Top Catal (2014) 57:215221

DOI 10.1007/s11244-013-0171-6

ORIGINAL PAPER

Unifying Solution and Surface Electrochemistry: Limitations

and Opportunities in Surface Electrocatalysis
Venkatasubramanian Viswanathan

Heine Anton Hansen

Published online: 23 October 2013

Springer Science+Business Media New York 2013

Abstract The electrochemical transformations involving principle, be enabled by low-temperature electrochemical

small molecules could play a key role in energy storage and transformations.
utilization. In this work, we develop a link between surface These electrochemical transformations could be carried
and solution electrochemistry and this is accomplished by out through the activation of small molecules and con-
relating the correlated binding of hydrogenated species on verting them into high value fuels. However, it is seen that
metal surfaces to their formation energy in solution. This the efficiencies of electrochemical reactions involving
link provides a semi-quantitative treatment of the thermo- small molecules such as O2, N2 and CO2 are low [24].
dynamic landscape of multi-electron electrochemical These transformations involve bond making and bond
reactions. Using this simple analysis, we identify some of breaking processes that require multi-electron and proton
the fundamental limits of oxygen, nitrogen and carbon transfer reactions and it is generally seen that the effi-
electrochemistry. This analysis points to new avenues of ciencies become progressively worse with increasing
electrocatalyst design and two such avenues are discussed number of electronproton transfer reactions involved.
in detail. Although advances have been made in the theory of
single electron transfer reactions, [5] the theory of multi-
Keywords Electrocatalysis  Solar fuels  Volcano electron electrochemical reactions has not yet reached a
curves similar level. Recently, the thermodynamic treatment of
multi-electron electrochemical reactions has seen success
in understanding trends for simple reactions such as
1 Introduction hydrogen evolution reaction, oxygen reduction reaction
(ORR) and oxygen evolution reaction (OER) [68]. This
The use of solar and wind energy represents attractive treatment, largely following the principles laid out in the
sustainable energy options with negligible carbon foot- seminal work by Nrskov et al. [6] attributes the limita-
print, reasonable energy efficiency and minimal water tions of electrocatalysis to the correlated binding between
consumption. However, if solar and wind are to be used certain reaction intermediates. The scaling between these
as a primary energy source, the ability to store and dis- reaction intermediates immediately imposes fundamental
patch energy on demand to the consumer is necessary [1]. limits to surface electrocatalysis for multi-electron reac-
Efficient energy storage represents a major barrier and tions such as ORR and OER [8, 9]
one attractive avenue is to utilize natures approach and In this work, we explore the origin of these fundamental
store them in the form of chemical bonds. The storage of limitations of surface electrochemistry by forming a bridge
sun and wind energy in the form of fuels could, in between the concepts developed in the thermodynamic
theory of multi-electron reactions on surfaces to aqueous
solution electrochemistry. The establishment of this link
V. Viswanathan (&)  H. A. Hansen
points to certain new avenues which could, in principle,
Department of Chemical Engineering, Stanford University,
Stanford, CA 94305-3030, USA avoid some of the limits seen in aqueous surface
e-mail: [email protected] electrocatalysis.

