Evidence for Oxidative Decay of a Ru-Bound Ligand during Catalyzed Water Oxidation

Article abstract of DOI:10.1021/acscatal.6b03278

In the evaluation of systems designed for catalytic water oxidation, ceric ammonium nitrate (CAN) is often used as a sacrificial electron acceptor. One of the sources of failure for such systems is oxidative decay of the catalyst in the presence of the st

Full text of DOI:10.1021/acscatal.6b03278

Subscriber access provided by Queen Mary, University of London
Article
Evidence for Oxidative Decay of a Ru-Bound
Ligand During Catalyzed Water Oxidation
Husain Noman Kagalwala, Lianpeng Tong, Ruifa Zong, Lars Kohler, Mårten
S.G. Ahlquist, Ting Fan, Kevin J Gagnon, and Randolph P. Thummel
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b03278 • Publication Date (Web): 06 Mar 2017
Downloaded from http://pubs.acs.org on March 6, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 22
ACS Catalysis
1
2
3
4
5
6
7
8
9
Evidence for Oxidative Decay of a Ru-Bound Ligand
During Catalyzed Water Oxidation
Husain Kagalwala,a‡ Lianpeng Tong,a‡ Ruifa Zong,a Lars Kohler,a Mårten S. G. Ahlquist,b
Ting Fan,b Kevin J. Gagnonc and Randolph P. Thummela*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
aDepartment of Chemistry, 112 Fleming Building, University of Houston,
Houston, TX 77204ꢀ5003, bDivision of Theoretical Chemistry & Biology, School of
Biotechnology, KTH Royal Institute of Technology, 10691 Stockholm, Sweden,
cAdvanced Light Source, Lawrence Berkeley National Lab,
1 Cyclotron Rd., Berkeley, CA 94720
ABSTRACT
In the evaluation of systems designed for catalytic water oxidation, ceric ammonium nitrate
(CAN) is often used as a sacrificial electron acceptor. One of the sources of failure for such
systems is oxidative decay of the catalyst in the presence of the strong oxidant CAN (Eox = +1.71
V). Little progress has been made in understanding the circumstances behind this decay. In this
study we show that a 2ꢀ(2’ꢀhydroxphenyl) derivative (LH) of 1,10ꢀphenanthroline (phen) in the
complex [Ru(L)(tpy)]+ (tpy = 2,2';6',2''ꢀterpyridine) can be oxidized by CAN to a 2ꢀcarboxyphen
while still bound to the metal. This complex is, in fact, a very active water oxidation catalyst.
The incorporation of a methyl substituent on the phenol ring of LH slows down the oxidative
decay and consequently slows down the catalytic oxidation. An analogous system based on bpy
(2,2'ꢀbipyridine) instead of phen shows much lower activity under the same conditions. Water
molecule association to the Ru center of [Ru(L)(tpy)]+ and carboxylate donor dissociation was
proposed to occur at the trivalent state. The resulting [RuIII−OH2] was further oxidized to
[RuIV=O] via PCET process.
KEYWORDS: Mononuclear Ru catalysts, water oxidation, anionic ligands, 2ꢀ
carboxyphenanthroline, ligand decay
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 22
1
2
3
4
INTRODUCTION
5
6
7
8
During the past several decades the objective of using solar energy to convert water into its
elements has inspired a tremendous amount of work devoted to the design of effective
photocatalysts to drive this process.1ꢀ2 Special focus has been on water oxidation, which is
considered to be a bottleneck in achieving total water splitting.3 An important class of
compounds that is being developed for this purpose involves transition metal complexes of
polypyridineꢀtype ligands.4ꢀ5 Some of the earliest reported catalysts were mononuclear Ru(II)
complexes involving 2,2'ꢀbipyridine (bpy), 2,2'; 6,2''ꢀterpyridine (tpy), 1,10ꢀphenanthroline
(phen), and 2,9ꢀdi(pyridꢀ2'ꢀyl)ꢀ1,10ꢀphenanthroline (dpp).6ꢀ10
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
There are two fundamental features that are desirable in an effective water oxidation catalyst
(WOC). Firstly, the catalyst should exhibit a high turnover frequency (TOF) meaning that it
should facilitate water decomposition at a fast rate. Secondly, the catalyst should be very stable
under the reaction conditions. Ideally, the catalyst will provide many millions of turnovers
before becoming inactive. For a metalꢀpolypyridine catalyst a likely decomposition pathway
would involve oxidation of the polypyridine ligand, leading to fragmentation of the complex and
ultimately release of the metal.11ꢀ12 Sometimes the metal or metal oxide is generated as a
nanoparticulate material and these metal nanoparticles have been shown to serve as somewhat
sluggish but quite effective WOCs.13ꢀ14
In terms of catalyst design, it seemed important to have a water molecule bound to the metal
center.7, 15ꢀ16 This was further established by studies carried out using [Ru(tpy)(bpy)X]+ type
complexes.17ꢀ18 However, it was also discovered that water could add to coordinatively saturated
Ru when the metal is first converted to a higher oxidation state.19 The consequential involvement
