Computer-aided design and analysis of circuits and semiconductor – Nanotechnology related integrated circuit design
Reexamination Certificate
2001-03-29
2003-09-16
Siek, Vuthe (Department: 2825)
Computer-aided design and analysis of circuits and semiconductor
Nanotechnology related integrated circuit design
C716S030000, C716S030000
Reexamination Certificate
active
06622288
ABSTRACT:
BACKGROUND OF THE INVENTION
1. The Field of the Invention
This invention relates to the field of integrated circuit manufacturing. In particular, this invention relates to adjusting design layouts to eliminate phase-shift conflicts for shifters on masks used to fabricate integrated circuits.
2. Description of Related Art
Conventional integrated circuit (IC) fabrication involves many steps in common with other processes that impose physical structures in a layer on a substrate, such as laying ink in patterns on a page, or laying chrome in patterns on a quartz substrate. Some of the important steps viewed at a high level are depicted in FIG. 
1
.
In step 
110
, engineers use a functional computer aided design (CAD) process, to create a schematic design, such as a schematic circuit design consisting of individual devices coupled together to perform a certain function or set of functions. The schematic design 
115
 is translated into a representation of the actual physical arrangement of materials upon completion, called a design layout 
125
, with a physical CAD process 
120
. If multiple layers are involved, as is typical for an IC, a design layout is produced for each layer, e.g., design layouts 
125
a
, 
125
b
, etc. 
FIG. 2
 shows a sample design layout. A fabrication CAD process 
130
 produces one or more fabrication layouts 
135
, such as masks for each design layout 
125
a
. The one or more fabrication layouts 
135
 are then used by a substantiation process 
140
 to actually produce physical features in a layer, called here the printed features layer 
149
.
One recent advance in optical lithography called phase shifting generates features in the printed features layer 
149
 that are smaller than the features on the mask 
135
a 
projected onto the printed features layer 
149
. Such fine features are generated by the destructive interference of light in adjacent separated windows in the mask called shifters. 
FIG. 3
 shows two adjacent shifters, 
310
 and 
312
, in a mask 
300
. The shifters 
310
 and 
312
 are light transmissive areas on the mask separated by an opaque area 
311
 with a width of Wm 
313
 when projected onto the printed features layer 
149
. The projection of Wm onto the printed features layer 
149
 is limited by the resolution of the optical process. However, if the light of a single wavelength passing through one of the shifters, e.g. 
310
, is out of phase (by 180 degrees or &pgr; radians) with the light of the same wavelength passing through the other shifter, e.g. 
312
, then an interference pattern is set up on the printed features layer 
149
 during the substantiation process 
140
. This interference generates a printed feature 
350
 having a width Wp 
353
 that is less than the width Wm 
313
 of the opaque area projected onto the printed features layer 
149
. In other embodiments, the width 
313
 and width 
353
 are much closer and can be equal. In each case, the width 
353
 of the printed feature is less than can be produced by the same optical system without phase shifting.
The use of phase shifting puts extra constraints on the fabrication layouts 
135
, and hence on the design layout, e.g. 
125
a
. These constraints are due to several factors. One factor already illustrated is the need for finding space on the mask, e.g., 
135
a
, for the two shifters, 
310
 and 
312
, as well as for the opaque area 
311
 between them. This precludes the one mask from placing additional features on the printed features layer 
149
 in the region covered by the projection of the two shifters 
310
 and 
312
 and the opaque area 
311
. Another factor is that overlapping or adjacent shifters on a single mask, used, for example, to generate neighboring phase-shifted features, generally do not have different phases. Adjacent shifters with different phases will produce a spurious feature.
