Method of and apparatus for spectroscopically analysing samples.

0001The invention relates to a method of analysing one or more samples in succession using a single beam wavelength scanning spectrophotometer comprising the steps of <ul id="ul0001" list-style="none"><li>1) passing radiation from a radiation source to a radiation detector over a defined path which includes a wavelength selective element which is capable of being scanned across a selected wavelength range,</li><li>2) storing a first set of signals representative of the radiation falling on the radiation detector, the first set of signals being spread across the selected wavelength range to form a first spectral response,</li><li>3) inserting a sample in the defined path,</li><li>4) producing a second set of signals representative of the radiation falling on the radiation detector when the sample is inserted in the defined path, the second set of signals being spread across a selected wavelength range to form a second spectral response.</li><li>5) calculating the absorbance or transmittance of the sample at desired wavelengths using the first and second spectral responses, and</li><li>6) repeating steps 3) to 5) for each successive sample.</li></ul>

0002The invention further relates to a single beam wavelength scanning spectrophotometer comprising means for passing radiation from a radiation source to a radiation detector over a defined path which includes a wavelength selective element which is capable of being scanned across a selected wavelength range, means for storing a first set of signals representative of the radiation falling on the radiation detector, the first set of signals being spread across the selected wavelength range to form a first spectral response, means for inserting a sample in the defined path, means for producing a second set of signals representative of the radiation falling on the radiation detector when the sample is inserted in the defined path, the second set of signals being spread across a selected wavelength range to form a second spectral response and means for calculating the transmittance or absorbance of the sample at desired wavelengths from the first and second spectral responses.

0003The detector output of a single beam spectrophotometer varies rapidly with wavelength as it depends on the combined effects of the variation with wavelength of the source energy, grating efficiency, mirror reflectivity and detector photo-sensitivity. This means that a spectrum can only be obtained from simple single beam instruments by the tedious process of setting a baseline detector output with either a clear optical path or when analysing samples dissolved in a solvent a cell containing the solvent (blank or reference solution) in the optical path, moving the sample cell into the beam, measuring the detector output ratioing the detector outputs to calculate absorbance or transmittance of the sample, changing the wavelength, and repeating the process for each data point required.

0004Linking a computer to the spectrophotometer can take the tedium out of this process and produce the data faster than any operator. The computer is simply told the start and stop wavelenghs of the scan and the blank is placed in the beam. The computer then commands the spectrophotometer to drive to the start wavelength and to store the detector output in its memory. It then commands the spectrophotometer to go to the next wavelength and the process is repeated until the spectrum of the blank solution (and the cell containing it) is held in its memory. The blank solution is then emptied and the cell refilled with the sample solution and placed in the beam. The measurement sequences are repeated and at each wavelength the stored detector output when using the blank is ratioed by the computer with the new reading to yield the nett absorbance or transmittance of the sample which is then displayed and plotted.

0005This principle of storing a spectrum of the blank and ratioing it with subsequent sample spectra has, with the reducing costs of memory and processing power, now become a practicable proposition for purpose-built single beam scanning instruments. Such an instrument is currently manufactured and sold by Pye Unicam Limited under the type reference PU8600. However, the technique does have its limitations as it demands a high level of stability with time, both of the instrument and of the sample matrix. Little is gained if the instrumental readings drift between the time when the blank was measured and when the last sample is run. Similarly, if blanks and samples exhibit absorbance changes in time due to settling of particulate matter, variations in ambient temperature, sensitivity to light, etc., then the recorded differences will not be true. All electronic circuit drift, especially when warming to operating temperature. Consequently single beam scanning instruments tend to be left powered-up night and day with only motors and sources shut-down when the instrument is not in use (reasonable in a laboratory carrying out routine analyses but unwise in research laboratories with patchy and unpredictable instrumental usage). Scanning rapidly has been the only way of reducing the vulnerability to temporal changes in the sample. However, this involves other performance compromises and any apparent throughput gains are quickly swallowed by the necessity to produce hard-copy after the scan rather than simultaneously. Therefore, single beam scanning spectrophotometers have currently been confined to applications where low to medium performance is acceptable. For the highest levels of performance the double beam instruments have been used.

