) the history of Raman effect starts with the paper communicated to Nature by Raman on February 16th, 1928 (Raman and Krishnan, 1928a). In this paper we have studied the similarities in the instrumentation used to detect Raman scattering in liquids in the studies of Ramanathan( 1923), Krishnan ( 1925) and Raman and Krishnan ( 1928a). The scientific importance of the study of Ramanathan in April, 1923 is explicitly studied, a century after its publication ( Ramanathan, 1923).

We will first discuss the experimental results of Ramanathan ( 1923) and Krishnan ( 1925) related to depolarization ratio measurements in the sunlight scattered from pure liquids which led to the inference of inelastic scattering of light ( now understood as Raman scattering). The relation between complimentary filters and Raman shifts will be then described citing suitable examples. The conditions for observation of Raman scattering in liquid samples is shown to depend on the existence of Raman bands in the favourable region associated with the Raman shifts related to the complimentary filters used. For substances containing polar molecules this lies in the OH stretching and CH stretching region with Raman shifts between 2800-3500 cm-1. We find that out of 65 liquids studied by Krishnan ( 1925) 59 of them has Raman bands in the above region . The advantages and limitations in the report of Raman and Krishnan ( 1928a) related to observation of Raman scattering in about 60 liquids using simple instrumentation ( intense sunlight, complimentary filters and polarization detectors similar to Ramanan, 1923) is then discussed. The present paper is an improved version of our earlier study (Girish and Eapen, 2014).

2. Experimental Arrangements for Detecting ‘Weak Flurescence’ or Raman Sattering in Pure Liquids by Ramanathan ( 1923) and Krishnan ( 1925)

i)Source: horizontal beam of sunlight reflected by a heliostat

ii) complimentary color filters:-

a)Ramanathan ( 1923) used blue filter ( cuprammonium sulphate liquid filter) and green filter ( Wratten color filter) for his investigations

b) Krishnan ( 1925) used blue filter , green filter orange filters for his studies

iii) Vaccum distilled and highly purified liquids are used as samples for their investigations ( Ramanathan used 2 samples and Krishnan used 65 samples)

iv) Depolarization ratio measuring facilities ( polarisers and allied optics)

The depolarization ratio ( DPR hereafter) defines the extent to which incident light changes its polarization characteristics after scattering from samples. This ratio ( DPR) measures the intensity of scattered light that remain parallel to the incident light ( weak component) divided by the intensity of scattered light that remains perpendicular to the incident light ( strong component)

DPR = (polarization of weak component)/(polarization of strong component) X 100% (1)

3. Experimental Results of Ramanathan ( 1923)

In the experiments of Ramanthan ( 1923) conducted during April 2023 in Calcutta, the incident sunlight is passed through a blue filter and DPR is measured (Pi). Now light after scattering from the liquid sample ( water and ethyl alcohol) is passed through the blue filter and again DPR is measured ( Ps). The experiment is now repeated with the green filter. The results of these experiments is found to satisfy the following inequalities :-

Let ΔP = Ps - Pi , now

Ps < Pi or ΔP < 0 (1) for the blue filter

Ps >Pi or ΔP > 0 (2) for the green filter

The experimental results of Ramanathan ( 1923) is given in Table 1.

From these results Ramanathan( 1923) concluded that the scattered radiation from the sample after passing through the blue filter contain new radiations resembling ‘weak florescence’. His inference is based on the following facts:

i) The new radiation similar to flurescence has a wavelength longer than the the wavelength of radiation exciting the ‘weak flurescence’ and present in the incident sunlight.

ii) the intensity of the scattered radiation is observed to be much lower than the incident light

iii) the scattered radiation from the sample after passing through the green filter is polarized as evident from increase in DPR.

iv) there is no fluorescent impurities present in samples because they are subjected repeated distillation in vacuum. So the results of Ramanathan ( 1923) suggests that this is first observation of inelastic scattering of light in liquids ( now understood as Raman scattering).

4. Experimental Results of Krishnan ( 1925)

K.S Krishnan , one of the important associates of Raman in Calcutta conducted his experiments of scattering of light following Ramanathan ( 1923) in the spring and summer of the year 1924 to detect the phenomenon of ‘weak flurescence’. He measured DPR for 65 pure liquids of which only for 30 liquids Krishnan used complimentary filters for his investigations.He used blue( violet) filter , green filter and the orange filter for his studies. But we confine our discussion of his results based on blue(violet) and the orange filters since his measurements with green filter has lot of data gaps.Krishnan ( 1925) could infer weak fluorescent phenomenon for 17 liquids .

