Redox couple involving NOx in aerobic Pd-catalyzed oxidation of sp3-C-H bonds: Direct Evidence for Pd-NO3-/NO2- interactions involved in oxidation and reductive elimination

Article abstract of DOI:10.1021/jacs.6b10853

NaNO₃ is used in oxidative Pd-catalyzed processes as a complementary co-catalyst to common oxidants, e.g., Cu^II salts, in C-H bond activation and Wacker oxidation processes. NaNO₃ and NaNO₂ (with air or O_2<

Full text of DOI:10.1021/jacs.6b10853

Article
pubs.acs.org/JACS
Redox Couple Involving NOx in Aerobic Pd-Catalyzed Oxidation of
sp3‑C−H Bonds: Direct Evidence for Pd−NO3 /NO2 Interactions
−
−
Involved in Oxidation and Reductive Elimination
Margot N. Wenzel,† Philippa K. Owens,‡ Joshua T. W. Bray, Jason M. Lynam, Pedro M. Aguiar,
Christopher Reed, James D. Lee, Jacqueline F. Hamilton, Adrian C. Whitwood, and Ian J. S. Fairlamb*
Department of Chemistry, University of York, York YO10 5DD, United Kingdom
S
* Supporting Information
ABSTRACT: NaNO3 is used in oxidative Pd-catalyzed processes as a
complementary co-catalyst to common oxidants, e.g., CuII salts, in C−H
bond activation and Wacker oxidation processes. NaNO3 and NaNO2 (with air
or O2) assist the sp3-C−H bond acetoxylation of substrates bearing an N-
directing group. It has been proposed previously that a redox couple is
operative. The role played by NOx anions is examined in this investigation.
Evidence for an NOx anion interaction at PdII is presented. Palladacyclic
complexes containing NOx anions are competent catalysts for acetoxylation of
8-methylquinoline, with and without exogenous NaNO3. The oxidation of 8-methylquinoline to the corresponding carboxylic
acid has also been noted at PdII. 18O-Labeling studies indicate that oxygen derived from nitrate appears in the acetoxylation
product, the transfer of which can only occur by interaction of 18O at Pd with a coordinating-acetate ligand. Nitrated organic
intermediates are formed under catalytic conditions, which are converted to acetoxylation products, a process that occurs with
(50 °C) and without Pd (110 °C). A catalytically competent palladacyclic dimer intermediate has been identified. Head-space
analysis measurements show that NO and NO2 gases are formed within minutes on heating catalytic mixtures to 110 °C from
room temperature. Measurements by in situ infrared spectroscopy show that N2O is formed in sp3-C−H acetoxylation reactions
at 80 °C. Studies confirm that cyclopalladated NO2 complexes are rapidly oxidized to the corresponding NO3 adducts on
exposure to NO2(g). The investigation shows that NOx anions act as participating ligands at PdII in aerobic sp3-C−H bond
acetoxylation processes and are involved in redox processes.
Sanford and co-workers3 reported a remarkable methodology
INTRODUCTION
■
for the acetoxylation of unactivated C(sp3)−H bonds, i.e., in
The catalytic and selective C−H bond functionalization of
oxime ether and pyridine-type derivatives, with the nitrogen
organic substrates is of fundamental importance in synthetic
chemistry.1 Significant effort has been directed toward solving
atom playing the role of directing group to Pd. The reactions
utilized Pd(OAc)2 as a catalyst and dioxygen as the terminal
oxidant in acetic acid (AcOH)/acetic anhydride (Ac2O)
the issue of challenging C−H bond functionalization reactions
in a variety of substrates. While much has been accomplished in
(Figure 1). Vedernikov and co-workers reported a similar
the field of selective C(sp2)−H functionalizations, reactions of
oxidative catalyst process using PdII−pyridinecarboxylic acid
systems.4
less-activated C(sp3)−H bonds, e.g., as found in secondary and
tertiary alkyls and allylic and benzylic systems, has proven more
challenging. In driving the development of challenging C(sp3)−
H bond functionalization processes, mechanistic studies (both
experimental and computational) are essentialthe key role
played by a particular metal, ligand (neutral or anionic),
additive, or specific substrate class, under an eclectic array of
reaction conditions, cannot be overstated. These points are
evidenced from a series of recent studies exploiting the redox
Addition of NaNO3proposed to be a redox co-catalyst3
was important for effective catalysis (NO was detected, formed
by decomposition of NaNO3, and it is of note that NaNO2 can
play a similar activating role). Using 18O-labeled dioxygen, it
was deduced that oxygen incorporation into the acetoxylated
product derives from acetic acid (solvent) and not O2. A key
−
−
question that we asked was if NO3 and/or NO2 are critical
ligands (pseudohalides) at Pd within the catalytic cycle? If Pd−
NO3/Pd−NO2 species are present, then C−O bond formation
ought to be favored over C−N bond formation as acetoxylated
products are the only ones observed, despite reductive
elimination of NO2 being feasible at Pd.
−
−
activity of nitrate (NO3 ) and/or nitrite (NO2 ) anions, vide
infra. In C−H bond functionalization chemistry, the role of
such anions at Pd is poorly defined, requiring mechanistic
examination, particularly if the catalytic behavior of reaction
processes utilizing these anions is to be fully realized. Of
particular note is the work of Stahl and co-workers, who
reported selective benzene acetoxylation using NOx-based
redox mediators at Pd.2
Received: October 20, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 1. Synthesis of Pd−NO3/Pd−NO2 Complexes
Figure 1. Proposed mechanism for C(sp3)−H bond functionalization3
using a NOx redox-couple at Pd.
−
Examining the role played by NO3 /NO2− anions is arguably
essential for the development of other redox-coupled Pd
catalysis, particularly as nitrite (AgNO2) has been shown to be
directly involved in a recently developed catalyst-controlled
Wacker oxidation.5 Grubbs and Stoltz further reported Pd-
catalyzed aerobic intramolecular aminoacetoxylation of alkenes
enabled by Cu(NO3)2, the latter acting as an electron-transfer
mediator.6 Also, Xu and co-workers demonstrated that MNO3
(M = K or Ag) can act as a co-catalyst in C−H bond
fluorination chemistry.7 Xu proposed the involvement of PdII/
PdIV species based on ESI-MS analysis of the catalytic reaction
under working conditions. Moreover, they suggested [Pd-
(NO3)]+ as the active catalyst species, although direct evidence
for its existence was not presented. These recent findings are
underpinned by a series of long-standing studies on other
catalytic processes involving nitrate/nitrite anions at PdII. For
example, formation of ethylene glycol monoacetate from
ethylene in acetic acid solution containing LiNO3 and
Pd(OAc)2 is accompanied by oxygen atom transfer from the
oxidant to the carbonyl group of the product (confirmed by
17O-labeling).8 Henry reported that NaNO3 and NaNO2 can be
used as oxidants in the C(sp2)−H oxidation of various
aromatics, where the nitrate/nitrite can play a role similar to
that of Cu.9 In the 1980s, several groups explored the catalytic
properties of [Pd(Cl) (NO2) (CH3CN)] in the oxidation of
olefins to either epoxides and ketones, as well as glycol
18 h), the acetoxylated product (2) was formed in 80% yield
(Scheme 1).3 The Pd(OAc)2 used throughout this study is
Pd3(OAc)6 (verified by IR and NMR; stated as >99% purity by
the supplier).18 In our hands, we recorded >99% conversion to
2 by both GC and 1H NMR spectroscopic analysis of the crude
reaction, validating it as a starting point for further study.
The overwhelming evidence from the field is that the first
committed step in the catalytic cycle involves C−Pd bond
formation by C−H bond functionalization (cyclopalladation),19
likely by a concerted metalation deprotonation (CMD)
reaction.20 Stoichiometric reaction of Pd(OAc)2 with 1 (1:1)
in AcOH at reflux for 1.5 h gave the dinuclear PdII complex (3)
in 85% yield (Scheme 1).
monoacetate.10−14 Backvall and Heumann proposed that the
̈
nitrite ligand can be an active ligand at PdII, where the oxygen
from PdNO2 is transferred to the acetoxylated organic
product.15
The dynamic behavior of nitrite at PdII has been of particular
interest to our group16 and the group of Raithby,17 i.e., Pd−
NO2 and Pd−ONO linkage isomerization in various palladium
complexes. Indeed, in expanding the scope of our work to
understand the intimate role played by nitrite and nitrate at PdII
and higher oxidation state Pd species, in addition to the role
played by such anions in other catalytic Pd chemistry, we were
attracted by the underlying role played by NOx anions in
Sanford’s C(sp3)−H bond acetoxylation mediated by PdII/
NaNO3,3 in addition to the broader context of the work
reported by Grubbs et al.5,6 Herein, we describe our
mechanistic findings on the role played by NaNO3 and Pd−
NO2/Pd−NO3 species in the acetoxylation of C(sp3)−H
bonds.
A single crystal of 3 was grown from CH2Cl2/pentane and
found to be suitable for X-ray diffraction study, confirming the
structure of the cyclopalladated system as a mixture of syn/anti
isomers (Figure 2), which is consistent with the other
characterization data. A prominent feature of the X-ray
structure is the slipped π-stacking between quinolinyl groups,
which occurs in both syn/anti isomers. A palladaphilic
interaction [dashed line between Pd(1) and Pd(2)] is indicated
as the bond distance is within the sum of the van der Waals
−
radii for the two Pd centers. The introduction of both NO3 /
NO2 anions to the PdII centers could be achieved starting
−
from the Cl-bridged dinuclear PdII complex 4 (synthesized in
80% yield by treatment of 3 with LiCl). The mononuclear PdII
complex 5 was formed quantitatively on treatment of 4 with a
slight excess of PPh3. Pd−NO3 (6a) and Pd−NO2 (6b)
mononuclear complexes could be synthesized from 5 by
treatment with AgNO3 and AgNO2, respectively.
RESULTS
■
We directed our mechanistic studies to 8-methylquinoline (1).
Under the conditions described by Sanford and co-workers (5
mol % Pd(OAc)2, 1 equiv of NaNO3, AcOH/Ac2O, air, 110 °C,
B
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
isomers containing an O-bound nitrite ligand are found to be
ca. 21−29 kJ mol−1 higher in energy than those containing N-
bound nitrite ligands. In the latter case, however, the cis-
[Pd(NO2)(C∧N)CH3CN] complex is found to be of lower
energy than trans-[Pd(NO2)(C∧N)CH3CN] by 9 kJ mol−1.
Reductive elimination of the nitrated ligand backbone, which
requires a cis-configuration, is feasible from this cis-Pd complex
vide infra.
Taken together, the X-ray structures of 6a and 7b and the
spectroscopic data for 6b and 7a indicate that nitrite and nitrate
ligands can coordinate to a PdII center containing a
cyclopalladated 8-methylenylquinoline.
Aerobic Oxidation of 8-Methylquinoline (1). The
reference complexes gathered in Scheme 1, along with the
benchmark catalyst Pd(OAc)2, were evaluated in the aerobic
oxidation of 1 to give 2 (Figure 4). For Pd(OAc)2, there was
Figure 2. X-ray structure of 3. Note that the structure is disordered
over two positions (only the anti-isomer shown here); the C(23) and
C(24) atoms are not resolved enough to be displayed as thermal
ellipsoids. Selected bond distances (Å) and angles (deg): Pd1···Pd2 =
2.8408(4), O1−Pd1 = 2.176(3), O2−Pd2 = 2.042(3), O3−Pd1 =
2.058(3), O4−Pd2 = 2.172(4), N1−Pd1 = 2.010(2), N2−Pd2 =
2.005(3), C1−Pd1 = 1.996(3), C11−Pd2 = 2.003(3); C1−Pd1−N1 =
83.96(11), C11−Pd2−N2 = 83.68(12), N2−Pd2−O2 = 173.56(11),
N1−Pd1−O3 = 174.82(11).
