224 DENTON AND WHITEHEAD : AN AUTOMATIC APPARATUS [AndySt, VOl. 91 An Automatic Apparatus for the Determination of Titanium BY C. L. DENTON AND J. WHITEHEAD (British Titan Products Company Ltd., Billingham, Co. Durham) A method is described for the determination of titanium in solution. The titanium is reduced with cadmium in a column-type reductor and titrated with ferric alum solution to a potentiometric end-point. The determination is carried out automatically once the sample solution has been placed in the instrument, and the result is obtained in approximately 7 minutes. Results are given demonstrating the excellent precision obtainable with the instrument. IN the titanium pigment industry it is necessary to analyse a large number of samples, mostly liquid, for titanium content.The method most frequently used for this determination is to reduce the titanium to the tervalent state and oxidise with a standard oxidising agent.ls2 -4gents which have been used for reducing titanium include the following : aluminium, fusible alloys, cadmium, iron, zinc, bismuth, lead, slightly amalgamated zinc and liquid amalgams of zinc, bismuth, lead, tin or cadmium. Solid reducing agents are generally used in a column or Jones-type reductor, and liquid reducing agents in the separating-funnel type of reductor of which the Nakazono is a typical developed form? Solid reducing agents may also be added directly to the solution, a typical example of this being aluminium. After reduction is complete, excess aluminium dissolves in the solution and does not interfere in the subsequent t i t r a t i ~ n .~ Cathodic reduction has also been used, particularly in the coulometric generation of tervalent titanium.5 Reduced titanium solutions can be titrated with a variety of standard titrimetric reagents. Among the commonest of these are solutions of potassium dichromate, perman- ganate or bromate, ceric sulphate and ferric solutions. The end-point may be determined using indicators such as methylene blue, diphenylamine, indigo carmine or sodium or potassium thiocyanate. Solutions containing titanous salts are unstable and easily oxidised by the atmosphere, but this may be avoided by the use of an inert atmosphere or by adding the titanous solution to an excess of oxidising agent. Alternatively, the solution may be run into an excess of ferric sulphate solution and the ferrous equivalent titrated with standard oxidising agent.Even in the most rapid of the above methods, the manipulative time is still considerable, and it is evident that an automatic instrument would achieve a large saving in man-hours. It was envisaged that such an instrument would carry out automatically a determination of titanium in a solution which had been manually inserted into the instrument. At the end of the determination the result would be displayed in digital form and the instrument would be ready for another determination. The above methods were critically examined in order to select the most suitable for further consideration. Initially, attempts were made to use liquid amalgams since they are rapid and efficient in action.The titanium solution was agitated with the amalgam in a glass cylinder, by means of a mechanically operated perforated glass piston, until all the titanium was reduced to the tervalent form. The reduced solution was then run off into a titration vessel via a side arm, situated at the amalgam level. The reduction appeared to be complete, but slight oxidation took place during transference to the titration vessel despite the presence of an inert atmosphere of carbon dioxide and the presence of potassium thio- cyanate which stabilises tervalent titanium. In view of this and mechanical difficulties, no further work was done on this method. Electrolytic reduction is potentially very attractive, but it was shown that the current densities required to reduce the quantity of titanium in the average sample in a reasonable time were too high. The titration may also be completed potentiometrically.April, 19661 FOR THE DETERMINATION OF TITANIUM 225 The column-type of reductor was next investigated. Experience has shown that cadmium metal granules (1 to 2 mm) are superior to amalgamated zinc as reducing agent in a column- type reductor.A column-type reductor, packed with cadmium, was therefore