Influence of Confinement and Antiplasticization on the Dynamics of Polymers

1. Influence of Confinement and Antiplasticization on the Dynamics of Polymers Robert Riggleman Tommy Knotts and Juan J. de Pablo Center for NanoTechnology and Department of Chemical and Biological Engineering University of Wisconsin- Madison This work is based in part by a grant from the Semiconductor Research Corporation under Grant Number 2005-OC-985 task number 985.008.

2. Summary I. Computational resources available to our group II. Influence of Confinement and Antiplasticization on the Dynamics of Polymers III. Development of a Mesoscale DNA Model

3. Grid Laboratory of Wisconsin Initially funded as a collaboration of six research groups Each group contributes machines, has access to other group’s machines Group priority on owned machines Idle machines available to all Each site administers their own machines Resources: 1200 2.8 GHz Xeon cpus 200 1.8 GHz Opteron cpus 1000 ~1 GHz cpus in general Condor pool

4. Group Resources & Requirements Timeline of de Pablo group resources: 2002 – 130 private cpus 2003 – 170 private cpus 2004 – 160 private + 130 GLOW = 290 Total 2005 – 140 private + 190 GLOW = 330 Total 2006 – 150 private + 250 GLOW = 400 Total Use of Condor enables our group to function as well as it does Submit jobs and forget about them Substantial use of Standard Universe Typical number of jobs: 400-500 total in queue 350-400 running at any given time (including private, non-Condor pool) Introduction of GLOW in 2004 greatly expanded our resources

5. II. Influence of Confinement and Antiplasticization on the Dynamics of Polymers

6. Introduction Moore’s Law predicts that the number of transitors on a chip doubles every 18-24 months Gordon MooreChairman Emeritus, Intel Corporation This requires smaller features to be produced on cpu chips There are many challenges facing the semiconductor industry as they attempt to fabricate next-generation chips

7. Chemically amplified resist technology at the sub-32 node Acid Reacted polymer Expose Bake (120~140 oC) Develop Etch Photoacid generator (PAG) Pawloski, AJ, et al, J. of Vac. Sci. and Tech., 22, 869 (2004) Polymer chain The semiconductor industry has the goal of producing chips with features smaller than 45 nm within the next five years Segmental motion Properties of photoresist materials change when they are confined to such small geometries Current technology will have to be modified in order to achieve stable structure on nanoscopic lengthscales PAG

8. Being able to reproduce stable structures with lengthscales less than 40 nm is critical Physical properties of polymers are known to change when they are confined to small lengthscales Decrease in Tg causes a decrease in the stiffness, increase in the dynamics in polymers Can lead to collapse during the rinsing step of photolithography! 300 nm Introduction - Mattson, J, Forrest, JA, Boerjesson, L, Phys. Rev.E 62 5187 (2000)- Stoykovich, MP, Cao, HB, Yoshimoto, K, Ocola, LE, Nealey, PF, Adv. Mat.15 1180 (2003)

9. Plasticizers and Antiplasticizers Antiplasticizer • Decreases Tg • Increases the density • Increases the elastic constants • Ex: water in poly(amide), organophosphates in an epoxy, Epon 825 (below) Plasticizer • Decreases Tg • Decreases the density • Decreases the elastic constants • Ex: Water in poly(styrene), adipic acid polyesters in poly(lactide acid) (shown below) Martino V.P., Proc. 8th Poly. Adv. Tech. Int. Sym. (2005) Zerda, AS, et al, Poly. Sci. and Engr., 11, 2125 (2004)

10. Multiscale Modeling Glass-forming polymer 30 - 50 nm ~0.5 nm 2.0 - 5.0 nm O C C CH2 O CH3 CH3 Coarse-grained polymer chain (500 - 1000mers) Monomer unit of PMMA (100g/mol) Photoresist nanostructures Statistical segment (~10mers) 40mers

11. Coarse-Grained Model Small inclusion solvent Glycerol polymer r ~2 nm U (Interaction energy) ~1 nm ~2 nm Molecular Dynamics • All interactions are assumed to be pairwise • Forces on all particles calculated at each timestep: • Fij = - (dU/drij) • Positions of each particle updated based on Newton’s equations of motion • Procedure repeated to obtain equilibrium results Interparticle distance, r [nm]

12. Procedure Computational Time Required Begin with 5 initial configurations per system and equilibrate at high temperature with molecular dynamics 3 weeks/system: ~30 weeks total Cool each system to the temperatures of interest (10 T’s / system x 5 configs x 2 systems) 1 week/config: ~100 weeks total Equilibrate at each temperature(10 T’s / system x 5 configs x 2 systems) 1 week/config: ~100 weeks total Production runs at each temperature, data analysis(10 T’s / system x 5 configs x 2 systems) 2 weeks/config: ~200 weeks total ~ 430 weeks or8.3 years Real time: ~ 11 weeks