Complex 6a was crystallized from CH2Cl2/cyclohexene, and
its structure was confirmed by single-crystal X-ray diffraction
(Figure 3). The NO3 ligand is found to be trans to the sp3-
Figure 4. Catalytic activity of Pd complexes 3, 6a, 6b, 7a, and 7b
compared against Pd(OAc)2 (5 mol %) with varying equivalents of
NaNO3 in the aerobic oxidation of 1 to give 2 (conditions: AcOH/
Ac2O, 110 °C, air, as for Scheme 1; conversion percentage relates to
2).
Figure 3. X-ray structures of 6a (left) and 7b (right). Selected bond
distances (Å) and angles (deg) for 6a: C1−Pd1 = 1.9900(18), N1−
Pd1 = 2.0818(14), O1−Pd1 = 2.1250(13), N2−O1 = 1.289(2), N2−
O2 = 1.240(2), N2−O3 = 1.232(2), P1−Pd1 = 2.2725(4); C1−Pd1−
N1 = 81.61(6), C1−Pd1−O1 = 171.11(6), N1−Pd1−P1 = 175.42(4),
O1−Pd1−P1 = 92.66(4). Selected bond distances (Å) and angles
(deg) for 7b: C1−Pd1 = 2.009(2), N1−O1 = 1.237(3), N1−O2 =
1.237(3), N1−Pd1 = 2.006(2), N2−Pd1 = 2.0355(19), N3−Pd1 =
2.1185(19); C1−Pd1−N2 = 83.67(8), C1−Pd1−N3 = 177.09(8),
N1−Pd1−N2 = 172.26(7).
>99% conversion using 1 equiv of NaNO3. The % conversion
dropped to 45% when 0.1 equiv of NaNO3 was used, and
negligible conversion was noted in its absence. The dinuclear
PdII complex (3) is a catalytically competent species, showing a
relatively similar profile to Pd(OAc)2. The PPh3-containing
complexes 6a and 6b are also viable catalysts, with 6b being
more active under conditions using 1 equiv of NaNO3.
Complexes 3, 6a, and 6b are not catalytically competent in
the absence of NaNO3. The stand-out result from Figure 4 is
the catalyst efficacy of complex 7a, which gave product 2 with
24% conversion (ca. 5 catalytic turnovers) in the absence of
NaNO3 and air as the oxidant. The reaction mediated by 7a was
found to be more effective under 1 atm of O2, leading to a 36%
conversion to 2 (ca. 7 catalytic turnovers). The effect of nitrate
at PdII is therefore confirmed by this observation. The reason
for the lack of activity of 6a and 6b is likely due to the
phosphine ligand altering the kinetics of the catalytic process,
particularly comparing directly 6a with 7a (CH3CN being more
labile than PPh3).23 Complex 7b, containing the nitrite ligand,
was inactive without NaNO3.
carbon center and cis to phosphorus and nitrogen (quinolinyl).
The Pd−ONO2 bond distance of Å = 2.125(13) is similar to
that of related complexes.21 In addition, the CH3CN adducts 7a
and 7b could be synthesized quantitatively from 4, which were
fully characterized. The X-ray structure of 7b is shown in Figure
3, which shows that the NO2 ligand is trans to nitrogen
(quinolinyl). The Pd−NO2 bond distance, 2.006(2) Å, is
similar to that of related Pd complexes reported by our group13
and others.22
It did not prove possible to obtain single crystals of either 7a
or 6b for X-ray crystallography studies. DFT calculations
therefore allowed us to assess the single-point energies of the
possible isomeric structures (see the Supporting Information
for details). The trans-configuration is preferred in complex 7a
(ΔGcis−trans = +13 kJ mol−1). For complex 7b, cis- and trans-
isomeric forms can possess either N-bound or O-bound nitrite
ligands, giving four isomeric possibilities. Both cis- and trans-
A further series of reactions (1 → 2) were conducted using
7a as the catalyst with varying equivalents of NaNO3 under an
atmosphere of either air or O2 (Figure 5; for completeness,
selected data from Figure 4 are included to reveal the essential
trends). From these data, complex 7a is found to be a viable
catalyst, especially when lower equivalents of NaNO3 are used.
C
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 2. Incorporation of 18O into Acetoxylation Product
a
Using NaN18O3 under Different Conditions (N2/air/O2)
Figure 5. Catalytic activity of Pd complex 7a, with varying equivalents
of NaNO3 in the presence of either air or O2 in the aerobic oxidation
of 1 to give 2 (other conditions as for Figure 4).
Finally, the effect of the nitrate equivalents on 7a revealed
that the relative drop in % conversion to 2 was more
pronounced under air, whereas an absolute drop is evidenced
under O2 (1 atm). Good % conversions were observed between
0.05 and 0.1 equiv of NaNO3.
It is worth noting that ceric ammonium nitrate (CAN;
[(NH4)2Ce(NO3)6]) can be used as a co-catalyst in place of
a
Mechanism shows one pathway for 18O incorporation.
•
NaNO3. CAN is known to generate NO3 radicals, which are
powerful one-electron oxidants.24 The reaction of 1 → 2
mediated by Pd(OAc)2 (5 mol %), AcOH/Ac2O (7:1), 110 °C,
24 h and air, in the presence of 0.25 equiv of CAN, gave 2 in
95% yield.
processes.26 The incentive for this part of the investigation was
to assess whether PdIII or PdIV species could be detected in
reactions involving a redox couple and nitrate. As PhI(OAc)2 is
a known powerful oxidant in Pd chemistry, we anticipated that
it was useful to assess whether higher oxidation state Pd species
formed from it were comparable with the Pd−NOx chemistry.
Using standard acetoxylation conditions, the nitrate oxidant
reacts with the in situ generated palladacycle. Having previously
confirmed palladacycle 3 as catalytically competent,27 we
envisaged that higher oxidation state Pd species, i.e., PdIII
dinuclear or PdIV mononuclear complexes as depicted in
Scheme 3, could be formed by reaction with typical oxidants.26
NaN18O3-Labeling Experiments. The reaction of 1 → 2
mediated by Pd(OAc)2 (5 mol %) was conducted using 1 equiv
of NaN18O3 (77.9% 18O-enriched) in AcOH/Ac2O at 110 °C
for 24 h under either N2, air or O2. Reactions were essentially
quantitative (by GC analysis). For the reaction under N2
(rigorous exclusion of air and degassing using Schlenk
technique), the presence of 1 equiv of NaN18O3 efficiently
promotes the oxidative process, with 16% 18O incorporation
(determined by MS) observed in product 2. A similar outcome
was recorded for the reaction conducted under air (15.6% 18O
incorporation; note with a vast excess of AcOH/Ac2O present,
if free 18O exchange occurred in solution then the %
incorporation would be negligible). In the case of the reaction
under an O2 atmosphere, negligible 18O incorporation into 2
was recorded. The results suggest that there is an exchange
process occurring; i.e., the 18O label is washed out of NaN18O3
in the reaction conducted under O2 (i.e., into AcOH/Ac2O).25
Under N2 or air the exchange process is diminished, leading to
some 18O localization into acetoxylation product 2. The
Scheme 3. Reaction of PhI(OAc)2 with Palladacycle 3
outcome confirms that 18O is transferred from NO3 to Pd
−
to product 2, confirming its role as an active spectator ligand
(note: we have not been able to determine whether the
remaining 18O has been transferred to AcOH/Ac2O under the
catalytic conditions). One potential pathway showing intra-
molecular 18O-transfer is given in Scheme 2; the formation of
Pd−O or Pd−OH species may explain the washing out of the
18O label from NaN18O3 in the presence of pure O2. In
Sanford’s study,3 18O derived from 18O2 (with unlabeled
NaNO3) was transferred into a different acetoxylation product
to a small extent (5% 18O incorporation). Taken together with
our results, this confirms that O2, AcOH/Ac2O and NaNO3 are
intimately connected through exchange processes.
The reaction of palladacycle 3 with PhI(OAc)2 in AcOH at −12
°C (note: at a substantially lower reaction temperature than the
reported catalyzed process3) was monitored by 1H NMR
spectroscopy (700 MHz). The consumption of 3 was noted
over ca. 1.5−2 h with the appearance of acetoxylation product 2
and PhI, but no other Pd intermediates were recorded (see the
Supporting Information for NMR spectra).
A stoichiometric reaction of 3 with NaNO3 was conducted
under air at 50 °C in AcOH-d4. In comparison with the
PhI(OAc)2 reaction conducted at −12 °C, detailed above, a
higher temperature was deemed necessary for this stoichio-
On the Involvement of Higher Oxidation State Pd
Species. Higher oxidation state Pd intermediates have been
isolated and characterized in related aerobic acetoxylation
D
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
metric reaction to (1) aid solubility of NaNO3 in AcOH and
(2) mirror the working catalytic reaction conditions more
closely (without excess 1 to slow down overall catalytic flux27).
1
The time course of the reaction was monitored by H NMR
1
spectroscopic analysis (see stacked H NMR spectra in Figure
6). For comparison, reference spectra for substrate 1, product
Figure 7. X-ray structure of trans-Pd(OAc)2(2)2 (8). Selected bond
distances (Å) and angles (deg) for 8: N1−Pd1 = 2.0404(14), O1−Pd1
= 2.0064(13); O1−Pd1−N1 = 88.24(5), O1(−X,−Y,1−Z)−Pd1-N1 =
91.76(5), N1−Pd1−N1(−X,−Y,1−Z) = 180.0.
Figure 6. Reaction of 3 with NaNO3 (1 equiv) in AcOH-d4 at 50 °C
monitored by 1H NMR spectroscopy (400 MHz). The 1H NMR
spectra for reference compounds 1−3 are given.
2, and complex 3 are included. Within 0.5 h of reaction, the
proton signals belonging to 3 began to broaden, forming other
species which are ill-defined on the NMR time scale. New
products were formed after 1.5 h. One of these appears to
correspond to 2 (identified by blue arrows) vide infra. Another
organic product is formed as indicated by a singlet at ca. δ 6.20
ppm (highlighted by an orange arrow; see the following for
confirmation of the structure). Other proton signals are
observed, the most prominent being that at ∼δ 9.4 ppm
(highlighted by a green arrow) the complete structure of which
could not be discerned. After 2 h, a new product (which is 9)
appears, possessing six correlated proton environments (high-
lighted by yellow inset boxes).
Figure 8. X-ray structure of 9 (top, single molecule; bottom, showing
the packing between molecules and Pd1−Pd1(2−X,1−Y,−Z) interaction).
Note that all hydrogens and three molecules of co-crystallized acetic
acid are omitted from this drawing. Selected bond distances (Å) and
angles (deg) for 9: Pd2···Pd1 = 2.7977(2), Pd1−N1 = 1.9931(16),
Pd2−N2 = 1.9868(17), Pd1−O3 = 2.0210(13), Pd2−O2 =
2.0189(15); Pd1−O5 = 1.9538(15), Pd2−O7 = 1.9455(13); N1−
Pd1−Pd2 = 102.60(4), N2−Pd2−Pd1 = 100.53(5), N1−Pd1−O3 =
177.04(5), N2−Pd2−O2 = 178.17(6).
Stopping the reaction at 4 h and crystallization gave two sets
of single crystals which were found to be suitable for X-ray
diffraction study. The first crystal recorded was found to be
trans-Pd(OAc)2(2)2, complex 8 (Figure 7). This indicates that
the species observed by 1H NMR (highlighted by blue arrows)
is more likely trans-Pd(OAc)2(2)2 (8) rather than 2 (confirmed
1
by an authentic H NMR spectrum of 8). The second crystal
was determined to be palladacycle 9, where the ligand
backbone had been fully oxidized to a carboxylate under the
conditions of the reaction (Figure 8; see below for further
crystallographic analysis). After 4 h (Figure 6), only one
dominant species remains, which is palladacycle 9.