0006It is an object of the invention to enable the use of a single beam scanning spectrophotometer while mitigating the effects of one or more of the disadvantages discussed hereinbefore.

0007The invention provides a method of analysing one or more samples in succession using a single beam scanning spectrophotometer as set forth in the opening paragraph characterised by the further steps of <ul id="ul0002" list-style="none"><li>7) monitoring the ambient temperature in the region of the radiation detector</li><li>8) modifying the first and/or second sets of signals in dependence on the ambient temperature when the first and second set of signals are produced.</li></ul>

0008By monitoring the ambient temperature in which the radiation detector is operating changes in detector sensitivity with ambient temperature may be compensated for and the time between calibrations of blank solution absorbance spectra increased.

0009The first and second sets of signals may be modified by passing them through a circuit arrangement having a temperature co-efficient of gain of substantially the same magnitude but of opposite sign to the temperature co-efficient of sensitivity of the radiation detector.

0010Alternatively the method may further comprise the steps of measuring the ambient temperature at a first instant immediately prior to, immediately subsequent to, or during the period during which the first set of signals is generated, storing the temperature measured at the first instant, measuring the ambient temperature at a second instant immediately prior to, immediately subsequent to, or during the period in which the second set of signals is produced, and modifying the first or the second set of signals in accordance with the difference in ambient temperature between the first and second instants in such a manner as to compensate for changes in radiation detector sensitivity with temperature.

0011In a further alternative the method may further comprise the steps of measuring the ambient temperature at a first instant immediately prior to, immediately subsequent to, or during the period during which the first set of signals is generated, modifying the first set of signals in dependence on the difference in ambient temperature measured at the first instant and a predetermined temperature, measuring the ambient temperature at a second instant immediately prior to, immediately subsequent to, or during the period in which the second set of signals is produced, and modifying the second signals in dependence on the difference in ambient temperature measured at the second instant and said predetermined temperature.

0012The selected wavelength range in step 2) may be the same as the selected wavelength range in step 4). The method may further comprise displaying the calculated absorbance or transmittance of the sample. The calculated absorbance or transmittance may be converted to concentration for display.

0013The invention further provides a single beam wavelength scanning spectrophotometer as set forth in the second paragraph characterised by means for monitoring the ambient temperature in the region of the radiation detector and means for modifying the first and/or second sets of signals in dependence on the ambient temperature when the first and second set of signals are produced.

0014The modifying means may comprise a temperature sensitive element located in the region of the radiation detector, the temperature sensitive element being arranged to modify the gain of an amplifier to the input of which the output of the radiation detector is connected in such a manner that the gain of the amplifier varies in an inverse manner to that of the output of the radiation detector with changes in ambient temperature.

0015Alternatively the monitoring means may comprise means for measuring the temperature at a first instant immediately prior to, immediately subsequent to, or during the period during which the first set of signals is generated, means for storing the temperature measured at the first instant, and means for measuring the ambient temperature at a second instant immediately prior to, immediately subsequent to, or during the period in which the second set of signals is produced, and the modifying means comprises means for modifying the first or the second set of signals in accordance with the difference in ambient temperature between the first and second instants in such a manner as to compensate for changes in radiation detector sensitivity with temperature.

0016In a further alternative the monitoring means may comprise means for measuring the temperature at a first instant immediately prior to, immediately subsequent to, or during the period during which the first set of signals is generated, and means for measuring the ambient temperature at a second instant immediately prior to, immediately subsequent to, or during the period in which the second set of signals is produced, and the modifying means comprises means for modifying the first set of signals in dependence on the difference between the ambient temperature measured at the first instant and a predetermined temperature and means for modifying the second set of signals in dependence on the difference between the ambient temperature measured at the second instant and the predetermined temperature.

