Opinion
No. 578.
September 18, 1926.
Pennie, Davis, Marvin Edmonds and John F. Neary, all of New York City, for plaintiff.
Clair W. Fairbank and Irving M. Obrieght, both of New York City, and William G. Mahaffy, of Wilmington, Del., for defendant.
In Equity. Patent infringement suit by the Air Reduction Company, Inc., against the Carbo-Oxygen Company and another. Bill dismissed.
Patents No. 959,563 and No. 957,170, for improvements in the method of separating gases from their mixtures, particularly oxygen and nitrogen from atmospheric air, constitute the basis of this infringement suit of Air Reduction Company against Carbo-Oxygen Company, in which the defenses are invalidity because of the prior art, insufficiency of disclosure, inoperativeness, and noninfringement.
The desired product of the separation is oxygen. Levy and Helbronner, the patentees, were not the first to obtain oxygen from the air. By the processes of their patents in suit, of which September 13, 1902, and June 22, 1906, respectively, are the agreed effective dates, they sought a yield of that product in excess of that obtainable by the processes of the prior art. Their method of producing that result, as set out in the former, hereinafter called the earlier, patent, all of whose claims are in issue, is stated in typical claim 5 to be:
"In the separation of oxygen and nitrogen by means of a liquefaction of the air and its rectification, in combination, a first rectification, the reliquefaction of the gas rich in nitrogen resulting from the said first rectification with a further rectification by means of the reliquefied mixture."
Liquefaction of air, the conversion from a gaseous to a liquid state, has long been the first step in obtaining oxygen from the air. It is accomplished by a reduction of temperature. The extent of the reduction required is great. It is lessened by increasing the pressure upon the air to be liquefied. The reason for this step, liquefaction, is that, in the partial revaporization, called fractionation, of the liquid, the gases given off are richer in nitrogen, because of its greater volatility, and the remaining liquid richer in oxygen, than the original liquid. But by simple fractionation the remaining liquid does not become pure or nearly pure oxygen until shortly before the complete vaporization of the original liquid, with the consequent loss, in the vapors, of practically all the oxygen. Rectification was added to recover some of the oxygen from the waste gases resulting from the fractionation. By this process the vaporization of the liquid is brought about in the bottom of a rectification column, wherein the warm vapors ascending from the boiling pool and passing up the column are brought into contact with a cold descending stream of oxygen-nitrogen liquid, which reliquefies the less volatile of the ascending vapors, oxygen, causing it to join the descending stream and again to enter the vaporizing pool, while the heat of the vapor and the latent heat set free by the liquefaction of the oxygen vaporize an equivalent quantity of the more volatile constituent, nitrogen, of the descending liquid. The products of liquefaction, fractionation, and simple rectification are liquid oxygen and an effluent waste gas containing 93 per cent. nitrogen and 7 per cent. oxygen. As the atmosphere is composed of nitrogen, 79 per cent., and oxygen, 21 per cent., a short mathematical calculation discloses that the oxygen waste of the simple rectification process represents 5.94 parts of the original 21, or almost 29 per cent. thereof. By the process of their earlier patent in suit Levy and Helbronner proposed to salvage or recover some or all of this waste by a reliquefaction and rectification of the effluent gases resulting from the first rectification.
