Li 2021

Review
www.advmattechnol.de
Design of Electrochemical Microfluidic Detectors: A Review

Tianyu Li, Jesús Adrián Díaz-Real, and Thomas Holm*
microfluidic devices have become an inte-

Miniaturized electrochemical systems integrated into microfluidic flow gral part of many scientific studies.
devices have been an active research field the last couple of decades. This For electrochemical studies, electrode
has resulted in many advanced microfluidic platforms for the conversion and integration with the microfluidic flow is
an attractive feature. The advantage of
detection of chemicals. The development is partly driven by the increased
controlled mass-transport and reaction
access to and decreased cost of fabrication. In the design of microfluidic elec- environment through electrochemistry
trochemical cells, several aspects need to be handled, such as the accurate has been utilized when making micro-
control of potential, electrolyte flow, pH, pressure, and temperature. This can fluidic fuel cells.[8–10] Similarly, the small
be further complicated by integrating photon traffic for spectro(electro) analyte volume necessary, combined with
high sensitivity and selectivity have made
chemical detection. Here, a comprehensive review of recent advances and
microfluidic electrochemical sensors an
approaches to the design of microfluidic flow devices for detection is pre- attractive application.[11] Many of the spe-
sented, first dealing with the design of electrodes and flow pathways, then cific applications have been summarized
highlighting some common challenges and how these have been dealt with in recent reviews, for example, microflu-
in the literature. This work aims to be a guide for researchers in the design of idic flow cells,[12–14] CO2 electroreduction
microfluidic electrochemical systems for detection of chemical species. flow cells,[15] microfluidic biosensors,[4,16,17]
and nanofluidic sensors and proces-
sors.[18–20] While attractive, the combina-
tion of several simultaneous processes
1. Introduction puts a large demand on the experimenter, where an under-
standing of the flow pattern, current and potential distribution,
Microfluidic cells are a type of device that operate with a small and the reaction environment is necessary. As a consequence,
volume of fluid and nominally has at least one dimension on microfluidic devices require optimization for specific applica-
the micrometer scale.[1,2] Due to the small scale and accurate tions. This means that it is hard to extract general trends, and it
control of experimental conditions, these have found applica- is often not a good advice for researchers to directly use designs
tion in the field of, for example, medicine, pharmaceuticals, and from other systems. For example, a microfluidic fuel cell will be
sensors.[3–5] Besides, the small size resulting in a short optical optimized based on low ohmic losses over the channel, mini-
pathway and the readily maintained homogeneous reaction mized mixing of reactants between the anode and cathode, and
environment has made it possible to integrate optical detectors high energy output.[21] In contrast, a sensor will work at sev-
that access the solution volume or electrodes, that is, in situ or eral orders of magnitude smaller currents and be optimized for
in-operando characterization of the reaction environment.[6,7] high detection limit and selectivity.[22]
These advantages have led to increased academic interest, and Microfluidic electrochemical detection involves the use of
an analyte, that is, a solution containing the target species to
be quantified, and the quality of the method is assessed in
T. Li terms of sensitivity, lowest concentration that can be quanti-
Department of Mechanical and Industrial Engineering
University of Toronto fied, and selectivity toward the target species over other species
Toronto, ON M5S 3G8, Canada in the solution. The advantage of microfluidic methods is often
T. Li the portability and ease of use, and not necessarily the sensi-
Department of Chemistry tivity, for example, in the biomedical field, the PCR method
University of Victoria represents the gold standard, while a microfluidic detector is
Victoria, BC V8W 3V6, Canada
rather a faster and less accurate method that can be applied in
J. A. Díaz-Real
Centro de Investigación y Desarrollo Tecnológico en Electroquímica
the field (analogous to the PCR vs the rapid antigen test used
S. C., Parque Tecnológico Sanfandila, Querétaro C.P. for Covid-19 detection). In a similar manner, for (metal) ion
Pedro Escobedo 76703, Mexico detection in the solution, a microfluidic method can usually
T. Holm not reach the detection limit of, for example, ICP-MS, but can
Institute for Energy Technology rather be used as a rapid and cheap screening tool for large
P.O. Box 40, Kjeller NO-2027, Norway amount of samples before a more precise method is required
E-mail: thomas.holm@ife.no
for accurate quantification of a sub-selection of the samples.
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/admt.202100569.
As such, while improved sensitivity is always a goal, the
cost aspect should always be kept in mind when looking at
DOI: 10.1002/admt.202100569 commercialization paths.
Adv. Mater. Technol. 2021, 2100569 2100569 (1 of 19) © 2021 Wiley-VCH GmbH
www.advancedsciencenews.com www.advmattechnol.de
This review focuses on the design of microfluidic electro- to adsorb enough of the target species for analysis. The trans-
chemical devices for the purpose of detection and aims to be a position of ultramicroelectrodes in microfluidic chip has been
guide for the readers in the design process. First, examples from explored by Horny et al., by using a thiol-labeled DNA probe (P)
the literature for placement of electrodes, the configuration of and complementary target (T)[32]. The target DNA used mimics
flow in the device, design and optimization using numerical the fragment of interest miRNA-122 for the diagnosis of liver
modeling, and manufacturing methods are presented. Then, cells in case of injury (hepatitis, alcoholism, obesity). Several
some common challenges, such as the integration of photo- electrochemical microfluidic devices have been assembled for
spectroscopic methods, and the control of operational para detection of complimentary synthetic DNA,[33,34] human sam-
meters are highlighted, and examples from the literature of how ples of human leukocyte antigen,[35] salmonella,[36] GSTP1 and
these are dealt with are discussed. Finally, a design guideline is EFEMP1 spiked in urine,[37] among other targets. Other studies
presented to help the readers in the design of new microfluidic have used enzyme immobilization for detection of biomarkers,
electrochemical cells for detection purposes. like serotonin, to be detected by monoamine oxidase-A enzyme
immobilized on multi-walled carbon nanotubes (MWCNT).[38]
The size and shape of the electrodes is also an important var-
2. Design of Microfluidic Devices iable, and in general, small electrodes are preferred to reduce
the system size. An important consideration in regard to the
An electrochemical microfluidic system for sensing purposes geometry is the detection method. If an amperometric method
consists of one or more working electrodes, counter electrodes is used, the highest current will be near one of the edges of
and reference electrodes. In addition, the microfluidic system the electrode. If deposition of a species is the purpose, the elec-
also has flow channels that transport the analyte or electrolyte trodes can be designed to largely take advantage of edge effects,
to the electrodes. In this section, the electrodes and flow chan- like a microelectrode with close to hemispherical diffusion.
nels are first discussed separately. Then, fabrication and the use Here, faster diffusion rates are obtained due to the electrode
of numerical modeling are discussed. size. However, the overall signal is small, and often, larger
electrodes are preferred where the response per area is more
averaged over the whole surface. Numerical modeling is par-
2.1. Working Electrodes ticularly suitable to deal with such problems, and more details
are given in Section 2.6.
The working electrode is the electrode that does the sensing The electrodes can be manufactured through several routes
of the target species. In electrochemical systems, this involves depending on the type of detection and available manufac-
either amperometric, potentiometric, or impedimetric turing methods. Some of the main ones are photolithog-
sensing.[23,24] Amperometric sensors measures the current as a raphy, 3D printing, computerized numerical control (CNC)
function of a potential or a potential program. Examples here machining, and injection molding. Combined with these fab-
are chronoamperometry, cyclic voltammetry, and various pulse rication methods, the electrode materials can be made through
voltammetry methods such as square-wave voltammetry, stair- among other physical vapour deposition (PVD), electrodeposi-
case voltammetry, and differential pulse voltammetry. This type tion, and wet impregnation methods. The choice of fabrication
of sensing is the most common for quantification of species in route largely depends on what type of electrode material is used
an analyte. Potentiometric sensors instead measure the poten- for detection and the resolution necessary for the fabrication of
tial while the current is controlled on the electrode. A target the device. For example, photolithography gives higher resolu-
molecule can be detected by the influence it has on the poten- tion of less than 1 μm,[44] and down to 10s of nm in some cases,
tial of the electrode. For example, a very sensitive DNA sensor while CNC machining is at about 10 μm and 3D printing is at
and sensors to measure the pH and redox potential has been about 100 μm are at least 1 and 2 orders of magnitude higher
made in this way. Impedimetric sensors use electrochemical than photolithography, but cheaper and more available. The
impedance spectroscopy or other ac techniques to detect spe- fabrication methods are dealt with in more detail in Section 3.
cies. This can work in several ways, for example, through the
detection of diffusion as a function of flow rate[25] or the meas-
urement of the electrode capacitance as a measure of target 2.2. Electrode Arrays
concentration.[26]
In addition to the consideration of detection method, the When designing a microfluidic electrochemical cell, there is
geometry of the electrode must be carefully considered. The little hindrance to incorporating several working and counter
working electrode can be either 2D or 3D, it can be smaller electrodes in the system. This allows for more application of
or larger, and it can have several shapes, for example, round the system. Firstly, one can use several electrodes to run spe-
or straight like a channel electrode. The simplest electrode to cialized electrochemical sensing methods, for example by using
use and manufacture is a planar electrode,[18] see example in two electrodes as a generator-sensor system. Secondly, one can
Figure 1a.[27] While this is useful for many purposes, 3D elec- use electrode arrays to have multiple detector electrodes for
trodes are commonly used, for example, for detection of DNA different target species or several electrodes of the same kind
markers,[28–30] see Figure 1b, or the use of electrode columns to for simultaneous detection. Figure 1d shows an example of a
increase the detection surface,[31] see Figure 1c. In the case of generator-sensor system, where the redox species FcTMA+ is
DNA detection, enzymes or other sensing molecules are depos- first oxidized to FcTMA2 +, and subsequently detected at the
ited on a porous structure and long deposition times are used downstream electrode.[40] By comparing the electric signals of
Figure 1. a) A channel flow cell with a 1.5 mm × 1.0 mm channel and two band electrodes. Reproduced with permission.[39] Copyright 2018, Elsevier
B.V. b) Microfluidic biochip for electrochemical detection of DNA hybridization. Reproduced with permission.[31] Copyright 2015, AIP Publishing LLC.