0017The selected wavelength range for the first set of signals may be the same as the selected wavelength range for the second set of signals.

0018The spectrophotometer may comprise means for displaying absorbance or transmittance of the sample. The spectrophotometer may comprise means for converting the calculated absorbance or transmittance to concentration for display.

0019Embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings, in which:- <ul id="ul0003" list-style="none"><li>Figure 1 shows in block schematic form a spectrophotometer according to the invention,</li><li>Figure 2 is a plan view of the optical layout of a spectrophotometer according to the invention,</li><li>Figure 3 is a circuit diagram of one embodiment of the radiation detection unit shown in Figure 1,</li><li>Figure 4 is a block schematic diagram of one embodiment of the signal processing circuit shown in Figure 1, and</li><li>Figures 5, 6 and 7 show flow diagrams illustrating a method of sample analysis according to the invention.</li></ul>

0020Figure 1 shows in block schematic form a single beam wavelength scanning spectrophotometer comprising a radiation source 1, the radiation from which is directed to a monochromator 2 and thereafter passes through a sample cell 3 before falling on a radiation detection unit 4. The radiation detection unit 4 comprises a radiation detector device 5, such as a photomultiplier tube or a photodiode, which produces an electrical signal representative of the radiation falling on it and a temperature sensor 6, such as a thermistor, which is placed so as to sense the ambient temperature in which the radiation detector 5 is operating. The output of the device 5 is passed to a signal processing circuit 7 either directly or after modification in dependence on the sensed temperature. If the output of the radiation detector 5 is passed directly to the signal processing circuit 7 then a signal representative of the sensed temperature is also passed to the signal processing circuit 7. The output of the signal processing circuit 7 is fed to a display unit 8 which may be of any convenient form, for example, an alpha numeric display, a video display unit, or if a hard copy of the results is required a chart recorder or printer.

0021In operation the sample cell 3 is first filled with a blank solution, e.g. the solvent in which the sample is to be dissolved, the monochromator is scanned between pre-selected wavelength limits and the output of the radiation detector with the sample cell 3 inserted in the optical path is measured at intervals across the wavelength range. The measured output signals are stored within the processing circuit 7. The number of intervals is arbitrary and will depend on the amount of storage provided in the processing circuitry 7 and the wavelength resolution required. The ambient temperature in the region of the radiation detector 5 is sensed by the sensor 6 and the output of the radiation detection unit 4 may be modified in dependence on the temperature sensed. Alternatively the temperature sensor 6 may also produce a signal which is fed to the processing circuitry 7 which may then include means for modifying the received signal from the radiation detector 5 in dependence on the signal received from the temperature sensor 6. In the latter case it will be necessary also to store the signal representing the ambient temperature. As a further alternative the processing circuitry 7 may be arranged to normalise the received signal to that which would be produced at a reference temperature, in which case it is not necessary to store the signal representative of the sensed ambient temperature.

0022A sample cell 3 containing a sample is then placed in the radiation path and the monochromator 2 is caused to scan a desired wavelength range, either the same as or part of the previous range, and the radiation received by the radiation detection unit 4 is measured and processed by the processing circuit 7. The ambient temperature in the region of the radiation detector 5 is sensed by the temperature sensor 6 and the output of the radiation detector 5 is either modified in dependence on the sensed ambient temperature or a measure of the temperature is passed to the processing circuit 7 together with the output of the radiation detector 5. In the latter case either the stored values of radiation detector output are corrected for the change in temperature between measurement of the blank and sample solutions or the sample values are corrected. When the further alternative is adopted the sample values are normalised to the reference temperature using the measured ambient temperature when the sample values are generated. Subsequently the processing circuitry calculates the transmittance or absorbance of the sample from the stored and sample radiation signals and causes them to be passed to the display unit 8. If the output of the radiation detector 5 is modified in the radiation detection unit 4 then it is not subsequently corrected in the processing circuitry 7.

