Predicting Electrocatalytic Properties: Modeling Structure-Activity Relationships of Nitroxyl Radicals

Article abstract of DOI:10.1021/jacs.5b11252

Stable nitroxyl radical-containing compounds, such as 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO) and its derivatives, are capable of electrocatalytically oxidizing a wide range of alcohols under mild and environmentally friendly conditions. Herein, we examine the structure-function relationships that determine the catalytic activity of a diverse range of water-soluble nitroxyl radical compounds. A strong correlation is described between the difference in the electrochemical oxidation potentials of a compound and its electrocatalytic activity. Additionally, we construct a simple computational model that is able to accurately predict the electrochemical potential and catalytic activity of a wide range of nitroxyl radical derivatives.

Full text of DOI:10.1021/jacs.5b11252

Article
pubs.acs.org/JACS
Predicting Electrocatalytic Properties: Modeling Structure−Activity
Relationships of Nitroxyl Radicals
David P. Hickey,†,∥ David A. Schiedler,†,∥ Ivana Matanovic,‡,§ Phuong Vy Doan,† Plamen Atanassov,‡
Shelley D. Minteer,*,† and Matthew S. Sigman*,†
†Department of Chemistry, University of Utah, Salt Lake City, Utah 84112, United States
‡Department of Chemical & Biological Engineering, University of New Mexico, Albuquerque, New Mexico 87131, United States
§Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States
S
* Supporting Information
ABSTRACT: Stable nitroxyl radical-containing compounds,
such as 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO) and its
derivatives, are capable of electrocatalytically oxidizing a wide
range of alcohols under mild and environmentally friendly
conditions. Herein, we examine the structure−function
relationships that determine the catalytic activity of a diverse
range of water-soluble nitroxyl radical compounds. A strong
correlation is described between the difference in the
electrochemical oxidation potentials of a compound and its
electrocatalytic activity. Additionally, we construct a simple
computational model that is able to accurately predict the
electrochemical potential and catalytic activity of a wide range
of nitroxyl radical derivatives.
reduced TH species.29,30 The active form of the catalyst is
INTRODUCTION
■
then electrochemically regenerated by an overall one proton
Since their discovery in 1960, stable nitroxyl radicals have
attracted tremendous attention due to their unique para-
coupled-two-electron oxidation of TH.31 In aqueous con-
ditions, the second oxidative step is presumed to be turnover
magnetic and redox characteristics.1,2 The applications of these
limiting.30 However, the exact nature of the oxidative
nitroxyl radical-containing compounds are diverse including
their use as mechanistic tools for the detection of free
regeneration of T+ is not well understood.32
radicals,3−6 contrast agents for magnetic resonance imaging
A wide range of hypotheses have been proposed to correlate
(MRI) in biological settings,7−10 analytical tools for the
the structure of TEMPO derivatives to their catalytic activity
detection of reactive oxygen species,11−13 and, perhaps, most
with >100 structural analogues synthesized and studied to
date.17,33−35 Many of the early reports suggest explanations for
extensively as oxidative dehydrogenation catalysts.3,14−23 In
structure−activity relationships, which involve a predominantly
particular, readily available (2,2,6,6-tetramethylpiperidin-1-yl)-
steric argument with respect to the geminal dimethyl groups.36
oxyl (TEMPO) has been exploited for its ability to electro-
However, more recent efforts show a correlation of the T*/T+
oxidation potential with catalytic activity.37,38 These proposed
explanations are able to accurately account for defined subsets
of TEMPO derivatives but fail to describe trends for broader
classes of nitroxyl radical catalysts due in part to the
examination of relatively small data sets.
catalytically oxidize alcohols24−27 and amines28 under mild and
environmentally benign reaction conditions. Despite the vast
wealth of literature on TEMPO-mediated oxidations, there
remain central mechanistic aspects and structure−activity
relationships about this class of compounds that are not well
understood.
