Solid-Electrolyte Niobium Capacitors Exhibit Similar Performance to Tantalum
For a quarter of a century, niobium was tested as a capacitor substrate, with disappointing results that led many to believe it never would provide equivalent performance to tantalum. However, in September 2001, we achieved a breakthrough, sampling the industry's first solid-electrolytic niobium capacitors — with a conventional MnO2 cathode — that can serve as less expensive, plentifully available replacements for tantalum devices.
The first experiments in developing niobium capacitors with properties similar or equal to tantalum capacitors began in the United States and former USSR in the mid-1970s. The key stumbling block in developing an effective niobium capacitor was that thermal and electrical shock damaged the dielectric during manufacture, destroying the capacitor.
Our breakthrough in developing a niobium device with performance comparable to tantalum is the result of a unique, patent-pending set of techniques that improve the temperature stability of the niobium dielectric while circumventing the processes that cause damage. The capacitors resulting from this new niobium manufacturing technology offer the same operating temperature range, dc leakage, and ESR as tantalum capacitors (Table 1).
This breakthrough is particularly significant because niobium is the only viable substitute for tantalum. It also happens to be available in large reserves and is one-tenth as expensive (Table 2). Tantalum has become scarce and so costly that many designers are reluctant to use it. Niobium capacitors offering equivalent performance to tantalum will provide electronics manufacturers with a plentiful and affordable supply of high-performance components, while removing a persistent source of uncertainty from their supply chains.
Like niobium, ceramic and aluminum are also possible substitutes for tantalum; however, some factors limit their usefulness in this respect. Inherent in ceramic capacitors is a microphonic effect that creates noise in electrical circuits. This is a natural physical phenomenon. On the other hand, very steep temperature dependence of ac parameters, including equivalent series resistance (ESR) and impedance, typify aluminum capacitors.
In short, the behavior of ceramics and aluminum when used as anode material is very different from that of tantalum. By contrast, niobium and tantalum are next to each other within the same column on the periodic table. Indeed, they share so many like characteristics — including a similar crystalline structure — that many thought they were the same element (Table 3, on page 69). Both exist in coarse granitic rocks (pegmatites) rich in lithium and phosphorus minerals. You can find niobium in the mineral columbite, concentrated at the edges of the pegmatite, and tantalum in the mineral tantalite, enriched in the core.
There are many myths surrounding tantalum and niobium (see sidebar). For example, since niobium oxide (Nb2O5) has a 50% higher dielectric constant than tantalum oxide, this means that devices built on niobium feature a 50% greater capacitance. However, don't overlook the ratio of capacitance per unit of surface (C/S), which is identical between the two substances. Because the dielectric thickness of niobium oxide also is 50% greater for any given voltage, its C/S ratio remains the same as that of tantalum: 1.7V/A (Table 4).
A similar misperception exists with respect to niobium's density. Since the density of niobium is half that of tantalum, an assumption exists that a niobium capacitor with equivalent performance to tantalum will require twice as much raw material. This is like comparing two parcels of real estate on the basis of what they weigh rather than according to their respective square footage. As with real estate, it's not the weight that counts but the available surface area — and for the same volumes of niobium and tantalum, the available surface area is identical. So working with niobium, you can produce twice as many capacitors from the same amount of raw material as measured by weight. Furthermore, these niobium capacitors are lighter and cheaper than their tantalum equivalents. As a result, it will cost manufacturers less to produce high-performance capacitors that reduce the weight of end products.
One of the disadvantages attributed to niobium in the past is its high capacitance dependence on bias and temperature. Fig. 1 shows that high capacitance dependence on bias is characteristic of wet niobium capacitors, while in solid niobium capacitors it's in the range of ± 10%, similar to tantalum.
The capacitance dependence on temperature varies according to case size. The increased capacitance dependence on temperature that becomes evident in larger case sizes is the result of the morphology of the niobium powder now in use. In contrast to tantalum, which features larger and more open pores, niobium powder exhibits quite fine pores. When there is a smaller diameter of the pore, it's more difficult to impregnate it with a manganese dioxide cathode. This effect is magnified in the larger packages (Fig. 2). As the production of good quality, high-CV niobium powder continues to develop, the resemblance between niobium and tantalum powder morphology will continue to improve, as will capacitance change values for niobium devices of all sizes.
One of the benefits niobium capacitors bring to manufacturers is a lower ignition failure mode than tantalum capacitors exhibit. Although the specifics of this behavior have yet to be quantified, manufacturers can reasonably expect that the use of niobium capacitors will provide a greater measure of safety, in addition to better availability and lower cost compared to devices built on tantalum.
The Myth Behind Tantalum and Niobium
The names tantalum and niobium derive from Greek Mythology, wherein which King Tantalus angers the gods and is condemned to live with eternal thirst and hunger, forever tantalized by water and fruit just beyond his grasp. His daughter Niob, too, angers the gods and is turned to stone. With this history, perhaps it was inevitable that the relationship between tantalum and niobium would continue to be surrounded by various myths a few thousand years later.
For more information on this article, CIRCLE 338 on Reader Service Card