Currently, design layouts 
125
 may provide the space needed for placement of phase shifters through design rules, but shifters are actually placed and simultaneously assigned a phase in the conventional fabrication design steps, not shown, in attempts to produce the fabrication layouts. As complex circuits are designed, such as by combining many standard cells of previously designed sub-circuits, shifters of different phases may overlap or become adjacent in the layouts, leading to phase-shift conflicts. It is generally recognized that resolving phase-shift conflicts should be done globally, after the whole circuit is laid out, because swapping the phases of a pair of shifters to resolve one conflict can generate a new conflict with another neighboring feature already located in the design or one added later. The conventional IC design systems try to reassign phases of individual pairs to resolve the conflicts at the end of the design process when all the phase conflicts are apparent. For example, iN-Phase™ software from NUMERICAL TECHNOLOGIES, INC.™ of San Jose, Calif., uses this conventional technique.
For example, 
FIG. 4
 shows a T-junction element 
440
 that is desirably formed with narrow phase-shifted features 
443
, 
442
 and 
444
 as well as with wide non-critical features 
441
 and 
445
. 
FIG. 4A
 shows a pair of shifters 
410
 and 
420
 needed to form the vertical phase-shifted feature 
443
 of element 
440
. 
FIG. 4A
 also shows another shifter 
415
 disposed opposite shifter 
410
 to form the left half 
442
 of the horizontal phase-shifted feature of element 
440
. Similarly, 
FIG. 4A
 also shows a fourth shifter 
425
 disposed opposite shifter 
420
 to form the right half 
444
 of the horizontal phase-shifted feature of element 
440
. Shifters 
415
 and 
425
 are so close that they violate a design rule requiring at least a minimum spacing X between adjacent shifters. That is, separation 
427
 is less than X.
In the conventional fabrication CAD process, not shown, the shifters 
410
, 
420
, 
415
 and 
425
 are placed as shown and assigned phases, but the phase-shift conflict is not addressed until all the elements of the design layout have been accounted for. Then the design rule is applied in which shifters 
415
 and 
425
 are replaced by a single shifter 
430
.
However, there is no assignment of phase for shifter 
430
 that can simultaneously be opposite to the phases assigned to shifters 
410
 and 
420
, because shifters 
410
 and 
420
 are already opposite to each other. Thus such a design has a conflict that cannot be solved by changing the phases assigned to the shifters. Some re-arrangement of shifters or features or both is needed. In this example, however, the feature 
440
 from the physical design layout does not allow shifter 
430
 to be moved and does not allow another shifter to be inserted. Thus the fabrication CAD process 
130
 cannot move or change the shifters enough to resolve the conflict.
When a phase-shift conflict is irresolvable by the fabrication CAD process 
130
, then the physical CAD process 
120
 is run again to move or reshape the features, such as those of element 
440
. Process flow with an irreconcilable phase-shift conflict is represented in 
FIG. 1
, which shows that fabrication layouts 
135
 are produced along the arrow marked “Succeed” if the fabrication CAD process 
130
 succeeds, but that control returns to the physical CAD process 
120
 along the arrow marked “Fail” if the fabrication CAD process 
130
 fails, such as if it fails to resolve all phase conflicts.
While suitable for many purposes, the conventional techniques have some deficiencies. As designs, such as designs for IC circuits, become more complex, the time and effort involved in performing the physical CAD process 
120
 and the fabrication CAD process 
130
 increase dramatically, consuming hours and days. By resolving phase-shift conflicts at the end of this process, circumstances that lead to irresolvable phase-shift conflicts are not discovered until the end of these time consuming processes. The discovery of such irresolvable phase-shift conflicts induces the design engineers to start over at the physical CAD process 
120
. The processes 
120
 and 
130
 are repeated 
Pierrat Christophe
Richardson Kent
Sanie Michael
Wang Yao-Ting
Wu Shao-Po
Bever Hoffman & Harms LLP
Dimyan Magid Y
Harms Jeanette S.
Numerical Technologies Inc.
LandOfFree
Conflict sensitive compaction for resolving phase-shift... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Conflict sensitive compaction for resolving phase-shift..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Conflict sensitive compaction for resolving phase-shift... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3034152