A goal of our program is to utilize nitroxyl radical catalysts
for a diverse range of electrocatalytic applications including fuel
cells and batteries; therefore, the ability to predict the
performance (oxidation potential and catalytic activity) of any
nitroxyl radical catalyst would streamline our efforts. Addition-
ally, understanding the mechanistic foundations associated with
the structure function relationships would provide the
Three electrochemically accessible oxidation states exist for
TEMPO and its derivatives including a hydroxylamine (TH), a
nitroxyl radical (T*), or an oxoammonium cation (T+). During
catalytic oxidation of an alcohol (as shown in Figure 1A),
TEMPO is electrochemically oxidized from T* to the
corresponding T+, which operates as the catalytically active
species. An alcohol-containing substrate undergoes nucleophilic
attack on the nitrogen of T+ to form a TEMPO-substrate
intermediate complex. A subsequent two-electron oxidation of
the substrate yields the corresponding aldehyde and the
Received: October 27, 2015
Published: December 3, 2015
© 2015 American Chemical Society
16179
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186

Journal of the American Chemical Society
Article
Figure 1. Presumed catalytic cycle for alcohol oxidation by TEMPO (A) that highlights the two external oxidation steps, Ea1 and Ea2; a representative
cyclic voltammogram of a nitroxyl radical (B) demonstrating how Ea1 and Ea2 are observed experimentally, and; the pH profile of catalytic activity
(jmax) toward glycerol and oxidation potentials Ea1 and Ea2 for TEMPO (C). The pH profile of TEMPO indicates a correlation between the
difference in Ea1 and Ea2 with the catalytic activity.
community a clearer picture of how to select an appropriate
catalyst. Herein, we describe the synthesis, electrochemical
characterization, and computational modeling of a range of
nitroxyl radicals to reveal the salient features responsible for
performance. These studies provide the basis to correlate and
predict both the electrochemical potential and catalytic activity
for new nitroxyl radical-based electrocatalysts under aqueous
conditions.
of T* to T+, while the second occurs at an oxidation potential
defined as Ea2 corresponding to the proton-coupled oxidation
of TH to T*. A representative cyclic voltammogram (CV)
shown in Figure 1B highlights the experimental determination
of Ea1 and Ea2 for a nitroxyl radical species. The observed values
for Ea1 follow expected Nernst kinetics; however, values for Ea2
were highly dependent on scan rate and frequently overlapped
with the peaks for Ea1. Therefore, Ea2 was determined
experimentally using variable scan rates and extrapolating to a
standard scan rate (a more complete discussion on the
experimental determination of electrochemical parameters can
be found in the Supporting Information Section S2.1).
RESULTS AND DISCUSSION
■
pH-Dependent Electrochemical Properties of TEMPO.
We previously reported on the anomalous catalytic activity and
expanded pH range of 4-amino-TEMPO.39 It was observed that
the reversible oxidation potential of amino-TEMPO is strongly
pH dependent whereas the oxidation potential for unmodified
TEMPO is independent of pH. This data compelled us to
suggest that the uncharacteristic activity of amino-TEMPO
must be related to its unique pH dependence, and consequently
that TEMPO activity may be described as a function of redox
potentials. To probe this hypothesis, we first examined the
electrochemical characteristics and catalytic activity of TEMPO
derivatives as a function of pH by using glycerol oxidation as a
model reaction and unmodified TEMPO as an initial
electrocatalyst.
The examination of both Ea1 and Ea2 of unmodified TEMPO
as a function pH (shown in Figure 1C) indicates that, as
mentioned above, Ea1 is independent of proton concentration
but that Ea2 is inversely proportional to pH and thus
dramatically decreased under basic conditions. This is due to
the equilibrium that exists between the hydroxylamine form of
TEMPO (TH) and its conjugate acid hydroxylammonium
40
+
+
(TH2 ). Under acidic conditions, the TH2 form is favored
and corresponds to an increased oxidation potential. In this
manner, the pH profile of Ea2 serves as a titration curve for the
+
formation of TH2 and suggests a pKa ≈ 5.5. This value is in
good agreement with the reported literature pKa for similar
hydroxylamines.41−43 Due to the uneven dependence of Ea1
and Ea2 on proton concentration, the equilibrium between TH
and TH2+ can be described in terms of the difference of the two
As described above, the general mechanism for the
electrocatalytic oxidation of alcohols by TEMPO is thought
to proceed through two electrooxidative steps; the first occurs
at an oxidation potential of Ea1 corresponding to the oxidation
16180
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186

Journal of the American Chemical Society
Article
Figure 2. Summary of the nitroxyl radical catalysts screened for this study. A representative training set (indicated by the dashed enclosure, top left)
was used to determine trends for both experimental and computational properties. All values for jmax were determined voltammetrically using a 200
mM Robinson buffer at pH 7.0 and 25 °C.
potentials. Therefore, as Ea2 − Ea1 decreases and eventually
becomes negative, the equilibrium shifts to favor TH.
