Tantalum Capacitor History Part 2: Improving The Solid Tantalum Capacitor

2025-01-06

Philip Lessner, James Q. Chen, and Hideaki Sato

This post is based on a presentation given at the Tantalum and Niobium International Study Center (T.I.C) 65^th General Assembly in Tokyo, Japan, September 2024. After the invention of the solid Tantalum capacitor, covered in the previous post, it went though rapid improvement in performance and reliability. This post covers these improvements which were the result of investigations into the fundamental science of the capacitor materials and engineering application of those studies.

Improving the Solid Tantalum Capacitor

The invention of the solid tantalum capacitor came at a propitious time. The transistor had been introduced a few years earlier which meant lower voltages and circuit miniaturization both of which utilized the strengths of the solid tantalum capacitor. It was also the beginning of the space age and satellites, and manned space missions required highly reliable electronics which could survive a vacuum and both high and low temperature conditions. This made capacitors with wet electrolytes like aluminum electrolytic unsuitable for these applications.

Starting in the early 1960’s, many improvements were made to the solid tantalum capacitor process. One set of improvements was focused on the amorphous Ta₂O₅ dielectric which is formed by electrochemical anodization of the tantalum anode. However, the Ta₂O₅ is susceptible to losing oxygen transforming to Ta₂O_5-x which is not as good an insulator (the oxygen vacancies created in this transformation are conducting) which causes leakage current increases and eventual failures. The other mechanism of failure is crystallization of the amorphous dielectric. The crystals grow perpendicular to the dielectric plane and ‘punch through’ the dielectric causing high leakage currents and eventual failure.

Early on, it was recognized, that the high temperature of the conversion process of manganese nitrate to manganese dioxide had a destabilizing effect of on the dielectric by driving oxygen from the dielectric into the Ta anode creating oxygen vacancies. The reform step was introduced to partially solve this issue. Smyth, Shirn, and Tripp ¹ at Sprague Electric discovered that heat treating (at around 400°C) the Ta/Ta₂O₅ system and then subjecting it to a second anodization step could stabilize the dielectric against some of the damage done by the deposition of the solid cathode. Some oxygen is driven into the tantalum metal from dielectric during this heat treatment stabilizing that interface during subsequent high temperature exposure (either during MnO₂ deposition or exposure to temperature/voltage during capacitor operation). The second dielectric anodization step replaces oxygen lost from the dielectric during the heat treatment (reducing the number of oxygen vacancies). Later, Gerhart Klein at P. R. Mallory and Co. extended the fundamental understanding of what was occurring in the dielectric and tantalum metal during heat treatment. ²

A key to increasing capacitance or ‘downsizing’ (or putting the same capacitance in a smaller package size) is increasing the surface area of the tantalum powder. Instead of surface area, powder charge, CV/g or µCoulombs/g is used as measure.

Figure 1: Evolution of Powder Charge in Capacitor-Grade Tantalum Powders

Figure 1 ³ shows the evolution of powder charge. It has increased more than an order of magnitude since the 1960s which has resulted in tantalum capacitors with smaller case sizes and higher capacitances. Powder purity has also improved which contributes to higher quality dielectrics and more reliable capacitors. In addition, careful control of tantalum particle size has allowed more surface area to be retained as anodization voltage increases resulting in higher capacitance ratings at higher rated voltages.

To use the powder in a tantalum capacitor, the particles must be sintered together to form a porous anode that can be impregnated with the cathode material. Tantalum forms a passivating film of a few nanometers of Ta₂O₅ on the surface. During the anode sintering process in vacuum, oxygen from this film is driven into the tantalum metal and after the anode is cooled and re-exposed to air a highly exothermic reaction forms a new passivating film on the surface of the sintered particles. Control of the oxygen content is critical as high oxygen can lead to crystallization during subsequent dielectric formation and uncontrolled passivation can lead to formation of a crystalline passivation layer and potentially powder burning. This becomes especially problematic as powder charge increases. Dr. Yuri Freeman studied this issue while he was at Vishay (who had acquired Sprague Electric’s tantalum capacitor business in 1992) along with his collaborators at Tel Aviv University ⁴. Dr. Freeman invented ways of practically applying this knowledge to improve tantalum anodes made from high charge tantalum powders ⁵. His book on tantalum and niobium capacitors is the definitive reference in the field ⁶, and the first edition was awarded the first Ekeberg Prize from the T.I.C. in 2018.

The quality of the dielectric is also determined by the chemistry and process of anodization of tantalum, to tantalum pentoxide (or formation as it is often referred to). The reaction can be written:

5Ta + 5H₂0 → Ta₂0₅ + 5H₂                     [1]

While the reaction is simple, there are many pitfalls in forming a high-quality dielectric. The best quality dielectrics formed by anodization are amorphous and the wrong process conditions can lead to formation of crystals in the film which leads to either high leakage current or eventual catastrophic failure under voltage and temperature. Fundamental work in the 1950s by Vermilyea at General Electric established the key role of the purity of the tantalum starting material and its effect on crystallization of the amorphous dielectric⁷. Early investigations showed that, if formed in a phosphoric acid containing electrolyte, phosphorous species were incorporated into the dielectric which improved the performance of tantalum capacitors under voltage and temperature conditions. Randall, et al. at Sprague Electric carried out some of the early fundamental studies on the anodization of tantalum in phosphoric acid electrolytes ⁸, and later work by Sloppy and Dickey at Penn State provided more insight into the role of phosphorous incorporation into the dielectric ⁹.

