The following selection is from: T. Thorne Baker, The Spectroscope and its Uses in General Analytical Chemistry. William Wood and Co. New York (1923).
 

Copyright © 1998 Richard A. Paselk

 

Baker - Table of Contents

 

 

CHAPTER VIII

 

Absorption bands in inorganic and organic substances, and their relation to chemical constitution - The spectrophotometer - Spectrographic methods of measurement of absorptions

 

. . .

 

150

 

. . .

 

158 THE SPECTROSCOPE

. . .

 
The study of absorption bands is of importance in physical chemistry for investigating molecular structure, and as an auxiliary aid to certain types of organic analysis. Laboratory methods of measuring absorptions both optically and photographically will now be described.
 
Pronounced absorption bands in the visible spectrum can, of course, be seen and approximately measured, but except as a rough qualitative test a visual method is of little value. The spectrophotometer provides a means of comparing one light source
 
 
 

SPECTROPHOTOMETERS

159

 
with another - a standard source - and by comparing the two spectra in successive regions and plotting the results against wave-length as a curve, the absorption may be accurately represented. The beam of light from the standard source has to be 'cut down' in each region until the two spectra appear of equal brightness, and this is done by means of crossed Nicol prisms. A ray of light transmitted by a Nicol prism is plane polarized, the ordinary ray being totally reflected, and the extraordinary ray passing through, the plane of polarization being perpendicular to the principal plane. If the plane polarized light be now passed through a second Nicol, the intensity of the transmitted light will vary

as the latter is rotated. If the principal plane of the second Nicol be inclined at an angleto the plane of polarization of the light passing through the first Nicol, the intensity of the transmitted light will be I cos2, I being the intensity of the incident light, so that when = 0°, the transmitted light is I, and when = 90°, it is zero.
 
A photometer depending on crossed Nicols, designed by Nutting, may be used with any spectrometer, and is made of standard dimensions by Hilger for use with the constant deviation instrument. The general arrangement is seen in Figs. 78 and 79. Here L is the light source; the writer uses a Pointolite lamp, which has proved very satisfactory, provided it be lighted up several minutes before it is required, so that the metal sphere assumes a fixed position.
 
 
 
160 THE SPECTROSCOPE

Two beams of light are cast by means of two achromatic deviating prisms, P1 and P2, upon two Nicol prisms, N1 and N2. The beams emergent from the Nicols pass one direct to the lens A, the other to a rhomb, from which it is reflected to A in line with the first beam. The beam from P2 then passes through the analyzing Nicol N3, which is rotated by means of a large divided disc C, so that if a solution of a dye, for instance, be placed in front of P1, and the intensity of the beam thereby diminished at any particular wave-length, the Nicol N3 can be rotated until the intensity of the beam from P2 is equal to that from P1. The two beams emergent from the Nutting photometer are concentric, and are focussed upon the slit S of the spectroscope.

The eye-piece of the spectrometer E is fitted with shutters, so that only a narrow band of the spectrum need be viewed at one time; the images appear thus as tripartite bands, the standard beam (from P2) being in the centre, and the controlled beam (from P1) being above and below.
 
By passing through the spectrum, and matching each narrow region consecutively the absorption of any coloured body can be computed, as compared with a standard. For the measurement of colours, such as dyes, the standard should be the screened 'daylight' acetylene described on p. 141, or else a metal filament lamp, run from accumulators at a constant voltage obtained by the use of a potentiometer, with a 'daylight' filter fixed in front. This is con-
 
 
 

COMPARING LIGHT SOURCES

161

 
veniently arranged in the manner indicated in Fig. 80. A metal filament lamp L is run from an accumulator (12 Volts is sufficient); a well-made and carefully graduated voltmeter G is placed in shunt with the lamp. In series with the battery is a small regulating resistance or potentiometer P, so that the resistance of P can be regulated until the voltage in the lamp circuit is 11.5. The lamp is thus under-run, and will last for a very considerable time, giving a constant amount of light if always run at exactly 11.5 volts. F represents a filter, which compensates for the richness of the green, yellow, and red in the incandescent filament spectrum, and gives an approximation of daylight. It may be prepared by the use of two 1-centimetre wide glass cells, filled with the solutions given in the appendix.
 