these analogues have been previously prepared, the compre-
hensive electrochemistry in water has not been determined. A
summary of the compounds with their reversible oxidation
potential and absolute catalytic activity in the form of jmax are
depicted in Figure 2. A significant range of electrochemical
potential and catalytic activity is afforded with this group of
electrocatalysts. The results qualitatively agree with the
In addition to Ea2, the electrocatalytic activity of TEMPO
toward glycerol oxidation is also dependent on pH (as shown
in Figure 1C), where the maximum catalytic current density
(jmax) increases exponentially at higher pH. A similar trend can
be drawn between Ea2 − Ea1 and jmax in which the catalytic
activity of TEMPO increases as Ea2 decreases with respect to
Ea1. The pH-dependence of TEMPO activity has been reported
+
proposed hypothesis that destabilization of TH2 should result
in higher activity. For example, amino-TEMPO (7), the initial
inspiration for this study, possesses an ammonium cation under
neutral aqueous conditions. This presumably disfavors the
further protonation of its corresponding TH form. Con-
sequently, 7 exhibits much higher activity than 5, which
possesses a carboxylate anion under the same conditions and
previously; however, the simultaneous correlation between jmax
,
pH, and Ea2 − Ea1 indicates that the equilibrium between TH
+
and TH2 is directly linked to the electrocatalytic activity of
TEMPO. In other words, this correlation suggests that the
catalytic activity of TEMPO is retarded by the formation of
+
+
TH2 and the extent of this deactivation can be observed
can act to stabilize the resultant TH2 species. Consistent with
through the measurement of Ea1 and Ea2.
these observations, the incorporation of a permanent
ammonium cation in 29 results in the largest jmax for all
monocyclic TEMPO derivatives.
Additionally, qualitative analysis of several commonly used
polycyclic nitroxyl electrocatalysts such as derivatives of 9-
azabicyclo[3.3.1]nonane N-oxyl (ABNO, 12), 2-azaadamantane
N-oxyl (AZADO, 13), and 2-azanoradamantane N-oxyl
(norAZADO, 14) were studied with and without geminal
Expansion of Potential−Activity Model. To test the
validity of the hypothesis that jmax and Ea2 − Ea1 are correlated,
we synthesized (for synthetic procedures and characterization
see the SI) and electrochemically evaluated a range of nitroxyl
radical-containing compounds. The set included various
functionalized TEMPO derivatives with different substitution
patterns, bicyclic cores, and functional groups. While many of
16181
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186

Journal of the American Chemical Society
Article
methyl groups relative to the nitroxyl nitrogen atom. A lower
jmax for all geminal methyl-containing compounds is observed
relative to the corresponding unmethylated derivative (compare
13 to 19, 14 to 20, 15 to 21, and 16 to 22). This trend can be
explained by steric effects as others have alluded;36 however, an
alternate explanation can be offered through stabilization of
shown in Figure 3B, demonstrates an excellent linear
correlation between (ipa/ipc)cat and (ΔEa+0.18)−1 with R2 =
0.92.
The plots in Figure 3 indicate that there is a lower limit to
the extent to which Ea2 can be lowered with respect to Ea1. The
precise source of this limit is not completely understood;
however, the value of this limit (−0.180 eV = −4.15 kcal
mol−1) closely corresponds with the energy associated with
hydrogen bond interactions between an aliphatic alcohol and a
single water molecule.44 This suggests that the source of the
non-zero limit is a lowering of Ea2 caused by hydrogen bonding
assistance with the proton-coupled oxidation. Therefore, the
catalytic activity is maximized when the thermodynamic
potential for Ea1 is equal to that of Ea2. This assertion is
speculative and further studies are needed to conclusively
determine the source of the observed limit. Nevertheless, the
experimental correlation between activity and ΔEa provides a
means to predict catalytic activity as a function of two
computationally accessible parameters, Ea1 and Ea2.
Computational Prediction of Electrochemical Poten-
tial and Activity. Using the correlation described above, we
developed a computational model to predict both the
electrochemical potentials Ea1 and Ea2 and subsequently the
catalytic activity for nitroxyl radical catalysts. A training set of
TEMPO derivatives (highlighted in Figure 2) was used to
derive trends in the computational modeling, and the remaining
compounds were used to validate or refute the preliminary
models. The training set was selected to represent the structural
and electronic diversity of the nitroxyl radical species studied as
well as a range of electrocatalytic activities observed.
+
TH2 . The enhanced basicity of the methyl-containing
+
structures stabilize TH2 and accordingly result in a lower
catalytic activity for these analogues.