of a Ru species with seven ligands demanded geometric features that allowed the formation of
the required pentagonal bipyramid geometry. It was found that the two fused 5ꢀmembered
chelate rings of a Ruꢀbound tpy would not allow the formation of a pentagonal bipyramid and
thus [Ru(tpy)2]2+ was not an active catalyst for water oxidation. However, by expanding the size
of one of the chelate rings to 6 atoms, a complex such as 1(Scheme 1), after losing two electrons,
would allow the attack of water and the consequent 7ꢀcoordinate geometry.20
Electronic effects have also been shown to play a key role in the design of effective water
oxidation catalysts. In particular, anionic ligands are expected to stabilize higher oxidation states
2
ACS Paragon Plus Environment

Page 3 of 22
ACS Catalysis
1
2
of the metal and thus lower the overall energy demands of catalyzed water oxidation.21ꢀ22 This
concept has been elegantly demonstrated by Sun and coworkers who examined Ru complexes of
6,6’ꢀdicarboxybpy and 2,9ꢀdicarboxyphen.23 Both systems were effective water oxidation
catalysts. Herein we present our findings on a family of [Ru(tpy)(N^N^O)]+ complexes (Scheme
1) used as water oxidation catalysts, with CAN as the sacrificial electron acceptor. We
incorporated both the steric advantage of 6,5ꢀchelation as well as the stabilizing effect of an
anionic ligand by examining the 2ꢀphenoxyphen system 2a and comparing it against its 5,5ꢀ
chelated 2ꢀcarboxyphen analog, 4 (Scheme 1). Expecting to see a difference in O2 evolution
activity, we instead obtained strong evidence for the oxidative decay of the 2ꢀphenoxyphen
ligand and its subsequent conversion into 2ꢀcarboxyphen while still bound to the metal. Using
the methyl derivative 2b slows down the rate of this conversion. Complex 4 was thus revealed to
be the authentic catalyst as supported by 1HꢀNMR, ESIꢀMS, Xꢀray and theoretical analysis. The
related bpyꢀbased complexes 3a,b and 5 showed much lower activity, highlighting the
importance of designing stable catalysts as well as moving towards more benign oxidizing agents.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Complex 1, and family of [Ru(tpy)(N^N^O)]+ complexes 2ꢀ5 studied in this work
RESULTS AND DISCUSSION
Preparation and Characterization of the Catalysts
The phenanthrolinyl phenol ligand 10a was prepared in three steps. First o-bromoanisole (6a)
was converted to the corresponding boronic acid 7a which was then coupled with 2ꢀchloroꢀ1,10ꢀ
phenanthroline (8) to provide the methoxyphen derivative 9a. Cleavage of the ether with HBr
gave the phenol 10a which then was used to prepare the heteroleptic complex 2a. The analogous
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 22
1
2
3
4
5
6
7
8
9
methyl substituted ligand 10b and its RuII complex 2b were prepared in a similar fashion. A
parallel route was followed for the preparation of the analogous phenol substituted bpy
ligands13a and 13b. These ligands, in turn, provided the heteroleptic complexes 3a and 3b. The
synthesis of complex 5 was reported in an earlier work24 and the complex 4 resulted from the
treatment of [Ru(tpy)Cl3] with 2ꢀcarboxyꢀ1,10ꢀphenanthroline (14). All these synthetic routes are
outlined in Scheme 2.The ligands, complexes and intermediates all gave clear 1H NMR spectra
that allowed their unambiguous structure assignment (see Figure S1 and S2 for 1H NMR of all
phen and bpyꢀbased complexes, respectively).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Synthesis of complexes 2-5.
The electronic absorption spectra of the complexes 2-5 were measured in acetonitrile and the
data is recorded in Table 1 as well as Figure S3. The phenoxy complexes 2a,b and 3a,b show a
metalꢀtoꢀligand charge transfer (MLCT) absorption in the 525ꢀ529 nm range as well as a very
weak low energy absorption at 714ꢀ718 nm. The carboxyphen and carboxybpy complexes 4 and
5 show a somewhat stronger MLCT absorption at 488ꢀ491 nm.
4
ACS Paragon Plus Environment

Page 5 of 22
ACS Catalysis
1
2
3
4
5
6
7
8
9
The electrochemical properties of the complexes were measured by cyclic voltammetry and
the redox potentials are recorded in Table 1. All four phenoxy complexes show very similar and
reversible first oxidation potentials in the range of +0.42ꢀ0.46 V. This potential is substantially
below the potential of +1.28 V recorded for [Ru(tpy)2]2+,20 meaning that the electronꢀdonating
phenoxy substituent helps to stabilize RuIII allowing RuII/III oxidation to proceed much more
readily. The reductions of 2a,b and 3a,b involve an electron being added to the most
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
electronegative ligand which is tpy. The first reduction occurs at ꢀ1.45 to ꢀ1.47 V and the second
reduction is observed at ꢀ1.79 to ꢀ1.80 V, somewhat more negative than the parent [Ru(tpy)2]2+.
Table 1. Electronic absorptiona and electrochemical propertiesb of Ru complexes.
