Electrometric Titration Device for Automatically Stopping at a Predetermined End Point W. E. SHENK and F. FENWICK, Research Laboratory, United States Steel Corporation, Kearny, N. J.
T
HIS device makes use of
The amplifying circuit, shown A device is described which automatically schematically in Figure 2, is of the p o t e n t i a l difference stops an electrometric titration by utilizing the two-tube bridge circuit type b e t w e e n t h e two electhe sudden change developed in the poten(8) which has t h e a d v a n t a g e trodes involved in a n e l e c t r o tial difference between two dissimilar metal that the operating point a t which metric titration as a means of electrodes at the end point to actuate a stopping the addition of the tithe output voltage is zero is not affected by fluctuations in supply trating agent at a set end point. shut-off. The instrument is made of easily It is positive in action, is made voltage. Since the thyratron obtainable parts and requires no source of used (FG-57) has a cut-off grid up largely of standard parts, and power besides the lighting circuit and a in o p e r a t i o n r e q u i r e s little potential of -1 volt, and 1 volt single dry cell; it is practically independent trained attention; the reproducicorresponds to about 5 millivolts of fluctuations in line voltage. Its usefulbility of its end point equals, or input, any change in the supply voltage changes the cut-off point closely approaches, that obtained ness is demonstrated by application to with careful manual operation three titrations-the titration of Fe++ of the circuit by much less than 5 millivolts o n t h e e l e c t r o d e requiring continuous attention with potassium dichromate, of CrSO,-potential. I n order to make during the whole course of the with Fe++, and of Zn++ with potassium even this error small, the volttitration. The only analogous ferrocyanide. age regulator tube m a i n t a i n s automatic device of which the the somewhat critical controlauthors are aware is that described by Muller and Partridge (I), who used the color-change grid and screen-grid potential nearly constant for a change of an indicator as a means of actuating a photoelectric cell to of * 10 per cent in supply voltage. Since some small current is drawn from the titration cell, it shut off the flow of the titrating agent. is advantageous to avoid the use of a constant half-cell such If the potential between the two electrodes is to be used to as the calomel or silver chloride electrode, and instead to use actuate the control, the circuit must be such that the current two dissimilar metal electrodes, both in direct contact with drawn from the titration cell is exceedingly small; otherwise the titrated solution. The authors chose the electrode systhere will be serious polarization of the electrodes which is tem, platinum-tungsten. I n presence of a vanishing active likely to alter the electromotive force of the cell so much as to ion concentration, the potential change of tungsten is usually make uncertain the determination of the true end point. much less than that of platinum, and so tungsten may replace Moreover, there is the condition, applicable to any accurate the more cumbersome half-cell commonly used as a reference electrometric titration, that the development of the electrode electrode in potentiometric titrations. The selected elecpotential must very nearly keep pace with the addition of trode system is readily obtained, convenient, and easy to the titrating agent. This requires that the electrodes come keep in order; an occasional ignition of the platinum wire in rapidly into substantial equilibrium with the solution in the flame of a burner and a rubbing of the tungsten with fine immediate contact with them, and there must be stirring so emery suffices. Very little attention need be paid to the thorough that the ion concentrations in the solution a t the electrodes when in continuous service; but it is advisable ta electrodes are a t any instant substantially the same as in the treat them as described if they have been in disuse for any solution as a whole. Unless these conditions are met, the length of time. For electrometric work generally ignition electrodes will not actuate the cut-off until some excess of the of the platinum electrode is so much the best pre-treatment titrating agent has left the buret; yet with proper calibration to insure a sharp end point that it is wise always to employ even this need not cause an appreciable error in the result. a simple wire electrode which may easily be removed and An apparatus which meets the requirements, whose only ignited and to avoid more complex, difficult to ignite, elecsource of power is the lighting circuit and a single No. 6 dry trodes of larger surface, such as foil sealed into glass. cell, was built about two years ago and proved to be satisThe amplifying circuit is arranged to operate a t any