"This is the peer reviewed version of the following article: The Behavior of the Ru-bda Water
Oxidation Catalysts at Low Oxidation States, which has been published in final form at
https://onlinelibrary.wiley.com/doi/pdf/10.1002/chem.201801236

This article may be used for non-commercial purposes in accordance with Wiley Terms and
Conditions for Self-Archiving."

Accepted Article

The Behavior of the Ru-bda Water Oxidation Catalysts at Low Oxidation
States
Roc Matheu,a Abolfazl Ghaderian,a Laia Francàs,a Petko Chernev,b Mehmed Z. Ertem,c Jordi BenetBuchholz,a Victor Batista,d,* Michael Haumann,b Carolina Gimbert-Suriñach,a,* Xavier Sala,e,* and
Antoni Llobeta,e,*
a

Institute of Chemical Research of Catalonia (ICIQ), Barcelona Institute of Technology (BIST),
Avinguda Països Catalans 16, 43007 Tarragona, Spain.
b

Institute for Experimental Physics, Free University Berlin, D-14195 Berlin, Germany.

c

Chemistry Department, Brookhaven National Laboratory, Upton, NY 11973, USA.

d

Department of Chemistry, Yale University. P.O. Box 208107, New Haven, CT 06520-8107, USA.

e

Departament de Química, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, 08193
Barcelona, Spain.

Corresponding authors: cgimbert@iciq.es; victor.batista@yale.edu; xavier.sala@uab.cat;
allobet@iciq.es.

1

Abstract
The Ru complex [RuII(bda--N2O2)(N-NH2)2], 1, (bda2- = (2,2'-bipyridine)-6,6'-dicarboxylate; N-NH2 =
4-(pyridin-4-yl)aniline) is used as a synthetic intermediate to prepare new Ru-bda complexes that
contain the NO+, acetonitrile (MeCN) or H2O ligands at oxidation states II and III. Complex 1 in acidic
solutions reacts with excess NO+ (generated in situ from sodium nitrite) to form a new Ru complex
where the aryl amines ligand N-NH2 in 1 are transformed into diazonium salts (N-N2+ = 4-(pyridin-4yl)benzenediazonium)) together with the formation of a new Ru-NO group at the equatorial zone,
to generate [RuII(bda--N2O)(NO)(N-N2)2]3+, 23+. Here the bda ligand only acts in a tridentate manner
as -N2O with a dangling carboxylato group. Similarly complex 1 can also react with a coordinating
solvent, such as MeCN, at room temperature leading to complex [RuII(bda--N2O)(MeCN)(N-NH2)2],
3. Finally in acidic aqueous solutions a related reaction occurs where solvent water coordinates the
Ru center forming {[RuII(bda--(NO)3)(H2O)(N-NH3)2](H2O)n}2+, 42+, that is strongly hydrogen bonded
with additional water molecules at the second coordination sphere. Furthermore, under acidic
conditions the aniline ligands are also protonated forming the corresponding anilinium cationic
ligands N-NH3+. We have additionally characterized the one electron oxidized complex {[RuIII(bda-(NO)3.5)(H2O)(N-NH3)2](H2O)n}3+, 53+. The fractional value of  notation, indicates the presence of an
additional contact to the pseudo-octahedral geometry of the Ru metal center. The coordination
mode of the complexes has been studied both in the solid state and in solution through single-crystal
X-ray diffraction, X-ray absorption spectroscopy, variable-temperature NMR and density functional
theory calculations. While the -N2O is the main coordination mode for 23+ and 3, an equilibrium
that involves isomers with -N2O and -NO2 coordination modes and neighboring hydrogen bonded
water molecules is observed for 42+ and 53+.

Keywords
water oxidation catalysis, water splitting, Ru complexes, water coordination, frustrated
coordination site, Ru-bda complexes

