MICROPROCESSOR CONTROLLED TITRATOR
THIS INVENTION relates to a method and apparatus for chemical analysis of samples which method and apparatus is suf ficiently simple and fast to be used in process control . It is particularly suitable for determining,
SO 4. or Ca
Whilst the method and apparatus may be used for a large number of chemical analyses of samples, the following description and specification by way of example only relates to SO
The known colori etric precipitation procedure (two indicators - to detect endpoint of precipitation = first small excess of titrant - have been used; they are either 'Thorin' or mainly now 'Dimethylsulfonazo III') has been used for pure or purified samples matrices. - Both indicators are subject to a host of interferences. These interferences by anions and cations must be prevented by suitable pretreatment. These interfering (foreign) ions may interfere with the mechanism of the precipitation reaction, and they may also consume reagent and/or react with the indicator itself. The separation (purification) steps required in order to prevent interference, usually require extensive manipulative steps (e.g. selective separation using ion exchangers) which are often lengthy and not necessarily quantitative.
Various circumventive titration (manipulation) steps have been tried in order to mainly prevent such interferences; they include back titration (after a given precipitation time) with complexing agents; back titration of deliberately added excess of precipitation (titrating) agent; and displacement or substitution titrations where, after a necessary reaction time, an excess titrant
(usuaXXy barium) can be determined by displacement of magnesium- from added Mg-EDTA complex by free barium in an alkaline medium. The magnesium released is then titrated.
All these circumventive steps and rather exotic titration methods produce time consuming factors and a rate of, manipulative juggling that the purpose of a fast, reliable, accurate and especially simple procedure which can be employed on almost any sample matrix (free from almost any interference) , is lost completely.
Even in relatively pure sample matrices both indicators are unsuitable for colour change observation with the naked eye, mainly because of the very faint colour change (dimethylsulfonazo III changes from a reddish blue to purple blue) . Often very slow returns of involved premature colour changes, plus slight inclusions of interfering radicals which play the most -^important part (in terms of system interference) can lead to erroneous results (even - and often - up to +50% of the actual contents) .
Optically observed precipitation titrations using colorimetric endpoint detection overcome obstacles in reference to indistinct colour changes but it must be remembered that they are still subject to the aforementioned interferences .
In one form the invention resides in a method of analysing a sample utilising a colorimetric endpoint resulting from a reaction between the sample and a titrant or from the presence of an indicator which comprises adjusting the pH of the sample as necessary, ions in the sample, titrating the sample with the titrant which .will produce a colour change to indicate the end form of a reaction between the titrant and the component of the sample being determined
using a constant titrant addition time manipulated by a computer/microprocessor, sensing the colour change with means for producing a stable e.m.f. acceptable to the computer/microprocessor, feeding the e.m.f. to the computer/microprocessor to enable the endpoint to be determined by the Keller-Richter equation or an equivalent equation.
In analysing some samples, it is necessary to add an organic solvents or mixture of solvent to minimise the effect of interference.
In another form the invention resides in an apparatus for determining a component of a sample comprising a computer/ microprocessor controlled titrator, a titration vessel fitted with a constant speed stirrer and colour sensing means capable of producing an electric signal to be fed to the computer/microprocessor in response to a colour change indicating the endpoint of a reaction involving a component of the sample and titrant.
The colour change sensing means may be a photo electric cell connected via an optical fibre lead and electrical lead to a spectrophotometer or a photo electric probe modified by connection to two accurate potentiometers allowing coarse and lockable fine control adjustments.
The invention will be better understood by reference to the following description in conjunction with the accompanying drawings wherein:-
Fig. 1 is an enlarged titration curve showing the end portion of the curve;
Fig. 2 is a block diagram illustrating the apparatus;
Fig. 3 is a circuit diagram of the modified photo electric probe; and
Fig. 4 shows the hook up details of modified probe.
Where possible commercially available instrumentation is used and consists of a computer/microprocessor controlled semiautomatic titrator (A) (Fig. 2) the titration arm (B) which holds the titration vessel and titrant delivery jet plus the stirrer (C) and the photocell (D) is connected via an optical fibre lead and electrical leads to the, spectrophotometer (E) ; plus a suitable (but dedicated) printer (F) .
In a preferred embodiment a commercially available modified photo electric probe replaces the rather expensive spectrophotometer (E) and the optical fibre lead plus electrical leads. This photo electric probe in its unmodified, form is meant to be used mainly for comparitive and not absolute measurements of light transmittance (transformed to a suitable e.m.f. acceptable to the electronics of the titrator (A) ) . It is also subject to vibrational interferences which are transmitted from the stirrer (see titration arm B in Fig. 2) from the continuously stirred solution to be titrated. Space limitations forced the manufacturer to miniaturise the required output potential adjustment circuit (nullpoint adjustment) and required potentiometers plus adjustment knob.