c) Example of a microfluidic device with a flow-through electrode. Reproduced with permission.[28] Copyright 2018, MDPI. d) A dual band electrode
within a flow channel of height 2h containing a generator and collector electrode of length xg, and xc, respectively, separated by a distance, d. EC reaction
scheme for the FcTMA+/2 +/AA system is displayed on the same scheme. Reproduced with permission.[40] Copyright 2013, Elsevier B.V. e) Schematic of the
fabricated three electrode arrays with two electrodes (A), four electrodes (B) and four electrodes mesh (C) is shown here with the microfluidic channel
across the active part of the electrodes. Reproduced with permission.[41] Copyright 2014, Public Library of Science. f) Example of a plastic-based design for
a microfluidic platform for multiple, standardized biosensor chips (i.e., graphene FET sensor). Reproduced with permission.[42] Copyright 2015, Elsevier
Ltd. g) A PCB-based electrochemical microfluidic imaging platform. Reproduced with permission.[43] Copyright 2013, The Royal Society of Chemistry.
the generator (ig) with the sensor (is), the concentration of the reactive in the range of applied potentials.[8,40,48] The co-reactant
targeted compound can be estimated.[45–47] This can also be molecules are then generated, and a part of them react with the
done by quantifying the amount of a chemically reactive spe- targeted molecules. The unconsumed co-reactant molecules
cies flowed into the channel, which is not electrochemically are then flowed downstream and captured by the collector
electrode. The relationship between concentration and reaction the working electrode without carefully considering the cur-
rate leads to the evaluation of rate constant k. rent distribution, the potential distribution is dependent on
The other application is to have several similar or different the current obtained and hard to predict.[60] To alleviate the
detectors integrated into the same microfluidic cell. These problems of external reference electrodes, it is often a better
detector electrodes can either be along the same channel, see choice to integrate the reference electrode inside the microflu-
Figure 1e, or the flow can be split into several independent idic flow cell.[61] This brings in additional challenges, especially
detectors, see Figure 1f. This can be done either to increase in regards to chemical compatibility of the reference electrode
the accuracy of the technique[49–51] or to have the ability to with the electrolyte.
detect multiple species.[52,53] It can also help during operation Internal reference electrodes can be of many types. Ag|AgCl
to identify single electrode failure by comparing the measured has been successfully integrated,[63,64] and is perhaps the most
behavior with several similar electrodes under the same condi- common system used for this purpose. Palladium hydride
tions. In the case of electrode arrays, care should be taken in (PdH) has also been used and shown stability for several
the design process to ensure that the products of one electrode hours.[60] Due to the influence of the electrolyte, the refer-
do not influence the response of other electrodes. ence electrode is often placed in a side channel separate from
A special use of electrode arrays is the mapping of the con- the flow and with an undisturbed electrolyte or separate elec-
centration distribution in a channel.[43,54] An example of con- trolyte supply, see example in Figure 2b. Pseudo-reference
centration mapping is shown in Figure 1g, where an array of electrodes can also be used, where the potential is not as well-
electrodes is used to sequentially map the concentration distri- known or as stable as standard reference systems. Examples
bution in the microfluidic channel.[42] Another special use of an here are particularly noble metal electrodes, for example Ag,
electrode array is that these can be used to filter out electrodes Au and Pt,[65] due to their chemical inertness in many elec-
that are undesired later on at the sensing electrode. Compounds trolytes. For the cases with extremely small currents (typically
with electrochemical reversibility can be filtered out from other between the scale of fA and nA within an electrochemical
compounds without this feature, as the molecules electrochem- sensor), integration of pseudo-RE and CE into one electrode is
ically oxidized upstream can be restored at the downstream possible.[66–68] In general, for accurate potential control, it
working electrode.[48] This method is very useful if the onset is recommended that an internal reference electrode placed
potentials of wanted and unwanted molecules are very close upstream of the sensing electrode is used.
to each other. With a one-way flow, the current at the down-
stream electrode will increase when the mass transfer toward
it is improved.[55] If the two electrodes are positioned very close 2.4. Counter Electrodes
to each other, in situ redox cycling (RC) effect could amplify the
redox current (e.g., 6500 times by Straver et al.[56]). The ratio The purpose of a counter electrode is to drive the opposite cur-
of a redox current to the current from non-reversible reactions rent of the working electrode, and as such, close the circuit. In
will be significantly enhanced, therefore chemicals with electro- order to have a well-controlled potential at the working elec-
chemical reversibility are distinguished from others.[44,56] trode, a few principles should be applied in the design of a
counter electrode. First, it should be larger than the working
electrode, preferably an order of magnitude, such that the reac-
2.3. Reference Electrodes tion occurring on the counter electrode does not influence the
potential of the working electrode. Secondly, the counter elec-
Reference electrodes (REs) are needed for electrochemical trode should be positioned in such a way that the products
sensing systems to accurately control the potential on the produced at the counter electrode can not reach the working
working electrode(s). The purpose is to have a well-determined electrode. Third, a potential gradient will occur between the
reduction-oxidation reaction occurring that has a known poten- working and the counter electrode during measurement, and
tial. Many such systems exist, for example standard or reversible the counter electrode should be placed in such a way that it
hydrogen electrodes (SHE or RHE), the silver-silver chloride does not interfere with the measurement of potential between
system (Ag|AgCl), mercury-mercury sulfate electrodes, and the working and reference electrode. The first point is a gen-
standard calomel electrodes (SCE). For integration in micro- eral principle in electrochemistry done by having a much larger
fluidic systems, two strategies can generally be used, firstly to counter electrode. This is a little more challenging in a micro-
have an external reference electrode that is connected through fluidic flow cell because the electrodes are confined to a small
the electrolyte or a salt bridge to the working electrode, often in channel and the space may be restricted in the design, in par-
the outlet reservoir,[57–59] see Figure 2a, or to have an internal ticular for array electrodes. The second point is commonly dealt
reference electrode embedded in the microfluidic cell,[60] see with by placing the counter electrode either on the opposite end
Figure 2b. The first approach is more flexible in terms of the of the channel of the working electrode or downstream in the
choice of reference electrode system and can generally be used channel. The third point is a problem particularly challenging
for all types of reference electrodes, also involving gases such as in microfluidics because the electrolyte volume is small,
the SHE and RHE systems. However, the ionic pathway to the leading to the potential and current gradients being restricted
reference electrode is often long, and due to the small volumes, to a small volume.
significant voltage correction may be necessary. It is also critical In regards to the current and potential distribution in the
to consider the placement of an external reference electrode, channel, the electrode placement is key. For example, for a uni-
for example, if the counter electrode is placed downstream of form current distribution on the electrode, it is advantageous
Figure 2. a) Example of a reservoir reference electrode. Reproduced with permission.[25] Copyright 2016, Elsevier Ltd. b) Example of an embedded refer-
ence electrode. Reproduced with permission.[60] Copyright 2016, Elsevier Ltd. c) Electrode positions: bottom wall, side wall, bottom and top walls, side,
top, and bottom walls. d) Potential and electric field lines modeled using COMSOL Multiphysics 4.3b for a 100 × 100 μm cross-section of a microchannel
where (left) represents electrode pads (5 × 20 μm), (middle) represents wire electrodes (radius 10 μm), and (right) represents an ideal, conformally
coated electrode on the side walls of the channel. Reproduced with permission.[62] Copyright 2014, Elsevier B.V.
to have the working electrode and counter electrode placed on potential distribution in the channel, for example, in a recent
each side of the channel, similar to a fuel cell. However, practi- work by Fanavoll et al.[60], it was argued that the measurement
cally, it is easier to fabricate electrodes placed on one side of of the potential at the working electrode was not feasible when
the channel, especially through photolithography, and these can the counter electrode was downstream of the working electrode
either be side by side when looking at a cross-section, or down- and the reference electrode was placed in the outlet chamber.
stream of each other in the flow direction, such as is common As such, they implemented an internal Pd-H reference elec-
for channel electrodes, examples in Figure 2a,b. Figure 2c trode that was placed upstream of the working electrode to alle-
shows some examples of placement in the cross-section of the viate this problem. In general, it is recommended to have the
channel (the flow direction is perpendicular to the drawing).[69] counter electrode placed on the opposite side of the channel of
Here, each will have current going between the electrodes, the sensing electrode if the fabrication method allows it.
and this can be numerically modelled. Figure 2d shows the
potential distribution for such a system, as modeled by Kadilak
et al.[62] Here, it is clear that while the side-by-side configura- 2.5. Flow Patterns and Channels
tion is easy to manufacture, this gives a complex potential and
current distribution and not necessarily even distribution of the In microfluidic flow detectors, the supply of reactants is pro-
reaction on the electrode. On the other hand, the top-bottom vided by the channels. In general, the limit of detection can
configuration in the right of Figure 2d gives straight current be improved by increasing the flow rate and thus the supply
lines (black lines), which is both ideal for even current distri- of analyte to the electrode. The flow interaction with the elec-
bution on the electrodes and establishes a well-defined poten- trode can be classified mainly in two ways. Firstly, a “flow-
tial distribution in the electrolyte. This alleviates problems with through” configuration, see Figure 1c, where the electrolyte
Table 1. Common shapes of microfluidic channels.
