Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2019 Apr 30;10(10):2487–2492. doi: 10.1021/acs.jpclett.9b00773

Water Ordering on the Magnetite Fe3O4 Surfaces

Eman Zaki †, Zdenek Jakub ‡, Francesca Mirabella †,‡, Gareth S Parkinson ‡, Shamil Shaikhutdinov †,*, Hans-Joachim Freund †
PMCID: PMC6550440  PMID: 31039610

Abstract

graphic file with name jz-2019-007732_0005.jpg

The interaction of water with the most prominent surfaces of Fe3O4, (001) and (111), is directly compared using a combination of temperature-programmed desorption, temperature-programmed low energy electron diffraction (TP LEED), and scanning probe microscopies. Adsorption on the (√2 × √2)R45°-reconstructed surface of Fe3O4(001) is strongly influenced by the surface reconstruction, which remains intact at all coverages. Close to the completion of the first monolayer, however, the ad-layer adopts a longer-range (2 × 2) superstructure. This finding is discussed in the context of a similar (2 × 2) superstructure recently observed on the (111) facet, which exists over a significantly larger range of temperatures and coverages. In both cases, the long-range order is evidence that water–water interactions exert a significant influence on the structure already prior to the nucleation of the second layer. We conclude that the stability differences stem from the smaller unit cell on the (111) surface, and the ability of water to more easily form stable hexagonal ice-like structures on the hexagonal substrate.

The interaction of water with oxides, in particular with iron oxides, plays an important role in geology, electrochemistry, corrosion, and catalysis.1,2 Magnetite (Fe3O4) is one of the most explored oxide materials that have been studied using a “surface-science” approach.3−5 Prepared either as single crystals or thin films, the (111) and (001) surfaces of Fe3O4 were investigated intensively, primarily by scanning tunneling microscopy (STM) and low energy electron diffraction (LEED), in combination with density functional theory (DFT) calculations.4,6−12 For the (111) surface (Figure 1b), the results favor the single metal surface termination, i.e., the outmost layer consists of the tetrahedrally coordinated Fetet ions placed with a (2 × 2) periodicity in the 3-fold hollow sites over the close packed O-layer. (Since this is a bulk-truncation at the Fetet layer, we will refer to it as (1 × 1) in this Letter.) As far as the Fe3O4(001) surface is concerned, it reconstructs upon preparation in ultrahigh vacuum (UHV) into a (√2 × √2)R45° structure, which was debated in the literature for a while.4 To date, it is generally accepted that the surface reconstruction originates from an ordered array of subsurface iron vacancies and interstitials (Figure 1c).9 The subsurface rearrangement distorts the surface lattice and causes two surface oxygen atoms without a subsurface Fetet neighbor (labeled with a yellow star in Figure 1c) to be significantly more reactive than the others. Adsorbates including atomic H and various metals have been found to bind strongly in this location.4

Figure 1.

Figure 1

Open in a new tab

(a) Unit cell of Fe3O4 has the spinel structure. The lattice is based on a close packed oxygen lattice with octahedrally coordinated Fe (Feoct) and tetrahedrally coordinated (Fetet) cations. (b) Fe3O4(111) surface is terminated by Fetet cations above a close packed layer of oxygen. (c) Fe3O4(001) surface terminated with an intact layer of oxygen and Feoct atoms. However, a rearrangement of the subsurface cations distorts the structure leading to a (√2 × √2)R45° superstructure. The surface oxygen atoms marked with a yellow star differ from the others as they have no subsurface Fetet neighbor.