0023The foregoing description has assumed that the sample to be analysed is dissolved in a solvent. However, as is well known to those skilled in the art, spectroscopic analysis may be applied to solid samples, for example filter glasses, or to liquid substances which are themselves the samples and are not dissolved in a solvent. In the former case the initial scan is made with no sample in the radiation path i.e. the path is completely clear, while in the second case an empty sample cell is inserted in the radiation path for the initial scan.

0024In this way the disadvantage of changing response of a single beam scanning instruments due to changing sensitivity of the radiation detector with temperature changes can be significantly reduced. This allows longer intervals between calibration and enables the instrument to be used with a much shorter "warm-up" period.

0025Figure 2 shows the layout of a spectrophotometer having a Gillieson Monochromator. The spectrophotometer comprises a first section 801 which includes a source of radiation, a second section 802 housing the monochromator, a third section 803 in which the sample to be analysed is placed, and a fourth section 804 which includes a radiation detector.

0026The spectrophotometer shown includes a tungsten filament lamp 805 and a deuterium lamp 806 to enable a wavelength range which covers both the visible and ultra violet regions of the spectrum to be covered. A mirror 807 is movable between a first position (as shown) in which radiation from the deuterium lamp 806 is reflected onto a concave mirror 808 and a second position (shown dotted) in which radiation from the tungsten lamp 805 is allowed to fall on the concave mirror 808.

0027Radiation reflected from the concave mirror 808 is focussed on an entrance slit 809 of the monochromator 802 and passes through the entrance slit to fall onto a concave mirror 810. The radiation is reflected by the mirror 810 as a converging beam onto a diffraction grating 811 where it is diffracted and focussed on an exit slit 812 of the monochromator 802. The radiation then passes through the sample compartment 803 and falls on a detector 813 which may, for example, be in the form of a photomultiplier tube, a vacuum photo tube or a photo diode in the fourth section 804. The electrical output of the detector 813 is fed to signal processing circuitry (not shown in Figure 2) which converts it into suitable form for display. In order to cause the radiation passing through the sample compartment 803 to scan a selected wavelength range the diffraction grating is rotated about an axis perpendicular to the plane of the drawing and passing through the centre of a pivot 814. This is achieved by rotating the arm 815 which carries the grating 811 about the pivot 814.

0028A stepper motor 816 is mounted on a bracket 817 in the monochromator 802 and its output is coupled to a threaded shaft 818 by a universal joint 819. A nut 820 is pivotally mounted in an aperture in the arm 815 and rides on the shaft 818. Thus as the motor 816 steps the shaft 818 rotates and the nut 820 advances along the threaded shaft causing the arm 815 to rotate about the pivot 814.

0029The radiation beam passes over the top of the pivot 814. It is, of course, necessary to arrange the entrance and exit slits, mirror 810 and diffraction grating 811 in appropriate orientation to be in the radiation path and to enable that path to pass over the pivot 814.

0030Within the fourth section 804 the radiation detector 813 is mounted on a printed circuit board 830 which also carries a temperature sensor 831. Thus the ambient temperature within the further section, which is preferably fully enclosed apart from an aperture 832 to enable radiation from the monochromator exit slit 812 to fall on the detector 813 after passing through the sample, is sensed by the temperature sensor 831 and can be used either directly to modify the radiation detector output before it is passed to the signal processing circuitry or can be modified in the processing circuitry if a signal representative of the ambient temperature is passed to the signal processing circuitry.