The X-ray structure for palladacycle 9 is found in the P-1
space group (Z = 2). There is evidence of a PdII···PdII
interaction; the Pd1−Pd2 distance is 2.7977(2) Å, which is
shorter than the Pd1−Pd2 distance in palladacycle 3
(2.8408(4) Å). In addition, it is pertinent to mention the
packing between molecules, which shows that Pd1−
Pd1(2−X,1−Y,−Z) is separated by 3.0755(3) Å and Pd2−
Pd2(1−X,1−Y,−Z) by 3.0631(3) Å.
A preparative-scale reaction of 3 with NaNO3 was conducted,
which gave 9 in quantitative yield (Figure 9). Palladacycle 9 was
also accessed by reaction of 8-quinolinecarboxylic acid 10 with
Pd(OAc)2 (1:1) at 110 °C for 1.5 h, affording 9 in 68% yield.
E
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
with other species). The 15N signal at δ 269 ppm correlates
with δ 5.91 and δ 8.29 from the diastereotopic hydrogens of a
CH2 (vide infra) and 8.63 (HQuin). Both of these nitrogen shifts
are within the expected range for M−ONO2, M−NO2, and R−
NO2 species,28 and the observed signal-to-noise ratio requires
that both species are derived from the Na15NO3 and cannot be
from the quinoline ligand which is at natural abundance (see
the Supporting Information).
Figure 9. Direct reaction of palladacycle 3 with NaNO3.
A repeat experiment to monitor the reaction using natural
abundance NaNO3 reveals the H signal at 6.08 ppm to be a
1
1
The minor species present by H NMR (see Figure 6) were
singlet rather than a doublet (2.1 Hz) and the signal at 5.91
ppm also shows evidence of additional unresolved coupling in
the reaction with Na15NO3.
of concern to us, particularly as they could not be satisfactorily
characterized with the data available. 15N labeling was therefore
employed to pursue possible Pd−NOx interactions in these
minor species. In an otherwise equivalent reaction, palladacycle
To assist with the characterization of this species, 14/15N
NMR spectral data (natural abundance) for a series of
compounds (in both CDCl3 and AcOH-d4) of interest to this
study were collected (Table 1).
1
3 was reacted with Na15NO3. A H{15N} HMQC experiment
allowed two clear nitrogen chemical environments (at δ 269
and 340 ppm) to be proton correlated. The major signal
observed at δ 6.08 ppm correlates with the 15N signal at δ 340
1
ppm (Figure 10). A weaker long-range H−15N coupling was
Table 1. 15N NMR Data for Selected Compounds in This
Study (298 K)
seen between δ 340 ppm and δ 8.85 ppm (HQuin), overlapping
Figure 10. (a) 1H{15N} HMQC NMR spectrum (700 MHz) from the
reaction of 3 with Na15NO3 in AcOH-d4 after 100 min at 80 °C. A 1D
1H spectrum (b) provides a more accurate reflection of the
populations of the two 15NOx species; δ(1H) 6.08 ppm (red square,
11); δ(1H) 5.91 ppm (blue circle, Pd−NOx complex) and δ(1H) 5.72
ppm (green triangle, 2). Note: the signal at 5.72 ppm is shown off
scale.
a
b
No 15N signal observed due to dynamic behavior. Nitrogen shift
c
measured from 1D 14N spectrum. Insoluble material.
F
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The 15N chemical shift of NO3 is δ 374 ppm in AcOH-d4.
conducted. Dissolution of reaction samples in MeOH before
electrospray injection facilitated effective ionization. Analysis of
a sample after 1 h reaction revealed the presence of a sodiated
dinuclear Pd ion I that contains the cyclometalated backbone at
Pd, suggesting that it possesses μ-1,3-NO3 and two terminal
OMe ligands (Table 2). The observed and calculated isotope
patterns for the proposed structures are gathered in Table 2.
−
The quinoline nitrogen (Nquin) is typically found in the region δ
230−233 ppm (in AcOH-d4) in compounds 1, 3, and 10,
whereas it is δ 274 ppm for compound 2. The 15N chemical
shifts were higher in CDCl3 than AcOH-d3 as at the lower pH
the quinoline is in equilibrium with the quinolinium ion.29
Therefore, the 15N chemical shift of complex 9 of δ 180 ppm in
CDCl3 is likely lower in AcOH-d4; the latter solvent could not
be used due to the poor solubility of 9. The 15N NMR chemical
shifts are in keeping with those recently described by Stahl and
Table 2. ESI-MS and LIFDI-MS Analysis of the Reaction of 3
a
with NaNO3 in AcOH at 50 °C at 1 h
co-workers.30 At natural abundance, no H−15N correlation
1
could be observed for the NO2 of 7b; however, the 1D 14N (see
the Supporting Information) in CDCl3 reveals a chemical shift
of 312 ppm for the Pd−NO2. In AcOH, 7b is catalytically active
1
and slowly produces 2 (1−2 days). H{15N} HMQC spectra
(under these conditions revealed a signal δ 340 ppm (15N),
which correlates with the methylene protons at δ 6.08 ppm.
The observation of the same intermediate, as in the reaction of
3 with Na15NO3, indicates that this species must be formed via
a Pd−NO2 species.
1
The spectrum of 11 showed H−15N correlations at δ 6.10,
340 ppm (CH2, NO2) and δ 9.14, 284 ppm (o-Hquin, Nquin).
1
Despite the small difference in the H shift (ascribed to pH
differences of the solution), we assign the (δ 6.08, 340 ppm)
signal to compound 11.
The signal at δ 5.91 ppm (Figure 10) arises from one of the
hydrogens of a diastereotopic CH2 group. The signal appears as
a doublet with a splitting of 15.5 0.2 Hz at 16.4 and 9.4 T
1
1
(700 and 400 MHz H, respectively). 1D H exchange spectra
(tmix = 0.3 s) upon selective irradiation of the signal at 5.91 ppm
revealed a signal at δ 8.29 ppm which appears as a doublet with
a splitting of 15.5 Hz and is of the same phase as the selectively
irradiated signal indicating that it arises due to chemical
(EXSY) rather than spin−spin exchange (NOE). A multiplicity-
edited 1H{13C} HSQC (see the Supporting Information),
indicated that both of these hydrogens were attached to “CH2-
type” carbons and within the resolution of the experiment, both
were attached to the same carbon (78.0 0.3 ppm). The very
high chemical shift of δ 8.29 ppm could be explained by H-
bonding with the NOx anion. The only other compound where
the methylquinoline CH2 was found to be diastereotopic was
the dimeric palladacycle 3 (anti-isomer: δ 2.56 and 3.47 ppm, 2J
a
The structures are suggested for key molecular ion peaks observed by
ESI-MS and LIFDI (I = [C22H22N3O5Pd2Na]+ and I′ =
[C22H22N3O5Pd2]•+, and it is feasible that the methoxy/NO3 ions
could occupy different positions and offer bridging or terminal
coordination to Pd.
A radical cation I′ was observed by LIFDI-MS, which
provides support for these higher oxidation state Pd species
containing a nitrate ligand. At 1.5 h, other ions were observed,
including a partially oxidized dinuclear Pd species by ESI-MS,
which supports the formation of complex 9 under working
reaction conditions and consistent with the NMR studies (see
the Supporting Information). After 2 h, the major ions
observed by ESI-MS and LIFDI-MS were derived from 9.
Other ions were detected under the working reaction
conditions, for which attempts to fit reasonable structures
and chemical formulas were unsuccessful.
Reaction Kinetics and Appearance of Gases under
Working Catalyst Conditions. The originally reported
reaction of 1 → 2, mediated by Pd(OAc)2 and 1 equiv of
NaNO3, in air-saturated AcOH, was conducted for 18 h at 110
°C (Figure 11).3 We learned that this reaction was over within
minutes at this temperature. Lowering the reaction temperature
to 80 °C allowed the evolution of product 2 to be monitored by
GC analysis (temperature at t = 0 min was 80 °C). In all cases,
classical homogeneous behavior is apparent and there is no
evidence to suggest the involvement of Pd nanoparticles
(heterogeneous processes) under these oxidizing conditions.33
The reaction mediated by Pd(OAc)2 reaches 94% conversion
after 10 min. The catalytic competency of the presynthesized
complexes 3, 7a, 7b, 8, and 9 was established under conditions
otherwise identical to those of Pd(OAc)2. It is apparent that the
kinetics observed for the reaction mediated by Pd(OAc)2 are
different from that mediated by 9. The appearance of a defined
induction period suggests that 9 is a precatalyst. The initially
2
= 12.9 Hz, syn-isomer: δ 3.41 and 3.77 ppm, J = 8.5 Hz). In
order to gauge the relative size (solvodynamic radius) of the
compound giving rise to the signal at 5.91 ppm, a 2D diffusion-
ordered (DOSY) spectrum was collected. The known species in
the reaction mixture (2, 11, and AcOD-d4) had diffusion
coefficients of 4.21 × 10−10 m2 s−1, 4.92 × 10−10 m2 s−1, and
8.04 × 10−10 m2 s−1, respectively. The signal at 5.91 ppm was
found to arise from a complex with a much lower diffusion
coefficient, D = 2.69 × 10−10 m2 s−1, indicating it is of
significantly larger size.
ESI-MS and LIFDI-MS analysis. Electrospray ionization
mass spectrometry (ESI-MS31) is a useful tool for character-
izing charged organometallic compounds. Liquid injection field
desorption ionization mass spectrometry (LIFDI-MS32) can
also be used to characterize neutral organometallic species.
Both ESI-MS and LIFDI-MS are soft ionization methods,
making them suitable for detecting Pd species present under
working reaction conditions. Xu and co-workers successfully
detected Pd intermediates by ESI-MS in their C−H bond
fluorination chemistry.7 The direct ESI-MS analysis of the
reaction of palladacycle 3 with NaNO3 in AcOH at 50 °C was
G
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 12. Head-space analysis of the reaction 1 → 2, mediated by
Pd(OAc)2 (5 mol %) and NaNO3 (1 equiv) in AcOH/Ac2O (t0 = 25
°C), with heating to 110 °C (over 15 min) (note: NOx = NO + NO2).
Figure 11. Kinetic plots for the reaction of 1 → 2 catalyzed by
Pd(OAc)2 (5 mol % Pd) and complexes 3, 7a, 7b, 8, and 9 (1 equiv of
NaNO3, 80 °C, air-saturated AcOH). Reactions with dinuclear Pd
complexes employed 10 mol % Pd. Each data point is an average of
three independent GC injections (error bars for each conversion point
are shown against each curve).
light affected the catalytic chemistry (cf. only a marginal
difference in acetoxylation reaction kinetics was recorded).
Moreover, the formation of O3 was not detected under these
reaction conditions.
Following on from the results gained by the headspace
analysis, we tested the reaction 1 → 2, mediated by Pd(OAc)2
(5 mol %) and NaNO3 (1 equiv) in AcOH at room
temperature, which gave 2 in 90% conversion after 3 days.
A preparative-scale reaction of 2, catalyzed by 5 mol %
Pd(OAc)2, 1 equiv of NaNO3 in AcOH/Ac2O at 110 °C, gave
liberated acid 10 in 23% yield (Figure 13). Future studies will
formed palladacycle 3 and trans-Pd(OAc)2(2)2 8 share catalytic
activity that is commensurate with Pd(OAc)2. Within the series,
the most active catalyst was 7b, containing an NO2 anion,
which was found to be more active than the corresponding
catalyst 7a containing an NO3 anion.