fitted with a solenoid-operated outlet valve and platinum level probes in the top of the column. The cadmium was supported on a sintered-glass disc protected by a plug of glass-wool. With this system, the complete reduction and washing of a sample solution took about 15 to 20 minutes, but this time progressively increased as the granules decreased in size and packed into the base of the column.This was clearly unsuitable for an automatic method and consequently the upward flow of liquid was investigated. A simple apparatus was assembled to investigate this technique. The sample was pumped from a beaker into the bottom of the column using a mechanically operated nylon syringe and two glass non-return valves. Level probes were used to control the reduction and washing processes, and the solution was finally titrated manually. Consistent reduction efficiencies and constant reduction times were obtained over many determinations, indicating that the upward flow was fluidising the column and preventing it from packing. This reduc- tion procedure was accordingly adopted and incorporated into the final apparatus. Ferric alum solution was chosen as titrant in view of the fact that many of the sample solutions contained ferrous iron which would react with such oxidising agents as potassium dichromate and ceric sulphate.The nature of the samples, which varied appreciably in colour, precluded the use of photpmetric methods for end-point detection, and a potentiometric method was therefore selected for use. The change in potential at the end-point was enhanced by the presence of potassium thiocyanate, and this was added in all titrations. It has the further advantage of stabilising the tervalent titanium to atmospheric oxidation and the colour change at the end-point provides a visual check on the accuracy of the instrument. Several methods of adding titrant were examined. The standard burette with location of meniscus by light source and photoelectric cell was rejected because it is too complex.Preliminary work with an early form of peristaltic pump indicated that it was not sufficiently accurate and it would be difficult to incorporate in an automatic system. Frequent calibration would also be necessary. A glass syringe pipette, operated by a precision calibrated leadscrew driven by a syn- chronous motor, was next investigated. A cam on the top of the leadscrew operates a mechanical counter, which is adjusted so that the full traverse of the syringe, i.e., 50m1, is equivalent to 1000 revolutions. This method, which was proved to be completely satisfactory for dispensing accurate volumes of liquid, was incorporated into the automatic titrator. Very careful lining up of the syringe and leadscrew was found to be necessary to prevent breakage of the syringe.This was avoided by using a shorter piston made from Teflon with an O-ring seal fitted to the end. These have proved to be extremely satisfactory in use, and breakages have been almost eliminated. GENERAL DESCRIPTION OF THE JNSTRC‘MENT END-POIKT DETECTOR- The end-point detector is essentially the same as that described by Brown and Weir,6 which is incorporated in the Model 34 Titromatic Analyser made by Electronic Instruments Ltd. It has, however, been modified to give a direct reading in pH units. The modified circuit is given in Fig. 1, and a list of components can be found in Appendix I.The modifications are as follows: A 100-0-100 microammeter with series resistor R,, (5600 ohms) is connected across the cathodes of V4. R, is omitted, and the value of R,, is changed to 22,000 ohms. R4oA and RdOB form part of a twin-ganged potentiometer, connected so that as one increases the other decreases. R41 is a potentiometer to give the required pH scale length, and R,, is a 25,000-ohm helical potentiometer that gives a direct indication of the pH of the solution. The manual-check auto switch is converted to a two-position rotary switch giving manual and auto positions only. The neon lamps, V1, and V,,,!normally in the Brown - Weir unit) that are used to indicate “Titrate on” and “Titrate off, are omitted, and have been replaced by the meter connected across the cathodes of V,.This is the set buffer control.226 DENTON AND WHITEHEAD 1 AN AUTOMATIC APPARATUS [AutalySt, VOl. 91 2 0 7 - > m i - - A d - N d + d r 0 Z M M al r L I 0-April, 19661 FOR THE UETERMINATION OF TITANIUM 227 The value of R,,, the anticipation control, is change.d to 25,000 ohm, with R45, 180,000 ohm, connected from the top end of R,, to R21. The E l contact is now made normally closed to operate the sequence unit, The end-point detector is capable of being used for pH or potentiometric titrations. C I I_ O F I .6 cm - A = 5-cm diameter ring with several holes to run wash liquor into beaker B = Platinum probe fused through glass C = 610 socket D = PVC tubing E = Outlet to glass filter F = Air hole G = QVF 614 socket H = PTFE cone to fit 614 socket J = I x 25-s.w.