The required special application rendered this photo electric probe unsuitable in reference to stability of e.m.f. and in reference to its adjustment range. It was modified by placing the adjustment circuit outside and away from the photo electric probe and replacing required potentiometer with two more accurately working potentiometers allowing .coarse and lockable fine, control adjustments (see Figs. 3 and 4).
This modified, output-stabilised photo electric probe now also allows very exact measurements in light transmittance of any coloured solution within the visual range, which,
together with the accurate nullpoint adjustment facility, widens the application range of the instrumental set-up. The titrator together with photo electric probe and printer could, for instance, now be used as a spectrophotometer.
Extensive manipulation of available dedicated instrumental computer-memory facilities plus additions of 'inhibition' chemicals (crystal growth of precipitate formed is enhanced but also restricted at the same time) have made this semi-automated procedure possible.
The titration is based on a constant titrant addition (at constant stirring speed) colour indicated precipitation reaction, which is time-manipulated by computer control of various involved parameters. The colour change towards and at the endpoint of the titration proceeds from reddish blue to pusrple blue.
The enlarged, graphically displayed titration curve (end portion of curve only) shown in Fig. 1 should help to explain some of the criteria involved.
Each addition of titrant produces a large spike in e.m.f. (see Fig. 1) and a premature colour change. The computer has to wait until the e.m.f. is reduced to a lower but steady signal, provided the colour changes back to reddish blue or a mixture of colours which are predominant before the actual endpoint. Note that the actual height of the e.m.f. at this stage is greater than the e.m.f. at the endpoint. The premature endpoint is due to localised precipitation and gross-inclusions of interfering radicals (they are always present unless a chemically pure system is used) . The humps indicated by an x on one spike of the titration curve (Fig. 1) are due to these inclusions. The
selected pH plus titration medium (the titration -is performed in a selected ratio mixture of cooled ethanol and glycerol) ensures that these - otherwise interfering radicals - are dissolved from the freshly formed crystals. Only after waiting until the above has actually happened, gives the computer a signal to the titrator, so that aη. additional titrant addition is made. The lengths of these waiting periods vary not only from sample to sample and from system, but also within an actual titration (see section A to B of Fig . 1) . The selection of computer parameters cater for all these differences . They can be explained by the fact that involved intensity signals, stemming from light scattering effects, are greatest when the barium sulphate sol has a particle diameter of 0.2 microns. Since there is a constant shift in particle size distribution, any difference would be greater at the beginning of the titration where also the electrokinetic potential, hence inclusions of 'foreign bodies', is greatest.
As the titration proceeds, the differences between momentary premature and lower, but steady, signal e.m.f. 's become smaller and smaller until a steady and actually slightly falling {excess of titrant = small dilution) e.m.f. is observed. The computer shuts off the titrator at this point, calculates the result in %S0 by Keller- Richter equation, prints out the result and resets the titrator which is then ready for the next determination.
The above appears to be a complicated set-up. It is in fact very simple and could be competently learned by an Analytical Chemist within a 3-4 hour training period. All parameters are memory preset and one single method caters for analyses of SO in all phosphoric acids and slurries met in a commercial phosphoric acid manufacturing plant.
It should be noted that the above set-up is almost universally adoptable and can be used in almost any sample matrix. Only very slight computer parameter changes and only slight changes in sample dissolution modes will facilitate fast adaption for many analytical requirement (for the colorimetric determination of a large number of anions and cations .
The following table lists the ions which have been determined using this invention:-
An example of the determination for SO proceeds as follows:
A usually acidic sample aliquot of suitable size is pipetted into a titration vessel; some distilled water is added to make the total volume to 40ml approximately. Some {a few drops) 2:dinitrophenol pH indicator solution is added and the solution is then 'neutralised* by dropwise addition of sodium hydroxide solution until a greenish yellow colour persists. Diluted perchloric acid is added until the colour is discharged; 4 to 6 drops of diluted perchloric acid are then added in excess. The pH of the resulting solution is about 2.0 (1.8 to 2.3). Some dimethylsulphonazo III barium indicator (about 1ml of a 0.1% solution in water) and about 70ml of a pre-cooled ethanol/glycerol mixture are added. The mixture is shaken for a few moments and the titration vessel is clipped to the titrator.
