As patterning of implant layers becomes increasingly challenging it is clear that the standard resist/Top Antireflective
Coating (TARC) process may be soon be limited in terms of its ability to meet implant targets at future nodes. A
particularly attractive solution for patterning implant levels is the use of a Developable Bottom Anti-Reflective Coating
(DBARC). Similar to a conventional BARC, a DBARC controls reflectivity from the underlying substrate by absorbing
the incident radiation thereby minimizing detrimental effects of reflected light. However, unlike a conventional Bottom
Anti-Reflective Coating (BARC) which requires a BARC open etch step, the DBARC is developed with the resist in a
single step leaving the substrate ready for implantation. These properties make DBARC very attractive for implant
In this paper, we report on the development of KrF and ArF DBARCs for implant applications. Our primary interest is
in developing solutions for patterning Post-Gate implant levels. We briefly describe our fundamental design concepts
and demonstrate the concepts are robust as we develop ARTM602 DBARC to address the criteria for a production worthy
DBARC. This includes data on EBR performance, drain line compatibility, sublimation and footing coverage over
topography. In terms of lithographic performance, we demonstrate improved capability over the incumbent SLR/TARC
process in many key areas. This includes through pitch performance, process window and profile integrity over
topography for both KrF and ArF DBARC solutions. Several strategies to enhance profile by resist/DBARC matching
are also demonstrated. From a platform robustness standpoint, we show that AR602 DBARC is ready for high volume
manufacturing in terms of batch to batch control and shelf life.
As device scaling continues according to Moore's Law, an ongoing theme in the semiconductor industry is the need for
robust patterning solutions for advanced device manufacture. One particularly attractive solution for implant lithography
is the use of a developable BARC (DBARC) to improve reflection control while still affording an "implant ready"
substrate following development. Going forward, these two features of DBARC technology are key to successful
implant patterning as the industry standard TARC process begins to falter due to poor substrate reflection control leading
to profile degradation, shrinking process windows and poor CDU.
In this paper, we report our progress in the design and development of production worthy DBARCs for implant
lithography. In addition to outlining our general design concepts, we describe our fundamental approach to
characterizing DBARCs and share key performance data showing our DBARC technology is surpassing the capability of
a traditional TARC process for both KrF and ArF implant applications. Key performance metrics include CD swing, CD
control over varying oxide thickness, active to field CD bias and footing over topography.
Developable bottom anti-reflective coating (DBARC) technology holds promise in two main areas of lithography. The
first application of DBARC is in implant lithography where patterning implant levels would greatly benefit from
improved reflection control such as provided by a conventional BARC. However, implant layers cannot withstand
BARC open etch thereby making DBARC an attractive solution as the resist and DBARC are simultaneously dissolved
during the development step leaving the underlying substrate ready for implantation. In comparison to current implant
processes with top anti-reflective coatings (TARC), DBARCs are anticipated to offer improvements in reflection control
which would translate to improved CDU and increased process window for both KrF and ArF implants. Indeed, this area
has long been considered the ideal insertion point for DBARC technology.
The second area where DBARC technology can make a significant impact is in non-implant lithography. In this large
segment, the ability to replace a conventional BARC with a DBARC affords the device maker the ability to simplify
both lithographic and integration processes. By replacing the BARC with a DBARC, the BARC open etch is negated.
Furthermore, by applying this strategy on multilayer stacks it is possible to greatly simplify the process by avoiding both
CVD steps and pattern transfer steps thereby easing integration. In this area, DBARC technology could have merit for
low k1 KrF and ArF (dry) lithography as well as in immersion ArF processes.
This paper describes our results in designing production worthy DBARCs for both implant and non-implant applications.
A newly developed KrF DBARC platform is evaluated for logic implant applications and compared to a standard TARC
implant process. Post develop residue and defectivity are checked for the new platform and the results compared to
production worthy BARC and implant resists. A new ArF platform was also developed and initial lithographic results
are reported for an implant application. Several non-implant applications were also investigated and results are reported
for high resolution KrF and ArF (dry) lithography as well as an immersion ArF process.
Immersion lithography for the 32nm node and beyond requires advanced methods to control 193 nm radiation
reflected at the resist/BARC interface, due to the high incident angles that are verified under high numerical aperture
(NA) imaging conditions. Swing curve effects are exacerbated in the high NA regime, especially when highly reflective
substrates are used, and lead to critical dimension (CD) control problems. BARC reflectivity control is also particularly
critical when underlying surface topography is present in buried layers due to potential reflective notching problems. In
this work, a graded spin-on organic BARC was developed to enable appropriate reflectivity control under those
conditions. The graded BARC consists of two optically distinct polymers that are completely miscible in the casting
solution. Upon film coating and post-apply baking, the two polymers vertically phase-separate to form an optically
graded layer. Different characterization techniques have been applied to the study of the distribution of graded BARC
components to reveal the internal and surface composition of the optically graded film, which includes Variable Angle
Spectroscopic Ellipsometry (VASE) and Secondary Ion Mass Spectroscopy (SIMS). Also, optical constant optimization,
substrate compatibility, patterning defectivity and etch feasibility for graded BARC layers are described. Superior 193
nm lithographic performance and reflectivity control of graded BARC beyond 1.20 NA compared to conventional
BARCs is also demonstrated.