GB1582228A - Colorimetric analysis - Google Patents

Description

(54) IMPROVEMENTS IN OR RELATING TO COLORIMETRIC ANALYSIS (71) We, BRITISH GAS CORPORATION, of 59 Bryanston Street, London, W1A 2AZ, a British Body Corporate, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed to be particularly described in and by the following statement: This invention relates to colorimentric analysis, and more particularly to methods and apparatus for carrying out automatic and continuous colorimentric analysis of materials such as boiler feed waters.

The change from manufactured gas to natural gas has reduced the numbers of trained chemists within the Industry. In the interim period between complete conversion and adequate reserves of natural gas becoming available from offshore wells, supplementary supplies may be required.

This contingency necessitates the possibility of automating some of the routine chemical analyses needed for control on a reforming plant.

Many of the tests done on plants concern the quality of feed and boiler water and have colorimetric end points. Plant effluent is another aspect of gas manufacture that has to be considered, as discharge to rivers and public drainage authorities is now strictly controlled. Effluent analysis is particularly important where coal is used as a feedstock, as quantities of phenol, cyanide, ammonia and sulphides are produced as major contaminants. These can also be determined colorimetrically.

The present invention proposes a method and apparatus for continuous on-line colorimetric analysis which is different from the normal concept. The usual way in which an Auto Analyser is used is to analyse a fixed number of samples per hour. Used this way it is necessary to take samples at fixed intervals and transfer them to an automatic sample handling device. This method is ideal for many discrete samples, such as clinical specimens, but it defeats our intention of continuity of measurement.

In accordance with the present invention there is provided apparatus for the continuous analysis of components in a liquid comprising: 1. a photometric absorptiometer comprising a plurality of modules, each module com prising a flow cell, means for supplying an influent stream of a coloured complex of said liquid, means for illuminating the flow cell, photometric means for detecting light passing through said flow cell and a differential amplifier having a pair of input terminals coupled to said photometric means in addition to a zero-potential reference terminal, said supply means including 60 means for introducing air into an influent stream of said liquid thereby, in use, to portion said influent stream and form a segmented influent stream composed of discrete samples of liquid of predetermined volume, interposed by bubbles of air, means for introducing complex reagent into the liquid portions of said segmented stream and means for removing the air from the influent stream after formation of the coloured complex of said liquid, 2. means for sequentially selecting output signals from each amplifier and, 3. means for displaying and recording the selected output signals.

The liquid to be analysed is conveniently supplied to the cell modules by means of a proportionating pump. Preferably, the proportionating pump is constructed to provide mixing of the non-complex with a reagent which upon mixture with the liquid forms a coloured complex.

Prior to passage in the cell modules, the temperature of the coloured complex is preferably adjusted and controlled at a predetermined level. This may be accomplished by passing the liquid through coils immersed in a thermostatically controlled bath.

The apparatus of the invention may be used to analyse any liquids, preferably aqueous liquids which will form a coloured complex.

The colorimetric analysis of liquids is well knawn and suitable methods are described for example in British Standards 2690, Parts 2, 3 and 6. Although the apparatus of the invention may be employed for the analysis of any liquid, for ease of understanding the invention will be described with reference to tests for the analysis of silica, chloride, phosphate and hydrazine.

The invention will be described in greater detail with reference to the accompanying drawings, in which: Figure 1 is a block schematic diagram of apparatus in accordance with the invention, Figures 2 to 5 are flow diagrams for the manifold operating conditions associated with the measurement of silica, phosphate, chloride and hydrazine, Figure 6 illustrates the provision of the wash cycle, Figure 7 illustrates apparatus for the prevention of pulsing of the fluid through the cell module, Figure 8 is a circuit diagram of a differential input amplifier suited for use in the present invention, Figure 9 is a circuit diagram showing the signal switching circuit for selecting amplifier output signals, Figure 10 is a circuit diagram for manual operation of the apparatus, Figure 11 is a circuit diagram of the print and space command circuit, and Figures 12 and 13 are alternative circuit diagrams for a digital volt meter hold circuit.