~.platinum wire fused through end of glass tubing and soldered to tinned-copper wire Fig. 2. Sample beaker and probes SAMPLE BEAKER- This is a modified 250-ml squat-type beaker, and a diagram of it is shown in Fig. 2. The top of the beaker is widened to take the wash liquor manifold. This consists of a 7 to 8-mm diameter tube, with holes at 1-cm intervals through which wash liquor is passed to rinse the sides of the beaker. The tube is recessed to prevent solution spraying straight into the beaker and not washing the sides. The beaker has a sloping base so that a minimum volume of solution is retained in the beaker when a change is made to the next sequence in the operations. The volume o f wash solution is determined by three probes which are placed in a side arm to avoid accidental wetting by the wash solution. As a further precaution against spurious contact, the connecting wires are encased for most of their length in glass- tubing.SAMPLE TRANSFER PUMP- The syringe is driven by a 250-volt, a.c,, 50 c/s geared induction motor, running at 4-75 r.p.m., connected to the syringe by means of a crankshaft. The arrangement can be seen in Fig. 3. REDUCTOR- A 20-ml nylon syringe is used for this purpose. A diagram of the apparatus is shown in Fig. 4. The solution is pumped from the sample beaker to the reductor through a non-return valve system, consisting of hollow, ground-glass valves weighted with mercury, fitted to the base of the reductor column so that the solution can only pass from the beaker to the column.Since small particles of solid material interfere with the operation of the glass valves, all228 DENTON AND WHITEHEAD: AN AUTOMATIC APPARATUS [Ana&d, VOl. 91 solutions used must be free from suspended matter, and a KO. 1 porosity sintered-glass filter is fitted in the sample line as an additional precaution. The column has a series of indentations to improve distribution, and is fitted with a 75-mm No. 0 porosity sintered-glass filter a t the base to support the cadmium reducing agent. A similar filter is fitted at the upper end to prevent cadmium passing into the titration beaker. Two clip-on joints are fitted to the column to facilitate re-packing. The space between the top joint and the sintered-glass filter is also packed with glass-wool as an additional precaution.To the upper end is attached an outlet tube leading into the titration beaker. The glass reductor column is 20 mm in diameter and 35 cm long. A To motorised syringe To titration beaker C E D A = QVF ball joint, m.s. 12/2 BS B = 75 x 0 sintered disc C = Glass valves weighted with mercury D = BIO cone and socket E = BIO cone F = Constrictions ground t o form valve seat G = 7 x I sintered disc H = I-inch clip-on joint J = 20-mm bore with indentations K = I-inch clip-on joint Fig. 4. Keductor column and valvc assembly From reductor To beaker drain valve A = Calomel electrode B = Platinum earth electrode C = Titrant delivery D = Carbon dioxide inlet E = Carbon dioxide ring F = Stirrer G = Thiocyanate delivery H = Platinum indicating electrode K = Perspex beaker top Fig.5. Titration beaker and electrodes TITRATION BEAKER AND ELECTRODE SYSTEM- I t is made from a 1-litre beaker, the top of which is cut off to the dimensions shown, and a drain outlet fitted. The lid of the beaker is machined from solid Perspex, and is drilled to take the calomel reference and two platinum electrodes, which are held in position by clips attached to the centre bridge-piece. The lid also contains holes through which pass the stirrer, the sample solution, the inert purge gas, the potassium thiocyanate inlet and the titrant. The inert gas and drain outlet are both controlled by standard solenoid valves that