2

1. Introduction
The field of water oxidation catalysis (WOC) by molecular transition metal complexes has evolved
enormously since the early work of Meyer and co-workers on the blue dimer 1 back in the early
eighties up to now.2 Extremely rugged molecular WOCs based on Ru complexes with turnover
numbers (TON) in some cases over a million have been recently reported with confirmed molecular
structure maintained over the entire catalytic performance.3,4,5 A prominent family of these
catalysts are Ru complexes based on the tetradentate ligand 2,2'-bipyridine-6,6'-dicarboxylato
(bda2-; see Chart 1 for a drawing of all the ligands described in this work), named Ru-bda type of
complexes that have been recently reported.6,7,8,9,10 In some cases these complexes have catalytic
activity that is comparable to that of the natural oxygen evolving complex of photosystem II (OECPSII) in green plants and algae.11,12 Further, even faster catalysts have been reported recently
containing a related pentadentate ligand, such as 2,2':6’,2”-terpyridine-6,6”-dicarboxylate (tda 2-,
see Chart 1).13,14
The Ru-bda complexes of general formula [Ru(bda)(N-Ax)2] (N-Ax is a neutral monodentate pyridyl
type of ligand in axial position) constitute a family of complexes that are easy to prepare and whose
properties can be nicely tuned through the axial ligand. Indeed, the different nature of the axial
ligand can exert steric, electronic or supramolecular effects that influence the properties of the
catalyst.2,6,8,15,16 In addition, axial ligands with the adequate functionalities have been used to
immobilize this class of complexes on conducting solid surfaces.17,18,19,20,21
At oxidation state II the [Ru(bda)(N-Ax)2] complexes have a saturated coordination sphere with a
distorted octahedral type of symmetry. As such they are not water oxidation catalysts but actually
catalyst precursors. Those complexes enter the catalytic cycle upon coordinating a water molecule
and reaching higher oxidation states.6,20,22,23 The Ru-aqua functionality is important since it provides
the capacity to loose protons and electrons in a concerted fashion at relatively low potentials and
at the same time provides for a O-O bond formation site. For the particular case where the axial
ligand is 4-Methylpyridine (4-Me-py), an X-ray crystal structure of {[[RuIV(OH)(bda--N2O2)(4-Mepy)2](H2O)]2(-H)}3+ reveals a seven coordinate sphere for the Ru(IV) metal center in a pseudo
pentagonal-bipyramid geometry.24 In sharp contrast, at oxidation state II, the corresponding Ruaqua complexes are not well characterized. Interestingly, the coordination geometry of the d6 Ru(II)
ion is expected to be octahedral since a seven coordination would imply the violation of the 18
electron rule.25 Thus, assuming the axial ligands do not detach from the Ru-aqua complex, as
evidenced by 1H-NMR spectroscopy,23 the bda2- ligand must partially decoordinate to generate
either a dangling carboxylate or a frustrated coordination site via decoordination of one of Ru-N
bonds of the bda2- bipyridyl moiety.20,23,26 The latter is an interesting concept, recently developed,
with implications in a variety of processes including heterolytic H2 splitting. 27,28
The main goal of the present work is to shed light on the potential coordination modes of the bda 2ligand at low oxidation states (i.e., Ru(II) and Ru(III)) which is crucial for the aqua ligand coordination
that provides the entry into the catalytic cycle. For this purpose, we have prepared a family of Ru-

3

bda complexes (see complexes 1-53+ in Scheme 1) and characterized them by x-ray diffraction and
spectroscopic methods as well as by density functional theory (DFT) calculations.
2. Results and Discussion
2.1 Synthesis and solid state structure
We use complex [RuII(bda--N2O2)(N-NH2)2], 1,17 that contains 4-(pyridin-4-yl)aniline (N-NH2) ligands
at axial positions, as the starting material for all the complexes described in this work, following the
synthetic strategy depicted in Scheme 1. In this work we selected 1 because of its solubility both in
organic solvents and acidic water and because the amino functionality gives an easy synthetic access
to the family of Ru-bda complexes depicted in Scheme 1. This family allows to obtain a large variety
of spectroscopic properties that analyzed together provide a consistent and comprehensive
description of the phenomena occurring in Ru-bda complexes at low oxidation sates. Further, the
remote situation of the amino groups doesn’t influence significantly the redox potential of the Ru
complex.17
The solid-state structure of 1 was analyzed by X-ray diffraction (XRD), Density functional theory
(DFT) calculations and X-ray absorption spectroscopy (XAS). Monocrystals of 1 were obtained by
slow diffusion of diethyl ether into a methanol solution and an ORTEP drawing of its structure is
shown in Figure 1. The bonding distances and angles found in this structure are similar to related
Ru(II) complexes containing similar type of ligands (dRu-O/Ru-N = 1.9 - 2.1 Å).24,29,30,31 The most
interesting feature is the equatorial geometry of the first coordination sphere imposed by the bda 2ligand, generating an ORuO angle of 121.7 o. XAS at the Ru K-edge was carried on a powder of 1, and
the results are reported in Figure 2. The bonding distances obtained by Extended X-ray absorption
fine structure (EXAFS) simulation are presented in Table 1 and show very good agreement with the
XRD data. Further, DFT calculations were carried out for the geometrical optimization of 1 at the
M06 level of theory32 and the main geometric parameters are also displayed in Table 1, whereas the
complete data can be found in the Supporting Information (SI). It is worth noting here the high
degree of consistency of all the metric parameters obtained for 1 by these three methodologies
bringing a good degree of confidence when used for other related structural characterizations that
will be discussed further on.
Complex 1 reacts with excess NO+ in acid (generated in situ from sodium nitrite) 33 to form complex
[RuII(bda--N2O)(NO)(N-N2)2]3+, 23+, where the aryl amines of 1 are transformed into diazonium salts
(N-N2+ = is 4-(pyridin-4-yl)benzenediazonium)). In addition, a new Ru-NO group is also formed at the
equatorial zone according to equation 1. Addition of a saturated solution of KPF6 allows its isolation
as a powdery salt.17

[RuII(bda--N2O2)(N-NH2)2] + 3 NO+ -> [RuII(bda--N2O)(NO)(N-N2)2]3+ + 2 H2O
1
23+

(1)

