Iron Nitride Nanoparticles for Enhanced Reductive Dechlorination of Trichloroethylene

Nitriding has been used for decades to improve the corrosion resistance of iron and steel materials. Moreover, iron nitrides (FexN) have been shown to give an outstanding catalytic performance in a wide range of applications. We demonstrate that nitriding also substantially enhances the reactivity of zerovalent iron nanoparticles (nZVI) used for groundwater remediation, alongside reducing particle corrosion. Two different types of FexN nanoparticles were synthesized by passing gaseous NH3/N2 mixtures over pristine nZVI at elevated temperatures. The resulting particles were composed mostly of face-centered cubic (γ′-Fe4N) and hexagonal close-packed (ε-Fe2–3N) arrangements. Nitriding was found to increase the particles’ water contact angle and surface availability of iron in reduced forms. The two types of FexN nanoparticles showed a 20- and 5-fold increase in the trichloroethylene (TCE) dechlorination rate, compared to pristine nZVI, and about a 3-fold reduction in the hydrogen evolution rate. This was related to a low energy barrier of 27.0 kJ mol–1 for the first dechlorination step of TCE on the γ′-Fe4N(001) surface, as revealed by density functional theory calculations with an implicit solvation model. TCE dechlorination experiments with aged particles showed that the γ′-Fe4N nanoparticles retained high reactivity even after three months of aging. This combined theoretical-experimental study shows that FexN nanoparticles represent a new and potentially important tool for TCE dechlorination.

. XRD patterns of nZVI particles (NANOFER 25P) used for the synthesis of FexN particles (red) and used as reference material in aging and reactivity experiments (blue). Table S1. Rietveld XRD quantification of Fe-containing mineral phases in the XRD profiles of fresh nZVI particles (NANOFER 25P) used in this study.

Text S2. Particle Characterization Sample Preparation
Fresh particles were surface passivated before characterization. The surface passivation consisted of particle suspension in water using the same procedure as used for aging and reactivity experiments, i.e., 4 g of particles were stirred with 16 mL of deoxygenated ultrapure water using T25 ULTRA-TURRAX ® disperser (IKA, Germany) at 11 000 rpm for 2 min. Subsequently, the particles were separated with a magnet, the supernatant was discarded, the remaining pellet was either washed with ethanol three times and dried in an Ar-filled glovebox for 48 h (samples for water contact angle measurements, BET, and elemental analysis) or flash-frozen in liquid N2 (samples for other characterization methods). Samples were stored either in airtight tubes filled with an inert atmosphere inside the glovebox or in a freezer at -80 °C before analysis, respectively. Aged particles were separated from suspensions with a magnet, the supernatant was discarded, and the remaining pellet was flash-frozen in liquid N2 (at -196 °C). Samples were stored at -80 °C before analysis.

Powder X-Ray Diffraction
The phase composition of fresh and aged particles was determined by X-ray powder diffraction (XRD) with an Aeris diffractometer (PANalytical, B.V.) operating in Bragg-Brentano geometry. Defrozen samples were transferred on a zero-diffraction silicone sample holder and immediately covered with Kapton foil to protect the samples against degradation. The diffractometer was equipped with a CoKα radiation source, fixed divergence, diffracted beam anti-scatter slits, and PIXcel detector. The patterns were measured in the 2θ range from 5 to 105° and the data were processed using HighScorePlus software in conjunction with PDF-4+ and ICSD databases. Crystalline phases were quantified using the Rietveld refinement. 4

Mössbauer Spectroscopy
Iron speciation was analyzed using a low-temperature (-123 °C) Mössbauer spectrometer on isotope 57 Fe (MS2007 instrument), operating in a constant acceleration mode and equipped with a 50 mCi 57 Co(Rh) source. 5,6 Defrozen samples were transferred to a plastic in-house-built sample cuvette, immediately frozen in liquid N2 (-196 °C), and mounted in a cooling chamber (-123 °C) where the samples were analyzed. MossWinn software 7 was used for fitting the Mössbauer spectra. The values of the isomer shift were referred to an α-Fe foil sample at room temperature. The effects of non-ideal absorber thickness and variable recoil-free fractions for iron atoms in non-equivalent structural sites of S5 different phases were expected to be within experimental error (hyperfine parameters ±0.02 mm s -1 , relative spectral area ±2%).