There are several experimental and theoretical studies of water adsorption on these two surfaces that were discussed within existing structural models (see refs (13−18) and references therein). Our recent temperature-programmed desorption (TPD), single crystal adsorption calorimetry, and infrared reflection absorption spectroscopy (IRAS) results, corroborated by DFT calculations, provided compelling evidence that water first dissociates on the Fe3O4(111) surface, ultimately resulting in two surface hydroxyls. These then act as an anchor for another water molecule, resulting in a half-dissociated complex (“dimer”). At further increasing coverage, but before the formation of an amorphous solid water (ASW) film sets in, the water ad-layer forms an ordered Fe3O4(111)-(2 × 2) structure, as directly observed by LEED.15,16 The observation of water ordering was, in fact, inspired by the TPD spectra showing sharp desorption peaks at relatively high coverages (Figure 2a). (NB: In these experiments, water was dosed at 140 K solely to minimize the contribution of the ASW signal in TPD spectra.) All three peaks followed first order desorption kinetics indicating the presence of water molecules having discrete binding energies and desorbing almost simultaneously at certain temperature (i.e., 200, 220, and 255 K), hence implying a certain degree of ordering. LEED movies recorded while heating the sample with the same heating rate as for TPD measurements (henceforth referred to as temperature-programmed (TP) LEED)) allowed to link water ordering and the corresponding temperature, hence water coverage. This is illustrated in Figure 2b, where the intensity of the (2 × 2) superstructure spots is shown as a function of temperature together with the water desorption profile. Clearly, the (2 × 2) structure only forms after water desorbs via the peak at 200 K. Importantly, its formation only depends on water coverage and not on the adsorption temperature: the (2 × 2) LEED pattern appears on samples kept at all temperatures between 200 and 250 K at water exposures resulting in coverages above some threshold, although close to saturation coverage at this temperature. Therefore, the observed long-range ordering is thermodynamically driven and not kinetically limited and only occurs at average coverages between 1.2 and 1.8 ML as determined by calibration of a mass spectrometer.15 (Here, 1 ML is defined as one H2O molecule per Fe3O4(111) unit cell exposing one Fe ion, i.e., 3.2 × 1014 at cm–2.)

Figure 2.

Figure 2

Open in a new tab

(a) TPD spectra of D2O (20 amu) adsorbed on the Fe3O4(111) surface at 140 K at increasing exposures. The heating rate was 3 K/s. (b) Intensity of the selected diffraction spots taken from the TP LEED movie (at 64 eV) on the sample exposed to 1 L of D2O at 140 K. The snapshot at 220 K is shown in the inset, and the unit cells are indicated. The 1 L TPD spectrum is shown as dashed line.\

In addition, Figure 2b shows integral intensities of selected (0,1) and (0,1/2) spots, which mostly characterize the O- and Fe-layers in Fe3O4(111), respectively. In principle, the intensity of a diffraction spot depends on the incident electron energy. The so-called I–V curves for several different spots are used for a quantitative analysis of ordered surface structures. In fact, it was the LEED analysis that provided compelling evidence for the currently accepted models of both the Fe3O4(111)6,8 and Fe3O4(001)-(√2 × √2)R45° surfaces.9 Basically, the surface geometry determines the positions of the peaks in I–V curves, whereas their intensities additionally include inelastic losses and thermal vibrations. Obviously, the adsorbate layer may alter the I–V curve, both peak position and its intensity, thus leading to intensity changes at the energy monitored. Accordingly, thermal desorption results in reverse changes. Figure 2b shows that the intensity of the (0,1) spot gradually increases upon heating. This can reasonably be explained by water desorption from the surface. Meanwhile, the intensity of the (0,1/2) spot, related to the Fe sublattice and mostly to the Fe topmost layer, exhibits nonmonotonic behavior and somewhat “mirrors” the TPD curve. It rapidly decreases with the onset of the water desorption and starts to increase at temperatures above 250 K, following recombinative desorption of hydroxyl species that are only remaining at the surface at these temperatures. Clearly, the intensity drop obtained at 150–250 K is linked to the shift of the I–V curve in the region close to the energy used here (64 eV), which suggests substantial changes in the Fetet–O interlayer distance at the oxide surface. This is likely due to the bond of the terminal hydroxyl to the Fetet ion. To date, acquisition of the whole I–V data set during sample heating, which would shed more light on the TP LEED results, remains technically challenging.