are obtainable from Baird and Tatlock Ltd., Chadwell Heath, Essex. All the valves used in the instrument are of the energise - to - open type.The stirrer This is shown in Fig. 5.Fig. 3. General view of titrator [To face page 228April, 19661 FOR THE DETERMINATION OF TITANIUM 229 has a shaft of diameter 5 to 6 mm, and is fitted with four blades. The bearing is machined from PTFE, and the drive is by a flexible shaft from a'456 r.p.m., 250-volt ax., 50 c/s geared induction motor. kmj ,Ground toform I -- d7%Lva've Floating Seat I IzI I Fig. 6. Potassium thiocyanate vessel Fig. 6 shows the vessel from which 10 ml of potassium thiocyanate solution are trans- ferred to the beaker. Solenoid valves are used to control the filling and emptying of the vessel, the seal being effected by a floating glass valve which is simpler and more reliable than a system of level electrodes.The inert-gas supply is also controlled by a solenoid valve. A standard calomel electrode with a porous plug base is used. The platinum indicating and earth electrodes consist of 25 s.w.g.-platinum wire fused into a glass tube so that a length of 1 cm is exposed. TITRANT SYRINGE- This consists of a glass syringe of 20-mm i.d. barrel, and a piston made of PTFE with a rubber sealing ring. The piston is attached to a leadscrew driven by a 250-volt a.c., 50 c/s, 19-r.p.m. reversible geared induction motor. The leadscrew also drives a mechanical counter which indicates the amount of titrant added. The early form of syringe consisted of a glass piston and barrel, and considerable care had to be taken in the assembly of the drive mechanism to prevent fracture of the syringe.Replacement of the piston by one machined from Teflon, fitted with a rubber O-ring seal eliminated this trouble, and at the same time increased the accuracy of the delivery. The filling and emptying of the syringe is controlled by a parallel pair of solenoid valves mounted in a Perspex block. When the syringe is full, a micro-switch, Sw,, is operated which automatically closes the fill valve. The empty valve is opened when the drive motor starts to operate. If, in the preliminary experiments, a speed of delivery was selected such that over- shooting of the fill position or the end-point did not occur, the rate of filling of the syringe was very slow and the duration of determination was lengthened appreciably. Accordingly, a two-speed gearbox was designed and fitted to the drive, so that a fast rate could be used for most of the filling and emptying, and a slow rate when the end-point or the fill positions were being approached.A diagram of the gearbox system is shown in Fig. 7 . The change- over is operated by a micro-switch, Sw,, when the syringe is being filled, and by an anticipation control when the end-point is approached. CONTROLS- The titrator consists of three basic units, all of which are readily detachable and replace- able. The end-point indicator, set-buff er control, scale-length control, manual - auto switch and anticipation control are contained in the end-point unit which is housed in the upper part of the instrument. A switch is also incorporated to allow for change in the direction of the titration.230 DENTON AND WHITEHEAD : AN AUTOMATIC APPARATUS [AutdySt, VOl.91 I Sliding spindle ‘ ~ o g coupling Fig. 7. Two-speed gearbox The centre unit, which controls the sequence of operations, contains the mains on - off switch, the start button, the manual advance control and indicating lights. A fill-deliver control is also installed for use in priming the syringe when a new one is being fitted. The manual advance control enables wash stages to be by-passed if required. The lower section houses the titration assembly. SEQUENCE OF OPERATIONS The instrument has been designed so that many of the operations take place simul- taneously, resulting in a considerably reduced cycle time; this is summarised in Table I. Before starting a determination, the ferric alum, potassium thiocyanate and wash solution reservoirs should all be filled, and the ferric alum syringe should be primed by manual operation of the fill-deliver control.The instrument should be switched on for approximately 1 hour before use to allow the end-point system to become stabilised. The “stand-by” light, L,, should be illuminated to indicate that the Uniselector