123
216 Top Catal (2014) 57:215221

2 Surface Electrochemistry the last step that becomes downhill in free energy when the
overpotential is increased [6].
In this section, we introduce the concepts involved in the The free energy diagram for the oxygen reduction
thermodynamic theory of multi-electron electrochemical reaction on Pt(111) constructed using the approach detailed
reactions using ORR as an example. An ideal electrocata- above is shown in Fig. 1. On Pt(111), the limiting poten-
lyst for the four electron reduction of oxygen to water tial, UL is given by the free energy change associated with
should be able to facilitate the reaction just below the the reduction of OH* to H2O(l), reaction 1d. This analysis
equilibrium potential of 1.23 V. This requires, as a mini- suggests that directions for better catalysts are those that
mum requirement, that the four charge transfer steps bind OH* weaker than Pt(111). This principle has been
involved have reaction free energies of the same magnitude used to discover several new catalysts [14]. However, as
equal to the equilibrium potential of 1.23 eV [6]. we will discuss in the next section, there is a fundamental
Considering the associative mechanism shown below [6, limit to how much better these catalysts could be.
10]:
2.1 Scaling Relations
O2 (g) 4H 4e  ! OOH 3H 3e ; 1a
OOH 3H 3e ! O H2 O(l) 2H 2e ; 1b In a pioneering work, Abild-Pedersen et al. [15] showed
the existence of a linear scaling between the adsorption
O H2 O(l) 2H 2e ! OH H2 O(l) H e ; energy of hydrogenated species, AHx and the atom A. It
1c was further shown using simple bond counting principles
that the slope of the linear scaling between the hydroge-
OH H2 O(l) H e !  2H2 O(l); 1d
nated species, AHx and the atom A was only dependent on
where * refers to a surface site. In this mechanism, the the valency and did not depend on the specifics of metal
electrocatalytic activity for oxygen reduction to water is and the facet. The existence of these linear scaling relations
governed by the stability of the intermediates, OOH*, has an important implication for the oxygen reduction
OH* and O*. It is possible that the transfer of a proton and reaction. This implies that the binding of the oxygen
electron could be decoupled at high pH for weak binding intermediates, OOH*, OH* and O* cannot be changed
catalysts and this could lead to additional intermediates that independently and the activity is only governed by a single
could affect the electrocatalytic activity [11]. However, in parameter [9, 13].
acidic environment, we expect the concerted pathway From the analysis in the previous section, it was shown
involving the simultaneous transfer of a proton and elec- that a weakening of OH* could lead to better catalysts.
tron to be dominant. However, a weakening of OH* also weakens O* and OOH*.
The reaction free energies of a step in the mechanism The linear scaling between OOH* and OH* adsorbed on fcc
can be calculated at standard conditions, and a potential, U metal (111) and (100) facets is shown in Fig. 1b and this
= 0 V versus the reversible hydrogen electrode, using the indicates that beyond a certain weakening of OH*, the
relation: limiting step changes to reaction 1a. This leads to a
Sabatier-type volcano relationship with the optimal elect-
DG0 DEw;water DZPE  TDS; 2 rocatalyst having equal free energy difference for reaction
1a and reaction 1d and this is shown in Fig. 1c. This
where DEw;water is the reaction energy including the sta-
analysis suggests that the origin of the limitations of oxy-
bilization of water, calculated relative to H2O and
gen reduction reaction on metals is due to the unfavorable
H2 ; DZPE is the difference in zero point energies for the
scaling between OOH* and OH*. On an ideal electrocata-
reaction, and DS is the change in entropy for the reaction
lyst, the difference in free energy between these two
considered. A potential of U = 0 V versus the reversible
intermediates, OOH* and OH* should be 2 9 1.23 =
hydrogen electrode implies the relation, DGH DGe
aq 2.46 eV while according to the scaling relation, this dif-
1
2 DGH2 g :
This relation provides a computationally tracta- ference is 3.2 eV, indepedent of the metal surface. This
ble way of calculating the sum of the free energy of a immediately imposes a fundamental limit given by
proton and an electron [6, 12]. The effect of a potential U is gORR = 0.37 V, even on the best metal electrocatalyst
included by shifting the free energy of an electron by [16].
-eU. The reaction free energy is given by DG DG0 eU; In the next section, we will explore the possibility of
where U is the potential relative to the reversible hydrogen identifying the fundamental limits of electrocatalysis by
electrode. This procedure can be used to construct free making a connection to solution electrochemistry. An
energy diagrams and can be used to deduce an important analysis similar in spirit to the one carried out here is
parameter, the limiting potential, UL, which is defined as discussed in a recent work [11].

123
Top Catal (2014) 57:215221 217

Fig. 1 a Free energy diagram (a)

for the oxygen reduction
reaction on Pt(111) plotted at
UL = 0.76 V. The
potential limiting step is the
reduction of OH* to H2O(l). b A
plot of the binding energy of
OOH* versus OH* on fcc metal
(111) and (100) facets. The
scaling line,
DEOOH DEOH 3:2[eV]; is
shown in the figure. The red
dots and blue squares are
binding energies on (111) and
(100) surfaces respectively. c
Activity volcano of the limiting
potential, UL as a function of
DGOH ; the free energy of
adsorbed OH* is shown. The (b) (c)
source of the experimental data
points are listed in ref. [13]