defactory. sired potential difference between the electrodes whether the difference is increasing or decreasing as the end point is passed. Description of Apparatus For a given titration this potential difference is greater on The set-up, shown in Figure 1, consists essentially of two that side of the end point in which the reaction between the dissimilar electrodes, a vacuum-tube amplifier circuit, a solesolution and the platinum electrode is less reversible. If the direction of the titration is toward decreased reversibility, the noid-operated clamp which stops the flow of the titrating end point is marked by a sudden and very pronounced inagent from an ordinary buret by pinching a short rubber tube crease in the observed potential of the cell. This happens attached to the tip of the buret, and a suitable stirrer. The potential between the electrodes is introduced into the grid with many common oxidation reactions; but it does not follow circuit of the amplifying tube, and amplified so that a change that all oxidations are accompanied by a rise in potential or that the reverse is true of reductions; for, since the actual of approximately 5 millivolts produces a change in the grid potential at the thyratron sufficient to cause a positive interpotential difference depends upon the relative reversibility of ruption of its plate current at the next alternation of its plate the electrode reactions on the two sides of the end point, it voltage. This interruption of the output of the thyratron cannot be said that oxidation always makes for an increasing potential and reduction the reverse, I n order that a reaction deenergizes the solenoid and so closes the clamp which normally is held open by the current against a spring. may serve as an analytical tool, it is necessary that the re194
MAY 15, 1935
ANALYTICAL EDITION
19s
1. Plug in the power supply and allow the filaments to heat for 5 to 10 minutes, being sure that switch Sa is in the “off” position. 2. Throw switch SIto STD and switch S, to either I N C R or DECR, depending upon whether the electrode potential is increasing or decreasing at the end point. 3. Adjust potentiometer P until the voltmeter V reads 100 mv. 4. Adjust rheostat RI (coarse adjustment) until the plate current, as indicated by ammeter A , is within the range in which the tube is most sensitive to a change in grid voltage. This range varies with the type of the tube, the plate load resistance ( R p J , and the plate supply voltage. For the circuit shown a current of about 0.4 ma. is proper. 5. Close switch 8,and adjust R, until the pilot light just turns on. If in the course of this adjustment the plate current goes much above 0.4 ma., R, should be readjusted t o decrease it to 0.4 ma. and then R, reset. 6. Turn R1 (fine adjustment) until the pilot light just goes out. The cut-off of the nmaratus is now set at 100 mv. The setting may be checked bj; turning potentiometer P until the pilot light is illuminated and then turning P in the opposite direction until the light just goes out; the reading on the voltmeter then indicates the cut-off point and should be 100 mv. 7. With the solution to be titrated in position start the stirrer, throw switch SIto TZTN, read the buret, then force open the solenoid-operated clamp to start the titration. The solenoid is purposely made too weak to open the clamp but when the clamp is opened manually the solenoid is sufficientiy powerful to hold it open as long as the solenoid is energized. The pilot lamp is FICIJRE 1. ASSEMBLY FOR AUTOMATIC CONTROL OF ELECTRO- illuminated whenever the solenoid is energized. The rate of METRIC TITRATIONS flow from the buret may be controlled by the regular buret cock, but it is advisable to attach a rather fine glass tip to the lower end versibility of the potential of a n inert electrode in contact of the rubber tube leading from the buret,. The authors found a with the titrated solution change as the end point is passed; rate of addition of the titrating solution of 1 drop (0.025 ml. per second satisfactory. Much depends upon the efficiency o the this holds for the electrode pair adopted, which may, therestirring. The stirring device (shown in Figure l), equipped with fore, be used for electrometric titrations involving oxidation, a glass stirrer with two sets of blades, was operated as ra idly as reduction, neutralization. and mecititation reactions. feasible within a beaker somewhat more than half full o r liquid. The instrument is set to interkpt i h e current to the solenoid at a definite potential. This means that the potential a t the proper end point must be known for the electrode system selected, but this is not such a drawback as it may appear. On the more reversible side of the end point the observed difference in potential between tungsten and platinum is usually quite small, but in the near vicinity of the end point, it changes sharply with a small added amount of