0031Figure 3 shows an arrangement for modifying the output signal produced by the radiation detector in dependence on the ambient temperature in which the radiation detector is operating. The arrangement comprises a radiation detector 100 in the form of a photomultiplier tube whose output is connected to the input of an amplifier 101. The radiation detector 100 is equivalent to the radiation detector device 5 of Figure 1. The output of the amplifier 101 is connected to the one end of a network comprising a resistor 102 connected in parallel with the series arrangement of a thermistor 103 and a resistor 104. The thermistor 103 is an embodiment of the temperature sensor 6 of Figure 1. The other end of the network is connected to an inverting input of an amplifier 105 which has a feedback resistor 106 connected between its output and inverting input. A further resistor 107 is connected between a non-inverting input of the amplifier 105 and ground while the output of the amplifier 105 is connected to the output 108 of the arrangement.

0032In this example the photo multiplier tube is a multi alkali photocathode photomultiplier as sold by Hamamatsu TV Co. Limited. For these devices the temperature co-efficient of sensitivity over the wavelength range of 200nm to 800nm is almost constant at -0.4% per ʹC. The arrangement shown in Figure 3 includes an amplifying circuit having a temperature co-efficient of gain equal and opposite to that of the photomultiplier 100. Amplifier 101 forms a current-voltage preamplifier while the circuit to which its output is connected forms a modifying circuit which has a temperature co-efficient of gain of +0.4% per ʹC. In this example amplifier 105 is a high gain operational amplifier and the thermistor 103 is a disc type thermistor with a negative temperature co-efficient of resistance of -4.9% per ʹC and a resistance of 33K at 25ʹC. The values of resistors 102, 104, 106 and 107 were 7.8K Ohms, 18.7K Ohms, 6.8K Ohms and 3.3K Ohms, respectively. The characteristics of the arrangement would, of course, be modified to adapt to the properties of different radiation detectors.

0033Figure 4 shows in block schematic form the signal processing circuit 7 of the spectrophotometer shown in Figure 1 together with circuitry for controlling the operation of the instrument. The signal processing circuit has an input 200 to which the output 108 of the arrangement shown in Figure 3 may be connected, i.e. the input 200 receives the output signal from the radiation detection unit 4. This signal is converted into digital form by an analogue to digital converter (ADC) 201. The signal processing circuitry further comprises a microprocessor 202 and a memory 203 which are interconnected with the ADC 201 by means of a bus 204. Also connected to the bus 204 are a keyboard unit 205 by means of which an operator can enter appropriate operating instructions, for example wavelength range, scanning speed, number of samples, form of output, a display unit 206, for example chart recorder, printer, VDU, and a control interface unit 207 through which appropriate signals are passed from the microprocessor 202 for causing the monochromator 2 to scan across the selected wavelength range and from various elements within the instrument information to the microprocessor 202 to enable it to generate the appropriate control instructions.

0034Figure 5 is a flow diagram illustrating a method of analysis according to the invention.

0035Box 501 (ST) represents the start of the analysis i.e. the switching on and initial setting up of the instrument. The step represented by box 502 (CALIB) is a calibration scan which comprises inserting a sample cell containing a blank solution in the optical path, running a wavelength scan and storing the results. The step represented by box 503 (SAMP) is a sample scan comprising inserting a sample cell containing the sample dissolved in the blank solution in the optical path, running a wavelength scan, calculating the absorbance or transmittance of the sample from the results of the current scan and the calibration scan. The box 504 (DIS.RES) represents the presentation of the results of the analysis, for example the variation of absorbance or transmittance with wavelength, to an output device such as a chart recorder or a VDU. This may be done either after the completion of the sample scan or simultaneously with the sample scan as the values at each wavelength step become available. The box 505 (SP) marks the completion of the analysis.

0036As discussed earlier with respect to Figure 1 solid samples or liquid samples which do not contain dissolved samples may be analysed using the appropriate procedures for producing the calibration scan.

