Residual polymer stabiliser causes anisotropic electrical conductivity during inkjet printing of metal nanoparticles

The study addresses why inkjet-printed metal nanoparticle structures exhibit lower and anisotropic electrical conductivity compared to bulk metals and traditionally manufactured devices. While inkjet printing enables flexible, rapid, and multi-material fabrication of 2D and 3D electronics, the electrical performance is limited by processing and materials factors that are not fully understood. The authors hypothesize that residues of organic stabilizers used in nanoparticle inks, particularly polyvinylpyrrolidone (PVP), accumulate at interfaces during printing and low-temperature sintering and thereby hinder inter-particle and inter-layer electrical conduction. To test this, they combine electrical measurements with high-specificity 3D chemical mapping to reveal the distribution and chemical state of PVP within printed silver nanoparticle (AgNP) structures and correlate it with anisotropic conductivity.

Prior work shows widespread use of inkjet printing with colloidal materials (metal nanoparticles, quantum dots, polymers) for flexible/wearable electronics and optoelectronics, but printed conductors underperform due to suboptimal sintering and organic residues. Metal nanoparticle inks typically undergo solvent evaporation (pinning) followed by low-temperature sintering (120–200 °C) to form conductive tracks on polymer substrates. Anisotropic conductivity has been linked to thermal profiles, morphology, and potential organics, but the detailed mechanisms at low temperatures remain unclear. Organic capping agents like PVP are essential for dispersion and stability yet may persist and impair conductivity even at low concentrations. Advanced surface and depth-profiling techniques such as ToF-SIMS and XPS (augmented by gas cluster ion beams) have enabled nanometre-scale, chemically specific analyses of organic/inorganic interfaces and core–shell structures, motivating their use here to probe residual stabilizers within printed AgNP layers.

Materials and printing: Commercial AgNP ink (Advanced Nano Products, SilverJet DGP-40LT-15C; 38.85 wt% AgNPs in triethylene glycol monomethyl ether plus additives) was used as received. Printing was performed with a Fujifilm Dimatix DMP-2850 (10 pL cartridge), ~45 µm droplet diameter, 30 µm drop spacing, single nozzle to maximize uniformity. Substrates were Si/SiO2 (and ITO for vertical resistivity). Substrate temperature during printing was 90 °C to pin inks with minimal sintering. Multi-material devices used in-house TPGDA-based dielectric ink printed with AgNPs and UV curing/sintering on a PixDro LP50. Post-processing: Sintering performed in argon at temperatures up to 500 °C. Single-layer and multi-layer (100 and 200 layers) AgNP electrodes were prepared. Thermal analysis (TGA/DSC) was performed (TA Q600; 10 °C min−1; N2 100 mL min−1). Electrical measurements: Planar resistivity (ρxy) measured by four-probe Kelvin method using a Keithley 2400 sourcemeter and 7500 multimeter, with triplicate I–V scans and geometrical calibration via optical microscopy. Vertical resistivity (ρ⊥) measured on 50- and 200-layer stacks printed on ITO, sintered at 130, 200, 300 °C, using a PEN insulating spacer and a top contact with a glass slide and dried silver paste; contact resistance corrected by subtracting resistance from thinner stacks; thickness by SEM cross-sections; triplicate measurements. Morphology: SEM (JEOL 7100F, 5 keV, 5 mm WD) and TEM (JEOL 2000 FX, 200 kV). Cross-sections for TEM prepared by ultramicrotomy after epoxy embedding. Chemical analysis: ToF-SIMS surface and depth profiling using ToF-SIMS IV and 3D OrbiSIMS (Hybrid SIMS). Dual-beam mode with Bi+ analysis and argon GCIB sputtering; both high- and low-energy GCIB conditions employed to tailor depth resolution. OrbiSIMS used for high mass-resolving power in the 50–750 m/z range. Charge compensation via low-energy electron flood; data normalized to total ion counts; depth scale calibrated by optical profilometry (Zeta-20). Multivariate analysis (PCA) via simsMVA after automated peak finding and Poisson scaling. XPS (Kratos AXIS ULTRA, Al Kα, charge neutralization, high-resolution scans; peak fitting in CasaXPS with mixed Gaussian/Lorentzian components; charge-corrected to C 1s at 284.8 eV) to resolve N 1s components corresponding to non-interacting PVP (C–N ~400 eV) and interacting PVP (C–N–Ag ~398 eV). Device-level interfaces: ToF-SIMS depth profiling from top and bottom to probe PVP at TPGDA/AgNP interfaces in a fully 3D printed encapsulated device; TEM cross-sections used to examine interpenetration at interfaces.

