of 18
A guide to troubleshooting metal sacri
fi
cial anodes
for organic electrosynthesis
Skyler D. Ware,
a
Wendy Zhang,
a
Weiyang Guan,
b
Song Lin
b
and Kimberly A. See
*
a
The development of reductive electrosynthetic reactions is often enabled by the oxidation of a sacri
fi
cial
metal anode, which charge-balances the reductive reaction of interest occurring at the cathode. The
metal oxidation is frequently assumed to be straightforward and innocent relative to the chemistry of
interest, but several processes can interfere with ideal sacri
fi
cial anode behavior, thereby limiting the
success of reductive electrosynthetic reactions. These issues are compounded by a lack of reported
observations and characterization of the anodes themselves, even when a failure at the anode is
observed. Here, we weave lessons from electrochemistry, interfacial characterization, and organic
synthesis to share strategies for overcoming issues related to sacri
fi
cial anodes in electrosynthesis. We
highlight common but underexplored challenges with sacri
fi
cial anodes that cause reactions to fail,
including detrimental side reactions between the anode or its cations and the components of the
organic reaction, passivation of the anode surface by an insulating native surface
fi
lm, accumulation of
insulating byproducts at the anode surface during the reaction, and competitive reduction of sacri
fi
cial
metal cations at the cathode. For each case, we propose experiments to diagnose and characterize the
anode and explore troubleshooting strategies to overcome the challenge. We conclude by highlighting
open questions in the
fi
eld of sacri
fi
cial-anode-driven electrosynthesis and by indicating alternatives to
traditional sacri
fi
cial anodes that could streamline reaction optimization.
1. Introduction
Organic electrosynthesis has undergone a major revival in the
past several years.
1
12
The rapidly growing
eld enables new
reactivity that is not achievable with traditional synthetic
methods.
13,14
Furthermore, the use of electrons as reagents
o
ff
ers routes to greener and safer synthetic conditions, poten-
tially eliminating the need for harsh or hazardous chemicals.
11
Skyler D
:
Ware
Skyler Ware received her BS in
chemistry from The Ohio State
University in 2018. In 2023, she
earned her PhD in chemistry
from Caltech under the supervi-
sion of Prof. Kimberly See. Her
research focuses on electro-
chemical interfaces and non-
aqueous electrolytes in energy
storage systems and organic
electrosynthesis.
Wendy Zhang
Wendy Zhang received her BS
and MS in chemistry from the
College of William and Mary in
2015 and 2016 under the
supervision of Prof. William
McNamara. In 2023, she earned
her PhD in chemistry from Cal-
tech under the supervision of
Prof. Kimberly See. Her research
in the See lab focuses on under-
standing and manipulating
electrode interfaces to improve
the metal sacri
cial anode
performance for organic
electrosynthesis.
a
Division of Chemistry and Chemical Engineering, California Institute of Technology,
Pasadena, California 91125, USA. E-mail: ksee@caltech.edu
b
Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York
14853, USA
Cite this:
Chem. Sci.
,2024,
15
,5814
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
Received 22nd December 2023
Accepted 26th February 2024
DOI: 10.1039/d3sc06885d
rsc.li/chemical-science
5814
|
Chem. Sci.
,2024,
15
,5814
5831
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chemical
Science
PERSPECTIVE
Electrosynthesis also enables greater selectivity in synthetic
reactions
via
the ability to
ne-tune either the applied potential
to promote speci
c reduction or oxidation reactions, or the
applied current to select for the most kinetically labile reac-
tions.
15
Electrochemical experiments o
ff
er a handle by which to
probe the reaction mechanism, enabling greater mechanistic
insight than would be achievable with chemical reductants.
16
18
Though an attractive goal of organic electrosynthesis is to
develop paired electrolyses in which reactions at both the anode
and the cathode contribute to value-added products,
1,2,19
23
individual half reactions must be well understood if they are to
be combined into a larger process. In reductive electrosynthesis,
these half reactions frequently rely on charge balancing
via
the
oxidation of a sacri
cial anode, typically Mg, Al, Zn, or Fe.
24
Sacri
cial anodes have enabled the use of electrochemistry in
a variety of organic reactions.
7
10,22,24
26
During electrolysis, the
metal electrode is oxidized, releasing metal cations into solu-
tion as shown in Fig. 1. The cations are o
en thought to be inert
to the reductive chemistry of interest. In some situations, the
cations do participate in the reaction of interest; in these cases
the anode is not truly sacri
cial and plays a more substantial
role.
27
29
In an ideal case, electrolysis with sacri
cial anodes o
ff
ers
several bene
ts during early stages of reaction development.
First, the oxidation reaction itself is typically straightforward
Fig. 1
Sacri
fi
cial metal anodes enable reductive electrosynthesis by
charge-balancing the target reductive reactions at the cathode. During
a reductive reaction, the metal sacri
fi
cial anode is oxidized, releasing
metal cations into solution.