X-Ray Photoelectron Spectroscopy
Chemical states of major elements and their distribution on the particle surface were evaluated by Xray photoelectron spectroscopy (XPS) with a PHI 5000 VersaProbe II XPS system (Physical

Fe, Fe 0 , and N Content
The total Fe content in the FexN particles was determined on an HNO3-digested sample using electrothermal atomic absorption spectrometry (AAS) with a graphite furnace (ContrAA 600, Analytik Jena AG, Germany) equipped with a high-resolution Echelle double monochromator (spectral bandwidth, 2 pm at 200 nm). A xenon lamp was used as a continuum radiation source and an absorption line at 248.3 nm was used for quantification. The total N content was determined on a dry and homogenized sample using an Elemental Vario MACRO CHNS analyzer (Germany), with a laboratory-determined relative standard deviation of ± 0.8%. Before the analysis, the instrument calibration was adjusted to the daily ambient conditions using daily factor determination with the calibration standard sulfanilamide. The Fe 0 content in FexN particles was determined using the volume of hydrogen produced after adding an excess of 1:1 mixture of HCl (37%) and ultrapure water to the particle suspensions. The Fe 0 content was calculated using the measured volume of the evolved hydrogen and the molar volume of an ideal gas at laboratory temperature. The reported values are the average value of three replicates.

Electron Microscopy
The particle morphology was examined by a combination of transmission electron microscopy (TEM) and scanning electron microscopy (SEM). For routine particle size/morphology observations, we used a TEM JEOL 2100 instrument equipped with X- MaxN

Water Contact Angle Measurement
13 mm (diameter) by 2 mm (thick) pellets of both nitrided particle types were prepared using an H-62 laboratory hydraulic press (Trystom, Czech Republic). The pellets were dried in a vacuum oven at 60 °C for 8 h and then the pressure was gradually equalized with air over 10 minutes. The water contact angle of the vacuum-dried pellets was determined by axisymmetric drop shape analysis on a Krüss DSA 30 (Krüss GmbH, Germany) instrument. All measurements were performed with 2 μL of purified water (Millipore, USA) with a conductivity of 0.056 μS cm −1 and repeated 5× at 21 °C.

Brunauer-Emmett-Teller (BET) Specific Surface Area
The Brunauer-Emmett-Teller (BET) surface area was determined on dried samples using a NOVA

Agglomerate Size Distribution
A 125 mg L -1 particle suspension in MHW was characterized by laser diffraction analysis (Mastersizer 2000, Malvern Instruments, U.K.) equipped with the Hydro 2000S dispersion unit. The speed of the pump and stirrer was set to 1 750 rpm and the ultrasonic level to 100%. Samples were analyzed in triplicate with an acquisition time of 60 s.

Dissolved Nitrogen-Containing Species
Dissolved ammonia (NH3(aq)) and other nitrogen species (NO2and NO3 -) originating from FexN particle leaching and HCl digestion were determined using distillation and titration method (ISO 5664) and liquid chromatography (ISO 10304-1), respectively, after separation of the solids with a magnet and filtration using 0.1 µm PTFE syringe filter.

Text S3. TCE Dechlorination Experiments
Batch experiments with fresh particles were conducted in 42-mL glass vials capped with PTFE-lined were then placed on a horizontal shaker (125 rpm) at 22 ± 1 °C. An aliquot (25-100 µL) of headspace gas was periodically withdrawn using a gastight syringe and analyzed using a 7890B gas chromatograph (GC, Agilent Technologies, USA) for the amount of TCE, its C2-degradation products, and hydrogen for three weeks with a previously established method. 8 Control experiments with pristine nZVI were performed in parallel. A control experiment with fresh core-shell S-nZVI particles was also performed. The S-nZVI was synthesized according to our previous study with a S/Fe mass ratio of 0.94/100, which exhibited the highest electron efficiency and > 90% sulfur incorporation during synthesis. 8 Reactivity experiments with fresh and aged particles were done in four and two replicates, respectively. Measured amounts of analytes were compensated for overpressure and sampling-induced headspace losses as described elsewhere. 8,9 Total concentrations of all analytes, S8 observed pseudo-first-order reaction rate constants of TCE removal, kobs, and initial surface-area normalized rate constant, kSA, were calculated according to ref. 8  The adsorption energies of TCE and H2O were calculated as