3 Solution Electrochemistry Table 1 The acid dissociation constant, pKa, of different molecules
at 25 C
The free energy diagrams discussed in the context of sur- Name Formula pKa Ref.
face electrochemistry are known as LatimerFrost dia-
grams in solution electrochemistry. These diagrams, also Water H2O 15.7 [29]
known as oxidation potential diagrams, were originally Hydrogen peroxide H2O2 11.6 [29]
introduced by Latimer to study the stability of different cis-Diazene N2H2 34.3 [29]
oxidation states of oxides [17]. A refinement of these dia- Hydrazine N2H4 32 [30]
grams was later proposed by Frost, where it was suggested Ammonia NH3 38 [29]
that the use of free energy could be advantageous in better Formic acid HCOOH 3.75 [31]
understanding the reaction landscape [18]. These diagrams Formaldehyde-hydrate H2C(OH)2 13.27 [31]
can be used to study the stable phases of transistion metal Methanol CH3OH 15.5 [31]
oxides, identify stability windows in aqueous solutions and Methane CH4 60 [23]
reaction free energy diagrams. In this section, we explore
these diagrams in the context of small molecule electro-
chemisty namely oxygen, nitrogen and carbon. of oxygen electrochemistry, the 2e- reduction results in
H2O2 and 4e- reduction results in H2O. The free energy
3.1 Procedure of these molecules can be obtained from thermodynamic
tables. The free energy of the corresponding solution
The construction of Frost diagrams involves identifying species can be obtained using the acid dissociation con-
the free energy of intermediate solution species. In a Frost stant, pKa, listed in Table 1. This procedure is used to
diagram, the sequential addition of two protons and construct the Frost diagrams for oxygen, nitrogen and
electrons results in a molecule. For instance, in the case carbon electrochemistry.

123
218 Top Catal (2014) 57:215221

Fig. 2 Frost diagram for oxygen electrochemistry plotted at a Fig. 3 Frost diagram for nitrogen electrochemistry at a potential of
potential of U = 0 VRHE and pH = 0. The stability of OOH- aq and U = 0 VRHE and pH = 0. The stability of NNH- -
aq and NH2, aq is
OH- aq is obtained using the pKa of H2O2 and H2O respectively. The obtained using the pKa of N2H2 and NH3 respectively. The free
free energy difference between these two species in solution is 3.4 eV energy difference between these two species in solution is
2.52 0.24 eV. The error bar is estimated based on 2 units of
uncertainty in the pKa for ammonia and diazene
3.2 Oxygen Electrochemistry

Using the procedure outlined above, we construct the Frost We explore the origin of this large overpotential by
diagram for oxygen electrochemistry and this is shown in examining the Frost diagram for nitrogen electrochem-
Fig. 2. The free energy of the two key solution species, istry and this is shown in Fig. 3. The reduction of
OOH 
aq and OHaq is obtained using the pKa of H2O2 and nitrogen to ammonia proceeds through the addition of
H2O respectively, which are listed in Table 1. It is inter- six protons and electrons and thus there are two inter-
esting to note that the free energy difference between these mediate molecules involved in the sequential reduction
two intermediates is 3.4 eV. This is quite close to the of N2, namely the 2e- reduction to cis-diazene, N2H2
scaling found on metal surfaces and this is not entirely and the 4e- reduction to hydrazine, N2H4. The origin of
surprising as the scaling relations were found to be uni- the difficulty associated with nitrogen reduction becomes
versally independent of metal and facet. Thus, we would immediately apparent by looking at the Frost diagram. In
expect the solution free energy difference to give a good this analysis, we attribute the limitations in nitrogen
first approximation for the free energy difference between electrochemistry to the large free energy difference
these intermediates on a metal surface. This relation is between NNH 
aq and NH2;aq : The free energy differ-
given by: ence between these two species is 2.52 0.24 eV and
DGOOH  DGOH  DGOOHaq  DGOHaq 3 they are separated by four protons and electrons. In the
ideal case, this difference should only be & 0.22 eV.
This analysis can be used to get a preliminary idea of the Using the approximation,
possible energy landscapes of multi-electron electrochem- DGNNH  DGNH 2  DGNNHaq  DGNH2;aq 4
ical reactions. In the next section, we will explore this
methodology to nitrogen and carbon chemistry in an we immediately identify a fundamental limit for nitrogen
attempt to identify the fundamental limits for these electrocatalysis to ammonia given by g = (2.52 - 0.37
reactions. 0.24)/4 = 0.54 0.06 V. A similar limit was found in a
recent work using detailed density functional theory cal-
3.3 Nitrogen Electrochemistry culations of reaction intermediates associated with nitrogen
electrochemistry on stepped metal surfaces [20]. According
The electrochemical reduction of nitrogen to ammonia has to our analysis, the metal surfaces need to have sufficient
a reversible potential, U0 & 0.09 VRHE. However, there is a binding strength to activate N2 as NNH*. However on these
large barrier to carry out this transformation and typically surfaces, the binding of NH*2 will be so strong that the
high temperature and pressure is required [3]. However, in limiting step would be the reduction of NH*2 to NH3
biology, this process can be carried out with the help of forming the strong binding leg of the volcano. At suffi-
nitrogenase enzyme at room temperature [19]. This cient weakening of NH*2, due to corresponding weaken-
enzyme, however, requires the hydrolysis of 16 adenosine ing of NNH*, the limiting step transitions to the activation
triphosphate (nearly 5 eV) molecules to perform biological of N2 to NNH* forming the weak binding leg of the vol-
nitrogen fixation. cano. The predictions of our analysis agree well with the