To examine whether water adsorption on magnetite depends on the surface plane, we have carried out similar experiments on the Fe3O4(001) films grown on Pt(001).12 The prepared films all showed sharp LEED patterns of the Fe3O4(001)-(√2 × √2)R45° surface. Figure 3a shows TPD spectra obtained after D2O exposure at 140 K. Three desorption states at 200, 220, and 240 K are populated with increasing exposure following first-order (or close to) desorption kinetics. The spectra look almost identical to those recently reported by some of the present authors on a Fe3O4(001) single crystal,17 which was additionally characterized by STM to confirm its surface termination.9 Therefore, we may cross-correlate the results obtained in these two studies. For consistency, we label the desorption peaks α, β, etc., as in ref (17). Note that two more desorption states at around 180 K could be resolved in our experiments, which appear between the α and β peaks, henceforth denoted as α*. These were observed on many samples studied, independently of the heating rate and deposition flux. To some extent, they resemble the α′ state mentioned in the previous TPD study on the single crystal surface, which was performed with a heating rate of 1 K/s.17

Figure 3.

Figure 3

Open in a new tab

(a) TPD spectra of D2O (20 amu) adsorbed on the Fe3O4(001)-(√2 × √2)R45° surface at 140 K. The heating rate was 3 K/s. (b) Four snapshots taken from the TP LEED movie (at 47 eV) recorded after the sample was exposed to 1.75 L of D2O at 140 K. The unit cells are indicated. (c) Averaged intensity of the selected (circled) diffraction spots taken from the TP LEED movie (at 47 eV) on the film exposed to 1.75 L. W- and R-spots are the fingerprints of the (2 × 2) water ad-layer and (√2 × √2)R45° reconstructed surface, respectively. The TPD spectrum at 1.75 L exposure is shown as dashed line.

A comparison of the TPD spectra presented in Figures 2a and 3a reveals some similarities and some differences. Beyond the peak at ∼160 K, associated with a “multilayer” ASW film, three distinct desorption peaks are observed on both surfaces. Moreover, the position of the β and γ peaks on the (001) films almost coincide with desorption maxima at 200 and 220 K obtained on the (111) films, all assigned to molecular water adsorption. Substantial differences are found at higher temperatures. While the desorption from the (111) surface exhibits a maximum at 255 K and a long “tail” extending to 400 K showing a second-order desorption kinetics of dissociated species, water desorbs from the Fe3O4(001)-(√2 × √2)R45° surface via the 240 K peak following pseudo first-order desorption kinetics, showing a steep descending edge. This indicates that water adsorption is either nondissociative or that the hydroxyl groups resulting from dissociation do not separate on the surface during the TPD ramp and readily recombine.

As in the case of Fe3O4(111), the sharp desorption signals observed in TPD spectra on the Fe3O4(001) films suggest water ordering. To investigate further, we employed TP LEED, as above. Figure 3b shows four snapshots from the TP LEED movie recorded during heating the sample dosed with 1.75 L of D2O at 140 K. Additional diffraction spots, which are identified as the Fe3O4(001)-(2 × 2) structure, were observed at 175–187 (±3) K (Figure 3c), i.e., in a much narrower temperature region as compared to the Fe3O4(111) surface (i.e., 200–250 K, see Figure 2b). This renders its observation rather difficult and easy to overlook. Indeed, initial water coverage considerably affects the appearance of the (2 × 2) structure. At 1.5 L exposure and below, no such structure was detected, suggesting the water coverage is not sufficient to form it. The (2 × 2) structure is most markedly observed on heating of the sample exposed to 1.75 L water. At 2 L exposure, the (2 × 2) spot intensity is considerably reduced, and it further decreases at increasing exposures from 2.5 to 4, and 5 L. No LEED pattern could be observed on the 6 L sample. Comparison with the corresponding TPD spectra shows that the (2 × 2) pattern only appears after multilayer desorption and disappears before water starts to desorb via the β peak at 200 K (Figure 3c), which is in the region of the α* state. As a result, a temperature window for its observation becomes smaller at higher water coverages since the ASW desorption follows zero-order desorption kinetics, with the peak shifting toward higher temperatures. In comparison, long-range water ordering on the Fe3O4(111) surface occurs at considerably higher temperatures (Figure 2b), and its observation is therefore less sensitive to the initial water coverage above a certain threshold.

