Water Agglomerates on Fe3O4(001)
Matthias Meier, Jan Hulva, Zdenek Jakub, Jiri Pavelec, Martin Setvin, Roland Bliem, Michael Schmid, Ulrike Diebold, Cesare Franchini, Gareth S. Parkinson
WWater Agglomerates on Fe O (001) Matthias Meier,
Jan Hulva, Zdenĕk Jakub, Ji Pavelec, r rı ı Martin Setvin, Roland Bliem , Michael Schmid, UlrikeDiebold, Cesare Franchini, and Gareth S. Parkinson Institute of Applied Physics, Technische Universität Wien, Vienna, Austria University of Vienna, Faculty of Physics and Center for Computational Materials Science, Vienna, Austria
Determining the structure of water adsorbed on solidsurfaces is a notoriously difficult task, and pushes thelimits of experimental and theoretical techniques.Here, we follow the evolution of water agglomerateson Fe O (001); a complex mineral surface relevant inboth modern technology and the natural environment.Strong OH-H O bonds drive the formation of partially-dissociated water dimers at low coverage, but a surfacereconstruction restricts the density of such species toone per unit cell. The dimers act as an anchor forfurther water molecules as the coverage increases,leading first to partially-dissociated water trimers, andthen to a ring-like, hydrogen-bonded network thatcovers the entire surface. Unraveling this complexityrequires the concerted application of several state-of-the-art methods. Quantitative temperatureprogrammed desorption (TPD) reveals the coverage ofstable structures, monochromatic x-ray photoelectronspectroscopy (XPS) shows the extent of partialdissociation, and non-contact Atomic Force Microscopy(AFM) using a CO-functionalized tip provides a directview of the agglomerate structure. Together, these dataprovide a stringent test of the minimum energyconfigurations determined via a van der Waals densityfunctional theory (DFT)-based genetic search.
The ubiquity of water in the ambient environmentensures that its interaction with solid surfaces is offundamental importance [1]. To understand processes such as dissolution, corrosion, and weathering at themolecular level requires an understanding of how wateradsorbs on surfaces, and what governs their reactivity.Atomic-scale investigations on single-crystal sampleshave revealed that interfacial water almost never formsan ice-like structure [2], and aims to simultaneouslymaximize its interaction with the surface and inter-molecular hydrogen bonding (H-bond). The surface andH- bonds have similar magnitude on metals, and theadlayer is stabilized if some fraction of the waterdissociates, allowing the formation of strong H O-OH H-bonds [2].The situation is somewhat different on metal oxidesbecause the bonds to the surface dominate. The lonepair on the oxygen atom forms a dative bond with theelectron-deficient cation sites, while on more reactivesurfaces, dissociation gives rise to two distinct hydroxylgroups (terminal O water
H and surface O surface
H). Theenergetic difference between molecular and dissociativeadsorption can be extremely small, and some mixture isinevitably observed in equilibrium at finitetemperatures [3]. There is, however, increasing evidencethat partially-dissociated adlayers can also represent thelowest-energy configuration on metal oxide surfaces [4–6], and partially-dissociated water dimers have beenrecently proposed to be the most stable water specieson both RuO and Fe O (111) [7–9]. A key issue for the understanding of water adlayers hasbeen the difficulty of achieving molecular resolution ofater clusters and adlayers. While significant progresshas been made using STM in recent years [10], nc-AFMhas emerged as a technique capable of superiorresolution, particularly when the tip is functionalized bya CO molecule [11]. Recently, Shiotari andSugimoto [12] demonstrated spectacular images ofwater clusters adsorbed on Cu(110), and we resolved toapply this method to the particularly complex case ofwater adsorption on the (√2×√2)R45°-reconstructedFe O (001) surface. In combination with quantitativeTPD, high-resolution XPS, and state of the art theory, weare able to determine the evolution of stable waterstructures over the full range from an isolated moleculeto the completion of the first monolayer. Interestingly, although an isolated molecule adsorbsintact, significant energy is gained through the formationof partially-dissociated water dimers. The surfacereconstruction limits the coverage of such species to oneper (√2×√2)R45° unit cell, however, because only asubset of the surface O atoms can accept a proton toform an O surface H group. The partially-dissociated waterdimers act as an anchor for further water as thecoverage increases, leading first to partially-dissociatedwater timers, and then to a ring-like H-bonded network,which covers the entire surface. Interestingly, the nc-AFM images allow us to rule out one of two iso-energetic water trimers predicted by a thorough DFTsearch, and the data indicate that van der Waals DFTdoes not accurately handle the cooperative energybalance in this system.
Results
A key feature of the spectroscopic measurementsdescribed here is the ability to deposit an accuratelydetermined number of water molecules on the Fe O (001) surface using a calibrated molecular-beamsource [13]. Figure 1(A) shows TPD spectra obtained forvarious initial D O coverages ranging from 0 to 14molecules per (√2×√2)R45° unit cell (H O/u.c.). D O wasutilized to ensure that the measured signal originatessolely from the sample surface, but spectra obtained forH O are indistinguishable from those presented in Fig.1(A). A complex spectrum with 7 distinct desorptionfeatures was reproducibly observed from severaldifferent single crystal samples, and we label the peaksα, α’ β, γ, δ, ε and φ in order of ascending temperature.A plot of the integrated peak area versus exposure (Fig.1(B)) yields a straight line, consistent with the measuredsticking probability of unity at all coverages (Fig. S1). Theonset of multilayer ice desorption (peak α at 155 K)occurs for a coverage close to 8.5 molecules per(√2×√2)R45° unit cell (8.5 H O/u.c. = 1.28×10 H O/cm ),which is close to the density of an ice monolayer onclose-packed metal surfaces, and we thus considereverything desorbing at higher temperatures aconstituent of the first water monolayer. The saturationof the φ (550 K) and ε (310 K) peaks (inset, Fig. 1(A))occurs for coverages significantly less than 1 H O/u.c.,and we assign these states to surface defects. Peaks β, γ,and δ saturate at coverages close to 8, 6, and 3 H O/u.c.,respectively, which suggests that stable surface phasesare completed at these coverages. α’ is a small shoulderbetween the saturation of the β peak and the onset ofmultilayer desorption (peak α). 2 igure 1: Quantification of water adsorbed on Fe O (001) byTPD. (A) Experimental TPD spectra obtained for initial D Ocoverages ranging from 0 to 14 molecules per Fe O (001)-(√2×√2)R45° unit cell (inset: higher temperature rangeshowing desorption peaks ε and φ, which originate fromsurface defects). The colored curves indicate the coverages forwhich a particular desorption feature (labeled α’, β, γ, δ)saturates. (B) Plot of the integrated TPD peak areas as afunction of beam exposure. The colored data pointscorrespond to the colored curves in panel (A). Based on thesedata we conclude the β, γ, and δ peaks saturate at coveragesof 8, 6, and 3 molecules per (√2×√2)R45° unit cell, respectively.(C) Inversion analysis of the TPD data for D O on Fe O (001)for the different peaks. The filled area marks the uncertaintyrange of the coverage-dependent desorption energies for eachpeak. To extract information regarding the desorptionenergetics from the TPD data, we performed aninversion analysis [14]. Full details are contained in thesupporting information. Briefly, the analysis assumesthat the desorption follows first-order Arrhenius kinetics,and yields the coverage-dependent activation energy fordesorption, E d (Fig. 1(C)) by direct inversion of the well-known Polanyi-Wigner equation. The uncertainty in E d isrelated to the uncertainty in the pre-exponential factor, ν , which is unknown, but optimized during the analysis [14]. The resulting E d is equivalent to theadsorption energy ( E ad ) if the adsorption is a reversibleprocess with no activation barrier. The results, shown inFig. 1(C), show that E d decreases with increasingcoverage from a maximum of 0.85 ± 0.05 eV in the low-coverage limit to 0.52 ± 0.05 eV for the first moleculesdesorbing from the first monolayer. Interestingly, thecorresponding values of ν ( ν δ =10 s -1 , ν γ =10 s -1 ,ν β =10 s -1 , ν α’ =10 s -1 ) are relatively high, whichsuggests that the adsorbed state is highlyconstrained [14]. For comparison, utilizing ν = 10 s -1 (appropriate for a 2D gas) in Fig. 1(C) would see alldesorption energies lowered by approximately 0.15 eV. To understand the origin of the complex multi-peakdesorption profile we studied the adsorbed waterstructures with STM and nc-AFM. Figure 2(A) shows anSTM image of the as-prepared Fe O (001) surfaceacquired at 78 K. Rows of protrusions in the [110]direction are due to the octahedrally-coordinatedsurface Fe oct atoms of a stoichiometric surface layer (seesurface model in inset). The surface oxygen atoms arenot imaged because there are no O-related states in thevicinity of the Fermi level. The undulating appearance ofthe Fe oct rows and associated (√2×√2)R45° periodicity(white square) are linked to a subsurface rearrangementof the cation sublattice [15]. We have previously shownthat the O* atoms (i.e. surface oxygen without atetrahedrally coordinated Fe tet neighbor in the secondlayer) are active sites for adsorption. These atoms differelectronically from the others (DFT predicts a smallmagnetic moment [15]), and they stabilize metaladatoms to high temperatures [16]. Crucially for whatfollows, the O* atoms are also preferred sites for theformation of O surface H groups [17,18]. There is always a3mall coverage of O*H following in-situ preparation dueto the reaction of water from the residual gas withoxygen vacancies, and they cause pairs of surface Fe oct protrusions to be imaged slightly brighter in emptystates STM images. An example is highlighted by a whitearrowhead in Fig. 2(A). Recombination of the O*Hspecies with lattice O to desorb water is responsible forthe φ peak (550 K) observed in TPD [19].