• Residual PVP detected on printed AgNP surfaces: ToF-SIMS and XPS unambiguously identified PVP (e.g., C6H10NO+ ions; N 1s peaks), demonstrating that PVP persists after printing/pinning at 90 °C.
• Inter-layer accumulation: In four stacked AgNP layers, ToF-SIMS depth profiles showed PVP signal oscillations with ~282.8 nm period, peaking at ~300, 600, and 900 nm, corresponding to interfaces between printed layers; strongest PVP at the top surface.
• Molecular interaction and migration: XPS N 1s revealed higher non-interacting PVP (C–N at 400 eV) at the top surface and higher interacting PVP (C–N–Ag at 398 eV) in the bulk (~100 nm), indicating migration of weakly bound PVP to surfaces/interfaces during drying/sintering.
• Nanoparticle-scale organization: High depth-resolution ToF-SIMS with PCA identified an ~80.7 nm oscillatory pattern within a single printed layer, compatible with AgNP size (average 60 ± 22 nm). Alternating signals of silver (Ag+, 109Ag+, Ag2+) and PVP fragments (e.g., C4HNO+, C6H10NO+) support a core–shell AgNP/PVP structure with variable PVP coverage.
• Anisotropic resistivity: For multi-layer stacks sintered at 130 °C, vertical resistivity was three orders of magnitude higher than planar resistivity and independent of layer number: ρ⊥ = 23.6 ± 8.4 mΩ·cm (200 layers) and 23.5 ± 9.4 mΩ·cm (100 layers). Planar resistivity for single-layer samples dropped markedly after 100 °C and remained ~13 μΩ·cm between 100–200 °C but stayed above bulk Ag values by about an order of magnitude.
• High-temperature effects: Above 300 °C, single-layer planar resistivity increased due to formation of 200–500 nm voids; at 500 °C, AgNPs agglomerated into 1–5 μm macrostructures with poor continuity, degrading conductivity.
• Persistence and redistribution of PVP with sintering: PVP signals (ToF-SIMS) and XPS components indicative of PVP persisted after sintering at 100–230 °C. The fraction of non-interacting PVP increased at the surface but decreased in the bulk with higher processing temperatures, consistent with surface/interface enrichment of weakly bound PVP. TGA showed ~70% mass loss between 150–250 °C (solvent evaporation dominated), not complete removal of organics.
• Multi-material device interfaces: In a 3D printed encapsulated AgNP/TPGDA device, ToF-SIMS showed higher PVP at dielectric/conductor interfaces. At the top TPGDA/AgNP interface, silver appeared earlier than PVP in depth profiles, corroborated by TEM evidence of AgNP interpenetration into the dielectric; bottom interface showed less interpenetration, consistent with deposition order and mobility. Overall, residual PVP accumulates at layer and material interfaces, correlating with significantly reduced and anisotropic electrical conductivity.

The findings directly support the hypothesis that residual stabilizer (PVP) drives anisotropic electrical behavior in inkjet-printed AgNP structures. 3D chemical mapping reveals PVP enrichment at interfaces between printed layers and at dielectric/conductor boundaries, precisely where charge must traverse in the vertical direction. This interfacial PVP, especially in its non-interacting form that migrates during drying/sintering, likely impedes neck formation and electron transport across layers, explaining the three-orders higher vertical resistivity relative to planar. At nanoparticle scale, the observed core–shell AgNP/PVP arrangement and persistent PVP fragments further rationalize incomplete sintering at low temperatures. Processing at higher temperatures does not uniformly improve conductivity: while moderate temperatures (100–200 °C) promote bridging and reduce planar resistivity, excessive temperatures induce voiding and coalescence that degrade continuity. For multi-material 3D printing, the presence and redistribution of PVP at organic/inorganic interfaces, and interpenetration of AgNPs into the dielectric, complicate interface cohesion and charge transfer. These insights emphasize the need to control stabilizer chemistry and processing pathways to mitigate interfacial barriers and reduce anisotropy.

This work establishes that organic stabilizer residues, specifically PVP, persist and concentrate at critical interfaces in inkjet-printed AgNP structures, causing strong anisotropy in electrical conductivity. Using ToF-SIMS/OrbiSIMS and XPS with nanoscale depth resolution, the study correlates interfacial PVP distribution with elevated vertical resistivity and identifies nanoparticle-scale core–shell features. The results caution against relying solely on higher thermal budgets, which can worsen morphology and conductivity, and point to process-compatible strategies such as in situ UV/IR-assisted sintering, tailored thermal profiles, and reformulated inks with alternative stabilizers that decompose or cleave under low-temperature or photonic activation. The methodology and insights are broadly transferable to other nanomaterial inks and multi-material printed electronics. Future work should explore stabilizer chemistries engineered for facile removal or activation, real-time in situ processing within printers, and interface engineering to optimize charge transport in complex, co-printed device stacks.

• Temperature-dependent morphology and conductivity were characterized on Si/SiO2 (and ITO for vertical tests); temperature windows and morphology evolution may vary with different substrates and interfacial energies.
• Vertical resistivity measurements required contact resistance corrections and bespoke top-contact assemblies; residual contact effects may contribute uncertainty despite subtraction using thinner stacks.
• ToF-SIMS depth profiling of buried interfaces in multi-material devices required higher sputter currents and shorter acquisition per level, limiting depth resolution and signal-to-noise compared with single-material stacks.
• Persistence of organics was inferred from representative markers (e.g., C6H10NO+ and XPS N 1s components); absolute quantification across depths is challenging due to matrix effects and varying sputter yields.
• The study focuses on one commercial AgNP ink and one stabilizer (PVP); generalization to other inks/stabilizers requires further validation.