In reactions of 1, mediated by Pd(OAc)2 in the presence of 1
equiv of NaNO3, the evolution of NO is seen on heating
between 80 and 110 °C. The appearance of a red/brown gas is
visible after ∼2 min reaction time, which is attributed to NO2↑
vide infra. Infrared spectroscopic analysis of the catalytic
reaction 1 → 2 under working conditions at 80 °C (under air)
revealed that two new bands at 2213 and 2178 cm−1 grew in
after 2 h reaction time (see the Supporting Information). These
bands were attributed to N2O,34 the formation of which occurs
on full consumption of substrate 1.
Figure 13. Example demonstrating the feasibility of a catalytic reaction
2 → 10.
Gas headspace analysis, using a customized chemilumines-
cence NOx analyzer (see the Experimental Section), of the
reaction of 1 → 2 catalyzed by Pd(OAc)2 in the presence of 1
equiv of NaNO3, allowed the gases evolved to be characterized.
It was necessary to run a reaction using 2.3 mmol of 1 in
AcOH/Ac2O (40 mL, 7/1, v/v), giving a concentration in 1 of
0.0575 M. The gases were analyzed directly from the top of the
reaction condenser, equipped with a glass tubing-based piece
allowing the entrance of a flux of “zero-air” and the exit of the
gases to the detectors. The volume dilution was 15 L min−1.
The reaction was started at 25 °C, as the large volume of
AcOH/Ac2O would require ca. 20 min to reach 110 °C. The
experiment conducted in this way showed that mixing 1,
Pd(OAc)2, and NaNO3 in AcOH/Ac2O led to the evolution of
NOx gases at 25 °C (depicted as product evolution curves in
Figure 12). The concentrations of NOx gases reported are given
as mixing ratios, in parts per million. The total amount of NOx
formed is ca. 140 ppm (∼1.2 mmol), consuming over half of
the NaNO3 (2.3 mmol) initially present. The experiment
confirms that NO is formed rapidly under these reactions. NO2
also forms, which continues to increase at the expense of NO;
therefore, as demonstrated by this experiment, air (O2) is
playing the role of reoxidant.
address the feasibility of this catalytic transformation in more
general terms, e.g., in terms of substrate scope, as it has the
potential to be synthetically useful.
Lastly, reaction of 7b with NO2 gas was conducted as, based
on literature precedent,35 we envisaged that the NO2 ligand at
PdII could be oxidized to NO3 (Figure 14).
Figure 14. Reaction of 7b with NO2 (gas) monitored by in situ
infrared spectroscopic analysis (ReactIR): tentatively suggested
oxidized Pd complexes.
The interaction of nitrate with air in light could also lead to
the formation of O3. However, there was no clear evidence that
H
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Bubbling a CH2Cl2 solution of 7b with NO2 led to loss of
CH3CN from 7b. Reaction monitoring by in situ spectroscopic
analysis allowed us to visualize solubilization of NO2(g)
(saturated at ca. t = 5 min), loss of 7b, and formation of a
new compound (Figure 15). On the basis of the IR data, a
(f) Another intermediate from (e) was revealed by
stoichiometric reaction of 3 with Na15NO3, proposed
as being a Pd-containing species possessing a NOx ligand.
(g) The formation of nitrated product 11 has been observed
in stoichiometric reactions of 3 and NaNO3 in acetic
acid, which is converted to acetoxylation product 2, in
the presence of Pd at 50 °C, and in the absence of Pd at
110 °C.
(h) ESI-MS and LIFDI-MS provided indirect evidence for
the formation of dinuclear palladacycles containing NO3
anion (with AcOH/MeOH), which also confirmed the
evolution of an oxidized ligand backbone (leading to the
isolation of 9).
(i) Complex 9 is catalytically competent and exhibits an
induction period, contrasting with the catalytic behavior
of the other Pd complexes (3, 7a, 7b, and 8).
(j) The in situ infrared spectroscopic analysis of a catalytic
reaction catalyzed by Pd(OAc)2 at 80 °C showed the
evolution of N2O, after prolonged heating (up to 12 h).
(k) Head-space analysis of a catalytic reaction, started at 25
°C and heated to 110 °C, revealed the rapid formation of
NO and NO2 gases within 2 min. The formation of NO
and subsequent oxidation to NO2, confirms that NO is
being oxidized under the aerobic reaction conditions.
(l) The mechanistic observations made above allowed
identification of a room temperature catalytic reaction
(90% conversion to 2, over 3 days, was recorded).
Figure 15. In situ IR reaction monitoring of 7b → 7aa in the presence
of NO2 (g) (1251 cm−1) at 20 °C. The black dots show depletion of
7b at 1223 cm−1, and red dots show appearance of 7aa at 717 cm−1.
dinuclear Pd complex possessing two μ-1,3-NO3 ligands (7aa)
is tentatively proposed, based on a characteristic deformation
mode at 717 cm−1, assigned to μ-1,3-NO3, similar to a related
complex reported by Yu et al. (766 cm−1).21 Extensive analysis
by LIFDI was unable to reveal the molecular ion of 7aa,
although a very weak molecular ion for 7ab was found (see the
Supporting Information). The 1H NMR spectra of this complex
suggest that the quinoline protons have become highly
deshielded (distinct from nitrated organic product 11), in
keeping with the structure proposed containing NO3 anions.
Based on the above evidence, a catalytic cycle involving
dinuclear palladacycle 3 as an entry point is plausible. The
precise redox chemistry involving NO3 and NO2 anions at Pd is
undoubtedly complicated. Direct evidence for NO3 and NO2
anions acting as ligands at PdII has been obtained (by X-ray
diffraction, NMR spectroscopy, and MS analysis). Transfer
(exchange) of 18O from NaN18O3 into the acetoxylation
product 2 confirms that 18O exchange between NO3 and OAc
anions occurs at Pd, which is in keeping with the observations
DISCUSSION AND CONCLUSION
■
From this study, evidence supporting the role of nitrate and
nitrite anions at Pd in the aerobic Pd-catalyzed oxidation of
nonactivated sp3-C−H bonds has been gathered. 8-Methyl-
quinoline 1 was identified as a model substrate to conduct this
study, which affords acetoxylated product 2, and was originally
reported in Sanford’s work3 as part of an expanded substrate
scoping study.
made by Backvall and co-workers in the acetoxylation/
̈
hydration reactions of olefins.15
The presence of nitrated product 11 indicates that NO2 is
acting as a ligand at Pd and that reductive elimination is
feasible. Moreover, such products are intermediates to
acetoxylated product 2.
The formation of complex 9 indicates that the reaction of 1
→ 2 is running under strongly oxidizing conditions. Prolonged
heating will therefore lead to loss of product 2, impacting on
overall yield. The primary reason for this is that trans-
Pd(OAc)2(2)2 (8) goes on to form complex 9 under the
reaction conditions.
NOx gases readily evolve in the aerobic Pd-catalyzed
oxidation of nonactivated sp3-C−H bonds.3 Our headspace
measurements indicate that NO is rapidly formed under
working reaction conditions, with measurements starting at
room temperature. We believe that the solubility of NaNO3 is
critical here, where higher temperatures favor dissolution.
Oxidation of Pd−NO species to Pd−NO2 and Pd−NO3
species is reported,36 and there is an interesting complemen-
tarity here with iron−nitrato/iron−-nitrosyl chemistry.37,16b
The formation of N2O is noted on prolonged heating and
reaction times (by in situ IR measurements). We cannot
identify a clear role for N2O, other than it is a dead-end for
catalysis, which is in keeping with the timing of its appearance,
long after substrate turnover has completed. N2O formation
The findings include (all catalytic processes refer to the
reaction of 1 → 2 conducted at 110 °C in AcOH/Ac2O, unless
specified) the following:
(a) Cyclopalladation of 1 with Pd(OAc)2 affords a dinuclear
Pd complex 3 (in AcOH), which is catalytically
competent.
(b) Palladacycles containing NO3 and NO2 ligands (7a and
7b, respectively) are both viable (pre)catalysts.
(c) Palladacycle 7a functions as an active catalyst in the
absence of additional NaNO3 co-catalyst.
(d) Under either N2 or air, a catalytic reaction mediated by 1
equiv of NaN18O3 showed incorporation of the 18O label
into acetoxylated product 2 (in AcOH). Negligible 18O
incorporation was recorded under O2.
(e) Stoichiometric reactions of 3 with NaNO3 at either 50 or
80 °C revealed the intermediacy of several species −
trans-Pd(OAc)2(2)2 (8) and the oxidized Pd complex 9
could be characterized (in AcOH-d4). The latter complex
formed quantitatively, after overnight heating at 80 °C.
I
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
can be explained by dimerization of HNO as a termination
reaction.38
under the reaction conditions. Intermediate IV can also derive
from NO2 oxidation of validated dimer intermediate 3 (step 7).
It is further recognized that NOx-free intermediates could form
under the conditions, either by oxidation of 3 (step 6) or anion
metathesis with IV (step 8), giving V. Against this latter
proposal are the 18O-labeling studies indicating that NOx is at
Pd for oxygen exchange to occur intramolecularly with acetate
(18O appearing in product 2). Reductive elimination from V
gives acetoxylated product 2 (step 9) and a Pd species that
requires reoxidation to Pd(OAc)2 (step 10), which is feasible
under the oxidizing conditions of the reaction. The nitrated
organic product 11 can form directly from IV, in keeping with
that proposed.38 Here, a Pd reoxidation step is again necessary,
akin to step 10 (not shown). Step 11 shows how nitrated
organic product 11 can be converted into 2, confirmed by
experiment as being more rapid in the presence of Pd, although
it occurs without.
We propose a mechanism involving higher oxidation state Pd
species (Figure 16). It is unclear whether PdIV species are
We regularly encountered formation of complex 9 in our
studies, which can derive from IV by oxidation of the
methylene group to a carbonyl (step 13). The characterization
of the potential intermediate 12, from the 15N NMR labeling
studies, provides support for this hypothesis−the very high
chemical shift for a methylene proton (δ 8.29 ppm) indicates
an interaction of an NOx anion (at Pd), which can ultimately
assist this oxidation step.
To conclude, the principal aim of this investigation was to
delineate the role played by nitrate and nitrite anions in the
aerobic Pd-catalyzed oxidation of nonactivated sp3-C−H bonds.
Here, we were able to draw on the promising methodology
described by Sanford3 and other Pd-catalyzed processes taking
advantage of such redox anions.8,9 Recent synthetic develop-
ments,5,7 when taken together with the mechanistic work
described herein, serve to highlight the potential of redox active
anions such as nitrate and nitrite in oxidative Pd-catalyzed
processes. The approach taken to testing the catalytic viability
of suitable Pd−NOx complexes will, we hope, be more widely
adopted by others in studying Pd/NOx-mediated organic
transformations, which may also prove useful in understanding
regiochemical consequences in appropriate transformations.
The intimate role of NOx at Pd is undoubtedly complicated,
but through appropriate isotopic labeling and complementary
mechanistic studies, useful experimental evidence can be
gathered.
Figure 16. Postulated mechanism based on the experimental evidence
gathered in this study (where O∧O is shown as acetate, the bridging
ligand could also be nitrate).
involved in these transformationswith the experimental
evidence available, we propose that dinuclear PdIII species
play an important role in catalysis, which is based on the
combined NMR, MS, and IR data that we have available from
this study. The mechanism makes the reasonable assumption
that Pd3(OAc)6 breaks down to form Pd(OAc)2 under the
reaction conditions. The caveat to our hypothesis is that
structurally characterized PdIV species containing NO, NO2,
and NO3 and tris(pyrazolyl)borate ligands have been reported
EXPERIMENTAL SECTION
■
General Experimental and Instrumental Details. Reagents
were purchased from Sigma-Aldrich, Alfa Aesar, Acros Organics, or
Fluorochem and used as received unless otherwise stated. Isotopically
labeled chemicals were purchased from Icon Isotopes, Ltd. and used as
received. Pd(OAc)2 was obtained from Precious Metals Online. Dry
THF, CH2Cl2, and CH3CN were obtained from a Pure Solv MD-7
solvent system and stored under nitrogen. THF was also degassed by
bubbling nitrogen through the solvent with sonication. Dry methanol
was obtained by drying over 3 Å molecular sieves. Petroleum ether
refers to the fraction of petroleum that is collected at 40−60 °C.