Weiyang Guan
Weiyang Guan was born in Zhe-
jiang, China. A
er obtaining her
bachelor's degree from the
University of Wisconsin-Madison
in 2019, she became interested
in electrochemistry and joined
the Song Lin group at Cornell
University for further studies. Her
PhD research is focused on
applying reductive electrochem-
istry for strong bonds activation,
especially in the generation of
reactive intermediates for down-
stream transformations.
Song Lin
Song Lin grew up in Tianjin,
China. A
er earning a BS degree
from Peking University in 2008,
he pursued graduate studies at
Harvard University working
under the direction of Professor
Eric Jacobsen. He then carried
out postdoctoral studies with
Professor Chris Chang at UC
Berkeley. In the summer of 2016,
Song moved to Ithaca to start his
independent career. He was
promoted to Associate Professor
with tenure in July 2021 and
then to Full Professor in January 2023. He is currently an Associate
Editor for Organic Letters and serves on the Editorial Advisory
Board of Synlett, Chem, and Chemistry
A European Journal as
well as the Scienti
c Advisory Board of Snapdragon Chemistry.
The Lin Laboratory's research lies at the interface of electro-
chemistry and organic chemistry, with a main objective of using
fundamental principles of electrochemistry and radical chemistry
to discover new organic transformations and uncover new reaction
mechanisms.
Kimberly A
:
See
Kimberly See is an Assistant
Professor of Chemistry in the
Division of Chemistry and
Chemical Engineering at Cal-
tech. She was born and raised in
Colorado and received her BS in
Chemistry from the Colorado
School of Mines. Kim earned her
PhD in Chemistry at the
University of California, Santa
Barbara where she worked with
Prof. Ram Seshadri and Galen
Stucky. Kim was awarded the St.
Elmo Brady Future Faculty
Postdoctoral Fellowship at the University of Illinois at Urbana-
Champaign and worked with Prof. Andrew Gewirth in the
Department of Chemistry. Now, her group at Caltech studies
electrochemical systems with sustainability applications, ranging
from next-generation battery chemistries to developing electrolytes
and electrodes for green organic electrosynthesis. Her group
studies both interfaces and bulk materials properties in electro-
chemical devices and how they change as a result of charge transfer
reactions.
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci.
,2024,
15
,5814
5831 |
5815
Perspective
Chemical Science
and occurs at a known, constant potential. The oxidation of the
metal anode is su
ffi
cient to charge balance the reductive reac-
tion of interest, thereby preventing unwanted oxidation of
substrates or additives in the solution. The addition of metal
cations into solution over the course of the reaction is less likely
to limit the scope of the reductive reaction than the use of other
chemicals as sacri
cial reductants.
21
Furthermore, metal elec-
trodes are generally inexpensive and easy to store, and they
provide a less hazardous alternative to commonly used sacri
-
cial reductants.
30
The ideal sacri
cial anode should not limit or interfere with
the reductive reaction of interest, either through direct inter-
actions with reaction components or through its electro-
chemical performance. In particular, four major criteria must
be satis
ed to ensure that the reaction is not limited by the
sacri
cial anode:
(1) Both the metal anode and the cations generated during
electrolysis should not degrade any electrolyte components or
reagents used in the reductive reaction.
(2) Any inherent reactivity between the metal and the elec-
trolyte solution should not form an insulating surface
lm or
prohibit oxidation of the anode.
(3) The anode should permit metal stripping throughout the
reaction, meaning that it must not be passivated by products or
byproducts formed over the course of the electrolysis.
(4) Metal cations generated from anodic oxidation should
not undergo competitive reduction at the cathode.
These four criteria are outlined in Fig. 2 and serve as a basis
for the outline of our discussion.
Though many reductive electrosynthetic reactions assume
that each of these sacri
cial anode criteria are met, in reality
several processes can interfere with these assumptions and
prevent ideal sacri
cial anode behavior. The surface chemistry
at metal electrodes is extremely sensitive to the chemicals
present in the reaction solution, a fact that has been well
studied in the analytical electrochemistry community. In addi-
tion to the electrode materials, the success of an electro-
synthetic reaction depends on a host of factors that must be
optimized, including chemical parameters such as the solvent,
reactants, and additives, as well as electrochemical parameters
including the supporting electrolyte and the magnitude of the
applied potential or current.
26,27,31
36
Changes to any of these
parameters can induce nonideal behavior at the sacri
cial
anode, leading to failed reactions, low yields, hazardous short-
circuits, and/or extreme voltages that exceed the compliance
limits of the potentiostat and stop the reaction early. Thus, the
performance of the anode can impose limits on the available
reaction conditions, eliminating chemical space that would
otherwise be compatible with the reaction of interest.
The issues associated with sacri
cial anode performance are
compounded by a lack of reported observations of the anodes
themselves.