0037Figure 6 is a flow diagram illustrating the steps taken in one implementation of the calibration scan represented by box 502 of Figure 5. Box 600 (ST) represents the entry into this part of the method. Box 601 (SET W/L RNG) represents an input from the keyboard 205 which defines the upper and lower limits of the range of wavelengths over which the monochromator 2 is to be scanned. The microprocessor 202 takes this information and uses it to provide appropriate driving signals to the monochromator 2 through the control interface circuit 207 to drive the monochromator to the initial wavelength of the scan range, box 602 (S.IN W/L). The microprocessor then causes the output of the radiation detector to be digitised by the analogue to digital converter to produce a digital value representative of the magnitude of the radiation falling on the detector, box 603 (MEAS RAD). This value is then stored in the memory 203 at a location determined by the wavelength setting of the monochromator, box 604 (STR RAD). The microprocessor 202 then produces an instruction, which is fed via the control interface unit 207 to the monochromator, which causes the monochromator to step to an adjacent wavelength, box 605 (STP.W/L). Subsequently the output of the radiation detector is digitised by the analogue to digital converter, box 606 (MEAS RAD) and the digital value is stored in a location of memory 203 associated with the present wavelength to which the monochromator is set, box 607 (STR RAD). A decision is then taken, box 608 (FNL W/L?), as to whether the monochromator has stepped to the final wavelength, that is it has scanned across the complete wavelength range selected. If it has then the process ends, box 609 (SP), and the instrument is ready for a sample to be inserted into the radiation beam to allow its characteristics to be determined. If the monochromator has not completely scanned the selected wavelength range then path 610 is followed and the steps represented by boxes 605 to 608 are repeated until the wavelength scan has been completed.

0038Figure 7 is a flow diagram illustrating the steps taken in one implementation of the sample scan represented by box 503 of Figure 5. Box 701 (ST) represents the start of the procedure while box 702 (S.W/L RAN) represents the setting of the wavelength range over which the sample scan is to take place. This may take the form of information entered into the microprocessor 202 by an operator using the keyboard 205 and may include both the limits of the wavelength range of the scan and the increments by which the monochromator steps in completing the scan, i.e. the wavelength intervals at which measurements are taken. The microprocessor then generates an instruction to cause the monochromator to drive to its initial wavelength setting, box 703 (S.IN.W/L). This instruction is passed to a stepper motor in the monochromator which causes a diffraction grating to rotate by a given amount, the instruction being converted to an appropriate driving signal by the control interface circuit 207. The output of the detector is then digitised by the analogue to digital converter 201, this time with a sample inserted in the optical path i.e. the sample cell 3 contains a sample dissolved in the solvent used for the calibration scan, and the absorbance or transmittance of the sample is calculated by use of the stored digitised signal obtained in the calibration scan and the just obtained digitised signal, box 704 (MEAS ABS). The calculated value may then be fed to a display device or may be stored in memory for later display. The microprocessor then generates an instruction to cause the monochromator to step to the next wavelength, box 705 (STP W/L), the magnitude of the steps may be pre-programmed and constant or may be controllable by entry of information by the operator, for example using the keyboard. The signal produced by the radiation detector at the new wavelength is then digitised and used together with the stored calibration value to calculate the absorbance or transmittance, box 706 (MEAS ABS). Again the calculated value may then be fed to the display device or stored in memory for later display. The microprocessor then checks whether the monochromator has stepped to the final wavelength, box 707 (FNL W/L?), and if not then the path 708 is taken and the monochromator is stepped to the next wavelength. If the final wavelength has been reached the microprocessor may then check whether another sample is to be analysed, using an automatic sample changer, box 709 (FNLSAMP?), and if so then path 711 is followed. If no further sample is to be analysed then the process ends box 712 (SP). If the results are stored in memory then when the final wavelength of a scan is reached the resulting absorbance or transmittance scan may be fed to a display device. This may take place either at the end of each sample scan box 713 (O/PRES) or at the end of the analysis of a plurality of samples box 714 (O/PRES).

