Fig. 15—Illustrating the principle of the Lippich polarizer for double field.
The three-field alternative has an even greater sensitivity but, if the two outer fields are not polarized exactly in the same direction, it loses its advantage, since two balance-points are obtained. Under ordinary laboratory conditions, the simpler two-field balance is preferable.
The angle between the polarizations of the two fields is known as the half-shadow angle. Theory shows that, the smaller the half-shadow angle used, the greater is the sensitivity of the instrument. However, the smaller this angle, the closer the two sides of the field are to extinction at the balance point, and the less light there is available to judge the balance. Since the sensitivity also depends on the light intensity, when the light source is as bright as can be obtained, a compromise is required on the half-shadow angle between the loss of sensitivity due to too large an angle and the loss due to too little light. In practice angles of 1° to 10° are used. In the saccharimeter described later an angle about 7° to 8° has been found a good compromise for accuracy and available light.
There are several methods used to make polarizers that give a split field.
The simplest to understand are forms of the Jellet-Cornu polarizer. Imagine a polarizing prism with a narrow-angle V taken longitudinally from its centre.
The two separate prisms are then cemented together to give two adjacent polarizers with a fixed half-shadow angle, the angle of the V cut. Another polarizer with a fixed half-shadow angle is made from two long natural rhombs of calcite; this is one example of the use of the rhomb itself, instead of a prism, to separate the ordinary from the extraordinary ray.
The most common method of obtaining the split field is the Lippich polarizer, shown in Fig. 16. In front of the main polarizer is placed a smaller polarizer covering half the field. This is rotated through the half-shadow angle from the main polarizer. This rotation changes the direction of polarization across this half of the field and also slightly reduces the
26 OPTICAL INSTRUMENTS intensity. The reduction of intensity affects the posit- ion of the setting slightly; it is no longer exactly midway between the angles at which the two fields extinguished. The small polarizer is also tilted slightly so that the observer does not look along its face (and hence see a broad band separating the two fields) but sees only a sharp edge. When the triple field is used, two such small polarizers are employed.
The Lippich system has the advantage that the half-shadow angle is adjustable and can be altered to suit the intensity of the illumination. When it is altered, however, there is an alteration of the zero point of the analyser. But, in the Bates Fric sacchar- imeter a special set of gears is fitted so that, when the large polarizer is rotated to change the half- shadow angle, the analyser is rotated by the amount required to correct for the change in zero point.
The need to use monochromatic light with a polarimeter was a great practical disadvantage when it had to be obtained by feeding metal salts into a flame. Now spectral lamps, such as sodium and mercury, are readily available and easy to use. Most of the visible light from the sodium lamp is in a pair of orange lines, called the D line, and it is usually used with a yellow filter to cut out the light from a fainter pair of green lines. Sodium lamps have a fixed, rather low brightness.
Mercury lamps can be obtained with a wide range of brightness. The bright, high-pressure lamps, however, give broadened spectral lines and this can cause errors. Thus for polarimetry, a low-pressure mercury lamp is required with a filter to separate out the green e line.
The Saccharimeter
Formerly a saccharimeter was considered to be a polarimeter graduated not in angular degrees but in relative concentration of sugar or degrees of sugar oS. However some polarimeters today have both angle and sugar scale graduations and modern automatic polarimeters can be arranged to display the rotation in any chosen unit. This applies equally whether the sample rotation is compensated by turning the analyser prism, or by placing a suit- able amount of optically active substance, such as a piece of quartz, or a glass rod in a magnetic field, immediately before a fixed analyser.
Therefore it seems best to describe a polarimeter with a sugar scale merely as a sugar polarimeter and to confine the term saccharimeter to an instrument which by virtue of its principle of operation should be used only on sucrose solutions.
As a result, it is becoming common, therefore, to reserve the name saccharimeter for an instrument that uses quartz wedges for compensation.
When a polarimeter is designed specifically for use with sucrose solutions, that is, as a saccharimeter, it becomes possible to adopt an alternative means of eliminating the ill-effects of rotatory dispersion and the need to use rather Fig. 16—Showing the c o n s t r u c t i o n o f Lippich polarizer for
double field.