Figure 3c also shows a temperature profile for the intensity of the diffraction spots reflecting the reconstructed surface (R-spots), which seems to be even more complex than for the (111) surface. Nevertheless, it is clear that the (√2 × √2)R45° surface reconstruction is maintained on the water covered surface at our conditions. For comparison, previous LEED studies have reported that the reconstruction is lifted, either upon exposure to 2 × 10–6 mbar of H2O at 273 K for 2 min (i.e., ∼300 L) in ref (19) on a single crystal or even to 0.01 L at 165 K in ref (20) on thin films grown on Mo(001). This was interpreted as indirect evidence for water dissociation. Note, however, that the surface termination was not determined precisely in these studies. Turning back to Figure 3c, it is interesting that the intensity of the “reconstruction” spots maximizes simultaneously with the appearance of the (2 × 2) water ad-layer structure.

To shed light on the nature (thermodynamic vs kinetic) of the (2 × 2) structure observed during TP LEED measurements, we also looked at whether it may be obtained immediately upon water exposure. The results only showed faint (2 × 2) diffraction spots at low energies (25 eV) in a small range of water exposures between 1.5 and 2 L. According to the TPD results (Figure 3a), at this coverage the “monolayer” water film completes, and the formation of the ASW film sets in. However, the intensity of the (2 × 2) spots remarkably increases upon heating and subsequent desorption of ASW-related species. This finding suggests that (i) water ordering is promoted by surface diffusion and that (ii) the formation of the ASW film is not the prerequisite for the (2 × 2) ad-layer formation.

The formation of a (2 × 2) water ad-layer structure on the Fe3O4(001)-(√2 × √2)R45° surface observed here by LEED is further substantiated by scanning probe microscopy. Figure 4a shows scanning tunneling microscopy (STM) and noncontact atomic force microscopy (nc-AFM) images for a water coverage close to 6 H2O per (√2 × √2)R45° unit cell (u.c.) (coverage calibration based on the previously reported quantitative TPD and DFT17). The STM image is dominated by rows running in the ⟨110⟩ directions, consistent with molecules being adsorbed at the surface Fe cations. nc-AFM using a CO-functionalized tip achieves superior resolution, and individual protrusions are discernible along the rows, along with additional protrusions located between the rows. The spacing of ∼11.8 Å is the repeat distance of the (√2 × √2)R45° unit cell in this direction, suggesting the water adlayer structure is defined by the surface reconstruction at this coverage. A DFT-based search17 has previously shown that the most stable configuration at 6 H2O/u.c. is a ring-like structure containing both molecular and dissociated water, with four species bound to the surface Fe cations, with a further two molecules stabilized at the bridge sites solely by hydrogen bonds. Figure 4b shows STM and AFM images when the coverage is increased slightly to 7–9 H2O/u.c. Patches of the surface exhibit the (2 × 2) periodicity in STM and nc-AFM, as observed in LEED. Note that an additional protrusion is significantly offset from the center of the (2 × 2) cell. If it were in the center, this would correspond to (√2 × √2)R45° order. We infer that the new structure is not simply formed by the addition of water molecules to the existing 6 H2O/u.c. structure and that additional reconfiguration must take place to optimize the hydrogen bonding network. Together with the TP LEED data, we conclude that the structure of the water adlayer is highly sensitive to the coverage.

Figure 4.

Figure 4

Open in a new tab

STM (top) and nc-AFM (bottom) images of high-coverage water structures on Fe3O4(001) (√2 × √2)R45° surface. (a) At a coverage close to 6 H2O/u.c., the protrusions on the surface exhibit the same (√2 × √2)R45° periodicity as the underlying surface. (b) When additional water is adsorbed (coverage 7–8 H2O/u.c.), an array of bright protrusions with a (2 × 2) ordering emerges.

STM and AFM studies of water adsorption on the (111) surface at low temperatures (to achieve high coverages corresponding to the ordered structure) are still missing. Such studies could verify the DFT-derived model of the water ad-layer, which is based on the cooperative formation of two-dimensional network of H-bonded dimers.15 Nonetheless, in the light of the (2 × 2) structure formation, we mention previously reported STM images21 obtained at room temperature on an Fe3O4(111) single crystal exposed to ∼100 L at 110 K, which showed random short rows of water related protrusions separated by 12 Å, i.e., double the surface lattice constant. In addition, our own room-temperature STM images of Fe3O4(111) films revealed water-related protrusions often showing a 12 Å periodicity (see Supporting Information). Therefore, it cannot be excluded that such structures, already formed in the initial stages of water adsorption, behave as the template for the long-range water ordering observed by LEED only at considerably higher coverages.