All chemicals and samples are delivered to manifolds by a pump, e.g. a peristaltic pump.

A suitable pump is a 30 channel one manufactured by the Newton Instrument Company, which is of the "contrarotating" roller type. This is a continuous, clockwise chain-driven bank of rollers, each of which rotates counter-clockwise as it contacts the tubes. All tubes are stretched uniformly on a removable platen.

This type of contact causes the fluid to be drawn and then pushed through the tube.

A problem that causes difficulty in maintaining a steady flow through the manifold, and ultimately the optical cell, is that of pulsing.

To avoid this problem it is essential to introduce like quantities of fluid into the manifold through similar sized orifices. This eliminates most of the pressurising/depressu- rising effect (pulsing) and allows a smooth passage through the system.

The first step is to partition the flow of sample as it enters the manifold (see Figures 2 to 5) by injecting air at regular intervals.

This injection should be done in a horizontal or downward direction to ensure a regular bubble pattern. Segmentation with air ensures that the colour development takes place in a small discrete quantity of sample, typically 0.01 cc. The small quantities ensure complete dolour development in the shortest time and also a constant sample/reagent ratio.

Reagents are added separately to the sample to avoid any premature interaction.

After each addition the sample passes through a short (eg 14 turn) delay mixing coil. A longer delay is needed after the final reagent has been added to allow the coloured complex to develop. The delay time can be varied by changing the length(s) of coil(s) prior to entry to the colorimeter. The time required for the complex to develop is governed by the chemistry involved and is different for each analysis.

To assist colour development and maintain the reactions at a constant temperature the manifolds are sited in a water bath thermostatically controlled, eg. in the range 35 - 40"C.

This control is important when operating on a 24 hour basis as day/night temperature changes will adversely affect complex development times.

Before optical density measurements are made it is necessary to remove all bubbles from the flowing liquid. This may be done by passing the influent (defined as "Total Flow" in Figures 2 - 5) through a debubbler sited on top of the optical cell (see Figure 7). The bubbles pass upward and out to drain together with 60% of the influent. The remaining 40% (bubble free and defined as the "F/Cell Pull Through" in Figures 2 - 5) is drawn through the optical cell by the proportioning pump, measurement made, then discharged to drain.

The 60/40 split is an ideal figure but in practise it may be necessary to alter it. Alteration becomes necessary when long lengths (several feet) of transmission tubing are used, causing a high differential pressure across the system.

The higher the A P across the system, the slower the flow rate. Consequently, a "pull through" of less than 40% must be applied to avoid drawing bubbles into the optical cell.

To minimise the A P it is advisable to arrange the items of equipment with short lengths of connecting tubing between them. In this way it is possible to approach an ideal condition.

Low flow rates through the optical cell will not impair stability or accuracy of measurement, though the occasional air bubble in the cell will take longer to clear.

To assist the flow of reagents through the manifolds it is desirable to lower the surface tension of the fluids on the glass surfaces.

Suitable surfactants are commercially available, but ionic ones, such as Decon 90, "Teepol", "Byprox", [The words "Teepol" and "Byprox" are Registered Trade Marks] must be avoided at all costs. The non-ionic surfactants used in the prototype module are 'Nonidet' [Registered Trade Mark] P40 for the Silica, Chloride and Hydrazine assays and 'Lever' [Registered Trade Mark] IV for the Phosphate assay.

Over a period, say 1/2 days, the results from each assay will tend to drift upward and give a false indication of the true value. The problem is caused by the excess reagents, particularly the more coloured ones, "plating out" on the inner surfaces of the optical cell. This causes an apparent increase in the optical density, ie. less light falls on the photo-resistor detector, resulting in a higher output voltage, hence a higher (erroneous) answer will be displayed and recorded. To avoid this problem it is desirable to incorporate a wash cycle to clean the optical cell. The washing solution is different for each cell: Silica 0.2N Sodium Hydroxide Chloride 1.ON Nitric Acid Phosphate 10% Ammonium Hydroxide Hydrazine 10% Hydrochloric Acid The wash cycle is done automatically at pre-set times.For example the cycle can be initiated twice per day (12 hour interval) by a cam-timer which is programmed to trigger the operation of a solenoid valve (see Figure 6).