OPTICAL INSTRUMENTS 27 low intensity spectral lamps. (When white light is used with a simple polari-
meter, no extinction is obtained since different colours are rotated different amounts.) By chance, quartz has practically the same rotatory dispersion as sucrose solution. The rotation produced by the sugar is balanced out by a quartz compensator, wavelength by wavelength, and extinction can now be obtained with white light. The extinction is improved further if the blue end of the spectrum, where the dispersions match worst, is not used. The light is therefore filtered through a bichromate filter (15 mm thickness of a 6 per cent solution of potassium bichromate) or a plate of glass having similar trans- mittance characteristics.
The quartz compensator consists of two wedges of quartz of equal angle mounted so that one can be moved past the other, as shown in Fig. 17. The pair of wedges then acts as a parallel-sided plate of quartz of adjustable
Fig. 1 7—Showing the construction of single wedge quartz compensation.
I Dextrorotatory system. II Laevorotatory system.
thickness and it gives a controlled rotation to the light going through it.
This rotation is never zero, since the plate formed by the two wedges can never be zero thickness. To obtain zero rotation, the wedges are "backed off"
by a fixed plate of quartz of the opposite hand; i.e. if the wedges are made of left-handed quartz, this plate is right-handed. This system, known as a single-wedge compensator, is the one most commonly used in commercial saccharimeters.
The optical system of a saccharimeter is shown in Fig. 18. The lens a condenses white light from a clear filament lamp, with a ground glass disc in front of it, on the aperture in b; the light is brought to a focus at the objec- tive of the telescope by a lens c; d is the polarizer (with fixed half-shadow angle); e is a stop to limit the size of the light beam and / a glass protecting
Fig. 18—Illustrating the parts of a saccharimeter.
plate. The sugar solution under examination is contained in the cell g; h is a second protecting plate; i and m are stops for cutting out stray light; j, k, and I make up the single-wedge compensator; n is the analyser; o the objective of the viewing telescope; p a field stop in the focal plane of the eyepiece;
and q and r form the eyepiece of the telescope.
Two separate optical parts of the instrument are thus in dust-proof enclosures, protected from juice splashes by the optically inactive protecting glasses. The whole system is mounted in a rigid metal tube which is held horizontal on a stand. Formerly, saccharimeters were supplied for both
200 mm and 400 mm sample cells but the former has almost disappeared from the modern sugar-mill laboratory; the longer cell is needed for such solutions as bagasse extracts, which are of low optical activity. On the latest models (Fig. 19) the lamp housing is built on as an extension to the instrument so that the light source is fixed in relation to the instrument and is held in its correct position. With the Schmidt and Haensch instrument a small focusing disc is provided. This is placed at the end of the trough towards the analyser, and if the light be correctly placed, a sharp image of the filament of the lamp will coincide with the horizontal diameter marked on the disc. The ground glass disc with which the lamp is fitted should, of course, be removed when making this test.
The scale is usually graduated from —30 °S through zero to + 1 0 5 °S (with extended graduations at both ends), or occasionally, from —150 °S to +150 °S. The angular rotation that corresponds to 100 °S depends on the length of cell used, the normal weight specified for the instrument, and on the wavelength for which the rotation is measured.
Fig. 19—Illustrating a Schmidt and Haensch saccharimeter.
The scale is viewed through a low- power microscope, being illuminated by some of the light that has been deflected from the main path. Two types of scale are now in common use. The type em- ploying a vernier is illustrated in Fig. 20.
It will be observed that the main scale is graduated at intervals of one degree of sugar. A centre-zero vernier is provided, one side for positive readings and the other for negative, both divided to read to 0.1 °S. In Fig. 21 is illustrated the scale employed in the current Schmidt and Haensch saccharimeter. The scale moves vertically as opposed to the former hori- zontal scale and the main scale is divided into 10 °S divisions. A fixed engraved scale
Fig. 20—Illustrating the double vernier scale of a saccharimeter.
Reading 73.4° S.
OPTICAL INSTRUMENTS
Fig. 21—Direct reading scale employed in the Schmidt and Haensch saccharimeter.