4

Suitable single crystals for x-ray diffraction analysis of 23+ were obtained by slow evaporation of an
acidic aqueous solution of 23+. Figure 1 shows the ORTEP representation where the nitrosyl group
occupies one of the equatorial positions, while the bda2- exhibits-N2O coordination mode. Such
coordination mode releases the geometrical strain in the equatorial plane described in 1, and
produces a dangling carboxylate not bonded to the Ru metal center. The -N2O coordination mode
is also favored by the crystal packing of the molecule where the dangling carboxylate interacts with
the N atoms of the diazonium groups of another molecule, forming dimers as shown with green
dashes in Figure S9. With regard to the Ru-NO+ group, the near linearity of the Ru-N-O+ bond (175.8o)
and the N-O bond length (dN-O = 1.13 Å) indicate a N-O triple bond character. The DFT calculations
at the M06 level again gives geometric parameters that are very consistent with those obtained by
X-ray diffraction as can be seen in Figure 3 left and Table S2. In particular, 23+ is assigned as closedshell singlet with a near lineal Ru-N-O+ bond (174.7 o) and dN-O = 1.14 Å confirming the RuII-(N≡O+)
nature.
Complex 1 can also react with a coordinating solvent, such as MeCN at room temperature, leading
to complex [RuII(bda--N2O)(MeCN)(N-NH2)2], 3, as shown in equation 2. In this complex, the MeCN
coordinates the Ru center in the equatorial plane and again forces the Ru-bda moiety to rearrange
its coordination mode by releasing one carboxylate, similar to what is observed for 23+. Single
crystals of 3 were obtained by slow diffusion of diethyl ether to a 4:1 MeOH:MeCN solution of 1,
and its ORTEP plot is exhibited in Figure 1. The bda2- ligand coordinates the Ru center in a -N2O
fashion with a pendant carboxylate similar to the coordination mode exhibited by 23+. Here again
the computational calculations at the M06 level provided geometric parameters for 3 analogous to
those obtained by means of XRD (see Table S3)

[RuII(bda--N2O2)(N-NH2)2] + MeCN
1

[RuII(bda--N2O)(MeCN)(N-NH2)2]
3

(2)

The coordination of MeCN is subject to an equilibrium where the solvent molecule coordinates and
decoordinates the Ru center. This equilibrium can be controlled by the concentration of MeCN
solvent. For instance, the addition of a 20% of MeCN to a methanol solution of 1 allows the
quantitative generation of 3 as determined by 1D and 2D NMR spectroscopy (see Figures S1-S7).
However, even if the amount of 1 present in the equilibrium is negligible, as indicated by NMR, it
precipitates from the solution as a powder, probably due to its very low solubility.
At pH = 1.0, 1 reacts with the water solvent forming complex {[RuII(bda--(NO)3)(H2O)(NNH3)2](H2O)n}2+, 42+, as shown in equation 3. Again, the coordination occurs at the equatorial plane
with subsequent rearranging of the coordination mode of bda 2- ligand. In this case, the modification
of the first coordination sphere involves the presence of a network of strongly hydrogen bonded
5

H2O molecules as a second coordination sphere. The number of H2O molecules involved in this
network is designated as “n” and they are strongly interacting with the Ru-aqua group and the
carboxylate moieties of the bda2- ligand. In addition, the amino groups at this pH are protonated
forming the corresponding ammonium salts. The proposed label 42+, represents two limit isomeric
structures namely [4--N2O(H2O)n]2+ and [4--NO2(H2O)n]2+, (denoted as -(NO)3) whose relative
occurrence will be discussed in the following sections. As observed for 3, in the solid state the
equilibrium shifts towards the more insoluble isomer generating 1 again with no water coordinated
at the metal center.

[RuII(bda--N2O2)(N-NH2)2] + 2H+ + (n+1) H2O
1

{[RuII(bda--(NO)3)(H2O)(N-NH3)2](H2O)n}2+
42+

(3)

Finally, 42+ can easily be oxidized by atmospheric oxygen to the paramagnetic d5 Ru(III) complex
{[RuIII(bda--(NO)3.5(H2O)(N-NH3)2](H2O)n}2+, 53+, as presented in Scheme 1. The fractional value of 
notation here indicates the presence of an additional contact to the pseudo-octahedral geometry
of the Ru metal center at we consistently found for these type of complexes. This oxidation has been
monitored by UV-vis spectroscopy and takes place with a t1/2 of approximately 3.3 minutes at RT
according to a first order mechanism (see Figures S10-S11). The complex has been analyzed in
solution by Electron Paramagnetic Resonance (EPR) and XAS as it will be described in the following
sections and in the SI.
2.2. Dynamic behavior in organic solvents
The solution structure of 23+ and 3 were investigated by VT-NMR in organic solvents and by DFT
calculations and the spectra are shown in Figure 4 and in Figures S12-S13. Figure 4 shows the 1HNMR spectra of 23+ in acetone-d6 in the temperature range 273-193 K. At 273 K the 1H-NMR
spectrum of 23+ is not consistent with the -N2O binding mode of the bda2- ligand displayed in the
solid state X-ray structure. As the temperature is lowered the resonances due to the bda 2- ligand
broaden and split, now consistent with the asymmetry observed in the X-ray structure. At 193 K all
the resonances are unambiguously assigned based on 2D-NMR spectroscopy (see Figure S12) and
in agreement with the DFT calculated chemical shifts (see Figure 4). The broadening and splitting of
the bda2- resonances at lower temperatures points out to the presence of dynamic behavior where
the two carboxylates of the bda2- ligand coordinate and decoordinate very fast as graphically shown
in Scheme 2. This is further corroborated by the fact that the resonances for protons C and C’
contiguous to the carboxylate group, are by far the ones that suffer the largest shift of more than
0.4 ppm (see Chart 1). An energy of activation of 12.5 kcal/mol is obtained for this process from the
spectra (see Figure S14). We also obtained optimized structures for RuII-NO+ complexes via DFT
calculations in both -N2O and -(NO)3 modes of bda2- ligand and found the latter binding mode,
6

