This invention relates to the field of electrical energy storage and in particular to multilayer ceramic capacitors with high relative permittivity and high dielectric strength.
In 1965, Gordon Moore, one of the founders of Intel, first wrote what later became known as “Moore's Law”. Often misquoted, Dr. Moore in fact made the observation that the complexity for minimum component costs had increased at a rate of roughly a factor of two per year and postulated that this rate was likely to continue in the short term, as described by G. E. Moore in Electronics 38(8), 4 (1965). Moore's Law has paced the rate of semiconductor advances for almost half a century and has given rise to the mantra “smaller, faster, lighter, cheaper” frequently mentioned when pundits speak of the future of integrated circuits (ICs).
However, ICs are not the only electronic components that have witnessed a substantial decrease in critical component size: another example is the essential multilevel ceramic capacitors (MLCCs) often found in close proximity to them on printed circuit boards. In fact, the rate at which capacitance and volumetric efficiency for MLCCs has increased since 1994 has exceeded Moore's Law, doubling approximately every 13-14 months, as described by M. Randall, D. Skamser, T. Kinard, J. Qazi, A. Tajuddin, S. Trolier-McKinstry, C. Randall, S. W. Ko, and T. Dechakupt in CARTS 2007 Symposium Proceedings, Albuquerque, N. Mex., pp. 403-415, March 2007, while recent “Moore's Law” IC performance has doubled every ˜18 months. These rapid advances cannot be maintained ad infinitum.
In order to stay on pace, active layer counts and dielectric permittivity must increase, while dielectric and metal electrode thickness must decrease. This leads to a dilemma: the thick film techniques such as tape casting that are used to manufacturing most of today's MLCCs do not scale well to really small layer thicknesses, while thin film techniques such as sol-gel deposition, chemical vapor deposition (CVD) and physical vapor deposition (PVD) are too expensive to be used to fabricate an active layer count that is rapidly approaching 2,000. Clearly, revolutionary new processing techniques and/or new materials will be required if tomorrow's MLCCs are to achieve the rate of miniaturization needed in the “smaller, faster, lighter, cheaper” electronics of the future.
Another problem posed by decreasing capacitor dielectric layer thickness is increasing electrical leakage and/or dielectric breakdown. The latter can also lead to problems with cycle life. Typically, in order to maintain their electrical integrity, ceramic capacitors should not be subjected to voltages greater than ˜10% of the electric field required for breakdown. While in general, high quality films deposited with sol-gel, CVD or PVD usually exhibit an inverse relation between film thickness and dielectric strength, this only holds down to a certain film thickness. Films deposited with thick film techniques usually require a minimum of four grains across the thickness of the capacitor for reliability purposes.
It is the purpose of this invention to describe a manufacturing method that allows high performance MLCCs to be fabricated at a commercially acceptable cost. Moreover, the techniques described herein will enable capacitors so formed to store relatively large amounts of electrical energy in a small volume.
This object is achieved by the independent claims. Advantageous embodiments are detailed in the dependent claims.
Particularly, the object is achieved by a method for manufacturing a high performance multi layer ceramic capacitor, comprising the steps of:
According to another preferred embodiment of the invention, the method further comprises the steps of:
According to another preferred embodiment of the invention, the method further comprises the step of:
According to another preferred embodiment of the invention, the method further comprises the steps of
According to another preferred embodiment of the invention, the dielectric layers deposited during steps d) and f) are deposited such that the thickness of the low-k dielectric layer is ≦5% of the thickness of the high-k dielectric ceramic layer.
According to another preferred embodiment of the invention, the thick-film deposition method comprises screen printing and/or tape casting.
According to another preferred embodiment of the invention, the thin-film method deposition comprises sol-gel deposition, sputtering, evaporation, ion plating, pulsed laser deposition, atomic layer deposition, chemical vapor deposition, plasma-enhanced chemical vapor deposition, electrografting, electroplating and/or electroless plating.
According to another preferred embodiment of the invention, the substrate comprises a metal, a ceramic and/or a glass, preferably alumina, mullite, quartz, silicon, a refractory metal foil, most preferably nickel or nickel alloys.
