amounted to only disintegrations per second. It has been possihle by the above technique t o determine independent fission yields of tin fssion products in thermal neutron fission of uranium-235, tin half-lives, generic relationships, and also prominent gamma photopeak energies from pulse-height distributions taken of the rapidly separated tin fractions. The results in Table I have been obtained from many different tin qeparations made at times after fission varying from 10 secoiids to 30 minutes. 1ndependent evperiniental values of these tin half lives were obtained through continuous wolution of SnH4 from a basic KaBII4scllution of uranium235 during neutron irradiation and collection of the tin descendents on a charged mire. Tentative values by this method are 44 seconds for tin-131, 50 seconds for tin-132, 43 seconds
Table I.
Half Lives and Fission Yields of Tin Isotopes
Half life, Fission Suclide seconds yield, yo Snl31a 65 zt 10 0.7 Sn132 50 f 10 0.5 SnlS3 39 d= 15 0.2 The 1.6-hour isomer is not observed. Its fission yield would be less than 0.037,.
for tin-133, and 20 sec,onds for tin-134. LITERATURE CITED
(1) Gaylord, N. G., J . Chem. Educ. 34, 367 (1957). ( 2 ) Gaylord, ?;. G., “Reduction With Complex RIetal Hydrides,” pp. 1013, 1022. Interscience. Kew York. 1956. ( 3 ) Glendenin, L. E., RIetcalf; R. P., National Nuclear Energy Series, Vol. 9, Book 3, p. 1625. (4) Greendale, A. E., Love, D I,., ANAL. CHEU.35, 632 (1963).
(5)-Greendale, A. E., Love, D. L., U. S. haval Radiological Defense Laboratory Technical Rept., USNRDL-TR-601, Nov. 13. 1962. ( 6 ) Jolly,-’W. L., J. Am. Chem. SOC.83, 335 (1961).
( 7 ) Latimer, W. AI., Hildebrand, J. H., “Reference Book of Inorganic Chemistry,” p. 259, Macmillan, New York,
1936. _ ._. (8) Luke, C. L., A N ~ LCHEM. . 28, 1276 (1956). (9) Luke, C. L., Ibid., 31, 1803 (1959). (10) Mellor, J. W., “A Comprehensive Treatise on Inorganic and Theoretical Chemistry,” Vol. VII, pp. 324-5; Vol. X I , pp. 394-400, Longmans, Green and Co., London, 1947.
(11) Nervik, W. E., “The Radiochemistry of Tin,” Sational Academy of Science, Xuclear Science Series, Monograph NAS-NS 3023, 1960. (12) Schaeffer, G. W., Emilius, M., J . A m . Chem. SOC.7 6 , 1203 (1954). (13) Stetser, J. B., Norton, R. H., Iron Age 102,443-5 (1918).
RECEIVEDfor review Rfay 20, 1963. Accepted August 5, 1963.
Auto matic, Am pero metric, Cupfe r ron Titration of Zirconiium in Highly Radioactive Solutions HISASHI KUBOTA arid
J. G. SURAK’
Analytical Chemistry Division, Oak Ridge National laborafory, Oak Ridge, Tenn.
b The amperometric titration of zirconium with cupferron reported b y Olson and Elving hcis been modified and adapted for hot cell analysis. A titration has been devised in which the current is monitored continuously while the titrant i s fed in a t a constant rate. A stirred mercury pool i s the indicating electrode which senses the point where cupferron comes inyo excess, and a platinum wire i s used for the reference. Precision of 1 relative standard deviation i s obtainable for benchtop analysis and 1 .5% for titrations in a hot cell. This titration was specifically devised for the analysis of zirconium in molten salt reactor fuel, and the titration can be performed directly following dissolution without the need for separations.