37,38
Many electrosynthetic works simply select the
sacri
cial anode that provides the highest yield in optimization
experiments without investigating the anodic chemistry, even
when a failure at the anode is observed. Not only does this
strategy limit the range of compatible conditions for the
reductive reaction, it also ignores potential mechanistic
contributions from the cations generated during sacri
cial
anode oxidation. Fortunately, problems at the anode can o
en
be diagnosed and recti
ed with a few brief experiments, leading
to higher product yields and an expanded set of compatible
reaction conditions.
Understanding and characterizing processes at the anode
requires knowledge of both electrochemical and surface char-
acterization techniques that can interrogate the chemistry at
the electrode. Here, we integrate lessons from electrochemistry,
interface characterization, and organic synthesis to elucidate
strategies for overcoming issues related to sacri
cial anodes.
We highlight four common but underexplored challenges that
cause reactions to fail. For each, we suggest electrochemical and
surface characterization experiments to diagnose the problem,
and we present experiment design strategies for trouble-
shooting sacri
cial anodes. The fundamentals of electro-
chemical techniques in organic synthesis, as well as classes of
synthetic reactions enabled by sacri
cial anodes, have been
reviewed extensively elsewhere and will not be covered
here.
15
17,24
27,39
42
We conclude with an outlook for the future of
this multidisciplinary
eld, including open challenges related
to sacri
cial anode optimization and potential new directions
for developing reductive half reactions with minimal interfer-
ence from anodic processes. We hope that the strategies pre-
sented here will improve the process of screening and
optimizing electrosynthetic reactions, expand the chemical
space in which reductive reactions can occur, enable more
Fig. 2
Sacri
fi
cial anodes can deviate from ideal behavior and limit the
reductive reaction of interest, even under conditions that would nor-
mally be compatible with said reaction. Chemical reactions between
the electrode and substrate, formation of an insulating surface
fi
lm on
the anode, anode passivation by byproducts formed during the
reductive reaction, and competitive plating of cations from the sacri-
fi
cial anode at the cathode can all prevent ideal anode performance.
5816
|
Chem. Sci.
,2024,
15
,5814
5831
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chemical Science
Perspective
robust yields from reductive electrosyntheses, and encourage
collaborations at the intersection of electrochemistry, organic
synthesis, and interface chemistry.
2. Side reactions
As in all synthetic procedures, side reactions between reagents
must be considered when optimizing a reaction. In the case of
electrosynthesis, care must be taken to avoid side reactions
between components of the electrochemical system as well as
between the substrates and additives common to traditional
organic synthesis. In particular, reactions between any combi-
nation of the electrodes, solvent, supporting electrolyte,
substrates, and intermediates must be considered. Even if not
directly related to the reductive chemistry, chemical reactions
between the sacri
cial anode and any other component of the
electrolyte solution can lead to low yields of the desired product.
2.1 Examples of side reactions with the anode
In many cases, reactions that occur at the sacri
cial anode are
chemical reductions of a reactant by a strongly electropositive
anode, such as Mg. The canonical example of such a detri-
mental side reaction is the reaction between Mg metal and
organic halides (R
X) to form Grignard reagents, RMgX, and
related compounds in solution.
43,44
Given that organic halides
are frequently used as substrates for reductive electrosynthesis,
such side reactions are likely problematic during the optimi-
zation stage of many methodologies. It should be noted that the
formation of most Grignard reagents with non-activated Mg
requires an induction period, in which the native MgO surface
layer is removed from the Mg source to expose reactive Mg, but
the length of this induction period depends on the other
compounds in solution as well as any pretreatment steps, such
as polishing the Mg metal. In electrochemical systems, the
anode is usually chemically or mechanically polished to ensure
e
ff
ective Mg stripping, but the solution composition will be
reaction-dependent. In addition to a
ff
ecting the substrates and
mechanistic pathway, the formation of Grignard reagents can
contribute to the growth of a high-impedance passivation layer
at the Mg anode,
45,46
which will be discussed in more detail in
Section 4. In all cases, the electrolyte should be checked for
potential Grignard-forming conditions before a Mg sacri
cial
anode is employed.
The formation of Grignard reagents is not the only undesired
side reaction that can occur at sacri
cial anodes. In addition to
alkyl halides, Mg can also react with esters and ketones. Condon
et al.
observed that in an attempted alkylation of decyl tri-
chloroacetate with triethylborane, the ester group was reduced
when a Mg sacri
cial anode was used, leading to the formation
of decanol as an undesired side product. The researchers noted
that when a Mg or Zn sacri
cial anode was employed, the
electrolysis time was shortened from the expected 1 h to 30 min,
suggesting that some chemical reduction of the substrate
occurred at the anode. No chemical reduction was observed
when an Al or Fe anode was employed.