Complex
λmax / nm (ε / Mꢀ1cmꢀ1)
270(43650), 308(70794),
475(16596)
E1/2ox (ꢁE)
1.28
E1/2red (ꢁE)
ꢀ1.27 (ref 15)
[Ru(tpy)2]2+
235 (55100), 273 (48100),
0.45 (66), 1.00d,
ꢀ1.47 (70), ꢀ1.79 (71)
2a
2b
294 (35300, sh), 313 (31700), 1.55d, 1.79d
320 (29300, sh), 376 (18300),
470 (7970, sh), 525 (10000),
716 (1580)
234 (45600), 274 (42900),
292 (32200, sh), 312 (28600), 1.35d, 1.70d
322 (25800, sh), 382 (14400),
469 (7290, sh), 525 (9020),
718 (1370)
0.42 (66), 1.05d,
ꢀ1.46 (70), ꢀ1.79 (70)
3a
3b
237 (44000), 273 (33900),
315sh (30800), 324 (33500), 1.24d, 1.60d
348sh (15800), 527 (8740),
715 (1820)
0.46 (57), 1.03d,
ꢀ1.45 (63), ꢀ1.79 (70)
ꢀ1.45 (71), ꢀ1.80 (73)
238 (43700), 272 (36700),
313 (30700), 324 (30000),
342sh (16800), 529 (8710),
714 (1390)
0.46 (60), 1.05d,
1.27d, 1.47d
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 22
1
2
3
4
5
6
7
8
9
4
5
230 (51800), 272 (67900),
313 (40100), 357 (4820),
416 (6280, sh), 488 (15700)
0.86 (97), 1.10d
ꢀ1.36 (71)c, ꢀ 1.63 (63)c
ꢀ1.37 (67), ꢀ1.64 (73)
233 (29200), 244sh (25900), 0.87 (61), 1.05d
274 (31300), 296 (40200),
314 (36300), 364 (6880),
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
491 (12600)
aMeasured using CH3CN solutions (10 ꢂM). bRecorded in CH3CN under Ar atmosphere,
using 1 mM analyte and 0.1 M N(Bu)4PF6 as supporting electrolyte; E1/2 = (Epa + Epc)/2 in
volts versus SCE, ꢁE = (Epa – Epc) in mV. cQuasiꢀreversible. dIrreversible, depicts Epa
values only.
The carboxy complexes 4 and 5 are also selfꢀ consistent with RuII,III oxidation occurring at +0.86ꢀ
0.87 V and reductions at ꢀ1.36 to ꢀ1.37 V and ꢀ1.63 to ꢀ1.64 V. It is interesting to note that even
though their redox properties are almost identical, 4 and 5 show very different WOC activity, as
demonstrated in the next section. All complexes display irreversible oxidations at potentials
between +1.00 V and +1.79 V, which could be tentatively assigned to oxidation of the chloride
counter ion,25 as well as oxidation of the phenolate ring (for 2a,b and 3a, b).26
If one considers the pKa of phenol of 10.0 compared with the pKa of 2ꢀ picolinic acid at 1.07,
it is clear that the latter is much more acidic meaning that the conjugate base is more stable and
thus a weaker donor. A weaker donor, in turn, should be less able to stabilize a higher oxidation
state of a bound metal, leading to the observed higher oxidation potentials for the 2ꢀcarboxyphen
and 2ꢀcarboxybpy complexes 4 and 5 (see pages S2ꢀS9 for experimental details).
Water oxidation catalysis
The complexes 2a and 4 were tested for their ability to catalyze water oxidation in the
presence of ceric ammonium nitrate (CAN) as a sacrificial electron acceptor in aqueous nitric
acid (10 mL, pH = 1). Given the steric advantage of the 6,5 chelation in 2a compared to the 5,5
chelation in 4, we expected a difference in activity for the two systems. Curiously, almost
superimposable oxygen evolution curves were obtained for both catalysts as can be seen in
Figure 1A (black and red lines for 2a and 4, respectively). The oxygen evolution was rapid
6
ACS Paragon Plus Environment

Page 7 of 22
ACS Catalysis
1
2
3
4
5
6
7
8
during the first 2 hours after which it declined gradually. A kinetic study was carried out (Figure
1B) and an initial rate of oxygen evolution of 0.13 sꢀ1 was measured for catalyst 2a. This value
is marginally less than the 0.15 sꢀ1 measured for catalyst 4.
To help better understand the catalyzed oxidation process, we examined the ESI mass spectra
of 2a in presence of increasing equivalents of CAN. We noticed that with 8 equivalents of CAN
a cluster of mass signals at m/z = 558.14 emerged (Figure S4), whose mass value and profile
corresponded to the mass signals of the [4]+ cation (m/z = 558.05). It appeared that the
phenoxide portion of 2a had been destroyed by oxidation. Such oxidation would likely proceed
through the formation of a p-benzoquinoneꢀtype intermediate that could further oxidize to a
COOH substituent.27ꢀ28 The huge excess of CAN (10,000 equivalents) over 2a would result in an
instantaneous conversion to 4, explaining the identical O2 evolution curves in Figure 1A.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. A: O2 evolution versus time and corresponding TON plots for catalysts 4 (red line), 2a
(black line) and 2b (green line). Conditions: 20 ꢂM catalyst and 0.2 M CAN in 10 mL HNO3
(pH = 1.0). B: Initial rate of O2 evolution versus concentrations of catalysts.