In the "open" position the valve has the normal flow through it to the colorimeter and the wash solvent continuously circulating to a reservoir. At the pre-set time the valve "closes".

Normal flow is diverted to drain and wash solvent diverted into the flow line to the cell.

The wash period is for 30 minutes after which time the valve returns to its "open" position and resumes formal flow.

It is advisable to wash the manifold with the appropriate solvent on a weekly basis.

This is most easily and conveniently done at the time of the routine tube change. The instrument will require recalibration once the normal flows have been restored. Full details of calibration are4supplied with each instrument.

The signal handling system can be considered in two main parts: I Singal processing II Presentation of results Signal Processing Each channel is first amplified, eg. by a factor of five by independent, differential input amplifiers (see Figure 8). These bring the signal levels up to the order of 1 volt, and mean that the gain settings on the colorimeter can be kept low, a condition which we find will extend the linear range of any particular channel.

The output from the selected channel is then passed to a digital voltmeter (D V M) and subsequently to the digital printer. The gains of the various channels, and their associated amplifiers, are set up on standards so that the print-out will be a direct digital representation of the concentration in, say ppm. Channel selection may be made in three ways:1. Manual Here, an individual channel is selected by a front panel push-button control. The main timer must be switched off when monitoring a manually selected channel.

2. Auto In this mode, every thirty minutes the main timer will start a scan cycle which selects each channel in sequence. It also gives the printer a signal to print out each individual result. Each channel is selected for approximately 1 minute, and the whole cycle for five channels takes 6 minutes.

3. One shot (See Figure 10) Pressing the 'one shot' button will start the scan cycle, described above under auto control. At the end of the sequence the scanning mechanism stops automatically. The main timer must be turned off when making one shot scans.

The scanning mechanism (programmer) is basically a cam-timer. It incorporates a relay circuit which ensures that each.time a signal is received from either the main timer or the one shot button, the timer will complete one full revolution and then stop until another signal is received.

The signal from a particular colorimeter channel is switched to the D V M by reed relays. These are energised either by the programmer or by one of the channel selector push buttons. The energising current to the reed relay also lights a miniature filament lamp to give identification of the selected channel. When the programmer is running, a spare cam is used to give the printer a command to print during each channel selection period.

If a channel is being monitored manually, the reading can be printed by a separate button on the printer itself.

The signal processing equipment also includes the timer for the wash procedure as hereinbefore described. The timer is a separate, 1 revolution per day, cam timer. The drum is divided into 60 segments, allowing great flexibility in the duration and time of start of the wash period. The timer can be switched off if not required, and a manual wash switch is also provided for occasional extended wash periods.

Both the wash and main timers must be switched off during extended manual washes.

Although the wash timer is separate, it is interlocked with the main timer, so that a scan cycle cannot be initiated during a wash period.

Presentation of Results Two systems are described:- A multi-print recorder, and a D V M linked to a digital printer.

The recorder has its own mechanism, and so does not need the complete signal handling equipment. A separate wash timer is necessary.

The recorder may require buffer amplifiers due to a mismatch of impedances from the colorimeter to the recorder. A recorder which has a high input impedance (ie. greater than 1 M.Ohm) should not need these.

The operation of the D V M/Printer is best considered in steps:1. D V M receives an analogue signal, digitises it, and displays it. This cycle occurs at a rate of twice per second (sample rate).

2. The digital information is passed to the printer at the sampling rate of the D V M.

3. The printer is given a "Print" command. The printer employs standard logic mode. The 5V pulse required in this mode is obtained from a SN74121 integrated circuit Schmidt trigger, operated from the microswitch on the programmer (see Figure 11).