is in the first or “stand-by” position. Reference should be made to the sequence unit circuit shown in Fig. 8, for details of the circuit, and to the appendix for component details. The sample solution is transferred to the sample beaker and the start button pressed. This closes Swla and Swlb which energises relay RLS. Closure of Sla maintains relay RLS in an energised state when the “start” button is released.Energising relay RLS moves the Uniselector to position 2, indicated by the lighting of L,. UNISELECTOR POSITION 2- Contact Slb is closed, thereby opening the potassium thiocyanate delivery valve, and potassium thiocyanate is added to the beaker. Slc is closed and the circuit through RLB to earth is completed via electrode probes. The wash syringe motor is started when S2 is closed and S3a opens causing the beaker drain valve to close. Relay RMF is then energised by change over of S3b. Closure of MF1 energises the “fill titrant” valve, causing it to open.Uni- selector position 1 2 3 4 Potassium thiocyanate deliver Closed Open. Solution added to titration beaker Open TABLE I SUMMARY OF OPERATIONS CARRIED OUT AT EACH UNISELECTOR POSITION > 3.L + W 45 45 U Inert-gas valve Closed Open Open Beaker drain valve Open Closed Closed Ferric alum feed valve Closed Open Closed Closed Open Open Closed Closed Open Closed. Vessel fills Open Open Closed Closed Closed Closed Titration syringe Stopped Syringe fills with ferric alum a t fast rate until the last 50 revs., which are added a t the slow rate. When the zero end stop is reached, the ferric alum feed valve is closed Titration starts a t slow rate for first 50 revs., then a t fast rate until the anticipation is reached. At the end of the titration, i.e., when the end-point potential is maintained for 20 seconds, the Uniselector moves to position 1 Wash solution valve Closed Closed Open. Solution added to top level probe, then valve closes (first wash) Open.Solution added as above (second wash) Open. Solution added as above (third wash) Closed Wash syringe motor Pumps sample solution from beaker, through the reduction column to the titration beaker. When the liquid level falls to the lower level probe, the Uni- selector moves to position 3 Solution pumped as above, and when the liquid level falls to the lower probe the Uniselector moves to position 4 As above. When the lower probe is reached, the Uniselector moves to position 5 As above. When the lower probe is to reached, position the 6 Uniselector moves Stopped232 DENTON AND WHITEHEAD: AN AUTOMATIC APPARATUS [Ana&St, VOl. 91 h I aApril, 19661 FOR THE DETERMINATION OF TITANIUM 233 The titrant motor is started when MF2 changes over and MF3 opens. The phase difference ensures that the titrant motor starts rotating in the correct direction and titrant is pumped to the syringe.The rate of filling the syringe is increased by changing the gear drive. This is accomplished when MF4 is closed and relay RLG is energised. GI is closed which energises the gearbox solenoid and the gear is changed to the fast rate. Immediately before the syringe is full, the rate of filling is decreased to avoid over- shooting the end-point. This is achieved by the opening of Sw, when the leadscrew reaches a pre-set position. The gearbox solenoid is de-energised and the gear is changed back to the slow rate. When the titrant syringe is full, Sw, is opened; this de-energises RMF. Contact MF2 reverts to the normal position and the titrant motor stops, the braking components ensuring that the motor stops instantaneously. The above operations are all carried out on position 2 of the Uniselector. The sample is gradually pumped from the sample beaker to the reductor column and the beaker is washed out with three lots of wash solution.The detail of how this is done is as follows: When the liquid level falls below probe 2, relay RLB is de-energised and closing of B2 energises the Uniselector coil. B1 closes which energises relay RLC. This relay remains energised because C2 is closed. C1 closes, thus causing wash solution to pass into the beaker. UNISELECTOR POSITIONS 3 TO 5- When the liquid makes contact with probe 2, relay RLB is energised which opens B2. This advances the Uniselector to stage 3, lamp 2 is switched off and lamp 3 switched on.Kote that B1 is opened, but owing to C2 being