These preliminary results suggest that CAN serves a dual purpose: it promotes ligand
oxidation of the preꢀcatalyst (2a), and acts as a sacrificial oxidant for the actual catalyst
(presumably 4), which generates O2. To gain further insight, we initiated a methodical
investigation of this unusual process. We first synthesized and examined the catalytic activity of
the methylꢀsubstituted derivative 2b. It was expected that a 4ꢀmethyl substituent on the phenoxy
ring would impede ligand oxidation. Indeed 2b still functioned as a WOC but generated lesser
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 22
1
2
3
4
5
6
7
8
9
quantities of oxygen (Figure 1A, green line) at a considerably slower initial rate (0.06 sꢀ1, Figure
1B). We then examined the reaction using a higher catalyst concentration (0.2 mM) for each of
the phenꢀbased systems, 2a, 2b, and 4. In contrast to the activity seen at 0.02 mM (Figure 1A), a
slight difference in gas evolution (both initial and final) could be observed between 4 and 2a
(Figure S5A). This difference became significant upon using an even higher catalyst
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
concentration (2 mM) (Figure 2A). If we consider that each molecule of oxygen will require the
removal of 4 electrons by 4 CAN molecules, 10 mL of a 0.2 M CAN solution can theoretically
produce 500
catalyst 4 and the abrupt halting of the reaction after less than 10 minutes. For catalyst 2a we
observe a slower rate and the production of lesser quantities of gas (320 mol). The less reactive
μmol of oxygen. In Figure 2A we observe the generation of 410 μmol by the
μ
methylꢀsubstituted catalyst 2b produces the same amount but at an even slower rate.
This drastic change in behavior at higher catalyst concentrations can be explained by taking
into account the relative concentration of the sacrificial oxidant. Compared to the initial catalytic
conditions (Figure 1A), there are now lesser equivalents of CAN (100 ꢀ 1000 eq.) per catalyst.
The apparent consumption of CAN during ligand oxidation and catalyst conversion into 4 creates
a further deficit of the oxidant. This limits the overall O2 evolution reaction, resulting in the
lower gas evolution from reactions containing 2a and 2b. This was further illustrated by preꢀ
treating 2a,b with 8 equivalents of CAN and subjecting them to the conditions used in Figure
S5A. A slight improvement in performance was observed in terms of initial rate and overall O2
evolution. However the difference in activity compared to 4 still persists (Figure S5B), indicating
that excess CAN is necessary for complete conversion of 2a into 4.
8
ACS Paragon Plus Environment

Page 9 of 22
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Results from water oxidation studies using catalysts 2a, 2b and 4 (2 mM each) in
presence of 0.2 M CAN, in aqueous HNO3 (pH = 1). A: Gas evolution traces for catalysts 4 (red
line), 2a (black line) and 2b (green line). B: Chromatograms from GC analysis illustrating the
presence of CO2 in the reaction headspace of catalysts 2a and 2b.
Curiously enough, while 4 and 2b demonstrate an increase in gas evolution upon
changing catalyst concentration from 0.2 mM to 2 mM, no further increase was observed for 2a
(Table S3). Additionally, as noted previously, both 2a and 2b evolve the same amount of gas at a
2mM concentration. These results are again a consequence of the structural difference between
2a,b (since the relatively quicker conversion of 2a into 4 consumes CAN much faster than 2b),
and clearly suggest that at high catalyst concentrations, the activity is CAN limited.
We analyzed the head space gas at the end of the catalytic reaction (at 2mM catalyst
concentration) using a Porapak Q gas chromatography column and a thermal conductivity
detector. We found that CO2 was generated in the reaction of 2a and to a lesser extent in the
reaction of 2b, but no CO2 was generated in the reaction of 4 (Figure 2B). We also investigated
the behavior of the free ligands 10a,b and 2ꢀcarboxyphen (14) in the presence of CAN. Similar
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 22
1
2
3
4
5
6
7
8
9
to their respective complexes, ligand 10a with the 2ꢀphenoxy substituent showed much higher
quantities of CO2 evolution than ligand 10b with the 4ꢀmethylphenoxy group. The 2ꢀ
carboxyphen ligand 14 did not show any evidence of CO2 formation (Figure S6). All these
results further suggest degradation taking place only in the 2ꢀphenoxyꢀbased compounds. Since
other undetectable oxidation products may also be formed, it is difficult to quantify these
CO2 evolution experiments.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Treatment of the reaction solutions with KPF6 after water oxidation allowed precipitation and
isolation of the catalysts. 1HꢀNMR of these solids recovered from the reactions of 2a and 2b
revealed them to be identical to 4 (Figure S7). Additionally, ESI analysis of the solids revealed
the disappearance of the expected m/z peaks for catalysts 2a (Figure 3A) and 2b (Figure 3B) and
the presence of the peak belonging to the [4]+ cation instead, clearly indicating that both
phenoxyꢀbased complexes undergo a transformation to 4 through ligand oxidation. Consistent
with the GC and 1HꢀNMR analysis, complex 4 did not appear to undergo any change in its
original structure after the reaction (Figure S8), highlighting its robustness.
