Fluorinated acid amplifiers for EUV lithography

Article abstract of DOI:10.1021/ja901448d

(Chemical Equation Presented) Five new compounds were synthesized for use as acid amplifiers in EUV (13.5 nm) photoresists. Four compounds act as acid amplifiers and decompose by autocatalytic kinetics to generate fluorinated sulfonic acids, essential for the simultaneous improvement of resolution, sensitivity, and line edge roughness (LER) in EUV photoresists. The decomposition rates were studied using ¹⁹F NMR in the presence and absence of 1.2 equiv of tri-tertbutylpyridine. Three acid amplifiers decomposed 490, 1360, and 1430 times faster without base than with base. Preliminary lithographic evaluations show that cis-1-methyl-2-(4-(trifluoromethyl) phenylsulfonyloxy)cyclohexyl acetate simultaneously improves the resolution, LER, and sensitivity of an EUV photoresist. Copyright

Full text of DOI:10.1021/ja901448d

Published on Web 07/01/2009
Fluorinated Acid Amplifiers for EUV Lithography
Seth Kruger, Sri Revuru, Craig Higgins, Sarah Gibbons,† Daniel A. Freedman,† Wang Yueh,‡
Todd R. Younkin,‡ and Robert L. Brainard*
College of Nanoscale Science and Engineering, UniVersity at Albany, Albany, New York 12203
Received February 24, 2009; E-mail: rbrainard@uamail.albany.edu
High resolution chemically amplified photoresists1 are central
to the manufacture of today’s integrated circuits. These materials
employ strong acids, generated during exposure to 248, 193, or
13.5 nm light, to catalyze the transformation of photoresist polymers
from insoluble to soluble in alkaline developers. Critical to the
microelectronics industry is the need to develop chemically
amplified photoresists for use with extreme ultraviolet (EUV, 13.5
nm) light. These resists must simultaneously exhibit three properties:
high resolution, low line edge roughness (LER),2 and high
sensitivity.3
Generically, these AAs consist of three parts, a body, an acid-
sensitive trigger (T, either hydroxyl or acetate), and a sulfonic acid
precursor (A). During autocatalysis, the trigger undergoes acidoly-
sis12 yielding an olefin. In successful AAs, this olefin forms an
allylic sulfonic acid ester allowing the sulfonic ester to decompose
via an E1 or E2 elimination reaction more rapidly than the starting
AA, yielding a second double bond and a sulfonic acid. Ideally,
this autocatalytic pathway should occur 100-1000 times faster than
the uncatalyzed thermal decomposition of the AA in the absence
of acid.
We have proposed that the best way to simultaneously improve
these three properties in an EUV resist is to increase the number
of strong acids generated during exposure4 and assert that acid
amplifiers may be one of the best ways to achieve this goal. Acid
amplifiers (AAs) are compounds that decompose via acid-catalyzed
mechanisms to produce more acid.5 When the product acid is strong
enough to catalyze the decomposition of the AA, the decomposition
occurs autocatalytically (Scheme 1).6
Figure 1. We report four acid amplifiers (1a, 1b, 2, 3b) and one thermal
acid generator (3a). 1a ) R1 ) p-C6H4CF3; 1b ) R1 ) o-C6H4CF3; 2 ) R1
) p-C6H4CF3; 3a ) R1 ) p-C6H4CF3, R2 ) H; 3b ) R1 ) p-C6H4CF3, R2
) Ac; 4 ) Proposed transition state for the decomposition of 3a.
Scheme 1. Proposed Reaction Mechanisms for the Autocatalytic
(Acid Catalyzed) and Uncatalyzed Decomposition of Acid
Amplifiers 1a or 1b (R1 ) p-C6H4CF3 or o-C6H4CF3, respectively)
We evaluated the thermal stability of the compounds in a positive
photoresist matrix, by baking a blend of the AA in the resist matrix
at 70 and 110 °C for 150 s, followed by development in a standard
aqueous developer (Table 1).13 Stable AAs retain most of their film
thickness (losing <10 nm), while thermally unstable compounds
decompose to create acid which converts the resist polymer to its
soluble form allowing rapid dissolution in the aqueous developer.
Table 1 shows that AAs 1a and 3b are stable to 110 °C anneal,
AAs 1b and 2 are stable to 70 °C anneal, and 3a decomposes at 70
°C.
Table 1. Film Stability and 19F NMR Kinetics Results14
k × 103, 100 °C (s-1
)
film
stability
E0 test/
control
1.2 equiv
of base
no added
base
kAUTO/kBASE
100 °C
To be successful in a modern photoresist, AAs must (1) be
thermally stable in the absence of catalytic acid in phenolic polymer
matrices at 70-120 °C; (2) rapidly decompose autocatalytically in
the presence of the acids they produce; and (3) produce strong
(fluorine-containing sulfonic) acids capable of participating in
photoresist chemistry. To date, 26 AAs have been reported, but
none of these would be appropriate for use in photoresist applica-
tions. Only two of the AAs produce fluorinated acids;7 the others
are either too difficult or expensive to synthesize,8 do not decompose
autocatalytically,9 or are unstable under photoresist conditions.10
Here, we report on the design and kinetics of four AAs and one
thermal AA (Figure 1).11 We compare the relative rates of
autocatalyzed and uncatalyzed reactions and show the unprec-
edented ability of 3b to simultaneously improve resolution, sensitiv-
ity, and LER of EUV photoresists. Scheme 1 shows the proposed
mechanism for the autocatalytic decomposition of 1a and 1b.
AA
1a
1b
2
3a
3b
>110 °C
>70 °C
>70 °C
fail
2.6/ 4.4
–
5.0/ 6.0
–
0.033
0.053
4.3
1.1
0.0050
48
73
105
0.98
2.50
1425
1359
25
0.9
490
>110 °C
3.9/ 4.4
The thermal decomposition kinetics in solution for the AAs were
measured using 19F NMR.14 Solutions of AAs (18 mM) were
monitored in 50/50 C6D6/ethylphenol (to simulate the environment
of a phenolic polymer matrix) and in the presence and absence of
1.2 equiv of added tri-tert-butylpyridine. The sterically hindered
base was added to consume acid as it formed, so that the
uncatalyzed reactions could be studied independently.15 All reac-
tions evaluated in the presence of base showed first-order kinetics,
typically with R2 > 0.98 over two half-lives. In the absence of base,
the kinetic data were consistent with the autocatalytic mechanism
illustrated in Scheme 1. Rate constants were determined by fitting
† Chemistry Department, SUNY New Paltz, New Patlz, NY 12561.
‡ Intel Corporation, Hillsboro, OR 97124.
9
9862 J. AM. CHEM. SOC. 2009, 131, 9862–9863
10.1021/ja901448d CCC: $40.75  2009 American Chemical Society