All in all, our results show that water forms ordered structures on both the Fe3O4(111) and the Fe3O4(001)-(√2 × √2)R45° surfaces. Although ordered structures have been reported for a few water/oxide interfaces,22−25 our study is the first to demonstrate the same phenomena on two major surfaces of the same oxide. While qualitative analysis of the LEED pattern allows us to identify structures with a surface symmetry different from the underlying oxide surface, we note that the intensity of the diffraction spots from the bare oxide surface (i.e., the (1 × 1) spots in Figure 2b and the R-spots in Figure 3c) remain almost the same at water coverage close to the onset of the ASW formation. This suggests that the final water adlayer most likely exhibits a “(1 × 1)” ordering at the highest coverages before multilayer growth sets in. Such structure have been predicted by DFT to be stable on the (111) surface.15

The (2 × 2) ad-layer structures observed on these two surfaces differ depending on the coverage and temperature at which they form. Apparently, the (2 × 2) structure on the Fe3O4(001)-(√2 × √2)R45° surface is less stable than on Fe3O4(111) in terms of temperature window, i.e., 175–185 K versus 200–250 K, measured under the same experimental conditions. According to water coverage calibration done in ref (17), the Fe3O4(001)-(2 × 2) structure is observed at a coverage of about 8 H2O/u.c., or 1.2 × 1015 H2O/cm2. For comparison, the Fe3O4(111)-(2 × 2) surface possesses about 1.75 molecules per (1 × 1) unit cell, or 5.6 × 1014 H2O/cm2, i.e., about two times smaller.

Water adsorption on solids is usually rationalized in terms of competition between bonding to a (metal, oxide) surface and intermolecular H-bonding. Strong adsorption is often accompanied by water dissociation as in the case of Fe3O4(111), for which both microcalorimetry and TPD results measured the adsorption energy about 100 kJ/mol.16 Further adsorption occurs essentially on the hydroxyl precovered surface and results in half-dissociated dimers. According to the DFT calculations, H-bonding between water dimers further stabilizes the structure through the formation of a two-dimensional network, ultimately completing the first layer (monolayer). In top view, the network structure resembles the basal plane of hexagonal ice, observed on several (111)-oriented noble metal surfaces.26−28 Note that on metal substrates the ice-like layer formation occurs from the onset as water adsorbs weakly, at least at low temperatures, so that H-bonding dominates in the formed structure.

According to the TPD results and also microcalorimetry data (to be published), initial water adsorption on the Fe3O4(001)-(√2 × √2)R45° surface is much weaker, about 70 kJ/mol, suggesting that water may adsorb molecularly at first. This would be consistent with the most recent DFT calculations,17 which utilize the subsurface cation vacancy model for the surface structure. Isolated molecules were predicted to be slightly favored (by about 5 kJ/mol) over a dissociated molecule (Eads = −57 kJ/mol). Moreover, the DFT analysis favors a partially dissociated water dimer (and trimer) when two (three) water molecules meet at the surface. Partial dissociation causes even stronger interaction with the metal cations such that the formation of a hexagonal ice layer on this surface seems unfavorable due to the square symmetry of the underlying oxide surface. Instead, a ring-like structure covering the surface follows the periodicity of the surface, suggesting water bonding to the metal dominates over H-bonding. According to the DFT-derived structures at the highest coverages, where (2 × 2) ordering is observed (see Figure 4), many water molecules coordinate without forming direct bonds to a substrate. It seems, therefore, that the observed ordered structure is metastable and intermediate between mono- and multilayer water films and forms when hydrogen bonds begin to dominate over the water–surface interaction.

In contrast, the formation of ice-like layer is rather natural on the Fe3O4(111) surface having a proper surface symmetry. However, recent DFT calculations suggest that neither the symmetry of the OH patterns nor the similarity between a substrate and ice correlate well with the ice nucleation ability.29 Instead, the OH density and the substrate–water interaction strength can be used as descriptors for ice to nucleate on a solid surface. Our results may shed some light on this issue.