Figure 3. ESIꢀMS (CH3OH as solvent) of A) catalyst 2a and B) catalyst 2b before and after
water oxidation, depicting conversion to catalyst 4.
10
ACS Paragon Plus Environment

Page 11 of 22
ACS Catalysis
1
2
3
4
5
6
7
8
9
As was mentioned earlier, Sun and coworkers examined water oxidation catalysis involving
Ru(II) complexes of 6,6’ꢀdicarboxybpy and 2,9ꢀdicarboxyphen.23 They observed very different
behavior between these two closely related complexes with the dicarboxybpy system showing
remarkably high activity in water oxidation. For this reason we decided to examine the series of
complexes 3a, 3b, and 5 which are simply bpy analogs of the series 2a, 2b, and 4. The bpy
compounds were prepared in a manner directly analogous to the method used for the phen series
as outlined in Scheme 1. When tested for water oxidation using CAN as the sacrificial reagent,
under the same conditions (see Figure 1) as were initially employed for the phen catalysts, all
three bpy complexes, 3a, 3b, and 5 were essentially inactive (Figure S9). If the experiment is
repeated at 100ꢀfold higher concentration of the catalyst, similar to what is shown in Figure 2,
gas evolution is observed (Figure 4) but CO2 is also observed in all three cases. After gas
evolution ceases, an attempt was made to recover and identify the depleted catalyst by adding
KPF6 to precipitate the complex. Dark insoluble material was observed and is believed to be the
result of decomposition of the catalyst.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 22
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Results from water oxidation experiments carried out at higher catalyst concentrations
(2.5 mM) with CAN (0.2 M) in aq. HNO3. A) O2 evolution curves. B) Corresponding gas
chromatograms, depicting presence of CO2 in the headspace.
This inconsistent catalytic behavior of 3a, b and 5 suggests that decay of the catalyst may be a
competing pathway. Therefore, even though the rate of decay is slow it may be comparable to
the rate of O2 evolution, especially in the presence of a large excess of CAN. At low
concentration of 3a,b and 5, the catalyst rapidly decomposes and loses activity before very much
O2 evolution can be detected (see Figure S9). Accordingly, at higher concentrations of catalyst,
water oxidation can compete with catalyst decay. To investigate further, the catalytic activity of
5 was compared to the wellꢀknown analog [Ru(tpy)(bpy)H2O]2+ 18. The latter was found to
generate much more O2 than 5 at lower concentrations, where catalyst 5 was essentially inactive
(Figure S10). These results suggests that the low activity of 5 (and 3a,b) probably results from
the poor rate of water binding to the metal center, leaving the catalysts prone to oxidative
degradation.
Based on the experimental evidence, the carboxyphen complex 4 does not appear to
decompose readily upon treatment with CAN. As shown at the top of Figure 2B, essentially no
CO2 is detected in the headspace after reaction. By contrast, the carboxybpy complex 5 seems to
12
ACS Paragon Plus Environment

Page 13 of 22
ACS Catalysis
1
2
3
4
5
6
7
8
9
undergo decomposition as evidenced by CO2 formation (Figure 4). One could imagine that the
bpy moiety might be more prone to oxidative decomposition since the two pyridine rings are
connected only by a σꢀbond while a fused central benzo ring unites the two pyridine moieties in
phenanthroline. Oxidative decomposition of phen usually involves oxidation of the C5ꢀC6
bridging bond to first give a quinone that in turn oxidizes to a 3,3'ꢀdicarboxybpy.29 One might
imagine ways in which 5,6ꢀsubstitution on phen might slow down this oxidative decay.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A successful water oxidation catalyst is one that consumes the entire sacrificial reagent
(CAN) before the catalyst itself is destroyed by the strongly oxidizing conditions. We examined
this interplay between the catalyst (4) and CAN concentration in three parallel experiments
where the CAN concentration was kept constant at 0.2 M and the catalyst concentration was
varied from 50
had essentially stopped, more catalyst was added to one flask and more CAN was added to the
other. For the run at lowest catalyst concentration (50 M), about 200 mol of oxygen were
μM ꢀ 1 mM. Each experiment was run in duplicate and when oxygen evolution
μ
μ
produced after 5 hours. All the catalyst had been consumed so that the addition of excess CAN
had no effect but the addition of catalyst caused the resumption of oxygen evolution at a
somewhat slower rate (Figure 5A). With a 20ꢀfold increase in the catalyst concentration, the
reaction produced about 350 μmol of oxygen after 15 minutes. In this run all the CAN had been
consumed so that the addition of more catalyst had no effect while the addition of more CAN
caused oxygen evolution to resume (Figure 5C). In the intermediate experiment where the
catalyst concentration was 0.25 mM, both the catalyst and CAN were consumed (or destroyed) in
2 hours and the addition of either species independently did not result in more oxygen evolution
(Figure 5B). These results help assign an optimal CAN/catalyst ratio (800:1, Figure 5B) where
the effects of degradation balance out, and also ascertain that the reactions are catalyst or CAN
limited, depending on the deviation from this optimal CAN/catalyst ratio.
