recting for the 10% interference of nitrogen oxides. In the daytime the stations generally have shown more oxidant than was observed previously without the absorbers. Interesting new information on oxidant pollutants is being obtained. ACKNOWLEDGMENT
The assistance of George A. Jutee, Kirk E. Foster, and Richard J. Lewis
with the field work and in providing reports of absorber operation in Continuous Air Monitoring Stations is acknowledged. LITERATURE CITED
(1) Adarns, D. F., J . Ai7 Pollution Control Assoe. 13, k8 (1963). (2) Bravo, H. A,, Lodge, J. P., Jr., ANAL. CHEM. 36, 671 (1964). (3) Chahk, J., Schafer, L. J., Yeager, D . W., J . Air Pollution Control Assoe. 5 , 227 (1956).
(4) Hersch, P., Deuringer, R., ANAL.
CHEM.35, 897 (1963). H.,J . Geophys. Res. 69,
( 5 ) Regener, ,.,,”,> V. ~~
37bn (,YO*,. (6) Saltmmn, B. E., Gilbert, N., Am. Ind. H y g . Assoc. J . 20, 379 (1959). (7) Salternan, B. E., Gilbert, N., ANAL. CHEM.31, 1914 (1959). (8) Wartbnrg, A. F., Brewer, A. W., Lodee. J. P.. Jr.. Intern. J . Air Water P d l & n 8, Zl~(l964).
Division of Water and Waste Chemistry, 145th Meeting, ACS, New York, N. Y . , September 13, 1963.
Determination of Oxygen in Uranium Dicarbide Beads Having Highly Impervious Pyrolytic Carbon Coatings M. E. Smith, J. M. Hansel, and G. R. Waterbury, Lor Alamor Scientific Laboratory, University of California, Lor Alamor, HANDLING DEVICE is defor crushing and loading materials into a crucible within an inert atmosphere of a furnace which is an integral part of an inert carrier gas apparatus for determining oxygen. The crushing operation is prerequisite to the satisfactory release of oxygen at 2600’ C. from uranium dicarhide beads having highly impervious pyrolytic carbon c o a t ings. Increasing the operating temperature of the inert carrier gas method to 2900° C. also releases oxygen from these beads, hut operation at the higher temperature is impractical because of excessively high apparatus blanks and the rapid failure of available crucible materials. Crushing must be done in an inert atmosphere because the exposed cores are extremely reactive and are readily contaminated with oxygen. Some duplex-coated beads and all triplex-coated beads investigated require crushing prior to the determination of oxygen. In the investigation of the effectiveness of the sample handling device, the inert carrier gas apparatus equipped with a current-concentrator furnace, as previously described (21, was used. Other types of furnaces could have heen used for this purpose if they had been modified to support the device.
Ascribed .
SAMPLE
EXPERIMENTAL
Apparatus. The sample handling device, shown inserted in the furnace (Figure 1) and in isometric view (Figure 2), is made of stainless steel, with the exception of the crushing surfaces (Figure 2) which are hardened tool-steel disks cemented to the stainless steel with epoxy resin. The bottom of the device is hollow and forms a funnel which may be rotated by a lever at the top t o provide either a crushing surface or an opening through which the crushed samplr may be poured into the open crucible within the furnace (Figure 2). Cruding is accomplished by screwing down the stainless steel plunger while 782
.
ANALYTICAL CHEMISTRY
Figure 1 . Furnace with sample handling device in position
the sample rests on the crushing surface of the disk. A hole is drilled in the side of the sample handling device to permit argon Ron into the plunger cavity while the sample is added by the longstemmed funnel made of hypodermic tubing. The device is lowered into the furnace against a Row of argon to prevent oxygen contamination. The device is supported in the furnace by an adjuse able brass collar fixed in position with a set screw (Figures 1 and 2). Procedure. The operation of the device is shown schematically in Figure 3. The shaded area helow the supporting collar represents the sight glans adapter a t the top of the furnace through which the device is inserted. In the flushing position, the crushing plunger is removed and the funnel rotated t o permit the argon from the furnace t o flow through the funnel stem. After a few seconds, the funnel is rotated to provide a crushing surface. The weighed sample is added through a long-stemmed funnel while argon continues t o flush the plunger cavity through a hole in the