It should be noted that some of the compounds in this study
exhibited either autocatalytic behavior or comproportionation,
which led to an artificially inflated jmax (the absolute jmax is
reported in Figure 2). In order to test the quantitative
correlation between catalytic activity and destabilization of
+
TH2 (via Ea2 − Ea1), we used the ratio of the peak oxidative
current, ipa, and the peak reductive current, ipc, to describe the
normalized catalytic activity as (ipa/ipc)cat. A plot of (ipa/ipc)cat
versus ΔEa (where ΔEa = Ea2 − Ea1) for all nitroxyl radical
species at pH 7 is shown in Figure 3A. An asymptotic
relationship between electrocatalytic activity and ΔEa for
nitroxyl radical catalysts is observed where the catalytic activity
approaches infinity as ΔEa approaches −0.180 V. This
correlation can be subsequently linearized by taking the inverse
of the sum of ΔEa plus the asymptotic limit. The resulting plot,
Several reports have previously conveyed the computational
prediction of Ea1 from the solvated free energy cycle;45,46
however, only a limited number of studies have been performed
to demonstrate the prediction of Ea2.47 In recent work by
Hodgson et al., the irreversible one-electron reduction
potentials (T* → T−) for several TEMPO derivatives were
computationally predicted, but their report did not account for
+
multiple protonation states of the reduced species (i.e., TH2 ,
TH, T−).47 Additionally, the experimentally determined
Nernstian kinetics derived previously for the oxidation
potentials associated with TEMPO derivatives do not take
into consideration intramolecular protonation states (such as
those associated with TEMPO-carboxylic acid, 5, or amino-
TEMPO, 7). Therefore, we derived modified Nernstian kinetic
equations to account for the possible protonation states, which
were used for all computational modeling described herein. A
more detailed discussion of these kinetic equations along with a
complete derivation can be found in the Supporting
Information.
Electrochemical potentials were determined using DFT
calculations to compute the thermally corrected free energies
for optimized structures of derivatives in each of their
oxidation/protonation states (TH2 , TH, T*, T+). Free
+
energies were calculated using Gaussian 09W with B3LYP/6-
31+G(d,p) level of theory and basis set and CPCM continuum
solvation model.48,49 For Ea1, the free energy difference
between T+ and T* (ΔGa1, in kJ mol−1) was converted into
an electrochemical potential and a correction factor was added
to normalize the calculated potential to a saturated calomel
reference electrode by
Figure 3. Plot indicating an asymptotic relationship between ΔEa and
(ipa/ipc)cat for the nitroxyl radicals studied (A) and linearized plot (B)
○
▲
highlighting the 4-substituted ( ) and 4N-substituted ( ) TEMPO
■
derivatives, polycyclic species ( ). The values shown in these plots
ΔG
a1
were determined voltammetrically in 200 mM Robinson buffer at pH
7.0 and at 25 °C.
Ea1 = −
− 4.68
nF
16182
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186

Journal of the American Chemical Society
Article
Where n is the number of electrons transferred (for all
oxidations described herein, n = 1), F is Faraday’s constant (F =
96.485 kJ mol−1) and 4.68 corresponds to the combined
normalization factors to report Ea1 relative to SCE.50 A plot of
predicted versus experimental values for Ea1 (Figure 4A)
activity for the validation set. The resulting plot of the predicted
versus experimental values of (ipa/ipc)cat (Figure 4B) demon-
strates that both the training set and validation set provide a
good linear correlation between predicted and observed activity
(R2 = 0.89 and R2 = 0.80 respectively). A single deviation from
the expected activity (shown as a red point in Figure 4B) is the
result of a lack convergence for the TH2+ species of compound
29. This is likely due to both the size of the catalyst and the
presence of a permanent cation. Nevertheless, the model
demonstrated here provides a convenient means of predicting
both electrochemical potential and catalytic activity of nitroxyl
radical oxidation catalysts under aqueous conditions.
Structure−Function Relationships from Computa-
tional Analysis. In addition to a quantitative model for
predicting redox potential and catalytic activity of nitroxyl
radicals demonstrated in the previous section, we sought to use
DFT calculations to determine simple structure−function
relationships that would provide a chemically intuitive model
to aid in the rational design of novel TEMPO derivatives. We
focused on calculated infrared (IR) vibrational stretching
frequencies for the NO bond in each oxoammonium T+
species as these parameters have been an excellent platform for
correlating structure to function in complex catalysis.51,52
The correlation between the calculated frequency of the N
O stretch (υNO) and the experimentally observed Ea1 (Figure
5A) demonstrates that structures with a low Ea1 also exhibit a
relatively low υNO. This trend is corollary rather than
causational. A lower Ea1 oxidation potential is caused by
stabilization of the oxoammonium cation of the T+ species.