The defendant takes the position that the method of reliquefaction and rectification disclosed and claimed in the earlier patent in suit is but a duplication of the simple rectification process brought into the air separation art by Linde and published in the Zeitschrift on August 9, 1902. In carrying out his process Linde conveyed air, brought to a pressure of from two to three atmospheres, through a countercurrent heat exchanger (consisting of coils of pipes of two diameters, having a common center, through the larger of which the cold effluent gases of his process flow counterwise to the compressed air in the smaller, chilling it almost to the point of liquefaction) to the lower chamber, K¹, of a vessel at the bottom of a rectification column. This lower chamber is separated from the upper part of the vessel by a tube plate forming the support of numerous metal tubes, open at their lower and closed at their upper end, so that there is a flow of the compressed air into, but not through, them, which project into the upper portion of the vessel. This upper portion, V¹, is an open vessel, which, when the apparatus is in its normal state of working, is filled with liquid rich in oxygen. This liquid, whose boiling temperature at atmospheric pressure is about that of oxygen, — 182.5° C. or 90.5° on the absolute scale (which has its zero at — 273° C., a point approximating that of complete absence of heat), is brought to a boil by the compressed air in the chamber K¹. The latent heat required by the vaporization of the liquid is supplied by the compressed air, and, as air under a pressure of three atmospheres liquefies at a temperature higher than — 182.5° C. (see Olzewski's table), the extraction of the latent heat of the compressed air by the vaporizing oxygen-rich liquid brings about the liquefaction of a quantity of air approximately equal to the evaporated quantity of the oxygen-rich liquid. The vapors pass up the rectification column, made of a cylindrical casing filled with glass beads and resting upon the vessel V¹, while the liquefied air is conducted in a pipe to the top of the column and there discharged, through a pressure-reducing valve, at atmospheric pressure. The liquid trickling down is met by the vaporization products from V¹ which now deliver oxygen to the liquid while an equivalent quantity of nitrogen which vaporizes, boils, at atmospheric pressure at — 195.5° C., 77.5° absolute, or 13 degrees lower, and so more readily, than oxygen, passes from the liquid to the gas current. At the top of the column this current, then containing about 93 per cent. nitrogen and 7 per cent. oxygen, passes out to the atmosphere through the countercurrent heat exchanger, where it chills, abstracts the sensible heat of, the incoming compressed air. The descending oxygen enriched liquid falls into the vessel, V¹, where the greater part is vaporized, while the smaller part, oxygen, or a liquid having a very high oxygen content, flows into a vessel V², designed like the vessel V¹, and is there gasified by a stream of chilled compressed air passing through a bottom chamber K², in the same manner as is the liquid in the vessel, V¹. This gas is the ultimate product of the process. The air liquefied in the chamber, K², like that liquefied in K¹, is made to pass into the top of the rectification column. In British Liquid Air Company, Ltd., v. British Oxygen Company, Ltd., 25 R.P.C. 577, Lord Justice Fletcher Moulton, speaking in 1908 of Linde's introduction of the rectification principle into the air separation art, said:
"By this invention, Linde solved the problem of mechanically separating commercially pure oxygen from the air, and laid the foundation of what will no doubt become an important industry. He himself regarded the invention as an addition to the method of obtaining oxygen from liquid air by means of partial evaporation. * * * In my opinion, he was justified in so regarding it. The failure of partial evaporation as a method of obtaining oxygen arose from the fact that when the mixture became rich in oxygen, the gas given off by further evaporation became also so rich in oxygen that it occasioned a serious loss of the very substance that the process was designed to produce. But, by using rectification, the oxygen that was evaporated off was no longer lost. On the contrary, it was arrested and returned with a large increase derived from the descending liquid air, the nitrogen of which had been substantially driven off by the innumerable partial evaporations which had taken place in the rectifying column."
The apparatus described in their earlier patent by Levy and Helbronner as suitable for carrying out their claimed process consists of an air compressor and countercurrent heat exchanger to chill the incoming compressed air and of a cylindrical rectification column divided horizontally into two main compartments each containing, at the bottom, coils in which the incoming compressed oxygen-nitrogen gases (air in the lower and the effluent, 93 per cent. nitrogen and 7 per cent. oxygen, of the lower in the upper) are liquefied by the vaporization of the oxygen-rich liquid immersing the coils. Each compartment is subdivided into a number of small chambers by plates or trays provided with passages for the upward flow of the vaporization products and for the downward flow of the liquefied oxygen-nitrogen gas admitted from the respective liquefying coils to the top of each compartment. Beneath the lower main compartment is a smaller one, A, into which the liquid oxygen flows, from the bottom of the lower main compartment, for conversion by the compressed air in the immersed air liquefaction coils into the oxygen gas there drawn off as the ultimate product of the process. Between the two main compartments is another small compartment, A', through which and the pipe M the liquid product of the upper main compartment, operated at atmospheric pressure, is intermittently "locked" down to the lower main compartment, which is operated at a pressure of about four atmospheres. Most of the air liquefied in the oxygen compartment, A, joins that flowing from the liquefaction coils at the bottom of the lower main compartment, and empties on the top tray of that compartment; but, to compensate the upper main compartment for the loss of "cold" resulting from the delivery of some of its product in liquid form through A' and the pipe M to the lower compartment, some of the air liquefied in A is delivered to an intermediate tray of the upper main compartment. In each of the main compartments the warm ascending oxygen-nitrogen vapor and the cold descending oxygen-nitrogen liquid are rectified by their action upon each other. The waste gas passing from the upper compartment to the countercurrent heat exchanger to chill the incoming compressed air is composed of 98 per cent. nitrogen and 2 per cent. oxygen.