47
In the electrochemical
allylation of carbonyl compounds, Durandetti
et al.
observed
side reactions between carbonyl species and a Mg sacri
cial
anode resulting from enolization of the ketone substrate. The
authors attribute this reactivity to the reducing power of Mg;
similar side reactions were not observed when a Zn anode was
used in place of Mg.
48
Mg can chemically reduce a variety of
substrates, including activated alkenes, pyridine derivatives,
and cyanoarenes (Fig. 3), which could lead to undesired
byproducts in an electrochemically driven reaction.
49
52
Zn can
chemically reduce carbonyl halides, potentially causing similar
issues.
53
Substrates that can be chemically reduced by a mildly
reducing metal can o
en also be reduced by more strongly
reducing metals, necessitating the careful choice of sacri
cial
anodes in an electroreductive reaction. The choice of solvent
can introduce potential side reactions as well. Saboureau
et al.
observed overconsumption of a Mg sacri
cial anode during
electrolyses carried out in dimethylformamide (DMF) solvent,
which was traced back to chemical corrosion of the Mg anode
via
reductive decomposition of the solvent. Zn and Al sacri
cial
anodes were not subject to the same corrosion reaction.
54
Furthermore, Zn sacri
cial anodes can chemically reduce Ni
Fig. 3
Examples of substrates and catalysts that can be chemically reduced by various metal sacri
fi
cial anodes.
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci.
,2024,
15
,5814
5831 |
5817
Perspective
Chemical Science
and Pd catalysts for reductive coupling reactions, possibly
complicating reaction optimization and mechanistic
studies.
55,56
2.2 Examples of side reactions with anodically generated
cations
In addition to side reactions between the metal anode and
reaction components in solution, the cations generated during
oxidation of the anode can also interfere with the reductive
reaction. Nedelec
et al.
demonstrated that the nature of the
cation generated from the sacri
cial anode dictates the extent of
electrochemically driven cross-coupling products of alkyl
halides.
35
Introduction of metal cations and salts
via
oxidation
of the anode can also lead to diminished yields. For example,
Peters
et al.
observed that anode-derived Mg salts limited the
success of an electrochemical Birch reaction. Addition of Mg
salts directly to the electrolyte reduced yields of the desired
diene from 74% to 30%. In cases when the solution was not
stirred during electrolysis, smaller anode
cathode distances
were correlated with lower yields, suggesting that metal salts
generated at the anode negatively impacted yields.
57
During
electrochemical atom transfer radical polymerization (ATRP)
reactions catalyzed by Cu
amine complexes, anodically gener-
ated Al
3+
can hinder the catalyst's reactivity, providing
a competitive coordination pathway for the amine ligands, and
the Al sacri
cial anode can chemically reduce the Cu catalyst to
Cu(0).
58
60
2.3 Symptoms of side reactions with the anode
As detailed above, there are several indicators of side reactions
between the sacri
cial anode and another component of the
system. Similarly to traditional chemical reactions, high
conversion of the starting material but low yields of the desired
product serve as key markers of background reactivity. Side
products may appear in
1
H-NMR spectra of the post-electrolysis
solution, either as discrete and identi
able species or as an
intractable mixture formed from decomposition of the
substrate, additives, or solvent. In certain cases, degradation of
the electrolyte solution may be visible through a color change or
the formation of precipitates.
2.4 Diagnosing the problem
A few experiments can be undertaken to determine whether the
formation of an undesired side product is related to the sacri
cial
anode. Control experiments without electricity, such as placing
the anode in the electrolyte solution without applying a current or
potential bias, can indicate whether chemical reactions between
the anode and electrolyte are occurring. Perhaps the simplest
electrochemical experiment is to exchange the anode for
adi
ff
erent metal. A less reducing metal, such as Zn, is less likely
to undergo detrimental side reactions with electrolyte compo-
nents, but the nature of any observed side reactions will be
solution-dependent and must be optimized as such. However, if
a speci
csacri
cial anode must be used, or if an understanding
of the side reaction is important to mechanistic development,
further electrochemical characterization can be carried out. A
divided cell is an excellent tool that can be employed to decouple
any observed reductive decomposition from reactions occurring
at the cathode. Saboureau
et al.
adopted this approach when
determining the nature of the Mg corrosion reaction in DMF
during the electrosynthesis of carboxylic acids from organic
halides. A Mg anode was
tted to the anodic compartment, which
was then charged with an organic halide (4-chloro-
tri
uoromethylbenzene) and several supporting electrolytes in
succession. The cathodic reaction was the straightforward elec-
trochemical reduction of 1,2-dibromoethane, an easily reducible
species that serves as an e
ffi
cient counter reaction. Analysis of the
anodic compartment a
er electrolysis indicated that the
aromatic halide was not consumed at the anode and that only
DMF had degraded.
54
This set of divided cell experiments
conclusively demonstrated the nature of the side reaction
between anodically polarized Mg and DMF.