When working with a continuous light source the slit width should be as small as possible, in any case not more than 0.1 millimetre.

For comparing two light sources, A and B (Fig. 81), the dimmer of which should be in the B position, each light must be examined by itself, and its position adjusted so that its image is central and comes within the object glass. Then letting through light from both sources the images should be central in the object glass. Both sources should
 
 

162 THE SPECTROSCOPE
 
be equidistant from the photometer box. The ratio of intensity of the light sources B/A will be found by the value 1 / antilog D, where D is the density reading when the sources are equidistant from the photometer box. By means of the shutter eye-piece the value of 1 / antilog D can be found for
successive regions throughout the spectrum.
 
The colours reflected from fabrics, samples of paper, pigments and so on can be compared or measured against the standard white light by means of reflection. A piece of fabric is placed on a slab of plaster of Paris, and over the fabric and the white slab is placed a piece of dead black paper or a metal plate, with two apertures about 2.5 centimetres in diameter and 38 millimetres apart, the fabric showing through one aperture and the white plaster through the other. This is placed in front of the photometer box, so that the plaster of Paris is opposite the further aperture and the fabric opposite the nearer aperture, and so that the slab is sloping upwards at an angle of 45°. A brilliant light source is placed immediately above the mid-point between the apertures, and the measurements are then taken on the density scale. The percentage reflection of the fabric for each region examined will be given by
 

100 / antilog D'

 
We come now to the methods of measuring absorption spectra by means of photography; characteristic absorption curves are used, made in the ultra-violet for general convenience, a quartz spectrograph with wave-length scale being employed. A series of successive photographs is taken, the same exposure being given for each one, with a cell in front of the slit containing the solution or liquid under examination, a different concentration of the solution,
 
 
 

ABSORPTION SPECTRA

163

 
or a different thickness of liquid, being used in each case, these differences being constant. Very convenient for the purpose is the absorption tube of Professor E. C. C. Baly, shown in Fig. 82, a graduated tube with a bulb reservoir, into which an inner tube slides through a rubber ring.
The ends of the outer and inner tubes are fitted with quartz discs, and the tube is placed end on to the slit. As the inner tube is drawn out the thickness of liquid traversed by the rays is increased, and the limits of absorption are measured in each photograph and plotted against the thickness of liquid. Baly recommends lining the inner

tube with black paper in order to stop reflections from the glass wall. He also suggested using logarithms of the concentrations or relative thicknesses of liquid in place of the values themselves.
 
A series of photographs of the dye tartrazin are shown in Fig. 83, each spectrum being taken through an increasing width of solution from .5 to 8.5 centimetres. It will be seen that by drawing carefully a freehand curve over the limiting positions of the spectra, a characteristic curve is obtained of the absorption of the dye. The limits of the absorptions are more usually set out on squared paper against the logarithms of the thickness (or concentration) of the substance.
 
As a source of light the author finds an oscillatory spark discharge between aluminium or iron electrodes the most
 
 

164 THE SPECTROSCOPE
 
convenient, using a spark gap of about 1 centimetre, the exposure with a 10-inch quartz spectrograph being about three seconds for each photograph. Gerlach and Koch* have suggested an iron wire .031 millimetre diameter, exploded by the passage of the discharge from a high potential condenser across a gap of 26 millimetres. The light given

is always constant, and depends on the capacity, the gap, and the diameter of the wire.
 
Amongst the more recent sources of ultra-violet light may be mentioned the tungsten arc. Luckiesh** used an enclosed tungsten arc burning in commercial argon, containing 20 per cent. of nitrogen, the arc. voltage varying from 28 to 16 volts, the consumption being 5 to 8o watts. The best results were obtained when using a current of 1 ampere upwards. The spectrum between 3000 and 4000 A.U. is practically con-
 
*Berichte, 1922, 55B, 695-7.
 
**Journ. Franklin Institute, 1921.
 
 
 

ABSORPTION SPECTRA

165


tinuous, and somewhat fluted in character; beyond 3000 A. U. it consists of closely adjacent lines.
 