Shape Purpose Application(s) Typical Reference(s)

Straight/U-shaped Simple microfluidic cell Detectors; Analytical Cells; [27, 77–79]
Reactors; CO2 reduction cells
A-shaped/H-shaped Improving the conductivity between electrodes placed separately on main Micro-reactors with MS [74, 80]
channels and also minimizing the electrolyte crossover studies
S-shaped/Coiled Long channels for better adsorption of analytes; larger contact area between Detector; Electrolytic cells [66, 81, 82]
electrolyte and electrode for higher reactant conversion or better upstream mixing
Y-shaped/T-shaped Multiple inlets for seperating buffer/analyte or fuel/oxidant; Multiple Detectors; Fuel cells [73, 83]
outlets for implementing various downstream measurements; A thin
bottle neck may work as a mixer
Symmetrical Separation of anolyte and catholyte downstream Fuel cells [84]
Y-shaped/X-shaped
F-shaped Compatibility with air-breathing electrodes Fuel cells; CO2 reduction cells [85]
K-shaped Side channels for hydrogen and oxygen collection Water electrolyzers [86]
L-shaped A side channel for accmmodating reference electrode Detectors [60, 72]
H-shaped One channel for conversion from gas phase to aqueous phase; CO2 reduction cells [15]
The other channel for conversion from aqueous phase to gas phase
flows through a porous or mesh electrode. Secondly, a “flow-by” coefficient and t is time), it can exhibit thin layer behavior
or “flow-over” configuration, Figure 1a and d, where the flow and a concentration gradient only exists in two dimensions.[87]
direction is parallel to the planar electrode.[70] In addition to Another consideration is the implementation of numerical
these, some special cases exist, for example where the flow goes modeling, where it is generally preferred to model the 3D
toward a planar electrode, a wall-jet configuration.[71] channel as a 2D representation. To do this, one dimension in
For the vast majority of microfluidic devices, the shape of the the cross-section of the channel, for example the channel width
channels strongly reflects the purpose of the device. Table 1 lists should be about an order of magnitude larger than the channel
different types of channels and their respective applications. height.[88]
For detectors, the most common are straight, L/F-shaped, Y/T- In addition to the flow classification given in Table 1, many
shaped, and S-shaped. The straight type, example in Figure 3a, examples exist for specialized flow systems. In Figure 3e, a spe-
where all electrodes are placed along a single channel is the cialized cell aimed at generating drug metabolites is shown.[74]
simplest and most common approach. If supply of reactant is Here, two separate channels, one for the working electrode and
necessary during detection, a L-shaped channel can be used, one for the counter electrode were manufactured so that the
for example for the acidification of the analyte,[72] see Figure 3b. reaction products of each do not interact. In addition, several
A similar design, the F-shape, is used where gas evolution is connecting channels were made to provide conductivity between
expected, where a gas outlet can be placed along the channel the channel thus providing an even current distribution along
to release for example O2 or CO2 into the atmosphere through the working electrode. Another specialized example is given
a porous electrode.[12] The F-shape side channel can also be in Figure 3f. Here, a device to sense chemical gradients was
used for positioning of an internal reference electrode, to keep designed by using a multi-layer microfluidic device to measure
the reference electrode free from the reactions and flow in the analyte concentration.[75] This can also be done by using a multi-
main channel.[60] A second flow can be applied within this electrode microfluidic cell, where the flow can be guided to var-
side channel to counteract the mixing of the electrolyte from ious section of the cell for concentration sensing,[89] or by having
main channel. Y or T-shaped channels are commonly used in several outlets branching off the main channel.[90]
microfluidic fuel cells,[65,73,76] but can also be used in detectors
to have co-laminar flow and thus avoid mixing of the working
and counter electrode products. An example of this design is 2.6. Numerical Modeling
given in Figure 3c. Finally, to extend the detection zone, an
S-shaped channel can be used to allow the installation of larger Numerical modeling is a valuable tool when designing micro-
electrodes, see Figure 3d. This design is mainly used when the fluidic cells, and can help assess the influence of geometry,
analyte has a very low concentration of the target species.[66] flow rates, and applied potential on the detection system prior
There are other flow channel configurations developed for spe- to or after fabrication of the devices. A starting point with
cific purposes listed in Table 1. Since these are not for the pur- numerical modeling is to have clearly defined targets in mind,
pose of detection, they are not discussed further. for example, expected concentration of species in the analyte,
Aside from the shape of a channel, the size of the channel sensitivity of the used potentiostat, and available fabrication
also have impacts on the device. A small channel is generally techniques, in particular tolerances of the fabrication method
preferred to minimize the consumption of reactant. However, chosen. Several software packages are available where Comsol
if the channel is very thin, (<< 2 Dt , where D is the diffusion Multiphysics is the most widely applied because several physics,
Figure 3. a) Example of a electrochemical microfluidic detector with straight channel. Reproduced with permission.[27] Copyright 2019, Frontiers Media
SA. b) Carboxylation of benzyl halides using a microreactor (L-shaped channel) with an electrolysis and mixing compartment for subsequent acidifi-
cation. Reproduced with permission.[72] Copyright 2015, The Royal Society of Chemistry. c) A flow-over type counter flow MFC. Arrows represent the
flow direction. Reproduced with permission.[73] Copyright 2017, Elsevier B.V. d) Microfabricated microfluidic platform for desalination. The printed Ag
working electrode has a serpentine shape corresponding to the fluidic channel. Reproduced with permission.[66] Copyright 2017, The Electrochemical
Society. e) Scheme of an electrochemical cell with frit channels. Reproduced with permission.[74] Copyright 2012, American Chemical Society. f) Top view
of a three-layer (gradient layer, analysis layer and electrode layer) microfluidic flow device with four sets of three-electrode setup installed separately.
Reproduced with permission.[75] Copyright 2013, The Royal Society of Chemistry.
such as electrochemistry and computational fluid dynamics to reduce calculation time. For channel electrodes, the criteria
can be combined in a single simulation. Other examples are for doing representative 2D calculations have been assessed,
the Matlab FEATool Multiphysics package, and MFEM, where and can be stated explicitly as that one dimension in the cross-
the partial or ordinary differential equations of for example section is at least 10 times as large as the other.[91] In addition,
electrochemistry must be stated explicitly. For only flow simula- it is faster to do a calculation as a steady-state calculation com-
tions, computational fluid dynamics software such as Autodesk, pared to a transient calculation, and if the geometry can be
OpenFOAM, Ansys and SolidWorks can be used to assess the assessed satisfactorily using a steady-state calculation, that is
flow pattern, for example where pressure is used to drive the often preferred. However, numerical modeling examples exist
flow. The principle of a finite element method software, as all both for steady-state,[92–97] transient,[98] and even ac current
of these are, is that the cell is represented in 0D, 1D, 2D, or 3D, applications.[25,45,46] Care should also be taken when selecting
a mesh is created to discretize the problem, and a set of equa- the mesh size in the numerical modeling, where a preliminary
tions are solved with appropriate boundary conditions either as mesh sensitivity analysis is useful. In that case, consecutively
steady-state or as a transient signal. finer mesh is used for the same calculation to assess when the
A numerical simulation can easily become complex and calculated response stabilizes as a function of mesh size.
time-consuming if not suitable limits are applied. For example, The value of numerical modeling is in assessing the response
it is often an advantage that a 3D system can be assessed in 2D as a function of applied flow, geometry, and experimental
Figure 4. a) Computational proof-of-concept: steady-state Ru(NH3)63+ concentration distributions obtained from the electrochemical model in the sym-
metry plane of a flow cell operating at negative WE overpotential. The magnitude of the WE current (I) increases with mean velocity (U). Conditions:
c0 = 5 mM, η = −0.2 V, h = 100 μm, w = 500 μm, d = 1 mm. Reproduced with permission.[88] Copyright 2006, Springer-Verlag. b) Scheme and flow simu-
lation of the microfluidic 2-cell microchannel used in the mixed-reactant μDMFC stack. Reproduced with permission.[99] Copyright 2017, Elsevier B.V.
parameters. In particular, where the electrodes are fairly passing the electrode, while at high flow rate, 10 mm s−1, only
simple, that is, planar, the geometry can be simplified from a small fraction of the original species has reacted. Another
3D to 2D, and where amperometric or impedimetric detec- example is given in Figure 4b where numerical modeling was
tion methods are used, numerical simulations can accurately used to assess the flow conditions of a microfluidic flow reactor.
estimate the electrochemcial response of the electrode. For 3D In this case, only the flow conditions were considered and not
electrodes, for example in the biomedical field, or where poten- the electrochemical reaction making it feasible to do the model
tiometric methods are used, numerical modeling can largely in 3D. A third example is to model only the current distribu-
be used qualitatively as a rough approximation of the response. tion, and Figure 2d is an example of this. Finally, it is strictly
Figure 4a shows an example of a channel electrode, where feasible to solve all of these problems; the electrochemical
the concentration profile and current is calculated as a func- reaction, the flow in the cell, and the current distribution in
tion of the flow rate.[88] Here, one can clearly see the electrode the same model, but this is costly and time-consuming. It also
reducing the concentration and the change in concentration is carries considerable risk of obtaining no convergence or non-
distributed downstream of the working electrode. Also, the flow physical solution, so it is recommended that each aspect of the
rate heavily influences this profile, where for the low flow rate, microfluidic flow cell is first understood individually before a
0.1 mm s−1, about 60% of the reacting species is consumed after fully integrated model is designed.
Figure 5. Fabrication of electrodes for microfluidic devices by photolithography and laminar flow lithography. a) Microelectrode and microchannel
fabrication steps including photolithography, metal etching, electrodeposition of Au, Pd/XC-72 and Pt/XC-72 electrode spraying, and PDMS moulding.
Reproduced with permission.[102] Copyright 2013, Elsevier Ltd. b) Process flow diagram outlining the applied usage of LFL to create both tri-metallic
(three electrodes, left) and bi-metallic (two electrodes, right) devices. Reproduced with permission.[103] Copyright 2014, The Royal Society of Chemistry.