that is proposed to be the transition state in the interconversion, to be higher in energy by 6.2
kcal/mol (see Scheme 2 and Figure 3).
A similar behavior is observed for the RuII-MeCN complex 3 dissolved in a 1:4 acetonitriled3:methanol-d4 mixture but with subtle differences (see Figure S13). At room temperature the main
isomer observed is the -N2O that upon heating to temperature close to 320 K, its resonances
coalesce. In this case, the energy of activation obtained from the spectra is 15.0 kcal/mol (see Figure
S15).
The X-ray structures and VT-NMR results, just described, clearly show the preference for a -N2O
coordination mode for the bda2- ligand in 23+ and 3, in the absence of as second coordination sphere
hydrogen bonding. The preference for the -N2O coordination mode for Ru-bda complexes
coordinating the MeCN ligand has been observed in related Ru-bda complexes,34,35 This is in
agreement with the HSAB theory,36 where Ru at low oxidation states prefers soft -acceptor ligands
such as N-pyridyl over hard O-carboxylate ligands.

2.3. Dynamic behavior in acidic aqueous solutions
The 1H-NMR spectrum of 1 was recorded in methanol-d4 at room temperature and with increasing
amounts of a pD = 1.0 solution as indicated in the spectra shown in Figure 5. In pure methanol-d4
the 1H-NMR spectrum of 1 shows the typical behavior of a six coordinated Ru(II) complex with C2v
symmetry. All resonances could be unambiguously assigned based on 2D NMR spectra 17 and also
supported by computed NMR chemical shifts at M06 level of theory. However, the addition of
CF3SO3D dissolved in D2O, significantly broaden resonances and some of them are strongly shifted
as observed in Figure 5. The F labelled protons (Chart I), which are close to the amino group, suffer
the highest shift due to the protonation of the amino group. The small shift of the D protons
indicates that the Ru center is almost unaffected by the protonation of the amines. Further, at 25%
of pD = 1.0 solution, we carried out VT-NMR all the way to 240 K where the resonances sharpen
again, as shown in Figure 5 (right), consistent with the presence of a single symmetrical isomer. In
contrast to the previous VT-NMR spectra for 23+, now both B and C protons are similarly shifted
which advocates for a different type of dynamic behavior where both the Ru-N and Ru-O bonds are
simultaneously being formed and broken as proposed in Scheme 3. Further, the VT-NMR results are
consistent with the coordination of a water molecule to the Ru metal center upon addition of an
acidic aqueous solution. The formation and braking of the Ru(II)-OH2 bond occurs simultaneously
with a dynamic behavior involving species [4--N2O(H2O)n]2+ and [4--NO2(H2O)n]2+, that
interconvert through the transition states TS-A, TS-B and TS-C (Scheme 3). The sharpening at low
temperature of the 1H resonances is consistent the absence of the coordination/decoordination of
water to the Ru center but with all the equilibria indicated in Scheme 3 occurring very fast. In sharp
contrast, when 1 is dissolved in neutral aqueous solutions, methanol-d4:D2O (4:1), the VT 1H-NMR
experiment reveals that the aqua coordination/decoordination to Ru equilibrium is still present
(Figure S16).
7

