S
P
O
T
L
I
G
H
T
O
N
L
I
T
H
O
G
R
A
P
H
The Problem with PSM The phase shifting mask, when first introduced more than 20 years ago, was truly an idea before its time. While the resolution and depth of focus advantages of phase shift masks (PSMs) were evident from the beginning, the classical lithography scaling of lower wavelengths and higher numerical apertures proved more practical and powerful. Twenty years later, however, things look very different. Lower wavelengths below 193 nm present a host of practical difficulties, from the unavailability of high quality optical materials to the difficulty of making a reasonably transparent photoresist. Simple geometry prevents the numerical aperture from increasing forever, though the introduction of immersion lithography technology provides some room to squeeze more from higher angles. Today, the potential of PSM seems more compelling than ever. First, not all phase shift masks are created equal. Attenuated PSM (also called the embedded phase shift mask, or EPSM) has been widely adopted for contact and via printing, and is becoming fairly mainstream for other critical lithography layers as well. This type of PSM—generally called a “weak” shifter—provides only a portion of the full resolution and depth of focus potential of PSM. Its great advantage is the simplicity and low cost of replacing chrome on glass (COG) masks, the non-phase shift alternative. Essentially, an existing design based on COG can be converted to an EPSM by simply recalibrating the optical proximity correction (OPC) models used to apply OPC to the design. Mask manufacturing, while certainly more difficult than COG, is not dramatically different (the chrome is replaced by a more complex absorber such as molybdenum silicon) and only somewhat more costly. In short, the transition from COG to EPSM-based lithography presents no major hurdles.
30
Winter 2005
Yield Management Solutions
“Strong” phase shift masks, on the other hand, have seen only limited use in manufacChris A. Mack, KLA-Tencor turing, despite their potential for nearly doubling resolution and extending depth-of-focus (DOF) even more. Strong shifters, exemplified by the alternating PSM (also called the Levenson PSM after Marc Levenson, the first to publish on the use of PSM in lithography), involves a much more difficult mask making process. Strong shifters invariably use etched quartz to create the phase shift and require two mask writing steps (including an alignment in between). While the much greater complexity and expense of making alternating PSMs have certainly delayed their introduction, it is not mask manufacturing that is the biggest problem, but mask design. Alternating PSM works by shifting the phase of the clear region to one side of a small line by 180º relative to the phase of the light coming from the other side of the line. While simple in concept, attempting to phase shift an arbitrary layout of lines will invariably lead to phase conflicts. There are two basics types of phase conflicts, as seen in Figure 1: no phase shift where you want it, and a phase shift where you don’t want it. The first type (Figure 1a) results in a lack of phase shift across a critical feature when there is an odd wrapping of phase assignments. This “non-shifted” feature will not properly print. The second type (Figure 1b) is also called the termination problem since it usually occurs at the end of a line. Alternating phase across each side of a line will result in those two phases meeting at the line end. Whenever two opposing phases meet, a dark interference line is created, causing an unwanted resist line to print. (In fact, when printing small lines, the use of chrome is almost superfluous—it
Y
S
P
O
T
L
I
G
H
T
O
is the 0° to 180° phase transition that causes the dark lines to print.) The phase conflict problem does not have an obvious solution. Taking an arbitrary layout and coloring the clear regions with two phase colors will lead to numerous conflicts in all but a few special cases. One solution, though not yet proven, is to force the original layout to be “phase friendly”, a layout where no phase conflicts can occur. It is unclear whether such a solution can in fact be developed and if so, what the design trade-offs will be. The alternative is costly but more practical: use two exposures from two masks. The phase shift mask is used to define the critical features using a darkfield background to avoid phase conflicts. A second exposure uses a brightfield mask that prints the non-critical feaa)
L
N
I
T
H
O
G
R
A
P
H
Y
tures as well as the open areas not exposed with the initial darkfield mask. Besides the obvious cost disadvantage due to the much lower lithography tool throughput, this double exposure approach also tends to have less than optimum transistor density (you can print small lines, but you cannot put them too close to each other). The double exposure PSM process has been used successfully for polysilicon gate layers on microprocessors, where gate CD control has a profound effect on the value of the device. Further refinements could make this PSM approach more practical for other critical levels and for other types of products. Today, phase shifting masks are already having an important and profound impact on lithographic capabilities, but there is still much unexplored potential with PSM applications. b)
180 º
180º
0º
180º
0º
180º
Figure 1. Types of phase conflicts: (a) no phase shift across a critical pattern, and (b) the phase termination problem producing an unwanted phase edge.
Winter 2005
www.kla-tencor.com/magazine
31
The 16th Annual IEEE/SEMI
Advanced Semiconductor Manufacturing Conference and Workshop April 11–12, 2005 International Conference Center Munich, Germany www.semi.org/asmc
ASMC 2005
ASMC is the world’s premier technical conference addressing the challenges and trends of semiconductor manufacturing. From 300 mm to nanoscale devices, from factory automation to yield enhancement, advance your semiconductor manufacturing knowledge with real solutions— direct from the fab. Register for ASMC 2005 today!
Sponsors:
Media Sponsors: