P is for upper m e r c u r y contact Q i s for lower m e r c u r y contact
I
S i s connected to m e r c u r y
1wkn
Min. Neon Panel Lamp W i t h Side Opened Ciarex T m
-Relay Line - Live
Figure 2.
Wiring diagram of relay
adjusted pressure to the chromatographic system. As soon as the mercury in A drops below the lower electrode, the solenoids are de-energized, and the pump assumes operation at the normal mode. Spring-loaded check valves in the solvent lines ensure the correct direction of flow. Gravity-type check valves above the cylinders prevent mercury from getting out but pass solvents in both directions. We have found it necessary to introduce a n impedance to flow between the output lines and the solvent reservoirs, among others, to prevent these check valves from slamming shut during the change between the cycles. Microporous disks of Kel-F have been found satisfactory for this purpose. Figure 1 shows the locations of these disks. The electrodes in A are 0.047-inch
stainless steel rods which are mounted through a modified Conax packing gland assembly which is sealed a t the base with epoxy cement. To provide for the cycling between the two mercury levels in A , we have modified a commercially available "Solid State Relay," obtained from Scientific Kit Co., so the relay can be locked up. This alteration is required because the initial electrical contact, which starts the refilling of A , is broken as soon as the refilling operation is initiated. The wiring diagram is given in Figure 2. As the solvent in C should account for both the filling of A and the delivery to the chromatographic column during the time of filling, the displacement in this vessel should be larger than the amount displaced in A between the two electrodes. For the stated dimensions, flows u p to 100 ml. per minute have
been attained. For larger capacities, C and D should be larger than indicated. No major problems should be encountered during scaling up. The output pressure will vary with the heights of the mercury levels. To keep this variation negligible in relation to the output pressure, increase in size should preferably be found in increasing the reservoir diameter instead of the length. The limit is reached when the pressure rating of the vessel falls below the desired operating pressure. The 2-inch stainless steel nipples are rated at 3000 p.s.i. All the Swagelok connections used throughout the pump and the thick-walled l/r-inch stainless steel tubing have ratings far in excess of this number. The maximum operating pressure is determined by the accessory equipment-Le., the solenoids and the flow controller. Commercially available items are rated up to 250-500 p.s.i.; and to go to higher pressures, specially constructed equipment may be necessary. It is, however, not difficult to achieve operating pressures up to 1000 p.s.i. I n comparison, a piston pump with the equivalent rating would have to be much heavier in construction. Operation of the pump has been carried out a t pressures up to 1000 p s i and flow rates up to 100 ml. per minute with no discernible changes in output flow or output pressure during the different cycles of the pump. ACKNOWLEDGMENT
The authors are indebted to J. F. Johnson for helpful discussions and to W. C. Eaton for assembling and testing this pump during initial development stages.
The Purity, Properties, and Analytical Determination of 4-sec-Butyl-2-(ac-methylbenzyl)phenol 8. Z. Egan and W.
D. Arnold,
Oak Ridge National Laboratory, Oak Ridge, Tenn.
