A ReaxFF molecular dynamics study of molecular-level interactions during binder jetting 3D-printing
Summary¶
Binder jetting 3D printing (BJP) of AISI 316L stainless steel is modeled with ReaxFF MD using Cr-oxide nanoparticles and aqueous diethylene glycol (DEG) binder, following a print → cure → burn-out → sinter thermal sequence. Simulations relate hydrogen-bond networks, DEG oxidation, and Cr–O bond formation to a computed breaking strength proxy (restraint potential separating two nanoparticles). Varying water/DEG composition and comparing 2-ethoxyethanol, DEG, and 1-(2,2,2-trihydroxyethoxy)ethane-2,2,2-triol probes the role of hydroxyl content during early stages.
Methods¶
Force-field lineage (A)¶
Cr/C/H/O ReaxFF from Shin et al. for Cr₂O₃-rich passivated nanoparticle surfaces.
Molecular dynamics: system, stages, and mechanical probe (B)¶
Particles: Cr-oxide spheres ~22 Å diameter (~500 atoms) after annealing in H₂O + O₂ at 1273 K. Print reference: two particles, 160 H₂O, 60 DEG, 80 Å periodic cube, 300 K, NVT, Berendsen thermostat (100 fs damping).
Binder-jetting sequence: print 300 K → cure 393 K in 200 Å cell (per article/ESI) with dissociated water removed → burn-out 900 K (volatiles stripped) → sintering 1900 K; direct heating between stages.
Stage goals mirror BJP practice: print/cure preserve H-bonded green strength, burn-out removes organics that would outgas in furnace sintering, and 1900 K sintering welds particles through oxide bridges rather than metallic necking in this ReaxFF model.
Mechanical proxy: bell-shaped restraint separating 50 atom pairs (25 per particle); strain rate via R₁₂ updates (Eq. 3). Sets A/B vary water/DEG (Table 1). ESI figures S1–S3 for geometries.
Static QM (C)¶
Not a standalone DFT study—ReaxFF application to BJP.
Constant-volume stages. All BJP staging trajectories summarized above remain NVT without NPT barostat or target GPa pressure—N/A for hydrostatic pressure control beyond the stress implied by the restraint probe.
Findings¶
Mechanisms¶
H-bond networks from DEG + water link particles in print/cure, raising restraint strength when both polar components are balanced. Burn-out consumes DEG and disrupts H-bonds; sintering forms Cr–O bridges. Water contributes little to late-stage strength; an optimal binder content emerges for post-sinter cohesion. Hydroxyl count (2-ethoxyethanol < DEG < triol) tracks early strength in this workflow.
Sets A/B in Table 1 vary water/DEG ratios at fixed particle count to show non-monotonic cohesion: too much water dilutes DEG H-bond donors, while DEG-rich mixes retain organic linkers through cure but still pyrolyze during burn-out, so post-sinter Cr–O bridges dominate ultimate strength in the model.
Limitations¶
Nanoparticle size and nanosecond MD windows do not map one-to-one to industrial BJP time/length scales; the study aims at mechanistic trends rather than quantitative part-scale prediction.
Relevance to group¶
Penn State mechanical engineering / van Duin-group ReaxFF application to additive manufacturing binders and oxide powder beds.
Citations and evidence anchors¶
papers/Gao_PCCP_binder_jetting_3D_printing.pdf(PCCP article PDF).