Enhanced corrosion resistance of additively manufactured stainless steel by modification of feedstock

Austenitic 316L stainless steel is widely used across industries but producing complex structures conventionally is costly and time-consuming. Selective laser melting (SLM), a powder bed fusion AM technique, enables complex geometries but reported corrosion behavior of SLM 316L is inconsistent, attributed to defects (e.g., porosity) and microstructural heterogeneity. Post-processing to remedy these issues can be expensive and may even degrade corrosion performance. Improving corrosion resistance in the as-printed condition via feedstock modification offers a potential solution. Prior approaches to modify feedstock include mechanical alloying, surface deposition/coating, and mixing of additives, yet corrosion-focused studies are limited and none showed improved corrosion performance of SLM-316L using modified feedstock. Because inclusions (e.g., MnS) and microstructural inhomogeneity drive corrosion in stainless steels, and nitrogen is known to enhance corrosion resistance, the authors hypothesize that adding a suitable nitrogen-containing additive that suppresses deleterious inclusions, refines microstructure, and increases nitrogen could enhance corrosion resistance. Chromium nitride (CrN), with a melting point below SLM processing temperatures, was selected to become incorporated into the matrix during SLM.

The literature reports both superior and inferior corrosion resistance for SLM-316L relative to wrought 316L, with inconsistencies linked to AM-induced defects and heterogeneities. Post-processing (HIP/heat treatments) can worsen corrosion behavior. Research on modified feedstocks in AM has largely targeted microstructure and mechanical property improvements, with few studies on corrosion of stainless steels from modified feedstock and none demonstrating improved corrosion in SLM-316L. Inclusion chemistry (e.g., MnS, mixed sulfide-oxide inclusions) and inclusion size/distribution critically affect pitting initiation. High-nitrogen stainless steels generally exhibit improved pitting resistance, repassivation, and reduced metastable pitting via mechanisms including NH4+ formation, enhanced passive film stability, nitrate formation, and Cr enrichment in the passive film. These insights motivate additive selection and the feedstock modification strategy to engineer corrosion-resistant SLM 316L.

Feedstock modification and sample manufacturing: Gas-atomized 316L feedstock (max particle size 63 µm) was ball-milled with 1 wt.% CrN additive (1–3 µm) using stainless steel balls/jars at 180 RPM for 5 h, BPR 1:1. Wrought 316L sheet was used for comparison. SLM was performed on a Renishaw AM400 (400 W Yb fiber laser, ~70 µm spot) in argon with O2 setpoint <2000 ppm. Cylindrical coupons (15 mm diameter, 5 mm height) were printed from commercial and modified feedstocks, denoted SLM-316L and SLM-316L/CrN respectively.

Characterization: Density by Archimedes method. Bulk composition by ICP-OES; C/S by LECO carbon/sulfur analyzer; O/N/H by LECO ONH836 (ASTM E1019 and AL0025). XRD using Rigaku SmartLab (Cu Kα). Microstructure by FEI Verios 460L SEM; specimens ground to 1200-grit and polished to 0.05 µm, then electro-etched in 10% oxalic acid at 15 V for 60 s (ASTM A262 practice A). Sub-cell areas measured via ImageJ. STEM (Talos F200X G2, 200 kV) in BF and HAADF modes with Super-X EDS; site-specific lamellae prepared by FIB (Quanta 3D FEG), with Pt capping and low-current thinning. Passive films formed by immersion in 3.5 wt.% NaCl for 24 h were analyzed by XPS (SPECS, 300 W Mg anode, hv 1253.6 eV; survey 0–1100 eV, high-resolution O1s, Fe2p, Cr2p, Ni2p, Mo3d; CasaXPS with Shirley background; C1s at 285 eV calibration) and ToF-SIMS (negative ions; pulsed 25 keV Bi3, 0.32 pA over 100×100 µm2; sputter with Cs+ 1 keV, 7.2 nA over 200×200 µm2; profiles acquired within crater to avoid edge effects).