the titrating solution. The precise setting of the cut-off may, therefore, vary within a considerable range without affecting appreciably the corresponding volume of the titrating soluti~onrequired. The best way to ascertain the proper cut-off point for a particular titration is to plot the titration curve as obtained with the identical electrodes and a good potentiometer. The inflection of this curve, about the mid-point of the abrupt change in potential, fixes the proper cut-off. Fortunately the position of the inflection is little affected by differences in concentration of ions other than those entering directly into the titration reaction. In many cases the automatic titrator may also be set by using a color indicator, determining the FIGURE 2. SCHER~ATIC CIRCUITDIAGRAM cut-ojlf potential when the color change occurs, V . Weston model 301 switchboard direct current voltmeter, double scale, 0 to 200 mv., and setting the instrument accordingly for sub0 to 1000 mv. No. 6 dry cell placed across terminals through voltage divider sequent titrations; but this is a less desirable A . Weston model 301 switchboard milliammeter 0 t o 1 ma. RCA-57. Triple-grid amplifier tube Radio Cordoration of .4merica procedure. UX-874. Volta,ge regulator tube, uied here t o help maintain constant the screen grid and control grid biases The method of setting the apparatus to cut FG-57. Thyratron tube, General Electric Corn any off at any desired potential is quite simple, and RCA-82. Full-wave mercury vapor rectifier tu!, Radio Corporation of America Ri and R z . Rheostats of about 400 ohms; R2 rheostat of about 50 ohms once the setting is made it need not be readRo. 1-megohm fixed resistance C. Condenser 0.02 mfd. justed unless the apparatus is shut down or a Filter is compdsed of two 30 henry inductances and 4 mfd condensers different titration is to be performed. For any RPi and RPx. 500,000-ohm fixed resistors; the potentiometer connected between them has 200.000 ohms total resistance particular cut-off point, say 100 mv., the proPower transformer American Transformer Company; 4 secondaries: 2.5, 2.5, 870, and 5 volts, rating 75 ;a. cedure for adjusting and using the apparatus Insulating transformer American Transformer Company; 1 to 1 ratio of sufficient sise to is as follows (Figure 2) : supply current requhed by solenoid
1
ISDUSTRIAL AND ENGIKEERING CHEMISTRY
196
8. When the pilot lamp goes out. the flov of reagent from the buret is cut off, and the buret is read. If the pilot lamp should happen to remain lighted though the clamp has closed, the power supply has probably been interrupted during the course of the titration, and it is necessary t o open the clamp t o allow more reagent to flow until the true end point is reached.
The amplifier as constructed uses 115 volts alternating current as power supply a t any frequency from 25 to 60 cycles per second a n d , c o n s e q u e n t l y , is slightly larger and heavier than would be necessary for 60-cycle o p e r a t i o n alone. The only parts which are influenced by the frequency are the two transformers, which have to be designed for the lowest frequency that is to be used. A 25-cycle transformer will work satisfactorily a t 60 cycles but a 60-cycle transformer will probably burn. out a t 25 cycles. The main items required are readily obtainable with the exception of the solenoidoperated clamp, which must be made. This is illustrated in some detail in Figure 3. The amplifier proper was built for about $125, e x c l u s i v e of case.
Applications These reactions were selected for
FIGURE 3. DETAIL the purpose of testing the automatic OF CLAMP Awuthe titration of Fe++ ATED BY SOLENOID titrator-namely, Jaws Constructed of 0.16-cm. (0.0625-inch) brass. Solenoid core about 0.926 om. (0.375 inch) diameter.
w i t h p o t a s s i u m d i c h r o m a t e , of CrzO7-- with Fe++, and of Zn++ with potassium ferrocyanide.
OF FE++W T H POTASDICHROMATE. It is especially
TITR.4TION
SIUM
desirable to control the dichromate titration of iron automatically because it is frequently necessary to determine the iron content of a large number of samples of ores and slags under conditions that make a reasonable degree of accuracy with a minimum of skilled attention essential. A 25-nil. portion of a stock solution of ferrous sulfate was acidified with hydrochloric acid, and the usual reduction with stannous chloride was carried out, followed by addition of mercuric chloride; it was then titrated with a 0.1 N solution of potassium dichromate (which had been recrystallized three times and dried to constant weight a t 200" C,), and proved to be 0.0667 N . The full line in Figure 4a is the electrode potential curve of this titration, in the near vicinity of the end point, as followed with an appropriate potentiometer; this serves to standardize the iron solution as well as to fix the proper endpoint potential between the platinum-tungsten electrodes. The