Figure 2: Water monomers, dimers and chains on theFe O (001) surface imaged by low-temperature (78 K) STM (A) The as-prepared Fe O (001) surface. The (√2×√2)R45°periodicity is indicated by the white square, and the whitearrow highlights an O*H group. (inset) Top view of theFe O (001)-(√2×√2)R45° surface structure with the subsurfacecation vacancy structure. Only the Fe oct atoms are imaged inSTM. (B) STM image acquired after 0.05 L water adsorbed andheated to 255 K. The surface is clean, except for protrusions located at surface defects including antiphase domainboundaries in the (√2×√2)R45° reconstruction (cyan arrow).(C) STM image following adsorption of 0.1 L water at 120 K.Isolated single protrusions (yellow arrow), double protrusions(red arrow) and longer chains (green arrow) are due to watermolecules adsorbed on the Fe oct rows. To confirm the ε TPD peak at 310 K was defect related,we exposed the as-prepared Fe O (001) surface to 0.05 Lwater, heated to 255 K, and imaged the surface usingSTM. Figure 2(B) shows bright protrusions adsorbed atan antiphase domain boundary in the (√2×√2)R45°reconstruction [20], and there is also evidence foradsorption at Fe related point defects and step edges(Fig. S3). Similar behavior was recently observed formethanol on this surface [21]. In the current paper, weare primarily interested in water adsorbed at regularlattice sites.Figure 2(C) shows an STM image of the Fe O (001)surface after exposure to 0.1 L (1 L = 1.33x10 -6 mbar.s)H O at 120 K. At this temperature, far below thedesorption threshold, surface mobility is low, and weobserve a non-equilibrium state. The image, acquired at78 K, exhibits isolated, bright protrusions on the Fe oct rows due to adsorbed water (yellow arrow). It is notstraightforward to determine whether the molecules areintact or dissociated from this image, but several waterdimers are observed already at this coverage (redarrow). Interestingly, dimers have two apparent heights,so there may be two types of water dimers under theseconditions. 4 igure 3: Imaging water agglomerates on Fe O (001) with nc-AFM using a CO-functionalized tip. Nc-AFM images obtained afterexposing the as-prepared Fe O (001) surface to (A) 2.5 ± 0.5 H O/u.c., (B) ≈6 H O/u.c. and (C) ≈8 H O/u.c. In each case, water wasdosed at 105 K, and the sample preheated to ≈
155 K prior to imaging at 78 K. The coverages in panels (A), (B), and (C)correspond roughly to saturation of the δ, γ, and β peaks in TPD, respectively. Partially-dissociated water dimers and trimers onthe Fe oct rows are indicated by red and cyan arrows in (A), respectively, and yellow arrows highlight protrusions bridging the Fe oct rows in panel (B). Additional water deposited on the surface appears as bright protrusions (yellow star), suggesting it protrudessignificantly above the ring-like structure (C). The (√2×√2)R45° surface unit cell is shown by a white square.
Finally, there are instances of longer water chains (greenarrow), but it is difficult to know how much water isinvolved, and these could simply be two dimers.Nevertheless, the STM data suggest that watermolecules can diffuse already at 120 K, and interactattractively should they meet. STM images of higherwater coverages were acquired (see Supplement),revealing limited additional information. The Fe oct rowsare increasingly occupied by extended protrusions, but itis not possible to resolve the internal structure (Fig. S3). To learn more about water in the sub-monolayer regime,we imaged the surface using nc-AFM. The best imageswere obtained in constant-height mode using a CO-functionalized tip (Fig. 3). This experimental setup wasrecently utilized to image water clusters on differentsurfaces [12,22–24], and the observed image contrastwas attributed to electrostatic interaction between the CO quadrupole field and strongly polar watermolecules [24]. This mechanism provides stable,molecular resolution at relatively large tip-sampledistances, where the tip does not interact with the waterclusters.Figure 3(A) shows an nc-AFM image of the Fe O (001)surface after 2.5 ± 0.5 H O/u.c. H O was adsorbed at 105K. Prior to imaging, the sample was heated to 155 K,which is short of the desorption onset of the δ peak. Theimage, acquired at 78 K, exhibits a bi-modal distributionof double (red arrow) and triple (cyan arrow) protrusionsaligned with the [110] direction, which we assign towater dimers and water trimers, respectively. The brightspots originate from repulsive electrostatic interactionbetween the CO tip and the O atom of the watermolecule or OH group. The distance measured betweenneighboring protrusions within each dimer/trimer is 0.35m, consistent with adsorption at the surface Fe oct cations on the underlying surface (see structural modelin Fig. 2(A)). Again, it is impossible to know from theAFM images alone whether the species within thedimer/trimer are intact or dissociated.
Figure 4: O1s XPS data showing that the water agglomeratesformed on Fe O (001) are partially dissociated. The as-prepares surface exhibits a single peak at 530.1 eV due to thelattice oxygen atoms. The 2.6 D O/u.c. data should becompared with the surface shown in Fig. 3(A), and showsroughly equal contributions from OD and D O, consistent withone dissociated molecule per water dimer/trimer. Most of theadditional water adsorbed at a coverage of 7.7 H O/u.c. ismolecular. Data were measured at 95 K, with monochromaticAl Kα radiation and at a grazing exit of 80° for the emittedphotoelectrons.