Experimental Section

For studying reactivity of magnetite surfaces, we made use of well-ordered Fe3O4(111)6,7,30 and Fe3O4(001)12 thin films grown on Pt(111) and Pt(001), respectively. The experiments were performed in several ultrahigh vacuum (UHV) chambers (background pressures were below 2 × 10–10 mbar) in FHI (Berlin) and TU (Wien). In Berlin, the setup was equipped with standard facilities necessary to grow oxide films. TPD spectra were recorded using a quadrupole mass spectrometer (Hiden 301 3F) having a gold-plated cone shield in order to minimize signals from the heating stage. Water (D2O, Sigma-Aldrich) was dosed using a directional doser. All samples were flashed in UHV to 900 K prior to water exposure.

The STM/nc-AFM experiments were performed on Fe3O4(001) single crystals of natural origin (SurfaceNet GmbH). The samples were prepared by cycles of sputtering (1 keV Ar+, 10 min) and annealing (900 K in UHV, with every other annealing step in a partial pressure of 1 × 10–6 mbar O2 for 20 min). The STM image shown in Figure 4a and all nc-AFM images were acquired at 78 K using an Omicron LT STM equipped with a qPlus sensor. The STM image shown in Figure 4b was acquired in a different vacuum system using an Omicron LT STM with a standard STM setup. In both cases, sample preparation and water exposure were performed in a separate preparation chamber with a base pressure below 1 × 10–10 mbar. Water (H2O) was exposed by backfilling the chamber using a high precision leak valve with the sample held at 150 K. Functionalization of the AFM tip with CO was performed in the manner described in ref (17).

Acknowledgments

This work has been supported by the Deutsche Forschungsgemeinschaft through SFB 1109 and by the Fonds der Chemischen Industrie. E.Z. thanks the International Max-Planck Research School “Functional Interfaces in Physics and Chemistry” for a fellowship. G.S.P. and Z.J. gratefully acknowledge funding from the Austrian Science Fund FWF (START-Prize Y 847-N20) and the Doctoral College TU-D (to Z.J.).

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.9b00773.

  • STM images recorded at 300 K of water-contaminated Fe3O4(111) surfaces (PDF)

The authors declare no competing financial interest.

Supplementary Material

jz9b00773_si_001.pdf (501.1KB, pdf)