The presence of the [4--NO2(H2O)n]2+ species is favored by solvent H2O molecules strongly
hydrogen bonded to the Ru-H2O group and the carboxylato moiety of the bda2- ligand. This hydrogen
bonding will push the bpy part of the bda2- ligand away from the metal center and explains why the
intuitively preferred Ru-N coordination mode for a soft low spin d6 Ru(II) ion is not always favored.
All the NMR measurements were performed under strict N2 conditions to avoid the oxidation of 42+
to 53+ (see NMR methods in the SI for further details). In order to further prove that 42+ remains at
oxidation state II, we prepared solutions of 42+ identically to those prepared for NMR analysis and
carried out EPR measurements. The absence of EPR signal discards the formation of any traces of
the one electron oxidized species, 53+, and thus confirms that the nature of the NMR resonance
broadening is solely due to the dynamic behavior. The same behavior also occurs in other Ru-bda
catalyst such as [Ru(bda--N2O2)(isoq)2] (isoq is isoquinoline) under exactly the same conditions as
for 1, as shown in Figure S18 in the supporting information. The comparable shift and broadening
of the resonances observed for the [Ru(bda--N2O2)(isoq)2] complex suggests that this is a general
phenomenon occurring in Ru-bda complexes.
We carried out DFT calculations for both the [4--NO2(H2O)n]2+ and [4--N2O(H2O)n]2+ isomers using
either two or four solvated water molecules. We have chosen a limited number of water molecules
as a model although it is obvious that a much larger number of water molecules will be involved as
second and third coordination sphere to the Ru center. The structures with two or four solvated
molecules yielded similar results in terms of geometrical features and relative energies of the
isomers (see Figures S21-22 and Tables S4-S5). To simplify the discussion we present only [4-NO2(H2O)4]2+ and [4--N2O(H2O)4]2+ isomers and the optimized geometries and relevant metric
parameters of which are depicted in Figure 6 and Table 1. The first interesting feature of the
obtained optimized structures for [4--NO2(H2O)4]2+ and [4--N2O(H2O)4] 2+ is the presence of a
hydrogen bonding network between the Ru-aqua group and the carboxylate groups of the bda 2ligand that resembles the crystal structure of {[[RuIV(OH)(-N2O2-bda)(4-Me-py)2](H2O)]2(-H)}3+.24 It
is also interesting to compare the distances between Ru and the N and O atoms of the bda 2- ligand
in both isomers. For the [4--N2O(H2O)4]2+ isomer these distances are very similar to those of 23+
with a the characteristic distance between the Ru and the O atom of the dangling carboxylate at
3.45 Å. On the other hand, for [4--NO2(H2O)4]2+, the most unusual distances are the Ru-N that are
2.26 Å and 2.82 Å, which are 0.1 Å and 0.7 Å longer than the typical Ru-N bonds for related
complexes. The computed chemical shifts at M06 level for the optimized structures (Figure S25) are
not consistent with either those of the [4--N2O(H2O)4]2+ isomer (Path A, Scheme 3) or those of the
[4--NO2(H2O)4]2+ (Path B, Scheme 3) but with a mixture of all of them. The combination of all these
equilibria is depicted in Scheme 3 (Paths A, B and C), where explicit water molecules are not shown
for the sake of simplicity.
The isomers of the one electron oxidized complex, [5--N2O(H2O)4]3+ and [5--NO2(H2O)4]3+, were
also optimized at M06 level of theory and the corresponding structures and most relevant geometric
parameters are presented in Figure 6 and Table 1. For the one electron oxidized isomers, the bond
distances of the first and second coordination sphere are shortened compared to those found for
8

the 42+ counterparts as expected.37 However, the most relevant distances for both isomers, Ru-N of
2.46 Å for [5--NO2(H2O)4]3+ and the Ru-O distance of 3.06 Å for the dangling carboxylate in [5-N2O(H2O)4]2+ remain comparable to its related complex in the oxidation state II.
Another very interesting feature of these complexes is their relative energies. While at oxidation
sate II, the [4--N2O(H2O)4]2+ isomer is favored by 20.3 kcal/mol with regard to the [4--NO2(H2O)4]2+
isomer, at oxidation state III this difference is drastically reduced to only 2.1 kcal/mol. While the
energy comparison is only indicative because of the possibility of several conformers associated with
solvent molecules hampering a reliable comparison, the trend is certainly significant.
In order to extract further experimental evidence about the potential combination of equilibria
acting in the Ru-aqua complexes bearing the bda2- ligand in aqueous solution, we carried out XAS
spectroscopy of frozen solutions of 53+. The results obtained are shown in Figure 2 and Table 1
together with those of 1, [RuIIICl3(trpy)] (trpy is 2,2’:6’,2”-terpyridine) and RuIVO2 that were used as
reference materials.
Table 1 presents the metric parameters 53+, extracted from EXAFS simulations (Figure 2B). As can be
observed the EXAFS spectral changes for 53+ compared to 1 involve an overall shortening of the RuN/O bond lengths, leading to homogenization of the first coordination sphere of Ru-ligand distances.
Very importantly a ~2.54 Å distance was detected that improves the quality of the EXAFS fit by a
factor of about 2 (Table S1). This additional distance is thus indicative that the isomer [5-NO2(H2O)n]3+ is present in the mixture of isomers at this oxidation state. In addition, the metric
parameters obtained from EXAFS coincide well with those obtained from the computed structures
for [5--NO2(H2O)4]3+ confirming the presence of these species in frozen aqueous solution (Table 1).
To the best of our knowledge, this is the first time to experimentally demonstrate that the -NO2
coordination mode for the bda2- ligands is relevant in the combination of potential isomers involved
in Ru-bda type of complexes at low oxidation states, although it had been suggested based on DFT
calculations.26 In addition, we also found a similar distance for the [Ru(tda)(py) 2] catalyst precursor
at oxidation state III, where the two isomers, [RuIII(tda--N2O2)(py)2]+ and [RuIII(tda--N3O)(py)2]+,
are found to coexist in the crystal structure.13
This counterintuitively behavior on the coordination modes of bda2- in 42+ and 53+, which is different
from their homologues to 23+ and 3 in organic solvents, is mainly associated with the strong
hydrogen bonding belt along the aqua and carboxylato ligands. The latter provides the driving force
to push away the bpy part of the bda2- ligand producing a decoordination of one of the N-atoms.26
In summary, our results based on VT-NMR, XAS and DFT calculations provide a detailed description
of the nature of the species involved in water oxidation catalysis by Ru-bda type complexes at low
oxidation states, where the critical Ru-aqua group that provides access to the catalytic cycle is
formed. The results show the fundamental need for a flexible and adaptable ligand, as is the case
for bda2-, both from an electronic and geometrical perspective, to comply with the demands of the
Ru metal centers at the different oxidation states. Further, for the first time we have shown
9