Figure 3(B) was acquired after the water coverage wasincreased to ≈6 H O/u.c., and the sample again heated to155 K prior to imaging at 78 K. The image exhibits fullrows of bright protrusions along [110]; four protrusionsare observed per unit cell, consistent with adsorption onall surface Fe oct atoms. In addition, protrusions areobserved in between the rows (yellow arrows). In mostcases, the distance between these bridging protrusionsalong [110] is 1.19 nm, which corresponds to theperiodicity of the (√2×√2)R45° reconstruction in that direction. Finally, when the coverage is increased (Fig.3(C)), the contrast becomes dominated by new features,which protrude further from the surface than the rest ofthe water layer. This suggests that there are additionalstable binding sites available on the 6 H O/u.c. structure,or that the layer restructures above this coverage. To ascertain the chemical state of the water within theadlayers we performed XPS experiments. Figure 4 showsthe O 1s region for the as-prepared Fe O (001) surface(black curve), and after 2.6 (blue curve) and 7.7(magenta curve) D O/u.c. was adsorbed and the samplewas heated to 175 K with a 1 K/s ramp. The as-preparedsurface exhibits a single, slightly asymmetric peak at530.1 eV due to the lattice oxygen [25]. Exposure towater creates a clear peak at 533.4 eV due to D O, whichshifts slightly to lower binding energy with increasingcoverage. Fitting the 2.6 D O/u.c. data with Voigtfunctions, we find that (at least) one additional peak at531.5 eV is required to accurately model the data. Thispeak position is close to that observed previously forO surface
H groups (531.3 eV) [26]. Of course, the XPSbinding energy of O water
H groups could be slightlydifferent, particularly when it is part of an agglomerate,but calculated core level shifts [27] for the O surface
H andthe O water
H of the linear water trimer (see Fig. 5) found adifference of 0.1 eV. Since D O dissociation yields two ODgroups, the similar peak areas at 533.4 eV and 531.5 eVsuggests that approximately half of the D O isdissociated. At the higher coverage, the area of the D Opeak increases significantly, and shifts to lower bindingenergy. The peak area in the OD region remains constantwith respect to the substrate peak (fit not shown), whichsuggests that the additional water adsorbs molecularly. 6o understand the formation of different waterstructures we now turn to our computational results. Asexplained below, we employed a systematic approach todetermine the lowest energy configuration of watermolecules in the coverage regime 0-8 H O/u.c. It isimportant to note that this is not an automated geneticalgorithm, but rather proceeds by identifying factors thatcertain trial structures more stable than others at eachcoverage, and using this information to build subsequentgenerations. A complete account of the theoreticalapproach, and discussion of all structures computed willbe published separately. Selected results relevant to thediscussion here are shown in Fig. 5. Before continuing, itis important to note that our calculations utilize theGGA+U approach (U eff =3.61 eV) [28,29] with the optPBE-DF exchange-correlation (Xc)-functional [30–32] which ismodified to include long-range vdW interactions, andthe so-called subsurface cation vacancy (SCV) model ofFe O (001)-(√2×√2)R45° [15]. Thus, our setup differsmarkedly from the prior work of Mulakuluri et al. [5,33],who utilized a standard GGA+U functional and a bulk-truncated surface model, and only calculated coveragesof 1, 2 and 4 H O/u.c.. Interestingly, we find that an isolated water moleculeprefers to adsorb molecularly on the Fe O (001) surface(Fig. 5(A), E ad =-0.64 eV). The optimum configuration hasthe O atom close to atop a substrate Fe oct cation, withthe molecule in the plane of the surface and orientedsuch that the H atoms interact with nearby surface Oatoms via very weak H bonds, 2–2.2 Å. This configurationis 0.05 eV more stable than a dissociated molecule (E ad =-0.59 eV), where the OH group adsorbs upright atop aFe oct cation, with the proton deposited at theneighboring O* forming a surface hydroxyl (O*H). The most stable configuration of water on the Fe O (001)surface occurs at a coverage of 2 H O/u.c. with theformation of a partially dissociated water dimer (E ad =-0.92 per molecule). This species comprises one terminalOH and one H O, bound to neighboring surface Fe oct atoms along the row, connected by an inter-molecular H-bond (1.41 Å). The H + atom liberated by the dissociationforms an O*H group. Further details of the adsorptiongeometry are included in the discussion section, wherewe explain the cooperative origin of this species’stability. The DFT-based search at 3 H O/u.c. yields two partiallydissociated water trimers degenerate in energy (see Fig.5(A), labeled E and E ). Both species are based on thepartially-dissociated water dimer described above, butdiffer in the location of a third molecule. In the linearH O-OH-H O trimer, the third molecule binds on thesurface Fe oct row, and donates an H-bond into the OH. Inthe alternative non-linear isomer trimer, the third watermolecule binds by H-bonds only. It receives an H-bondfrom the surface O*H, and donates an H-bond to thenearby, unoccupied O* atom. Electrostatic repulsionrenders the adsorption of a proton at both O* sitesenergetically unfavorable at low coverage, and thusdissociation is limited to one molecule per (√2×√2)R45°unit cell. 7 igure 5: Top view of the minimum-energy structuresdetermined by DFT for water coverages of 1, 2, 3, 6 and 8H O/u.c.. (A) An isolated molecule adsorbs intact, but partially-dissociated water dimers and trimers areenergetically preferred. Two partially-dissociated trimerstructures are calculated to be energetically degenerate. Featoms are blue, and O are red. (B) DFT-based model at 6H O/u.c. showing a ring-like structure based on full occupationof the Fe oct rows with OH or H O, and water molecules bridgingthe O* sites. These bridging molecules are adsorbed partlythrough H-bonds to surface O*H groups. The O*H groupsbeneath the adsorbed molecules are shown in the rightmostwhite circle. Alternatively, the structure can be viewed asbased on a pair of H O-OH-H O timers (labeled 1 and 2). (C)DFT-based model at 8 H O/u.c. showing a complex structureutilizing dangling bonds in the 6 H O/u.c. structure to form asecond bridge in the region of the yellow star. All adsorptionenergies are given in eV. The (√2×√2)R45° unit cell and bothO* are highlighted.
The lowest-energy structure determined by our DFTsearch at 6 H O/u.c. exhibits a ring-link appearance, inagreement with the nc-AFM image shown in Fig. 3(B).This is the first coverage at which all adsorbed moleculesare involved in a H-bonded network that covers thesurface. All four Fe oct sites in each (√2×√2)R45° unit cellare occupied by either H O or OH, and the rows arebridged by two further water molecules attached solelythrough H-bonds. In general, the structure ischaracterized by H O-OH-H O trimers, and facilitatesnear-ideal bonding angles of 122-124° for intact watermolecules. Interestingly, the repulsive behavior of thetwo O*H species observed at lower coverage is mitigatedthrough the additional H-bonding with the bridgemolecules. The structure at 8 H O/u.c. is shown in Fig.5(C). It is rather complex, but essentially water utilizesthe remaining dangling H-bonds in the 6 H O/u.c.structure to form a second bridge of molecules near thecenter of the previously ring-like feature (in the vicinity8f the yellow star in Fig. 5(B)). However, additionalreorganisation occurs to optimise the H-bonding,including a modification of the original bridge structureformed at 6 H O/u.c.. Since the coverage at 8 H O/u.c. isalready close to that of a close-packed ice layer, it isstraightforward to understand why further wateradsorption results in the adsorption of multilayer ice. Allstructures shown in Fig. 5 can be downloaded as part ofthe SI.