This stabilization is a result of either geometric constraints or
the presence of functional groups that can donate electron
density to the empty p-orbital of the oxoammonium cation,
which in turn gives the NO bond more sp3 character.
Ultimately, this structural feature lowers υNO. Oxoammonium
cation stabilization is exemplified in pairs of polycyclic
structures that differ only by the presence of geminal methyl
groups relative to the nitrogen atom (such as 13 vs 19, 14 vs
20, or 16 vs 22, Figure 5C). In all cases, the structures with the
geminal methyl group(s) exhibits a lower Ea1 due to the
stabilization of the resulting oxoammonium cation through
hyperconjugation with the electron-donating methyl groups. In
contrast to oxoammonium stabilization to lower Ea1, the
opposite effect is observed in the presence of functional groups
that destabilize the oxoammonium cation. Compounds that
contain either a temporary or permanent cation (such as 7 or
28, Figure 2) effectively increase the oxidation potential of the
nitroxyl radical through coloumbic repulsion of cationic charges
(as illustrated in Figure 5C). Similar effects have been observed
for the oxidation potential of amine-containing ferrocene
compounds and are highly dependent on the distance between
the cationic moiety and the site of oxidation.53
Figure 4. Plots of the computationally derived values of Ea1 (A) and
(ipa/ipc)cat (B) versus their corresponding experimental values.
Calculated r2 values for the training set (blue) and validation set
(black) are provided for each graph. Computations were performed
using B3LYP/6-31+G(d,p) level of theory and basis set with a CPCM
solvation model and assumed conditions of pH 7.0 and 25 °C as
described in the Supporting Information.
demonstrates a good linear correlation between expected and
observed values for the reversible oxidation potential between
T+ and T* of the training set and validation set (R2 = 0.91 and
R2 = 0.80 respectively). Additionally, the predicted trend
provides an accurate prediction of the validation set. As
mentioned above, values of Ea2 were calculated similarly from
the free energy cycle associated with the oxidation of TH to T*.
However, the calculation of Ea2 from the free energy of the
oxidation step required that various protonation steps be
accounted for and that the Nernst equation be modified
accordingly. Complete analysis of these computations and
derivations of the modified reaction kinetics are provided in the
Supporting Information.
Many of the same factors that affect the Ea1 oxidation
potential also affect the Ea2 oxidation potential. As described
earlier, Ea2 is highly dependent on the protonation state of the
TH species; so that Ea2 increases under acidic conditions as the
+
protonation equilibrium favors the TH2 species and decreases
under basic conditions where TH is favored. Therefore, many
of the features that decrease Ea1 through stabilization of the
oxoammonium cation of T+, will effectively increase Ea2
+
through stabilization of the hydroxylammonium cation in TH2 .
The computationally predicted values of Ea1 and Ea2 were
−1
used to draw a correlation between (ΔEa,computational
)
and the
A correlation exists between Ea2 and the vibrational
stretching frequency of the N−O* (υN−O*) bond of the T*
state (as shown in Figure 5B), which is similar to that between
catalytic activity of the compounds in the training set. This
trend was subsequently used as a model to predict the catalytic
16183
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186

Journal of the American Chemical Society
Article
Figure 5. Correlation plots of computationally derived oxoammonium NO (A) and nitroxyl radical N−O* (B) IR stretching frequencies versus
Ea1 and Ea2 respectfully, and a chart indicating the effects of hybridization geometry of nitrogen, electronic character of the substituents and
Coulombic repulsion on oxidation potential (C). In all cases, stabilization of the resulting ammonium cation results in a decreased oxidation
potential. Structures are depicted to highlight the effect of geometric strain at NO.
υNO and Ea1. However, the υN−O*/Ea2 correlation arises from
stabilization of the resulting radical species. Additionally, it
should be noted that the correlation between Ea2 and υN−O* is
weaker than that of Ea1 and υNO. This is likely due to the
additional dependence of Ea2 on the protonation state of the
TH species as described earlier.
as demarcation lines, each of the compounds studied can be
separated into one of four quadrants based on their
electrochemical characteristics. A quadrant graph highlighting
these categorizations is depicted in Figure 6. The purpose of
this graph is to visualize the types of applications for which each
of the nitroxyl radical structures might be suitable.