N. M.
side. The long-stemmed funnel is removed, and the plunger is inserted and screwed down to crush the sample. The device is then lowered, and the funnel stem inserted into the crucible. The crushing pressure is released, and the funnel is again rotated to the position where the sample falls into the crucible, A small electrically operated engraving tool is held q a i n s t the side of the sample handling device to vibrate the funnel and assist in the transfer of material. The device is then removed from the furnace. During this operation, argon flow prevents contamination of the sample with oxygen. The furnace is capped and alloved to flush for 1 minute before the oxygen is determined according t o the previously described method (2). The amount of sample left in the sample handling device is determined by disassembling the device, rinsing any adhering particles with methyl chloroform onto a weighed filter paper contained in a suction funnel, and reweighing the dried paper. This amount, which is subtracted from the original sample weight, generally is less than 3 mg. and often less than 1 mg. RESULTS A N D DISCUSSION
The device was first applied to a sample of uncoated uranium dicarbide beads. Oxygen was determined in portions of the uncrushed beads using the original method (2) and the values were compared with those obtained for portions of the beads crushed with the sample handling device prior to analysis. Agreement among the results (Table I) indicates that no observable oxygen contamination was introduced by crushing the sample in the handling device. When this same material was crushed in an inert atmosphere enclosure and then exposed to air, the powder was ohserved to be pyrophoric. Application of the sample handling device to uranium dicarbide beads coated with a single layer of pyrolytic carbon did not increase the values obtained for the oxygen (Table I). This confirmed Drevious results with similar
FUNNEL POSITIONER-.,
CRUSHING PLUNGER\
c
3
SAYPLE DUYPER
CRUSHING PLUNGER
GRAPHITE CRUCIBLE
Figure 2.
Sample handling device (isometric view)
CRUCIBLE
I
FLUSHINO
samples in which analysis of portions crushed in an inert atmosphere enclosure were in good agreement with values obtained for uncrushed portions of the same material ( 2 ) . Results obtained for two samples of duplex-coated beads (Table I) show that more oxygen was released from the first sample when the beads were crushed using the sample handling device, but no appreciable difference in results was observed between the crushed and uncrushed beads for the second sample. The data also show some heterogeneity in oxygen content in these samples which were among the best available. Metallographic examination showed the two samples to be different in appearance, but both possessed a smooth inner coating and a rough outer coating. There was no physical evidence that
Figure 3.
explained the difference in behavior of the two samples during the oxygen determinations. The analyses of two samples of triplexcoated beads showed that more oxygen was released from the crushed portions of the samples than from the uncrushed portions (Table I). Apparently, the application of the sample handling device to beads with duplex coatings of pyrolytic carbon is beneficial for some bead samples and in no case is it detrimental. Results for all of the samples of triplex-coated beads analyzed to date have indicated that crushing the beads with the sample handling device is desirable.
Table 1. Oxygen Found in Uranium and Thorium-Uranium Carbide Beads 30. Rel. N o . Av. oxygen Est. std. Sam- of pJe poatSample of found,* std. dev., dev., h o . ingsa Core material treatment detns. p.p.m. P.p.m. % 14 0.11% 0.01% 10 U Carbide Cncrushed 1 0 14 0.107~ 0.017~ i o U Carbide Crushed 1 0 14 155 26 17 U Carbide Uncrushed 2 1 26 17 Crushed 14 155 U Carbide 2 1 6 65 30 46 V Carbide Uncrushed 3 2 5 225 26 11 U Carbide Crushed 3 2 5 275 30 11 U Carbide Uncrushed 4 2 5 255 32 13 U Carbide Crushed 4 2 14 45 13 29 C Carbide Uncrushed 5 3 Crushed 14 100 17 17 U Carbide 5 3 14 45 20 44 (Th, U) Carbide Uncrushed 6 3 10 11 Crushed 14 95 (Th, V) Carbide 6 3 a Coatings are of pyrolytic carbon. * For samples 3 and 4, values for oxygen found expressed as median values rather than average values, and standard deviation estimated according to a statistical method for small numbers of observations ( 1).
Table 11.
Oxygen Found in Duplex Coated Beads at Operating Temperatures of 2600" C. and 2000" C.
Sample treatment Uncrushed Crushed Uncrushed Crushed
Operatjng temp., C.
No. of
detns.
Av. oxygen found, p.p.m.
2600 2600 2000 2000
6 14 7 14
75 200 45 185
Re1. std. dev.,
9%
37 20 24 30
I
SAYPLE ADDITION
I
CRUSHIN6
LOADING
Operation of sample handling device
To test the efficiency of the crushing of the beads, a series of determinations was made to compare the values for oxygen obtained a t 2600" C. and 2000" C. for crushed and uncruehed portions of a sample of duplex-coated beads. If the crushing operation is effective in exposing the bead cores, which can be analyzed satisfactorily a t 2000" C., then there would be no significant difference between the oxygen values obtained a t the two temperatures. The results (Table 11) verify that the cores are effectively exposed hy the crushing operation. .Ilthough this sample apparently is somewhat heterogeneous with respect to oxygen, the difference between the average values determined a t each operating temperature is not statistically significant. Operation a t the lower temperature (2000" C.) offers the advantage of lower apparatus blanks (3 to 4 pg. of oxygen); however, RUCcessful operation a t the higher temperature is substantiated by many determinations and is recommended until more data at the lower temperature are obtained. Whether all the oxygen in the beads is completely released by this method is an important question that can be answered only by the analysis of appropriate standards. These standards are not available, and the probability of obtaining them is small. However, to circumvent this lack of standard material, oxygen concentrations were determined in highly impervious triplex-coated beads in the uncrushed state a t 2900" C. and the values compared to those obtained when portions of the same bead sample mere crushed with the sample handling device and analyzed a t 2600" C. The few results obtained indicated that essentially the same quantity of oxygen was released by both methods. The sample handling device has been applied to many bead samples with satisfactory results, and the application of this technique to the determinat,ion VOL. 37, NO, 6 , M A Y 1965
0
783
of oxygen in other highly refractory materials is being investigated. ACKNOWLEDGMENT
The authors thank C. F. LIetz, under whose supervision this work was per-
formed, for his valuable advice, and G. E. Nagy for his assistance in fabricating the sample handling device. The samples of triplex-coated beads were obtained from the Oak Ridge National Laboratory through the courtesy of W. R. Laing and J. L. Cook.