COMPOUND 4-sec-butyl-2- (aTzthy1benzyl)phenol (BAMBP) has been used as a reagent in the analytical determination of cesium and rubidium (18), in process applications for the extraction of cesium from reactor waste solutions ( 2 , 11, 16, 17) , and for the recovery of cesium and rubidium from ore leach liquors ( I ) . These applications require that the purity and properties of the reagent be known, and that a reliable method be available for the determination of its concentration in solution. This paper is concerned with the identification of impurities occurring in largescale (10 to 50 lb.) preparations of BAMBP, the estimation by distillation and gas-liquid chromatography of the concentrations of the impurities, the determination of some of the chemical and physical properties of BAMBP, the development of a spectrophotometric method for accurately measuring the
950
ANALYTICAL CHEMISTRY
concentration of BAMBP in various diluents, and a comparison of this method with an existing potentiometric method. DISTILLATION O F BAMBP A N D IDENTIFICATION O F IMPURITIES
An Aerograph Autoprep Model A-700 gas chromatograph was used to obtain chromatograms of 1M solutions of BAMBP in benzene or in diisopropylbenzene. The 1/4-inch X 20-foot copper column was packed with 60/80 mesh Chromosorb, treated with hexamethyldisilazane, and then with 20% silicone SE 30 rubber. Column temperature was programmed from 140" to 245' C. Helium flow rate was 65 ml./minute a t 50 p.s.i. The similarity of the chromatograms of different batches of BAMBP indicated that the same impurities occurred, but in different amounts, in the batches tested. A chromatogram of the most impure batch of BAMBP is
shown in Figure la. The different batches contained 2-401, of impurities more volatile than B.4MBP and 2 4 % of impurities less volatile than BAMBP. Samples of BAMBP were fractionated by distillation a t reduced pressure in a Podbielniak distillation apparatus, equipped with a 1- X 36-inch column, packed with Heli-Grid packing. I n two separate runs, a colorless B h h I B P product was collected a t 186" f 1" C. (6 mm. of Hg) and 207" i 2" C. (15 mm. of Hg). Distillation was discontinued during collection of the BAhlBP product, leaving unrecovered BARIBP and any higher boiling impurities in the still pot residue. Comparison of the chromatograms of commercial BAMBP and the distilled product shows that distillation removed most of the impurities (Figure 1). The purity of the distilled product was estimated to be greater than 98% BAMBP. During the synthesis of BARIBP,
it is possible to form isomers in which the 4-see-butyl group is replaced by n-butyl, iso-butyl, or tert-butyl groups. The impurity which gives Peak I1 in Figure l a is 4-tert-butyl-2-(a-methylbenzy1)phenol. This was established by comparing chromatograms of distilled BAhIUP to chromatograms of B=lhlBP containing added amounts of the 4-tert-butyl iaomer. The height of Peak I1 in distilled UAMBP changed in proportion to the amount of 4-tertbutyl-2-( a-methylbenzy1)phenol added. Chromatograms of mixtures of BAMBP and 4 - n - butyl - 2 - (a-methylbeney1)phenol showed no coincident peaks, indicating that this isomer was not one of the impurities. The GLC emergence time for the 4-n-butyl-2-(amethylbenzy1)phenol was greater than either of the peaks shown in Figure 1. Since pure 4-iso-butyl-2-(a-methylbeney1)phenol was not available, neither its presence nor absence has been established, but it is considered a likely possibility as a higher-boiling impurity (Peak IV). X small amount (less than 1% of the starting material) of 4-see-butyl phenol, not evident in the chromatograms, was separated by the distillation, and was identified by its infrared spectrum, melting point (47-8' C.), and equivalent weight (130). This compound is a starting material in the synthesis of BAMUP. PHYSICAL PROPERTIES OF DISTILLED BAMBP
Figure 1 . Gaschromatograms of BAMBP
W
z v)
a: Undlstilled b: Distilled
B v) (L
w
LT W LI: 0
8
LT W
+ INCREASING TEMPERATURE
for normal liquids, indicates association of the phenol in the liquid state. This is supported by the infrared spectrum of distilled BAMBP (Figure 2). Even though there is some steric hindrance in the BAMBP molecule, hydrogen bonding of the phenolic 0-H readily occurs (6), as evidenced by the broad 0-H absorption band around 3500 cm.-' Free 0-H absorption, which generally appears as a sharp band a t 3600 cm.? or higher, is observed only as a weak shoulder. DETERMINATION OF BAMBP CONCENTRATION IN SOLUTION
For process control and in analytical procedures, it is necessary in the solvent extraction steps to have a reliable method for measuring the extractant concentration in the diluent. Several general methods have been reported for the analytical determination of weak acids such as phenols. These include potentiometric titrations (9, 10, 15)' spectrophotometric measurements (8, 12, 14, and bromination procedures (19). = 1.004 gram/ml. (26' C.) Density Refractive index, n P J = 1.5570 Viscosity (relative to water) = 390 centipoises (25' C.) Boiling point = 186' C. (6 mm. of Hg) = 207" C. (15 mm. of Hg) = 325' C. (calculated for 760 mm. of Hg) = 20 kcal./mole/degree AH,,, AH,,, Trouton's constant = - = 31.7 cal./mole/degree T
BAMBP is completely miscible with many organic solvents, including carbon tetrachloride, chloroform, toluene, benzene, diisopropylbenzene, n-octane, kerosene, and hmsco 125-82 (a highly refined naphtha). It is essentially insoluble in dilute acid solution but slightly soluble in alkaline solutions (1, 3). Physical properties of BAMBP which were determined on the distilled product are :
The values for the normal boiling point, heat of vaporization, and Trouton's constant were calculated from the vapor pressure a t 186' and 207' C., using the integrated form of the Clausius-Clapeyron relationship log p = - AHV,,/2.303RT Constant and assuming AH,., to be constant a t higher temperatures and pressures. Because the vapor pressure was measured only a t two different temperatures, the calculated quantities are considered to be estimates. The high Trouton's constant value, compared to the value of 21
+
The potentiometric and spectrophotometric methods were investigated for possible application to the determination of BAMBP in several diluents. The spectrophotometric method proved to be simpler, faster, and more accurate than the potentiometric method, though limited to diluents which do not absorb strongly a t 278 and 284 mp. Phenolic impurities were measured as BAMBP with both methods. Spectrophotometric Determination.