Is the process of Levy and Helbronner but a duplicate of that of Linde? In finding the answer to this inquiry we are not concerned with the substitution by Levy and Helbronner of equivalent plates and trays for the glass beads of Linde, with the particular location of the liquefaction coils in the lower main compartment, with the particular means shown in the drawings and described in the specification as appropriate for conducting the rectified liquid product of the upper compartment to the lower, with the failure to indicate on the drawings pressure-reducing valves at points where the specification makes it obvious such valves are to be placed, or with any or all of these in combination. They are details not constituting or affecting the claimed steps. They may be pertinent, possibly, to the defenses of insufficiency of disclosure and inoperativeness, but not to the broad defense of anticipation and want of patentable novelty, in view of the prior art. What steps, if any, of Linde's process have Levy and Helbronner employed? In the bottom of their upper compartment Levy and Helbronner, as did Linde in his single rectification column, liquefy under pressure an oxygen-nitrogen gas through an exchange of heat, of an oxygen-rich liquid bathing the heat transferring walls of the compressed-gas container. In each the mixture thus liquefied is emptied through a pressure-reducing valve into the top of the column, operating at atmospheric pressure, and is brought into close and intimate contact with the ascending vapors which give up their oxygen to the liquid in exchange for its nitrogen. In each the vaporizing liquid which brings about the liquefaction of the compressed oxygen-nitrogen gas is a rectified reliquefaction product. In each the pressure differential maintained between the liquefaction coils or chamber and the rectification column serves the same dual purpose. It raises the liquefaction temperature of the gas to be liquefied above the boiling or vaporizing temperature of the immersing liquid, and thus makes possible the liquefaction of the gas by the immersing liquid. Again, at the temperatures here employed, the discharge at a lower pressure of a compressed oxygen-nitrogen liquid brings about through expansion a marked fall in its temperature (known as the Joule-Thomson law or effect), and thereby the liquid is not only reduced below its liquefaction temperature at the reduced pressure, and thus enabled to remain a liquid at the reduced pressure, but it is thereby made cold enough as a descending stream to reliquefy the oxygen of the ascending vapors in the rectification column.
It is, of course, true that the grist of Linde's process was air, an oxygen-nitrogen mixture of 79 per cent. nitrogen and 21 per cent. oxygen, while that of Levy and Helbronner's upper column is an oxygen-nitrogen mixture of 93 per cent. nitrogen and 7 per cent. oxygen; but there is no evidence to show that a change in the relative proportions of the constituents of an oxygen-nitrogen mixture, gas or liquid, brings about a reversal, or even a radical change, of its properties. In British Liquid Air Company, Ltd., v. British Oxygen Company, Ltd., Lord Justice Moulton there said:
"* * * It must be borne in mind that rectification does not depend on any particular proportions in the mixture which is to be operated upon. * * * I therefore fail to see what relevance there is in the suggestion that the plaintiffs deal with a different mixture in the process of rectification."
It is likewise true that the pressure in the liquefaction coils of the upper rectification column of the earlier patent in suit is somewhat greater than the pressure employed by Linde. But Baly by his tables and his chart, known as "Baly's Curves," published in March of 1900, had shown the boiling or vaporizing temperature at atmospheric pressure of oxygen-nitrogen liquids of every possible relative oxygen-nitrogen proportion, while Olzewski years earlier had disclosed the boiling, or, conversely stated, the liquefaction, temperature of air of various pressures. It required no invention, then, to ascertain the pressure to which the oxygen-nitrogen mixture in the liquefaction coils of Levy and Helbronner's upper column should be subjected in order to make possible the liquefaction of the mixture at the boiling temperature of the liquid immersing the coils, or, otherwise expressed, to produce the difference of boiling temperature necessary for the transfer of heat. But how should the pressure upon the gases in the liquefaction coils of the upper column be brought about? Raising their pressure from atmospheric (at which Linde operated his column and let his effluent gases escape) by renewed compression constituted, of course, one method. But that would have permitted the gases having the frigidity of the lower or first column to lose their "cold," become heated, and require, not only their recompression, but, as well, their being conducted anew through a heat exchanger before being sent to the liquefaction coils of the upper column.
Expense and inconvenience would attend the use of that method; so the patentees employed another. They pass the gases directly and without recompression from the lower column to the coils of the upper. To do this the lower column is operated under pressure, but it is so operated for no other purpose. The direct passage of the gases from the lower column to the coils of the upper is, however, without effect upon the extent to which the pressure upon the gas in the coils of the upper column exceeds the pressure upon their immersing liquid in order to boil the liquid and to liquefy the gas. That pressure differential is ascertainable, approximately, from Baly's curves and Olzewski's table, and remains the same for a gas of any specific oxygen-nitrogen composition, without regard to whether the columns are operated independently or are connected with a view to multiple effect. The functioning of a column is dependent, not upon the pressure at which it is operated, but upon the maintenance of the required differential between the pressure in the column and that in the liquefaction coils of that column. Consequently, to obtain through the lower column, from the compressor for the air in the coils of that column, the pressure required in the coils of the upper column, it was but necessary to increase the original pressure by the amount required in the second column, and still maintain the required differential in pressure between the first column and its coils. That is what the patentees did. Of that method an inevitable incident is that the consecutive columns will be under a pressure diminishing from the first to the last.
With the exception of an increase of pressure in that column and a like increase in its coils, Levy and Helbronner operated their lower column and its coils by Linde's steps, for his purpose, with his result and upon his raw material. The evidence does not disclose that an increase of pressure to the same extent in the column and in its coils had any effect upon Linde's process or his result. The increase in the oxygen yield obtained by the patentees was brought about solely by the retreatment, by the Linde process, of the effluent gas resulting from a first treatment of the air by his process. The total oxygen recoverable from a definite volume of air is the same, so far as the record shows, whether the treatment of the effluent gases of a first process is carried on in wholly independent rectification columns, or in Linde columns obtaining their pressures from a common source. By obtaining the pressure for the coils of their second column through the first column from the compressor for the coils of the first column, Levy and Helbronner eliminated the necessity for a separate compressor and heat exchanger for the second column. But their method of making the compressor and the heat exchanger for the first column serve as well for the second column does not in itself give patentable novelty to the claims, unless the skill required to make one compressor and one heat exchanger serve for two columns rose to the dignity of invention. As I view it, skill of that character is not there found.
They likewise eliminated the need for a separate oxygen compartment for the upper column. Did that require invention? This elimination is made possible by causing the products of the two columns, or rather their combined product, to pass into the one oxygen compartment, there to be vaporized, and to divide between the two columns the air liquefied by the vaporization of the combined product; the portion going to each column not being stated, but being, presumably, equal to the volume of the liquid product of that column. The return to the upper column of the amount of liquid withdrawn from it, which is necessary to maintain its thermal balance, is in no sense different from the return to Linde's column, from the liquefaction coils of his separate oxygen compartment, of the amount of liquid oxygen conducted from Linde's column to his oxygen compartment, for the volume of air there liquefied is equal to the volume of liquid oxygen there gasified. As the method of bringing about the elimination of the separate oxygen compartment for the upper or second column is by increasing the capacity of the oxygen compartment for the first column, and not otherwise, there is here, too, as I view it, a want of invention. The method of obtaining the air for the coils of the upper column and that employed to vaporize the liquid product of the second column are wholly independent and without joint effect. Consequently, when considered together, they are as devoid of invention as when viewed separately. Furthermore, the patent is purely a paper patent, whose application remained long in the patent office. It has made no imprint upon the art. As I view it, it is a duplication of Linde's process and without patentable novelty.
By the process of the second patent in suit, No. 957,170, the vapors passing beyond or rising from the tray in the column upon which the air liquefied in the coils at the bottom of the column is discharged are not piped to a second column for reliquefaction and rectification. Instead they continue to ascend in the vertical continuation or extension of the column likewise divided into compartments by trays. Passing to the top, some of the gas comes into contact with a multitubular heat-conducting plate like that used by Linde, separating the rectification column from an upper chamber into which the liquid oxygen or oxygen-rich liquid withdrawn from the bottom of the column is emptied. The remaining gases pass out of the column near the top to the heat exchanger. That part of the gas which comes into contact with the tubular plate gives up its heat to the oxygen, boiling it, and is in turn liquefied. The liquid produced falls and becomes the rectifying stream, richer in nitrogen than is liquid air, descending from the top of the column. The oxygen compartment is operated at atmospheric pressure, the column at a higher, and the liquefaction coils at a still higher pressure. The first four claims are in issue. The second, which may be taken as representative, is:
"A process for the separation of a gaseous mixture into its constituents consisting in separating the same after liquefaction into its constituents by rectification effected under pressure, reliquefying the more volatile gaseous constituent resulting therefrom while still under pressure, by causing it to vaporize under a lower pressure the less volatile constituent resulting from the separation, and utilizing the reliquefied constituent to purify by rectification the gaseous constituent."