Taken together, these two structure-function relationships
demonstrate that the catalytic activity is increased through
simultaneous destabilization of the oxoammonium cation, T+,
and the hydroxylammonium cation, TH2+. However, it is
Both quadrants I and II contain nitroxyl radicals that exhibit
high catalytic activity. However, quadrant I compounds display
a lower Ea1 and are potentially well-suited as anodic oxidation
catalysts in the context of a fuel cell, while compounds in
quadrant II may be more suited for electrosynthetic
applications in which the oxidation potential is less important
within the range described in Figure 6. Quadrant III
compounds are easily oxidized but relatively unreactive as
oxidation catalysts; these properties are desirable in the design
and selection of effective antioxidants or radical traps.47 Finally,
quadrant IV derivatives are not easily oxidized and their
oxidized form is relatively inactive. Therefore, these compounds
are considered the most stable nitroxyl radical species and
would be useful for applications in which an unreactive radical
species is needed such as in vivo EPR imaging.41 It should be
noted that the most widely used oxidation catalyst and EPR
label is TEMPO (1), likely due to ease of access. However,
TEMPO is neither the best nitroxyl radical for either
application.
+
important to note that destabilization of the off-cycle TH2
species is significantly more important for increasing catalytic
activity in aqueous systems because of the deactivating nature
+
of the TH2 formation. Therefore, the most effective strategy
for enhancing catalytic activity of nitroxyl radicals involves the
+
destabilization of TH2 either by geometric constraints,
hyperconjugative effects, or through Coulombic repulsion. Of
particular note, these parameters are easily considered and used
to anticipate the performance of proposed nitroxyl radical
electrocatalyst.
Application of Nitroxyl Radical Compounds. As
mentioned above, stable nitroxyl radical compounds can act
as efficient and environmentally friendly catalysts for the
oxidation of alcohols and amines; however, their use is not
limited to synthetic applications. Exploiting the statistical mean
of both the catalytic activity (described in terms of jmax) and Ea1
16184
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186

Journal of the American Chemical Society
Article
X-ray crystal data for selected synthetic intermediates.
(CIF)
Materials, synthetic procedures, assignment of Ea1 and
Ea2, derivation of kinetic equations, pH dependence,
computational methods, optimization of computational
level of theory and basis set, sample CVs, spectra,
summary of electrochemistry for each compound. (PDF)
AUTHOR INFORMATION
■
Corresponding Authors
*minteer@chem.utah.edu
*sigman@chem.utah.edu
Author Contributions
∥DPH and DAS contributed equally.
Notes
The authors declare no competing financial interest.
Figure 6. Plot of Ea1 versus (ipa/ipc)cat separated into four quadrants
based on the corresponding statistical averages of all nitroxyl radical
species studied. 4-Substituted TEMPO derivatives ( ), 4N-substituted
ACKNOWLEDGMENTS
○
■
■
▲
The support and resources from the Center for High
Performance Computing at the University of Utah are
gratefully acknowledged. Part of the calculations were done
using computational resources of Theoretical division, LANL,
which is supported by the Office of Science of the U.S.
Department of Energy under Contract No. DE-AC52-
06NA25396 and Center for Advanced Research Computing
at University of New Mexico. The authors would like to thank
the Army Research Office MURI (#W911NF1410263). This
paper was designated LA-UR-15-29076.
TEMPO derivatives ( ), and polycyclic species ( ) are represented
by different shapes. This is meant to aid in determining practical
applications for any given species based on its relative electrochemical
and catalytic characteristics.
CONCLUSION
■
We have reported a correlation between the difference in the
oxidation potentials of nitroxyl radical-containing compounds
and their activity as electrooxidation catalysts. Specifically, the
catalytic activity is directly proportional to the inverse of the
oxidation potentials, Ea1 and Ea2, such that (ipa/ipc)cat ∼ 1/(Ea2
− Ea1). The value of Ea1 was demonstrated to be directly
dependent on the stability of the resulting oxoammonium
cation, T+, while the value of Ea2 was shown to be reliant on the
REFERENCES
■
(1) Neiman, M. B.; Rozantzev, E. G.; Mamedova, Y. G. Nature 1962,
196, 472−474.
(2) Golubev, V. A.; Rozantsev, E. G.; Neiman, M. B. Bull. Acad. Sci.
́
+
USSR, Div. Chem. Sci. 1965, 14, 1898−1904.
destabilization of the hydroxylammonium cation, TH2 , in the
(3) Ganem, B. J. Org. Chem. 1975, 40, 1998−2000.
(4) Keana, J. F. W. Chem. Rev. 1978, 78, 37−64.
(5) Kelen, T.; Tudos, F.; Balint, G.; Rockenbaur, A. In Polymer
Stabilization and Degradation; American Chemical Society: Wash-
ington, D.C., 1985; Vol. 280, pp 109−117.
catalytic cycle.