References

  1. Henderson M. A. The Interaction of Water With Solid Surfaces: Fundamental Aspects Revisited. Surf. Sci. Rep. 2002, 46, 1–308. 10.1016/S0167-5729(01)00020-6. [DOI] [Google Scholar]
  2. Thiel P. A.; Madey T. E. The Interaction of Water With Solid Surfaces: Fundamental Aspects. Surf. Sci. Rep. 1987, 7, 211–385. 10.1016/0167-5729(87)90001-X. [DOI] [Google Scholar]
  3. Weiss W.; Ranke W. Surface Chemistry and Catalysis on Well-Defined Epitaxial Iron-Oxide Layers. Prog. Surf. Sci. 2002, 70, 1–151. 10.1016/S0079-6816(01)00056-9. [DOI] [Google Scholar]
  4. Parkinson G. S. Iron Oxide Surfaces. Surf. Sci. Rep. 2016, 71, 272–365. 10.1016/j.surfrep.2016.02.001. [DOI] [Google Scholar]
  5. Kuhlenbeck H.; Shaikhutdinov S.; Freund H.-J. Well-Ordered Transition Metal Oxide Layers in Model Catalysis – A Series of Case Studies. Chem. Rev. 2013, 113, 3986–4034. 10.1021/cr300312n. [DOI] [PubMed] [Google Scholar]
  6. Ritter M.; Weiss W. Fe3O4(111) Surface Structure Determined by LEED Crystallography. Surf. Sci. 1999, 432, 81–94. 10.1016/S0039-6028(99)00518-X. [DOI] [Google Scholar]
  7. Shaikhutdinov S. K.; Ritter M.; Wang X. G.; Over H.; Weiss W. Defect Structures on Epitaxial Fe3O4(111) films. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 60, 11062–11069. 10.1103/PhysRevB.60.11062. [DOI] [Google Scholar]
  8. Sala A.; Marchetto H.; Qin Z. H.; Shaikhutdinov S.; Schmidt T.; Freund H. J. Defects and Inhomogeneities in Fe3O4(111) Thin Film Growth on Pt(111). Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 86, 155430. 10.1103/PhysRevB.86.155430. [DOI] [Google Scholar]
  9. Bliem R.; McDermott E.; Ferstl P.; Setvin M.; Gamba O.; Pavelec J.; Schneider M. A.; Schmid M.; Diebold U.; Blaha P.; Hammer L.; Parkinson G. S. Subsurface Cation Vacancy Stabilization of the Magnetite (001) Surface. Science 2014, 346, 1215–1218. 10.1126/science.1260556. [DOI] [PubMed] [Google Scholar]
  10. Novotny Z.; Mulakaluri N.; Edes Z.; Schmid M.; Pentcheva R.; Diebold U.; Parkinson G. Probing the Surface Phase Diagram of Fe3O4(001) Towards the Fe-rich Limit: Evidence for Progressive Reduction of the Surface. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 195410. 10.1103/PhysRevB.87.195410. [DOI] [Google Scholar]
  11. Pentcheva R.; Moritz W.; Rundgren J.; Frank S.; Schrupp D.; Scheffler M. A Combined DFT/LEED-Approach for Complex Oxide Surface Structure Determination: Fe3O4(001). Surf. Sci. 2008, 602, 1299–1305. 10.1016/j.susc.2008.01.006. [DOI] [Google Scholar]
  12. Davis E. M.; Zhang K.; Cui Y.; Kuhlenbeck H.; Shaikhutdinov S.; Freund H.-J. Growth of Fe3O4(001) Thin Films on Pt(100): Tuning Surface Termination With an Fe Buffer Layer. Surf. Sci. 2015, 636, 42–46. 10.1016/j.susc.2015.02.004. [DOI] [Google Scholar]
  13. Dementyev P.; Dostert K.-H.; Ivars-Barceló F.; O’Brien C. P.; Mirabella F.; Schauermann S.; Li X.; Paier J.; Sauer J.; Freund H.-J. Water Interaction with Iron Oxides. Angew. Chem., Int. Ed. 2015, 54, 13942–13946. 10.1002/anie.201506439. [DOI] [PubMed] [Google Scholar]
  14. Li X.; Paier J. Adsorption of Water on the Fe3O4(111) Surface: Structures, Stabilities, and Vibrational Properties Studied by Density Functional Theory. J. Phys. Chem. C 2016, 120, 1056–1065. 10.1021/acs.jpcc.5b10560. [DOI] [Google Scholar]
  15. Mirabella F.; Zaki E.; Ivars-Barcelo F.; Li X.; Paier J.; Sauer J.; Shaikhutdinov S.; Freund H.-J. Cooperative Formation of Long-Range Ordering in Water Ad-layers on Fe3O4(111). Angew. Chem., Int. Ed. 2018, 57, 1409–1413. 10.1002/anie.201711890. [DOI] [PubMed] [Google Scholar]
  16. Zaki E.; Mirabella F.; Ivars-Barceló F.; Seifert J.; Carey S.; Shaikhutdinov S.; Freund H.-J.; Li X.; Paier J.; Sauer J. Water Adsorption on the Fe3O4(111) surface: Dissociation and Network Formation. Phys. Chem. Chem. Phys. 2018, 20, 15764–15774. 10.1039/C8CP02333F. [DOI] [PubMed] [Google Scholar]