The ultraviolet spectra of BAMBP in carbon tetrachloride and in n-octane are shown in Figure 3. Two absorp-
tion peaks occur a t 279.0 and 284.5 mp in carbon tetrachloride, and 277.5 and 284.5 mp in n-octane. The absorption spectrum of BAMBP in Amsco 125-82 was almost identical to that in n-octane with absorption peaks a t 278.0 and 284.5 mp. The spectra were obtained for 5.0 X 10-4Jf solutions, using a 10-mm. cell in a Cary Model XIVPM instrument. A Beckman DU spectrophotometer was used to ascertain that Beer's law holds in the concentration range 1 X to 6 X 10-4X BAPIIBP in the three solvents when absorption is measured at either 278.0 or 284.0 mp. The 278-mp peak is considered more reliable for measurements due to its broadness and higher intensity (1 3). Also, the 284-mp peak is not well resolved in carbon tetrachloride. Average molar absorptivities are given in Table I. The accuracy of the spectrophotometric method of determining BAMBP concentration is shown in Table 11. Solutions of BAhlBP were prepared by adding weighed amounts of BAMBP to Amsco 125-82. A sample of the solvent Amsco 125-82 was used for the comparison cell and for diluting the samples
Table 1.
Molar Absorptivities for BAMBP
Solvent Amsco 125-82
n-Octane CCla
A218
h a
2545 2550 2618
2270 2310 2474
Table II. Accuracy of Spectrophotometric Determination of BAMBP Concentration
Phenol present (M) 0.504 0.614 0.736 0.858 1.014
Phenol concn. found ( M ) 284mp Av.
278 mp 0.501 0.615 0.737 0.856 0.998
0.505 0.618 0.727 0.861 1.017
VOL. 38, NO. 7, JUNE 1966
0,503 n.617
0.7a2 0.858 1.008
951
4000 3000 2500
2000
I
I
/
(600
1400
I
I
I
FREQUENCY ( c d l 1200 (400 (000
I
I
1
900 850
/
1
1
800
750
1
I
,
700
WAVELENGTH (microns)
Figure 2.
Infrared spectrum of BAMBP
Pure liquid, 0.025-mm. cell
2000-fold before absorbance measurements. Absorbance was measured at 278 and 284 mp, and an average concentration was calculated. The spectrophotometric method presumably could be used for other phenols after establishing appropriate absorptivities. Those given are specific for BAMBP. This method could not be used for determination of BAMBP in diluents having high ultraviolet absorbance at 278 and 284 mp-e.g., kerosene. Potentiometric Titration. Weak acids, including phenols, have been titrated with tetrabutylammonium hydroxide in solvents such as pyridine ( 4 , 5 ) , tert-butyl alcohol ( 7 ) , and ethylenediamine. Several other titrant-solvent systems have also been used (4, 9, 15). Some of these systems were evaluated for titrating BAMBP solutions. Best results were obtained with tetrabutylammonium hydroxide titrant and either pyridine pr tert-butyl alcohol solvent (Table 111). Sodium hydroxide and sodium ethoxide were unsatisfactory titrants because of extremely small potential changes a t the end point. The potential change a t the BAMBP end point was less than 75 mv. under the best conditions. For comparison, the more acidic 4-
Table 111.