123
Top Catal (2014) 57:215221 219

adsorbates that bind through the carbon end and those that
bind through the oxygen end. [22] These details are crucial
in developing a detailed treatment of surface electrocatal-
ysis of CO2 reduction or methanol oxidation.
In this work, we consider a simplified analysis, consid-
ering bonding primarily through the oxygen atom as the
acid dissociation constants can be experimentally deter-
mined only for the hydrogen attached to the oxygen atom.
The acidity of the hydrogens on the carbon atom are so
weak that these can not be directly measured and can only
be obtained through extrapolation methods [23]. However,
we expect that this analysis could prove to be useful on
Fig. 4 Frost diagram for carbon electrochemistry at a potential of surfaces that have high oxygen affinity. Additionally, the

U = 0 VRHE and pH = 7. The stability of HCOO aq ; H2 COHOaq ; species considered in our analysis, namely HCOO-,
-
and CH3Oaq is obtained using the pKa of formic acid, formaldehyde- CH2COO- and CH3O-, form the intermediates in one of
hydrate and methanol respectively. The free energy of H2 COH 2;aq
the possible pathways for methanol synthesis on Cu sur-
was calculated using an effective Henrys law coefficient of
kH = 3.0 9 103 accounting for hydration [27]. The free energy of faces in heterogenous catalysis [24]. It has been shown that
CH3 OHaq was calculated using Henrys law coefficient, binding energy of the oxygen-bound species, HCOO*,
kH = 2 9 102 [28]. The free energy difference between HCOO- CH2COO* and CH3O* scale with the binding energy of
and H2C(OH)O- is 0.52 eV and the free energy difference between OH* with the same slope further validating the link to
H2 C(OH)O and CH3 O is -0.42 eV
solution electrochemistry [25]. In our analysis, we find that
detailed density funcational theory treatment of N2 reduc- the limiting step is the reduction of CO or HCOO- to
tion [20]. H2C(OH)O- and we estimate a limiting potential of
A similar discussion can be made for the oxidation of UL = -0.26 VRHE. However, it is worth pointing out that
ammonia to nitrogen. The overpotential for ammonia oxi- this analysis assumes the activation of CO2 through the
dation on Pt(100), which is one of the best catalysts for this oxygen end. The binding of these intermediates through the
reaction, is around 0.50.6 V, close to the fundamental carbon end would be competing and the barriers associated
limit identified [21]. The mechanism suggested for with these processes are expected to play an important role
ammonia oxidation to molecular nitrogen is in good in determining the exact mechanism on different metals. In
agreement with the Frost diagram analysis discussed here. enzyme catalysis, however, it is known that carbon mon-
Our simple analysis suggests that electrochemical oxida- oxide dehydrogenase enzyme can reversibly catalyze the
tion of hydrazine could also be difficult, but a more transformation of CO2 to CO [26]. Thus, in principle, it
detailed analysis is necessary to identify the fundamental could be possible to find catalysts that mimic the active
limitations associated with this reaction. centers of these enzymes. The understanding of trends in
electrochemical reduction of CO2 on metal surfaces is still
3.4 Carbon Electrochemistry at its early stages and our present analysis does not rule out
the existence of efficient electrocatalysts for CO2
CO2 reduction may result in a wide range of useful products. reduction.
However, there are no known metal electrocatalysts that can
catalyze CO2 reduction at low overpotentials [4]. The best
known elemental catalyst that can reduce CO2 to hydrocar- 4 Opportunities in Surface Electrocatalysis
bons is copper, which requires an overpotential of &1 V. A
particular interesting fuel which may be formed by CO2 In this section, we try to explore possible ways to enhance
reduction is CH3OH. Similar to nitrogen reduction, the the efficiency of electrocatalytic transformations involving
reduction of CO2 to CH3OH occurs at an U0 & 0.03VRHE. small molecules. We discuss in detail two approaches
We construct the Frost diagram for carbon electro- which could avoid the limits of electrocatalysis identified
chemistry and this is shown in Fig. 4. The treatment of in this work.
carbon electrochemistry is complicated by the fact that the
initial reactant is CO2. Thus, the addition of a proton and an 4.1 Electrochemistry in Non-aqueous Electrolytes
electron could occur at either the carbon atom or the oxygen and Ionic Liquids
atom. The approach of linking solution intermediates to
surface intermediates relies on similar characteristics of the We use oxygen reduction reaction as the prototypical
metal-adsorbate bond and this does not hold between reaction to explore approaches to avoid the fundamental