Reading 66.3° S.
of 10 °S subdivided into 100 divisions is also provided whereby the reading may be made directly to 0.1 °S and estimated to 0.02 °S. The zero adjustment for each scale is carried out as follows:—
In the vernier type scale the field is set to the balance position with the trough empty and the zero of the vernier is adjusted, with the key provided, to the zero of the main scale. With the direct reading scale the zero on the movable scale is set to the zero on the fixed scale first and the field is then balanced for equal intensity by the knurled knob situated at the base of the analyser housing. At the balance point the two halves of the field should appear identical. The appearance of a difference in colours at the balance point, one side appearing yellowish and the other nearly white indicates the need for internal adjustment. This should not be attempted by unskilled technicians.
Effect of Illumination
As stated earlier, the rotatory dispersion of sucrose solution is close, but not exactly equal to that of quartz, the sugar having the greater dispersion.
Since the difference in the two dispersions is greatest for blue light, the quartz-wedge saccharimeter is designed for use with white light filtered to remove the blue end of the spectrum. A movable glass filter, that approx- imates closely to the characteristics of a six per cent potassium bichromate solution of 15 mm thickness, is now usually built into the saccharimeter.
This filter transmits red, orange, and yellow light but absorbs the rest of the spectrum; the transmitted radiation has a mean wavelength of about 6000 A.
If white light is used without a filter, a saccharimeter will give readings that are in error by about +0.12 °S at the 100 °S point. Only when the solution is coloured and acts as its own filter should the filter be omitted.
If a sodium lamp is used with a quartz-wedge saccharimeter, there is again a small error, now about 0.03°S at 100 °S, whether a filter is used or not.
Automatic Polarimeters
The modern tendency in optical measuring instruments is to replace the eye by some photoelectric detector. Such instruments do not require as highly skilled an observer and are less fatiguing to use. In addition, the 29
30 OPTICAL INSTRUMENTS
results obtained are more reliable and often more accurate and, being in the form of an electrical signal, can be recorded by means of the large variety of data-recording equipment now available. If calculations are made on the results, this is done by connecting in the appropriate calculating circuits and the result is obtained with very little delay. An automatic polarimeter is normally used with a flow-through cell so that samples can be readily introduced and flushed away; many installations use an automatic sample feeder which introduces samples to the instrument at regular intervals of, say, 60 seconds and actuates the read-out device. Certain instruments allow the polarisation to be recorded continuously as the sample flows through the cell. However, the precision of the measurement usually surfers seriously as a result of striations.
A photoelectric polarimeter could be made by using a conventional split-field polarizer and taking the light from each half of the field to a separate photocell. At the balance point, the two electrical signals from the photocells would be equal. Such a system would give continuous d.c. signals from the photocells and would require d.c. amplifiers, which are notoriously more unreliable and more unstable than a.c. amplifiers.
Modern automatic polarimeters, therefore, use a.c. balancing. Instead of a field split in space and two photocells, one photocell is used with a field
"split in time". The plane of polarization changes backwards and forwards between the two positions it would have for the split field, either in jumps or continuously. The electrical signal from the photocell then consists of a d.c. background superimposed on which is an alternating current of the same frequency as that at which the polarization is being switched. This a.c. component of the signal is an error signal: it becomes zero at balance.
The instrument balances itself by using the error signal to drive the balancing system; when the error signal vanishes, this drive stops. It is thus a servo-system.
To oscillate the direction of polarization one of three methods is used.
The first, employed by Schmidt and Haensch and also by Perkin Elmer in their automatic saccharimeters and polarimeters, uses a synchronous motor coupled to the polarising prism, which is caused to rotate backwards and forwards so that the direction of polarisation oscillates. The second, used by the National Physical Laboratory (N.P.L.) in the standard polarimeter that they use to calibrate quartz control plates and also by Jobin-Yvon, has a rotating plate, around the edge of which is a series of holes, each covered by a quartz plate. These plates are of equal thickness and alter- natively left- and right-handed. As the plate rotates, these quartz plates pass in turn in front of the polarizer to give a direction of polarization that switches first to the left, then to the right. Hilger and Watts use a similar system consisting of a vibrating reed supporting and oscillating two pieces of quartz side by side, one left-handed and the other right-handed, and oscillating them across the light beam.
