ANALYTICAL CHEMISTRY
534
.Is with xylene as entrainer the expulsion of the water from the sample contained in the test tube would be too violent if the boiling entrainer itself were in contact with the test tube, it cannot be placed directly in the boiler flask. Care must be taken that the boiler flask is never more than half filled with entrainer. Otherwise, the strongly boiling xylene might reach thc hottom of the test t,ube, with the same effect as direct immersion. The sample holder should never be filled more than half full. Samples which are liquids or liquefy below the boiling point of xylene (about 136" C.) should fill only one third of the test tube it' the water content is above 10%. The speed a t which the water is removed from the sample is the same as for the standard Dean and Stark proredure--e.g., for sodium sesquicarhonate: Xyithout adapter 40 minutes; a i t h adnpter 45 minutes: in oven at 110" C. 150 minutes. The speed a t which the a-atrr is removed depends largely on the nature of the sample: 30 minutes in a draft-frec position are usually sufficient,. Best results were observed under conditions u-hich obtain mostly with samples for v-hich the procedure was developed, w c h as checking intermediaries quickly in cases u-here the more accurate Karl Fischer determination is not justifietl. Samples 3hould contain more than 3 to 4yo water, hut the readiness with which the water is givm u p depends to a certain extent on whether the sample remains a solid, liquefies, or is a liquid t,hroughout. In the last tbyo cases, determinations ahove 2% can be carried out satisfactorily. Since only 3 ml. of xylene will he found on average to have condensed in a sample holder during dehydration of the sample, the problem of spent entrainer recovery does not arise. ' The loss incurred is negligible when compared with the saving in manipulation, as 1 liter of commcwial, wat,er-free xylene will treat over 300 successive samples. The kind of sample which can be treated is limited only by t'wo considerations, one of which applies also to the standard Dean and Stark-its boiling point must be above that of xylene (toluene was too slow under the conditions of the modification). Very finely divided samples with comparatively high water content (above 10% on average) are likely to he l)lown out of the test tube as soon as the water contained in thcni I)egins to boil-e.g., slaked lime. ACKNOWLEDGMENT
The writer desires to record his grateful acknowledgment to the managing director of Beckrrs Pty.. Limited, in whose Research Laboratories a t Dudley Park. South Australia, this work ha.: heen carried out, for permission to puhlish this paper. LITERATURE CITED (1) Danes, V., and Rund, B., Chem. Listy, 41, 133 (1947). ( 2 ) Dean, E. W., and Stark, D. D.. J . Ind. Eng. Chem., 12, 486
(1920).
( 8 ) Fetzer, W. R.,
A i v . 4 ~ .CHEM.,23, 1062 (1951). (4) J a m , P. E., and Zimrnerman, H. K.. Jr., Chemist-dnaZyst, 39,
72 (1950).
Modification of the Beckman Spectrophotometer with External C Battery Supply and Voltage-Checking Arrangement. Louis P. Cecchini and Maynard Eicher, Division of Chemistry and Instrument Laboratoiy, Naval lledical Research Institute, Sational Naval Medical Center, Bethesda. \Id. the C batteries of the Beckman spectrophotometer (Model DU) need checking or replacement, the necessity of inverting the instrument presents several difficulties. The size and weight of the spectrophotometer are such that unless extreme care is exercised during this manipulation, jarring of the instrument and damage may result. -4fter replacement of the dry batteries, it is frequently necessary to change the plate and 'or screen voltages to the amplifier tubes. Improper voltage. are detrcted HENEVER
N-
by observing the action of the galvanometer needle under operating conditions while the spectrophotometer is inverted. This procedure is inconvenient and awkward and can be avoided by locating the dry cells outside of the instrument housing. More serious objections arise when the Beckman spectrophotometer is employed with accessories for which the optical alignment of the parts is of paramount importance. In the commercially supplied accessories-e.g.. the flame photometer attachment-an attempt is made to ensure optimum alignment by providing detents for the feet of the spectrophotometer. Howevt.1, in the use of some auxiliary units which have been developed for specific purposes in our laboratories, it is imperative that the instiument not be mobed during a relatively long series of dependent measuiements (4, 6). Therefore, a method was devised to eliminate the necessitv of inverting the spectrophotometer. It is recognized that the desiccant in the monochromator housing cannot be reached without movement of the instrument; however, new desiccant will be effective for long periods in normal lahoratory environment, and indefinitely if the spectrophotonietei i i operated in a humidity-controlled room DESIGN
In order to relocate the C batteries in a more convenient and accessible popition, a hole 0.75 inch in diameter was cut in the side of the spectrophotometer case a t the lower left end of the battery compartment (as the operator views the instrument). One end of a shielded 12-conductor, rubber-covered cable was passed through this hole and connections made by lugs to the 12 terminals on the two strips in the s ectrophotometer With the 10-terminal strip numbered by tge manufacturer, the 2terminal strip was arbitrarilv numbered 11 and 12, with povitioii 12 adjacent to terminal 1. ST0R 4 G E B4TTERY CABLE
.I,
&TEW CABLE
Figure 1.