broken line is the curve for a similar titration, except that phosphoric acid and the color indicator diphenylamine sulfonic acid had been added. The difference in potential a t the two inflections, 120 and 100 mv., is not regarded as significant for it is a t most equivalent to two drops of the titrating solution. The organic color indicator does, however, lower the observed oxidation potential in presence of excess of dichromate. The true electrometric and the color end points are very nearly coincident, the former occurring just before the latter. Table I presents typical results of a number of subsequent titrations of the stock iron solution, carried out by the automatic titrator set at the cut-off potential stated. Addition of the color indicator showed that the titrator did not stop the flow of the dichromate solution, about one drop (0.028
VOL. 7, NO. 3
ml.) per eecond, until the latter Ti-as present in some excess; with the cut-off set a t 100 my., which is very close to the point of inflection of the titration curve, the excess was in one case as much as 0.20 ml. The over-running is due, not to any lag in the amplifying circuit, but to the fact that the attainment of complete equilibrium between the electrode system and the solution as a whole is not instantaneous, the actual lag depending upon the condition of both electrodes and especially upon the efficiency of the stirring. With the authors' set-up the error due to this lag would be negligible if the rate of flow were about one-fourth of that adopted; but a titration carried out wholly a t this rate would require an hour or more, which would raise a serious question as to the practical usefulness of the automatic titrator. Indeed a question as to its usefulness arises from the fact that the point a t which it cuts off is not as close to the true end point as is the color-change of the indicator. On the other hand, it is unaffected by the presence of highly colored ions, such as C r + + + ,M i + , and Cu++, and in any case requires no attention during the whole course of the titration. Moreover, for a given set-up and a given rate of flow of the solution from the buret, the error, being substantially constant, can be brought within the limits ordinarily required in analytical work by application of a small correction, about 0.10 ml. in this case. According to the data in Table I, the error was reduced somewhat by setting the cut-off a t 75 mv., and became negative with the cut-off set a t 50 mv.; it may therefore be partly taken care of by setting the circuit to actuate the cut-off somewhat in advance of the potential difference a t the true end point as determined by preliminary titration with a good potentiometer. I n a series of similar titrations, therefore, the use of the automatic titrator would yield results of sufkient accuracy with a great saving in time and attention required on the part of a skilled operator. TITRATION OF Crz07-- WITH Fe++. Thc titration of Cr,O,-- with F e + + is the reverse of the titration iust discussed. The coior indicators of t h e d i p h e n y l 4" amine g r o u p c a n n o t b e added to the dichromate 400 solution because they are d e s t r o y e d b y the strong oxidizing agent; so that it is n e c e s s a r y t o a d d a n excess of Fe++, then the 303 i n d i c a t o r , and to determine the excess F e + i with a s t a n d a r d s o l u t i o n of potassium dichromate. T h i s complication is 200 avoided by the use of the automatic titrator. Figure 4b shows t h e e n d - p o i n t curve of two typical titrat i o n s of t h e s t a n d a r d lcci dichromate solution; the solutions as titrated coiitained about 20 per cent by volume of reagent hydrochloric acid. The potential change in the new vicinity of the end point FIGURE4. ELECTRODE Pois much larger for a given TENTIAL CURVEIN NEAR VICINITYOF END POINT i n c r e m e n t of the titratFOR TITRATION OF: ing s o l u t i o n t h a n is the case for the reverse titraF e + + solution with 0.1 N (3rs.or.prT- solution: full line, tion. The mean 'of several color indicator absent. broken line, oolor indicator p;esent. ( 6 ) t i t r a t i o n s in close agreeCrz07-- solution with 0.1.N t e + + m e n t showed t h e Fe++ ~ o l u t i o n ; two typical titrations
AK-4LYTICA4LEDITIOS
S'IAY 15, 1935 T.4BLE
I. TITRATIOX OF Fe+T WITH 0.1 A- POTASSIUM DICHROMATE
Cut-off Setting
(True end point at 16.68 ml.) Volume Shut Off .4utomatically
MV.
MI.
100 100 100 100
16.66 16.84 16.86 16.84
-0.02 tO.16
75 75 i5
16.76 16.61 16.81
$0.08 -0.07 +0.13
50
16 62 16.61
-0.07
50
Error M1. +0.20
4-0.16
-0.06
TABLE 11. TITRATION OF Crz07-- WITH Fe++ cut-off
Ratting M V.
250
100
Equivalent Volume 0.1 N FeSO Shut off Automatically M1. 20.04 20.05 20.09 25.10 25.14
Error M1. +0.04 $0.05 +0.09 +0.10 +O. 14
20.15 20.10 20.09 30.02
$0.15 $0.10 +0.09 +0.02
20.00 20.00 20.00
20.04 20.10
+o.
20.00
20.21
$0.21
Volume 0.1 N K2Cr107 Taken MI.