Base
ACKNOWLEDGMENT
We thank R. A. Zingaro, W. E. Oxendine, and W. B. Howerton for assistance in obtaining data for this paper
Potentiometric Titration of BAMBP
Titrating medium
NaOH NaOCZH6 NaOCzH6 NaOC2H6 Bu~NOH
7070 EtOH 70% EtOH 95% EtOH
BUNOH
ethylenediamine pyridine
BurNOH
chloro-2-benzylphenol gave a sharp end point, and the potential change was about 150 mv. I n a typical titration, 0.250 ml. of a 1M BAMBP solution was pipetted into 150 ml. of tertiary-butyl alcohol and stirred with a magnetic stirrer. Standard 0.078-l4 tetrabutylaninionium hydroxide (Eastman) solution in 80% benzene-20yo isopropanol was added slowly, and the potential was measured with a p H meter, using calomel and glass electrodes. The titration end point was obtained from a plot of potential us. volume of titrant. When titrated by this procedure, weighed 1. O O M solutions of BAMBP and 4-chloro-2-benzyl-phenol in Amsco 125-82 and in diisopropylbenzene diluents gave values ranging from 1.01M to 1.06M. The potentiometric titration has the advantage of being a more general method of determination, applicable to different phenols in various diluents. However, this method is more timeconsuming and is not as accurate as the spectrophotometric method.
ethylenediamine tert-butyl alcohol
BAMBP diluent DIPB“ DIPB Cc4 CClr DIPB, ccl4, Amsco 125-82d DIPB DIPB, CCla, ~
b c d
DIPB: diisopropylbenzene. Insufficient potential change a t end point. Unstable potential. Amsco 125-82: a highly refined naphtha.
952
ANALYTICAL CHEMISTRY
S
C
125-82 O
Remarks Poorb Poorb Poorb Erratice Good Fair Good
250
260
270
280
290
300
340
WAVELENGTH (mp)
Figure
3.
Ultraviolet
spectrum
of
BAMBP 0.0005M, 1 0-mm. cell
LITERATURE CITED
( 1 ) Arnold, W. D., Crouse, D. J., Brown, K. B., Ind. Eng. Chem., Process Design Develop. 4 , 249 (1965). ( 2 ) Bray, L. A., Nucl. Sci. Eng. 20, 359 (1964).. \ _ _ . _
U. S. At. Energy Comm. Rept. ORNL-TM-449(1963). (4) Bruss, D. B., Harlow, G. A., ANAL. CHEM.30, 1836 (1958). (5) Cundiff, R. H., Markunas, R. C., Ibid., 28, 792 (1956). ( 6 ) Egan, B. Z., Zingaro, R. A,, Benjamin, B. Rl., Inorg. Chem. 4 , 1055 (1965). (7) Fritz, J. S., Marple, L. W., ANAL. CHEM.34, 921 (1962). ( 8 ) Harlow, G. A., Bruss, D. B., Ibid., (3) Brown, K. B.,
30, 1833 (1958). (9) Harlow, G. A., Noble, C. M., Wyld, G. E. A., Ibid., 28,784 (1956). (10) Harlow, G. A., Noble, C. XI., Wyld, G. E. A., Ibid., 28,787 (1956). (11) Horner, D. E., et al., Nucl. Sci. Eng. 17, 234 (1963). (12) Hummelstedt, L. E. I., Hume, D. N., ANAL.CHEM.32, 1792 (1960). (13) Lothian, G. F., Anal@ 88, 678 (1963). (14) McKinney, R. W., Reynolds, C. A,, Talanta 1, 792 (1956). (15) Moss, M. L., Elliott, J. H., Hall, R. T., ANAL.CHEM.2 0 , 7 8 4 (1948). (16) Richardson, G. L., Hanford Laboratories Rept. “-80686 (1963). (17) Ibid., HW-SA-3211 (1963). (18) Ross, W. J., White, J. C., ANAL. CHEM.36, 1998 (1964). (19) Stope, K. G., “Determination of Organic Compounds,” p. 121, McGrawHill, New York, 1956.
RESEARCH sponsored by the U. S. Atomic Energy Commission under contract with the Union Carbide Corp.