4. Printer instructs the D V M to hold the displayed reading and digital information for the duration of the print cycle. The output from the printer is normally high (+5V) and goes low for the duration of the print cycle. Similarly, the control input to the D V M is normally high and has to be made low to hold the display. The high level of this D V M is, however, at 9.1 V instead of the normal +5V. When connected to the printer, the +9.1V on the D V M will fall to the +5V high level of the printer, and this fall is enough to hold the display. An inter face between the voltage levels of the D V M and printer is necessary. Our solution was to use a 'Dual in-line' reed relay supplied from a buffer stage (see Figure 12). The buffer stage is simply a transistor, with a separate +5V power supply.This buffer stage is necessary to prevent excessive load ing of the printer circuits. The reed relay is normally open circuit, and grounds the hold input to the D V M when the printer output goes low. An alternative circuit is to use an Opto-Isolator instead of a reed relay (see Figure 13).

5. Printer prints the displayed information.

6. Printer allows the D V M to continue its regular sampling.

The use of an 'Anadex' printer has the advantage of user options for the control inputs.

Briefly, this means that a control input or output can be either standard logic or complementary logic.

In the case of the print command, for instance, there are two ways of operating it: In complementary logic form, the input is normally high and must be grounded - by a switch closure for instance, to indicate a print command. In standard logic form, the input is normally low, and needs a +5V pulse to initiate a print command.

A 'Gralex' D V M is preferred for its very legible 'Sperry' [Registered Trade Mark] display and the availability of binary coded decimal (B C D) output for driving the printer.

Power supplied for the various sections should be kept separate to minimise interference and noise spikes.

The results may be presented in two ways:a. A multi-point recorder to give a visual indi cation of the trend of the Analysis.

b. A numerical display of concentration on a digital voltmeter which subsequently drives a digital printer.

It may be preferred to run these systems in parallel for two reasons; a. The trend recorder to give the Plant Engineer a visual warning of changing states, and b. the D V M/Printer system to present not only a numerical and written record but ultimately to provide a compatible signal to enable the automation loop to be closed.

This would mean that repeated "off-spec" results would trigger the remedy, such as increase/decrease of the dosing pump stroke, increase/decrease boiler blow-down, whichever is necessary. In practice it may be decided that only one of the display systems is required. Option (a) is a cheap form of display but does not have the ability to deliver a compatible signal for full automation.

In the analysis of liquids it is necessary to obtain a linear response with respect to concentration. Thus, it is essential to work in the region where the solutions obey Beer's Law.

The assays for silica and hydrazine are linear up to the maximum concentrations quoted and will present no problems. The chloride and phosphate assays are not, however, linear over teh required range. To overcome this problem it is necessary to determine the concentration at which Beer's Law ceases to be obeyed. Some commercially available colorimeters will indicate a deviation at approximately 2.5 ppm for chloride and 25 ppm for phosphate. Thus, to enable chloride and phosphate to exhibit a linear response up to the concentrations required, the sample must be diluted with water. For the chloride assay a 9:1 dilution will increase the effective linear range up to 25 ppm and a 4:1 dilution will increase the effective phosphate linear range to 125 ppm.

The analytical methods used are essentially those documented in BS 2690 "Methods of Testing Water used in Industry", Parts 2, 3 and 6. A brief description of the chemistry involved is:1. Silica: The method determines reactive silica (monomeric and dimeric silicic acid) and is based on the blue colour (molyb denum blue) of reduced silico-molybdic acid. The reducing agent is 1-amino-2 naphthol 4-sulphonic acid.

2. Chloride: Chloride ions react with mercuric thiocyanates to release thiocyanate ions.

These ions react with ferric perchlorate to produce the coloured complex (Fe(SCN)2+.

3. Phosphate: The method is based on the formation of the yellow vanadamolybdo phosphoric acid.

4. Hydrazine: Hydrazine reacts with 4-dimethylamino benzaldehyde to give acid salts of the corresponding azine. These salts are intensely red in strong solution and yellow in dilute solution.