Electrochemistry: Cyclic potentiodynamic polarization (CPP) in 3.5 wt.% NaCl at room temperature using a three-electrode flat cell (SCE reference, Pt mesh counter). Specimens ground to 1200-grit. OCP monitored for 1 h, then polarized from OCP −0.20 V at 1 mV/s; forward scan stopped and reversed at either 1.5 VSCE or current density 100 µA/cm2. Breakdown potential (Eb), repassivation potential (Erep), and maximum current density (imax) extracted. Potentiodynamic polarization (PDP) similarly performed but without reverse scan and with an upper current limit of 1000 µA/cm2. Potentiostatic polarization (PSP) for 90 min at Eapplied of 700 mVSCE (SLM-316L) and 900 mVSCE (SLM-316L/CrN) on 0.079 cm2 exposed area; post-test SEM used to assess metastable/stable pits.

• Densities (g/cm3): W-316L 7.90±0.01; SLM-316L 7.90±0.01; SLM-316L/CrN 7.89±0.01.
• Phases (XRD): SLM-316L and SLM-316L/CrN showed austenite only; W-316L showed austenite plus a minor peak consistent with martensite or ferrite.
• Microstructure: Similar melt pool dimensions for SLM-316L and SLM-316L/CrN, indicating CrN did not alter melt pool size. Both exhibited equiaxed and columnar cellular structures with sub-cells. Sub-cell boundaries were wider in SLM-316L. Sub-cell area distribution refined markedly in SLM-316L/CrN: SLM-316L had most sub-cells in 0.05–0.15 µm2 range; SLM-316L/CrN had most <0.005 µm2.
• Oxide inclusions (STEM-EDS): Along cellular and sub-cell boundaries. SLM-316L inclusions (30±15 nm) contained Si, Mn, O, and S; SLM-316L/CrN inclusions (53±14 nm) contained Si, Mn, O (some with Cr) and were not S-enriched.
• Elemental segregation: Cellular boundaries in both SLM materials showed segregation of Cr, Ni, Mo, S and Fe depletion; nitrogen segregation not resolved by STEM-EDS.
• Nitrogen contents (LECO): SLM-316L 0.0820 wt.% N; SLM-316L/CrN 0.165 wt.% N.
• CPP in 3.5 wt.% NaCl: • W-316L: metastable pitting from ~150 mVSCE; Eb 546±20 mVSCE; reverse imax 1510±62 µA/cm2; Erep 129±13 mVSCE. • SLM-316L: metastable pitting from ~250 mVSCE; Eb 795±63 mVSCE; reverse imax 2276±97 µA/cm2; Erep −113±30 mVSCE. • SLM-316L/CrN: no metastable pitting; Eb 1018±38 mVSCE; reverse imax 104±20 µA/cm2; Erep 994±69 mVSCE. • Eb improvement: SLM-316L/CrN +220 mV vs SLM-316L (+20%) and +470 mV vs W-316L (+45%).
• Post-corrosion morphology (PDP): SLM-316L showed pits with corroded cells and intact cell boundaries; SLM-316L/CrN showed no pits; at periphery, cell boundaries corroded while cells remained intact (behavior consistent with crevice-like attack localized to periphery).
• PSP (90 min): At 700 mVSCE (SLM-316L) exhibited metastable pitting with current transients ~60–700 µA/cm2 and subsequent stable pitting; SEM revealed multiple pits 15–30 µm (low peaks) and larger pits ~200 µm (higher peaks). At 900 mVSCE, SLM-316L/CrN showed no metastable pitting and no noticeable corrosion by SEM. Metastable pit frequency λ: SLM-316L 0.063 cm−2 s−1; SLM-316L/CrN 0 cm−2 s−1.
• Passive films (XPS/ToF-SIMS): Both SLM-316L and SLM-316L/CrN showed similar passive film chemistries. XPS deconvolution indicated Fe0/FeOOH/Fe2O3/Fe3O4, Cr0/Cr(OH)3/CrO3/Cr2O3, Ni0/NiO, Mo0/Mo4+/Mo6+. Ratios: Crox+hy/F eox+hy 2.66 (SLM-316L) vs 2.85 (SLM-316L/CrN); Totalox/Totalhy 3.58 vs 3.03. ToF-SIMS depth profiles showed outer hydroxide-rich zone (OH−, FeO2−, MoO2−) and inner oxide-rich zone (NiO−), with CrO2− across the film and Ni enrichment at the metal/film interface. No nitrogen species detected by XPS/SIMS at these N levels.
• Bulk composition (wt.%): SLM-316L vs SLM-316L/CrN: Cr 18.02 → 20.17; Ni 13.36 → 15.00; Mo 2.55 → 3.05; Mn 0.56 → 1.52; Si 0.81 → 0.24; Cu 0.16 → 0.006; P 0.020 → 0.010; S 0.013 → 0.010; C 0.028 → 0.018; N 0.0820 → 0.165. W-316L for reference: Cr 16.55, Ni 10.03, Mo 2.02, Mn 1.18, Si 0.27, C 0.016, N 0.062, etc.