2.5 Troubleshooting the problem
First, we discuss solutions to the anode chemically reducing
components in the electrolyte. A simple solution would be to
use a less reducing metal; however, another metal could intro-
duce new side reactions or di
ff
erent challenges. Alternatively,
a divided cell can be used to carry out the electrolysis itself, not
just in a diagnostic role, to prevent crossover from the cathodic
chamber into the anodic chamber. A more complicated but
exciting solution would be to generate a surface
lm on the
anode metal that conducts the corresponding metal cation but
is electronically insulating, preventing the metal from reacting
directly with the organic substrate
via
electron transfer. The
lm
would be akin to the solid-electrolyte interphase (SEI) formed
on anodes in Li-ion batteries; however, such a strategy has not
been pursued in the context of sacri
cial anodes and will thus
be discussed in the Summary and future outlook section.
Next, we discuss issues of reactivity between the metal cation
generated at the anode and components in the solution. A
divided cell can again be used to prevent the generated cation
from reacting with anything in the reaction mixture. If an
undivided cell is required, a strongly coordinating binding
agent could sequester the generated cation, provided that the
bound complex does not react further with the electrolyte or
passivate the anode.
It is important to acknowledge that not all reactions between
the sacri
cial anodes and the electrolyte solution are delete-
rious. The cations and salts produced by oxidation of the anode
can participate in the reductive reaction, stabilize products
formed at the cathode, or function as products in their own
right. For example, Mellah and coworkers have explored the use
of a samarium sacri
cial anode to directly generate Sm(
II
)
reagents for C
C bond formation in solution.
61
63
The Sm metal
anode is oxidized to Sm
2+
, and various Sm
2+
salts including
SmCl
2
, SmBr
2
, SmI
2
, and Sm(OTf)
2
(OTf
=
tri
ate, CF
3
SO
3
) were
formed
in situ
through the addition of
n
Bu
4
X salts.
63
Cations or
salts formed from oxidation of the anode can also function as
in
situ
-generated reactants participating in the reduction reaction.
Lu
et al.
demonstrated low product yields from an electro-
reductive radical silylation reaction when a divided cell was
5818
|
Chem. Sci.
,2024,
15
,5814
5831
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chemical Science
Perspective
used, in part because anodically generated Mg
2+
is proposed to
participate in the overall reaction mechanism.
64
Manabe
et al.
also demonstrated such reactivity in the electrochemical
reduction of triphenylphosphine oxide to triphenylphosphine.
AlCl
3
, generated
via
oxidation of an Al sacri
cial anode, facili-
tated selective cleavage of the P
O bond.
65
Gosmini
et al.
demonstrated that the presence of Fe(
II
) salts, formed from pre-
electrolysis of a sacri
cial Fe anode, was necessary to enable
electrochemical cross-coupling of aryl halides and 2-chlor-
opyrimidine; the same is true for the electroreductive coupling
of aldehydes or ketones with gem-dichloro complexes.
66,67
Fe(
II
)
salts were also found to facilitate the electrochemical cyclo-
propanation reaction of halogenated compounds and activated
ole
ns.
68
Even if the generated cations do not participate
directly in the reaction, they can stabilize products or species of
interest formed during the reductive reaction. For example,
Mg
2+
and Al
3+
cations coordinate to carboxylate anions formed
as the products of electrocarboxylation reactions.
25,69,70
This
coordination stabilizes the carboxylate and forms a precipitate,
which can easily be extracted from the organic solution. Simi-
larly, Zn
2+
cations can stabilize intermediates formed during
the reduction of quinolines.
71
Furthermore, metal anodes can
chemically reduce species in solution to form reagents that are
otherwise di
ffi
cult to access. For example, Hilt and Smolko
observed the chemical reduction of In(
III
) to In(0) at an Al anode.
The In(0) was then oxidized at the anode to form In(
I
), a catalyst
for the allylation of aldehydes, ketones, and esters.
72
The metal
anode can also in
uence the selectivity of the reaction. In an
electrochemical thiolation reaction
via
cross-electrophile
coupling of alkyl bromides with functionalized thiosulfonates,
Ang
et al.
determined that the cross-coupling reaction only
proceeds when a Mg sacri
cial anode is used. Attempts to run
the reaction with a Zn, Fe, or Cu anode resulted in homocou-
pling of the thiosulfonate to form diphenyldisul
de as the sole
product a
er 3 h electrolysis.
73
3. Anode passivation by inherent
metal reactivity
Reactions between the sacri
cial anode and reaction compo-
nents can a
ff
ect the electrochemical behavior of the anode
itself, in addition to altering the electrolyte solution. Decom-
position of certain electrolyte components can form ionically
insulating surface layers on the sacri
cial anode, prohibiting
further contact between the electrolyte and the metal and pre-
venting oxidation of the anode. Many anode materials
particularly electropositive metals such as Mg and Al
contain
native oxide layers that similarly limit e
ffi
cient metal oxidation.