  17. Meier M.; Hulva J.; Jakub Z.; Pavelec J.; Setvin M.; Bliem R.; Schmid M.; Diebold U.; Franchini C.; Parkinson G. S. Water Agglomerates on Fe3O4(001). Proc. Natl. Acad. Sci. U. S. A. 2018, 115, E5642–E5650. 10.1073/pnas.1801661115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Parkinson G. S.; Novotný Z.; Jacobson P.; Schmid M.; Diebold U. Room Temperature Water Splitting at the Surface of Magnetite. J. Am. Chem. Soc. 2011, 133, 12650–12655. 10.1021/ja203432e. [DOI] [PubMed] [Google Scholar]
  19. Mulakaluri N.; Pentcheva R.; Wieland M.; Moritz W.; Scheffler M. Partial Dissociation of Water on Fe3O4(001): Adsorbate Induced Charge and Orbital Order. Phys. Rev. Lett. 2009, 103, 176102. 10.1103/PhysRevLett.103.176102. [DOI] [PubMed] [Google Scholar]
  20. Liu S.; Wang S.; Li W.; Guo J.; Guo Q. Water Dissociation on Magnetite (001) Films. J. Phys. Chem. C 2013, 117, 14070–14074. 10.1021/jp403799n. [DOI] [Google Scholar]
  21. Cutting R. S.; Muryn C. A.; Vaughan D. J.; Thornton G. Substrate-termination and H2O-coverage Dependent Dissociation of H2O on Fe3O4(111). Surf. Sci. 2008, 602, 1155–1165. 10.1016/j.susc.2008.01.012. [DOI] [Google Scholar]
  22. He Y.; Tilocca A.; Dulub O.; Selloni A.; Diebold U. Local Ordering and Electronic Signatures of Submonolayer Water on Anatase TiO2(101). Nat. Mater. 2009, 8, 585–589. 10.1038/nmat2466. [DOI] [PubMed] [Google Scholar]
  23. Meyer B.; Marx D.; Dulub O.; Diebold U.; Kunat M.; Langenberg D.; Wöll C. Partial Dissociation of Water Leads to Stable Superstructures on the Surface of Zinc Oxide. Angew. Chem., Int. Ed. 2004, 43, 6641–6645. 10.1002/anie.200461696. [DOI] [PubMed] [Google Scholar]
  24. Włodarczyk R.; Sierka M.; Kwapień K.; Sauer J.; Carrasco E.; Aumer A.; Gomes J. F.; Sterrer M.; Freund H.-J. Structures of the Ordered Water Monolayer on MgO(001). J. Phys. Chem. C 2011, 115, 6764–6774. 10.1021/jp200112c. [DOI] [Google Scholar]
  25. Zhao X.; Shao X.; Fujimori Y.; Bhattacharya S.; Ghiringhelli L. M.; Freund H.-J.; Sterrer M.; Nilius N.; Levchenko S. V. Formation of Water Chains on CaO(001): What Drives the 1D Growth?. J. Phys. Chem. Lett. 2015, 6, 1204–1208. 10.1021/acs.jpclett.5b00223. [DOI] [PubMed] [Google Scholar]
  26. Guo J.; Bian K.; Lin Z.; Jiang Y. Perspective: Structure and Dynamics of Water at Surfaces Probed by Scanning Tunneling Microscopy and Spectroscopy. J. Chem. Phys. 2016, 145, 160901. 10.1063/1.4964668. [DOI] [PubMed] [Google Scholar]
  27. Tatarkhanov M.; Ogletree D. F.; Rose F.; Mitsui T.; Fomin E.; Maier S.; Rose M.; Cerdá J. I.; Salmeron M. Metal- and Hydrogen-Bonding Competition During Water Adsorption on Pd(111) and Ru(0001). J. Am. Chem. Soc. 2009, 131, 18425–18434. 10.1021/ja907468m. [DOI] [PubMed] [Google Scholar]
  28. Michaelides A.; Morgenstern K. Ice Nanoclusters at Hydrophobic Metal Surfaces. Nat. Mater. 2007, 6, 597. 10.1038/nmat1940. [DOI] [PubMed] [Google Scholar]
  29. Pedevilla P.; Fitzner M.; Michaelides A. What Makes a Good Descriptor for Heterogeneous Ice Nucleation on OH-patterned Surfaces. Phys. Rev. B: Condens. Matter Mater. Phys. 2017, 96, 115441. 10.1103/PhysRevB.96.115441. [DOI] [Google Scholar]
  30. Weiss W.; Ritter M. Metal Oxide Heteroepitaxy: Stranski-Krastanov Growth for Iron Oxides on Pt(111). Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 59, 5201–5213. 10.1103/PhysRevB.59.5201. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

jz9b00773_si_001.pdf (501.1KB, pdf)

Articles from The Journal of Physical Chemistry Letters are provided here courtesy of American Chemical Society

RESOURCES