123
220 Top Catal (2014) 57:215221

chosen should be such that it has weaker interactions with

OH- than water. Hard anions such as OH- have localized
charge and the solvation of these anions requires hydrogen
bonding. As a result, the solubility of these anions in aprotic
solvents is low. Thus, one approach is to add a mix of aprotic
solvents such as dimethylformamide, dimethyl sulfoxide,
acetonitrile, etc. to tune the nature of solvent interactions.
These solvents show a destabilization of about 0.10.3 eV
relative to water for hard anions [35].
Another approach is to use room-temperature ionic
liquids to tune the solvent interactions. The exponential
growth in the number of synthesized ionic liquids with
Fig. 5 Scaling between OOH* and OH* in different medium and wide stability windows makes them attractive candidates as
surfaces. The free energy difference between OOH 
aq and OHaq is solvents [36, 37]. With tremendous control over size of
3.4 eV. The free energy difference of adsorbed OOH and OH* on
*
cations and anions, these ionic liquids could be used to tune
metals and metal oxides is 3.2 and 3.0 eV respectively. The free
energy difference between OOH- - interactions with adsorbates. However, the understanding
(g) and OH(g) is 2.79 eV. The free
energy difference on an ideal electrocatalyst is 2.46 eV of electrodeionic liquid interface is at its infancy and
several efforts have been made to understand the structure
of the double layer of the electrodeionic liquid interface
limits of surface electrocatalysis. In the case of the oxygen [3840]. Recently, it has been shown using a linear
reduction reaction, the limit to surface electrocatalysis is response treatment of molten salts that due to correlations,
imposed due to the unfavorable scaling of 3.2 eV between the phenomenon of overscreening is observed where the
OOH* and OH*. Improvements in surface electrocatalysis first layer at the electrode has more countercharge than on
is possible if this scaling is lowered closer to the ideal value the surface with the second layer having a smaller net
of 2.46 eV. This is possible through the selective stabil- charge of the opposite sign, which again overscreens, and
ization of OOH* relative to OH*. Through detailed density so on, until electroneutrality is achieved [41, 42]. Through
functional theory calculations of the stabilization of water careful choice of cations and anions, it appears to be pos-
on OH* and OOH*, it is known that water stabilizes OH* by sible to tune the adsorbateionic liquid interactions to
about 0.5 eV and OOH* by only about 0.25 eV [10]. One enhance electrocatalysis. These ideas have seen pre-
approach to change the scaling is through the tuning of liminary sucess in the work of Freisen and co-workers who
solvent interactions. An instance of this is through loss of have shown that through suitable tuning of the pKa of the
solvation due to water which is typically the case on metal ionic liquid, control over oxygen electrocatalysis can be
oxides. On oxide surfaces, the scaling between OOH* and achieved [43]. They have demonstrated that the turn on
OH* shifts to about 3 eV [16] and it is not surprising that potential for oxygen reduction on Pt can be tuned using the
this value is closer to the difference between these inter- pKa of the aprotic ionic liquid, with the optimal ionic
mediates in gas phase of 2.79 eV where there is no sol- liquid having a turn-on potential of 1.0 V generating a
vation. The scaling between these two intermediates in current of nearly 1 mA/cm2 [43]. However, a more detailed
different media and surfaces is shown in Fig. 5. This points understanding is necessary to design catalyst-ionic liquid
to the opportunity of tuning surface electrochemistry combinations and the development of detailed design
through suitable catalyst-solvent combinations. principles in this area could be useful in computational
The nature of solvent interactions are characterized based identification of candidate combinations.
on donor numbers (DN) and acceptor numbers (AN). The
concept of DN, introduced by Gutmann and Mayer [32] 4.2 Bifunctional Electrocatalysts
provides a measure of the donor properties of solvents. This
is defined as the numerical value of the heat of adduct for- Another approach to avoid the limitations imposed by the
mation between the donor molecule and the reference scaling relations is through suitable structural changes at
acceptor [33]. An analogous definition is used for the the electrocatalyst surface. This could be achieved through
acceptor number and the usefulness of these numbers were alloying with the surface containing two different metals.
shown by correlating shifts in redox potentials to donor and The presence of two different metals could help in selective
acceptor numbers [34]. These can also be used to predict the stabilization of one intermediate relative to another [22].
thermodynamics of ions in solution which is the key metric in Another route is through patterning of electrode surfaces
designing solvents for electrocatalysis. In the case of oxygen with adsorbates that can subsquently affect the thermody-
reduction electrocatalysis, as discussed above, the solvent namics of reaction intermediates. This strategy has been