The study demonstrates that modifying the SLM 316L feedstock with 1 wt.% CrN markedly enhances pitting corrosion resistance, addressing the inconsistency challenge in AM corrosion performance without relying on post-processing. Compared with wrought 316L, SLM-316L shows improved Eb due to the absence of MnS inclusions typical in wrought alloys; however, SLM-316L still contained sulfur-rich Si–Mn–S–O oxide inclusions that act as cathodic sites for pit initiation and, coupled with cellular chemical inhomogeneity, promote metastable pit growth to stability. In contrast, SLM-316L/CrN lacked S-enriched inclusions, exhibited a refined cellular/sub-cell structure, and had altered bulk chemistry (higher Cr, Ni, Mo, and N), collectively eliminating metastable pitting and yielding substantially higher Eb and Erep with much lower repassivation currents. Passive film composition and structure (XPS/ToF-SIMS) were broadly similar between SLM-316L and SLM-316L/CrN, suggesting that improvements arise primarily from microstructural inclusion control, refined sub-cell structure, and beneficial alloying (notably increased N, Cr, Ni, Mo) that stabilize passivity and increase local pH within incipient pits via nitrogen chemistry. The findings support feedstock modification as a viable route to engineer corrosion-resistant AM 316L components in the as-printed condition.

Feedstock modification of 316L stainless steel for SLM by adding 1 wt.% CrN produced coupons with significantly superior corrosion performance in 3.5 wt.% NaCl: no metastable pitting, higher breakdown potentials (≈1018±38 mVSCE), faster and higher repassivation (Erep ≈ 994±69 mVSCE), and markedly lower repassivation current densities compared to both SLM-316L and wrought 316L. The improvement is attributed to removal of sulfur-rich oxide inclusions, refinement of the cellular/sub-cell structure, increased nitrogen content (0.165 wt.% vs 0.0820 wt.%), and elevated Cr/Ni/Mo contents. Passive film chemistry was similar across SLM variants, indicating microstructure and bulk composition dominate the corrosion enhancement. The work establishes a practical route—feedstock modification—to bolster corrosion resistance of AM stainless steels and motivates exploration of other additives, particularly nitrogen-bearing compounds, and deeper mechanistic studies into sub-cell refinement and inclusion evolution during SLM.

• The precise mechanism by which CrN addition refines sub-cell structure and alters solidification remains unresolved and is identified for future study.
• Nitrogen species were not detected in passive films by XPS/ToF-SIMS, likely due to low overall N content; thus, direct spectroscopic evidence of nitrogen’s role in passivity was not obtained.
• Only a single additive (CrN) at one loading (1 wt.%) and one chloride environment (3.5 wt.% NaCl) were evaluated; broader additive types, concentrations, and environments were not assessed.