During the first 35 years of solid tantalum capacitor commercialization, anodization electrolytics of phosphoric acid either in water or water-ethylene glycol mixtures were adequate to form high quality tantalum pentoxide dielectric films. These systems began to show their limitations as higher charge powders were deployed, and ethylene glycol emissions from formation tanks were becoming an issue. Brian Melody and his group at KEMET Electronics were pioneers in developing new electrolytes such as near neutral electrolytes for dielectric formation on high charge powders ¹⁰, replacing ethylene glycol with polyethylene glycols, and new electrolyte solvents such as tetragylme that allowed very thick (high voltage capability) dielectric films to be formed ¹¹.

Manufacturers of tantalum capacitors made significant progress during the 1960s and 1970s in electrical testing and screening of tantalum capacitors. Early on the discovery that tantalum capacitors had not a constant failure rate but a declining failure rate upon application of voltage and temperature led to ‘grading’ of tantalum capacitors in the factory before shipment and established reliability levels for high reliability applications like military and aerospace ¹².

Entering the Digital Age

The 1960s and 70s saw the continuing growth of the electronics industry and electronic component sales. A step change, however, occurred in the 1980s and 1990s first with the introduction of the personal computer and then the mobile phone. Electronic devices were penetrating more and more into our daily lives and production needed to ramp up and costs needed to come down. This led to the development of surface mount components which increased throughput and decreased costs of circuit board assembly. Tantalum capacitor manufacturers followed this trend. Figure 2 shows a cut-away view of a surface mount solid tantalum capacitor. The manufacture of the silver coated pellet is like the through hole version, but instead of being connected to wire leads, the tantalum positive lead is welded to one side of a lead frame and the negative silver side is attached to the other side of the lead frame with a conductive silver adhesive. The assembly is over-molded with an epoxy compound and the leads are then bent around the underside of the case.

Figure 2: Cut Away View of Surface Mount Solid Tantalum Capacitor

Other forms of surface mount solid tantalum exist such as conformally coated manufactured by Vishay¹³ and TACmicrochips® ¹⁴ manufactured by AVX, but the construction shown in Figure 2 is the dominant type in use today. Surface mount solid tantalum chips account for about 90% of the tantalum market today ¹⁵. Thru-hole components are still used in some legacy and specialized applications.

The transition from analog to digital electronics resulted in more stringent requirements for powering devices especially as clock speed is increased and voltage levels decreased. Clock speed increases meant higher currents needed to be supplied to the chip and voltage level decreases meant that margin between logic high and logic low was smaller and voltage needed tighter regulation. One of the main uses of solid tantalum capacitors is as a decoupling capacitor in power delivery networks, and to be effective they need to deliver their energy quickly. The rate of energy delivery (power) is limited by the capacitor’s Equivalent Series Resistance (ESR). The semiconductive MnO₂ used as the cathode and responsible to the self-healing reliability mechanism in the solid tantalum capacitor has a resistivity of about 0.1-1Ω-cm which is orders of magnitude less conductive than metals such as silver or copper.

Starting in the 1970s concerted efforts were made to reduce the ESR of solid tantalum capacitors. Piper ¹⁶ proposed connecting several tantalum capacitors in parallel inside the molded case. This was commercialized by KEMET Electronics in 1997 (Figure 3) as the Multiple Anode Tantalum (MAT) and resulted in availability of solid tantalum capacitors with ESR below 30mΩ.

Figure 3: Multiple Anode Tantalum (MAT) Chip

Due to having to place multiple capacitors and having to discard parts if one of the capacitors was defective, the MAT solution was expensive. Manufacturers such as KEMET and AVX developed alternatives such as the fluted anode design shown in Figure 4 which did not give as much ESR reduction as the MAT but was good enough for certain applications¹⁷.

Figure 4: Solid Tantalum Capacitor with Fluted Anode Construction

Starting in the 1950s there was a concerted effort to characterize the contributions of the different material components of the solid tantalum capacitor to its ESR¹⁸. The conclusion was that in the frequency range of 10s to 100s of kHz (the primary range of interest for capacitive filtering/bulk decoupling in switching power supplies) the resistance of the MnO₂ material dominated. There was significant work during this period to improve the conductivity of MnO₂ recognizing the β-MnO₂ form was the most conductive¹⁹. Combined with the constructions advances shown in Figures 3 and 4 this resulted in large case size (7.3x4.3mm) solid tantalum capacitors with ESRs in the 10’s of milliohms. The electrical properties of solid tantalum capacitors were eventually translated into Spice models to make it convenient for electrical design engineers to model their behavior in their circuits. ²⁰.

Digital circuit speeds continued to increase and voltage levels continued to drop. By the mid 1990s, it was apparent that even the most advanced Tantalum capacitors with manganese dioxide cathodes were struggling to meet circuit design requirements. Multi-layer ceramic capacitors (MLCC) were begining to challenge the dominance of Tantalum capacitors in portable devices especially as the expensive silver/palladium electrodes were being replaced by much less expensive nickel (base metal electrode technology). Tantalum capacitor use would probably have significantly declined, except for a radical breakthrough in conductive materials that occured in the 1970s and was applied to solid Tantalum capacitors starting in the 1990s. We'll cover that breakthrough and how it completely reshaped the Tantalum capacitor landscape in the next post.

References