Beckman Spectrophotometer Cable Connection Diagram (2) '
-4few simple changes were made in the battery Compartment wiring which did not in any wav alter the original main circuit of the instrument. Figure 1 s h o w how the spectrophotometer cable connection diagram ( 2 ) wac: modified by an external relocation of the C batterie.. The lfodel D U spectrophotometer wiring diagram (3) lists the lead numbers. With referenre to that diagram, the wiring changes are as follows: 1 Jumper between terminals 11 and 12 is removed. 2. Juniper between terminals 7 and 12 is removed. 3. Lead KO, 30, which connects battery No. 1 to terminal 3 , is removed. 4. Lead S o . 29 (red), originally connected to battery No. 4, is changed to terminal 11. 5. Lead So. 34 (red), originallp connected to battery No. 6, is changed to terminal 12. The 12-conductor cable which extended about 18 inches from the spectrophotometer was terminated a t a 12-prong Jones plug. The corresponding chassis connector was mounted in a 6 X 6 X 6 inch metal cabinet in which the six C batteries were placed with suitable nonabsorbent insulation (Bakelite). (Batteries 5 and 6 should also be insulated from the remaining batteries in the box.) The chassis connector m-as wired to the terminals of a 12-position selector switch (Mallory S o . 321125). The C batteries were also connected to this switch via a binding post terminal block to facilitate battery changing. -4pair of binding posts was mounted on the cabinet panel, connected to ground and to the selector switch arm. This permitted voltage readings to be made selectively from the 12 terminals within the spectrophotometer. The procedure for checking theqe voltagep is recommended by the manufacturer, and
535
V O L U M E 25, NO. 3, M A R C H 1 9 5 3 EXTERNAL BINDING POSTS FOR VOLTMETER
9 9
d
CHASSIS
CONNECTOR
I
SELECTOR SWITCH
1 8 9 10
12 BLOCK
I BATTERY #I
Figure 2.
j
#2
Wiring Diagram of External Hatter> Supplj for Beckniaii Spectrophotometer
for periodic disassembly of the apparatus to iefill the steam generator with water. It has been used by the authors successfully for several years.
A water inlet, A , connected to a three-way stopcock, R , is attached to the steam generator as shown in the diagram. Stopcock B is connected to the water line and to the condenser inlet, C. A straight adapter, D, is attached below the outlet, E , of the steam trap and connected to a length of tubing leading to the sink. The condenser outlet, F , is connected to a small piece of glasq tubing, G, slightly flattened a t one end, which is attached to D by means of a wire. I n use, B 19 normally turned bo as to let water flow through the condenser jacket. At the end of each determination, the stopcock is turned briefly in the other direction. This causes cold water to flow into the steam generator, which results in the immediate creation of the partial vacuum necessary to suck hack the -pent sample. At the same time, the steam generator has I i t w i it4illed with water for the next determination. The flame nrwt iiot tw removed, a t i t l boiling is not interruptpd too long.