According to another preferred embodiment of the invention, the electrode layer comprises nickel, copper, platinum, iridium, rhodium, palladium and/or alloys of palladium and/or of silver.
The object of the invention is further addressed by a high performance multi layer ceramic capacitor, comprising
According to another preferred embodiment of the invention, the capacitor further comprises
According to another preferred embodiment of the invention, the capacitor further comprises a plurality of first and second layer sets arranged each on top of each other.
According to another preferred embodiment of the invention, the thickness of the low-k dielectric layer is ≦5% of the thickness of the high-k dielectric layer.
According to another preferred embodiment of the invention, the low-k dielectric layer is deposited by sol-gel deposition, sputtering, evaporation, ion plating, pulsed laser deposition, atomic layer deposition, chemical vapor deposition, plasma-enhanced chemical vapor deposition, electrografting, electroplating and/or electroless plating.
These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.
In the drawings:
It is instructive to look at the figures to understand the invention described herein.
If desired, the bottom electrode can be now be heat treated to increase its density and/or to remove the organic and volatile components of the inks and binders used, for example, in the screen printing process. Where base metal electrodes such as Ni and/or Cu are present, this must be performed in vacuum or in another reducing environment. Alternatively, the heat treatment steps for the electrodes and the ceramic dielectric layers can be combined. Electrode layers deposited by sputtering at several hundred degrees Celsius are usually of high density and low resistivity and typically will not require a post-deposition heat treatment prior to depositing the dielectric. In a preferred embodiment, the bottom electrode should be deposited so as to extend all the way to one edge of the substrate but should not extend completely to the opposite edge: this pattern is easily achieved by screen printing or by PVD through a shadow mask.
Following deposition of the bottom electrode, the ceramic dielectric is deposited. Again, this can be by thick film techniques such as screen printing or tape casting that are preferred for cost reasons but where dielectric layer thinness is paramount, thin-film techniques including but not limited to sol-gel deposition, sputtering, evaporation, ion plating, pulsed laser deposition, atomic layer deposition, chemical vapor deposition, plasma-enhanced chemical vapor deposition and “electrografting” can be employed. In most cases, a post-deposition heat treatment of the ceramic dielectric will be required—a high temperature firing in vacuum or in a reducing environment to remove the organic and volatile components of the inks and binders used, for example, in the screen printing process and also to form the desired crystal and grain structures for high-k materials such as doped barium titanates that must be converted to their perovskite phase. Often, a second heat treatment at lower temperature in an oxidizing ambient is performed to anneal out any oxygen vacancies formed in the dielectric during high temperature processing and that can give rise to electrical leakage in the capacitor. It should be noted that if noble metal electrodes are used, it is not necessary to perform a two step anneal—a single high temperature firing in a controlled oxidizing ambient is usually sufficient. In addition, the introduction of certain dopants into high-k dielectrics such as barium titanate can obviate the requirement for the second firing step described above by compensating for any oxygen vacancies in the lattice. Furthermore, high temperature, typically >600° C., sputtering and CVD methods usually deposit doped barium titanates in the desired perovskite phase, thereby reducing the maximum post-deposition firing temperature of the dielectric. In principle, the use of noble metal electrodes with certain doped dielectrics deposited by high temperature PVD or CVD can dispense with the requirement for heat treatment entirely.
Continuing the sequence onto
C
Total
={C
Hi-k
+C
Lo-k
}/C
Hi-k
·C
Lo-k
where CTotal is the total capacitance of the two layer structure, CHi-k is the capacitance due to the high-k dielectric and CLo-k is the capacitance due to the low-k dielectric. However, in cases where the thickness of the low-k dielectric is ˜5% or lower than the thickness of the high-k layer, then modeling of the composite structure by analytical and numerical methods predicts that the overall composite will behave primarily as a capacitor with a large capacitance, rapidly approaching CHi-k as the volume fraction of the high-k dielectric tends to 100%. This result has been verified experimentally by measurements on composite capacitors that contained large volume fractions of high-k material in a low-k matrix that comprised <5% of the total dielectric volume. For fabricating the optimum structure, atmospheric CVD is again preferred because thermal CVD is able to penetrate into very small spaces, even between the gaps of the individual high-k grains. In this way, an internal barrier-layer type capacitor dielectric is formed with a large capacitance but with reduced leakage and increased dielectric breakdown strength. Due to the required thinness of the high breakdown material, it can be deposited relatively quickly and therefore relatively inexpensively.