Discussion
Based on the experimental and theoretical evidencepresented above, we conclude that partially-dissociatedwater dimers are the most stable species on theFe O (001) surface, closely followed by structurallyrelated, partially-dissociated water trimers. Our nc-AFMimages clearly show the adsorbed dimers and trimers,and XPS spectra reveal them to be partially dissociated.Moreover, the theoretically determined adsorptionenergies agree remarkably well with the E d valuesobtained from an inversion analysis of the δ-peak, andthe highly constrained adsorption geometry predicted byDFT is consistent with the high pre-exponential factor ( = 10 s -1 ). For higher coverages, the inversion analysisreveals the E d necessary to desorb the most weaklybound molecule(s), and thus should not be compared tothe average adsorption energies calculated by DFT. Clearly, the (√2×√2)R45° reconstruction plays a crucialrole in the adsorption behavior. At low coverages, thepartially-dissociated water dimers and trimers order with(√2×√2)R45° symmetry, while at high coverages thestructure of the H-bonded network also belies theperiodicity of the underlying substrate. Ultimately, thisstems from the strong preference to form surface O*Hgroups, which limits the density of dimers/trimers to one per unit cell. Later, the O*H groups provide a hydrogenbond to bridge the Fe oct rows and complete the H-bonded network.Despite the importance of the O* sites, the primarycontribution to the adsorption energy at low coveragearises from the Fe -O water bond. An isolated watermolecule binds strongly atop an Fe oct row atom (-0.64eV), and prefers this state to dissociation by 0.05 eV. Thisresult differs from the prior calculations of Mulakuluri etal. [5,33], and after extensive testing, we have found thatthe discrepancy originates in the structural model used,and not the functional applied. As suggested byMulakuluri et al. [33], the Fe cations in the subsurfacelayers of a bulk-truncated structure interact with theadsorbates and promote dissociation. The SCVreconstruction contains only Fe cations in theoutermost 4 layers, and molecular adsorption ispreferred.Given the lack of dissociation in the monomer case, it issomewhat surprising that partially-dissociated waterdimers form on the Fe O (001) surface. Recently,Freund’s group [8] proposed that partial-dissociationrequires two molecules to meet on the Fe O (111)surface, but later revised their IRAS analysis in favor ofthe “traditional picture” where dissociation occurs firstin isolation on an under-coordinated anion-cationpair [34]. Our STM images show that dimerization occursalready at very low coverages on Fe O (001), and there isno evidence for monomer dissociation in the form ofadditional isolated O*H groups. It is, however, difficult toknow if the dimers are molecular or partially dissociatedfrom STM alone. This ambiguity does not exist for highercoverages: nc-AFM images of 2 H O/u.c. (Fig. 3(A)) showdimerization occurs already at 155 K, and analysis of XPS9pectra suggests that roughly one molecule peragglomerate is dissociated. What then, drives the partial dissociation of waterdimers in the water/Fe O (001) system? To answer thisquestion we first analyze the DFT results for a molecularwater dimer (Fig. 6(A)). Somewhat surprisingly, theenergy gain of molecular dimerization is small; anisolated molecule has a binding energy of -0.64 eV, whilethe average binding energy in the molecular dimer is just-0.66 eV per molecule. This difference is significantly lessthan the binding energy of an H-bond in a gas-phasewater dimer (-0.10 eV per molecule). Since the H-bondlength in the present system (1.89 Å) is significantlyshorter than that of a gas-phase water dimer (2.0 Å),some energy must be lost in the final structure. In thisregard, we consider the Fe oct -O water bond lengths. Thewater that donates an H-bond has an Fe oct -O bond of2.20 Å, comparable to the isolated water monomer (2.22Å), but the acceptor molecule has an Fe oct -O bond lengthof 2.34 Å. This suggests that receiving an H-bondweakens the interaction of a water molecule with thesubstrate, consistent with the idea that forming Fe oct -Obonds and receiving H-bonds both involve the lone pair(O 2p) orbitals [35], and that the competition leads tothe marginal total-energy gain.The situation is very different for a partially-dissociatedwater dimer (Fig. 6(B)). Here, the average adsorptionenergy per molecule is -0.92 eV, so the small energeticcost of dissociating one molecule (0.05 eV for an isolatedmonomer) is easily compensated. The intermolecular H-bond is significantly shorter (1.41 Å) in the partially-dissociated water dimer, as expected on electrostaticgrounds (the OH species is negatively charged).Moreover, the H-bond donating water molecule has a significantly shorter Fe oct -O water bond (2.06 eV) than in themolecular water dimer, suggesting a stronger interactionwith the substrate. This phenomenon has been observedin gas-phase water clusters [36], and on metalsurfaces [35,37], and is known as cooperativity [38].Essentially, water molecules seek a balance in their H-bonding interactions. If a molecule donates a strong H-bond, it accepts stronger H-bonds. Since H-bondacceptance utilizes the same O orbital as the Fe oct -O water bond, the balance can be achieved through an enhancedinteraction with the substrate. Thus, the formation ofthe negatively charged OH group induces water todonate a strong H-bond, which in turn induces astronger water-surface interaction. The energy gain is sosubstantial that the system can accommodate aweakened terminal-OH Fe oct -O bond, which is 0.12 Ålonger than for an isolated terminal OH (not shown). Next we turn our attention to the partially-dissociatedwater trimer (Fig. 6(C)). The linear trimer is a naturalconsequence of the arguments outlined above, as thereis an under-coordinated Fe oct atom available to which awater molecule can bind and simultaneously donate anH-bond into the OH group of the partially-dissociateddimer. Of course, the Fe-Fe separation along the Fe oct rows is considerably larger (3 Å) than the sum of OH andH-bond lengths, so the partially dissociated water trimerforms with the OH group directly atop an Fe oct atom,with both water molecules leaning in toward the OHfrom their favored position atop an Fe oct (see Fig. 6). TheH-bonds are slightly longer, as is the OH Fe oct -O bond.This results in the slightly lower adsorption energy of-0.88 eV per molecule compared to the OH-H O waterdimer. 10o this point in the discussion, the DFT search predictswhat is observed in experiment, and allows us tounderstand why the partially-dissociated wateragglomerates form, and are so strongly bound. However,DFT finds a non-linear partially dissociated water trimer(E in Fig. 5(B)) with a comparable adsorption energyto the partial dissociated linear trimer, which is notobserved in the experiments. The fact that we canresolve water bridging the Fe oct rows upon the formationof the H-bonded network with nc-AFM in Fig. 3(B) givesus confidence we could detect the non-linear trimer if itwere present. This indicates that the energy balanceinvolved in adsorbed water agglomerates is not perfectlyhandled by DFT. To investigate further, we comparedseveral alternative functionals, with and without vdWcorrections, and obtained similar trends, albeit with avariation of +/- 0.1 eV in the absolute energies. Weconclude that DFT is insufficiently accurate to predict therelative stabilities of water adlayers, and that calculatingreliable structures requires guidance from experiment orthe adoption of superior approaches such as hybridfunctionals or the random phase approximation [39,40].
Figure 6: The geometry of partially-dissociated water dimersand trimers reveals a cooperative binding effect. (A) Amolecular water dimer exhibits a relatively long inter-molecular H-bond, and the H-bond acceptor has a weakenedinteraction with the surface compared to an isolatedmolecule. (B) The partially-dissociated water dimer exhibits astrong inter-molecular H-bond, and the H-bond-donatingwater molecule binds more strongly to the substrate. (C) In thepartially-dissociated water trimer, as second water moleculedonates an H-bond to the OH group, further weakening itsbond to the substrate. All bond lengths are given in Å andenergies in eV.
