The invention is generally related to the field of forming capacitors in semiconductor devices and more specifically to forming high density capacitors at the top metal interconnect level.
As semiconductor technology continues to scale, the supply or operating voltage of the integrated circuit becomes lower and lower. The nominal supply voltage has decreased from 5V to 3.3V to 1.8V and below. Transistors with operating voltages of 1.1V are currently being developed.
As the supply voltage decreases it becomes increasingly important to limit the voltage swing on the supply voltage lines. This is due to the fact that smaller voltage swings can cause unacceptable amounts or current leakage and even unintentionally switch the state of the transistor. Voltage swing may be minimized by providing capacitance on the power supply. Typically this is accomplished with off-chip decoupling capacitors. As the amount of voltage swing that can be tolerated is reduced, more and more decoupling capacitance is required.
The invention is an on-chip decoupling capacitor and method of fabrication. The decoupling capacitor is integrated at the top metal interconnect level and may be implemented with only one additional masking layer.
An advantage of the invention is providing on-chip decoupling capacitance.
This and other advantages will be apparent to those of ordinary skill in the art having reference to the specification in conjunction with the drawings.
In the drawings:
The invention will now be described in conjunction with copper damascene process utilizing an aluminum cap layer. Those of ordinary skill in the art will appreciate that the benefits of the invention can be applied to other metal interconnect processes.
As semiconductor devices continue to scale, the decoupling capacitance requirements increase significantly. The parasitic resistance of off-chip capacitance can result in performance penalties. Placing the decoupling capacitors on-chip can reduce or even avoid these performance penalties. However, integrating the decoupling capacitors on-chip can cause other concerns. For example, using a gate oxide capacitor (in which a MOSFET gate oxide layer is also used as the capacitor dielectric) consumes active area. There are also leakage concerns with gate oxide capacitors. Adding a capacitor between contact and M1 (the first level of metal interconnect) adds a mask, may require routing restrictions above the capacitor and causes planarity and thermal budget concerns. Adding a capacitor between metal interconnect lines also causes planarity and thermal budget concerns.
To alleviate some of these concerns, the preferred embodiment of the invention incorporates a high density capacitor at the top metal interconnect level. The top metal interconnect level is generally used for routing power and ground lines. As such, the interconnect routing is not as dense as the lower metal interconnect levels and there is more space available for forming the decoupling capacitors. Furthermore, since it is the uppermost interconnect level. planarity is not as much of a concern. Thermal budget is also less of a concern because there are fewer remaining steps and those steps remaining are generally performed at lower temperatures.
A preferred embodiment of the invention is shown in
Top metal interconnect level 104 is formed over semiconductor body 102. Semiconductor body 102 comprises a semiconductor substrate, transistors and other devices, as well as other metal interconnect levels. Only the top metal interconnect level 104 is shown for simplicity. In this embodiment, the top metal interconnect level 104 comprises copper interconnect lines 104a, 104b with appropriate barrier layers.
High density capacitor 106 comprises a bottom electrode 108, capacitor dielectric 110, and top electrode 112. The top 112 and bottom 108 electrodes comprise an electrically conductive material. In the preferred embodiment, TaN is used. TaN is often used as a barrier in copper interconnects and offers good compatibility with copper processes. Other electrically conductive materials, such as TiN, Ir, Ru, Ta and sandwiches/multi-layer combinations thereof may alternatively be used. Although TaN is used for both he top and bottom electrodes in the preferred embodiment, different materials may in fact be used for the top and bottom electrodes. The capacitor dielectric 110 preferably comprises a high dielectric constant dielectric. In the preferred embodiment, tantalum-oxide is used. Other high-k dielectrics such as hafnium oxide may alternatively be used. While high-k dielectrics are preferred, less high-k materials such as SiN can alternatively be used.
The high density capacitor 106 is located over first copper interconnect line 104a such that first copper interconnect line 104a is electrically connected to the bottom electrode 108. Connection to the top electrode 112 is made via aluminum cap layer 118. A portion of the aluminum cap layer 118 is electrically connected between the top electrode 112 and a second copper interconnect line 104b. So, for example, first copper interconnect line 104a may be designed as power supply line for which capacitive decoupling is desired and second copper interconnect line 104b may be designed as a ground line.
A method for fabricating the decoupling capacitor of
Referring to
Next, a capacitor dielectric 110 is formed over bottom electrode layer 108. Preferably, capacitor dielectric 110 comprises a high dielectric constant material such as tantalum-oxide. A high dielectric constant material allows for a larger capacitance value to be formed in a smaller area. Other high-k dielectric materials such as hafnium oxide or less high-k dielectrics such as SiN may alternatively be used. In the preferred embodiment, a layer of tantalum-oxide 130 is deposited over bottom electrode material 108, as shown in
Still referring to
Next, a pattern 132 is formed over top electrode material 112, as shown in
Referring to
Referring to
Processing then continues to package the semiconductor devices. During packaging, ball bonds and other bonding methods are used to connect the bondpads of a semiconductor device to the external pins of the integrated circuit.
While this invention has been described with reference to illustrative embodiments, this description is not intended to be construed in a limiting sense. Various modifications and combinations of the illustrative embodiments, as well as other embodiments of the invention, will be apparent to persons skilled in the art upon reference to the description. It is therefore intended that the appended claims encompass any such modifications or embodiments.
The following co-pending patent applications are related and hereby incorporated by reference: U.S. patent application Ser. No. (TI-35260), ______ filed to Burke et al. U S. patent application Ser. No. (TI-36382), ______ filed to Papa Rao et al.
Number | Date | Country | |
---|---|---|---|
Parent | 10696816 | Oct 2003 | US |
Child | 11862632 | Sep 2007 | US |