WHAT WE CLAIM IS: 1. Apparatus for the continuous analysis of components in a liquid comprising: a. a photometric absorptiometer comprising a plurality of modules, each module com prising a flow cell, means for supplying an influent stream of a coloured complex of said liquid, means for illuminating the flow cell, photometric means for detecting light passing said flow cell and a differential amplifier having a pair of input terminals coupled to said photometric means in addi tion to a zero-potential reference terminal, said supply means including means for introducing air into an influent stream of said liquid thereby, in use, to portion said influent stream and form a segmented in fluent stream composed of discrete samples of liquid of predetermined volume, inter posed by bubbles of air, means for intro ducing complex reagent into the liquid portions of said segmented stream and means for removing the air from the influent stream after formation of the coloured complex of said liquid, b. means for sequentially selecting output signals from each amplifier and, c. means for displaying and recording the selected output signals.

2. Apparatus as claimed in Claim 1, wherein a proportionating pump is provided to supply influent stream and complexing reagent to each flow cell.

3. Apparatus as claimed in Claim 2, wherein said proportionating pump includes provision for mixing liquid with a coloured complex forming reagent.

4. Apparatus as claimed in any one of the preceding Claims, including means for adjusting and controlling the temperature of the liquid prior to entry into the flow cell.

5. Apparatus as claimed in any one of the preceding Claims, wherein said selecting means is operated by a timer.

6. Apparatus as claimed in any one of Claims 1 to 5, wherein said selection means is a manually operated means.

7. Apparatus as claimed in any one of the preceding Claims including provision for periodically passing only water through the flow cell to wash the interior of the cell, thereby, in use, to prevent plating out of the complex on the surfaces of the cell.

8. Apparatus as claimed in any one of the preceding Claims and substantially as hereinbefore described with reference to the accompanying drawings.

9. A method for the continuous analysis of liquids, employing apparatus as claimed in any one of the preceding Claims and substantially as hereinbefore described.

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (9)

**WARNING** start of CLMS field may overlap end of DESC **. intensely red in strong solution and yellow in dilute solution. WHAT WE CLAIM IS:

1. Apparatus for the continuous analysis of components in a liquid comprising: a. a photometric absorptiometer comprising a plurality of modules, each module com prising a flow cell, means for supplying an influent stream of a coloured complex of said liquid, means for illuminating the flow cell, photometric means for detecting light passing said flow cell and a differential amplifier having a pair of input terminals coupled to said photometric means in addi tion to a zero-potential reference terminal, said supply means including means for introducing air into an influent stream of said liquid thereby, in use, to portion said influent stream and form a segmented in fluent stream composed of discrete samples of liquid of predetermined volume, inter posed by bubbles of air, means for intro ducing complex reagent into the liquid portions of said segmented stream and means for removing the air from the influent stream after formation of the coloured complex of said liquid, b. means for sequentially selecting output signals from each amplifier and, c. means for displaying and recording the selected output signals.

2. Apparatus as claimed in Claim 1, wherein a proportionating pump is provided to supply influent stream and complexing reagent to each flow cell.

3. Apparatus as claimed in Claim 2, wherein said proportionating pump includes provision for mixing liquid with a coloured complex forming reagent.

4. Apparatus as claimed in any one of the preceding Claims, including means for adjusting and controlling the temperature of the liquid prior to entry into the flow cell.

5. Apparatus as claimed in any one of the preceding Claims, wherein said selecting means is operated by a timer.

6. Apparatus as claimed in any one of Claims 1 to 5, wherein said selection means is a manually operated means.

7. Apparatus as claimed in any one of the preceding Claims including provision for periodically passing only water through the flow cell to wash the interior of the cell, thereby, in use, to prevent plating out of the complex on the surfaces of the cell.

8. Apparatus as claimed in any one of the preceding Claims and substantially as hereinbefore described with reference to the accompanying drawings.

9. A method for the continuous analysis of liquids, employing apparatus as claimed in any one of the preceding Claims and substantially as hereinbefore described.