0039The procedure described hereinbefore with reference to Figures 5 to 7 is appropriate when the radiation detection unit 4 takes the form shown in Figure 3, that is the output of the radiation detection unit 4 is made substantially constant with ambient temperature variations by means of a compensation circuit including a temperature sensing element. It is alternatively possible to correct for the temperature co-efficient of sensitivity of the photomultiplier tube, or other detector, within the microcomputer circuit shown in Figure 4. In this case the output of the amplifier 101 (Figure 3) would be fed to the input 200 (Figure 4) while an output signal representative of the ambient temperature in the region of the radiation detector would be fed to an input 210. The input 200 is then fed to a first input of a multiplexer 211 while the input 210 is fed to a second input of the multiplexer 211. The output of the multiplexer 211 is fed to the analogue to digital converter 201 in place of the input 200.

0040The previously described flow diagram in Figure 6 is now modified in that the ambient temperature in the region of the radiation detector is measured and stored in the memory for subsequent use during the measurement scan. In one scheme illustrated in Figure 6 after the step represented by box 601 (SET WIL RNG) of setting the wavelength range over which the scan is to be performed the microprocessor 202 causes the control interface unit 207 to produce signals which cause the multiplexer 211 to select input 210 and apply the signal thereat to the analogue to digital converter 201. The ADC 201 digitises this signal which represents the ambient temperture in the region of the radiation detector, box 620 (MEAS TEMP). The microprocessor 202 then causes the digitised signal to be stored in memory 203, box 621 (STR TEMP). The digitised signal may be processed before being stored by the microprocessor if necessary. If desired the temperature may be continuously monitored during the calibration scan and a temperature reading stored at each wavelength setting during the scan. However, this is not normally necessary as the ambient temperature is unlikely to change significantly during a scan at any normally used scanning rate. Clearly the instant at which the temperature is sensed can be at any desired time provided that the temperature does not change significantly between the time that the temperature is sensed and the time at which the calibration scan is run. It is clearly advantageous to sense the temperature at some instant during the scan to give the greatest likelihood that the actual temperature for the whole scanning period is as sensed but provided that the ambient temperature is not changing quickly the temperature sensing could take place prior to or subsequent to the running of the calibration scan.

0041The previously described flow diagram shown in Figure 7 is modified in one scheme as follows. When the monochromator has been driven to the initial wavelength the microprocessor causes the multiplexer to select input 210 and fed a signal representative of the ambient temperature in the region of the radiation detector to the analogue to digital converter 201 which produces a digital representation on the bus 204, box 720 (MEAS TEMP). The microprocessor then causes this representation of the temperature to be stored in the memory 203, box 721 (STR TEMP), either in the form received from the output of the analogue to digital converter 201 or after further processing. This value is then used to apply a correction to the calibration scan, box 722 (UPD CALIB), to compensate for the temperature change between the time the calibration scan was made and the time of the current measurement scan. The original calibration scan and temperature may remain stored in the memory 203 or the calibration scan and temperature may be continuously updated as new temperature measurements are made.

0042Alternatively the calibration scan results may be normalised to a reference temperature and the normalised results stored. The measurement scan is then also normalised to the same reference temperature. This eliminates the need to store the temperature at which the calibration scan takes place since it is not necessary to know the difference in temperature between the times of the calibration and measurement scans as they are normalised to the same temperature.

0043The correction of the calibration scan may be derived from the characteristics specified in the data sheets for the particular radiation detector, may be derived from a temperature run using a typical detector or may be derived for each radiation detector when fitted into the instrument by first performing a temperature run and storing the results. The use of the microprocessor for performing the correction of the calibration scan allows a more accurate and flexible correction to be made since the microprocessor can take into account different values of temperature co-efficient at different wavelengths whereas the arrangement shown in Figure 3 can only compensate for a single value of temperature co-efficient and consequently can only give complete correction if the value of the temperature co-efficient is constant with varying wavelength radiation. This is approximately true for the multialkali photocathode photomultiplier tubes manufactured by Hamamatsu TV Co. Limited which over the wavelength range of 200 to 800nm hve a temperature co-efficient of sensitivity of -0.4% per `C.