~~~n~~ T. Marvin and Larance M, bansford, U. S. Fish and Wildlife Service, Galveston, 5bx. of high purity is required for investigating the chemistry of the nutrient3 and trace elements in sea water. It is needed in large quantity where much experimental culture work using artificial sea water iS in progress. To meet this demand, the Gulf Fishery Investigations laboratory has constructed a pair of all-glass stills. Operating in tandem, these will produce in excess of 50 gallons a day of double-distilled water having a specific conductivity of 0.7 X 10-8mho. This water is used in preparing culture and nutrient solutions, running chemical analyses, and, equally important, in cleaning laboratory glassware. A diagram of the two stills is shown in Figure 1. Borosilicate glass construction is used throughout. The boilers, P, consist of 72,0(#)-ml. flasks with 102/75socket neck joints, C. Adapters, D, connect the 7-foot columns, G, to the boilers. The adapters aiid cdunins are made of 3-inch glnss pipe, held
together with glass pipe connectin clamps, E. The columns are acke! with borosiIicate b s c h i g rings, geld in place with retaining plates, F, constructed of block tin or high temperature polyethylene (I/, inch thick). These are profusely perforated with 1/4-inch holes. The top of the columns connect to the heat exchangers, K , by eonnecting arms, I. These are made of 1-inch glass pipe with 35/25 socket joints on each end as shown (Hand J) and are wrapped with 700-watt Briskeat heating tapes. The heat exchangers are high speed condensers similar to H. S. Martin's Catalog No. M-7032 but are twice as long and have 35/25 socket joints as shown. Each still contains a siphon tube, B, made of block tin or high temperature polyethylene tubing, which enters the system through one of the six entrance holes in adapter 1). Electrical eonductors T, conductor braces U , nnd water inlet 0 also pass through this adopter. The first still contains two 2 "4 x 6 inch carbon electrodes, A , and the second, a heating coil, I?, made by winding 17
ISTILLED WATER
COOLINP WATER INLET-
-7'1
feet of 0.051-gage Alloy K (C. 0. JelM Mfg. Corp., Southport, Conn.) on a I/&ch mandrel, The condensate from each condenser is conducted to storage tanks, M (53.5gallon polyethylene drums, obtained from A. Daigger and Co., Chicago IO, IL), by means of polyethylene pipes, L, each of which contains a steam vent, S. The operation of the first still depends upon direct conduction of electricity through the water in the boiling pot. Initially, the pot water must be fortified with enough sodium chloride to permit passage of a desired amount of current. Enough salt solution is added through B, so that a t a water temperature equal to room temperature, or perhaps slightly higher, the conductivity will be increased sufficiently to permit passage of 25 to 30 amperes under n potential pressure of 220 volts. As the temperature increases to boiling, the conductivity also increases. The final current can be controlled by adjusting the level of the siphon overflow, N . The water
CDOLINO W A l C R INLET-
K
-COOLIN@ WATER INLET
L
'0DRAIN
P
Figure 1.