experimentally the presence of isomers in Ru-bda type of complexes where the ligand acts in a NO2 fashion.
3. Associated content
Supporting Information (SI). Synthetic procedures; X-ray crystallographic data in CIF format (CCDC #
1014176 for 1 and 1014177 for 23+); additional experimental, spectroscopic, electrochemical and
computational data.
Corresponding Authors:
victor.batista@yale.edu; xavier.sala@uab.cat; allobet@iciq.es.

5. Acknowledgments
RM, AL and XS acknowledge MINECO and FEDER (CTQ2016-80058-R, CTQ2015-64261-R, SEV 20130319, ENE2016-82025-REDT, CTQ2016-81923-REDC), AGAUR (2017-SGR-1631) for financial support.
RM thanks “La Caixa” foundation for a PhD grant. M.H. thanks the Deutsche
Forschungsgemeinschaft for financial support (grant Ha3265/6-1) and for a Heisenberg Fellowship
and the German Bundesministerium für Bildung und Forschung for funding within the RöntgenAngström Cluster (grant 05K14KE1). We thank S. Reschke and M. Görlin for help in XAS data
collection and M. Nachtegaal at SuperXAS of SLS for excellent technical support. The work at
Brookhaven National Laboratory (M.Z.E.) was carried out under contract DE-SC00112704 with the
U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, and utilized resources
at the BNL Center for Functional Nanomaterials. V.S.B. acknowledges financial support as part of
the Argonne-Northwestern Solar Energy Research (ANSER) Center, an Energy Frontier Research
Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences
under Award Number DE-SC0001059.

10

Table 1. Metric parameters obtained for 1, 42+ and 53+ from X-ray diffraction (XRD) crystallography, EXAFS and DFT.
42+

1
Distancea

DFTc

DFTc

DFTc

[4--NO2(H2O)4]2+

XRD

EXAFS (N)b

53+

EXAFS (N)b

[4--N2O(H2O)4]2+

DFTc

DFTc

[5--NO2(H2O)4]3+

[5--N2O(H2O)4]3+

Ru-OH2

--

--

--

2.09

2.18

1.92 (1)

2.03

2.13

Ru-NO

--

--

--

--

--

--

--

--

Ru-N1

1.92

1.92 (2)

1.93

2.82

1.94

2.54 (1)

2.46

1.99

Ru-N2

1.93

…

1.93

2.26

2.12

2.11 (3)

2.32

2.14

Ru-O1

2.16

2.17 (2)

2.17

2.05

2.12

…

2.03

2.05

Ru-O2

2.18

...

2.17

2.23

3.45

…

2.08

3.06

Ru-N3

2.07

2.07 (2)

2.08

2.04

2.08

2.01 (2)

2.10

2.11

Ru-N4

2.09

…

2.09

2.06

2.11

…

2.10

2.14

20.3

0.0

2.1

0.0

G (kcal/mol)
a

, the labeling scheme is the same as the one used for XRD data of 1 depicted in Figure 1. All distances in Å.
, distances between Ru and the atoms in its first and second coordination sphere. In parenthesis, N, is the coordination number defined as the number of atoms
associated with a particular distance. Additional EXAFS fit parameters (Debye-Waller factors and error sums) are given in Table S1. N and O coordination is
indistinguishable in EXAFS analysis as are the locations of ligands in equatorial or axial positions; in the table the distances from EXAFS therefore were ordered to
comply with the DFT results.
c
, M06 functional, see SI for details.
d
, in bold are depicted Ru-N or Ru-O distances relevant for the -NO2 vs. -N2O isomer discussion.
b

11

Chart 1. Ligands used and discussed in this work together labels for NMR assignment.

12

Scheme 1: General synthetic scheme and labels of complexes. The dashed lines at the first
coordination sphere of the Ru metal center indicate bonds that are being simultaneously formed
and broken. See Chart 1 for the detailed structure of the axial and equatorial ligands. The dashed

lines in 42+ and 53+ at the second coordination sphere represent hydrogen bonding from solvent
water molecules

13

Scheme 2: Species involved in the fast equilibria associated with the dynamic behavior of 23+.
Dashed lines indicate the bonds that are simultaneously formed and broken at the transition state
TS-A. The DFT calculates TS structure for this complex is depicted in Figure 3 right.