The preferred embodiment described that proposes the use of a continuous processing atmospheric, or near atmospheric, CVD deposition system with multi-zone furnaces lends itself very well to fully automated manufacturing. Following deposition of the ceramic, the substrates can be introduced into a multi-zone furnace where the first high temperature zone incorporates a reducing ambient, the second zone incorporates a controlled oxidizing ambient and the third zone incorporates the deposition process: suitable gas curtains separate each zone from the one prior. Different zones can be regulated at different temperatures as practiced by those skilled in the art.
After deposition the capacitor stack is allowed to cool and a second layer of Ni or other suitable metal electrode material is deposited according to the techniques already described. Optionally, this layer can be subjected to post-deposition heat treatment if desired. At this point, the capacitor structure is introduced into a suitable apparatus for etching. This is most economically performed with a laser, with or without chemical assist. The laser power and raster speed should be adjusted so as to cut a trench through the Ni, or other metal electrode layer, and through the thin layer of low-k dielectric. In this manner, a series of parallel trenches can be etched across the whole substrate, thereby preparing the structure for eventual singulation into MLCCs of the requisite size. Alternatively, the capacitor stack can be lithographically patterned and etched in a reactive ion etcher or plasma etcher, or by wet chemical methods but this increases complexity and hence overall cost. Moreover, both chemical and plasma etching of noble metals is difficult; plasma etching is also not well-suited to Ni and Cu electrodes as few sufficiently volatile etch products are known for these metals.
Referring to
Opposing faces of these strips are then metallized by methods known to those skilled in the art, including but not limited to sputtering, evaporation, ion plating, pulsed laser deposition, atomic layer deposition, chemical vapor deposition, plasma-enhanced chemical vapor deposition, electroplating and electroless plating, —see
In addition to increasing the overall dielectric strength and reducing the electrical leakage of conventionally fabricated MLCCs, the method described here that combines thick film and thin-film deposition techniques can also be used to advantage in other ways. For example, MLCCs fabricated with thick film means reportedly require at least four grains of high-k dielectric between adjacent electrodes, as described by M. Randall, D. Skamser, T. Kinard, J. Qazi, A. Tajuddin, S. Trolier-McKinstry, C. Randall, S. W. Ko, and T. Dechakupt in CARTS 2007 Symposium Proceedings, Albuquerque, N. Mex., pp. 403-415, March 2007. Thin-film processes such as CVD and/or PVD could be used to deposit high quality, high integrity, small-grained layers on top of similar thicker layers deposited by tape casting or screen printing, thus allowing the overall thickness of the dielectric in each layer to be reduced. Depositing the whole dielectric layer by thin-film techniques would be much slower and therefore more costly.
The method described herein also allows for the use of novel high-k capacitor dielectrics that would otherwise prove too leaky. For example, the material CaCu3Ti4O12 (CCTO) has a reported relative permittivity close to 100,000, as described by C. C. Homes, T. Vogt, S. M. Shapiro, S. Wakimoto and A. P. Ramirez in Science 293, 673, 2001, but it has too high an electrical conductivity for application as a capacitor dielectric. By building structures with very thin electrically blocking layers. e.g., SiNx, SiO2, Al2O3, etc., ideally deposited by thermal CVD or by techniques that will allow the high strength dielectrics to infiltrate into the grain structure, MLCCs using CCTO, related doped compounds or other ultra-high k materials could be envisioned.
An alternative embodiment to the method described herein is the use of chemical vapor infiltration (CVI) to create a composite structure consisting of high-k dielectric material surrounded by a matrix of high breakdown strength, electrically insulating material. This CVI process could be performed before, during or after one or both of the high temperature firing/anneal steps of the high-k ceramic dielectric described earlier.
While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments. Other variations to be disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting scope.
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/EP2011/074317 | 12/30/2011 | WO | 00 | 10/8/2013 |
Number | Date | Country | |
---|---|---|---|
61429577 | Jan 2011 | US |