3. Fabrication of Microfluidic Devices strong adhesion.[101] For the electrodes, a negative or positive
photoresist is spin-coated on top of the substrate. The elec-
A microfluidic cell can be made through several manufacturing trode pattern is then made by selectively exposing a section of
methods, where lithography and machining are the most the layer to UV-light using a pre-designed mask. Subsequently,
common. For lithography, the cells typically consist of two parts, the photoresist is developed and/or etched, exposing the
a substrate with the electrodes and a polymer with the channel, pattern on the substrate. The selected electrode material can
often made of polydimethyl siloxane (PDMS), examples in then be deposited by sputtering or vapour deposition, often by
Figure 5 and 6a.[100] For machining, typically a sandwich struc- first depositing an adhesive layer, for example Ti or Cr.[65,89,102]
ture is used, where the channel and electrode substrate(s) are This works most commonly for metals and some metal oxides,
made separately including a holder that can seal the device, see and is especially suited to make planar electrodes.[69] As it has
examples in Figure 6b–f. 3D machining can also be used, where been developed in the semiconductor industry, finely resolved
the examples in Figure 7 show 3D printed flow devices with electrode patterns can be created, up to half the wavelength
electrodes installed separately or printed at the same time as of the radiation used (about 1 μm).[44] A specific type of fab-
the devices were made. In most cases, polymeric tubes deliver rication is laminar flow lithography (LFL),[103], illustrated in
liquids into the channel through the inlet, and a reservoir Figure 5b. Here, two or three layers of different metals (for
is placed at the outlet which can serve both as a holder of the example Ag, Au, and Ni) are deposited onto the bottom and
liquid passing through and as a holder for the reference elec- sides of a microfluidic channel with electrically insulating
trode. The manufacturing methods greatly differ in the choice layers between them. By co-flowing etchant and water at spe-
of materials, possibilities for electrode design, and resolution, cifically selected rates, the depth and pattern of etching can be
and these aspects are dealt with separately below. controlled. Therefore, different metals are exposed by etching
at different positions.
For the channel, this can be made in a similar way, shown on
3.1. Photolithography the right in Figure 5a. Here, a thicker photoresist is used, often
SU-8,[104,105] and a negative of the channel is made by depositing
Both the electrodes and the channel can be made through pho- this layer on a substrate, glass or Si. A polymer is then moulded
tolithography, resulting cell shown in Figure 5a. As a substrate, on top of the photoresist negative, and finally, the substrate with
microscope glass slides and silicon wafers are favored by most the electrode is combined with the channel creating the micro-
of the microfluidic works due to their chemical inertness and fluidic cell. The polymer is most commonly PDMS, and thanks
Figure 6. Electrochemical microfluidic devices with different configurations. a) A typical three-electrode microfluidic flow device made by casting PDMS
on a glass substrate. Reproduced with permission.[100] Copyright 2017, Elsevier B.V. b) Exploded view of a typical two-electrode microfluidic flow device
made by sandwiching the channel pad and electrode pads with a pair of PMMA substrates. Reproduced with permission.[111] Copyright 2015, Elsevier
Ltd. c) Exploded view of a co-laminar flow cell with electrode pads positioned on the side walls of the channel. Reproduced with permission.[114] Copy-
right 2016, Elsevier Ltd. d) Top view of the counter-flow MFC showing the flow directions of anolyte (yellow arrow) and catholyte (blue arrow) and e)
exploded view of the (d). Reproduced with permission.[73] Copyright 2017, Elsevier B.V. f) Scheme of the stack micro-reactors (three cells). Reproduced
with permission.[119] Copyright 2015, Wiley-VCH.
to its chemical inertness, PDMS is compatible with most of the 3.2. Machining
solvents and electrolytes,[106] However, if online physical charac-
terization methods are needed, PDMS will generate undesirable Machining of microfluidic flow cells has been a common
background signals reducing the accuracy of the optical tech- choice in the last decades and is usually made through a sand-
nique, as reported in online mass spectrometry (MS) studies.[6] wich structure, see Figure 6b-f. Here, separate layers can be
Also, a great expansion of PDMS in all dimensions when made with a channel cut in the middle layer, typically by using
swelled with various organic solvents were repeatedly measured a CNC machine.[73,111] Then, electrodes can either be directly cut
by Adamiak et al.[107], which shows that its use as microfluidic from a metal or for example a graphite substrate, or be depos-
building blocks is subject to the property of solvents. PDMS is ited directly on the substrate or side walls of the channel in a
also permeable by oxygen, which may be a disadvantage in some sandwich structure.[112–116] Machining from graphite, glassy
experiments. Finally, the seal between the channel and the sub- carbon, or other carbon-based materials are often used due to
strate with electrodes is usually achieved with plasma assisted their high chemical compatibility and high electrical conduc-
bonding or UV-assisted thermal bonding.[108] The endurance tivity.[117] Machining methods have a resolution limit of about
of the connection to the pressure caused by flowing or external 10 μm, at least an order of magnitude higher than photolithog-
forces can be predicted theoretically by calculating the binding raphy. However, these methods are more widely available and
energy between the substrate and the building blocks.[108,109] In offer more choice in terms of the geometry and electrode types.
terms of the mechanical strength, PDMS can tolerate about 1 Common material choices for machining are acrylic
MPa pressure, resulting in a flow rate limitation.[6,102,110] (PMMA),[58,61] polyether ether ketone (PEEK),[87,118]
Figure 7. 3D printed microfluidic devices. a) Microfluidic thread-based electroanalytical devices (μTED). Reproduced with permission.[145] Copyright
2021, Elsevier B.V. b) 3D Microfluidic Channel for Supercapacitor and Electrochemical sensing. Reproduced with permission.[146] Copyright 2020,
Elsevier B.V. c) 3D printed devices with inserted screen-printed electrodes. Reproduced with permission.[147] Copyright 2016, American Chemical Society.
d) 3D-printed PET device with reference, working, and counter electrodes incorporated in the fluidic channel. Reproduced with permission.[148] Copyright
2016, Wiley-VCH. e,f) 3D printed device with an embedded gold electrode placed in the middle of the channel. Reproduced with permission.[149] Copy-
right 2020, The Royal Society of Chemistry. g) 3D-printed fluidic cell with integrated carbon-loaded PLA electrodes. Reproduced with permission.[150]
Copyright 2021, Elsevier B.V.
polytetrafluoroethylene,[119] cyclic olefin copolymers (COC, goal is to keep homogeneous pressure along the channel. The
Topas),[108,114,120] polyvinyl chloride,[79,121] and polycarbonate mechanical seal can be further secured by the use of o-ring type
(PC).[70,122] Material compatibility is an important consid- seals, for example silicon rubbers,[58] Silastic™, or molten wax
eration, for example, PC and polyethylene terephtalate (PET) in the case of porous substrates.[126–128] Permanent seal can be
may dissolve in certain acids or organic solvents.[123] Simi- achieved by using chloroform if the whole structure is made
larly, PMMA is often preferred because it deforms less with of PMMA, as chloroform dissolves PMMA and evaporates
flow, at least compared to PDMS.[55] In order to seal the quickly leaving no particulate traces. The idea behind it comes
setup, it is often mechanically and reusably sealed by a pair of from the solubility of the polymer and a proper solvent. A few
steel plates,[124] aluminum plates[82] or PMMA substrates, which examples include acrylonitrile butadiene styrene (ABS) and
have screw holes for screw-tightening and sealing.[57,84,99,124,125] tetrahydrofuran (THF), PEEK and chloroform + dichloroacetic
When screws or other kind of mechanical aid is used, the acid, polyurethane and THF, polycarbonate and THF or dichlo-
position where these elements are placed becomes important. romethane, etc.[129] Also epoxy-gluing and resins can be used
As there are innumerable possible designs of microfluidic to achieve a good adhesion between the components of a sand-
devices, there is no real guideline for placing screws, but the wich structure.[124]
After fabrication, the carbon-based electrodes can be cov- ment (vertical). The device can therefore be printed out from
ered by the desired electrode material or catalyst. This can be bottom to top. For devices made by FDM, a filament of ther-
done by deposition methods, such as sputtering or vapour moplastic materials is fed into the machine and melted at the
deposition like for photolithography, but also through sev- extruder. While filament is extruded, it hardens and makes the
eral other methods, such as electrodeposition, direct writing, devices layer by layer. Polylactic acid (PLA)[150,153] and ABS[151]
or wet impregnation. Electrodeposition consists in reducing are commonly used materials for FDM. Bishop et al. reported
metal ions by applying a potential and depositing these on a 3D printed device with transparency using PET.[148] SLA
electrodes.[43,129–131] This technique work for many metals that is a different technique, which uses UV light to activate the
can be dissolved, including noble metals such as Pt. Due to crosslinking of photo-sensitive resins (thermoset materials).[147]
the method, edge effects may arise as higher currents arise Its advantages include fine features, visible transparency and
at the edges of the electrode and uneven electrodes may be smooth surface finish.[154,155] As reported by Gross et al. the
the result,[132,133] or micro-strain in the deposit may lead to SLA-based devices can be as thin as 17 μm.[156]
mechanical failure or instability.[134] Therefore, small currents In most of the 3D-printed microfluidic flow devices, elec-
are chosen to produce a smooth coating. A cheaper and simpler trodes are made separately. This can be seen as an advantage
method to fabricate a noble metal electrode pad on a glass slide compared with conventional electrochemical microfluidic
or silicon wafer is called the “direct write technique”.[134] Here, devices. In the conventional ones, electrodes are integrated to
metal nanoparticles are mixed with organic solvents to make devices or substrates. Therefore, the devices or substrates have
pastes that are deposited on top of a rigid substrate.[135] After to be disposed once an electrode is damaged or needs replace-
heating, the obtained electrode is much thicker with higher ment, which is not a concern for the 3D-printed ones.[157] It is
mechanical strength than thin films made by lithography. How- noticeable that a few of the 3D printed electrochemical micro-
ever, this technique is at the expense of electrode uniformity fluidic devices do not have channels for fluid flow. Instead,
as the accurate control over electrode thickness is very chal- they use threads to transport electrolytes over a sensor chip
lenging. Wet impregnation methods[121,136–139] (e.g., by dipping, (Figure 7a). This type of device are called microfluidic thread-
spraying, sponging, brushing, casting, pressing and so on) are based electroanalytical devices (μTED).[145] For the majority of
usually adopted to accommodate nano-metallic electrocatalysts 3D-printed devices which use channels for fluid flow, electrodes
with high specific surface area.[140] These are usually deposited are inserted (Figure 7b,c),[146,147,153,158,159] threaded (Figure 7d)[160]
on conductive carbon supports, such as graphite, graphene,[141] or sandwiched (Figure 7e,f)[149] into the devices. Among them,
glassy carbon,[142] and carbon nanotubes.[76,111] To avoid aggre- the inserted electrodes from the sides are the most common,
gation and deformation, the amount of catalyst loading has to including three bar-shaped electrodes (Figure 7b) inserted and
be carefully determined and controlled,[126] and surfactants are chips with screen-printed electrodes (Figure 7c). Figure 7d
often added as they can attach to the surface of primary par- shows that electrode can stay in the middle of the channel,
ticles. Wet impregnation can also be used for 3D electrodes, sandwiched by two pieces of 3D printed materials.[149] Most
where added problems such as securing the electrodes during recently, 3D-printed electrodes are reported to facilitate the
rapid flow may arise.[143] To improve the connectivity between easy and customized positioning of electrodes and one-step
deposited metals and supports, binders play a critical role, fabrication of devices (Figure 7g). Since 2019, a few approaches
where for example the addition of carbon nanotubes have have been reported, as exemplified by carbon-loaded PLA,[150]
shown promise in increasing the electrical connection between Graphene-PLA composite,[39] and carbon nanotube doped PLA
deposited particles and carbon supports, and also to improve filament.[151]
the tolerance for harsh voltages.[118] The use of binders compli- Being a newly developed strategy, the most commonly
cates the overall design, and excessive use of binders may also reported analytes of these devices include heavy metal ions
increase the contact resistance.[144] like Pb(II) and Cd(II),[147,150] glucose,[158] and compounds like
catechol.[39,149] In most of the microfluidic detectors, chemical
conditions are mild. Materials for 3D printing which can resist
3.3. 3D Printing harsh chemical conditions are expected to be developed in
the future.