closed relay RLC remains energised. When the liquid level makes contact with probe 1, relay RLA is energised which opens Al. This de-energises relay RLC which opens C1 and the flow of wash solution stops. This wash procedure is repeated until three washes have been completed and the uni- selector is on position 5. UNISELECTOR POSITION 6- T1 is opened which ( a ) , closes the potassium thiocyanate deliver valve, ( b ) , keeps the drain valve closed and (c), de-energises relays RLS and RLC. This stops the wash syringe motor. T2 is closed, so starting the stirrer motor. T3 opens to maintain the Uniselector in position 6. T4 changes over and relay RMD is energised which ( a ) , opens the titrant delivery valve and ( b ) , starts the titrant motor in the delivery direction. Note that D4 is always closed at the beginning of a titration.Titrant is then delivered to the beaker a t the slow rate. During the titration the potassium thiocyanate vessel is filled, since the fill valve is directly in contact with position 6 on the Uniselector. When the relay RLY is energised owing to Sw, being closed by the lead screw, the contact Y1 maintaining the circuit through RLY when the Sw, circuit is broken, Y2 closes, energising relay RLG and so causing the gear to change so that titrant is delivered at a fast rate. At the anticipation setting, D4 opens, relay RLY is de-energised and Y1 and Y2 open. Relay RLG and hence the gearbox solenoid are de-energised and the titrant addition reverts to the slow rate.Closure of D4 causes addition of titrant to occur. Titrant is added in short bursts until the potential difference between the electrodes is the same as the end-point potential and has remained unchanged for a pre-set period. El opens and de-energises relay RLT so that T3 closes and the Uniselector operates on its motoring contacts and returns to position 1 indicated by lamp L,. In this position T1 closes and the beaker drain valve opens. Since position 1 of bank 1 on the Uniselector is not earthed, the carbon dioxide valve closes. When the Uniselector is moved to position 6, relay RLT is energised. Relay RMD is de-energised and the titration stops. The titration is controlled by the signal from the end-point detector.METHOD CALIBRATION AND STANDARDISATION- millivolts is required. millivolt supply and proceed as follows : To standardise the end-point detector a high-impedance potential source of 0 to 500 Disconnect the electrodes from the end-point detector, connect the234 DENTON AND WHITEHEAD: AN AUTOMATIC APPARATUS [Analyst, VOl. 91 Set the end-point pH control to 400 and the potential supply to 100 millivolts. Adjust the set-buffer control until the meter reads zero, Change the potential supply to 300 milli- volts and rotate the end-point pH control until the meter reads zero. The end-point pH control should read 800. If the reading is not 800, adjust the scale length control and repeat the previous operations until the correct scale length is obtained.Disconnect the potential source and connect the electrodes to the end-point detector. Transfer to the titration vessel a titanium solution which has been titrated to the end-point. Set the end-point pH control to 250 and adjust the set-buffer control until the meter reads zero. Set the “anticipation” control to a value which will cause the first anticipation to occur about 5 ml from the end-point. This value can only be found by trial and error. Weigh accurately to 0.1 mg a catch weight of approximately 0.9 g of titanium dioxide (99.9 per cent, purity) which has been previously dried in an air-oven at 105” C. Transfer to a 250-ml beaker, and dissolve by heating with 40 ml of concentrated sulphuric acid and 30 g of ammonium sulphate. Cool and dilute to 500 ml in a calibrated flask with distilled water.Pipette a 50-ml aliquot into the sample beaker, set the manual - auto switch to the manual position and press the “start” button. When the titanium solution has been reduced and transferred to the titration beaker, continue the titration manually and construct a curve of titrant versus potential difference. Select the setting for the end-point by inspection of the graph and use this value for all subsequent titrations. Set the manual - auto switch to auto and repeat the titration using a 50-ml aliquot each time. The calibration should be carried out in triplicate, and the volume of ferric alum used in each case noted in revolutions. The scatter between replicate titrations should be within k0.1 per cent. The weight of titanium dioxide per revolution can then be calculated.N.B.