w
Diagram of two rHDlr
input through 0 should be slightly greater than the production oi dntilled water. This wil1 prevent excessive accumulation of solids in the boiling pot and permit the siphon regulator to maintain suitable equilibrium conditiona for long periods of time without attention. Electrolytic action of the raw water in the first still necessitates the use of carbon electrodes. These simplify the maintenance of a constant water level. As the level tends to increase, the resulting increast in current consumption builds up the internal steam pressure in the system. This, in turn, increases the flow of water through the siphon until equilibrium conditions are attained. Similarly, the reverse of these conditions tends to prevent the level from dropping. As rccommended by Ballentine [ANAL.CHEM,26, 549 (1954)], the two. connecting arms, I , are wrapped with flexible heating tapes to prevent forniation of wakr films that might carry impurities into the condensers. The rate of flow of cooling water
of Films from
through the condensers is regulated 80 thst the condensak emergee with a fair amount of vapor, which escapss through 8, carrying with it much of the volatile impurities. The outlet water from the condenser is tspped just before entering the drain to provide a warm water feed source for the boiling pot. The feed water, on entering the boiling pot, must be allowed to run down the siphon tube, This prevents vapor Iocking within the tube, which would prevent siphon regulation. The storage tank, M , supplies water to the second still. This is identical to the first, with the exception of the heatiag element, R. This consists of a coil of resiststnce wire which limits current consumption to about 40 amperes at 220 volts. As current through this coil is in no way dependent upon the liquid level in the boiling pot, the regulation of the siphon is not so precise as in the first still. However, if the flow rate through 0 is adjusted so that it is about 1 m 2% greater than that of the output, a fairly constant level can be maintained, The pot is cleaned whenever
~
the distillate conductivity rises above about 0.8 X 104 mho. Every 24 hours potassium permanganate is added to oxidize any organic impurities, No specis1 provision has beenpade for the removal of carbon dioxide. The .&e of production of the first still a t a current consumption of 70 amperes at 220 volts is about 5 gallons per hour of distillate having a conductivity of a p p r o x h t e l y 1.3 X lo4 mho, and that of the second, at a current consumption of 40 amperes at the same voltage, is just under 2.5 gallons per hour of distillate having a conductivity of about 0.7 X 10- mho. The excess water from the first still is used wherever high purity distillate is not required. If an attempt is made t o increase current consumption in the first still, the column may flood and seriously impair the purity of the product. Operating at the suggested amperages, the product of the second still compares favorably with that of smaller all-glass laboratory stills having outputs of 1 to 2 liters per hour and no special provision for elimination of carbon dioxide.
~ f6r Infrared ~ ~ Analysis l $
M. A. Szyrnanski and R. 0. Conley, The Spectroscopy Laboratories, Canisius College, Buffalo, N. Y. it i w s neccwrry to obtain the infrared spectra of a variety of films conted on metals. KOsntisfactory method could be found in the literature The potas4um b omidc technique may be used if the film c:ui be scraped from the metal and finely poutlcred, but for films that cmnot bc powkred, other methods of snmple prclnration are required, Wiile engaged in a $ t i d y of the curing and nginy of varnislics and enamels on copper, the authors ntlopted a method of lifting the film electralytically froni the metal. By this method, film suitable for spectroscopic mensurenients can be obhined. Comparison with films cast on salt pl:ites, p o t w i m bromide disks, and tmnsmiwion curves obtnined using a reflectance nttnchmeiit indicirted that the electrolytic processes did not affect. the spcctml properties of the film. The only limiting feature of the rncthod was found with filnis nbout 0 . W S inch thick. In these cnaes fringc patterns were obtained in the wave length regions corresponding to the film thickncss, c:rusing Borne difficulty in obtaining satisfactory spcctra. How'th films thinner or thickcr than inch, evcellent spectra have k e n obtained, The method has shown g e a t promise in its applicability to a I;CI:STLY
variety of coatings on other metals and in the preparation of large sections of pure film. In the apparatus described, the films prepared were 1 inch long and 0.6 inch wide. These dimensions are suitable for use of the films in a potas-
sium bromide pellet holder or between salt plates in a microcell.
A direct current is passed through the film-coated plate, which is the negative side of the electrolytic cel1. The cathode is completely insulated with rubber, except for the uniform layer of enamel or varnish on the copper late. The entire apparatus is suspende in a large water bath to prevent undesirable heat effects at the cathode. The anode is a metal stirrer to maintain sufficient water circulation. Distilled water containing a trace of acid is the electrolyte. Only a small amount of current is used, to prevent cracking of the film by too rapid removal. Currents of the order of 20 ma. were used in these experiments, However, experimentation to find the desired current for films of varying thickness is necessary to obtain good clear film specimens. B using a n aspirator vacuum t o hold tle filmcoated metal firml in place on the contact plate, oompfete surface contact is obtained, thus avoiding the formation of hot spots on the metal which could lead to deterioration of the films.
B
Schematic diagram of apparatus A, B.
Motor and stirrer Cathodr.coated metal platr
C, Rubber template
ACKNOWLEDGMENT
D. Large rubber rtopper E. Contact plate F. Copper tubing G, Vacuum tubing lnsulahn H. Water bath
The authors wish to acknowledge the financial support of the Westinghouse Corp.