yo
T
of the Xolten Salt Reactor ExperLnent (commonly referred to as PVISRE: program brought out the need for a method to determine zirconium in highly radioactive solutions. The activity oi’ the sample material makes it mandatory to perform all the operations in a hot cell, thereby imposing certain limitalions on the procedure to be wed. Zirconium analysis is routinely carried out by x-ray fluorescence. The relative e:ase of this analysis coupled with the good results obtainable have given little incmtive to develop chemical methods for zirconium. Since HE INTIATION
the installation of an x-ray fluorescence unit into a hot cell involves serious problems related to cost, operation, and maintenance, attention was directed to the development of a chemical method of analysis. Titrimetric procedures are preferred to gravimetric or spectrophotometric procedures in hot cell work from the standpoints of the ease of installation, manipulation, and replacement. Furthermore. procedures utilizing electrometric end point devices are best because the components within the hot cell require a minimum of space and moving parts, and the bulk of the instrumentation can be located outside. Therefore, it was natural that the search for a suitable method was directed toward a \-olumetric method utilizing an electrometric end point device. Most of the available volumetric procedures for zirconium depend upon the complexation reaction with EDTA. The direct titration with E D T A a t a pH of 1.5 has been described by Fritz and Fulda (3). At a pH of 1.5, the hydrolysis of zirconium becomes a serious problem and a back-titration of excess E D T A (4, 5) is generally more reliable. Olson and Elving (8) reported the amperometric titration of zirconium using cupferron for titrant and a dropping mercury electrode for the indicator electrode. The amperometric titration has many
attractive features which can be utilized to good advantage for this analysis. The titration is performed in a 2iM sulfuric acid medium in which only a few metals beside zirconium can react with cupferron. The high acid medium also minimizes errors due to the partial hydrolysis of zirconium. Best of all, this titration is feasible in the presence of moderate amounts of fluoride, and even large excesses of fluoride can be tolerated by the simple expedient of adding some aluminum ion. This is important t o this particular analysis because the starting material is a mixture of fluoride salts, and the repeated fuming necessary to remove the last traces of fluoride can be avoided. On the other hand, the entire process is tedious and lengthy and is not suited for adaption to hot cell use. -4small increment of titrant has t o be added, the solution deaerated, and the current determination made. This series has to be repeated several times for a single titration. Finally, the current us. titrant volume graph has t o be constructed for the location of the end point. It was thought that if the principle of this titration could be adapted to one in which the titrant is fed in at a constant rate and the current monitored Present address, Department of Chemistry, Marquette University, Milwaukee 3, Wis. VOL. 35, NO. 1 1 , OCTOBER 1963
1715
TO VACUUM
continuously, a procedure well suited for hot cell use would result. The re sults of this investigation showed that by the proper choice of electrodes and working conditions such a titration was possible and could he e a d y performed in a hot cell with good precision. The electrode system can be utilized for many other titrations in which the titrant contains a substance which is reducible a t the dropping mercury electrode. EXPERIMENTAL
Reagents. Cupfrwon solution, 0.0511. Dirsolvc 775 log. of eupferron (Fastman U'hite L ~ h l in j water that has been deoxygenated. with a stream of inert gas for 10 minutes. Dilute to 100 ml. Zirconium standard solution, 5 mg. Zr per ml. Accurately weigh about half a gram of high purity zirconium metal (99.9%) into a platinum dish. Cover the metal with I0 ml. of nitric acid and place the dish on a hot plate. Add a few drops of hydrofluoric acid to the warm solution to start dissolution. Add hydrofluoric acid dropwise whenever the dissolution reaction slows down. When dissolution is complete, carefully add 10 ml. of sulfuric acid and heat till brown fumes are completely gone. Transfer the sulfuric acid solution to a volumetric flask. Use 1N
1716
ANALYT~CAL CHEMISTR~
