Forming attached features on a semiconductor substrate

Details Forming attached features on a semiconductor substrate Forming attached features on a semiconductor substrate

1999-11-09

2001-07-24

Smith, Matthew (Department: 2825)

Active solid-state devices (e.g., transistors, solid-state diode

Physical configuration of semiconductor

Mesa structure

C156S922000, C438S690000

Reexamination Certificate

active

06265757

ABSTRACT:

FIELD OF THE INVENTION
This invention relates generally to the fabrication of semiconductor devices and more particularly to the fabrication of a smaller, shallow feature adjacent to a deeper feature.
BACKGROUND OF THE INVENTION
Liquid phase etching of semiconductor substrates provides one of the core technologies of the microelectronics industry. As the demands for smaller devices and more densely packed designs have increased, new methods of liquid phase etching have been developed to meet the technological challenge of manufacturing these smaller devices.
One feature of certain liquid phase etching processes is that they produce different etch rates along different crytallographic directions in the semiconductor substrate. For example, in an etch process with ethylene diamine pyrocatechol (EDP), an etching process which yields an etch rate of 60-70 microns per hour on an Si(100) surface may only etch an Si(111) surface at 1-3 microns per hour. Other orientation dependent etchants, such as KOH, also exhibit a disparity in etch rates for silicon substrates. The discussion below will focus on the use of orientation dependent etchants on silicon substrates. Those skilled in the art, however, will understand that this background discussion can be broadly applied to substrates of other materials which exhibit orientation dependent etching.
Due to the anisotropy of etch rate with respect to crystal direction, orientation dependent etchants (ODEs) are well-suited for creating pyramidal etch features on (100) surfaces. This is particularly true when an etch feature is created by exposing a rectangular portion of a (100) surface to an orientation dependent etchant. If the sides of the rectangle are properly aligned with the <110> directions, the orientation dependent etch will produce an etch pit with (111) sidewalls. The angle between the (111) surfaces and the (100) surface is 54.7 degrees. Based on this relationship, if the etch is allowed to proceed for a sufficient length of time, an etch feature will be produced which is defined by the intersection of (111) surfaces. If such a feature is etched to completion, the depth of the feature will be one half of the product of the smallest side of the original exposed rectangle and the tangent of 54.7. This calculated depth excludes any additional depth from overetching of the feature. Such overetching usually results in only a modest increase in depth, however, due to the greatly reduced etch rate for ODEs on (111) surfaces. This allows for a great deal of process flexibility, as an etch process may be continued for a significant length of time without appreciable overetching of a feature.
Due to the self-limiting nature of this type of etch process, features of disparate sizes may often be etched in a single step. As an example, it may be desirable to etch a small feature and a large feature during the same step. If both features are exposed to an orientation dependent etchant, less time will be required to form the limiting pyramidal shape of intersecting (111) surfaces in the small feature. Because of the slow etch rate on (111) surfaces, however, additional overetch will lead to only a modest amount of additional etching in the small feature. As a result, allowing the etch to proceed for a longer time in order to create a large, deeper feature will not lead to an appreciable change in the smaller etch feature. Thus, smaller and larger features can be created during a single etch step.
While this technique for creating smaller and larger features during a single etch step may be applicable in many situations, problems arise when it is desirable to etch a small, shallow feature, such as a notch, which is attached to a larger, deeper feature.
FIGS. 1-3
schematically shows the difference between features which are near to each other and features which are attached. In
FIG. 1
, features
20
and
22
are only near one another. The height of barrier
24
between the features is substantially the same as the height of the substrate material
28
. On the other hand, features
30
and
32
in
FIG. 2
are attached. Although a barrier
34
is still present in the transition region between features
30
and
32
, the height of barrier
34
is well below the height of the substrate material
38
. Features
40
and
42
in
FIG. 3
are also attached, although now no barrier is present. Instead, features
40
and
42
are joined via transition region
44
.
For example, it may be desirable to control the shape and the depth of both a large feature and an attached small feature with a depth profile like the one shown in FIG.
3
. The large feature and the attached small feature cannot be etched in the same step without losing control over the shape of at least one of these features. Because the features are attached, (111) surfaces will be unable to form in the transition region between the large feature and the attached small feature. Instead, a variety of surfaces will be exposed that etch rapidly in an ODE. This will lead to fast etching in the transition region and at the bottom of the small feature, resulting in a deeper small feature than desired and loss of a great deal of material from the transition region and the sidewall of the large feature where it meets the transition region.
One alternative to etching the large feature and attached small feature in a single process step is to etch the large feature during a first etch step and then to etch the entire attached small feature, including the transition region, during a second etch step. Splitting the etch into these two steps provides some additional control over the process, but difficulties still remain. A variety of crystallographic faces are present in the transition region between the attached small feature and the large feature. If the transition region is exposed to an ODE, the etch rate in the transition region will be faster than the etch rate on a (111) surface. As a result, it is difficult to control the etch depth in both the transition region and the attached small feature at the same time.
In U.S. Pat. No. 4,601,777, a method for creating a small channel near a large cavity is disclosed for use in making ink jet printer heads. The method involves a two step etching process. In the first step, the large cavity and the small channels are created using an anisotropic etch, such as KOH, but a barrier is left between the cavity and the channel. In the second step, the barrier is removed by either dicing, a mechanical means, or by an isotropic etch followed by an anisotropic, or orientation dependent, etch. The isotropic etch used in the invention is a mixture of HF, HNO
8
, and C
2
H
4
O
2
. This disclosure also notes the unsuitability of using only an orientation dependent etch to remove the barrier between the cavity and the channel.
U.S. Pat. No. 4,810,557 discloses a method for creating a tandem V-groove where a portion of the groove is shallow and narrow relative to the remainder of the groove. In this method, the attached shallow groove and deep groove are created by applying two etch masks. The pattern in the first mask determines the location of the grooves. The second mask is then deposited to prevent etching in the shallow groove during the initial stages of the etch.
U.S. Pat. No. 4,863,560 discloses a method for creating both large, coarse features and small, fine features without having to resort to lithography steps in between etching steps. Several etch masks are deposited prior to etching, with the mask for the largest, coarsest features at the top of the stack of masks. The masks are patterned as they are deposited without etching the underlying substrate. After all of the etch masks are formed, the first etch process is carried out. After the first etch process the coarse mask is removed, leaving behind a second etch mask. This process can be repeated to achieve finer and finer control over the etch features created. U.S. Pat. Nos. 5,131,978 and 5,277,755 disclose other methods for using multiple etch masks to create attached small and large features

Affiliated with

Also associated with

Rating

Forming attached features on a semiconductor substrate does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Forming attached features on a semiconductor substrate, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Forming attached features on a semiconductor substrate will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2525055