3.1 Examples of passivation by native oxides
The native surface oxide layers that form on Mg and Al develop
before exposure to the electrolyte solution. While the oxide layer
can be bene
cial in preventing undesired chemical reactivity
between the anode and substrate before electrolysis, in many
cases the surface
lm is ionically insulating and does not permit
oxidation or dissolution of the anode material.
74
79
Solid-state
conductivity of multivalent cations is very di
ffi
cult due to their
high charge density and large size.
80
Once the oxides are
formed, they are very thermodynamically stable, which is
obvious from their position on the Ellingham diagram. In fact,
MgO and Al
2
O
3
are two of the most thermodynamically stable
oxides relative to the corresponding metal.
81
Mg is so oxophilic
that even if it is sputtered in ultra-high vacuum, the surface is
still covered by MgO.
82
The surface
lm thus limits the elec-
trochemical reaction by preventing the anodic stripping from
charge-balancing the cathodic reaction at low overpotentials.
The oxide layer forms whenever Mg or Al is in contact with air or
moisture.
74,83
As such, a rigorous electrode polishing procedure
is required to remove the passivating oxide
lm. Polishing the
anode has the added bene
t of removing any impurities or
oxidized products remaining on the anode from previous
reactions. If the anode has been used in prior reactions,
macroscale deposits of oxidized product can be removed
via
sonication, electropolishing, or an acid rinse.
57,84
The electrode
should then be mechanically polished under an inert atmo-
sphere using a razor blade,
ne-grit sandpaper, or a rotary tool
to minimize growth of the oxide surface
lm.
Even if the sacri
cial anode is rigorously polished, electro-
positive metals can still form an insulating surface
lm when
placed in contact with organic electrolytes. Even with rigorous
drying procedures, electrolytes can still contain ppb to ppm
amounts of trace water that will react with the metal. Addition-
ally, solvent decomposition or reactions between the anode and
the supporting electrolyte can form insulating surface
lms that
prevent oxidation of the anode.
64,65,84,85
Importantly, these insu-
lating surface
lms may not be immediately apparent or visible to
the naked eye. Even an electrode that appears shiny and metallic
with no obvious corrosion could experience di
ffi
culties with
anodic oxidation, preventing the cathodic reaction from
proceeding to completion and limiting reaction scale-up.
65,86
3.2 Examples of passivation by supporting electrolyte
Reactions between the anode material and the supporting
electrolyte are particularly insidious as these are two compo-
nents that are optimized independently but are in fact code-
pendent. Selecting either an anode material or a supporting
electrolyte too early in the optimization process can lead to
inadvertent exclusion of compatible reaction conditions if the
only factor limiting the cathodic reaction is the non-obvious
evolution of an insulating surface
lm on the anode. Fortu-
nately, the formation of such surface
lms has been studied
extensively in the battery community. Several supporting elec-
trolytes commonly used in organic electrosynthesis have
already been screened for metal deposition and stripping for
use in battery applications, and many are found to be incom-
patible with Mg and Al stripping.
65,87
89
In particular,
uoride-
containing electrolytes
including those containing BF
4
,
PF
6
, AsF
6
, or others with anions that can hydrolyze in the
presence of trace water to form HF
have been shown to form
ionically insulating MF
x
-rich surface
lms on Mg and Al.
65,87
Other electrolytes, including those containing CF
3
SO
3
,
(CF
3
SO
2
)
2
N
(bistri
imide or TFSI), or ClO
4
, can also interfere
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci.
,2024,
15
,5814
5831 |
5819
Perspective
Chemical Science
with Mg anodes.
87
89
Using supporting electrolytes containing
any of the aforementioned anions with a Mg or Al sacri
cial
anode is likely to lead to passivation of the anode and unsat-
isfactory electrochemical performance.
3.3 Symptoms of passivation by inherent reactivity
Several physical and electrochemical signatures could indicate
formation of an insulating surface
lm on the anode. During an
electrolysis experiment, a sharp increase in the overall cell
voltage might be observed, especially at the start of the reaction.
Depending on the instrument used, the voltage may increase
until the compliance limit of the device is reached, causing the
reaction to stop prematurely. In many cases, the insulating
surface
lm is not readily apparent and no visible electrode
fouling is observed because the surface
lms can be very thin.
3.4 Diagnosing the problem
The
rst step in determining if a reaction fails due to the
formation of an insulating surface
lm on the anode is to search
the literature for prior reports of reactivity between the anode
and the solvent or supporting electrolyte. For Mg anodes in
particular, research on SEIs in Mg metal batteries is a good
starting point for exploring inherent reactivity between the
anode and electrolyte.
87,89,90
If no prior reports of passivation
exist, several electrochemical experiments can point to the
existence of an ionically insulating surface
lm. Linear sweep
voltammetry (LSV), in which the sacri
cial anode is the working
electrode and the potential is swept anodically, can indicate
whether metal stripping is observed at the expected potential.