Equipment. Constant delivery buret assembly. A portable infusionwithdrawal pump (Catalog No. 1100, Harvard Apparatus Co., Inc., Dover, Mass.) equipped with a 2-r.p.m. motor and 10-ml. syringe is used to deliver the titrant a t a rate of 0.34 ml. per minute. The delivery tube from the s a polyethylene tube of syringe i inch i.d. and '/,-inch 0.d. drawn to a tip a t the terminal to the solution. A '/rinch hole is drilled into the side of the syringe barrel near the far withdrawal end (Figure 1). This portion of the barrel is covered with a machined plastic piece that fits snugly over the barrel. A hole is drilled into this plastic in line with the hole in the barrel to receive the outlet tubing for a vacuum connection. For filling the buret the plunger is backed to the rearmost position, the delivery tip is immersed in the titrant solution, and the vacuum connection is made until the barrel of the syringe is filled up to the vacuum line level. The vacuum is turned off and the plunger moved forward till the vacuum line is sealed off. Titration cell. The titration cell assembly is shown in Figure 2. The cell itself is a 50-ml. lipless beaker. A plastic cap (Teflon) is machined to fit over the beaker and is provided with openings for the buret tip, stirrer, reference electrode, lead to the mercury pool, and inert gas inlet. Since there
Polarograph. A recording polarograph with a 10- or 20-pa. full scale sensitivity range is used. Procedure. Place 10 ml. of mercury and 10 ml. of I N sulfuric acid in the titration cell. Transfer 1 ml. of standard Zr solution (5 mg. per ml.) by means of a micropipet followed by one drop of 0.1% gelatin solution. Position the titration cell under the cap and deoxygenate t h e solution for 10 minutes. Start t h e stirrer. Apply -0.5 volt to t h e mercury pool vs. the platinum electrode and start the current recorder and the infusion pump. Stop the titration when an abrupt and nearly vertical change in current is indicated. Locate the end point by extrapolation and, subsequently, determine the titer by comparing the volume of titrant delivered us. the amount of zirconium taken. In a similar manner, titrate an unknown sample. Use the titer obtained from the standardization to calculate the zirconium content of sample. RESULTS
A typical titration curve is shown in Figure 4B. Precision studies conducted over a period of 10 working days showed that a
Li, Be, Z r , U C A T I O N S IN MSRE SALT
1 CJPFERROU
Figure 4.
Effect
I
Figure 5.
Elements precipitated by cupferron in
1-4N
acid
C5M)
Table 1.
of potential on shape of titration curve
relative standard deviation of 1% or better is attainable for benchtop analysis. With the assemtly in place in the hot cell, the precision is 1.5%. The precision is the same for titrating zirconium either as the pure solution or as a component in a simulated molten salt reactor material. DISCUSEJON
The titration asseinbly was devised with the problems of the ease of installation, manipulation, and replacement as foremost considerations. The setup that is described is comprised of equipment that is easily available or material simple to fabricate and occupies a very limited area within a hot cell. The manual operations that have to be performed with the master slave manipulators have been reduced to transferring the requisite reagents to the titration cell, positioning the titration cell under the Teflon cap, and moving the buret delivery tip to the proper site for either the 6lliig operation or the titration. All other controls are performed from outside the hot cell. The equipment is of a simple nature such that in the event of a major malfunctioning, a new setup can be installed rapidly and economically. The critical problem in this adaptation was the search for a suitable electrode system. The dropping mercury electrode is not practical for this work because the solution has to be stirred throughout the titration. The rotating platinum electrode and a coated mercury electrode were also found to be unsatisfactory. The stirred mercury pool that is used in many coulometric procedures was found t o be admirably suited for this p u r p o s ~ . A calomel electrode making contact with the solution via a sulfate bridge% is an excellent reference electrode. I n view of some difficulties experienced a t this laboratory on the erratic behavior of calomel electrodes under the effects of radiation,
other electrodes were tried, and a simple platinum wire electrode was found to be suitable. It is thought that this electrode pair (with either the calomel or platinum) can be used to detect the end point of many titrations in which the titrant itself is reducible at a dropping mercury electrode, provided the titration reaction is feasible at the operating potential. The end point mechanism is dependent upon the electroactive nature of cupferron. The polarography of cupferron was studied in detail by Kolthoff and Liberti (6) and Elving and Olson ( 2 ) . The reduction is thought to be a stepwise affair involving six electrons giling a final reduction product of phenylhydrazine. The compound was slightly decomposed by acid solutions below pI1 3; however, the degree of this decomposition is relatively small such that cupferron can be used for amperometric titrations in acid media where changes in current flow rather than the absolute magnitude of the current are important. The half-wave potential for this reduction in an acid medium varies according to the following relation. Ell2
=
-0.58 volt
(US. S.C.E.) -0.128
pH.
