SENSITIVE PHOTOMETER USING A MODULATED LIGHT FLUX AND ITS APPLICATION IN A URANIUM FLUORIMETER*)

1. PHOTOMETER

The photometer described has a sensitivity of \(10^{-10}\) lumens, but it represents only one part of a fluorimeter intended for determining very small quantities of uranium. The sensitivity of this fluorimeter is such that it makes it possible to establish the presence of uranium in an amount of about \(5 \cdot 10^{-10}\) g.

The photometer consists of a photomultiplier enclosed in a special housing and connected by a cable to the rest of the photometer, which includes an amplifier, a measuring instrument for reading the value of the measured light quantity, and a power supply for the amplifier. The power supply for the photomultiplier is located separately. This design makes it possible, during measurements, to place the photomultiplier at any distance from the rest of the instrument.

When using a photomultiplier as the sensitive element of the photometer, it was necessary to overcome difficulties arising from the appearance of noise in the latter.

In the investigation it was established that the noise level per unit of a specified frequency band increases as the frequency decreases. To reduce the noise in the photomultiplier, the light source is supplied with alternating voltage at a frequency of 50 c/s and, consequently, produces an alternating light flux with a frequency of 100 c/s.

At this frequency the noise level of the photomultiplier is low and does not interfere with the high sensitivity of the photometer.

The photometer uses a three-stage amplifier with stabilized negative feedback. The gain of each stage is about 40. Selectivity is achieved by introducing into the circuit a single-section T-shaped filter, connected between the anode and the grid of the second stage.

Thus, strong negative feedback is obtained at all frequencies, except the resonant frequency of the filter, at which the amplifier’s full gain is practically obtained.

) C. D. Florida and C. N. Davey, J. Sci. Instr. 30*, No. 11, 409 (1953).

The filter passes a frequency band of 10 cps with half-amplitude. This solution is a compromise between the contradictory requirements of a narrow pass band for a low noise level, the necessity of taking into account the influence of changes in the mains frequency on the frequency of changes of the light flux, and tolerances on the parameters of the elements making up the circuit. The sensitivity is changed by switching the value of the input resistance of the first stage.

The measuring instrument at the output is a combination of a magnetoelectric-system instrument with a rectifier. The measuring limit of the instrument is 25 volts rms. The power supplies for the amplifier and the photomultiplier do not differ in any essential way in their design from analogous devices ordinarily used.

With the aid of the photometer it is possible to detect a light flux down to \(10^{-10}\) lumens; in this case the signal level is approximately equal to four times the noise.

The sensitivity of the instrument is such that quantities of light can be measured which would cause blackening of a photographic plate in several hours. The response time (\(\sim 1\) sec.) is limited mainly by the instructions to the measuring instrument.

A further increase in sensitivity can be achieved in three ways:

1) by using a phase-sensitive rectifier, which could serve as an obstacle to disturbances from noises that are not in phase with the oscillations of the light flux;

2) by improving the selectivity of the amplifier by a sharper cutoff of frequencies outside the 10 cps pass band;

3) by reducing the magnitude of the dark current: first, by rationally choosing the voltage for each stage of the photomultiplier so that its noises would be equalized in magnitude with the noises of the amplifier, and, second, by choosing the appropriate temperature of the surrounding medium, since it is known that the magnitude of the dark current increases as this temperature rises.

2. FLUORIMETER

Determination of the amount of uranium is of great importance in the analysis of ores. The fluorescent method of determining the amount of uranium in ore is well known; it is based on the luminescence of a mixture of uranium ore and sodium fluoride when irradiated by a source of ultraviolet radiation. The intensity of this luminescence is almost directly proportional to the amount of uranium contained in the sample.

The general appearance of the fluorimeter is shown in Fig. 1, and its optical diagram in Fig. 2.

The source of ultraviolet radiation \(A\) is located in the housing \(B\), below which the sample \(F\) under investigation is placed.

To eliminate errors of the instrument which may arise due to the nonuniform distribution of brightness in the source \(A\), quartz optics \(C\) and \(E\) are used. The filter \(D\), placed in the path of the rays, selects from them radiation with a wavelength of 3650 Å.

The ultraviolet radiation reflected from the sample and the visible light produced as a result of the fluorescence of the sample pass through the combination of filters \(G\) and \(H\), which transmits visible light with wavelengths from 4600 to 5600 Å. This region corresponds to the greater part of the visible fluorescence spectrum of uranium.

The system of lenses \(J\) and \(L\) projects the image of the fluorescing sample onto the cathode of the photomultiplier.

The samples under investigation are placed on a special support in the lower part of the instrument. The inner cavity of this part of the instrument is made of gra-

phite, which gives insignificant fluorescence while strongly absorbing scattered ultraviolet radiation.

When the sample holder is removed from the instrument, the opening in it is automatically closed in order to prevent significant light flux from outside from reaching the photocathode of the photomultiplier, since otherwise fatigue of the photomultiplier may occur (because of the large current flowing through it), and its sensitivity is restored no sooner than after an hour.

The photomultiplier window is placed opposite lens \(L\).

The photomultiplier, the amplifier, and the voltage divider for the photomultiplier are located in the upper part of the instrument, as shown in Fig. 1.

Fig. 1. \(A\)—ultraviolet source; \(B\)—photomultiplier with amplifier.

The instrument is used to determine the amount of uranium contained in the sample under investigation by comparison with a sample containing a definite amount of uranium. To prepare a sample, uranium is deposited by evaporation onto a platinum plate; a definite quantity of a mixture containing sodium fluoride is added to it. This mixture is melted and then cooled. Six plates with samples are inserted into the fluorimeter, and each sample is measured in turn by the instrument. Since the surface of each sample is not ideally diffuse, and also for a number of other reasons, the reproducibility of measurements of the same sample does not exceed 3%.

Measurements of two identically prepared samples give a divergence in the values of up to 10%. From this the minimum value of the amount of uranium that is measured by the instrument can be obtained.

Below are given the average figures obtained as a result of measurements carried out on 10 fluorimeters.

The first line gives the readings of the instrument with the lamp switched off. These readings can serve as a measure of the noise level of the photomultiplier.

The second line gives the readings obtained with the lamp switched on and with no sample in the instrument. In this case the vessel in which the sample is placed during measurements is covered with colloidal graphite. Thus these readings make it possible to judge the fluorescence of the very inner part of the instrument. The third line gives the instrument reading under the condition that only the mixture is placed in the sample vessel, while the uranium itself is absent.

Fig. 2.

The fourth line gives the reading of the instrument when the mixture with uranium is placed in it.

It is then clear that the minimum amount of uranium reliably detectable with the aid of the fluorimeter is determined as a quantity corresponding to 10% of the reading of the instrument when measuring the fluorescence of the mixture without uranium, multiplied by 3. This quantity, as is seen from the readings given above, is equal to \(5\cdot 10^{-10}\) g.

It should be noted that the 10% error in measurements of the mixture without uranium is, in magnitude, almost equal to the results of measurements with the lamp switched on. This indicates that the error in measuring the fluorescence of the mixture without uranium is almost equal to the magnitude of the fluorescence of the innermost part of the instrument itself.

M. Kannik