Reactions requiring anhydrous conditions were carried out using
Schlenk techniques (high vacuum, liquid nitrogen trap on a standard
in-house built dual line). Room temperature upper and lower limits are
stated as 13−25 °C, but typically 21 °C was recorded. Brine refers to a
saturated aqueous solution of NaCl.
by Cam
́
pora and co-workers.39 These model systems are of
potential relevance to the catalytic C−H acetoxylation reaction
mechanism, particularly as oxygen was shown to oxidize a
PdIV−NO complex to PdIV−NO3.
Cyclopalladation can occur through either monomer or
dimer complexes (Figure 16, steps 1 and 5), a proposal
supported by Musaev’s recent study.40 NO2 /NO3− metathesis
−
is facile at Pd, as demonstrated by experiment (step 2). Species
II can either react with liberated NO2 to give III or directly
form 11 by reductive elimination (indicated as step 12). Species
III can then be directly oxidized under the reaction conditions
to give IV (step 7), a species speculated by Liu and co-workers
as being formed in reactions employing tert-butyl nitrite.41 The
experimental evidence, particularly the 15N NMR evidence,
tentatively supports such type of an intermediate being formed
Thin-layer chromatography (TLC) was carried out using Merck
5554 aluminum-backed silica plates (silica gel 60 F254), and spots
were visualized using UV light (at 254 nm). Where necessary, plates
were stained and heated with one of potassium permanganate,
anisaldehyde or vanillin as appropriate. Retention factors (Rf) are
J
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
reported in parentheses along with the solvent system used. Flash
column chromatography was performed according to the method
reported by Still et al.42 using Merck 60 silica gel (particle size 40−63
μm) and a solvent system as reported in the text.
measuring at 1 Hz on both channels simultaneously. Channel 1 (ch1)
measures NO and has a nominal sensitivity of 3.5 counts per second/
parts per trillion (cps/ppt). Channel 2 (ch2) measures NOx (NO +
NO2) by dissociating NO2 to NO with a photolytic converter (blue
light converter) that irradiates the sample gas at 390 nm. The nominal
sensitivity on channel 2 is 4 cps/ppt. NO2 is obtained by subtracting
ch1 from ch2. The effective limit of detection is ∼2.5 ppt/min NO and
∼6.5 ppt/min NO2. Note: ppm values are given in Figure 12.
Prechamber zero measurements are taken every 5 min for 30 s. The
whole system is kept at 4 Torr vacuum pressure. The sample flow rate
is 1 sLpm per channel, MFC controlled. Calibration for sensitivity is
done by standard addition of NO; converter efficiency is determined
by gas-phase titration (GPT) of NO to NO2 with O3. Artifact is
determined in zero air, which is made by taking dried (−40 °C dew
point) compressed air, passing it through 13x molecular sieves, then
sofonofil to remove NOx and through activated carbon to remove
ozone and VOCs. The 2B Technologies Model 202 Ozone Monitor is
designed to enable accurate and precise measurements of O3 ranging
from low ppb (precision of ∼1 ppb) up to 100000 ppb (0−100 ppm)
based on the absorption of UV light at 254 nm.
All DFT calculations were performed using the TURBOMOLE V5
package using the resolution of identity (RI) approximation.43 Initial
optimizations were performed at the (RI-)BP86/SV(P) level, followed
by frequency calculations at the same level. All minima were confirmed
as such by the absence of imaginary frequencies; energies, geometries,
and vibrational frequencies are presented. Single-point calculations on
the (RI)BP86/SV(P)-optimized geometries were performed using the
hybrid PBE0 functional and the flexible def2-TZVPP basis set. The
(RI)PBE0/def2-TZVPP SCF energies were corrected for their zero-
point energies, thermal energies (ΔE), and entropies (obtained from
the (RI-)BP86/SV(P)-level frequency calculations at 298.15 K, ΔG
298.15). In all calculations, a 28 electron quasi-relativistic ECP
replaced the core electrons of Pd. No symmetry constraints were
applied during optimizations. Calculated XYZ coordinates, single-point
energies, and vibrational spectra are reported in the Supporting
Information.
NMR spectra were obtained in the solvent indicated using a JEOL
ECX400 or JEOL ECS400 spectrometer (400, 101, and 162 MHz for
1H, 13C, and 31P, respectively) or a Bruker 500 (500, 126, and 202
1
MHz for H, 13C, and 31P, respectively). Chemical shifts are reported
in parts per million and were referenced to the residual nondeuterated
solvent of the deuterated solvent used (CHCl3 TMH = 7.26 and TMC
= 77.16 (CDCl3), CDHCl2 TMH = 5.31 and TMC = 54.0 (CD2Cl2),
1
(CHD2)SO(CD3) TMH = 2.50 and TMC = 39.52 {SO(CD3)2}, H
and 13C respectively). AcOD-d4 was referenced to 2.04 ppm (1H
only). Spectra were typically run at a temperature of 298 K. All 13C
NMR spectra were obtained with 1H decoupling. 31P NMR were
1
referenced internally and recorded with H decoupling. NMR spectra
were processed using MestrNova software (versions 5.3, 7.03 and 8.1).
The spectra given below were typically saved as .emf files in
MestrNova and inserted into a Microsoft Word document. For the
1H NMR spectra, the resolution varies from 0.15 to 0.5 Hz; the
coupling constants have been quoted to 0.5 Hz in all cases for
1
consistency. H NMR chemical shifts are reported to two decimal
places; 13C and 31P NMR chemical shifts are reported to one decimal
place (note: for some 13C signals it was necessary to report to two
decimal places).
High-field NMR spectra were acquired at 16.4 T on a Bruker
Avance II spectrometer equipped with 5 mm TXI or BBO probes.
Spectra of natural abundance samples (1, 2, 3, 7b, 10, and 9) and the
monitoring of the reaction of 3 to 2 were acquired using the TXI
probe. The monitoring of the reaction of 1 to 2 in the presence of
Na15NO3 was performed with the BBO probe to facilitate direct
observation of 15N species. 1H NMR spectra (700.13 MHz) were
acquired using a Bloch-decay (single-pulse) sequence with a 5 s recycle
delay, 15 ppm spectral widths, and 3 s acquisition times that are the
1
sum of 32 coadded transients. H{15N} correlation spectra (700.13
and 70.96 MHz, respectively) were acquired using an HMQC pulse
sequence with pulsed field-gradient based coherence selection. Spectra
were acquired with different spectral widths and/or transmitter offsets
in the 15N dimension in order to ensure detect for possible leasing of
signals in the indirect dimension. 1H chemical shifts are reported
relative to TMS, and 15N chemical shifts are reported relative to liquid
ammonia.
Infrared spectra were obtained using either a Unicam Research
Series FTIR (KBr IR) or a Bruker ALPHA-Platinum FTIR
Spectrometer with a platinum−diamond ATR sampling module.
Where indicated, reactions were monitored in situ using a Mettler
Toledo ReactIR ic10 with a K6 conduit SiComp (silicon) probe and
MCT detector.
MS spectra were measured using a Bruker Daltronics micrOTOF
MS, Agilent series 1200LC with electrospray ionization (ESI and
APCI) or on a Thermo LCQ using electrospray ionization, with <5
ppm error recorded for all HRMS samples. LIFDI mass spectrometry
was carried out using a Waters GCT Premier MS Agilent 7890A GC
(usually for analysis of organometallic compounds when ESI or APCI
are not satisfactory ionization methods). Mass spectral data are quoted
as the m/z ratio along with the relative peak height in brackets (base
peak = 100). Mass to charge ratios (m/z) are reported in daltons.
High-resolution mass spectra are reported with <5 ppm error (ESI) or
<20 ppm error (LIFDI). For clarity, LIFDI data are reported for 106Pd,
the most abundant natural isotope of Pd.
Melting points were recorded using a Stuart digital SMP3 machine.
Gas chromatographic analysis was carried out using a Varian CP-
3800 GC equipped with a CP-8400 autosampler. Separation was
achieved using a DB-1 column (30 m × 0.32 mm, 0.25 μm film
thickness) with carrier gas flow rate of 2 mL min−1 and a temperature
ramp from 50 to 250 °C at 20 °C min−1. The injection volume was 1
μL with a split ratio of 10 (details of the kinetics measurements are
given below).
Characterization of Pd Complexes. Dinuclear PdII−PdII
Complex 3 [Pd(8-MQ) (OAc) Dimer]. 8-Methylquinoline (186 mg,
1.3 mmol) and Pd(OAc)2 (291 mg, 1.3 mmol) were suspended in
glacial acetic acid (20 mL), refluxed for 1.5 h, and then filtered. Water
(150 mL) was added to the filtrate, and the mixture was allowed to sit
overnight. The precipitated product was collected by filtration and
recrystallized from CH2Cl2/hexane to yield 3 as a bright orange solid
as a 2:1 mixture of isomers (342 mg, 85% yield). Crystals suitable for
X-ray diffraction were grown by layering of a CH2Cl2 solution with
pentane: mp 193−195 °C; 1H NMR (major isomer) (400 MHz,
CDCl3) δ 8.51 (dd, J = 5.0, 1.5 Hz, 2H, H2), 7.87 (dd, J = 8.4, 1.5 Hz,
2H, H4), 7.22 (dd, J = 8.3, 5.0 Hz, 2H, H3), 7.11−6.99 (m, 2H, H5),
6.89 (dd, J = 8.1, 7.1 Hz, 2H, H6), 6.70 (dd, J = 7.1, 1.1 Hz, 2H, H7),
3.45 (d, J = 13.8 Hz, 2H, CH2), 2.55 (d, J = 13.8 Hz, 2H, CH2), 2.14
(s, 6H, CH3); 13C NMR (101 MHz, CDCl3) δ 181.5 (CO), 152.7
(C9), 148.9 (C2), 148.2 (C10), 136.5 (C4), 127.3 (C6), 126.8 (C7),
122.9 (C5), 121.1 (C3), 120.5 (C8), 41.4 (CH2), 24.5 (CH3); LIFDI-
MS m/z 615.94 (calcd for C24H22N2O4Pd2 615.97); ESI-MS m/z
614.9621 (calcd for C24H21N2O4Pd2 ([M − H]+) 614.9620; IR (KBr,
cm−1) 3390, 1566, 1504, 1408, 1050, 817, 776, 671. Anal. Calcd for
C24H22N2O4Pd2: C, 46.85; H, 3.60; N, 4.55. Found: low C (46.17); H,
3.59; N, 4.48.
Dinuclear PdII−PdII Complex 4 [Pd(8-MQ) (Cl) Dimer]. To a
solution of 3 (265 mg, 0.43 mmol) in acetone (15 mL) was added
lithium chloride (73 mg, 1.72 mmol) in distilled water (10 mL),
resulting in immediate precipitation of the product. The precipitate
was collected by filtration, and washed with methanol/water (1:1 v/v)
to yield 4 as a pale yellow powder (195 mg, 80% yield). Pyridine-d5
was added to an NMR tube containing a sample of complex 4 in
CDCl3 in order to increase its solubility in common NMR solvents.