Fig. 4a shows an example of an LSV experiment designed to
assess metal stripping at an Al electrode.
91
When a TBABF
4
supporting electrolyte in tetrahydrofuran (THF) is used, the
oxidative current density is extremely low and no Al oxidation is
observed due to the formation of a passivating
lm resulting
from a chemical reaction between Al and the electrolyte. The
current density also decreases upon subsequent scans, indi-
cating that Al stripping becomes more di
ffi
cult due to the
growth of the passivating
lm. It should be noted that a sepa-
rate reference electrode is required for such experiments to
eliminate confounding e
ff
ects that may arise from the possible
passivation of the counter electrode in a two-electrode cell.
Several reference electrodes suitable for use in nonaqueous
systems have been developed.
92
95
The potential at the anode during a constant current elec-
trolysis can also be monitored to con
rm that the high voltage
observed in an electrolysis experiment is related to processes at
the anode and not the cathode. Fig. 4b shows the potentials of
an Al anode and a graphite cathode during reduction of
t
BuBr.
The potential at the cathode is steady and constant, suggesting
that the reduction proceeds as expected. However, the voltage at
the Al anode increases to >10 V within a few seconds and quickly
reaches the compliance limit of the potentiostat. This sharp
polarization and extreme overpotential at the anode indicate
that the oxidation reaction does not proceed smoothly, likely
due to the presence of a passivating surface
lm.
91
Monitoring
both the cathode and anode voltage during electrolysis requires
the use of a nonaqueous reference electrode and a potentiostat
equipped with the hardware necessary to record both the
working and counter electrode potentials.
Electrochemical impedance spectroscopy (EIS) may indicate
changes in the anode surface resistance over the course of
a reaction. To diagnose an ionically insulating surface
lm, an
EIS experiment in a two-electrode cell using the electrolyte of
interest and with both electrodes made from the sacri
cial
anode metal will provide the resistance associated with
oxidizing the anode; a notably high resistance (more than a few
k
U
) suggests that an insulating surface
lm has formed, pre-
venting metal stripping and limiting the anode's performance
in the reductive reaction.
96
Note, however, that a low resistance
does not necessarily mean that no surface
lm has formed, as
an electronically conductive but ionically insulating surface
lm
can exhibit low impedance.
If the nature of the reaction that causes surface passivation is
not known, various surface characterization techniques can
pinpoint the electrolyte component that reacts with the anode.
Scanning electron microscopy (SEM) and energy-dispersive X-
ray spectroscopy (EDS) provide information about the
morphology of the anode a
er reaction and a spatially resolved
elemental distribution map, respectively. These techniques
require high vacuum and thus will only probe solid products at
the surface. Therefore, if elements exclusively present in the
electrolyte are observed, then we can assume the electrolyte has
reacted with the metal.
97
For a more in-depth understanding of
the surface reaction, X-ray photoelectron spectroscopy (XPS)
provides information about the chemical environment of each
element within the top 5
10 nm of the surface
lm.
91,98
XPS is
useful for determining the identity of the surface species and for
characterizing extremely thin surface
lms, such as the oxide
layers formed when Mg and Al are exposed to oxidants.
3.5 Troubleshooting the problem
Research in the
elds of corrosion science and energy storage
have demonstrated that passivating metal oxide layers on Mg or
Fig. 4
(a) Linear sweep voltammograms of an Al sacri
fi
cial anode in
THF with 0.1 M TBABF
4
supporting electrolyte. The voltammograms
were collected at 5 mV s
1
scan rate with 85%
iR
compensation. (b)
Voltage pro
fi
les of Al and graphite electrodes during galvanostatic Al
stripping in THF with 0.5 M TBABF
4
supporting electrolyte. The Al
stripping experiment was conducted with
t
BuBr as a sacri
fi
cial
reductant. Adapted from ref. 91 with permission from the Royal Society
of Chemistry.
5820
|
Chem. Sci.
,2024,
15
,5814
5831
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chemical Science
Perspective
Al can be removed by adding halide salts to the electrolyte
solution. In an oxidizing environment, chloride ions migrate
through the oxide
lm that forms on pure Al metal, breaking
down the oxide layer and forming
pits
in the metal; corrosion
by such pitting reactions has been well studied and is not
limited to chlorides.
99
104
Chloride ions have also been shown to
break down non-oxide insulating surface
lms on Al. Manabe
et al.
demonstrated that an insulating AlF
3
surface
lm which
formed when Al was in contact with PF
6
-containing electrolyte
could be removed by introducing TBACl as a co-supporting
electrolyte, along with the chelating amine tetramethylethyle-
nediamine (TMEDA) to promote Al stripping.
65
Bromide additives can also break down the insulating
surface
lm on Al anodes. Zhang
et al.
observed that adding
TBABr as a co-supporting electrolyte enabled Al oxidation in
electrolytes that would otherwise passivate the metal (Fig. 5a).