With the help of computer-aided design (CAD) software like
AutoCAD and SolidWorks, 3D printing, also known as addi-
tive manufacturing, has been adopted by researchers to make 3.4. Other Methods
microfluidic flow devices in the most recent decade, especially
since 2014.[149] It is regarded as an advanced methodology to Printed circuit board (PCB) has also emerged as substrate
produce small devices with good reproducibility, low cost and material for electrochemical microfluidic devices as reported
short fabrication time.[39,150,151] Based on the way devices are by Greener et al.[43] PCBs have unique advantages in detection,
made, 3D printing can be categorized into three types: Fused as electrode arrays on the sub-millimetric scale (about 25 μm)
Deposition Modelling (FDM), Stereolithography (SLA), and can be printed on top as ”pixels”, forming a 2D electrochemical
Microjet Modelling (MJM),[152] while most of the works use imaging platform. Lastly, for mass production of microfluidic
FDM and SLA. devices, injection molding and hot embossing is a feasible
With a 3D printer software (e.g., Ultimaker Cura), a 3D path. These methods have a resolution approximately in the
design (STL file) can be converted to a mapping of the extrud- 100 μm range, and the cost of creating the mold is only justified
er’s movement (horizontal), as well as the stage’s move- for large quantities.
4. Challenges and Solutions Along with the optical path, the irradiation configuration is a
choice to consider to design the experimental setup. The most
In addition to the components discussed in the previous sec- common irradiation configurations are the front-illumination
tion, several additional or specific challenges may arise due to and back-illumination configurations. The front-illumination
the nature of the experiment. Here, the design for light inci- configuration, also known in photoelectrochemistry as electro-
dence and emission, control of temperature, pressure and pH, lyte-electrode (EE) configuration, represents an optical pathway
and durability of the electrodes are dealt with separately. where light crosses directly the body of the cell, the electrolyte
layer and finally reaches the photoactive material. Figure 8a
presents an example of front-illumination configuration where
4.1. Light-Induced Reactions the authors used a microfluidic fuel cell with a PMMA-body
using visible-light sensitive TiO2 nanotubes decorated with
Light stimuli and measurements are often used in micro- Ni and NiO nanostructures. Electrodes are often prepared on
fluidic devices with the intention of inducing a specific reac- a transparent conductive substrate with high transmissivity in
tion.[99,127,161] Formally, light-triggered reactions take place at the the spectral region where the photocatalyst absorbs light, such
interface of the photoactive material, which is supported on a as fluorine (FTO), indium (ITO), or antimony-doped tin oxide
substrate. Microfluidic cells using light-triggered reactions can (ATO). Deposition of the photocatalysts must ensure good
operate in photocatalytic (PC) or photoelectrocatalytic (PEC) ohmic contact and mechanical stability against the electrolyte
conditions. The photocatalytic mode is the simple interaction flow, often achieved by annealing. Figure 8b,c presents a case
of the photocatalyst with light, inducing a reaction with a mol- where an ITO substrate was used as optical window when
ecule in the electrolyte. In the photoelectrocatalytic mode, the BiVO4, a visible-light absorbing material, was used as photo-
previous conditions are preserved but the photocatalyst is sub- catalyst. Thinner layers of photocatalyst typically have higher
jected to external polarization thus, becoming a photoelectro- efficiencies in PEC experiments due to improved kinetics in
catalyst. For electrochemical experiments, PEC conditions are electron-hole separation. However, there is still a range where
more relevant. The photoactive materials (photocatalyst) require layer thickness can be explored to optimize the electrode effi-
the stimulation of light of a particular minimum wavelength to ciency. The back-illumination or substrate-electrode (SE) con-
generate charge-carriers (electrons and holes). The goal in these figuration, corresponds to an optical pathway where the pho-
experiments is to induce a photoreduction by injecting an elec- toactive material is supported on a transparent substrate and is
tron; or a photo-oxidation by the electron-scavenging action irradiated directly, avoiding the losses from the electrolyte and
of holes. Numerous challenges arise, and we can distinguish other phases. In Figure 8c, a back-irradiated (SE) configura-
internal from external factors determining their efficiency. We tion was used in a microfluidic reactor for the photo-oxidation
call internal factors those related to the photocatalyst itself. of methylene blue (MB).[161] For water purification purposes, a
Since the literature is vast detailing these phenomena,[162–167] front-illumination approach was performed by Wang et al.[169]
this section will only describe the external factors that are where methylene blue (MB), was oxidized using blue LED illu-
related to the microfluidic cell and the light sources. mination and BiVO4, finding higher degradation effects for
One major consideration is the effective irradiance reaching longer residence times. For enhanced charge-carrier separation,
the photocatalyst. This leads to increased complexity of the a back-illumination configuration was used by Diaz-Real et al.
device, as the optical path should be considered. For the sake to perform the same goal.[161]
of this discussion, the optical path represents all the elements
that electromagnetic radiation (i.e., light) interacts with from
the emission point to the photoactive material. In the context 4.2. Optical Detection Methods
of microfluidic flow cells, the body of the cell, electrolyte,
optical window and substrates where the photoactive material Several optical methods are suitable for integration in microflu-
is deposited, should be included in the optical path. The goal idic cells,[171] as the photons from excitation sources only need
is to have a selection of materials in the optical path that mini- to penetrate a thin layer of electrolyte.[172] Among the several
mizes power intensity losses due to absorbtivity or reflectivity advantages they integrate are quantitative detection, no external
of each of the phases. Some materials that are transparent to high pressure or voltage is required, and no clean-room facility
visible light can absorb in other spectral ranges, like the case or vacuum system is needed. UV-Visible Spectroscopy, Fourier
of PMMA that absorbs in the UV range. Thicker layers further Transformed Infrared (FTIR) Spectroscopy and Raman Spec-
reduce transmissivity and, thus, thinner layers are preferred. troscopy are among the more commonly used methods that
Also, an optical window with enhanced transmissivity in the can provide quantitative or qualitative information about the
UV region (e.g., quartz glass) can be used, and the material species in the electrolyte. Other methods, such as colorimetry,
choice of the cell body and substrate should have as low auto- fluorescence and electro-chemiluminescence (ECL) has also
fluorescence as possible.[108] Alternatively, in proof-of-concept been integrated in microfluidic cells,[170,173,174] however, these
experiments, higher irradiation power from the light source methods are very dependent on the chemical system and not
or a biased-irradiation source can be used. An example of the widely applicable.
previous approach is mentioned by Wang et al.,[168] where they UV-Vis spectroscopy can be used in either absorbance or
increased the operating voltage for their LED light source to transmittance mode, where quantitative information (con-
increase the irradiation power from 80 mW cm−2 (at 10 V) to centration) can be obtained from the analyte. An example of
140 mW cm−2 (at 11 V). the previous case is the work of Diaz-Real et al.,[161] where the
Figure 8. Microfluidic devices involving light incidence. a) A microfluidic methanol fuel cell using Ti/TiO2-NT/Ni as anode and Pt/toray paper as
cathode. Reproduced with permission.[83] Copyright 2016, Elsevier Ltd. b) 3D diagram of the reaction chamber of a microfluidic photoelectrocatalytic
reactor. The reactor consists of a blank ITO glass layer as the cover, a BVO-coated ITO glass layer as the substrate and an NOA81 adhesive layer as the
spacer. Reproduced with permission.[169] Copyright 2012, The Royal Society of Chemistry. c) Microfluidic photo electrochemical reactor scheme used for
the dye degradation. Reproduced with permission.[161] Copyright 2016, Elsevier B.V. d) Mechanism of redox reaction coupling on a BPE array spanning
two physically separated micro-channels. Left: Top view of the layout of the BPE array relative to the PME and the microfluidic channels. Right: Top
view at the opposite poles of a single BPE digit coupled to the WE digits in the analytical channel and to the quasi-reference electrode (QRE) digits
in the reporter channel. The regions of interest (ROI) used for fluorescence measurements are indicated on BPEWE and BPEQRE. Reproduced with
permission.[170] Copyright 2016, Wiley-VCH.