Upon addition of pyridine-d5, a yellow solution was formed,
presumably due to the formation of the monomeric complex
[PdCl(8-MQ) (pyr-d5)]. Residual pyridine signals have been omitted
The chemiluminescence NOx analyzer is a custom dual channel
instrument from Air Quality Design (AQD, Inc., Golden, CO)
1
from the reported NMR data for clarity: mp 249−251 °C; H NMR
K
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(400 MHz, CDCl3) δ 9.57 (dd, J = 5.1, 1.5 Hz, 2H, H2), 8.10 (dd, J =
8.3, 1.5 Hz, 2H, H4), 7.54−7.50 (m, 2H, H5) 7.47−7.42 (m, 2H, H7),
7.38 (dd, J = 7.3, 7.0 Hz, 2 H, H6), 7.31 (dd, J = 5.1, 8.3 Hz, 2H, H3)
3.54 (s, 4H, CH2); 13C NMR (101 MHz, CDCl3) δ 153.1 (C9), 151.8
(C10), 147.3 (C8), 137.2 (C4), 128.7 (C5), 127.8 (C6), 127.6 (C7),
123.5 (C2), 121.2 (C3), 30.1 (CH2); LIFDI-MS m/z 567.90 (calcd for
C20H16Cl2N2Pd2 567.88); IR (KBr, cm−1) 3059, 2919, 2283, 1563,
1505, 1385, 1375, 1316, 1056, 825, 787, 762, 676. Anal. Calcd for
C20H16Cl2N2Pd2: C, 42.28; H, 2.84; N, 4.93. Found: C, 41.92; H, 2.65;
N, 4.71.
(CN), 23.0 (CH2), 3.6 (CH3); LIFDl-MS m/z 309.98 [M − NCCH3]
(calcd for C10H8N2O3Pd 309.96); IR (KBr, cm−1) 2924, 2853, 2253,
1732, 1508, 1438, 1384, 1290, 1027, 824, 783, 327.
Mononuclear PdII Complex 7b [Pd(8-MQ)(NO2)(NCCH3)]. Method
A, starting from 4 and AgNO2: Complex 4 (140 mg, 0.25 mmol) and
silver nitrite (91 mg, 0.6 mmol) were suspended in a mixture of
CH2Cl2 and CH3CN (8 mL, 3:1 v/v) and stirred at room temperature
for 24 h. The reaction mixture was filtered, and the filtrate
concentrated in vacuo to afford 7b as a yellow solid in quantitative
yield (235 mg). Method B, starting from 3 and NaNO2: Complex 3
(105 mg, 0.171 mmol) and sodium nitrite (118 mg, 1.710 mmol) were
suspended under argon in distilled CH3CN (5 mL) and stirred at
reflux for 6 h. The reaction mixture was filtered while hot under argon.
Yellow crystals suitable for X-ray diffraction were grown in the filtrate
when cooled. The crystals were isolated by filtration and dried to yield
Mononuclear PdII Complex 5 [Pd(8-MQ) (PPh3)Cl]. A solution of
Pd(8-MQ) (Cl) dimer 4 (290 mg, 0.5 mmol) and PPh3(288 mg, 1.1
mmol) in CH2Cl2 (20 mL) was stirred at room temperature for 15
min under nitrogen. The reaction mixture was filtered through Celite,
and hexane was added to induce precipitation. The product was
collected by filtration and recrystallized from CH2Cl2/hexane to yield
5 as fine yellow crystals in quantitative yield (579 mg): mp 205−209
1
7b in 61% yield (71 mg): mp 239−242 °C; H NMR (400 MHz,
CDCl3) δ 8.53 (dd, J = 5.0, 1.5 Hz, 1H, H2), 8.29 (dd, J = 8.4, 1.5 Hz,
1H, H4), 7.66−7.57 (m, 2H, H5+H7), 7.49 (t, J = 7.6 Hz, 1H, H6),
7.43 (dd, J = 8.4, 5.0 Hz, 1H, H3), 3.94 (s, 2H), 2.01 (s, 3H); 13C
NMR (101 MHz, CDCl3) δ 152.6 (C2), 149.6 (C9), 147.2 (C8),
138.7 (C4), 129.4 (C6), 129.2 (C7), 128.8 (C5), 127.9 (C10), 124.4
(C3), 121.8 (CN), 27.2 (CH2), 2.1 (CH3); LIFDl-MS m/z 589.85
[2M − 2CH3CN] (calcd for C20H16N4O4Pd2 589.93), 389.06 [M −
NO2 + 8-MQ] (calcd for C20H16N2Pd 390.03); IR (KBr, cm−1) 2919,
2849, 1505, 1435, 1374, 1216, 1055, 1018, 818, 778.
1
°C; H NMR (400 MHz, CDCl3) δ 9.79 (m, 1H, H2), 8.27 (dd, J =
8.3, 1.6 Hz, 1H, H4), 7.86−7.77 (m, 6H, Ar−H), 7.58 (d, J = 7.9 Hz,
1H, H5 or H7), 7.52 (ddd, J = 8.3, 5.0, 1.2 Hz, 1H, H3), 7.46−7.39
(m, 9H, Ar−H), 7.39−7.36 (m, 1H, H6), 7.29 (dq, J = 7.2, 1.3 Hz, 1H,
H5 or H7), 2.85 (d, J = 3.8, 2H); 13C NMR (101 MHz, CDCl3) δ
152.2, 150.4, 148.2, 137.9, 135.0 (d, J = 11.9 Hz), 131.6, 131.1, 130.5
(d, J = 2.6 Hz), 129.1, 128.4 (d, J = 10.8 Hz), 127.7, 123.8, 121.6, 33.4;
31P NMR (162 MHz, CDCl3) δ 35.2; LIFDI-MS m/z 545.05 (calcd
for C28H23ClNPPd 545.03). Anal. Calcd for C28H23ClNPPd: C, 61.56;
H, 4.24; N, 2.56. Found: C, 61.70; H, 4.12; N, 2.21.
Mononuclear PdII Complex 8. Method A, starting from 3 and
NaNO3: Complex 3 (31.7 mg, 0.052 mmol) and NaNO3 (4.4 mg,
0.052 mmol) were suspended in glacial acetic acid (5 mL) and heated
at 50 °C for 1 h. The resulting solution was filtered and concentrated
under vacuum. Layering a solution of the previous powder dissolved in
CH2Cl2 by pentane led to cocrystallization of 8 as orange crystals and
9 as green crystals. Method B, starting from Pd(OAc)2 complex and 2:
Pd(OAc)2 (53.9 mg, 0.24 mmol) and 2 (91 mg, 0.48 mmol) were
suspended in distilled CH2Cl2 (5 mL) under argon and mixed at room
temperature for 6 h. The resulting solution was layered by distilled
pentane (15 mL), and the flask was allowed to sit overnight at 4 °C.
The precipitate was filtered and dried under vacuum to afford complex
9 as a green-brown powder. Crystals suitable for X-ray diffraction were
directly obtained by layering a solution of 9 in CH2Cl2 by pentane.
Selected data from a single crystal (from X-ray): 1H NMR (400 MHz,
CDCl3) δ 8.96 (dd, J = 4.2, 1.7 Hz, 2H, H2), 8.18 (dd, J = 8.3, 1.7 Hz,
2H, H4), 7−85−7.75 (m, 4H, H5+H7), 7.55 (dd, J = 8.2, 7.1 Hz, 2H,
H6), 7.44 (dd, J = 8.3, 4.2 Hz, 2H, H3), 5.86 (s, 4H), 2,28 (s, 3H),
2.21 (s, 6H), 1.96 (s, 3H).
Mononuclear PdII Complex 6a [Pd(8-MQ)(NO3)(PPh3)]. A solution
of 5 (250 mg, 0.46 mmol) and silver nitrate (389 mg, 2.29 mmol) in
CH2Cl2 (20 mL) was stirred at room temperature for 24 h. The
reaction mixture was filtered and the filtrate concentrated in vacuo to
yield 6a as a pale yellow solid in quantitative yield (280 mg). Crystals
suitable for X-ray diffraction were grown by layering of a CH2Cl2
1
solution with cyclohexane: mp 180−183 °C; H NMR (400 MHz,
CD2Cl2) δ 8.75−8.71 (m, 1H, H2), 8.39 (dd, J = 8.3, 1.2 Hz, 1H),
7.77−7.64 (m, 3H), 7.55−7.36 (m, 16H), 3.04 (d, J = 4.2 Hz, 2H);
13C NMR (101 MHz, CD2Cl2) δ 151.1, 148.8, 147.7, 138.4, 134.2 (d, J
= 11.9), 133.9 (d, J = 16.0 Hz), 130.9 (d, J = 2.5 Hz), 129.2, 128.8,
128.7 (d, J = 11.0 Hz), 128.0, 124.1, 121.8, 26.8; 31P NMR (162 MHz,
CD2Cl2) δ 33.0; LIFDI-MS m/z 510.04 [M − NO3] (calcd for
C28H23NPPd 510.06); IR (KBr, cm−1) 3054, 1505, 1447, 1384, 1286,
1096, 1021, 825, 786, 751, 696.
Mononuclear PdII Complex 6b [Pd(8-MQ)(NO2)(PPh3)]. A solution
of 5 (250 mg, 0.46 mmol) and silver nitrite (352 mg, 2.29 mmol) in
CH2Cl2 (20 mL) was stirred at room temperature for 24 h. The
reaction mixture was filtered and the filtrate concentrated in vacuo to
yield 6b as a light brown solid in quantitative yield (315 mg). We
noted that complex 6b is sensitive to air and moisture (OPPh3 formed
over time in CDCl3 solutions of 6b): mp 178−180 °C; 1H NMR (400
MHz, CD2Cl2) δ 8.68 (s, 1H), 8.35 (d, J = 8.2 Hz, 1H), 7.77−7.68 (m,
5H), 7.51−7.37 (m, 12H), 7.36−7.30 (m, 1H), 2.72 (s, 2H); 31P
NMR (162 MHz, dry CD2Cl2) δ 32.4; LIFDI-MS m/z 510.03 [M −
NO2] (calcd for C28H23NPPd 510.06); IR (KBr, cm−1) 3055, 2870,
1684, 1569, 1503, 1481, 1431, 1340, 1095, 825, 752, 696. Complex 6b
retained 0.25 equiv of AgCl. Further purification led to its
decomposition (we were unable to obtain a satisfactory 13C NMR
data for 6b). Anal. Calcd for C28H23N2O2PPd·Ag0.25Cl0.25: C, 56.74; H,
3.91; N, 4.73. Found: C, 56.77; H, 3.95; N, 4.68 (average of two runs).
Mononuclear Pd(II) Complex 7a [Pd(8-MQ)(NO3)(NCCH3)].
Complex 4 (140 mg, 0.25 mmol) and silver nitrate (100 mg, 0.6
mmol) were suspended in a mixture of CH2Cl2 and CH3CN (8 mL,
3:1 v/v) and stirred at room temperature for 24 h. The reaction
mixture was filtered and the filtrate concentrated in vacuo to yield 7a
Dinuclear PdII−PdII Complex 9. Method A, starting from 3 and
NaNO3: Complex 3 (115 mg, 0.19 mmol) and NaNO3 (16 mg, 0.19
mmol) were suspended in glacial acetic acid (10 mL) and heated at 80
°C for 3 h. The resulting solution was filtered and concentrated under
vacuum to afford complex 8 as an orange powder in quantitative yield
(134 mg). Method B, starting from Pd(OAc)2 complex and 8-
quinolinecarboxylic acid 10: Pd(OAc)2 (104 mg, 0.46 mmol) and 8-
quinolinecarboxylic acid 10 (80 mg, 0.46 mmol) were suspended in
glacial acetic acid (10 mL) and heated at 110 °C for 1.5 h. The
resulting solution was filtered and poured into cold distilled water (200
mL). The flask was left overnight at 4 °C. The mixture was
concentrated under vacuum to induce precipitation of the product.