A spark discharge between aluminium electrodes placed 2 millimetres apart in running- distilled water, an alternating

Supply of 50,000 volts being employed, is stated to have provided a continuous ultraviolet spectrum.*
 
Judd Lewis's method of measuring absorption spectra combines the use of the spectrograph and the spectrophotometer. A number of pairs of spectra are taken, each pair consisting of a normal spectrum of the light source
 
* Schweizerische Elecktrotechnische Zeitschrift, 1919
 
 
 
166 THE SPECTROSCOPE
 
and a spectrum of the same light after it has passed through a layer of the substance under examination. The absorption spectra of the substance are of the same thickness throughout, the exposure only being varied, and in each pair the same exposure is given to the control - i.e., the light source - and the light source screened by the substance. The intensity of each normal band is adjusted to a desired value by mean of the spectrophotometer; for example, in one pair the whole of the normal band was adjusted to have an intensity equal to 22.16 per cent of the original, and therefore any particular line in the spectrum would have an intensity of 22.16 per cent of the original. The lines in the absorption spectrum that just coincide in intensity with lines in the control are then observed; suppose these were = 2597 and =2,891 A.U., then for this particular case the transmitted light between these limits is 22.16, and the absorbed light 87.84 per cent. Similarly points of equality are determined in each of the other bands, and one is thus enabled to work out a series of corresponding points for different light intensities, which, joined up, give a characteristic curve.
 
One of Dr. Judd Lewis's results in seen in Fig. 84.
 
A somewhat crude, but very simple, way of obtaining by photography an absorption curve of a dye or substance is to give one exposure, using a wedge of neutral grey glass mounted with a similar wedde of plain glass, so as to be plane parallel-sided. The opacity of the wedge may vary from 1 to 1,000 or 1 to 10,000 or more. L. H. Davidson and the author* made a number of analyses of chemical indicators in this way, showing the change in colour in a dye when present in an acid or alkaline solution - i.e., in a solution the hydrogen ion concentration (pH) of which was below or above 7. An example is seen in Fig. 85, where the absorption of thymol blue is seen: (A) in its
 

The Photographic Journal, August, 1922.

 
 
 

MICRO-SPECTRO OCULARS

167

 
normal condition; (B) with a pH of 1.8 (acid reaction); and (C) with a pH of 9.6 (alkaline reaction). Absorption spectra can be obtained in a similar way in the ultra-violet spectrum by substituting for the wedge a small sector, as suggested by E. Belin, which is driven by a small motor during the exposure.
 
A spectroscopic ocular is sometimes used in microscopic examination. Most of these oculars are fitted with a small

direct vision prism train, mounted just above the eye-lens of the ocular, the slit being placed in the plane of the eyepiece diaphragm. A comparison prism and a photographed scale are generally provided, so that an absorption can be roughly measured or estimated. In order to provide for cases where the coloured image under analysis is not of sufficient dimensions to flood the whole slit, some device for shortening the slit should be present.
 
Spectroscopic oculars, fitted with wave-length scales of considerable accuracy, are manufactured, notably by Zeiss and Leitz, the scale itself being adjustable by means of a micrometer screw, so that if monochromatic light, such as that of a sodium flame, be employed, the scale can be
 
 

168 THE SPECTROSCOPE
 
adjusted so that the D line is in accurate position, when the whole scale will automatically read correctly. Chamot,* however, suggests that such instruments are usually too crude to be of value to the chemist, and gives methods of calibrating an ocular, or using a wave-length scale prepared similarly to those already described in Chapter IV. He also describes the Andrews cell for measuring absorption spectra with the microscope, in which the thickness of liquid in a cell can be varied, so that the position of maximum intensity of an absorption band can be determined correctly i.e. by observing the situation of the vanishing point after repeated dilutions.
 
An illustration of the Leitz micro-spectroscope is seen in Fig. 86.
 

. . .

 

* 'Chemical Microscopy,' p. 135

 

 

 

FLUORESCENCE

169

 

. . .