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 22
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Effect of addition of excess catalyst 4 and CAN on O2 evolution, at different initial
catalyst concentrations. Initial conditions: [Cat] + CAN (0.2 M) in aq. HNO3 (10 mL, pH = 1).
All reactions involved addition of 1 mmol of CAN (red dotted lines) along with addition of the
catalyst (solid red lines) after the pressure traces ceased to increase. A sharp pressure increase is
always observed at the time of addition.
Six versus 7-coordination: mechanistic considerations
If a sixꢀcoordinate metal complex such as 4 is going to hydrate the metal center by forming a
RuꢀOH2 bond, the system must either loose a ligand or become 7ꢀcoordinate. The geometry
required by a 7ꢀcoordinate species is pentagonal bipyramid. In the case of 4 a pentagonal
bipyramid would be highly distorted and hence rather unlikely. When complex 4 in D2O is
acidified by D2SO4 the 1H NMR spectrum changes somewhat indicating that protonation of the
carboxylic acid has occurred. Nevertheless, the five protons on the tpy ligand that integrate for
2H apiece continue to integrate for 2H in the protonated complex, indicating that Cs geometry
has been retained (Figure 6). The implication for the reaction of 4 is that protonation must occur
14
ACS Paragon Plus Environment

Page 15 of 22
ACS Catalysis
1
2
3
4
5
6
on the carboxy group which then dissociates from the metal center and is replaced by a water
molecule.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
Figure 6. H NMR of 4(Cl) in (a) D2O and (b) D2O containing 0.05 M D2SO4. Marked peaks
with integral of 2H are assigned to the tpy ligand of 4(Cl) and indicate Cs symmetry.
The hydration and Ru oxidation processes are clearly evidenced by comparison of observed
and calculated mass spectrum signals as illustrated in Figure 7. The peak centered at 557.97 can
be assigned to [4]+ (m/z = 558.05). Loss of one electron to CeIV and the addition of H2O gives a
peak at 287.93 that corresponds to [4 + H2O]2+ (m/z = 288.03). Finally the loss of a second
electron and two protons gives a peak at 574.05 that corresponds to [4 + O]+ (m/z = 573.97). It
should be noted that the detection conditions of the mass spectrometer are different from the
catalytic O2ꢀevolving conditions. Protonation of the [4 + O]+ species cannot be excluded in very
acidic aqueous solution.
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 22
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Observed and calculated mass spectrum signals of rutheniumꢀcontaining species after
treatment of 4(Cl) with one and two equivalents of CAN.
The possibility of hydrogenꢀbonding interactions taking place between a water molecule and
the bound carboxylic acid was demonstrated by Xꢀray crystal structure determination. Figure 8
shows well organized octahedral behavior for 4 along with a water molecule Hꢀbonded to the
carboxy substituent. The single counterion (triflate) supported a Ru(II) structure. The outer Ruꢀ
N bonds for the tpy ligand were slightly longer than the inner RuꢀN bond and the RuꢀO bond was
even longer at 2.15 Å. The bond between this same oxygen and C16 is longer (1.29 Å) than the
O2ꢀC16 bond (1.24 Å), indicating more carbonyl character in the latter bond. The interior angles
around Ru were all similar with the angle involving the phen ring (79.90 °) being slightly larger.
The dihedral angles are small and consistent with planar ligand structures (Table 2).
Figure 8. Thermal ellipsoid plot representation of crystal structures of 4(OTf) at 50%
probability; nonꢀpolar hydrogen atoms, fractional solvent and counter ions are omitted for
clarity.