123
Top Catal (2014) 57:215221 221

used to design cyanide patterned Pt surfaces to enhance 13. Viswanathan V, Hansen HA, Rossmeisl J, Nrskov JK (2012) J
oxygen reduction reaction activity [44]. A detailed under- Phys Chem Lett 3(20):2948
14. Stephens IEL, Bondarenko AS, Perez-Alonso FJ, Calle-Vallejo F,
standing of the underlying principles for bifunctionality Bech L, Johansson TP, Jepsen AK, Frydendal R, Knudsen BP,
could aid in the successful development of efficient Rossmeisl J, Chorkendorff I (2011) J Am Chem Soc
electrocatalysts. 133(14):5485
15. Abild-Pedersen F, Greeley J, Studt F, Rossmeisl J, Munter TR,
Moses PG, Skulason E, Bligaard T, Nrskov JK (2007) Phys Rev
Lett 99(1):016105
5 Summary 16. Koper MTM (2011) J Electroanal Chem 660(2):254
17. Latimer WM (1938) Oxidation Potentials. Prentice Hall Inc.,
In this work, we have established a link between surface Englewood Cliffs
18. Frost AA (1951) J Am Chem Soc 73(6):2680
electrochemistry and solution electrochemistry by linking 19. Howard J, Rees D (1996) Chem Rev 96(7):2965
the correlated binding energies of hydrogenated species to 20. Skulason E, Bligaard T, Gudmundsdottir S, Studt F, Rossmeisl J,
their formation energies in solution. Using this link, we Abild-Pedersen F, Vegge T, Jonsson H, Nrskov JK (2012) Phys
have identified some fundamental limits of oxygen, nitro- Chem Chem Phys 14(3):1235
21. Rosca V, Koper MTM (2006) Phys Chem Chem Phys 8(21):2513
gen and carbon electrochemistry. The exact reaction 22. Peterson AA, Nrskov JK (2012) J Phys Chem Lett 3(2):251
mechanisms on electrocatalytic surfaces will be more 23. Jaun B, Schwarz J, Breslow R (1980) J Am Chem Soc
complex and this analysis could be used to get a pre- 102(18):5741
liminary idea of the landscape of some of the reaction 24. Behrens M, Studt F, Kasatkin I, Kuhl S, Havecker M, Abild-
Pedersen F, Zander S, Girgsdies F, Kurr P, Kniep BL, Tovar M,
pathways involved. Fischer RW, Nrskov JK, Schlogl R (2012) Science
We would like to emphasize that the principles devel- 336(6083):893
oped in understanding small molecule electrochemistry is 25. Tritsaris GA, Rossmeisl J (2012) J Phys Chem C 116(22):11980
expected to play an important role in reactions involving 26. Armstrong FA, Hirst J (2011) Proc Natl Acad Sci USA
108(34):14049
larger molecules as well. The reactive moieties of the large 27. Betterton EA, Hoffmann MR (1988) Environ Sci Technol
molecules are similar to those discussed in the context of 22(12):1415