Similar to earlier section the inequalities satisfied by the DPR measurements of Krishnan ( 1925) are

Ps < Pi or ΔP < 0 (1) for the blue(violet) filter

Ps >Pi or ΔP > 0 (2) for the orange filter

The relevant experimental results are given in Table 2.

The results of Krishnan( 1925) confirmed the results of Ramanathan ( 1925) related to detection of weak florescence or Raman scattering in pure liquids.

5. Relations Between Complimentary Color Filters Used in Light Scattering Experiments and Raman Shifts

Complimentary color filters were first used to detect fluorescence in organic liquids by scientists like Stokes ( Raman, 1928). This technique is followed in the experiments of Ramanathan ( 1923) and Krishnan ( 1925).

In Table 3 we have shown common colors and their wavelength ranges.

Let λe is the the wavelength of radiation exciting ‘weak flourescence’ and present in the incident sunslight . Similarly λs be the Raman signal detected after scattering from the sample and viwed through the long wavelength complimentary filter.

Now the Raman shift ( Δƴ) corresponding to the experiment is given by:-

Δƴ ( cm-1) = (1/ λe – 1/ λs ) X 10⁷ (5)

Here λe and λs are expressed in nm.

In Table 4 we have calculated the Raman shifts corresponding to certain complimentary color filters. Here λe correspond to the centre wavelength of the first complimentary color filter and λs correspond to the centre wavelength of the second ( long wavelength) complimentary color filter.We can find allowed raman shifts only for the green-red and blue-greencomplimentary filters.

We have to consider the passbands of color filters used by Ramananthan ( 1923) and Krishnan ( 1925). From their repective papers we can find that pass band of blue filter used by Ramanathan covered the entire blue region and similarly for the green filter.

The passband of blue filter used by Krishnan covered both violet and blue regions. The pass band of orange filter starts from greenish-yellow ane extends up to the red region.

Some allowed Raman shifts for the above complimentary filters is calculated as typical examples

i) λe= 470 nm ( blue filter)

Δs = 560 nm ( greenish yellow in orange filter)

Raman shift Δƴ = 3142nm

ii) λe= 500 nm ( blue filter)

Δs = 600 nm ( orange filter)

Raman shift Δƴ = 3380 nm

iii) λe= 480 nm ( blue filter)

Δs = 555 nm ( greenish yellow in orange filter)

Raman shift Δƴ = 2781 nm

These examples suggests that the complimentary filters used by Krishnan ( 1925) can detect Raman bands corresponding to CH stretching and OH stretching vibrations covering the region 2800-3500m cm-1 approximately. If Raman bands exists in the samples studied and are of sufficient intensity they can be detected.

6. Discussion

Two important features of Raman scattered light ( extreme weak intensity and higher polarization) which is clearly stated in the announcement of the discovery of Raman effect ( Raman, 1928) could be anticipated by earlier investigators ( Ramananthan, 1923, Krishnan, 1925))of this phenomena. During the complimentary filter era of Raman instrumentation before the spectroscopic detection of Raman spectra in liquids ( Raman and Krishnan , 1928b) the increase in polarization of light scattered from samples is the key for detection of Raman scattering . This is also the basis for the detection of ‘weak flourescece’ ( inelastic scattering of light) in pure liquids by Ramanathan ( 1923) and Krishnan ( 1925). It is clearly illustrated by CV Raman through the photographs of the polarization traces in the incident and the inelastic scattered light ( after passing through a green filter) from Toulene (see Figure 1 and 2 in Raman, 1928). The source used was sunlight filtered through a blue-violet filter. .From modern observations it is identified that the ratio of the intensity of scattered to incident light is of the order of 10-6. It is interesting to note that Krishnan(1925) estimated this ratio in the range 10^-4 to 10^-5 using the instrumentation of those days. The name ‘weak fluorescence ‘ given by the phenomena by Ramanathan ( 1923) is also thus appropriate .