16
ACS Paragon Plus Environment

Page 17 of 22
ACS Catalysis
1
2
Table 2. Selected bond lengths (Å), bond angles (o), and dihedral angles (o) for crystal structure
of 4(OTf)
3
4
5
6
7
8
Bond distance (Å)
9
Ru1−O1
Ru1−N1
Ru1−N2
Ru1−N3
Ru1−N4
Ru1−N5
O1−C16
O2−C16
O2−H1WA
O1−H2WA
2.146(1)
2.070(1)
1.969(1)
2.059(1)
1.979(1)
2.060(1)
1.289(2)
1.241(2)
1.969(3)
4.175
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bond angles (o)
O1−Ru1−N4
N1−Ru1−N2
N2−Ru1−N3
N4−Ru1−N5
76.39(4)
79.30(5)
79.33(5)
79.90(5)
Dihedral angles (o)
O1−C16−C17−N4
N4−C21−C25−N5
N2−C10−C11−N3
N1−C5−C6−N2
2.5(2)
1.1(2)
1.0(2)
2.8(2)
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 22
1
2
3
4
5
6
7
8
A theoretical study provides more insight on the specific interaction between complex 4 in
the trivalent oxidation state and water. We considered two possible coordination events of water
where either the carboxylate or a pyridyl is decoordinated. After testing different structures of
potential RuIIIꢀaqua isomers, we conclude that 4aq2+ (Scheme 3) is the most stable RuIIIꢀaqua
species, in which the carboxylate group dissociates from the metal center and forms a hydrogenꢀ
bonded network with ligated and solvated water molecules. In another scenario where pyridyl is
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2+
replaced by water, the resulting aqua species is ca 13 kcal/mol higher in free energy than 4aq
(Figure S11). The free energy of 4aq2+ is slightly higher compared to 42+, indicating feasible
equilibrium between these two species if the barrier between them is low. We then located
transition states as well as intermediates in the 42+ to 4aq2+ conversion and indeed found a very
low free energy barrier of merely 6.2 kcal/mol (Scheme 3), corresponding to a very fast rate of
the carboxylate/aqua ligand exchange. Therefore, 42+ and 4aq2+ should exist in equilibrium in the
aqueous solution. The importance of the hemilabile carboxylate ligand that facilitates water
association has been observed for other ruthenium complexes.30 It should be noted, however, that
of 42+ and 4aq2+ only 4aq2+ is capable of being oxidized to the RuIV state and the proposed active
RuV state via proton coupled electron transfer (PCET) processes. A recent report demonstrating
Feꢀcatalyzed water oxidation sheds light on an innerꢀsphere catalytic mechanism through the
identification of an active FeIV(O)(μꢀO)CeIV intermediate.31 As depicted earlier by inꢀsitu ESI
analysis of 4, the coordinatively saturated RuIV=O species is generated after addition of 2
equivalents of CAN (Figure 7), most likely proceeding via the transition states depicted below.
Given the sterics of the entire process, the formation of a similar heterobimetallic adduct is
highly improbable and thus CANꢀdriven oxidation of the Ru center most probably progresses via
an outerꢀsphere mechanism.
18
ACS Paragon Plus Environment

Page 19 of 22
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Calculated pathway for carboxylate/water exchange in trivalent complex 4. Relative
solvated free energies (brackets) are in kcal/mol. Bond distances of Ru−O(carboxylate) and
Ru−O(aqua) are shown in Å.
CONCLUSIONS
A series of 2ꢀhydroxphenyl derivatives of phen and bpy have been prepared using Suzuki
coupling methodology. These molecules deprotonate the hydroxyl group to bind with Ru(II) as
anionic ligands, forming a 6ꢀmembered chelate ring. When exposed to CAN in nitric acid (pH =
1), the parent phen system 2a rapidly oxidizes to a 2ꢀcarboxyphen system 4 without further
decay. This 2ꢀcarboxyphen complex shows good activity as a WOC, producing up to 80% of the
theoretical amount of dioxygen. The incorporation of a methyl group on the phenol ring leads to
10b and the corresponding Ru(II) complex 2b. Methyl substitution slows down the ligand
oxidation process but ultimately the same 2ꢀcarboxyphen system 4 is formed. An analogous
series of complexes were prepared based on bpy rather than phen (3a,b and 5). Contrary to what
has been observed for a somewhat similar dicarboxybpy system, this series of complexes showed
greatly diminished activity towards water oxidation under the same conditions used for the phenꢀ
based systems. An important lesson is that CAN is a powerful and sometimes selective oxidant,
often leading to catalyst decomposition. More realistic results are likely to be obtained if the
catalysts are activated by light rather than by CAN. However, in designing systems for
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 22
1
2
3
4
5
6
photochemical water oxidation, major focus must be on catalyst / photosensitizer stability to
ensure high activity under the oxidative reaction conditions.
7
8
9
ASSOCIATED CONTENT
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Supporting Information. Xꢀray crystallographic data for complex 4(OTf), experimental details,
1H NMR, mass and UVꢀvis spectra, gas chromatographs, oxygen evolution graphs, and a
description of DFT calculations.
AUTHOR INFORMATION
Corresponding author
*Email for RT: thummel@uh.edu
Author contributions
‡ These authors contributed equally.
ACKNOWLEDGMENTS
This material is based upon work supported by the U.S. Department of Energy, Office of Science,
Office of Basic Energy Sciences under Award Number DEꢀFG02ꢀ07ER15888, the Advanced
Light Source supported by the Director, Office of Science, Office of Basic Energy Sciences, of
the U.S. Department of Energy under Contract No. DEꢀAC02ꢀ05CH11231, and the Robert A.
Welch Foundation (Grant Eꢀ621).
REFERENCES
(1) Jafari, T.; Moharreri, E.; Amin, A. S.; Miao, R.; Song, W.; Suib, S. L. Molecules 2016, 21,
900.
(2) Andreiadis, E. S.; Chavarot Kerlidou, M.; Fontecave, M.; Artero, V. Photochem. Photobiol.
2011, 87, 946ꢀ964.
(3) Rüttinger, W.; Dismukes, G. C. Chem. Rev. 1997, 97, 1ꢀ24.
(4) Blakemore, J. D.; Crabtree, R. H.; Brudvig, G. W. Chem. Rev. 2015, 115, 12974ꢀ13005.