The precipitate was filtered and dried under vacuum to afford complex
8 as an orange powder with 68% yield (106 mg). Crystals suitable for
X-ray diffraction were directly obtained from the concentrated solution
of 8 in AcOD (NMR tube): 1H NMR (400 MHz, CDCl3) δ 8.76 (dd,
J = 5.5, 1.6 Hz, 2H, H2), 8.46 (dd, J = 8.2, 1.6 Hz, 2H, H4), 8.01 (dd, J
= 7.4, 1.6 Hz, 2H, H5), 7.86 (dd, J = 8.1, 1.7 Hz, 2H, H7), 7.58 (dd, J
= 8.2, 5.5 Hz, 2H, H3), 7.31 (t, J = 7.7 Hz, 2H, H6), 2.12 (s, 6H,
CH3); 13C NMR (101 MHz, CDCl3) δ 186.2 (CO, Bn), 172.3 (CO,
AcO), 166.8 (C10), 154.6 (C2), 141.8 (C4), 138.1 (C5), 132.8 (C7),
129.6 (C6), 127.4 (C3), 123.8 (C8), 122.0 (C9),24.0 (CH3); LIFDI-
MS m/z 450.93 (calcd for C20H12N2O4Pd1 ([M-Pd-2OAc]) 450.99);
ESI-MS m/z 698.9075 (calcd for C24H18N2O8Pd2 ([M + Na]+)
698.9079; IR (ATR, cm−1) 3399, 3064, 1722, 1563, 1416, 1333, 1304,
769, 697.
1
as a brown solid in quantitative yield (228 mg): mp 139−143 °C; H
NMR (400 MHz, CDCl3) δ 8.64 (dd, J = 5.2, 1.5 Hz, 1H, H2), 8.26
(dd, J = 8.4, 1.5 Hz, 1H, H4), 7.59 (dd, J = 7.8, 1.3 Hz, 1H, H5), 7.53
(dq, J = 7.1, 1.3 Hz, 1H, H7), 7.46 (dd, J = 7.1, 7.8, Hz, 1H, H6), 7.40
(dd, J = 5.2, 8.4 Hz, 1H, H3) 3.69 (s, 2H, CH2), 2.34 (s, 3H, CH3);
13C NMR (101 MHz, CDCl3) δ 152.5 (C2), 149.7 (C9), 146.5 (C8),
138.4 (C4), 129.0 (C6), 128.5 (C7), 128.3 (C5), 124.2 (C3), 121.7
L
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Catalytic Evaluation of the Complexes in the Acetoxylation
Reaction of 1. General procedure A, acetoxylation of 8-methylquino-
line 1 under air: To a microwave flask containing Pd precatalyst
(0.0125 mmol, 5 mol % with respect to substrate) and NaNO3 (0−1
equiv relative to substrate) was added 2 mL of AcOH/Ac2O (7:1). 8-
Methylquinoline 1 (34 μL, 0.25 mmol) was added via syringe. The
reaction mixture was heated at 110 °C, and the flask was flushed with
compressed air at hourly intervals. After being heated for 24 h, the
reaction mixture was cooled to room temperature and filtered through
a plug of Celite, and the solvent removed under reduced pressure.
Sodium bicarbonate (saturated solution in water) was added to
neutralize the residue, and the organic products were extracted into
ethyl acetate. The organic extracts were washed with brine, dried over
magnesium sulfate, and concentrated in vacuo. 1H NMR spectroscopic
(integration of the CH2 signal) and gas chromatographic analysis of
the crude product was used to calculate conversion (tR1 = 6.53 min, tR2
= 9.06 min), and the acetoxylated product 2 was then purified by flash
chromatography on silica gel (35% EtOAc in hexane). General
Procedure B, acetoxylation of 8-methylquinoline 1 under O2: To a
Schlenk tube fitted with a Young’s tap and magnetic stirrer bar were
added Pd precatalyst (0.0125 mmol, 5 mol % with respect to
substrate) and NaNO3 (0−1 equiv relative to substrate). The flask was
attached to a grease-free Schlenk line fitted with Young’s taps and was
then evacuated (aspirator vacuum) and backfilled with oxygen three
times to achieve 1 atm of O2. The flask was sealed and heated at 110
°C for 24 h. Reaction workup followed procedure A as described
above.
Compound 2 (73 mg, 0.36 mmol) was added. The reaction mixture
was stirred at 110 °C for 24 h, cooled to room temperature, and
filtered through a plug of Celite, and the solvent removed under
reduced pressure. CH2Cl2 was added, and the organic phase was
extracted three times with 2 M NaCl (in water). The organic phase
was dried over MgSO4 and filtered and the solvent evaporated under
vacuum (to give product 2). The aqueous phase was acidified with 2 M
HCl and extracted three times with ethyl acetate. The organic phase
was dried over MgSO4 and filtered and the solvent removed under
vacuum (product 10). Due to the difficulty associated with the analysis
and quantification of the presence of the overoxidized product (to give
acid compound 10), both by MS and GC, the proportion of the
acetoxylated and acid products are given in terms of yields following
purification (in Figure 13).
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b10853.
■
S
Additional experimental data, full product character-
ization details (including NMR spectra), X-ray diffraction
data, and DFT calculations are supplied (PDF)
X-ray data for complex 3 (CIF)
AUTHOR INFORMATION
Corresponding Author
*ian.fairlamb@york.ac.uk
■
8-Acetoxymethylquinoline 2, known compound: Rf 0.22 (EtOAc/
1
hexane, 1:3, v/v); H NMR (400 MHz, CDCl3) δ 8.97 (ddd, J = 4.2,
1.8, 0.9 Hz, 1H, H2), 8.19 (dt, J = 8.3, 1.7 Hz, 1H, H4), 7.85−7.74 (m,
2H, H5+H7), 7.55 (dd, J = 8.0, 7.2 Hz, 1H, H6), 7.45 (ddd, J = 8.3,
4.2, 0.9 Hz, 1H, H3), 5.86 (s, 2H, CH2), 2.16 (s, 3H, CH3); 13C NMR
(101 MHz, CDCl3) δ 171.6 (CO), 150.4 (C2), 146.6 (C9), 136.7
(C4), 134.1 (C8), 129.3 (C6), 128.8 (C7), 126.8 (C5), 124.0 (C10),
121.8 (C3), 63.2 (CH2), 21.6 (CH3); HRMS (ESI+) m/z 202.0872
[M + H]+ (calcd for C12H12NO2 202.0868); IR (ATR, cm−1) 3066,
2925, 1735, 1497, 1365, 1345, 1234, 1065, 1022, 827, 793, 764.
General Protocol for the Kinetics Study of the Complexes in
the Acetoxylation Reaction of 1. To a round-bottom flask
containing Pd(OAc)2, 3, 7a, 7b, 8, or 9 (5 mol %, 0.023 mmol) and
NaNO3 (1 equiv relative to substrate, 39 mg, 0.460 mmol) was added
AcOH/Ac2O (8 mL, 7:1 v/v, air-saturated and preheated at 80 °C). 8-
Methylquinoline (1 equiv, 61 μL, 0.460 mmol) was added via syringe.
The reaction mixture was heated at 80 °C. Aliquots (0.1 mL) were
taken from the reaction mixture every minute during 20 min. Aliquots
of the reaction were immediately quenched by addition of 0.5 mL of
NH4Cl saturated solution. The acetoxylation product was then
extracted with 0.5 mL of ethyl acetate, filtered through a Pasteur
pipet, and filled with cotton wool and Celite into a GC vial containing
1 mL of the mesitylene stock solution (12.4 mmol/L, used as a
standard reference for the GC measurements). The GC method
consists of three separate washes with three different solvents between
each injection. Each sample was injected three times, giving three
independent values for each time interval. Individual points, within
each kinetic curve, represent the average value of these data (errors
bars are shown in Figure 10).
ORCID
Joshua T. W. Bray: 0000-0003-2384-9675
Ian J. S. Fairlamb: 0000-0002-7555-2761
Present Addresses
†School of Chemistry, Cardiff University, Main Building, Park
Place, Cardiff CF10 3AT, U.K.
‡Thomas Graham Building, University of Strathclyde, 295
Cathedral Street, Glasgow G1 1XL, U.K.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The research leading to these results has received funding from
the Innovative Medicines Initiative Joint Undertaking under
Grant Agreement No. 115360, resources of which are
composed of financial contributions from the European
Union’s Seventh Framework Programme (FP7/2007-2013)
and EFPIA companies’ in kind contribution (to M.N.W. and
I.J.S.F.). We are grateful to the University of York for funding.
EPSRC funded the computational equipment used in this study
(ref EP/H011455/and EPSRC ‘ENERGY’ grant, ref no. EP/
K031589/1). We appreciate the constructive critical comments
provided by the reviewers of our paper.
Protocol for the Head Space Analysis of the Acetoxylation
Reaction Catalyzed by Pd(OAc)2. In a 500 mL round-bottom flask
equipped with a condenser were added Pd(OAc)2 (25.8 mg, 0.115
mmol, 0.05 equiv), NaNO3 (195.5 mg, 2.30 mmol, 1 equiv), AcOH/
Ac2O (40 mL, 7:1, v/v), and 8-methylquinoline (305 μL, 2.30 mmol, 1
equiv). Before the mixture was heated at 110 °C, the top of the
condenser was equipped with a glass tubing-based piece allowing the
entrance of a flux of “zero-air” and the exit to the NOx detectors. The
volume dilution was 15 L per minute, and concentrations are reported
in mixing ratios, in this case, in parts per trillion.
REFERENCES
■
(1) (a) C−H and C−X Bond Functionalization. Transition Metal
Mediation; Ribas, X., Ed.; RSC Catalysis, Series No. 11; Royal Society
of Chemistry, 2013. Recent papers in the area of C−H bond
functionalization involving Pd catalysts:. (b) Fan, J.-H.; Wei, W.-T.;
Zhou, M.-B.; Song, R.-J.; Li, J.-H. Angew. Chem., Int. Ed. 2014, 53,
6650. (c) Wang, L.; Wang, Y.; Liu, C.; Lei, A. Angew. Chem., Int. Ed.
2014, 53, 5657. (d) Yan, J.-X.; Li, H.; Liu, X.-W.; Shi, J.-L.; Wang, X.;
Shi, Z.-J. Angew. Chem., Int. Ed. 2014, 53, 4945. (e) Park, T.-H.;
Hickman, A. J.; Koh, K.; Martin, S.; Wong-Foy, A. G.; Sanford, M. S.;
Matzger, A. J. J. Am. Chem. Soc. 2011, 133, 20138. (f) Zhu, R.;
Buchwald, S. L. Angew. Chem., Int. Ed. 2012, 51, 1926. (g) Zhang, H.;
Shi, R.; Gan, P.; Liu, C.; Ding, A.; Wang, Q.; Lei, A. Angew. Chem., Int.
General Procedure for the Formation of the Overoxidized
Product 10 from Acetoxylated Product 2. To a microwave flask
containing Pd(OAc)2 (0.0181 mmol, 5 mol % relative to 2) and
NaNO3 (1 equiv relative to 2) was added 8 mL of AcOH/Ac2O (7:1).
M
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Ed. 2012, 51, 5204. (h) Shan, G.; Yang, X.; Ma, L.; Rao, Y. Angew.
Chem., Int. Ed. 2012, 51, 13070. (i) Saget, T.; Cramer, N. Angew.
Chem., Int. Ed. 2012, 51, 12842. (j) Wang, H.; Li, G.; Engle, K. M.; Yu,
J.-Q.; Davies, H. M. L. J. Am. Chem. Soc. 2013, 135, 6774. (k) Sharma,
A.; Hartwig, J. F. J. Am. Chem. Soc. 2013, 135, 17983. (l) Deng, R.;
Huang, Y.; Ma, X.; Li, G.; Zhu, R.; Wang, B.; Kang, Y.-B.; Gu, Z. J. Am.