LITERATURE CITED
( 1 ) Dean, R. B., Dixon, W. J., ANAL. CHEM.23, 636 (1951). (2) Smith, E., Hansel, J. ll.,Johnson, R. B., Waterbury, G. R., Ibtd., 35, 1502 (1963). performed under the auspices of
woRB
the U. S. Atomic Energy Commission.
An Improved Apparatus for Biochemical Oxygen Demand James C. Young, William Garner, and John W. Clark, Civil Engineering Department, New Mexico State University, University Park, N. M.
HE
CLARKMETHOD for continuous
T determination of oxygen uptake of a
large inhomogenous biological sample was first developed in 1959 (1). The technique was applied to the automatic determination of biochemical oxygen demand (BOD) of polluted water. The principle was to maintain pressure in the closed sample-containing vessel by replenishing the metabolically utilized oxygen by electrolysis of dilute acid. The metabolically produced COZ was absorbed with KOH. The internal pressure of the vessel was monitored by a device which would sense a pressure decrease and actuated the power supply to the electrolysis cell. A constant current was supplied to the electrolysis cell and “on” time could be monitored both by an elapsed time meter and, for short term kinetics, a strip chart recorder. The technique has undergone continual revision (2, 3). Originally, a standard 2-liter resin kettle was used as the reaction vessel. The electrolysis cell and pressure switch were separate components. COz was absorbed by circulation of the internal atmosphere through absorption towers. Improvements of the apparatus are herein reported which include reduced cost, simplicity of operation, and obviation of a large head space which with varying external barometric pressure was a source of error. The apparatus is illustrated in Figure 1. The reaction vessel is a standard 1liter, narrow-mouth, reagent bottle with a flat head stopper (Corning No. 1500). The ground neck accepts a 3- 29 joint. A horizontal slot was cut through the neck about halfway down the ground section to produce an opening 1 em. in length. The COz absorbent container was fabricated from a 29/42 to 3- 24/40 bushing type reducing adapter (Corning S o . 8825). d test tube 20 mm. in diameter and 50 mm. in length was welded on the bottom of the bushing. A small glass propeller was welded onto the lower end of the test tube to increase turbulence in the liquid sample. Two opposing, 784
ANALYTICAL CHEMISTRY
SECTION A A
i-ii 29
Figure 1. Apparatus for biochemical oxygen demand
Slot
Stirring Magnrt
Onr- Lihr Rraqrnt Bottle
horizontal slots are cut into the upper end of the tube just below the ground section of the bushing to permit circulation of the make-up oxygen and the COn produced in the digestion. A vertical groove is cut part way through the outer surface of the bushing which extends along the length a distance which permits alignment with the slot in the bottle neck. The electrolysis cell was fabricated by Micro Chemical Specialties Co., Berkeley, Calif., from borosilicate glass. The drawing is to scale. The base 3-24/40 joint is indicative of dimensions, most of which were not critical for operation. It was important that the innermost tube was a t least 5-mm. i.d. so that it did not get plugged with condensate. The hole connecting the two electrolyte compartments had to be at least 10 mm. below the electrode levels. The electrodes were made of 0.005 inch Pt foil sweat welded to No. 20 Pt wire. Graded seals were not used a t the Pt-to-glass seal. It was sufficient to reinforce the seal mechanically with a drop of epoxy resin on the outer surface.
Adaptor- C O , Absorbent Contalnw
Elrclrolyvir COll
magnetic stirring bar coated with Teflon (Du Pont). A wick of glass wool was placed in the COS absorbent container and sufficient 30Oj, KOH added to cover two thirds of the wick. The electrolysis cell was filled with 0.7N HzS04to the level of the switching electrode. The apparatus was assembled with silicone stopcock grease coating the ground surfaces. In practice, four or five such assemblies are placed along with stirring motors in a standard BOD incubator. The vessel is vented until thermal equilibrium is reached, after which the bushing is rotated 180’ to seal. The power supply is described elsewhere ( 4 ) . LITERATURE CITED
( 1 ) Clark, J. W., New Mexico State
University, Engineering Expt. Sta. Bull. No. 11 (1959). (2) Clark, J. W., Water Sewage Works 107 (1960). ( 3 ) Clark, J. W., Zbid., 108 (1961).
(4) Young, James C., Clark, J. W.,
Garner, W., New Mexico State University, Engineering Expt. Sta. Tech. Rept. No. 20 (1964).