197
gram of zinc per i d . Samples of this solution were titrated electrometrically with a solution of potassium ferrocyanide a t 65" C. To each sample 13 ml. of reagent ammonia and 3 ml. of excess hydrochloric acid, methyl orange serving as the indicator, were added before titrating. Figure 5 gives the end-point curve of two typical titrations; 400 mv. was adopted as the end-point potential. One milliliter of &Fe(CN)s was found equivalent to 1.034 ml. of Zn++ from the titration curves. The titrations made with the automatic titrator, which are listed in Table 111,were also made at 65' C. The results for this ordinarily rather troublesome titration are regarded as favorable. TABLE111. DETERMINATION OF Zn++ WITH POTASSIUM FERROCYANIDE
i.O.04 10
Cut:Ofi Setting >I\,. 4 00
?0.00
Equivalent Volume KIFe(CN)s Shut Off Automatically M1. 20.14 20.11 20.04
+0.14 $0.11 +0.04
50.00 50.00 50.00
50.22 50.17 50.18
+0.22 +O.li +o. 18
Volufne Zn++ Solution Taken M1.
20.00 20.00
Error
M1.
It may be pointed out that on all the titrations cited a correction of -0.10 ml. applied to the buret reading at the close of the titration brings the error involved within acceptable limits for most analytical work. This correction is largely a function of the authors' working conditions, particularly of the rate of addition of the titrating solution and the effectiveness of the stirring.
solution (not the same solution referred to above) to be 0.0669 X , and the end-point inflection to lie close to 250 mv. Table I1 presents the data obtained with the autoinatic Literature Cited titrator. A considerably greater range in the setting is permissible than with the reverse titration, as the curves indicate. (1) Muller and Partridge, IND.Ewo. CHEM.,20,423 (1928). A setting of 100 mv. is, however, clearly too low; the error (2) Nottingham, J . Fmnklin Inst., 209,287 (1930). with the higher settings is not serious but always positive. RECEIVEDMarch 19, 1935. DE'PERMINATION OF ZINC WITH POTAfiS1Uh.I FERROCYANIDE. The volumetric determination of zinc with potassium ferrocyanide is particularly suited to the autoBusy Year Planned by the Paper Section of matic titrator because the reacNational Bureau of Standards tion is i n h e r e n t l y slower than With the assistance of funds from other government agencies most of those utilized by the and from outside organizations, the paper section of the National analyst. The rate of addition of Bureau of St,andards expects to expand its research activit,ies the titrating agent may therefore during the ensuing year. A new project recently initiated is a study of the possible effect be several times as fast as that on papers of fumigating gases used t.0 rid written and printed maordinarily used, and the titrator terials of destructive insects. This was undertaken a t the rewill stop the flow 1 to 2 ml. bequest of the National Archives and with their financial assistance. fore the true end point by virtue A new venture, which will take the section into a new field, is a study of the stability of prints contained on motion picture films. of a temporary a c c u m u l a t i o n This form of record material has assumed great importance, liof ferrocyanide. The pilot lamp brarians, educators, and others interested in the spreading of relighls a few seconds after the knowledge and in preserving it being deeply interested in the cut-off is made. By manipulatpossibilities of film records. With the assistance of a fund granted for the purpose to the ing the buret cock t h e r a t e of Sational Research Council by the Carnegie Foundation, it is flow of the titrating solution is planned to make a year's study of the resistance of the film rccreduced to less than 1 drop per ords to various degrees of temperature, humidity, and light. second, the clamp is reopened, and It is hoped that this work can later be expanded into a comprehensive study of the many other problems concerning miniature the titration completed. Much records, and of problems related to sound recordings. time is saved by this procedure. Direction of the part'icipation in the standardizing activities of A standard solution of zinc was t,he T . A. P . P. I. Paper Testing Committee will be continued. prepared by dissolving zinc metal FIGURE 5. ELECTRODE The committee has developed 37 standard testing methods and has 14 more under development at the present time. Additionai POTENTIAL CURVESIN d i s t r i b u t e d by the Bureau of testing work consists of an at>temptto develop a device for measNEARVICINITYOF END Stanc!ardq for use in thermomeuring in numerical terms the rat,e of failure of printed currency POINT try (!39.993 p e r c e n t Z n ) i n under conditions that produce appearance and loss of strength Two typical titrations of a hydrochloric acid. The solution similar to those caused by actual service wear. This work is zinc solution with otassium supported hy the Treasury Department. as p r e p a r e d contained 0.01084 ferrocyani&