S10
where Ecomplex is the total electronic energy of the relaxed complex between TCE/H2O and the corresponding slab, and Eslab and Emol are total electronic energies of the relaxed slab and the isolated TCE/H2O molecule, respectively. The structure optimization of physisorbed TCE was performed for eight initial configurations on the γ′-Fe4N slab and four initial configurations on the α-Fe slab (containing TCE in different orientations at ca. 5 Ǻ above the slab) in two steps: first with the GADGET code 29 and subsequently with the conjugate-gradient algorithm as implemented in the VASP package. Only the physisorbed TCE-γ′-Fe4N surface complex with the lowest energy is further discussed in this work. In all calculations, the full geometry relaxation of α-Fe surfaces with TCE led to spontaneous dechlorination of TCE to chloroacetylene. The structure of TCE chemisorbed onto the γ′-Fe4N surface was optimized with the conjugate-gradient algorithm as implemented in the VASP package. The structure optimization of adsorbed H2O was performed also in two steps as described above with four initial configurations for each slab. The effect of solvent on the TCE adsorption was included in the calculations by using an implicit solvation model developed for solid-liquid interfaces 30 and implemented to the VASP program (VASPsol). 31,32 This approach is similar to quantum mechanical polarizable continuum models developed for molecules in solvents. 33 The VASPsol calculations were performed as single-point calculations on the gas phase optimized geometries.
The energy barriers of the γ′-Fe4N-mediated dechlorination of TCE were computed with the climbing image nudged elastic band (CI-NEB) method. 34 A total of seven intermediate images (excluding the initial reactant and final product structures) were used. Note that the calculation of energy barriers for the α-Fe-mediated TCE dechlorination was not possible due to spontaneous TCE dissociation during the geometry optimization, as discussed above. The transition states were verified by frequency analysis at the same computational level as used for geometry optimization, which showed exactly one imaginary frequency corresponding to the change in geometry along the reaction path.
The energy barrier (E ‡ ) was calculated as relative to the energy of the reactant (Ereact) where ETS is the energy of the transition state. Computed structures were visualized using VESTA 3. 35 The calculated homolytic bond dissociation energy (BDE) of the C-Cl bond in the isolated TCE molecule in the gas phase served as a reference energy barrier for the non-facilitated dechlorination reaction. To account for an artificial effect of the periodic boundaries on calculated BDE, we S11 performed these calculations with both VASP and the program TURBOMOLE 36 using the same PBE-D3 method and def2-TZVP basis set. 37,38 To compare the effect of solvent on the BDE energies, we performed also TURBOMOLE calculations involving a continuum solvation model, particularly the Conductor-like Screening Model (COSMO). 39 The results presented in Table S16 showed only a small difference between the BDEs calculated with VASP and TURBOMOLE and the same trend for different Cl atoms. Table S4. Rietveld XRD quantification of Fe-containing mineral phases in the XRD profiles of fresh and three-month aged suspensions of FexN and nZVI particles.  2  Table S5. Values of the Mössbauer hyperfine parameters, derived from the least-square fitting of the Mössbauer spectra of the fresh and three-month aged suspensions of γ′-FexN and ε-FexN nanoparticles, where δ is the isomer shift, ΔEQ is the quadrupole splitting, Bhf is the hyperfine magnetic field, and RA is the relative spectral area of individual spectral components identified during fitting. The spectra were recorded at 150 K.

Particle type Component
1.02 ± 0.12 --------* Calculated from the linear portion of the hydrogen evolution curve (t < 9 days). † Calculated from the amount of hydrogen evolved over three weeks of reaction. Figure S12. (A) TCE removal by fresh nZVI, FexN, and S-nZVI particles (S/Fe mass ratio 0.94/100); (B) corresponding hydrogen evolution during the TCE degradation experiment. The reactions were carried out at an initial TCE concentration of 20 mg L -1 and particle concentration of 1 g L -1 . S-nZVI was synthesized according to our previous study. 8 Whiskers indicate standard deviation (SD). The y-axis range for aged particles is three times bigger since the volume of the reaction mixture was three times greater than for experiments with fresh particles. Whiskers indicate standard deviation (SD). Table S14. Tentatively identified products at the end of the TCE dechlorination experiment with fresh γ′-FexN nanoparticles using non-target screening. Compound identification was performed in the Agilent MassHunter Unknown Analysis software. Only compounds with a match factor > 90 with the NIST 17 mass spectral library are shown. The non-target analysis was performed with four replicates.

Compound
Match factors 3-methylpent-2-ene 94. 5 Table S16. DFT-calculated homolytic bond dissociation energies (BDE) of C-Cl bonds in the isolated TCE molecule in the gas phase (VASP and TURBOMOLE) and in the solvent (COSMO model as implemented in TURBOMOLE). Energies are in kJ mol -1 .