Active solid-state devices (e.g. – transistors – solid-state diode – Field effect device – Having insulated electrode
Reexamination Certificate
1999-11-15
2001-04-24
Loke, Steven (Department: 2811)
Active solid-state devices (e.g., transistors, solid-state diode
Field effect device
Having insulated electrode
C257S306000, C438S253000
Reexamination Certificate
active
06222221
ABSTRACT:
BACKGROUND OF THE INVENTION
(1) Field of the Invention
The invention relates to the general field of electrical capacitors, more particularly to low voltage-coefficient capacitors in integrated circuits.
(2) Description of the Prior Art
As is well known, the capacitance of a parallel plate capacitor is normally a function only of the area of the electrodes, the dielectric constant of the dielectric and the thickness of the dielectric. This assumes that both electrodes are metallic conductors. If one of the electrodes is a semiconductor, a depletion region, extending into the semiconductor, is formed at the semiconductor-dielectric interface, even in the absence of any externally applied voltage bias, thereby rendering the capacitance of such a structure somewhat lower than if both electrodes had been metallic.
An important consequence of this phenomenon is that when voltage is applied across the above described device's plates, the depletion layer in the semiconductor grows or shrinks, depending on the polarity, thereby reducing or increasing the measured capacitance. In other words, such a device has a high voltage coefficient, in some cases as high as 120 ppm/volt. In many integrated circuits this is not a problem but in many analog circuits (such as analog to digital converters, for example) this cannot be tolerated.
During the manufacture of integrated circuits where capacitors must be included it is obviously an advantage if such capacitors do not take up real estate needed for other devices, notably transistors. Fortunately, the areas of field oxide that are used to electrically isolate active devices from one another, are available for this. It is also particularly advantageous if the capacitors can be formed without introducing any additional steps into an existing process, other than the modification of the etch masks. With this in mind, it has been common practice to form capacitors in integrated circuits by sandwiching a layer of the silicon oxide known as IPO (or inter-poly oxide) between two layers of polysilicon.
In earlier processes such as the standard 0.5 micron mixed-mode process, the two polysilicon layers used for the capacitors were sufficiently heavily doped so that they were electrically degenerate (that is, they exhibited metallic conductivity) and any depletion layer formed at the silicon-dielectric interface would be negligibly thin. More recently, with the development of the 0.35 micron process, changes in processing parameters make the polysilicon layers less heavily doped. This is unimportant for the first layer of polysilicon (polyl) because a layer of tungsten silicide, which is a metallic conductor, is formed on polyl so it has a depletion free interface with the IPO.
There is no corresponding layer of tungsten silicide between poly2 and the IPO, so a depletion layer forms within the poly2 and the resulting capacitor has a high voltage coefficient. Referring now to
FIG. 1
, we show a schematic cross-section of such a capacitor. Bottom electrode
3
, comprising poly1, has been formed on an area
2
of field oxide on silicon substrate
1
. Dielectric layer
4
, comprising IPO, lies on top of
3
and layer
5
, comprising poly2, forms the top electrode.
FIG. 2
shows a schematic plot of the normalized capacitance value of the device of
FIG. 1
as a function of applied voltage V. Typically, V covers the range of from −5 to +5 volts, with V applied to the poly2 layer and the poly1 layer being grounded. The voltage coefficient of capacitance, as defined by the slope of curve
21
is generally about −120 ppm/volt.
A number of prior art references describe capacitors for incorporation within integrated circuits but most do not address the voltage coefficient issue. Boerstler et al. (U.S. Pat. No. 5,389,832 February 1995) describe capacitors formed from forward biassed diodes. Sundaresan (U.S. Pat. No. 5,016,070 May 1991) describes a stacked CMOS SRAM with cross-coupled capacitors. Sato (U.S. Pat. No. 5,521,111 May 1996) teaches use of a trench-stacked capacitor.
Himes et al. (U.S. Pat. No. 4,731,696 March 1988) address the voltage coefficient issue and disclose a three plate capacitor structure wherein two layers of dielectric are sandwiched between three conductive plates, the center plate being the semiconductor. Voltage is applied between the center plate and the two outer plates (which are connected to each other). Since the upper and lower capacitors always see opposed electric vectors the voltage coefficients in the upper and lower dielectric layers will always carry opposite signs and will thus tend to cancel one another.
While this structure should exhibit low voltage coefficient, it requires three separate deposition steps for the electrodes and two separately deposited dielectric layers. The latter limitation can lead to serious problems in the absence of precise thickness and process control during deposition of the two dielectric layers. It also means that the standard manufacturing process will have to be modified if this structure is to be made part of an integrated circuit.
SUMMARY OF THE INVENTION
It has been an object of the present invention to provide a capacitor that has a very low voltage-coefficient.
A further object of the invention has been that said capacitor be part of an integrated circuit.
Another object of the invention has been to provide a process for manufacturing such a capacitor.
A still further object of the invention has been that said process be part of a standard process already in place for manufacturing integrated circuits and that no additional steps, other than modified etch patterns, need be introduced into said standard process.
These objects have been achieved by forming two parallel plate capacitors side by side on an area of field oxide and then cross-connecting them. The bottom plate of one of the capacitors is connected to the top plate of the other capacitor, and vice versa. This arrangement causes the two capacitors to be polarized in opposite directions at all times so that the individual voltage coefficients cancel each other and give the combined structure a value that is 2 ppm/V.
REFERENCES:
patent: 4731696 (1988-03-01), Himes et al.
patent: 4878151 (1989-10-01), Gallichio
patent: 5016070 (1991-05-01), Sundaresan
patent: 5208597 (1993-05-01), Early et al.
patent: 5347152 (1994-09-01), Sundaresan
patent: 5389832 (1995-02-01), Boerstler et al.
patent: 5470775 (1995-11-01), Nariani
patent: 5521111 (1996-05-01), Sato
Chen Ming-Jer
Hou Chin-Shan
Ackerman Stephen B.
Loke Steven
Owens Douglas W.
Saile George O.
Taiwan Semiconductor Manufacturing Company
LandOfFree
Cross-coupled capacitors for improved voltage coefficient does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Cross-coupled capacitors for improved voltage coefficient, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Cross-coupled capacitors for improved voltage coefficient will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2554055