C O M M U N I C A T I O N S
the data to second-order catalyzed kinetics.16 Rate constants at 100
°C and kAUTO/kBASE ratios are shown in Table 1.
The film stability and kinetic data show that structural factors
such as cyclic vs acyclic bodies, hydroxyl vs acetate triggers, and
primary vs secondary sulfonic esters significantly influence the
reactivity of the AAs. Compounds 1a, 1b, 3b are useful AAs. They
are relatively stable in polymeric films, have slow decomposition
rates in the presence of base, yet have an excellent rate ratio between
catalyzed and uncatalyzed decomposition. Preliminary lithographic
evaluation shows that they can increase sensitivity (E0) by
12-40%.17 Compound 2 shows greatly reduced thermal stability,
which we ascribe to the secondary attachment of the sulfonic acid
to the acyclic chain.
Figure 3. Scanning electron micrographs of dense line images printed using
a control resist formulation with and without 70 mM 3b.12
Acknowledgment. This work was supported by Intel Corporation.
Figure 2 shows thermal decomposition kinetics for compounds
3a and 3b. Remarkably, although these compounds only differ by
the identity of the trigger group (OH or OAc, respectively), their
thermal decomposition kinetics differ significantly. AA 3a is
thermally unstable, decomposing in the resist film <70 °C, and
decomposing rapidly in solution independent of the presence of
added base. Clearly, this compound does not decompose by an acid-
catalyzed reaction mechanism. In contrast, AA 3b is quite stable,
as it passed the film stability test at 110 °C. Molecular modeling
of the uncatalyzed thermolysis reactions for 3a and 3b provides
some possible insight. The modeling predicts a transition state, 4
(Figure 1), where the proton on the hydroxyl moiety of AA 3a
protonates the sulfonyl oxygen during C-O bond breaking.
Supporting Information Available: Experimental procedures and
spectroscopic data for starting materials and products. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Ito, H.; Breyta, G.; Hofer, D.; Sooriyakumaran, R.; Petrillo, K.; Seeger, D.
J. Photopolym. Sci. Technol. 1994, 7, 433–47.
(2) Shumway, M. D.; Naulleau, P.; Goldberg, K. A.; Bokor, J. J. Vac. Sci.
Technol. B 2005, 23, 2844–2847.
(3) Gallatin, G. M.; Naulleau, P.; Niakoula, D.; Brainard, R. L.; Hassanein,
E.; Matyi, R.; Thackeray, J.; Spear, K.; Dean, K. Proc. SPIE 2008, 6921,
69211E/1–69211E/11.
(4) Brainard, R. L.; Trefonas, P.; Lammers, J. H.; Cutler, C. A.; Mackevich,
J. F.; Trefonas, A.; Robertson, S. A. Proc. SPIE 2004, 5374 (Pt. 1), 74–
85.
(5) Ichimura, K. Chem. Record 2002, 2, 46–55.
(6) Gerdts, C. J.; Sharoyan, D. E.; Ismagilov, R. F. J. Am. Chem. Soc. 2004,
126, 6327–31.
(7) Lee, S.; Arimitsu, K.; Park, S.-W.; Ichimura, K. J. Photopoly. Sci. and
Tech. 2000, 13, 215–216.
(8) Park, S.-W.; Arimitsu, K.; Ichimura, K. Macromol. Rapid Commun. 2000,
21, 1050–1053.
(9) Arimitsu, K.; Kudo, K.; Ichimura, K. J. Am. Chem. Soc. 1998, 120, 37–
45.
(10) Kudo, K.; Arimitsu, K.; Ohmori.Ito, H.; Ichimura, K. Chem. Mater. 1999,
11, 2119–2125.
(11) We selected p- and o-CF3(C6H4)SO3-acid precursors because they yield
AAs with good thermal stability and give simple 19F spectra.
(12) Arimitsu, K.; Kudo, K.; Ichimura, K. J. Am. Chem. Soc. 1998, 120, 37–
45.