The current densities observed in the LSV with TBABr co-
supporting electrolyte are much higher than those observed
without TBABr (shown in Fig. 4a). The current density also
steadily increases upon subsequent scans, suggesting that
surface passivation does not limit Al stripping. In a follow-up
electrolysis experiment, the Al anode potential remained low
and constant for several hours (Fig. 5b), suggesting that Br
contributes to the formation of an ionically conductive surface
lm on Al. The bene
cial e
ff
ect of additives is not limited to
bromides in this case; Cl
-, Br
-, and I
-containing additives all
enabled Al oxidation.
91
Similar strategies can be applied in systems with Mg sacri-
cial anodes. Addition of Br
decreased the thickness of the
passivation layer formed on a Mg anode and formed an ionically
conducting surface
lm that permitted Mg stripping while
limiting chemical reactions between the Mg anode and the
electrolyte (see Section 4).
97
In a Mg battery system, addition of
MgCl
2
to a solution of Mg(TFSI)
2
in dimethoxyethane (DME)
resulted in signi
cant improvements in Mg oxidation compared
to the solution without MgCl
2
, likely due to destabilization of
the surface oxide
lm.
79
MgCl
2
was later shown to suppress the
passivation of Mg metal by PF
6
ions.
85
Li
et al.
showed that, as
an alternative to Cl-containing salts, small amounts of I
2
could
be added to an electrolyte consisting of Mg(TFSI)
2
and DME to
form an ionically conductive MgI
2
surface
lm that mitigated
the Mg oxidation overpotential.
105
If no satisfactory means of suppressing the formation of an
insulating surface
lm on Mg or Al can be found, less oxophilic
anodes such as Zn may be more applicable to the system at
hand. Zn anodes are chemically compatible with a wide range of
solvents and supporting electrolytes and are less likely to exhibit
high stripping overpotentials due to passivation
via
inherent
reactivity.
106
108
4. Passivation by products formed
during anodic stripping
Insulating surface
lms can also form due to processes that
occur during the electrolysis in addition to or
in lieu
of the
lms
that form immediately upon contact with the electrolyte, as
described in the previous section. During stripping, fresh metal
surface is exposed to the electrolyte and the corresponding
metal cation is generated, ideally in solution. The fresh metal
surface can react with the electrolyte components di
ff
erently
than the original metal surface because the surface layers
described in the previous section can be anodically destroyed.
The newly exposed metal can react with the electrolyte to form
a new surface
lm composed of decomposition products from
the supporting electrolyte, solvent, or organic substrate.
109,110
Further, though stripping produces metal cations that ideally
dissolve into the electrolyte, the metal cations can react at the
anode/electrolyte interface to form insulating deposits.
4.1 Examples of passivation by insulating salt nucleation
Passivation of the sacri
cial anode during electrolysis has been
observed in several synthetic reactions. Here we de
ne passiv-
ation as the evolution of a high impedance surface
lm that
shuts down electrochemistry at that electrode. In many cases,
this passivation is linked to the use of strongly reducing anodes
like Mg or Al.
First we focus on examples using Mg anodes. While
attempting to scale up reactions involving electrochemical
cross-electrophile coupling of alkyl halides, Zhang
et al.
initially
observed high cell voltage accompanied by visible formation of
a thick passivating
lm during the
rst few hours of electrol-
ysis.
98
Lu
et al.
observed similar passivation of the Mg sacri
cial
anode in an electrochemically driven three-component cross-
electrophile coupling reaction; though conversion of the start-
ing material exceeded 95%, the desired product was obtained in
low yields due to the growth of a thick passivating
lm at the
anode and accompanying high cell voltage.
111
Anode passiv-
ation can also be in
uenced by seemingly unrelated compo-
nents. In the electroreduction of epoxides, Huang
et al.
employed tripyrrolidinophosphoric acid triamide (TPPA) as
a cosolvent to improve the solubility of the LiCl supporting
electrolyte in THF and to prevent cathodic reduction of the
sacri
cial metal cations. When the electroreduction was run in
the absence of TPPA, the Mg sacri
cial anode was coated in
Fig. 5
(a) Linear sweep voltammograms of an Al sacri
fi
cial anode in
THF with 0.05 M TBABF
4
+ 0.05 M TBABr supporting electrolyte. The
voltammograms were collected at 5 mV s
1
scan rate with 85%
iR
compensation. (b) Voltage pro
fi
les of Al and graphite electrodes during
galvanostatic Al stripping in THF with 0.25 M TBABF
4
+ 0.25 M TBABr
supporting electrolyte. The Al stripping experiment was conducted
with
t
BuBr as a sacri
fi
cial reductant. Adapted from ref. 91 with
permission from the Royal Society of Chemistry.
© 2024 The Author(s). Published by the Royal Society of Chemistry
Chem. Sci.
,2024,
15
,5814
5831 |
5821
Perspective
Chemical Science