photo-oxidation of MB was followed via absorbance spectros- significantly improves the intensity of scattered photons, ena-
copy in the visible range. In this work, a photoradiometer was bling the detection of trace amounts of molecules. Also, the
used to measure the irradiation power coming from a Xe lamp, acquisition time can be shortened, which increases the pos-
and the response was calibrated so that the extent of photo-oxi- sibility of detecting intermediates of a reaction.[176] SERS has
dation was quantifiable. been applied to detection of glycerol oxidation products in a
Fourier-Transformed Infrared (FTIR) Spectroscopy and paper-based flow device, where a SERS detection zone was
Raman Spectroscopy could display the absorption and non- positioned downstream of the three electrodes.[176] However,
elastic scattering of photons with spectra with characteristic SERS is a good qualitative method rather than a quantitative
peaks linked to the different molecular bonds. Compared method because SERS signals strongly depend on the nano-
with many other physical characterizations, they are capable structure of the metallic substrates, which can be irreversibly
of working with concentrated electrolytes as the mono-atomic affected by the potentials applied during measurements. Also,
ions are not a source of background noises. In situ FTIR has it is hard to guarantee the structural uniformity over the whole
been already used in many electrochemical studies,[175] as it has electrode surface.
the potential of examining the surface-adsorbed species. The Bipolar electrode (BPE) system is another novel research
channel of microfluidic devices for in situ FTIR has to be very that involves the emission of photons,[170] of which the mecha-
thin, otherwise water absorption will be a source of noise that nism is illustrated in Figure 8d. A closed BPE system is able to
overwhelms all other signals.[176] separate two redox systems to eliminate their interactions, as a
Raman spectroscopy offers non-destructive detection of BPE contacts both channels and works as anode and cathode
many organic compounds, biomaterials and metal oxides.[177,178] simultaneously. For example, fluorescent reaction may happen
A unique advantage of Raman spectroscopy used in microflu- at one end of the BPE, reflecting the concentration of a product
idics is that water molecules do not have any strong character- of interest generated from the other end.
istic peak, therefore it would not convolute or mask the signals
collected from the targeted molecules. Also, it is possible that
quantitative detection can be achieved by pre-calibrating the 4.3. Control of Pressure, Temperature, and pH
characteristic peaks.[178] When used in electrochemical micro-
fluidic devices, noble metal electrodes (e.g., Ag or Au) can The control of parameters such as pressure, temperature and
offer a surface enhanced Raman scattering (SERS) signal that pH can be critical to ensure reliability of the overall detection
method. These problems can be further enhanced due to the 5. Summary

reactions occurring in microfluidic channels with limited space
and reactant to disperse the effect. Pressure control is a major From the above sections of design, fabrication and challenges,
challenge for electrochemical microfluidic flow devices, as an a general flow chart for the design of microfluidic electrochem-
over-pressurized device may damage the electrodes residing ical detectors can be made, see Figure 9. This goes through five
in the channel[134] and cause leaking.[102] Besides, flow patterns phases:
may be disturbed by the buildup of a pressure gradient within
the channel.[179] Therefore, modeling of the pressure buildup 1. Decide goal of the device and limitations
within the channel (with pre-set dimensions) before fabrication 2. Do a literature review of microfluidic electrochemical devices
is recommended.[12,44] solving similar or auxiliary problems
Temperature fluctuations can influence the electrochemical 3. Set up an initial design, and make this in a CAD software
measurement by changing the reaction rate or diffusion rate of 4. If available, set up a numerical model of the system and as-
species.[12] For devices that operate with large currents, ohmic sess the initial design based on the goals and limitations
heating and endothermic/exothermic reactions may induce 5. Fabricate the device and test
temperature gradients. Issues may arise and thus influence
the electrochemical measurements.[140,180] To prevent self- For each point, there are specific considerations. Firstly, in
heating of a device, immersion into a water tank or water bath deciding the goal, important aspects are to decide on the nature
with a well-controlled temperature is suggested by Ye et al.[57] of the experiments, what are the goals in terms of detection of
Besides, a copper heat exchanger can also be used to control the the device, the detection limit of the potentiostat or electronic
temperature.[181] control used, and the available fabrication methods to the exper-
Similar to temperature, the pH gradients can influence the imenter and their tolerance. Secondly, a literature review should
operation of electrochemical microfluidic devices, however it be conducted on similar systems. Thirdly, the initial design can
is not always a concern for electrochemical detectors working be made based on existing devices and optimized for the spe-
at low currents. Therefore, the investigation and modulation cific problem. Here, the experience of the experimenter plays a
of pH is always expected but only attempted by a few experi- role in having an idea of what is a useful design both in terms
menters.[182] For devices that involve the fast consumption or
production of protons and hydroxide, the pH near the surface
of the electrodes tends to be different from the pH in the bulk
solution,[109] and the change of pH can significantly drift the
reaction potential and subsequently impact the reactivity.[79]
If the system is planned to be weakly acidic or basic, using a
buffer solution is a common approach.
4.4. Electrode Stability and Regeneration
In microfluidic cells, the electrode itself and its stability deter-

mines the life-span of the device. There are several processes
that can deteriorate the electrode, such as mechanical (e.g., ero-
sion,[183] abrasion,[184] abusive handling,[185] delamination[77,134])
or physico/chemical processes (e.g., polarization stress,[186] oxi-
dation,[134] corrosion,[187] ion intercalation,[188] poisoning[172]). For
example, carbon-supported electrodes suffer from chemical
instability inducing electrode damage and undesired aggrega-
tion of nanostructured electrocatalysts.[118] With this in mind,
many microfluidic cells are designed to be disposable or only to
be used a few times. The cost of materials used in the electrode
will often dictate the overall cost of the microfluidic cell, and may
therefore limit the applicability. For microfluidic cells intended to
be used several times or involving expensive catalyst materials,
it is important to consider the durability. Some approaches have
been suggested in the literature. Herrasti et al.[189] used magnetic
confinement of single-walled carbon nanotubes onto magnetic
nanoparticles as a method to simplify electrode recovery and
protocol automation. Also, by applying specific voltages to the
electrodes, poisoning adsorbates can be removed and expose
the ”fresh” surface underneath.[77,172] It should be stressed that
no approach is general for a microfluidic cell, and an approach
should be designed for the specific electrode used. Figure 9. Design guideline for microfluidic flow detectors.
of sensitivity and in terms of possibility of fabrication. Also, if [3] W. J. Paschoalino, S. Kogikoski, J. T. Barragan,
a numerical model is to be made, it is very helpful if that it is J. F. Giarola, L. Cantelli, T. M. Rabelo, T. M. Pessanha, L. T. Kubota,
incorporated in the initial design, for example, if a 2D model is ChemElectroChem 2019, 6, 10.
to be used for a 3D system which results in design limitations [4] D. G. Rackus, M. H. Shamsi, A. R. Wheeler, Chem. Soc. Rev. 2015,
44, 5320.
to the geometry. Fourthly, a numerical model should be made
[5] J. C. Cunningham, P. R. DeGregory, R. M. Crooks, Annu. Rev. Anal.
with the initial design and allow for its assessment based on the Chem. 2016, 9, 183.
design criteria and goal. With such a numerical model, an itera- [6] F. T. G. van den Brink, W. Olthuis, A. van den Berg, M. Odijk, TrAC,
tive process where the initial design is gradually improved and Trends Anal. Chem. 2015, 70, 40.
optimized is feasible and lead to a much faster path toward the [7] A. F. Chrimes, K. Khoshmanesh, P. R. Stoddart, A. Mitchell,
goal of solving the initial problem. Lastly, the optimized design K. Kalantar-Zadeh, Chem. Soc. Rev. 2013, 42, 5880.
can be fabricated using the available fabrication methods and [8] A. Bîrzu, J. Coleman, I. Z. Kiss, J. Electroanal. Chem. 2014, 726, 27.
tested. This also allows for the verification and improvement of [9] W. Rösing, T. Schildhauer, J. König, C. Cierpka, Microfluid.
the numerical model. Nanofluid. 2019, 23, 1.
With the suggested flow chart in Figure 9, a fast design pro- [10] S. K. Yoon, G. W. Fichtl, P. J. Kenis, Lab Chip 2006, 6, 1516.
[11] Y. Sameenoi, K. Koehler, J. Shapiro, K. Boonsong, Y. Sun, J. Collett,
cess with minimal prototyping simplifies the path toward an opti-
J. Volckens, C. S. Henry, J. Am. Chem. Soc. 2012, 134, 10562.
mized microfluidic device. Furthermore, an observed drawback [12] E. Kjeang, N. Djilali, D. Sinton, J. Power Sources 2009, 186, 353.
of many microfluidic sensors is that while many designs are pro- [13] E. Kjeang, Microfluidic Fuel Cells and Batteries, Springer, Cham
posed and tested, very few reach the state of commercialization. 2014.
The authors believe that through a more thorough design pro- [14] Y. Wang, S. Luo, H. Y. Kwok, W. Pan, Y. Zhang, X. Zhao,
cess, which also incorporates the feasibility of cheap fabrication D. Y. Leung, Renew. Sustain. Energy Rev. 2021, 141, 110806.
methods, this can be alleviated in the future and lead to the suc- [15] D. M. Weekes, D. A. Salvatore, A. Reyes, A. Huang,
cessful commercialization of more microfluidic sensing devices. C. P. Berlinguette, Acc. Chem. Res. 2018, 51, 910.
Finally, in the field of microfluidic electrochemical sensors, [16] A. Fernández-la Villa, D. F. Pozo-Ayuso, M. Castaño-Álvarez, Curr
there are some trends that can be identified. Firstly, devices Opin Electrochem. 2019, 15, 175.