Olson and Elving operated a t -0.84 volt, a potential well on the plateau of the diffusion wave in 211f sulfuric acid, assuming that the above relation can be extended to such a strong acid medium. A 114' sulfuric acid medium was the medium used in this work, and it was best to perform the titration a t -0.55 volt (us. S.C.E.) or -0.50 volt (us. platinum), Cupferron decomposes slowly even in the dry state, and the rate of deterioration is even faster in aqueous solution. The cupferron solution was prepared fresh daily by disolving a weighed portion in deoxygenated water and used for that day only. The change in titer during the course of a day is not dis-
Compound
Composition of MSRE Salt
Mol %
Element
Weight
%
tinguishable; however, a slight discoloration is observed after 24 hours. It was not necessary to continue purging the cupferron solution once it was inside the syringe. Maxima effects are observed in the polarography of cupferron which can be removed very readily by the addition of gelatin. The addition of gelatin to this titration medium is necessary to ensure good titration curves. The two titration curves shown in Figure 4 show the effect of operating potential upon the shape of the titration curve. At a potential of -0.90 volt, a potential on the diffusion wave plateau, considerable current begins to flow from a little after the start of the titration (Curve A ) , and it is difficult to construct the proper line corresponding to the pre-end-point stage for the extrapolation of the end point. At -0.50 volt, a potential near the foot of the diffusion wave, the pre-end-point portion is nearly a straight line as shown by curve B. I n either case, the current rise after the end point is very sharp. The elements precipitated by cupferron in 1 to 4J acid are shown in Figure 5 (?). Oxalate, citrate, tartrate, phosphate (I), and EDTA are some of the commonly used complexers which can be present and not interfere. Moderate amounts of fluoride (up to 10 times zirconium) also will not interfere. Larger amounts of fluoride can be complexed with aluminum which is unreactive toward cupferron. I n this acid medium nitrate is a very bad interference because it reacts with cupferron, while chloride reacts with mercury. The composition of MSRE salt is VOL 35, NO. 11, OCTOBER 1963
1717
shown in Table I. The salt is dissolved in a nitric-sulfuric acid medium with the addition of a little boric acid and the nitric acid is later fumed away. Since the only possible cationic interference of U(IV) has been converted to V(VI) during the dissolution and no anionic interferences are present, the titration for zirconium can be made directly after the removal of the nitrate. The titration of 5 mg. of zirconium takes about 5 minutes using the equipment and reagents that have been described. The entire procedure including the makeup of the solution, deoxygenating, and the titration takes about 20 minutes. The 1.5% precision for de-
termining zirconium a t this level is within the limitations imposed upon this analysis. If necessary, the precision can be improved by taking larger samples or, to a limited extent, using more dilute titrant. The reproducibility in taking samples seems t o be the chief cause for the lower precision of the analysis in the hot cell.
(4) Fri+bz, J. S., Johnson, Marlene, Ibid., 27, 1653 (1955). ( 5 ) Go ldstein, Gerald, hIanning, D. L., Zittel, H. E., Ibid., 35, 17 (1963). ( f j) Koltoff. I. 11..Liberti. A,.’ J . Am. Chem. Soc. 70, 1885 (1948). (7) Lundell, G. E. F., Hoffman, J. I., “Outlines of Methods of Chemical Analysis,” p. 117, Wiley, Kew York, 14.18
(8)”6i$’on, E. C., Elving, P. J., ANAL. CHEW26, 1747 (1954).
LITERATURE CITED
. .
C H E M . ’ 1206 ~ ~ , (1954):
RECEIVEDfor review May 20, 1963. Accepted July 12, 1963. Division of Analytical Chemistry, 144th Meeting, ACS, Los Angeles, Calif., April 1963. Oak Ridge Xational Laboratory is operated by Union Carbide Corp. for the U. S. Atomic Energy Commission.