The statistical analysis of a relatively large structural data set
provides a convenient method for visualizing important
characteristics of several water-soluble nitroxyl radical com-
pounds. Using either the graph shown in Figure 6 or the
corresponding data list (provided in the Supporting Informa-
tion), it is easy to select a TEMPO derivative that is well-suited
to a given application.
(6) Villamena, F. A.; Zweier, J. L. Antioxid. Redox Signaling 2004, 6,
619−629.
(7) Spin Labeling; Berliner, L.; Reuben, J., Eds.; Springer: New York,
1989.
(8) Bobst, A. M.; Kao, S.-C.; Toppin, R. C.; Ireland, J. C.; Thomas, I.
On the basis of the described correlation, DFT calculations
were used to develop a model able to accurately predict the
catalytic activity of several TEMPO derivatives. This model
provides a computational basis for predicting the catalytic
activity of nitroxyl radical catalysts in aqueous systems and
thereby, allowing for a method of rational catalyst design
without the need for synthesizing large libraries of TEMPO
derivatives. Additional correlations between the computed N
O and N−O* IR stretching frequencies and Ea1 and Ea2,
respectively, provide further insights into the structure−
function relationships from which the electrochemical
potentials are derived. Ongoing studies are focused on
expanding this model to nonaqueous systems and applying
reported or newly designed nitroxyl radicals to various energy
related goals.
E. J. Mol. Biol. 1984, 173, 63−74.
(9) McConnell, H. M.; McFarland, B. G. Q. Rev. Biophys. 1970, 3,
91−136.
(10) Song, C.; Hu, K.-N.; Joo, C.-G.; Swager, T. M.; Griffin, R. G. J.
Am. Chem. Soc. 2006, 128, 11385−11390.
(11) Krishna, M. C.; Grahame, D. A.; Samuni, A.; Mitchell, J. B.;
Russo, A. Proc. Natl. Acad. Sci. U. S. A. 1992, 89, 5537−5541.
(12) Goldstein, S.; Merenyi, G.; Russo, A.; Samuni, A. J. Am. Chem.
Soc. 2003, 125, 789−795.
(13) Goldstein, S.; Samuni, A.; Hideg, K.; Merenyi, G. J. Phys. Chem.
A 2006, 110, 3679−3685.
(14) Rozantsev, E. G.; Sholle, V. D. Synthesis 1971, 1971, 401−414.
(15) Bobbitt, J. M.; Flores, C. L. Heterocycles 1988, 27, 509−533.
(16) Ma, Z.; Bobbitt, J. M. J. Org. Chem. 1991, 56, 6110−6114.
(17) Rychnovsky, S. D.; McLernon, T. L.; Rajapakse, H. J. Org. Chem.
1996, 61, 1194−1195.
(18) Schnatbaum, K.; Schafer, H. J. Synthesis 1999, 1999, 864−872.
̈
ASSOCIATED CONTENT
■
(19) Fabbrini, M.; Galli, C.; Gentili, P.; Macchitella, D. Tetrahedron
Lett. 2001, 42, 7551−7553.
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.5b11252.
(20) Hoover, J. M.; Stahl, S. S. J. Am. Chem. Soc. 2011, 133, 16901−
16910.
16185
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186

Journal of the American Chemical Society
Article
(21) Lauber, M. B.; Stahl, S. S. ACS Catal. 2013, 3, 2612−2616.
(22) Ryland, B. L.; Stahl, S. S. Angew. Chem., Int. Ed. 2014, 53, 8824−
8838.
(23) Zhao, M.; Li, J.; Mano, E.; Song, Z.; Tschaen, D. M.; Grabowski,
E. J. J.; Reider, P. J. J. Org. Chem. 1999, 64, 2564−2566.
(24) Steelink, C.; Britton, W. E. Tetrahedron Lett. 1974, 15, 2869−
2872.
(49) Fu, Y.; Liu, L.; Yu, H.-Z.; Wang, Y.-M.; Guo, Q.-X. J. Am. Chem.
Soc. 2005, 127, 7227−7234.
(50) Tripkovic, V.; Bjorketun, M. E.; Skulason, E.; Rossmeisl. Phys.
̈
́
Rev. B: Condens. Matter Mater. Phys. 2011, 84, 115452.
(51) Milo, A.; Bess, E. N.; Sigman, M. S. Nature 2014, 507, 210−214.