[17] M. Puiu, C. Bala, Trends Anal. Chem. 2020, 125, 115831.
are going toward finer and smaller details driven by the devel-
[18] M. Odijk, A. van den Berg, Annu. Rev. Anal. Chem. 2018, 11, 421.
opment of fabrication methods. Secondly, researchers are [19] S. Prakash, M. Pinti, B. Bhushan, Review article: Theory,
working on integrating more processes on a single-device (Lab- fabrication and applications of microfluidic and nanofluidic
on-chip) to automate and miniaturize several processes for biosensors 2012.
analysis. Thirdly, candidate materials for more selective sensing [20] Y. Song, J. Zhang, D. Li, Micromachines 2017, 8, 1.
electrodes are being developed. With this, it is expected that [21] J. A. Diaz-Real, M. Guerra-Balcázar, N. Arjona, F. Cuevas-Muñiz,
several highly specialized and selective microfluidic cells for L. G. Arriaga, J. Ledesma-García, in Advances in Microfluidics -
sensing purposes will arise in the next decade. New Applications in Biology, Energy, and Materials Sciences, InTech,
Rijeka, Croatia 2016.
[22] A. Escarpa, M. C. González, M. A. López, Agricultural and Food
Electroanalysis, Wiley, Chichester 2015.
Acknowledgements [23] K. Nakazato, in State of the Art in Biosensors (Ed: T. Rinken),
IntechOpen, Rijeka 2013, Ch. 7.
This research has been supported by CONACyT (FORDECYT-PRONACES,
“Ciencia de Frontera 2019-1” Grant Number CF-2095605). In addition, T.L. [24] N. Bao, J. Wang, C. Lu, Anal. Bioanal. Chem. 2008, 391, 933.
thanks the University of Victoria for financially supporting this research, [25] T. Holm, M. Ingdal, E. V. Fanavoll, S. Sunde, F. Seland,
J.A.D.R. thanks CONACyT for the financial support, and T.H. thanks the D. A. Harrington, Electrochim. Acta 2016, 202, 84.
Institute for Energy Technology for supporting this research. [26] Y. Yusof, K. Sugimoto, H. Ozawa, S. Uno, K. Nakazato, Jpn. J. Appl.
Phys. 2010, 49, 3.
[27] T. Abadie, C. Sella, P. Perrodin, L. Thouin, Front. Chem. 2019, 7, 1.
[28] E. W. Nery, M. Kundys-Siedlecka, Y. Furuya, M. Jönsson-
Conflict of Interest Niedziółka, Sensors (Switzerland) 2018, 18, 11.
[29] M. C. Horny, V. Dupuis, J. M. Siaugue, J. Gamby, Sensors
The authors declare no conflict of interest.
(Switzerland) 2021, 21, 1.
[30] E. O. Blair, D. K. Corrigan, Biosens. Bioelectron. 2019, 134, 57.
[31] A. S. Ghrera, C. M. Pandey, M. A. Ali, B. D. Malhotra, Appl. Phys.
Keywords Lett. 2015, 106, 19.
[32] M. C. Horny, M. Lazerges, J. M. Siaugue, A. Pallandre, D. Rose,
electrochemical detection, flow modeling, microfluidic design, F. Bedioui, C. Deslouis, A. M. Haghiri-Gosnet, J. Gamby, Lab Chip
microfluidic electrodes, microfluidics 2016, 16, 4373.
[33] L. Ianeselli, G. Grenci, C. Callegari, M. Tormen, L. Casalis, Biosens.
Received: May 13, 2021 Bioelectron. 2014, 55, 1.
Revised: July 7, 2021 [34] Q. Li, C. Zhao, Z. Zheng, S. Weng, Q. Chen, Q. Liu, X. Lin,
Published online: J. Electroanal. Chem. 2017, 788, 23.
[35] G. V. Soraya, J. Chan, T. C. Nguyen, D. H. Huynh, C. D. Abeyrathne,
G. Chana, M. Todaro, E. Skafidas, P. Kwan, Biosens. Bioelectron.
2018, 111, 174.
[1] J. Wang, Talanta 2002, 56, 223. [36] T. García, M. Revenga-Parra, L. Añorga, S.arana, F. Pariente,
[2] M. A. Schwarz, P. C. Hauser, Lab Chip 2001, 1, 1. E. Lorenzo, Sens. Actuat. B- Chem. 2012, 161, 1030.
[37] S. A. Hong, Y. J. Kim, S. J. Kim, S. Yang, Biosens. Bioelectron. 2018, [68] L. P. Zaino, N. M. Contento, S. P. Branagan, P. W. Bohn,
107, 103. ChemElectroChem 2014, 1, 1570.
[38] A. Becerra-Hernández, J. Galindo-de-la Rosa, Y. Martínez- [69] M. K. Rashed, M. A. Mohd Salleh, H. A. Abdulbari, M. H. Shah
Pimentel, J. Ledesma-García, L. Álvarez-Contreras, M. Guerra- Ismail, S. Izhar, ChemBioEng Rev. 2015, 2, 356.
Balcázar, A. Aguilar-Elguezabal, A. Álvarez, A. U. Chávez-Ramírez, [70] J. F. Pérez, J. Llanos, C. Sáez, C. López, P. Cañizares, M. A. Rodrigo,
V. Vallejo-Becerra, Biochem. Eng. J. 2019, 149, 107240. Electrochem. Commun. 2017, 82, 85.
[39] G. D. O’Neil, S. Ahmed, K. Halloran, J. N. Janusz, A. Rodríguez, [71] A. S. Munshi, R. S. Martin, Analyst 2016, 141, 862.
I. M. Terrero Rodríguez, Electrochem. Commun. 2019, 99, 56. [72] H. Tateno, Y. Matsumura, K. Nakabayashi, H. Senboku, M. Atobe,
[40] E. Bitziou, M. E. Snowden, M. B. Joseph, S. J. Leigh, RSC Adv. 2015, 5, 98721.
J. A. Covington, J. V. MacPherson, P. R. Unwin, J. ElectroAnal. [73] Y. H. Kwok, A. C. H. Tsang, Y. Wang, D. Y. C. Leung, J. Power
Chem. 2013, 692, 72. Sources 2017, 349, 75.
[41] D. Padmaraj, R. Pande, J. H. Miller, J. Wosik, W. Zagozdzon-Wosik, [74] M. Odijk, W. Olthuis, A. Van Den Berg, L. Qiao, H. Girault, Anal.
PloS One 2014, 9, 7. Chem. 2012, 84, 9176.
[42] S. Viswanathan, T. N. Narayanan, K. Aran, K. D. Fink, J. Paredes, [75] M. M. Mensack, J. B. Wydallis, N. S. Lynn, D. S. Dandy,
P. M. Ajayan, S. Filipek, P. Miszta, H. C. Tekin, F. Inci, U. Demirci, C. S. Henry, Lab Chip 2013, 13, 208.
P. Li, K. I. Bolotin, D. Liepmann, V. Renugopalakrishanan, Mater. [76] D. Morales-Acosta, H. Rodríguez G., L. A. Godinez, L. G. Arriaga,
Today 2015, 18, 513. J. Power Sources 2010, 195, 1862.
[43] A. Kara, A. Reitz, J. Mathault, S. Mehou-Loko, M. A. Amirdehi, [77] L. Qiang, S. Vaddiraju, J. F. Rusling, F. Papadimitrakopoulos,
A. Miled, J. Greener, Lab Chip 2016, 16, 1081. Biosens. Bioelectron. 2010, 26, 682.
[44] L. Rassaei, P. S. Singh, S. G. Lemay, Anal. Chem. 2011, 83, 3974. [78] O. A. Ibrahim, P. Alday, N. Sabaté, J. P. Esquivel, E. Kjeang, J. The
[45] T. Holm, M. Ingdal, J. R. Strobl, E. V. Fanavoll, S. Sunde, F. Seland, Electrochem. Soc. 2017, 164, A2448.
D. A. Harrington, Electrochim. Acta 2017, 229, 452. [79] X. Lu, D. Y. C. Leung, H. Wang, M. M. Maroto-Valer, J. Xuan,
[46] T. Holm, J. A. Diaz Real, W. Mérida, J. Electroanal. Chem. 2019, Renewable Energy 2016, 95, 277.
842, 115. [80] F. T. Van Den Brink, L. Büter, M. Odijk, W. Olthuis, U. Karst, A. Van
[47] C. Amatore, A. Oleinick, I. Svir, Electrochem. Commun. 2004, 6, Den Berg, Anal. Chem. 2015, 87, 1527.
1123. [81] A. Chałupniak, A. Merkoçi, ACS Appl. Mater. Interfaces 2017, 9,
[48] N. Dossi, R. Toniolo, E. Piccin, S. Susmel, A. Pizzariello, 44766.
G. Bontempelli, Electroanalysis 2013, 25, 2515. [82] J. Kuleshova, J. T. Hill-Cousins, P. R. Birkin, R. C. D. Brown,
[49] Y. Huang, Y. Han, Y. Gao, J. Gao, H. Ji, Q. He, J. Tu, G. Xu, D. Pletcher, T. J. Underwood, Electrochim. Acta 2011, 56, 4322.
Y. Zhang, L. Han, Electrochim. Acta 2021, 373, 137921. [83] J. A. Díaz-Real, E. Ortiz-Ortega, M. P. Gurrola, J. Ledesma-Garcia,
[50] D. P. Wasalathanthri, V. Mani, C. K. Tang, J. F. Rusling, Anal. Chem. L. G. Arriaga, Electrochim. Acta 2016, 206, 388.
2011, 83, 9499. [84] A. Dector, N. Arjona, M. Guerra-Balcázar, J. P. Esquivel, F. J. Del
[51] C. V. Uliana, C. R. Peverari, A. S. Afonso, M. R. Cominetti, Campo, N. Sabaté, J. Ledesma-García, L. G. Arriaga, Fuel Cells
R. C. Faria, Biosens. Bioelectron. 2018, 99, 156. 2014, 14, 810.
[52] G. H. Lee, J. K. Lee, J. H. Kim, H. S. Choi, J. Kim, S. H. Lee, [85] R. S. Jayashree, L. Gancs, E. R. Choban, A. Primak, D. Natarajan,
H. Y. Lee, Sci. Rep. 2017, 7, 1. L. J. Markoski, P. J. Kenis, J. Am. Chem. Soc. 2005, 127, 16758.
[53] L. Zhu, X. Liu, J. Yang, Y. He, Y. Li, Anal. Chem. 2020, 92, 11981. [86] S. Rarotra, T. K. Mandal, D. Bandyopadhyay, Energy Technol. 2017,
[54] B. Wolfrum, E. Kätelhön, A. Yakushenko, K. J. Krause, N. Adly, 5, 1208.