14

Scheme 3 Reaction pathways leading to a combination of equilibria involved in dynamic behavior of
42+ or 53+ in acidic aqueous solution. The dashed lines indicate bonds that are simultaneously formed
and broken in the transition states. Water solvation molecules are not shown for clarity purposes.
Three different type of pathways are identified depending on the transition state through which the
species are interconverted. Thus path A involves “TS-A” whereas paths B and C involve “TS-B” and
“TS-C” respectively.

15

Figure 1. X-ray ORTEP plots (ellipsoids at 50% probability) and labeling scheme of 1 (left), the cationic
part of 23+ (center) and 3 (right). Color codes: Ru, cyan; N, blue; O, red; C, black; H, light blue.

16

Figure 2. X-ray absorption spectroscopy analysis for 1 and 53+. Data for metallic Ru, RuO2, and
[RuCl3(trpy)] is shown for comparison. (A) Ru K-edge spectra. Inset: correlation between K-edge
energies (determined at 50 % edge magnitude) and Ru oxidation states. (B) Fourier-transforms
(FTs) of EXAFS spectra in the inset. FTs (experimental data) were calculated for k-values of 1.814.2 Å-1 and using cos2 windows extending over 10 % at both k-range ends. Spectra were vertically
shifted for comparison. Inset, EXAFS oscillations in k-space. Thin black lines are experimental data
whereas colored lines are simulations using parameters shown in Table 1 and SI.

17

Figure 3. Ball and stick representation of the optimized structures at M06 level of theory for [2-N2O]3+ and the transition state [2--(NO)3]3+ that interconverts the two limiting structures shown in
Scheme 2. Color code: Ru, cyan; C, gray; N, blue; O, red, H atoms of the ligands were omitted for
clarity.

[2--N2O]3+

[2--(NO)3]3+

Figure 4. Left, VT -1H-NMR spectra for 23+ in acetone-d6. Right, computed chemical shifts at M06
level of theory for the bda2- hydrogens in 23+. See Chart 1 for the labeling.

18

Figure 5. Left, 1H-NMR of 1 in methanol-d4 (cyan spectrum) and of 42+ with different amounts of a
solution of 0.1 M CF3SO3D in D2O (labeled as D2O, pD = 1.0) at 298 K. Right, VT 1H-NMR of 42+ in
methanol-d4 with 25% D2O, pD = 1.0. Samples prepared under rigorous N2 atmosphere. See Chart 1
for the labeling.

Figure 6. Ball and stick representation of the optimized structures of [4--N2O(H2O)4]2+, [4-NO2(H2O)4]2+, [5--N2O(H2O)4]3+ and [5--NO2(H2O)4]3+ at M06 level of theory. Dashed line indicates
elongated bond. Color code: Ru, cyan; C, gray; N, blue; O, red, H atoms of the ligands were omitted
for clarity.

[4--N2O(H2O)4]2+

[4--NO2(H2O)4]2+

[5--N2O(H2O)4]3+

[5--NO2(H2O)4]3+

19

Graphical Abstract

Lost & Found Coordination
Ru-bda catalysts are powerful water oxidation catalysts thanks to seven coordination provided by
the ligand framework. While the hepta-coordination provides stability when the Ru complex
reaches high oxidation states, the Ru center can only afford to coordinate 6 of these positions in
the first catalytic steps. Herein, we study this coordination rearrangement by characterizing new
Ru-bda complexes by several techniques including Variable temperature NMR, X-ray absorption
spectroscopy and single-crystal XRD.