-This procedure eliminates the need to standardise the ferric alum before use on the titrator. REAGENTS- The instrument is now set approximately to the required end-point. Cadmium granules-1 to 2 mm in size. Potassium thiocyanate solution, 20 per cent. w/v. Wash solution-Prepare 10 per cent. w/v ammonium sulphate in sulphuric acid (1 + 50). Ferric alum solution, approximately 0.0625 N-Dissolve 1200 g of ferric alum, A.R., in 400 ml of concentrated sulphuric acid and 7 litres of water. Dilute to 36 litres and mix well. Oxidise any ferrous iron by adding 0.1 N potassium permanganate solution until one more drop will just colour 50 ml of the solution. PROCEDURE- Any substance which is reduced by cadmium metal and re-oxidised by ferric alum will interfere in the determination. Interfering agents include chromium, vanadium, niobium, tungsten, uranium, platinum, molybdenum, arsenic and antimony, as well as nitric acid and certain organic compounds.Those which occur in titaniferous materials are chromium, vanadium, niobium and molybdenum. Chromium, vanadium and molybdenum reduce to a definite valence and may be allowed for in the determination if their concentration is known. To determine the titanium content of a solution, a suitable aliquot is transferred to the sample beaker, the manual - auto control is set to auto and the start button pressed. At the end of the determination the weight of titanium dioxide present in the solution is calcu- lated from the reading on the counter and the conversion factor for the ferric alum titrant.Note that there is no need to rinse the titration beaker between determinations because the residual solution is at the end-point condition and will not interfere in the titration of the next sample. RESULTS Then add 20 ml in excess. The accuracy of the instrument was assessed by making 30 determinations on the same sample. The following results were obtained- TiO,, per cent. w/v . . 15.35 15-36 15.37 15-38 15.39 Number of results . . .. 1 9 4 3 5 TiO,, per cent. w/v . . . . 15.40 15.41 15.42 15.44 Number of results .. . . 3 3 1 1 Average determination-15.38 per cent. Standard deviation-0.023 per cent. Coefficient of variation--0.145 per cent.April, 19661 FOR THE DETERMINATION OF TITANIUM 235 PERFORMANCE A complete determination takes 7 minutes, during which time the operator is available to carry out other work such as the preparation of the next sample.The instrument has resulted in a substantial saving of time and has also increased the accuracy of the deter- mination by the use of a potentiometric instead of a colorimetric end-point. It is natural that such an instrument requires careful maintenance, and a scheme of routine preventative maintenance has been drawn up which has greatly contributed to reducing the down-time of the instrument. It has been found that particular attention should be paid to the elec- trodes and to ensuring that the solutions are iree from insoluble particles. The cadmium granules should be sifted to remove any fine particles before use.The instrument has been in use in the laboratories of this Company for longer than a year and has given satisfactory service during this period. The total cost of the instrument, including labour, materials and commissioning, is approximately L550. We thank Mr. R. Hutton and Mr. D. L. Suttill for technical assistance, and the Directors of British Titan Products Company Ltd., for permission to publish this paper. Appendix I COMPONENTS LIST FOK END-POINT DETECTOR (FIG. 1) Unless otherwise stated, the resistors have a tolerance of k2 per cent. and are made R1, R2, R,6 R,i, = l.8-megohm, 0.25-watt resistors R3 = 27,000-ohm, 0.25-watt resistor R4, R,, R,, R,,, R,,, R3, = 1-megohm, 0-25-watt resistors R61 R 7 = 10-ohm, 0.25-watt resistors R t 7 p R19, R28, R36 = 22,000-ohm, 0.25-watt resistors R,,, R,,, R,,, R,,, R,,, R,, = 2.2-megohm, 0.25-u-att, resistors R151 Rl13 = 39,000-ohm, 0.25-watt resistors Rl, = 1800-ohm, 0.25-watt resistor R21 = 2-megohm, 0.26-watt resistor R24 = 10-megohm, 0-25-watt resistor R25 = 560,000-ohm, 