:orrect values for terminals 1 through 10 are listed i i i this rr,rvic,ing bulletin (1). Voltage values for terminals 11 anti 12 Fliould read $20 volts and “slightly positive,” respectively. T(~rininal 3 is ‘rground” and the battery cabinet case can be Connected tn it for shielding. Figure 2 is a wiring diagram of the battery calmest 1 s described. The external C battery supply has been used routinely in this ;ahoratory with success. -is a result, the time required for diagnosing and eliminating instrumental t,roubles arising from ( ’ 3at tery failure has been materiallv reduced by this niodification. ACKNOWLEDGMENT
The authors wish to express their appreciation to T. P. Sord,quid, Hospitalman, United States S a v y , for his technical assistance. The opinions evpresscd ale the authors’, and do not necessarily reflect those of the S a w 1)epaitment LITER.ATURE CITED
(1 I Beckmai: Instrumrnts, l n c . . South Pasadena, Calif., BILZI.l W D , 8, August 1950.
(2 Reckman Instruments, Inc., Xational Technical Laboratories, South Pasadena, Calif., Bd1. 150-A, 10, Drawing No. CD25C, August 1946. ( 3 ) I t i d . , p. 11. Drawing S o . WD 2500 D. ‘4) Cecchini, L. P., ANAL.CHEM., 25, 547 (1953). .~.5) Schiller. A . A , and Cecchini, L. P., Naval Medical Research Institute Project SM 001 008, Rept. 2 (July 20. 1948). I
Modification of Parnas-Wagner-Pregl Micro-Kjeldahl Apparatus. Edward H. Sheers, American Cyanamid Co., Linden, N. J., and Morton S. Cole, Sylvania Electric Products, Inc., Flushing,
T.Y. , As-Wagner-Pregl apparatus [Parnas, J. K., and Kagner, p r B i o c h e m . Ztg., 125, 253 (1931)l for the micro-Kjeldahl determination is undoubtedly the most widely used in the United Rtates. One of its drawbacks is the necessity for removing the flame a t the end of each distillation and waiting for the water in the stream generator to cool sufficiently to create the partial vacuum necessary to draw the spent sample into the steam trap for removal. I). 1,. Shepard and AI. 13. Jacobs have recently suggested the addition of a steam escape valve to the steam generator ‘ J . Am. Pharm. Assoc., 40, 154-5 (1951)]. This valve, which is closed during the distillation by means of a pinch clamp, is opened a t the end of the distillation and the pinch clamp is then placed between the steam generator and the steam trap. This modification considerably lessens the time and effort required to miry out a determination. The following modification cliiiiinate, completely the necessity
The addition of D allows the use of the spent condenser water to keep the outlet tube to the sink constantly clean. The apparatus as modified thus need rarely be disassembled in normal use. I n areas where the tap water contains titratable amounts of ammonia, it is suggested that the water inlet to the steam generator lie connected to a container of distilled water. Calculator for Selection of Sample Weights in Microdetermination of Carbon and Hydrogen. Ruth C. Thompson (present address, Kesleyan Universit,y, Middletown, Conn.), Laboratory of Advanced Research, Remington Rand, Inc., South Sorn.alk, Conn. niicroanal?-tical work the sample weight is Is r l r c t d :iccording to the percentage of the constituent to be determined ( 3 ) . If more than one constituent is t o be deterY QL-~STIT-~TIVE
mined, the srlcction becomes more complex. I n the microdetermination of cnrbon and hydrogen, for example, a sample weight which produces a suitahle amount of carbon dioxide for a carbon determination might produce so little mater that thr accuracy and precision of the hydrogen determination are impaired. 4 compromise saniplc size should be chosen which allows each element to be determined with the best accuracy possible. The precision and accuritcy of the hydrogen tktcrmination depend also on the maintt,nance of the lead dioside-watercarbon dioxide equilibrium (8). If sample sizes a w selected SO that approximately the same amount of water will be produced in succrssive analyses. this equilibrium will remain undisturbed (7). Some niicrocheniists recommend using saniples oi 4 to 5 mg. (61, 4 to 6 mg. ( d ) , or 10 to 15 mg. (j)thus , disregarding the aforenit,iitioiir.d con~idcrations, -4 v : i r r h of tht. 1iter:tture re-