13. Effect of Inclusions on Density and Tg 5 wt % Solvent Pure PMMA Tg Tg Density [gm/cm3] increase density decrease Tg Temperature [K]

14. Elastic Modulus of Bulk Systems with Inclusions 5 wt % Solvent Pure PMMA Shear Modulus (GPa) Solventincrease G’ at low T Antiplasticization! Temperature [K]

15. Molecular Motion Near Tg Initial position Final position Upon cooling, relaxation requires that larger domains to move cooperatively Size of domains increases as T decreases Cooperative regions shown to behave as 1-D strings Antiplasticized Polymer: - Smaller cooperatively rearranging regions - Weak T dependence Pure Polymer: - Larger cooperative regions - Strong T dependence Results of previous studies have shown: Example of 1D cooperative motion Hypothesis Particles replace the position of each other along a 1-dimensional string

16. Cooperatively Rearranging Regions Pure Polymer Polymer + Antiplasticizer Large, extended clusters of mobile particles Smaller, isolated mobile particles

17. Effect of Antiplasticization: Shorter string length Weaker temperature dependence Cooperative Rearrangements: Bulk ftotal = 0.296

18. Free-Standing Thin Film: Density Profiles Macroscopic 5 wt. % at T/Tg = 1.58 Air Air Polymer + inclusion Local density of Polymer [g/cm3] Local density of Inclusion [g/cm3] h ≈ 30 nm Nanoscopic h 0 4 8 12 16 20 24 28 32 36 40 z [nm] y z Periodic boundaries in x and y directions x

19. Thin Film Cooperative Regions Pure Polymer Cooperative regions exist near free surfacesFilms are strongly heterogeneous Air Surface Region Surface Region Air Polymer + Antiplasticizer Homogeneous distribution of cooperative regions

20. Cooperative Rearrangements: Films T = 1.5 Tg T = 1.9 Tg Probability Antiplasticized Polymer Pure Polymer -10 0 -5 5 10 z-position (nm) Glass Transition Temperature Ratios: Pure polymer films strongly heterogeneous Antiplasticizer homogenizes the freestanding film: eliminates surface effects Pure Antiplas. (5%) TG/TG,bulk: 0.72 0.99 R. A. Riggleman et al., Phys. Rev. Lett., 97 (2006) 045502

21. Elimination of Confinement Effects: Experimental Evidence Thin PS films with various amounts of additives Pyrene makes the system a strongerglass former More homogeneous system Smaller cooperative regions Pyrene shown to eliminate confinement effects in PS films Oligomeric PS did not eliminate confinement effects Specific additives must be used! 9% Pyrene 2% Pyrene 0.2% Pyrene Ellison C.J. et al., Phys. Rev. Lett., 92 095702 (2004)

22. Summary and Conclusions Chemically amplified resists contain a substantial amount of low molecular weight additivesPlasticizers or antiplasticizers Developed a unique model to study the effects of antiplasticization in polymeric melts Shown that antiplasticization: Stiffens the bulk material Reduces the size of the cooperatively rearranging regions Eliminates the effects of confinement in photoresists

23. III. Parameterization of a Mesoscale Modelfor Molecular Simulations of DNA

24. Time/Length Scales and Existing Models 1 nm 10 nm 1 mm 10 mm 100 mm 100 nm Atomistic Coarse graining xp contour length L persistence length Rg 3.4 nm radius of gyration 2 nm ? CHARMM, AMBER, etc. Jendrejack et al., J Chem Phys 116:7752-7759 (2002).

25. Gō-like Model • Objectives/Features of Molecular Model • Allow and describe hybridization • Reduce the number of sites • Reproduce thermal and mechanical experimental data • Include coulombic interactions and salt effects • Simple to understand and implement

26. Proposed Model Intramolecular Intermolecular Computational time to develop and characterize: >19 years

27. “Long” DNA and Mechanical Properties 0.5 mm 2 nm Smith, Cui, and Bustamante, Science, 271:795-799 (1996). • Persistence Length, lp • l-phage DNA • Experimental lp ~ 45 nm; sim ~30 nm • Model System • 1489 bp fragment of l-phage DNA • Digest l DNA with StyI • 0.5 mm Chen et al., Macromolecules 38:6680-6687 (2005).

28. Group Computational Demands Since April 2006, on UW-GLOW: >75 years of cpu time!

29. Acknowledgements Yioryos Papakonstantopoulos, Dr. Kenji Yoshimoto, Dr. Jack Douglas Tommy Knotts Prof. Juan de Pablo University of Wisconsin Condor Team Semiconductor Research Corporation through Grant Number 295-OC-985,NSF NIRT funding, and UW-NSEC Questions ?