Hydrogen Cell Assembly for Standard Electromotive Force Measurements PETER G. SIBBALD’ and GEORGE MATSUYAMA Scientific and Process Instruments Division, Beckman Instruments, Inc., Fullerton, Calif.
b A simple hydrogen cell assembly for the calibration of buffer solutions, glass electrodes, and reference electrodes is described. The design of the cell permits such calibrations to b e made in 20 to 30 minutes with a maximum standard deviation of *0.6 mv. (kO.01 pH).
M
p H instrumentation now permits refinement of the readout to 0.001 pH. Such a high degree of precision, for example, is desired in the measurement of the p H of blood and other body fluids where small changes of p H may have significance. The importance of high precision p H measurements has further been recognized by the assignment of revised standard p H values, given to three decimal places, for the seven rational Bureau of Standards (NBS) reference solutions which define the p H scale ( 5 ) . For very accurate p H measurements, frequent calibrations of a p H cell containing a glass electrode are necessary. This requires a reliable reference buffer, the p H of which has been determined previously with great accuracy. Second, the p H response of the glass electrode must be checked, preferably in the p H range in which it is to be used. It is known that even the best modern electrodes show small variations in response t o p H with time. Third, the ODERN
ELECTRONIC
1 Present address, Research and Development Division, Richfield Oil Corp., Anaheim, Calif.
1718
e
ANALYTICAL CHEMISTRY
stability of the reference electrode with time must be verified. This paper describes a simple hydrogen cell assembly in which standardization of buffers, glass electrodes, and reference electrodes can be made accurately and with relative ease. The hydrogen cell assembly is similar in principle to that described by Bates, Pinching, and Smith (6), modified to obtain more rapid thermal and chemical equilibrium and designed to fit the Beckman Model 28505 Thermomatic constant temperature block. The standardization of a buffer solution against a NBS primary standard is accomplished by measuring the e.m.f of cells with liquid junctions of the following type : Pt; HZ(gas), Primary Standard/Saturated KCl/Buffer X, Hz (gas); Pt The design of the cell assembly permits the formation of physically well defined liquid junctions. The residual liquid junction potential-Le., the difference of the two liquid junction potentials formed between the saturated potassium chloride solution and the primary standard buffer on one side and the test solution on the other-manifests itself as an indeterminate error in the derived pI-1. I n the ideal case, where the liquid junction potentials are numerically identical and opposite in sign, the residual liquid junction potential equals zero. It is, therefore, of great importance to be able to produce and reproduce closely identical liquid junctions at both sides of the salt bridge. h minimum of two buffer standards is needed for the calibration of a glass elec-
trode. Its p H response can be determined by measuring the e.m.f. of cells without liquid junctions of the type Pt, 1 x 2 (gas), Buffer Soln./Glass Electrode
Similarly, the stability of a reference electrode can be verified by measuring the e.m.f. of the cells: PtJ Hz (gas),
Buffer Soln./Reference Electrode EXPERIMENTAL
Apparatus. A sketch of the hydrogen cell assembly is given in Figure 1. It consists of two electrode compartments, a and a’, connected by a saturated KCl bridge, b. Each electrode compartment holds approximately 3 ml. of solution. The liquid junctions are formed at c and e’ and protected from turbulence caused by the bubbling hydrogen gas, by stainless steel balls, d and d’, that imperfectly separate the standard and test solutions from the KC1 solution. The bottom of each electrode compartment is ground a t c and e’ to match the contour of the stainless steel balls. Compartment e holds a stopper, f, that prevents displacement and mixing of the saturated KC1 solution in the capillary bridge with the solutions of the two electrode compartments. Stopper j is ground in place at g. The presence of saturated KC1 in this area maintains electrical continuity. Purified hydrogen gas enters the system through h and h’; i and i’ are platinum electrodes, and j and j ’ are outlets for hydrogen gas. Copper mire leads covered with phone tips, k and k’, are soldered t o the platinum electrodes for electrical connections.