(52) Milo, A.; Neel, A. J.; Toste, F. D.; Sigman, M. S. Science 2015,
347, 737−743.
(53) Benito, A.; Martinez-Manez, R.; Soto, J.; Tendero, M. J. L. J.
Chem. Soc., Faraday Trans. 1997, 93, 2175−2180.
(25) Semmelhack, M. F.; Chou, C. S.; Cortes, D. A. J. Am. Chem. Soc.
1983, 105, 4492−4494.
(26) Miyazawa, T.; Endo, T.; Shiihashi, S.; Okawara, M. J. Org. Chem.
1985, 50, 1332−1334.
(27) Miyazawa, T.; Endo, T. J. Org. Chem. 1985, 50, 3930−3931.
(28) Semmelhack, M. F.; Schmid, C. R. J. Am. Chem. Soc. 1983, 105,
6732−6734.
́
(29) Semmelhack, M. F.; Schmid, C. R.; Cortes, D. A. Tetrahedron
Lett. 1986, 27, 1119−1122.
(30) Bailey, W. F.; Bobbitt, J. M.; Wiberg, K. B. J. Org. Chem. 2007,
72, 4504−4509.
(31) Comminges, C.; Barhdadi, R.; Doherty, A. P.; O’Toole, S.;
Troupel, M. J. Phys. Chem. A 2008, 112, 7848−7855.
(32) Bobbitt, J. M. ChemInform 2011, 42, no.
(33) Rychnovsky, S. D.; Beauchamp, T.; Vaidyanathan, R.; Kwan, T.
J. Org. Chem. 1998, 63, 6363−6374.
(34) Shibuya, M.; Pichierri, F.; Tomizawa, M.; Nagasawa, S.; Suzuki,
I.; Iwabuchi, Y. Tetrahedron Lett. 2012, 53, 2070−2073.
(35) Doi, R.; Shibuya, M.; Murayama, T.; Yamamoto, Y.; Iwabuchi,
Y. J. Org. Chem. 2015, 80, 401−413.
(36) Shibuya, M.; Tomizawa, M.; Suzuki, I.; Iwabuchi, Y. J. Am.
Chem. Soc. 2006, 128, 8412−8413.
(37) Manda, S.; Nakanishi, I.; Ohkubo, K.; Yakumaru, H.;
Matsumoto, K.-i.; Ozawa, T.; Ikota, N.; Fukuzumi, S.; Anzai, K. Org.
Biomol. Chem. 2007, 5, 3951−3955.
(38) Rafiee, M.; Miles, K. C.; Stahl, S. S. J. Am. Chem. Soc. 2015, 137,
14751.
(39) Hickey, D. P.; McCammant, M. S.; Giroud, F.; Sigman, M. S.;
Minteer, S. D. J. Am. Chem. Soc. 2014, 136, 15917−15920.
(40) Nigam, S.; Asmus, K.-D.; Willson, R. L. J. Chem. Soc., Faraday
Trans. 1 1976, 72, 2324−2340.
(41) Fish, J. R.; Swarts, S. G.; Sevilla, M. D.; Malinski, T. J. Phys.
Chem. 1988, 92, 3745−3751.
(42) Kato, Y.; Shimizu, Y.; Yijing, L.; Unoura, K.; Utsumi, H.; Ogata,
T. Electrochim. Acta 1995, 40, 2799−2802.
(43) Ma, Y.; Loyns, C.; Price, P.; Chechik, V. Org. Biomol. Chem.
2011, 9, 5573−5578.
(44) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R.
W.; Klein, M. L. J. Chem. Phys. 1983, 79, 926−935.
(45) Rychnovsky, S. D.; Vaidyanathan, R.; Beauchamp, T.; Lin, R.;
Farmer, P. J. J. Org. Chem. 1999, 64, 6745−6749.
(46) Hodgson, J. L.; Roskop, L. B.; Gordon, M. S.; Lin, C. Y.; Coote,
M. L. J. Phys. Chem. A 2010, 114, 10458−10466.
(47) Hodgson, J. L.; Namazian, M.; Bottle, S. E.; Coote, M. L. J. Phys.
Chem. A 2007, 111, 13595−13605.
(48) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery Jr., J. A.;
Peralta, J. E.; Ogliaro, F. o.; Bearpark, M. J.; Heyd, J.; Brothers, E. N.;
Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.;
Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.;
Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.;
Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.;
Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A.
̃
D.; Farkas, A.-d. n.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D.
J. Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009.
16186
DOI: 10.1021/jacs.5b11252
J. Am. Chem. Soc. 2015, 137, 16179−16186