The ethanol/glycerol mixture is added pre-cooled in order to overcome the heat of dissolution (heat is evolved when ethanol is mixed into water) and thereby producing a sample matrix mixture which is roughly at normal room temperature and is instantly ready for immediate titration as far as temperature is concerned. The solution cannot be titrated at temperatures higher than normal room temperature because air bubbles containing ethanol fumes would otherwise continuously be evolved. They would interfere with the transmittance measurements and as such rendering the procedure of determination impossible to follow.
The ethanol content of the mixture to be titrated is deliberately kept at about the 60% level in order that the solubility product of the resulting barium sulphate is
kept at the absolute minimum. The water content of the mixture is kept at its possible maximum in order to allow the titration to be followed in sample matrices which contain relatively large amounts of solvent insoluble matter.
The ethanol/glycerol mixture is pre-cooled in a small refrigerator to about 60°C; it is pumped from there (as needed) via a small bore stainless steel tube, ending in a suitable delivery valve and outlet, using low pressure compressed air.
The ethanol/glycerol ratio of 1:0.1 and the amount used per titration (about 70 mis) ensures that gross-inclusions of interfering radicals which precipitate momentarily are kept away from the growing barium sulphate crystal. The action of the glycerol is not fully explainable - but it acts similar to a wetting agent - where the momentarily co-precipitating interfering radicals are kept ('washed') away from the surfaces of the growing barium sulphate crystal. They then dissolve instantly because the pH level is deliberately kept low.
Each titration begins with an immediate pre-dispensing of titrant and after a waiting time (stirring time) of 240 seconds the titration proceeds and finishes automatically. This titrant pre-dispensing and waiting time (stirring time) ensures that some of the sulphate present is precipitated and that during the involved waiting time the main bulk of the barium sulphate crystal grows to a size of about 0.2 microns. A special feature of the water/ ethanol/glycerol titration medium mixture is that at its operating pH and temperature it also restricts the barium sulphate crystal growth to about 0.2 microns which is
ideal for this type of spectrophotometric precipitation/ titration procedure of barium sulphate.
The already pre-precipitated barium-sulphate crystals act also as 'seeding agents' for the .precipitate which is forming after each titrant addition. It ensures immediate and almost quantitative (stochiometric) precipitation before the endpoint and, apart from a minimal but constant excess of titrant at the endpoint, a completely quantitative precipitation at the automatic finish of the titration.
After the solution to be titrated is clipped to the titrator the corresponding method number is pressed and the titration proceeds automatically to the end. The desired, result unit is printed out on the printer, the titrator resets itself and is then ready for the next titration.
The reaction kinetics involved are dependent on many factors (as discussed earlier) . This dependence is further amplified when the ratios of normally present interfering ions are shifting from sample to sample (this is almost always the case) . This situation is getting much worse when it is asked (as has been achieved by the above extensive manipulations of all involved integers) that one only pre-set method caters for a large number of differing sample matrices with largely differing contents of interferences. Only slight modifications of sample dissolution modes and/or computer memory settings will make this procedure almost universally adoptable.
The continuous changes in e.m.f. (during the actual titration) produce signals which are often much greater when the titration proceeds than the signal available at
the endpoint; they require extensive computer manipulations which are set into the memory facilities of the titrator.
Time based shifts and differing widths of e.m.f. 's produced during the actual titration (see Fig. 1) would normally render this type of procedure of determination unworkable. These difficulties have been overcome by:
(a) inclusion of glycerol in the titration matrix.
(b) adding propan-2-ol to the barium perchlorate titrant,
and (c) computer manipulations - where:
a - the amount of glycerol added to the solution to be titrated has the maximum effect of producing a barrier between the growing barium sulphate crystal and the momentarily co-precipitating interfering radicals (as discussed above) .
b - the titrant consumption automatically increases the propan-2-ol concentration level as the titration proceeds. The concentration level of propan-2-ol is about at its optimum just before and shortly after the endpoint is reached (the total titrant consumption is kept to between 15 and 20 mis) .
The resulting quarternary mixture (water-ethanol-glyσerol- propan-2-ol) ensures that:
I. The barium sulphate precipitate formed is crystalline but the crystal size is becoming smaller and smaller
towards the endpoint and as such producing a relatively large jump in e.m.f. towards and at the end of the titration.
II. The main bulk of barium sulphate precipitate (at the beginning of the titration) has a crystal size of about 0.2 microns (ideal in reference to minimise inclusions of foreign bodies in the crystals formed) .
The combined effects of a and b reduce time based shifts and reduce the differences in widths of e.m.f.'s - especially at the end of the titration.
c — These reductions are necessary because of computer limitations; they enable computer memory settings applicable to wide ranges of sample matrices containing wide ranges of
. 'interfering' anions and cations.
The titrant is 0.005 Molar barium perchlorate solution containing water plus propan-2-ol in a ratio of 1:4.