K.R Ramanathan could detect ‘weak fluorescence ‘ in highly purified water and ethyl alcohol in 1923 and this can be considered as the first observation of ineleastic scattering of light ( now understood as Raman scattering) . His results was confirmed by Krishnan ( 1925) in 17 liquids even though the studies with complimentary filters were conducted for 30 liquids. The chemical family relationships between these liquids is noticed by Krishnan and it included water, ethyl ether , all the monohydric alcohols , benzyl and bezal chlorides , ethyl methyl ketone, diethyl ketone , butyric acid and acetaldehyde( see Table 2). The probability of detection of Raman scattering in pure liquids using the simple instrumentation of those days ( sunlight source, complimentary color filters and polarization detectors) from the study of Krishnan ( 1925) can be estimated to be 0.57.

Several favorable physical conditions are required for the observation of Raman scattering for the above experiments conducted more than a century ago. First the source is intense sunlight ( restricting the observations to periods of good sunshine) and it is not monochromatic. The short wavelength part of the incident sunlight ( blue or violet) is capable of exciting Raman scattering ( λe) and we know that the intensity of Raman scattered light is proportional to (1/λe⁴). The choice of complimentary color filters and its pass bands will introduce further uncertainties in the observation as explained in section of this paper. Existence of Raman bands for the given sample corresponding to the allowed Raman shifts for the given complimentary color filters is another constraint. All the 17 liquids ( this included water and ethyl alcohol studied by Ramanathan) for which ‘weak fluorescence ‘ is detected by Krishnan ( 1925) had Raman bands in the CH stretching and OH stretching vibration region ( approx. 2800-3500mcm-1). This is in favour of detection of Raman scattering for this liquids through increase in polarization in the scattered light when observed through the long wavelength color filer ( green or orange). Now we know that Raman lines are polarized even though the magnitude of depolarization ratio will depend on the symmetry of the Raman vibrations under observation. This explains the probabilistic nature of observation of Raman scattering in pure liquids using the experimental conditions of Ramanathan ( 1923) and Krishnan ( 1925).

S.Venkiteswaran an associate of CV Raman in his lab in Calcutta was successful in observing new radiations ( ineleastic scattering or Raman scattering ) with enhanced polarization in the organic liquid Glycerine in early January 1928 ( Mallik, 2000). His experiments was conducted with complimentary filters and sunlight. This result inspired Raman to continue his investigations related to ‘weak flouescence’ in pure liquids to discover optical analogue of the Compton effect for which Nobel prize was awarded in the year 1927. With a slight modification in the use of complimentary filters and very intense sunlight Raman could detect ‘weak fluorescence ‘ or Raman scattering in about 60 pure liquids in Janurary 1928 with the help of K.S.Krishnan. This is reported to Nature on 16th Janurary, 1928 ( Raman and Krishnan, 1928a). They also found enhanced polarization in the light scattered by these liquids. Unfortunately the details of liquids and the associated polarization measurements carried out is not published by Raman in this paper for unknown reasons. It has to be verified from the archives of the Journal Nature whether the review of this paper is favorable for publication.

In Table 5 we have shown the major Raman bands for other liquids studied by Krishan which is not included in Table 2. From Table 2 and Table we can find that out of 65 liquids studied by Krishan ( 1925) 59 had Raman bands either in CH stretching or OH stretching region ( 2800-3500 cm-1) even though the intensity of the Raman lines is found to vary from sample to sample. This information is gathered from modern published literature and chemical data bases. This result in support to the observations of inelastic scattering ( Raman scattering) in most liquids by Raman and Krishnan ( 1928a) compared to 57% liquids by Krishnan ( 1925). The relatively higher intensity of sunlight used by Raman and Krishnan ( 1928a) for thie experiments and passing this light through the blue-violet filter before falling in to the sample ( near monochromatic light) are other advantages for them compared to previous investigators like Ramanathan or Krishnan. Still the probabilistic nature of observation of Raman scattering when we employ crude instrumentation still holds good for the results of Raman and Krishnan ( 1928a). It must be note that there are several liquids found in Table 5 which has no Raman bands in the favourable region for the complimantary filters used ( blue-violet and green-yellow filters) by Raman. So unless we could find the archives available if any of their original observations related to Raman and Krishnan ( 1928a) we can not confirm the Raman’s claim that it is possible to observe Raman scattering in all liquids with sunlight and complimentary filters in this paper.