Once the partially-dissociated water trimer (Fig. 6(C))forms, it is not possible to H-bond additional water alongthe Fe oct row. The TPD data suggests that the next stable11tructure occurs at a coverage of 6 H O/u.c., where thenc-AFM images (Fig. 3(B)) clearly resolve a ring-likestructure with additional protrusions bridging themolecules adsorbed at the Fe oct rows. This is in line withthe minimum-energy structure determined at 6H O/u.c., which is the first configuration to establish anH-bonded network extending over the whole surface.The stability of the proposed structure stems from thepresence of H O-OH-H O trimers, which utilize H Omolecules stabilized at the O* bridge sites. We notehowever, that the experimental data does not allow tounambiguously confirm the fine details of the structure,and that the XPS fitting at 7.7 D O/u.c. suggests that lesswater is dissociated. The same is true at a coverage of 8H O/u.c., the next coverage where the TPD dataindicates that a stable structure exists. The nc-AFMimages show that new, ordered protrusions emergewhen additional water is added to the ring-like structure(Fig. 3(C)), and DFT predicts that this water binds via theremaining dangling H-bonds present in the 6 H O/u.c.structure. That these molecules bind solely by H-bondsto other water explains why the desorption temperatureis so close to that of multilayer ice. The reason for theobserved strong AFM contrast is not yet known, and wecannot discount the possibility of additionalrearrangement at this coverage. What is clear, is that theβ and α‘ peaks observed in TPD are due to watersqueezed into the 6 H O/u.c. structure, leading to a re-optimization of the available H-bonds. As mentioned above, partially dissociated water dimershave recently been reported to be the most stablespecies on RuO (110) [7] and Fe O (111) [8,9], andappear to be common on metal oxide surfaces. Based onthe analysis presented here, we expect these species toform whenever there are under-coordinated surface cations sufficiently close together that a H O-OH bondcan be established, provided there are under-coordinated O atoms available to form a stable surfacehydroxyl. Here, the SCV reconstruction of Fe O (001)limits the availability of the latter sites, which is why thewater-dimer coverage saturates at one per unit cell. OnRuO (110) [7], for example, where the under-coordinated surface O atoms are homogeneous andplentiful, a complete coverage of H O-OH dimers isachieved.Before concluding, it is worth to consider whether thelow temperature/low pressure phenomena observedhere bear any resemblance to the adsorption/desorptionof water on Fe O (001) under more realistic conditions.At first glance, the adsorption threshold of 10 -2 mbarobserved by Kendelewicz et al. [26] at room temperaturein ambient-pressure XPS studies suggests a significantpressure gap. However, this threshold is entirelyconsistent with our assertion that isolated molecules areweakly bound, and that a strongly bound partially-dissociated water dimer species forms when twomolecules meet on the surface. Given the bindingenergy of the water monomer (-0.64 eV) determinedhere, the 10 -2 mbar threshold corresponds to aninstantaneous coverage of 0.2 H O/u.c. at 273 K. This issufficient to expect that two monomers can meet beforedesorbing. The as-formed dimer is more strongly bound,so a stable coverage will develop rapidly. Alternatively,the 10 −2 mbar threshold corresponds to a chemicalpotential of −0.78 eV, which agrees very well with theadsorption energy of the partially dissociated waterdimer determined by the inversion analysis (-0.82 eV). Assuch, the surface science approach utilized here appearsdirectly applicable to understand theadsorption/desorption of water at pressures relevant to12atalysis. To our knowledge, the reactivity of partially-dissociated water dimers have not been studied directly,and it will be fascinating to see if these species play anactive role in geochemical or corrosion processes, or incatalysis where metal oxides are frequently used as acatalyst or as a support for metal nanoparticles. Inparticular, it will be interesting to learn whether partiallydissociated species play a role in the water-gas shiftreaction, because industry currently utilizes an Fe O based catalyst [41–44], and partially dissociated specieshave now been directly identified on both major facets.In summary, the formation of partially-dissociated wateragglomerates on Fe O (001) is driven by the formation ofstrong intermolecular H-bonds and facilitated by theclose proximity of under-coordinated cations. Thepresence of the SCV reconstruction ensures thatpartially-dissociated water dimers and trimers remainisolated because there is only one O site that can accommodate a proton per unit cell. The partially-dissociated agglomerates act as an anchor to build aring-like H-bonded network as the coverage is increased,and the water layer completes by saturating dangling H-bonds within this stable structure. A similar evolutioncan be expected wherever a surface presents well-spaced active sites for dissociation. ACKNOWLEDGEMENTS: The authors gratefullyacknowledge funding through projects from the AustrianScience Fund FWF (START-Prize Y 847-N20 (MM, JH, RB &GSP); Special Research Project ‘Functional Surfaces andInterfaces’, FOXSI F4505-N16 and F4507-N16 (MS &UD)), the European Research Council (UD: ERC-2011-ADG_20110209 Advanced Grant ‘OxideSurfaces’), andthe Doctoral College TU-D (ZJ) and Solids4fun (W1243:RB). The computational results presented have beenachieved using the Vienna Scientific Cluster (VSC). [1] O. Björneholm, M. H. Hansen, A. Hodgson, L.-M. Liu, D. T. Limmer, A. Michaelides, P. Pedevilla, J. Rossmeisl, H. Shen, G. Tocci, E. Tyrode, M.-M. Walz, J. Werner, and H. Bluhm, Chem. Rev. , 7698 (2016).[2] A. Hodgson and S. Haq, Surf. Sci. Rep. , 381 (2009).[3] R. Mu, Z. Zhao, Z. Dohnalek, and J. Gong, Chem. Soc. Rev. , 1785 (2017).[4] Z. Zhang, O. Bondarchuk, B. D. Kay, J. M. White, and Z. Dohnálek, J. Phys. Chem. B , 21840 (2006).[5] N. Mulakaluri, R. Pentcheva, M. Wieland, W. Moritz, and M. Scheffler, Phys. Rev. Lett. , 176102 (2009). [6] D. Halwidl, B. Stöger, W. Mayr-Schmölzer, J. Pavelec, D. Fobes, J. Peng, Z. Mao, G. S. Parkinson, M. Schmid, F. Mittendorfer, J. Redinger, and U. Diebold, Nat. Mater. , 450 (2015).[7] M.-T. Nguyen, R. Mu, D. C. Cantu, I. Lyubinetsky, V.-A. Glezakou, Z. Dohnálek, and R. Rousseau, J. Phys. Chem. C , 18505 (2017).[8] P. Dementyev, K.-H. Dostert, F. Ivars-Barceló, C. P. O’Brien, F. Mirabella, S. Schauermann, X. Li, J. Paier, J. Sauer, and H.-J. Freund, Angew. Chemie , 14148 (2015).[9] X. Li and J. Paier, J. Phys. Chem. C , 1056 (2016).[10] S. Maier and M. Salmeron, Acc. Chem. Res. , 2783 (2015). 1311] L. Gross, F. Mohn, N. Moll, P. Liljeroth, and G. Meyer, Science , 1110 (2009).[12] A. Shiotari and Y. Sugimoto, Nat. Commun. , 14313 (2017).[13] J. Pavelec, J. Hulva, D. Halwidl, R. Bliem, O. Gamba, Z. Jakub, F. Brunbauer, M. Schmid, U. Diebold, and G. S. Parkinson, J. Chem. Phys. , 14701 (2017).[14] S. L. Tait, Z. Dohnálek, C. T. Campbell, and B. D. Kay, J. Chem. Phys. , (2005).[15] R. Bliem, E. McDermott, P. Ferstl, M. Setvin, O. Gamba, J. Pavelec, M. A. Schneider, M. Schmid, U.Diebold, P. Blaha, L. Hammer, and G. S. Parkinson,Science , 1215 (2014).