1
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 the resolution of additive manufacturing (AM, a.k.a. 3-D printing) and emerging reverse tomographic approaches discard the ‘rastering print head’ paradigm. The resolution and print quality of both approaches are limited by the chemistry and physics of the polymeric resin itself. Our recent work uses atomistic monomer models and combinations of classical and reactive molecular dynamics to bridge the gap between atomistic, particle-based models and existing continuum models. SYSTEMS & METHODS SIMULATED POLYMERIZATION DYNAMIC POLYMER TOPOLOGY In 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. REFERENCES 1. 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 cube 1 (left) and LLNL logo (right) fabricated by holographic patterning. 2 2 mm H 2 C O O OH O CH 2 O O O CH 2 O O H 2 C O O CH 2 O O O O O CH 2 O H 2 C 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% M c (Da) vinyl group conversion PETA HDDA PEGDA increasing crosslink density 0 1 2 3 4 5+ HDDA PETA links per monomer 0 200 400 600 800 1000 1200 1400 0% 5% 10% 15% 20% 25% 30% monomer in largest polymer vinyl group conversion dashed lines are estimated gel points PETA 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 O H 2 C O O CH 2 0 200 400 600 800 1000 1200 0% 10% 20% 30% 40% 50% cycles (raw count) 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% volume (vs. liquid monomer) 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

Toward photopolymer resin design for additive manufacturingPETA HDDA PEGDA 0.90 0.92 0.94 0.96 0.98 1.00 0% 10% 20% 30% 40% 50% 60% 70% r) vinyl group conversion PETA HDDA PEGDA Shrinkage

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Toward photopolymer resin design for additive manufacturingPETA HDDA PEGDA 0.90 0.92 0.94 0.96 0.98 1.00 0% 10% 20% 30% 40% 50% 60% 70% r) vinyl group conversion PETA HDDA PEGDA Shrinkage

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