Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Toward photopolymer resin design for additive manufacturing John J. Karnes (LLNL), Todd H. Weisgraber (LLNL), James S. Oakdale (LLNL), Maxim Shusteff (LLNL), and Juergen Biener (LLNL)
Two-photon polymerization (2PP) techniques continue to increase theresolution of additive manufacturing (AM, a.k.a. 3-D printing) andemerging reverse tomographic approaches discard the ‘rastering printhead’ paradigm. The resolution and print quality of both approaches arelimited by the chemistry and physics of the polymeric resin itself. Ourrecent work uses atomistic monomer models and combinations ofclassical and reactive molecular dynamics to bridge the gap betweenatomistic, particle-based models and existing continuum models.
SYSTEMS & METHODS
SIMULATED POLYMERIZATION
DYNAMIC POLYMER TOPOLOGYIn the present work we study three acrylate monomers that are commonly used as AM
photosensitive resins. Simulation boxes for each of these three systems contain 2000
monomers that were equilibrated in the NPT ensemble at 300 K using the OPLS-UA
force field.
Our goal: Designer photopolymer resins for additive manufacturing informed by molecular mechanics and graph theory.
Molecules with active radical sites (red) rapidly convert the liquid PETA monomer (cyan, transparent). Cubic simulation boxes are approximately 10 nm per edge.
The trifunctional PETA monomer (left) has a higher density of reaction sites. Its corresponding united-atom representation is shown on the right.
REFERENCES1. Oakdale , J. S.; Ye, J.; Smith, W. L.; Biener, J. Optics Express 2016, 24, 27077-27086.
2. Shusteff, M. et. al. Science Advances 2017, 3, eaao5496.
3. Gissinger, J. R.; Jensen, B. D.; Wise, K. E. Polymer 2017, 128, 211–217.
1 2 3
After equilibrating the monomer liquid, the radical sites are allowed to
react with nearby free monomer. At the end of the simulation, nearly all of
the 20 radicals are present within the final polymer molecule as trapped
radicals: active, but not within range of an unreacted vinyl group.
When a radical-bearing PETA monomer ‘reacts’ with another free PETA monomer the local bond, angle, and dihedral topologies are updated to reflect the new chemical environment.
To create the polymer network, 20 random vinyl groups (1% of the monomer molecules) are
mutated into radical sites at t = 0 to simulate photoinitiation. We use fix bond/react, a
proximity-based reaction scheme recently developed by Gissinger et. al.3 to model diffusion-
limited polymerization.
A 2PP-printed free-standing octet-truss cube1 (left) and LLNL logo (right) fabricated by holographic patterning.2
2 mm
H2C O
O
OH OCH2
O
O
OCH2
O
OH2C O
OCH2
O O O O
OCH2
OH2C
HDDA (left) and PEGDA-250 (right) are bifunctional acrylates of similar size with slightly different functionalities between the acrylate groups.
By monitoring the size of the
largest propagating polymer, the
inflection point is assigned to be
the gel point, indicating when
the largest percolating mass
spans the simulation box.
Mathematical graph theory quantifies the loop/cycle and crosslinking densities of
propagating polymer networks. These microscopic topological features correlate
with macroscopic observables and will be used in future monomer design.
800
700
600
500
400
300
200
0% 10% 20% 30% 40% 50%
Mc
(Da)
vinyl group conversion
PETAHDDA
PEGDA
incr
easi
ng c
ross
link
dens
ity
012345+
HDDA PETA
links per
monomer
0
200
400
600
800
1000
1200
1400
0% 5% 10% 15% 20% 25% 30%
mon
omer
in la
rges
t pol
ymer
vinyl group conversion
dashed lines are
estimated gel pointsPETA
HDDA
PEGDA
COARSE GRAINING
• Expand reactive templates to include other chemical systems of interest.
• Measure physical parameters for model verification and calibration.
• Adapt graph theory-based metrics to analysis of polymer networks.
This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
FUTURE WORK
Since our interest lies in larger length and time scales, validation of OPLS-based
models will be followed by a coarse graining (CG) approach. The same reactive
scheme will allow us to explore both MARTINI and Boltzmann Inversion approaches.
Initial MARTINI-based coarse-graining of HDDA reduces the number of particles by a factor of 8.5 and increases the time step by an order of magnitude.
0.00
0.05
0.10
0.15
2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
! (")
" (Å)
0.00
0.01
0.02
0.03
0.04
60 80 100 120 140 160 180
! (#
)
# (deg)
O
OH2C O
OCH2
0
200
400
600
800
1000
1200
0% 10% 20% 30% 40% 50%
cycl
es (r
aw c
ount
)
vinyl group conversion
PETA
HDDA
PEGDA
0.90
0.92
0.94
0.96
0.98
1.00
0% 10% 20% 30% 40% 50% 60% 70%vo
lum
e (v
s. liq
uid
mon
omer
)vinyl group conversion
PETA
HDDA
PEGDA
Shrinkage may be quantified on-the-
fly during the simulation as the
reaction progresses and covalent
bonds replace van der Waals forces.
A crosslinking ‘heat map’ provides molecular insight to polymer chemists. Differences between bi- and trifunctional acrylates are evident in initial simulations.
LLNL-POST-785020