[16] R. Bliem, R. Kosak, L. Perneczky, Z. Novotny, O. Gamba, D. Fobes, Z. Mao, M. Schmid, P. Blaha, U. Diebold, and G. S. Parkinson, ACS Nano , 7531 (2014).[17] G. S. Parkinson, N. Mulakaluri, Y. Losovyj, P. Jacobson, R. Pentcheva, and U. Diebold, Phys. Rev. B , 125413 (2010).[18] N. Mulakaluri and R. Pentcheva, J. Phys. Chem. C , 16447 (2012).[19] G. S. Parkinson, Z. Novotný, P. Jacobson, M. Schmid, and U. Diebold, J. Am. Chem. Soc. , 12650 (2011).[20] G. S. Parkinson, T. A. Manz, Z. Novotný, P. T. Sprunger, R. L. Kurtz, M. Schmid, D. S. Sholl, and U. Diebold, Phys. Rev. B , 195450 (2012).[21] O. Gamba, J. Hulva, J. Pavelec, R. Bliem, M. Schmid, U. Diebold, and G. S. Parkinson, Top. Catal. , 420 (2017).[22] J. Guo, X. Meng, J. Chen, J. Peng, J. Sheng, X. Z. Li, L. Xu, J. R. Shi, E. Wang, and Y. Jiang, Nat. Mater. , 184 (2014).[23] X. Meng, J. Guo, J. Peng, J. Chen, Z. Wang, J.-R. Shi, X.-Z. Li, E.-G. Wang, and Y. Jiang, Nat. Phys. , 235 (2015). [24] J. Peng, J. Guo, P. Hapala, D. Cao, R. Ma, B. Cheng,L. Xu, M. Ondráček, P. Jelínek, E. Wang, and Y. Jiang, Nat. Commun. , 122 (2018).[25] M. Taguchi, A. Chainani, S. Ueda, M. Matsunami, Y. Ishida, R. Eguchi, S. Tsuda, Y. Takata, M. Yabashi,K. Tamasaku, Y. Nishino, T. Ishikawa, H. Daimon, S.Todo, H. Tanaka, M. Oura, Y. Senba, H. Ohashi, and S. Shin, Phys. Rev. Lett. , 256405 (2015).[26] T. Kendelewicz, S. Kaya, J. T. Newberg, H. Bluhm, N. Mulakaluri, W. Moritz, M. Scheffler, A. Nilsson, R. Pentcheva, and G. E. Brown, J. Phys. Chem. C , 2719 (2013).[27] L. Köhler and G. Kresse, Phys. Rev. B , 165405 (2004).[28] A. Kiejna, T. Ossowski, and T. Pabisiak, Phys. Rev. B , 125414 (2012).[29] I. Bernal-Villamil and S. Gallego, J. Phys. Condens. Matter , 12001 (2015).[30] M. Dion, H. Rydberg, E. Schröder, D. C. Langreth, and B. I. Lundqvist, Phys. Rev. Lett. , 246401 (2004).[31] J. Klimeš, D. R. Bowler, and A. Michaelides, J. Phys. Condens. Matter , 22201 (2010).[32] K. Lee, É. D. Murray, L. Kong, B. I. Lundqvist, and D. C. Langreth, Phys. Rev. B , 81101 (2010).[33] N. Mulakaluri, R. Pentcheva, and M. Scheffler, J. Phys. Chem. C , 11148 (2010).[34] F. Mirabella, E. Zaki, F. Ivars-Barcelo, X. Li, J. Paier, J. Sauer, S. Shaikhutdinov, and H.-J. Freund, Angew. Chemie Int. Ed. , 1 (2017).[35] T. Schiros, H. Ogasawara, L.-Å. Näslund, K. J. Andersson, J. Ren, S. Meng, G. S. Karlberg, M. Odelius, A. Nilsson, and L. G. M. Pettersson, J. Phys. Chem. C , 10240 (2010).[36] S. S. Xantheas, Chem. Phys. , 225 (2000). 1437] T. Schiros, L.-Å. Näslund, K. Andersson, J. Gyllenpalm, G. S. Karlberg, M. Odelius, H. Ogasawara, L. G. M. Pettersson, and A. Nilsson, J. Phys. Chem. C , 15003 (2007).[38] H. S. Frank and W.-Y. Wen, Discuss. Faraday Soc. , 133 (1957).[39] A. Stroppa, K. Termentzidis, J. Paier, G. Kresse, and J. Hafner, Phys. Rev. B , 195440 (2007).[40] L. Schimka, J. Harl, A. Stroppa, A. Grüneis, M. Marsman, F. Mittendorfer, and G. Kresse, Nat. Mater. , 741 (2010).[41] C. Martos, J. Dufour, and A. Ruiz, Int. J. Hydrogen Energy , 4475 (2009).[42] M. Estrella, L. Barrio, G. Zhou, X. Wang, Q. Wang, W. Wen, J. C. Hanson, A. I. Frenkel, and J. A. Rodriguez, J. Phys. Chem. C , 14411 (2009).[43] D. S. Newsome, Catal. Rev. , 275 (1980). [44] C. Rhodes, B. P. Williams, F. King, and G. J. Hutchings, Catal. Commun. , 381 (2002).[45] H. Schlichting and D. Menzel, Surf. Sci. , 27 (1992).[46] J. Hulva, Z. Jakub, Z. Novotny, N. Johansson, J. Knudsen, J. Schnadt, M. Schmid, U. Diebold, and G. S. Parkinson, J. Phys. Chem. B , 721 (2018).[47] G. Kresse and J. Hafner, Phys. Rev. B , 13115 (1993).[48] G. Kresse and J. Furthmüller, Comput. Mater. Sci. , 15 (1996).[49] G. G. Kebede, D. Spångberg, P. D. Mitev, P. Broqvist, and K. Hermansson, J. Chem. Phys. , 64703 (2017).[50] M. J. Gillan, D. Alfè, and A. Michaelides, J. Chem. Phys. , (2016). Methods
The TPD and XPS experiments were performed in avacuum system optimized for the study the surfacechemistry of metal oxide single crystals. The system hasbeen described in detail elsewhere [13]. Briefly, thesingle crystal Fe O sample (6x6x1 mm, SurfaceNetGmbH) is mounted on a Ta backplate in thermal contactwith a L-He flow cryostat. The sample can reach a basetemperature of ≈30 K, and can be heated to 1200 K bydirect current heating of the sample plate. Temperaturesare measured by a K-type thermocouple spot welded tothe sample plate, and calibrated by the multilayerdesorption of simple gases [45]. D O was adsorbeddirectly onto a 3.5 mm diameter spot in the centre of thesample surface using an effusive molecular beam source.The beam has a close to top-hat profile and has aprecisely calibrated flux (9.2 ± 0.5 × 10 D O molecules/cm .s) at the sample. Coupled with sticking-probability measurements, this allows accurateprediction of the absolute water coverage [13,46]. This isparticularly straightforward to achieve here, because thesticking probability for water is unity at all coverages at100 K (see Fig. S1 and Fig. 1(B)). For TPD experiments thesample is exposed to water at 100 K, and then heatedwith a linear ramp of 1 K/s. XPS utilizes a SPECS Phoibos150 analyser with a monochomatized FOCUS 500 Al KαX-ray source. STM and nc-AFM measurements wereperformed in a separate vacuum system using anOmicron LT-STM equipped with a QPlus sensor. Here,water exposures were performed using a high-precisionleak valve. The water coverage is defined in H Omolecules per (√2×√2)R45° unit cell (H O/u.c.), where 1H O/u.c. is a coverage of 1.42×10 cm -2 . 15he Vienna ab initio Simulation Package(VASP) [47,48] was used for all density functional theory(DFT) calculations. The Projector Augmented Wave(PAW) method describes the electron and ioninteractions, with the plane wave basis set cut-off energyset to 550 eV. A Γ-centered k-mesh of 5×5×5 was usedfor bulk calculations, adjusted to 5×5×1 for (001) surfacecalculations. Convergence is achieved when theelectronic energy step of 10 −6 eV is obtained, and forcesacting on ions become smaller than 0.02 eV/Å. Thecalculations are based on the “subsurface cationvacancy” (SCV) reconstructed model of the Fe O (001)surface [12]. Adsorption energies E ad are corrected forthe zero-point energy (ZPE) (details in S.I.) and arequoted as an average (per molecule, if > 1 H O/u.c.),unless otherwise mentioned. The optPBE-DF [30–32] functional (details in S.I.) accounts for the van der Waalscorrections, and ultimately delivers results that correlatewell with the experiments. The same functional wasrecently used to simulate water adsorbed on NaCl(001)and MgO(001) surfaces [49] and water clusters [50].The optimum configuration for each water coverage wasdetermined via a systematic search inspired by geneticalgorithms. For each coverage, the results of at least 10trial calculations were analyzed to identify factorsleading to a low total energy. These insights, togetherwith those found at other coverages, were used to builda next generation of trial structures. This processeventually leads to the energetically lowest configurationthe system can reach for the given coverage of water. Inthe end, over 500 configurations have been investigated.16 ater Agglomerates on Fe O (001) Matthias Meier,
Jan Hulva, Zdenĕk Jakub, Ji Pavelec, r rı ı Martn Setvin, Roland Bliem , Michael Schmid, Ulrike Diebold, Cesare Franchini, and Gareth S. Parkinson Institute of Applied Physics, Technische Universität Wien, Vienna, Austria University of Vienna, Faculty of Physics and Center for Computational Materials Science, Vienna, Austria
Supporting informationD O dosing
For both, TPD and XPS experiments, D O was dosed by an effusive molecular beam with its intensity calibrated [1].During the dosing, D O molecules with a thermal energy of a room-temperature molecular beam were impinging atthe surface with the intensity of 0.065 molecules per surface unit cell per second. Figure S1 shows the D O signal(m/e = 20) measured by a mass spectrometer positoned line of sight to the sample during dosing at 100 K (blackcurve) and 680 K (gray curve). The stcking coefficient of thermal kinetc energy D O molecules at 100 K is typicallyclose to unity [2]. Therefore, all dosed molecules remain at the sample and the signal measured by massspectrometer does not change after the molecular-beam shutter is opened (black line). Prolonged exposure of thesample to the D O beam leads to growth of water multlayers since the ice sublimaton rate at 100 K is negligible.Taking advantage of the stcking coefficient being close to unity, we can directly calculate the absolute coverage foreach TPD curve by multplying the beam intensity by the dosing tme. When D O is dosed at 680 K, the mass spectrometer signal increases immediately after opening of the shutter andstays constant. This is a consequence of all available adsorpton states desorbing below 650 K (inset of Fig. 1 in themain text) so molecules dosed at this temperature are scattered off the sample and can be detected by the massspectrometer(gray line in Fig. S1).
Figure S1: D O signal measured while dosing by molecular beam for asample temperature of 100 K (black curve) and 680 K (gray curve),molecular beam intensity = 0.065 D O/u.c. per second.
Inversion analysis
Details of the inversion analysis presented in Fig. 1(C) in the main text are shown in Fig. S2. TPD curves of thesaturated peaks used to obtain coverage dependent desorpton energies for a range of pre-exponental factors ν areshown in Fig. S2(A) as thick solid lines. The high-temperature parts of individual peaks belonging to the nextdesorpton feature were cut off to perform the analysis separately for individual peaks. Desorpton curves for lowerinital coverages were simulated using the obtained coverage dependent desorpton energies. Simulated curves werehen compared with the experimental data to find the value of ν giving the best agreement. Examples ofcorresponding experimental and the best matching simulated curves are shown in Fig. S2(A) as thin solid and blackdashed lines, respectvely. Figure S2:
Inversion analysis of the TPD data. (A)
Desorpton curves corresponding to saturated desorpton peaks (thick solidlines) were used for inversion to simulate lower coverage curves (thin solid lines, only one low-coverage curve for each peak isshown here). The black dashed lines represent the simulated curves for the best-matching pre-exponental factors. (B)
Comparison of the experimental data (thin blue line) with the simulated data (dashed lines) from the inversion analysis for a pre-exponental factor ν =10 s -1 (purple), ν =10 s -1 (black, best fit) and ν =10 s -1 (orange). (C-F) Dependence of the total error χ between simulated and experimental curves on pre-exponental factor ν . Y-axis scales between individual figures were set todisplay the trends of χ . Figure S2(B) shows the peak δ in Fig. S2(A) in detail with two additonal simulated curves for ν =10 s -1 (dashedpurple), and ν =10 s -1 (dashed orange). We see that these two curves differ from the experimental curve (thin solidblue) more than the best matching curve for ν =10 s -1 (dashed black). To find the value of ν giving the best agreement we calculated the error χ for each peak defined as a square of thedifference between experimental and simulated curve summed for all curves belonging to the given peak (except thesaturated curve used for inversion) [3]. The results in Fig. S2(C-F) show the dependence of χ as a functon of ν, indicatng an optmal value of the pre-exponental factor. We note that the results of the analysis are less reliable for the α’ peak because only one low coverage curve wasavailable for the analysis.The results are insensitve to which curve is chosen for the analysis. In other words, we obtain similar results if wechoose a lower coverage curve instead of the curve for the saturated peak.The temperature range for the analysis is restricted as displayed in Fig. S2(A). Trailing edges of individual peaks are’cut-off’ so as not to include these data in the error analysis. This was done to prevent the influence of the hightemperature part of the curve belonging to a different desorpton feature on the error analysis. Nevertheless, Fig.S2(B) shows that the biggest error caused by the variaton of ν is located at the leading edge of the curve and on theigh-temperature side of the peak apex. In additon, as the trailing edges on all TPD curves overlap at the trailingedges (see in Fig. S2(A)) the error at the high temperature side is not significant. This allow us to justfy the above-described approach of the inversion analysis of the TPD spectra containing multple peaks. Scanning Tunneling Microscopy
Figure 2(A,B) (in the main text) shows a clean surface with some of the typical surface defects (ant-phase domainboundaries - APDB, surface hydroxyls - O surface
H). When water is dosed at low temperature and the sample is annealedto 255 K (above the desorpton temperature of the 1 st monolayer), water remains adsorbed at the surface defects(inset of the Fig. 1(A) in the main text). Water molecules adsorbed on surface defects are shown in Fig. 2(B) ( in themain text) as intense bright protrusions. The small desorpton peak φ at 550 K is straightforward to explain on thebasis of our previous room-temperature study of water adsorpton on Fe O (001) [4]. Water molecules react with asmall number of surface oxygen vacancies created during sample preparaton, and the O atom repairs the vacancy.The hydrogen atoms create two O surface H groups, and these species recombine to desorb water above 500 K leavingbehind oxygen vacancies. Similar behavior is well known on reduced TiO (110) surfaces above 490 K [5]. The ε peakat 310 K is also related to desorpton from surface defects such as step edges and antphase domain boundaries(APDB) [6,7]. Bright features corresponding to H O at the defect sites responsible for the ε peak (as seen in Fig. 2(B))are not observed in room-temperature STM images of Fe O (001) because the desorpton rate is extremely fastcompared to the speed of an STM measurement. Since the coverage is very low, it is difficult to positvely identfy thepresence of OH groups by XPS due to strong overlap with the O 1s peak from the Fe O substrate, but the lack ofmolecular water in the spectrum leads us to believe that water most likely adsorbs dissociatvely. This is in line withrecent experiments of methanol adsorpton, where dissociatve adsorpton (deprotonaton) occurred preferentally atstep edges, APDB and subsurface Fe defects, and the increased actvity of these sites was linked to the presence ofFe catons [7]. Figure S3: :
STM image of (A) O annealed to 185 K, (B) O annealed to 212 K, (C) O annealed to 225 K. All images were acquired at 78 K.
Figures S3(A-C) show STM images (acquired at 78 K) of a surface which was exposed to 5 L (1 L = 1.33x10 -6 mbar.s) ofwater at 120 K and then annealed to higher temperatures in a step-wise manner to desorb part of the molecules.After annealing to 185 K for 5 minutes (Fig. S3(A)) we see fuzzy features in the directon of the surface Fe oct rows.After annealing to ≈212 K (Fig. S3(B)) we can clearly see chain structures in the directon of the iron rows. Theiroverage strongly decreases after further annealing to ≈225 K (Fig. S3(C)) and isolated double-lobed protrusions canbe seen instead. At all measured coverages the resoluton of the STM does not provide any informaton about theinner structure of the imaged species. X-ray Photoelectron Spectroscopy
X-ray Photoelectron Spectroscopy (XPS) was done using a monochromatzed Al Kα radiaton source. To increase thesurface sensitvity, spectra were acquired at a grazing exit angle. The Fe 2p peak is not affected by water adsorpton(Fig. S4). A shoulder characteristc for Fe +2 , which would indicate a change of the oxidaton state of the surface Fe by acharge transfer from the proton to the surface iron atom [8], is not visible. The clean surface exhibits a slightlyasymmetric O 1s peak at 530.1 eV due to the lattice oxygen [9]. Figure S4 also shows that D O adsorbs mostly inmolecular form at 40 K but partly dissociates 175 K. To rule out the influence of X-ray irradiaton on thephotoelectron spectra, we compared the first and the last scan of the O 1s transiton. We found these to be identcal,suggestng that water was not dissociated by X-rays over the tme scale of the experiment.
Figure S4:
Fe 2p and O 1s of the clean surface (black), 2.6 D O/u.c. deposited at 40 K (orange, only for O 1s) and 2.6 D O/u.c.heated to 175 K (blue). Measured at a grazing exit of 80° at the sample temperature of 40 K.
Theoretical Details
Calculatons were performed on a slab of 13 planes (equivalent of 5 fixed and 2 relaxed Fe oct O layers), where 9 werekept fixed and 4 allowed to relax (similar to the setup used in Ref. [10]). To accurately model a metal oxide system, aneffectve on-site Coulomb repulsion term (Hubbard term) U eff =3.61 [10,11] (U=4.5, J=0.89 according to Dudarev etal. [12]) was added. Slight variatons of U and J are not affectng the conclusions of this work. Dispersion effects aretreated by adding a non-local correlaton term with a density-dependent kernel, to the base Xc functonal (vdW-DF byDion et al. [13]). Klimeš et al. [14] applied and tested the above implementaton of dispersion and proposedoptmized versions of the base functonal of PBE to compensate small discrepancies such as imprecisions in gas phasedimer bonds, resultng in the optPBE-DF and optB88-DF [14] functonals. Dipole correctons, as implemented in VASP(IDIPOL=3 and LDIPOL=.TRUE.), according to the following Refs. [15,16], are applied.The phonon density of states required for calculatng the ZPE (zero-point energy) correctons are obtained within theharmonic approximaton neglectng substrate-adsorbate interactons. Displacements are generated usinghonopy [17]. For phonon density of state calculatons, additonally, under-coordinated O surface atoms (labelled O*,see details in Fig. 2(A) in the main text) were also displaced, to take into account eventual changes induced byadsorbed protons.Convergence criteria on the adsorbates and O* were further refined (down to 10 −9 eV and 0.005 eV/Å) on slabswhere only the 4 topmost planes previously relaxed and optmized, remained, discarding the 9 bottom planes. Withthe excepton of the O* atoms, the remaining 4 planes were kept fixed for the phonon calculatons.The ZPE corrected E ad are then given by: E ad =( E nH O / surf + E n H O / surfZPE )−( E surf + E surfZPE )− n ( E H O + E H OZPE ) where E O/surf , E , E O are DFT total energies of n H O molecules adsorbed on the surface, the clean surface and freemolecule references, respectvely. E ZPEnH O/surf , E
ZPEsurf , E
ZPEH O are the corresponding ZPE correctons. In the case of theclean surface, only the two O* are displaced and contribute to the correcton. References [1] J. Pavelec, J. Hulva, D. Halwidl, R. Bliem, O. Gamba, Z. Jakub, F. Brunbauer, M. Schmid, U. Diebold, and G. S. Parkinson, J. Chem. Phys. , 14701 (2017).[2] M. J. Strniman, C. Huang, R. S. Smith, S. A. Joyce, and B. D. Kay, J. Chem. Phys. , 1295 (1996).[3] S. L. Tait, Z. Dohnálek, C. T. Campbell, and B. D. Kay, J. Chem. Phys. , (2005).[4] G. S. Parkinson, Z. Novotný, P. Jacobson, M. Schmid, and U. Diebold, J. Am. Chem. Soc. , 12650 (2011).[5] M. A. Henderson, Langmuir , 5093 (1996).[6] G. S. Parkinson, T. A. Manz, Z. Novotný, P. T. Sprunger, R. L. Kurtz, M. Schmid, D. S. Sholl, and U. Diebold, Phys. Rev. B , 195450 (2012).[7] O. Gamba, J. Hulva, J. Pavelec, R. Bliem, M. Schmid, U. Diebold, and G. S. Parkinson, Top. Catal. , 420 (2017).[8] G. S. Parkinson, Surf. Sci. Rep. (2016).[9] R. Bliem, E. McDermott, P. Ferstl, M. Setvin, O. Gamba, J. Pavelec, M. A. Schneider, M. Schmid, U. Diebold, P. Blaha, L. Hammer, and G. S. Parkinson, Science , 1215 (2014).[10] I. Bernal-Villamil and S. Gallego, J. Phys. Condens. Matter , 12001 (2015).[11] A. Kiejna, T. Ossowski, and T. Pabisiak, Phys. Rev. B , 125414 (2012).[12] S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys, and A. P. Sutton, Phys. Rev. B , 1505 (1998).[13] M. Dion, H. Rydberg, E. Schröder, D. C. Langreth, and B. I. Lundqvist, Phys. Rev. Lett. , 246401 (2004).[14] J. Klimeš, D. R. Bowler, and A. Michaelides, J. Phys. Condens. Matter , 22201 (2010).[15] G. Makov and M. C. Payne, Phys. Rev. B , 4014 (1995).[16] J. Neugebauer and M. Scheffler, Phys. Rev. B , 16067 (1992).[17] A. Togo and I. Tanaka, Scr. Mater.108