0.25-watt resistor R29, R3; = 33,000-ohm, 0.25-watt resistors R30 = 560@-ohm, 0.25-watt resistor R33 = 33-megohm, 2-watt resistor R341 R44 = 1000-ohm, O.25-watt resistors R35 = 470,000-ohm, 0.25-watt resistor R42 = 22-megohm, 0.25-watt resistor R45 = 180,000-ohm, 0.25-watt resistor %n = 25,000-ohm, 4-watt, 10-turn helically-wound Potentiometer, with a resis- tance tolerance of 5 per cent.and linearity tolerance of &0.5 per cent. R,, = 25,000-ohm, I-watt, wirc-wound potentiometer with a tolerance of h20 per cent.R3 1 = 3300-ohm, 5-watt, wire-wound resistor with a tolerance of 5 per cent. R32 = 22,000-ohm, 5-\vatt, wire-wound resistor with a tolerance of 5 per cent. R40.4, R40~ = 100,000-ohm, O-.Fi-watt, twin-gang potentiometer with a tolerance of R4 1 : 100,000-ohm, l-watt, wirc-wound potentiometer with a tolerance of Cll c, = 0.005-pF tubular paper capacitors, 500-volt, d.c., working ‘3, ‘4 = 0.001-pF tubular paper capacitors, 500-volt, d.c., working C5 = 0.25-pF tubular paper capacitors, 500-\dt, d.c., working C6 = 1-pF tubular paper capacitor, 500-volt, d.c., working of high stability carbon. 1 2 0 per cent. *20 per cent. R4,,* increases as R4,,~, decreases. = 0.05-pF tubular paper capacitor, 500-volt, d.c., working = 32-pF electrolytic capacitor, 450-volt, d.c., working = 2-pF paper capacitor, 600-volt, cl.c., working = 8-pF electrolytic capacitor, 500-volt, d.c., working = 0.02-pF tubular paper capacitor, 500-volt, d.c., working c, C8 C , ClO Cll Cl, = 20-pF metallised paper capacitor, 150-volt, d.c., working MR1, MR, = 250-volt, r.m.s., silicon metal rectifiers = Round, double-pole, double-throw toggle switch - $pole, %way wafer switch = ME 1400 valve = EF 37A valve = ECC 83 valve Sl sz v, v2 v3 v4, v,, v6, vll = 6060 valves236 DENTON AND WHITEHEAD [Artalyst, VOl.91 = E Z 80 valve = 108 C1 valves = Constant voltage transformer, input : 240 volts, 50 c/s; output: 6-0 volts = Main transformer, input: 200-250 volts; output: 250-0-250 volts, 60 mA; = 20,000-ohm coil, 4-pole change-over, Post Office type 3000 plug-in relay = 20,000-ohm coil, 2-pole change-over, Post Office type 3000 plug-in relay = Centre-zero microammeter, 2.5 inch diameter, 100-0-100 pA full-scale = Painton Multicon 12-pole socket 6.3 volts, 3 amps and 6.3 volts, 1 amp deflection RLD/4 RLE/1 M P Appendix I1 COMPONENTS LIST FOR THE SEQUENCE UNIT (FIG.8) MR,, MR,, MR,, MR, RLA. RLB. RLC MR3, MR, RLG; RLS,.RLT, RLY RMD, KMF 1. 2. 3. 4. 5. 6. 7. 8. = BjO-ohrn, l-watt, high-stability carbon resistors = 600-ohm, &watt, wire-mound resistors = 100-ohm, B-watt, high-stability carbon resistor = 1-pF, 500-volt d.c. working paper capacitors = 64-pF, each composing 2 32-pF, 450-volt d.c. working, electrolytic = 0-l-pF, 500-volt d.c. working, tubular paper capacitor = REC 23a metal bridge rectifier, 54 volts r.m.s., 1 amp input, 50 volts = REC 51A silicon rectifier, 250 volts r.m.s., 500 mA capacitors d.c. output = Miniature plug-in relays, 24-volt d.c., 650-ohm coil, 4-pole change-over = Miniature flag-in relays, 48-volt d.c., 1850-ohm coil, 4-pole change-over = Push to make, release to break, double-pole double-throw switch = Push to break, release to make, double-pole double-throw switch = Centre-off, double-pole double-throw toggle switch = Round double-pole double-throw toggle switches = Micro-switches = Micro-switch = $-inch 2-amp fuse = Transformer: primary, 250 volts; secondary, 50 volts, 2 amps = Standard filament transformer: primary, 250 volts; secondary, 6.3 volts = 3-bank, 12-position, plug-in type Uniselector unit, 250-ohm coil = 6-3 volts, l-watt, panel lamps contacts rated a t 3 amps a.c. contacts rated a t 250 volt, 3 amp a x . REFERENCES Wilson, C. L., and Wilson, D. W., Editors, “Comprehensive Analytical Chemistry,” Elsevier Kolthoff, I. M., and Elving, P. J ., Editors, “Treatise on Analytical Chemistry,” Interscience Nakazono, T., J . Chem. SOC. Japan, 1921, 42, 526, 761. Kahm, J. A., A n a l y t . Chem., 1962, 24, 1832. Malmstadt, H. V., and Roberts, C. B., Ibid., 1956, 28, 1884. Brown, J. F., and Weir, K. J., Analyst, 1958, 83, 491. Labovatory Equipment Digest, Volume 2, KO. 3, p. 11. British Patent 39,786, 1962. Publishing Company, Amsterdam, London, New York, 1962, Volume lC, p. 501. Publishers, New York and London, 1961, Part IT, Volume 5, p. 30. Received A p r i l 14t12, 1965