20

References
1

S. W. Gersten, G. J. Samuels, T. J. J. Meyer, J. Am. Chem. Soc. 1982, 104, 4029–4030.
P. Garrido-Barros, C. Gimbert-Surinach, R. Matheu, X. Sala, A. Llobet, Chem. Soc. Rev. 2017, 46,
6088-6098.
3
I. López, M. Z. Ertem, S. Maji, J. Benet-Buchholz, A. Keidel, U. Kuhlmann, P. Hildebrandt, C. J.
Cramer, V. S. Batista, A. Llobet, Angew. Chem. Int. Ed. 2014, 53, 205-209.
4
J. Creus, R. Matheu, I. Peñafiel, D. Moonshiram, P. Blondeau, J. Benet-Buchholz, J. García-Antón,
X. Sala, C. Godard, A. Llobet, Angew. Chem. Int. Ed. 2016, 55, 15382-15386.
5
R. Matheu, I. A. Moreno-Hernandez, X. Sala, H. B. Gray, B. S. Brunschwig, A. Llobet, N. S. Lewis, J.
Am. Chem. Soc. 2017, 139, 11345-11348.
6
L. Duan, F. Bozoglian, S. Mandal, B. Stewart, T. Privalov, A. Llobet, L. Sun, Nat. Chem. 2012, 4, 418423.
7
C. J. Richmond, R. Matheu, A. Poater, L. Falivene, J. Benet-Buchholz, X. Sala, L. Cavallo, A. Llobet,
Chem. Eur. J. 2014, 20, 17282–17286.
8
L. Wang, L. Duan, Y. Wang, M. S. G. Ahlquist, L. Sun, Chem. Commun. 2014, 50, 12947-12950.
9
M. Schulze, V. Kunz, P. D. Frischmann, F. Würthner, Nat. Chem. 2016, 8, 576-583.
10
F. Li, B. Zhang, X. Li, Y. Jiang, L. Chen, Y. Li, L. Sun, Angew. Chem. Int. Ed. 2011, 50, 12276-12279
11
H. Dau, M. Haumann, Biochim. Biophys. Acta 2007, 1767, 472-483.
12
N. Cox, M. Retegan, F. Neese, D. A. Pantazis, A. Boussac, W. Lubitz, Science 2014, 345, 804-808.
13
R. Matheu, M. Z. Ertem, J. Benet-Buchholz, E. Coronado, V. S. Batista, X. Sala, A. Llobet, J. Am.
Chem. Soc. 2015, 137, 10786-10795.
14
R. Matheu, M. Z. Ertem, M. Pipelier, J. Lebreton, D. Dubreuil, J. Benet-Buchholz, X. Sala, A. Tessier,
A. Llobet, ACS Catal. 2018, 8, 2039–2048.
15
L. Duan, L. Wang, F. Li, F. Li, L. Sun, Acc. Chem. Res. 2015, 48, 2084-2096.
16
L. Duan, L. Wang, F. A. K. Inge, A. Fischer, X. Zou, L. Sun, Inorg. Chem. 2013, 52, 7844-7852
17
R. Matheu, L. Francàs, P. Chernev, M. Z. Ertem, V. S. Batista, M. Haumann, X. Sala, A. Llobet, ACS
Catal. 2015, 5, 3422-3429.
18
Y. Gao, X. Ding, J. Liu, L. Wang, Z. Lu, L. Li, L. Sun, J. Am. Chem. Soc. 2013, 135, 4219-4222.
19
F. Li, B. Zhang, X. Li, Y. Jiang, L. Chen, Y. Li, L. Sun, Angew. Chem., Int. Ed. 2011, 50, 12276-12279.
20
J. J. Concepcion, D. K. Zhong, D. J. Szalda, J. T. Muckerman, E. Fujita, Chem. Commun. 2015, 51,
4105-4108.
21
M. V. Sheridan, B. D. Sherman, R. L. Coppo, D. Wang, S. L. Marquard, K.-R. Wee, N. Y. Murakami
Iha, T. J. Meyer, ACS Energy Lett. 2016, 1, 231-236.
22
X. Sala, S. Maji, R. Bofill, J. Garcia-Anton, L. Escriche, A. Llobet, Acc. Chem. Res. 2014, 47, 504-516.
23
Q. Daniel, P. Huang, T. Fan, Y. Wang, L. Duan, L. Wang, F. Li, Z. Rinkevicius, F. Mamedov, M. S. G.
Ahlquist, S. Styring, L. Sun, Coord. Chem. Rev. 2017, 346, 206-215.
24
L. Duan, A. Fischer, Y. Xu, L. Sun, J. Am. Chem. Soc. 2009, 131, 10397-10399.
25
P. Pyykkö, J. Organomet. Chem. 2006, 691, 4336-4340.
26
J. Nyhlén, L. Duan, B. Åkermark, L. Sun, T. Privalov, Angew. Chem. Int. Ed. 2010, 49, 1773-1777.
27
D. W.Stephan, Comp. Inorg. Chem. II. Eds: Reedijk, J.; Poeppelmeier, K. Elsevier, Oxford, April
2013.
2

21

28

P. P. Samuel, K. C. Mondal, N. Amin Sk, H. W. Roesky, E. Carl, R. Neufeld, D. Stalke, S. Demeshko,
F. Meyer, L. Ungur, L. F. Chibotaru, J. Christian, V. Ramachandran, J. van Tol, N. S. Dalal, J. Am. Chem.
Soc. 2014, 136, 11964-11971.
29
C. Di Giovanni, A. Poater, J. Benet-Buchholz, L. Cavallo, M. Sola, A. Llobet, Chem. Eur. J. 2014, 20,
3898-3902.
30
I. Romero, M. Rodríguez, A. Llobet, M.-N. Collomb-Dunand-Sauthier, A. Deronzier, T. Parella, H.
Stoeckli-Evans, J. Chem. Soc., Dalton Trans. 2000, 1689-1694.
31
C. Di Giovanni, L. Vaquer, X. Sala, J. Benet-Buchholz, A. Llobet, Inorg. Chem. 2013, 52, 4335-4345.
32
Y. Zhao, D. Truhlar, Theor. Chem. Acc. 2008, 120, 215-241.
33
J. Pinson, F. Podvorica, Chem. Soc. Rev. 2005, 34, 429-439.
34
L. Tong, L. Duan, Y. Xu, T. Privalov, L. Sun, Angew. Chem. Int. Ed. 2011, 50, 445-449.
35
L. Alibabaei, B. D. Sherman, M. R. Norris, M. K. Brennaman, T. J. Meyer, Proc. Natl. Acad. Sci.
U.S.A. 2015, 112, 5899-5902.
36
R. G. Pearson, J. Am. Chem. Soc. 1963, 85, 3533-3539.
37
H. C. Stynes, J. A. Ibers, Inorg. Chem. 1971, 10, 2304-2308.

22