Chem. Soc. 2014, 136, 4472. For recent mechanistic work, see: (m) Ji,
Y.; Plata, E. R.; Regens, C. S.; Hay, M.; Schmidt, M.; Razler, T.; Qiu,
Y.; Geng, P.; Hsiao, Y.; Rosner, T.; Eastgate, M. D.; Blackmond, D. G.
J. Am. Chem. Soc. 2015, 137, 13272. (n) Smalley, A. P.; Gaunt, M. J. J.
Am. Chem. Soc. 2015, 137, 10632. For recent papers involving other
NO3 → Pd−O + NO2, if the process is reversible. Metal−oxo
complexes formed from nitrates are known; see: (a) Jiang, J.; Holm, R.
H. Inorg. Chem. 2005, 44, 1068. Pd-oxidase catalysis has been
reviewed; see: (b) Stahl, S. S. Angew. Chem., Int. Ed. 2004, 43, 3400.
(26) For the lead references, see: (a) Powers, D. C.; Geibel, M. L.;
Klein, J. E. M. N.; Ritter, T. J. Am. Chem. Soc. 2009, 131, 17050.
(b) Racowski, J. M.; Dick, A. R.; Sanford, M. S. J. Am. Chem. Soc. 2009,
131, 10974.
(27) Davies reported that additional N∧CH ligand is required for
effective catalytic turnover in cyclometallated systems containing Ir
and Rh; see: Boutadla, Y.; Davies, D. L.; Jones, R. C.; Singh, K. Chem. -
Eur. J. 2011, 17, 3438.
(28) (a) Mason. Chem. Rev. 1981, 81, 205. (b) Mason, J.;
Larkworthy, L. F.; Moore, E. A. Chem. Rev. 2002, 102, 913.
(29) Witanowski, M.; Stefaniak, L.; Webb, G. A. Nitrogen NMR
Spectroscopy. In Annual Reports on NMR Spectroscopy; Webb, G. A.,
Ed.; Elsevier, 1981; p 494.
́
metals, see: (o) Whitaker, D.; Bures, J.; Larrosa, I. J. Am. Chem. Soc.
2016, 138, 8384. (p) Cambeiro, X. C.; Ahlsten, N.; Larrosa, I. J. Am.
Chem. Soc. 2015, 137, 15636.
(2) Zultanski, S. L.; Stahl, S. S. J. Organomet. Chem. 2015, 793, 263.
(3) Stowers, K. J.; Kubota, A.; Sanford, M. S. Chem. Sci. 2012, 3,
3192.
(4) (a) Wang, D.; Zavalij, P. Y.; Vedernikov, A. N. Organometallics
2013, 32, 4882. (b) Zhang, J.; Khaskin, E.; Anderson, N. P.; Zavalij, P.
Y.; Vedernikov, A. N. Chem. Commun. 2008, 3625.
(5) Wickens, Z. K.; Morandi, B.; Grubbs, R. H. Angew. Chem., Int. Ed.
2013, 52, 11257.
(30) White, P. B.; Jaworski, J. N.; Fry, C. G.; Dolinar, B. S.; Guzei, I.
A.; Stahl, S. S. J. Am. Chem. Soc. 2016, 138, 4869.
(31) ESI-MS has been widely used to study catalytic processes
involving organometallic species. For example, to examine inter-
mediates from the Heck reaction: (a) Sabino, A. A.; Machado, A. H.
L.; Correia, C. R. D.; Eberlin, M. N. Angew. Chem., Int. Ed. 2004, 43,
2514. In allylic alkylation, see: (b) Markert, C.; Pfaltz, A. Angew.
Chem., Int. Ed. 2004, 43, 2498.
(6) Li, J.; Grubbs, R. H.; Stoltz, B. M. Org. Lett. 2016, 18, 5449.
(7) Lou, S.-J.; Xu, D.-Q.; Xu, Z.-Y. Angew. Chem., Int. Ed. 2014, 53,
10330.
(32) LIFDI-MS has been used to study transition-metal fluoride
complexes: (a) Dransfield, T. A.; Nazir, R.; Perutz, R. N.; Whitwood,
A. C. J. Fluorine Chem. 2010, 131, 1213. (b) Nova, A.; Erhardt, S.;
Jasim, N. A.; Perutz, R. N.; Macgregor, S. A.; McGrady, J. E.;
Whitwood, A. C. J. Am. Chem. Soc. 2008, 130, 15499. Mass spectra
produced by LIFDI-MS show little or no fragmentation. Recorded
ions are normally dominated by molecular radical cations M+•; see:
Smith, F. D.; Schaub, M. T.; Rodgers, R. P.; Hendrickson, C. L.;
Marshall, A. G. Anal. Chem. 2008, 80, 7379.
(33) (a) Baumann, C. G.; De Ornellas, S.; Reeds, J. P.; Storr, T. E.;
Williams, T. J.; Fairlamb, I. J. S. Tetrahedron 2014, 70, 6174−6187.
(b) Tang, D.-T. D.; Collins, K.; Glorius, F. J. Am. Chem. Soc. 2013,
135, 7450−7452.
(34) Infrared and Raman Characteristic Group Frequencies, Tables and
Charts, 3rd ed.; Socrates, G., Ed.; John Wiley and Sons Ltd., 2004.
(35) Jones, C. J.; McCleverty, J. A.; Rothin, A. S.; Adams, H.; Bailey,
N. A. J. Chem. Soc., Dalton Trans. 1986, 2055.
(36) Park, W.; Na, Y.; Baek, D.-J. Bull. Korean Chem. Soc. 2006, 27,
2023.
(8) (a) Kuznetsova, N. I.; Likholobov, V. A.; Fedotov, M. A.;
Yermakov, Y. J. Chem. Soc., Chem. Commun. 1982, 973. For previous
work, see: (b) Tamura, M.; Yasui, T. Chem. Commun. 1968, 1209.
(9) Henry, P. M. J. Org. Chem. 1971, 36, 1886.
(10) Andrews, M. A.; Kelly, K. P. J. Am. Chem. Soc. 1981, 103, 2894.
(11) Andrews, M. A.; Cheng, C. W. F. J. Am. Chem. Soc. 1982, 104,
4268.
(12) Heumann, A.; Chauvet, F.; Waegell, B. Tetrahedron Lett. 1982,
23, 2767.
(13) Andrews, M. A.; Chang, T. C. T.; Cheng, C. W. F.
Organometallics 1985, 4, 268.
(14) Mares, F.; Diamond, S. E.; Regina, F. J.; Solar, J. P. J. Am. Chem.
Soc. 1985, 107, 3545.
(15) Backvall, J.-E.; Heumann, A. J. Am. Chem. Soc. 1986, 108, 7107.
̈
(16) (a) Bajwa, S. E.; Storr, T. E.; Hatcher, L. E.; Williams, T. J.;
Baumann, C. G.; Whitwood, A. C.; Allan, D. R.; Teat, S. J.; Raithby, P.
R.; Fairlamb, I. J. S. Chem. Sci. 2012, 3, 1656. For a critical review, see:
(b) Fairlamb, I. J. S. Angew. Chem., Int. Ed. 2015, 54, 10415.
(17) Warren, M. R.; Brayshaw, S. K.; Hatcher, L. E.; Johnson, A. L.;
Schiffers, S.; Warren, A. J.; Teat, S. J.; Warren, J. E.; Woodall, C. H.;
Raithby, P. R. Dalton Trans. 2012, 41, 13173.
(37) Wah, H. L. K.; Postel, M.; Tomi, F. Inorg. Chem. 1989, 28, 233.
(38) Aellig, C.; Girard, C.; Hermans, I. Angew. Chem., Int. Ed. 2011,
50, 12355.
(18) For a useful review on the structure of Pd3(OAc)6 and its
speciation in the presence of water/alcoholic solvents, see: (a) Bedford,
R. B.; Bowen, J. G.; Davidson, R. B.; Haddow, M. F.; Seymour-Julen,
A. E.; Sparkes, H. A.; Webster, R. L. Angew. Chem., Int. Ed. 2015, 54,
6591. Also, see two excellent papers about the purity of Pd(OAc)2:
(b) Carole, W. A.; Colacot, T. J. Chem. - Eur. J. 2016, 22, 7686. Carole,
W. A.; Bradley, J.; Sarwar, M.; Colacot, T. J. Org. Lett. 2015, 17, 5472.
(19) Ryabov, A. D.; Eliseev, A. V.; Sergeyenko, E. S.; Usatov, A. V.;
Zakharkin, L. I.; Kalinin, V. N. Polyhedron 1989, 8, 1485.
(20) (a) Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. J. Am.
Chem. Soc. 2005, 127, 13754. (b) Lapointe, D.; Fagnou, K. Chem. Lett.
2010, 39, 1118.
́
(39) Campora, J.; Palma, P.; del Río, D.; Carmona, E.; Graiff, C.;
Tiripicchio, A. Organometallics 2003, 22, 3345.
(40) Haines, B. E.; Berry, J. F.; Yu, J.-Q.; Musaev, D. G. ACS Catal.
2016, 6, 829.
(41) (a) Zhang, W.; Ren, S.; Zhang, J.; Liu, Y. J. Org. Chem. 2015, 80,
5973. Valuable mechanistic and synthetic studies have also been
reported recently: (b) Liang, Y.-F.; Li, X.; Wang, X.; Yan, Y.; Feng, P.;
Jiao, N. ACS Catal. 2015, 5, 1956.
(42) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923.
(43) (a) Csaszar, P.; Pulay, P. J. Mol. Struct. 1984, 114, 31.
(b) Ahlrichs, R.; Bar, M.; Haser, M.; Horn, H.; Kolmel, C. Chem. Phys.
Lett. 1989, 162, 165. (c) Eichkorn, K.; Treutler, O.; Ohm, H.; Haser,
M.; Ahlrichs, R. Chem. Phys. Lett. 1995, 240, 283. (d) Treutler, O.;
Ahlrichs, R. J. J. Chem. Phys. 1995, 102, 346. (e) Eichkorn, K.;
Weigend, F.; Treutler, O.; Ahlrichs, R. Theor. Chem. Acc. 1997, 97,
119. (f) von Arnim, M.; Ahlrichs, R. J. J. Chem. Phys. 1999, 111, 9183.
(g) Deglmann, P.; Furche, F. J. J. Chem. Phys. 2002, 117, 9535.
(h) Deglmann, P.; Furche, F.; Ahlrichs, R. Chem. Phys. Lett. 2002, 362,
511. (i) Deglmann, P.; May, K.; Furche, F.; Ahlrichs, R. Chem. Phys.
Lett. 2004, 384, 103.
(21) Yu, S. − Y.; Fujita, M.; Yamaguchi, K. J. Chem. Soc., Dalton
Trans. 2001, 3415.
(22) (a) Nonoyama, M. Synth. React. Inorg. Met.-Org. Chem. 1999, 29,
119. (b) Nonoyama, M.; Nakajima, K. Polyhedron 1998, 18, 533.
(23) Phosphine ligands typically slow down processes involving PdIV
species; see: Sehnal, P.; Taylor, R. J. K.; Fairlamb, I. J. S. Chem. Rev.
2010, 110, 824.
(24) For a useful review, see: Nair, V.; Panicker, S. B.; Nair, L. G.;
George, T. G.; Augustine, A. Synlett 2003, 156.
(25) It is feasible that the “washing out” of the 18O label in the
presence of pure O2 might be promoted by Pd−oxo species, i.e., Pd−
N
DOI: 10.1021/jacs.6b10853
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX