Method and apparatus for purification of hydride gas streams

Chemistry of inorganic compounds – Modifying or removing component of normally gaseous mixture

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

C095S116000, C095S117000, C095S138000, C095S139000, C423S215500, C423S219000, C423S230000

Reexamination Certificate

active

06241955

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention herein relates to the removal of contaminants from streams of ammonia and related gases. More particularly it relates to the production of substantially contaminant-free streams of ammonia and related hydride gases for use in the production of semiconductors and similar products which cannot tolerate the presence of such contaminants during manufacture.
2. Description of the Prior Art
Recent advances in integrated circuit semiconductor technology have included the development of semiconductors with copper interconnects instead of aluminum interconnects. Copper interconnects are advantageous in that copper has less resistance than aluminum, which leads to higher performance in microprocessors, microcontrollers and random access memories.
However, copper tends to migrate over a period of time, so it is necessary to construct barrier layers in the semiconductor to prevent the copper migration. Such boundary layers are typically made of nitrides such as tantalum nitride, titanium nitride or silicon nitride. These layers are commonly formed by deposition from a hydride gas, such as ammonia or silane gas.
Further, ammonia is widely used as a source of nitrogen for film development in some thin film applications. The ammonia allows for lower temperature film growth in chemical vapor deposition (CVD) processes.
In addition, there is current research to develop high performance light emitting diodes (LEDs). These are made by metal organic chemical vapor deposition (MOCVD). Trialkyl gallium is deposited as gallium nitride from a gaseous mixture of ammonia and hydrogen. Similar, gallium nitride is being considered for “blue lasers,” i.e. lasers which emit blue light. Because blue light has a shorter wavelength than red, yellow or green light, blue lasers are anticipated to be capable of forming compact disks which will have a much higher information density than is presently the case with compact disks produced with red laser light. Gallium nitride for such blue lasers would be manufactured from the same type of ammonia/hydrogen gaseous environment as described above for the high capacity LEDs.
The LEDs, blue lasers and integrated circuits are all manufactured with electron accepting p-type dopants. Such products are extremely sensitive to the presence of electron-donating n-type materials, and very small concentrations of such n-type material are sufficient to deactivate the p-type dopants and impair or destroy the performance and operability of the integrated circuits, LEDs and blue lasers. Oxygen is a particularly efficient n-type material, and the presence of molecular oxygen is detrimental to the desired band gap properties in the semiconductor material. Even very low concentrations of oxygen (<100 ppb, usually <50 ppb, even <10 ppb) can be sufficient to cause sufficient significant reduction in performance or operability so as to require discarding of the product after manufacture.
Water vapor and carbon dioxide gas are two other detrimental contaminants in hydride gases, and which lead to degradation of products which are formed by deposition of active layers of metals or metal compounds from a hydride gas environment. Water is one of the most common and yet most difficult impurities to remove from the gases. Water is of course ubiquitous in almost all ambient environments. Even systems which are nominally referred to as “dry” usually have significant amounts of water, and most drying processes can reduce the moisture content of a gas only to a “minimum” which is still in the parts per million (ppm) range. However, since for many purposes water contents in the ppm range are quite acceptable, there are numerous patents and articles in the literature dealing with such types of “ppm drying processes.”
In the manufacture of the aforesaid LED, blue laser and integrated circuit products, however, moisture contents of the depositing gases which are in the ppm range are excessively wet. To form satisfactory products, the water content of the depositing gases must be reduced to the parts per billion (ppb) range, usually down to no more than about 100 ppb. See Whitlock et al, “High Purity Gases,” in Ruthven, ed., ENCYCLOPEDIA OF SEPARATION TECHNOLOGY, vol. 1, pp. 987-1000 (1997).
Attempts to use materials such as reduced nickel or copper catalysts to remove contaminants such as oxygen, carbon dioxide and water from hydride gases have not been successful. While contaminant removal can be effected for short periods of time down to the 10 ppb level, the reactive effects of the hydride gases, especially ammonia, very quickly cause the materials to degrade and contaminate the gas stream with metal complexes. Though pre-existing impurities may be reduced, the introduction of new impurities to the manufacturing process is unacceptable.
Some prior art processes have been described in which oxygen has been removed from ammonia streams by metals serving as “getters.” However, these have been relatively ineffective at reaching sufficiently low levels of decontamination. In addition, the getters are deposited on substrates, such as silica or zeolites, which do not play a central role in the decontamination process, and also may themselves be degraded by the hydride gases. See, for instance, U.S. Pat. Nos. 5,496,778 (Hoffman et al.), 5,716,588 (Vergani et al.) and 4,976,944 (Pacaud et al.); PCT publication No. WO 97/06104 (SAES Getters S.p.A.); and European Patent No. EP 0 784 595 B1 (SAES Getters S.p.A.). In particular, some of these references teach that manganese:iron ratios of >2:1 as depositions on such substrates are detrimental to getter performance and are to be avoided. The references specifically teach that very low manganese:iron ratios, usually about 0.12-0.16:1, are to be preferred. Further, the reference processes are usually not effective for removal of carbon dioxide or water, as compared to oxygen, from ammonia gas streams.
Consequently, the problem of removal of contaminant levels down to ≦100 ppb from hydride gases remains a significant problem in the field of production of high purity LEDs, blue lasers, semiconductors, and the like. Those processes which are being used are expensive because of the very short service life of the decontaminating materials and the need for their frequent replacement. In addition, since it is difficult to determine the exact rate of deterioration of the decontaminating materials in the presence of the reactive hydride gases, users of such decontaminating materials must schedule their discard and replacement at intervals less than the shortest expected service life. To do otherwise would risk failure of a decontamination unit with the resultant loss of contaminated product when the excessive contaminant concentrations reaches the production chamber through the failed unit. Consequently, the current systems require that many if not most of the decontamination units must be discarded while they still have some degree of useful service life left, thus further increasing the expense of the system operations.
In most high-purity product manufacturing processes, it has been conventional for hydride gases to be stored in and supplied from standard gas cylinders. The volume of gas in each such cylinder is of course limited, so that in larger scale manufacturing processes, it becomes necessary for process operators frequently to replace emptied cylinders and replace them with fresh, full cylinders. This frequent handling and movement of heavy, awkward gas cylinders represents a safety hazard to the operators, as well as providing opportunities for gas leakage and increasing the cost of manufacturing. Also importantly, each time an empty gas cylinder is detached from the system and a new full cylinder attached, there is an opportunity for ambient contaminant gases, such as oxygen, carbon dioxide and water vapor, to enter the system, thus increasing the decontamination load on the system and accelerating the system degradation. The industry is beginning to require gases to be supplie

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Method and apparatus for purification of hydride gas streams does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Method and apparatus for purification of hydride gas streams, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for purification of hydride gas streams will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-2534534

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.