US1086088A - Manufacture of filaments for tungsten-lamps. - Google Patents

Description

UNITED STATES PATENT OFFICE. A

GEORGE P. SO HOLL, OF NEW YORK, N. Y.-, ASSIG-NOR TO WESTINGHOUSE LAMP COMPANY, Al CORPORATION OF PENNSYLVANIA.

No Drawing.

Specification of Letters Patent.

Patented Feb. 3, 1914.

Application filed October 6, 1908. Serial No. 456,473.

tric lamps, and it has for itsobject to pro-- vide pure tungsten filaments and a process of producing such filaments that can be practised expeditiously and with a minimum liability to oxidize the filaments.

Heretofore, the manufacture of tungsten filaments for incandescent lamps has been carried out by first producing a paste of tungsten or tungsten compounds and a carbonaceous binder and subsequently removing the carbon contained in such carbonaceous binder. The removal of the. carbon has been accomplished along two broadly difl'erent lines, namely, either by means of oxidizing gases, such as water vapor, or by chemical action within the filament by means of oxygen-containing .com ounds of tungsten. In the first case, it as been custom ary to use tungsten metal as the main ingredient of the paste, while, in the second case, the main ingredients of the paste were oxygen-containing compounds of tlmgsten.

I have found that, in order to obtain the removal of the carbon binder, it is not necessary to use oxidizing gases, but that the same-can be accomplished by means of hydrogen.

My process is broadly distinguished from previous processes by the fact that the carbon is not removed by an oxidizing .gas or by a tungsten compound containing oxygen,

but'that, under the influence of heat, it combines with the hydrogen and is eliminated as hydrogen-carbon gas.

I am aware that ithas been proposed to.

use hydrogen as one constituent. and sometimes as the main constituent,'of the atmosphere in which the removal of the carbon of the carbon was concerned, was the oxidizing gas and the hydrogen was merely present as a protective environment. In my process, on the other hand, the hydro en is the active element which alone accomp ishes absolutely necessary. was effected, but, in all such cases, the active 9 constituent of the gas, so far as the removal the removal of the carbon, and, while I may accomplish the removal of the carbon from the tungsten filament in an atmosphere containing other inert gases, such as nitrogen, besides the hydrogen, such other inert gases do not take any part in the removal of the carbon. I am thus enabled to carry out the manufacture of the tungsten filament/in a more expeditious manner because I have no j oxidizing gases present which may oxidize the filament. Om'dization frequently took place when removal of the hydrogen was accomplished by oxidizing gases, and it was necessary to be exceedingly careful in proportioning the amount of oxidizing gases in the atmosphere in which the removal took place in order to avoid having too much of the oxidizing gases present.

While I may use any suitable carbonaceous'binding material for carrying out my process, I prefer to use a binder consisting ofan ammoniacal solution of casein, as I have found that this particular binder enables me to get along with a comparatively small amount of binder. 80

In the practice of my invention, I obtain the tungsten metal in any suitable manner, such as by reducing tungsten tri-oxid by hydrogen or'in any other convenient manner. The tungsten metal is then made up into a paste by mixing it with casein. I have found that very good results are obtained by using a paste composed of one hundred grams of tungsten with twelve grams of a ten per cent. ammoniacal solu- 90 tion of casein. "The exact proportion of 1 casein to the tungsten depends upon the desired diameter of the finished filament, as larger diameter filaments do not require as much of the binding material as the smaller diameter filaments. The proportion given above, however, is very well adapted for average filament diameters. The above mentioned ingredients are thoroughly mixed and worked up together into theform of a very stifl plastic paste, the object being to employ as little of the binding material as is As a squirted ing, already in such condition that they are sufiiciently good conductors of electricity. I am, therefore, in mostcases, enabled to do away withlthe process of carbonizing the eneral rule, I find that filaments binder which had been a necessary step in 110 y om such 'a'paste are, after squirtprevious processes. I then subject the filament material tothe action of an electric current in an atmosphere of hydrogen which may or may not contain nitrogen. The admixture of nitrogen is adopted for the purpose of obtaining a safer, practical working mixture, as the explosive force of the gas, in case of leakage from the apparatus in which the heating is performed, is very greatly diminished. The admixture of nitrogen also allows of reducing the current by which the filament is heated, inasmuch as the radiation of heat from the filament is not as high, in case a mixture of nitrogen and hydrogen is employed as it is when pure hydrogen is used.

The passage of the current through the filamentary material carbonizes it and drives off the volatilizable materials of the binder and, as the heating takes place in an atmosphere of hydrogen, the carbon contained in 'the binder combines with the latter with the result that the completed filament is composed, of absolutely pure tungsten. As the chemical reaction takes place between the carbon of the binder and the hydrogen of the atmosphere and not between the ingredients of the paste, as occurs when a paste composed of tungsten oxid and casein is used, the particles of the completed filament are very much closer together and are in much more intimate contact.

A microscopic examination shows that the filaments prepared by my process are smooth and shiny and regular on the surface, whereas filaments prepared by the other process are dull in appearance and have deeply pitted surfaces.

The breakage in my process of forming filaments'in an atmosphere of hydrogen is considerably less than is the case in filaments made from other pastes, inasmuch as the shrinkage of the filament during the forming process isgr'eatly reduced.

I claim as my invention:

1. The process of preparing tungsten filaments for incandescent lamps which consists in making a paste of tungsten and casein, forming filaments from the paste, and passing electric current through the filaments While in a non-oxidizing atmosphere containing hydrogen in which the hydrogen is the essential constituent in removing the carbon of the casein.

2. A paste for filaments consisting of pure tungsten metal and casein.

Signed at New York, in the county of New York, and State of New York, this 28th day of September, A. D. 1908.

GEORGE P. SCHOLL.