The third method makes use of a Faraday cell and is employed by Zeiss and Jouan. It is also used in the polarimeter designed by N.P.L. which is made by Thorn Bendix. As stated earlier, if a glass rod with light passing through it is placed in a magnetic field, the field being in the direction of the light, the plane of polarization of the light is rotated by an amount that de- pends on the type of glass and the strength of the magnetic field. Very dense flint glasses give the largest rotation. The sense of rotation depends on the di- rection of the field and, if this is alternated, the rotation alternates. To give an oscillating direction of polarization the glass rod is enclosed in a solenoid through which passes an alternating current.
OPTICAL INSTRUMENTS 31 The polarimeter is balanced in one of three ways, corresponding to the
above methods of modulation: Either a conventional analysing prism is rotated (Hilger and Watts, Zeiss, Perkin Elmer), or compensating quartz wedges are driven up and down (Schmidt and Haensch, Jobin-Yvon), or a d.c. Faraday cell is used as a compensator to balance the rotation due to the sample (Bendix, Jouan). The rotating analyser is turned to balance by a motor that is driven by the amplified error-signal, and the rotation can be read from an angle scale by an electrical method e.g. by using a potentiometer.
In the Hilger and Zeiss instruments a digital output of the rotation in sugar degrees is obtained using a shaft encoder. The compensation quartz wedges are driven to balance by a motor in a similar fashion to the rotating analyser, and the sugar value of the rotation of the sample is read from a linear scale.
(The Schmidt and Haensch instrument uses moire gratings, for example).
In the Bendix polarimeter, the current in the compensating Faraday cell is a measure of the rotation.
In the Hilger and Zeiss instruments, the servo motor drives the rotating analyser at a constant rate, and so the time taken to reach a balance depends on the range to be traversed: for instance if a cane juice sample reads 90 °S, the instrument will take twice as long to give a reading if the previous sample read 70 °S than if it had read 80 °S. In the Bendix instrument the balancing is done electronically, not mechanically, and equilibrium is approached at an exponentially decreasing rate. Therefore the time taken to reach balance depends only slightly on the range to be traversed. The accuracy of the final setting depends, however, on the time allowed for the equipment to reach a balance; the longer the time permitted, the higher the accuracy that can be obtained, until the instrument's limit of accuracy is reached. To be precise, the accuracy increases as the square-root of the time, so that, to double the accuracy, the instrument would take four times as long to reach a balance.
The Hilger M560, Zeiss OLD 3 and Bendix-NPL 700 A automatic polari- meters have been built to comply with the Australian Standard Specification for an Automatic Sugar Polarimeter, AS K157 - 1968. Australian Standard Specification AS K157 - 1968 "Automatic Sugar Polarimeter" covers the requirements which are considered desirable for an automatic sugar polari- meter suitable for use in the analysis of cane juice and sugar products in Australian sugar factories. It was approved by the Standards Association of Australia in 1968. They have a range of —120 °S to +120 °S with a digital readout to 0.01 °S, and are suitable for cane juice, raw sugar and molasses.
Most of the other commercially available automatic polarimeters are built for the European beet sugar industry, and have a range of 0 to 30 °S or 0 to 100 °S, with a readout to the nearest 0.05 °S or 0.10 °S.
The Hilger and Zeiss instruments are basically automated verisons of conventional polarimeters. Their components are shown in Figs. 22 and 23, respectively.
The Hilger M560 polarimeter normally uses a mercury vapour lamp and an absorption filter to provide monochromatic radiation of 546 nm.
However, it can be fitted with a sodium light source. The light is linearly polarised by a calcite prism of the Lippich type. A small oscillating biplate of left- and right- rotating quartz modulates the beam, which passes through the sample, normally contained in a 200 mm tube, and on to the analyser prism and the photomultiplier. Jacketed flow-through cells are available, and the tube trough is also provided with a water jacket for more effective temperature control. The "out of balance" signal from the photomultiplier is amplified and used to drive a servo motor which rotates the analyser prism until balance is reached. The shaft of the servo motor also drives the electro-