The advantage of this process arises from the fact that the vaporization products of a mixed liquid whose constituents have different boiling points are richer in the more volatile ingredient than is the liquid from which the vapors pass. An oxygen-nitrogen liquid is of that character; the nitrogen being the more volatile constituent. The oxygen-nitrogen content of the gas given off by oxygen-nitrogen liquids of every composition from pure oxygen to pure nitrogen, when those liquids begin to vaporize, boil, is shown by Baly's chart. The upper curve of the chart shows the temperature on the absolute scale at which the mixtures of gas (the relative oxygen-nitrogen content of which is expressed at the bottom of the chart in percentages of oxygen) begin to condense into a liquid at atmospheric pressure, and the lower curve the temperature at which the liquefaction is completed, or, conversely expressed, the temperature at which the liquid begins to boil. The temperature of each point in the curves is determined by its elevation or horizontal position; in short, by its latitude. The points at which a vertical line intersect the lower and upper curves are the boiling (beginning) point and the condensing (beginning) point, respectively, of a liquid and a gas having the same oxygen-nitrogen composition. As a boiling liquid and the vapors given off by it have the same temperature, the oxygen-nitrogen composition of the gas given off by any oxygen-nitrogen liquid of a known composition when first subjected to a boiling temperature is that indicated by the point on the upper curve which has the same temperature (latitude) as the point on the lower curve representing the liquid. Liquid air is represented on the lower curve by the point crossed by a vertical line resting on the bottom line at 21. The point in the upper curve having the same temperature is in the vertical line 7. Hence the vapor given off by liquid air when it begins to boil contains 7 per cent. of oxygen and 93 per cent. of nitrogen. If the gas given off by, or in equilibrium with liquid air is reliquefied, the gas in equilibrium with that liquid contains 98 per cent. nitrogen and 2 per cent. oxygen, as shown by the chart, and if this gas is reliquefied and the resultant liquid fractionated its vaporization products will be almost pure nitrogen. In the process of rectification, which is, in a sense, fractionation, repeated on each tray by the heat of the vapors given off by the liquid on the next lower tray, the effluent gas is that in equilibrium with the liquid on the topmost tray. Since the liquid on the topmost tray of a rectification column operated by the process of patent No. 957,170 is (by reason of the reliquefaction and revaporization of the gas given off by the liquid air and its revaporization, followed by a reliquefaction of that gas and a renewed revaporization, repeated indefinitely) almost pure nitrogen little, if any, oxygen escapes in the effluent gas. It passes to the pool at the bottom, is carried to the oxygen compartment at the top, and there passes off in the form of gas as the ultimate product of the process. However, pure oxygen and pure nitrogen cannot be obtained simultaneously, as stated in the specification. If pure nitrogen is desired as the ultimate product, the liquid product of the process will not be pure oxygen.
But on March 20, 1906, Linde obtained his patent for an apparatus employing the same process. That was before the effective date, June 22, 1906, of this patent. It is, of course, true that Linde operated his oxygen compartment at the top of the column at subatmospheric and his column at atmospheric pressure, while Levy and Helbronner operate their column above atmospheric and their oxygen compartment at atmospheric pressure. Yet in each the crucial factor — the pressure differential — is the same. I think the disclosures of the Linde patent constitute a complete anticipation of Levy and Helbronner patent, No. 957,170.
To these two patents, whose respective effective dates follow hard upon Linde's announcement, through printed publication and patent, of his two epoch-making advances in the air separation art, the oft-quoted words of Mr. Justice Bradley, in Atlantic Works v. Brady, 107 U.S. 192, 199, 200, 2 S. Ct. 225, 27 L. Ed. 438, are, I think, peculiarly appropriate.
The bill of complaint must be dismissed.