small molecules and we expect the ideas presented in this 28. Snider JR, Dawson GA (1985) J Geophys Res 90(D2):3797
work to be generally applicable. 29. Warren JJ, Tronic TA, Mayer JM (2010) Chem Rev
110(12):6961
30. Bauer N (1960) J Phys Chem 64(7):833
Acknowledgements The authors would like to thank Alan Luntz 31. Haynes WM (ed) (2012) In: CRC handbook of chemistry and
for helpful discussions. physics, 93rd edn (Internet Version 2013). CRC Press/Taylor and
Francis, Boca Raton, FL
32. Mayer U, Gutmann V, Gerger W (1975) Monatsh Chem
106(6):1235
References 33. Mayer U (1979) Pure Appl Chem 51:1697
34. Mayer U (1980) In: Yeager EB, Schumm JB, Blomgren G,
1. Lewis NS, Nocera DG, Natl P (2006) Acad Sci USA Blankenship DR, Leger V, Akridge J (eds) Proceedings of the
103(43):15729 workshop on lithium nonaqueous battery electrochemistry, p. 13
2. Markovic NM, Ross PN (2002) Surf Sci Rep 45:117 35. Aurbach D, Weissman I (1999) Non-aqueous electrochemistry
3. Rosca V, Duca M, de Groot MT, Koper MTM (2009) Chem Rev chap. 1. Marcel Dekker, New York
109(6):2209 36. Welton T (1999) Chem Rev 99:2071
4. Hori Y, Kikuchi K, Suzuki S (1985) Chem Lett 14(11):1695 37. Hallett JP, Welton T (2011) Chem Rev 111(5):3508
5. Marcus R (1965) J Chem Phys 43(2):679 38. Kornyshev AA (2007) J Phys Chem B 111(20):5545
6. Nrskov JK, Rossmeisl J, Logadottir A, Lindqvist L, Kitchin JR, 39. Lockett V, Sedev R, Ralston J, Horne M, Rodopoulos T (2008)
Bligaard T, Jonsson H (2004) J Phys Chem B 108(46):17886 Chem Rev 112(19):7486
7. Nrskov JK, Bligaard T, Logadottir A, Kitchin JR, Chen J, 40. Islam MM, Alam MT, Ohsaka T (2008) Chem Rev 112(42):
Pandelov S, Stimming U (2005) J Electrochem Soc 152(3):J23 16568
8. Man I, Su H, Calle-Vallejo F, Hansen HA, Martnez JI, Inoglu 41. Fedorov MV, Kornyshev AA (2008) J Phys Chem B
NG, Kitchin J, Jaramillo TF, Nrskov JK, Rossmeisl J (2011) 112(38):11868
ChemCatChem 3:1159 42. Bazant MZ, Storey BD, Kornyshev AA (2011) Phys Rev Lett
9. Viswanathan V, Hansen HA, Rossmeisl J, Nrskov JK (2012) 106(4):046102
ACS Catal 2(8):1654 43. Zeller RA (2011) Effects of extrinsic and intrinsic proton activity
10. Karlberg GS, Rossmeisl J, Nrskov JK (2007) Phys Chem Chem on the mechanism of oxygen reduction in ionic liquids. Ph.D.
Phys 9:5158 thesis. Arizona State University, Phoenix
11. Koper MTM (2013) Chem Sci. doi:10.1039/c3sc50205h 44. Strmcnik D, Escudero-Escribano M, Kodama K, Stamenkovic
12. Viswanathan V, Hansen HA, Rossmeisl J, Jaramillo TF, Pitsch H, VR, Cuesta A, Markovic NM (2010) Nat Chem 2(10):880
Nrskov JK (2012) J Phys Chem C 116(7):4698

123