(5) Kärkäs, M. D.; Verho, O.; Johnston, E. V.; Åkermark, B. Chem. Rev. 2014, 114, 11863ꢀ
12001.
(6) Zong, R.; Thummel, R. P. J. Am. Chem. Soc. 2005, 127, 12802ꢀ12803.
(7) Concepcion, J. J.; Jurss, J. W.; Templeton, J. L.; Meyer, T. J. J. Am. Chem. Soc. 2008, 130,
16462ꢀ16463.
20
ACS Paragon Plus Environment

Page 21 of 22
ACS Catalysis
1
2
3
4
(8) Masaoka, S.; Sakai, K. Chem. Lett. 2009, 38, 182ꢀ183.
(9) Duan, L.; Xu, Y.; Gorlov, M.; Tong, L.; Andersson, S.; Sun, L. Chem. Eur. J. 2010, 16,
4659ꢀ4668.
5
6
7
(10) Tong, L.; Thummel, R. P. Chem. Sci. 2016, 7, 6591ꢀ6603.
(11) Limburg, B.; Bouwman, E.; Bonnet, S. ACS Catal. 2016, 6, 5273ꢀ5284.
(12) Limburg, B.; Bouwman, E.; Bonnet, S. Coord. Chem. Rev. 2012, 256, 1451ꢀ1467.
(13) Corbucci, I.; Ellingwood, K.; Fagiolari, L.; Zuccaccia, C.; Elisei, F.; Gentili, P. L.;
Macchioni, A. Catal. Today 2016, in press, http://dx.doi.org/10.1016/j.cattod.2016.10.030
(14) Hong, D.; Murakami, M.; Yamada, Y.; Fukuzumi, S. Energy Environ. Sci. 2012, 5, 5708ꢀ
5716.
(15) Polyansky, D. E.; Muckerman, J. T.; Rochford, J.; Zong, R.; Thummel, R. P.; Fujita, E. J.
Am. Chem. Soc. 2011, 133, 14649ꢀ14665.
(16) Concepcion, J. J.; Tsai, M.ꢀK.; Muckerman, J. T.; Meyer, T. J. J. Am. Chem. Soc. 2010, 132,
1545ꢀ1557.
(17) Yan, L.; Zong, R.; Pushkar, Y. J. Catal. 2015, 330, 255ꢀ260.
(18) Pushkar, Y.; Moonshiram, D.; Purohit, V.; Yan, L.; Alperovich, I. J. Am. Chem. Soc. 2014,
136, 11938ꢀ11945.
(19) Tseng, H.ꢀW.; Zong, R.; Muckerman, J. T.; Thummel, R. Inorg. Chem. 2008, 47, 11763ꢀ
11773.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(20) Kaveevivitchai, N.; Kohler, L.; Zong, R.; El Ojaimi, M.; Mehta, N.; Thummel, R. P. Inorg.
Chem. 2013, 52, 10615ꢀ10622.
(21) Kärkäs, M. D.; Åkermark, B. Chem. Rec. 2016, 16, 940ꢀ963.
(22) Duan, L.; Wang, L.; Li, F.; Li, F.; Sun, L. Acc. Chem. Res. 2015, 48, 2084ꢀ2096.
(23) Tong, L.; Duan, L.; Xu, Y.; Privalov, T.; Sun, L. Angew. Chem. Int. Ed. 2011, 50, 445ꢀ449.
(24) Norrby, T.; Börje, A.; Åkermark, B.; Hammarström, L.; Alsins, J.; Lashgari, K.; Norrestam,
R.; Mårtensson, J.; Stenhagen, G. Inorg. Chem. 1997, 36, 5850ꢀ5858.
(25) Wickramasinghe, L. D.; Zhou, R.; Zong, R.; Vo, P.; Gagnon, K. J.; Thummel, R. P. J. Am.
Chem. Soc. 2015, 137, 13260ꢀ13263.
(26) Iniesta, J.; Michaud, P.; Panizza, M.; Cerisola, G.; Aldaz, A.; Comninellis, C. Electrochim.
Acta 2001, 46, 3573ꢀ3578.
(27) Sridharan, V.; Menéndez, J. C. Chem. Rev. 2010, 110, 3805ꢀ3849.
(28) Li, X.ꢀY.; Cui, Y.ꢀH.; Feng, Y.ꢀJ.; Xie, Z.ꢀM.; Gu, J.ꢀD. Water Res. 2005, 39, 1972ꢀ1981.
(29) Eckhard, I.; Summers, L. Aust. J. Chem. 1973, 26, 2727ꢀ2728.
(30) Wang, Y.; Duan, L.; Wang, L.; Chen, H.; Sun, J.; Sun, L.; Ahlquist, M. S. ACS Catal. 2015,
5, 3966ꢀ3972.
(31) Codolà, Z.; Gómez, L.; Kleespies, S. T.; Que, L., Jr.; Costas, M.; LloretꢀFillol, J. Nat.
Commun. 2015, 6, 5865.
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 22
1
2
3
4
5
6
Table of contents graphic
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22
ACS Paragon Plus Environment