(13) Photoresist films were prepared containing 70 mM of an AA, 7.5% of the
photoacid generator bis(4-tert-butylphenyl)iodonium nonaflate and phenolic
terpolymer [poly(hydroxystyrene)-(styrene)(tert-buylacrylate) 65/20/15 mol
%] and 0.5 wt % of tetrabutyl ammonium hydroxide in 125 nm film on a
Si wafer. The film was then baked at 110 or 70 °C for 150 s and developed
in 0.26 N Me4N+ OH- for 45 s.
(14) Reaction rates (with and without 1.2 equiv of 2,4,6-tri-tert-butylpyridine)
were followed by integration of the CF3 peaks for the starting AA and the
product CF3C6H4SO3H (or its pyridine salt). Estimated error for rate
constants with and without base is 2.5% and 5%, respectively.
(15) Compound 1a was decomposed in the presence of 2,4,6-trimethylpyridine,
2,6-di-tert-butyl-4-methylpyridine, and 2,4,6-tri-tert-butylpyridine giving
Figure 2. Kinetic plots for the thermal decomposition of 3a and 3b in the
presence and absence of 2,4,6-tri-tert-butylpyridine.
first-order decomposition rate constants of k ) 9.7 × 10-5, 1.0 × 10-4
,
and 1.0 × 10-4 s-1, respectively. This result is consistent with the
interpretation that the base removes the acid from the reaction mechanism,
but the base does not actively participate in the reaction by extracting
protons from the AA bodies.
To assess the impact of our AAs on improving EUV patterning
performance, we calculated the Z-parameter18 of the control resist
along with resists prepared using 70 mM AAs 1a and 3b to be 74,
54, and 25 mJ*nm3, respectively (smaller is desired). These results
demonstrate simultaneous improvements in resolution, LER, and
sensitivity for 1a and 3b and as illustrated in the SEM images shown
in Figure 3 for 3b.19
1,3-Acyclic and 1,2-cyclic AAs expand the scope of available
methods for introducing strong fluorine-containing sulfonic acids
in chemically amplified photoresists enabling us to achieve
simultaneous improvements in resolution, LER, and sensitivity and
realize a 3-fold gain in the Z-parameter. These studies demonstrate
that small structural changes dramatically affect AA activity. Further
studies of this type will allow us to design even more effective
acid amplifiers.
(16) Second-order catalyzed reaction kinetics was evaluated by fitting data to:
Capellos, C.; Bielski, B. Kinetic Systems: Mathematical Description of
Chemical Kinetics in Solution; Wiley-Interscience: New York, 1972.
(17) E0’s were determined by exposing the resist film described in ref 15 to
stepwise doses of EUV light, baking at 110 or 70 °C for 90 s and then
developing for 45 s. The lowest dose to give zero thickness is E0.
(18) Z-parameters simultaneously evaluate the resolution, LER, and sensitivity
of a resist. Wallow, T.; Higgins, C.; Brainard, R.; Petrillo K.; Montgomery
W.; Koay C.-S.; Denbeaux G.; Wood O.; Wei Y. Proc. SPIE 2008,6921,
69211F/1-69211F/11.
(19) Four non-fluorinated AAs were evaluated in a 193 nm resist. Sensitivity
improved in all cases, and in one case resolution improved. (a) Naito, T.;
Ohfuji, T.; Endo, M.; Arimitsu, K.; Ichimura, K. J. Photopolym. Sci.
Technol. 1999, 12, 509–514. (b) Ohfui, T.; Takahashi, M.; Kuhara, K.;
Ogawa, T.; Ohtsuka, H.; Sasago, M.; Ichimura, K. Proc. SPIE 1997, 3049,
76–82.
JA901448D
9
J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009 9863