M. Hüske, P. Rinklin, Acc. Chem. Res. 2016, 49, 2031. [87] A. Power, B. White, A. Morrin, Electrochim. Acta 2013, 104, 236.
[55] J. H. Weng, C. Y. Lai, L. C. Chen, Electrochim. Acta 2019, 297, 118. [88] E. Kjeang, B. Roesch, J. McKechnie, D. A. Harrington, N. Djilali,
[56] M. G. Straver, M. Odijk, W. Olthuis, A. Van Den Berg, Lab Chip D. Sinton, Microfluid. Nanofluid. 2007, 3, 403.
2012, 12, 1548. [89] C. J. Huang, J. L. Lin, P. H. Chen, M. J. Syu, G. B. Lee,
[57] D. Ye, Y. Yang, J. Li, X. Zhu, Q. Liao, B. Zhang, 20th World Electrophoresis 2011, 32, 931.
Hydrogen Energy Conference, WHEC 2014, Committee of WHEC, [90] X. Liu, N. L. Abbott, Anal. Chem. 2009, 81, 772.
Gwangju, South Korea 2014, p. 522. [91] E. Kjeang, A. G. Brolo, D. A. Harrington, N. Djilali, D. Sinton,
[58] E. Sinkala, J. E. McCutcheon, M. J. Schuck, E. Schmidt, J. Electrochem. Soc. 2007, 154, B1220.
M. F. Roitman, D. T. Eddington, Lab Chip 2012, 12, 2403. [92] A. Ebrahimi Khabbazi, A. J. Richards, M. Hoorfar, J. Power Sources
[59] S. Sansuk, E. Bitziou, M. B. Joseph, J. A. Covington, M. G. Boutelle, 2010, 195, 8141.
P. R. Unwin, J. V. MacPherson, Anal. Chem. 2013, 85, 163. [93] Y. Kotb, S.-E. K. Fateen, J. Albo, I. Ismail, J. Electrochem. Soc. 2017,
[60] E. V. Fanavoll, D. A. Harrington, S. Sunde, G. Singh, F. Seland, 164, E391.
Electrochim. Acta 2017, 225, 69. [94] Y. Li, W. Van Roy, L. Lagae, P. M. Vereecken, Electrochim. Acta 2017,
[61] N. Wongkaew, S. E. K. Kirschbaum, W. Surareungchai, R. A. Durst, 231, 200.
A. J. Baeumner, Electroanalysis 2012, 24, 1903. [95] D. E. Molina, A. S. Medina, H. Beyenal, C. F. Ivory, J. Electrochem.
[62] A. L. Kadilak, Y. Liu, S. Shrestha, J. R. Bernard, W. E. Mustain, Soc. 2019, 166, B125.
L. M. Shor, J. Electroanal. Chem. 2014, 727, 141. [96] G. Pireddu, I. Svir, C. Amatore, A. Oleinick, Electrochim. Acta 2021,
[63] B. V. Chikkaveeraiah, H. Liu, V. Mani, F. Papadimitrakopoulos, 365.
J. F. Rusling, Electrochem. Commun. 2009, 11, 819. [97] C. Amatore, A. Oleinick, I. Svir, N. Da Mota, L. Thouin, Nonlinear
[64] S. Gu, Y. Lu, Y. Ding, L. Li, H. Song, J. Wang, Q. Wu, Biosens. and Anal. Model. Control 2006, 11, 345.
Bioelectron. 2014, 55, 106. [98] T. Holm, S. Sunde, F. Seland, D. A. Harrington, J. Electroanal.
[65] Y. Li, W. Van Roy, P. M. Vereecken, L. Lagae, Microelectron. Eng. Chem. 2015, 745, 72.
2017, 181, 47. [99] J. C. Abrego-Martínez, A. Moreno-Zuria, F. M. Cuevas-Muñiz,
[66] M. Fighera, P. D. Van Der Wal, H. Shea, J. Electrochem. Soc. 2017, L. G. Arriaga, S. Sun, M. Mohamedi, J. Power Sources 2017,
164, H836. 371, 10.
[67] S. Kang, A. F. Nieuwenhuis, K. Mathwig, D. Mampallil, [100] E. Ko, V. K. Tran, Y. Geng, W. S. Chung, C. H. Park, M. K. Kim,
S. G. Lemay, ACS Nano 2013, 7, 10931. G. H. Jin, G. H. Seong, J. Electroanal. Chem. 2017, 792, 72.
[101] R. M. Tiggelaar, F. Benito-López, D. C. Hermes, H. Rathgen, [135] Y. Sameenoi, M. M. Mensack, K. Boonsong, R. Ewing,
R. J. Egberink, F. G. Mugele, D. N. Reinhoudt, A. van den Berg, W. Dungchai, O. Chailapakul, D. M. Cropek, C. S. Henry, Analyst
W. Verboom, H. J. Gardeniers, Chem. Eng. J. 2007, 131, 163. 2011, 136, 3177.
[102] A. Déctor, J. P. Esquivel, M. J. González, M. Guerra-Balcázar, [136] F. Della Pelle, L. Vázquez, M. Del Carlo, M. Sergi, D. Compagnone,
J. Ledesma-García, N. Sabaté, L. G. Arriaga, Electrochim. Acta A. Escarpa, Chem. Eur. J. 2016, 22, 12761.
2013, 92, 31. [137] A. K. Rathoure, H. Pramanik, Int. J. Hydrogen Energy 2016, 41, 15287.
[103] B. Nearingburg, A. L. Elias, RSC Adv. 2014, 4, 29975. [138] Z. Wang, H. Fan, H. Liang, J. Ma, S. Li, Y. Song, R. Wang,
[104] R. R. Unocic, R. L. Sacci, G. M. Brown, G. M. Veith, Electrochim. Acta 2017, 230, 245.
N. J. Dudney, K. L. More, F. S. Walden, D. S. Gardiner, J. Damiano, [139] F. Terzi, B. Zanfrognini, S. Ruggeri, N. Dossi, G. M. Casagrande,
D. P. Nackashi, Microsc. Microanal. 2014, 20, 452. E. Piccin, Sens. Actuators, B 2017, 245, 352.
[105] H. Han, Z. Zheng, D. Pan, C. Wang, X. Hu, Y. Wang, Micro Nano [140] M. Guerra-Balcázar, F. M. Cuevas-Muñiz, F. Castaneda, R. Ortega,
Lett. 2018, 13, 302. L. Álvarez-Contreras, J. Ledesma-García, L. G. Arriaga, Electrochim.
[106] J. N. Lee, C. Park, G. M. Whitesides, Anal. Chem. 2003, 75, 6544. Acta 2011, 56, 8758.
[107] W. Adamiak, D. Kałuża, M. Jönsson-Niedziolka, Microfluid. [141] D. Li, H. Xu, L. Zhang, D. Y. C. Leung, F. Vilela, H. Wang, J. Xuan,
Nanofluid. 2016, 20, 9. Int. J. Hydrogen Energy 2016, 41, 10249.
[108] M. Matteucci, T. L. Christiansen, S. Tanzi, P. F. Ostergaard, [142] N. Arjona, A. Palacios, A. Moreno-Zuria, M. Guerra-Balcázar,
S. T. Larsen, R. Taboryski, Microelectron. Eng. 2013, 111, 294. J. Ledesma-García, L. G. Arriaga, Chem. Commun. 2014, 50, 8151.
[109] J. J. Lv, M. Jouny, W. Luc, W. Zhu, J. J. Zhu, F. Jiao, Adv. Mater. 2018, [143] T. Akyazi, L. Basabe-Desmonts, F. Benito-Lopez, Anal. Chim. Acta
30, 1. 2018, 1001, 1.
[110] C. A. Teixeira, G. F. Giordano, M. B. Beltrame, L. C. S. Vieira, [144] A. Martín, A. Escarpa, Microchim. Acta 2017, 184, 307.
A. L. Gobbi, R. S. Lima, Anal. Chem. 2016, 88, 11199. [145] A. Gevaerd, E. Y. Watanabe, C. Belli, L. H. Marcolino-Junior,
[111] A. J. Armenta-González, R. Carrera-Cerritos, A. Moreno-Zuria, M. F. Bergamini, Sensor. Actuat. B- Chem. 2021, 332, 2020.
L. Álvarez-Contreras, J. Ledesma-García, F. M. Cuevas-Muñiz, [146] M. Salve, A. Mandal, K. Amreen, P. K. Pattnaik, S. Goel,
L. G. Arriaga, Fuel 2016, 167, 240. Microchem. J. 2020, 157, 104973.
[112] E. R. Choban, J. S. Spendelow, L. Gancs, A. Wieckowski, P. J. Kenis, [147] Y. Hong, M. Wu, G. Chan, Z. Dai, Y. Zhang, G. Chen, X. Dong,
Electrochim. Acta 2005, 50, 5390. ACS Appl. Mater. Interfaces 2016, 8, 32940.
[113] E. R. Choban, P. Waszczuk, P. J. Kenis, Electrochem. Solid-State [148] G. W. Bishop, J. E. Satterwhite, S. Bhakta, K. Kadimisetty,
Lett. 2005, 8, A348. K. M. Gillette, E. Chen, J. F. Rusling, Anal. Chem. 2015, 87, 5437.
[114] B. Wouters, J. Hereijgers, W. De Malsche, T. Breugelmans, [149] A. D. Castiaux, E. R. Currens, R. S. Martin, Analyst 2020, 145, 3274.
A. Hubin, Electrochim. Acta 2016, 210, 337. [150] A. Altima, H. Panagopoulou, A. Economou, C. Kokkinos, Anal.
[115] E. Behroodi, H. Latifi, Z. Bagheri, E. Ermis, S. Roshani, M. Salehi Chim. Acta 2021, 1159.
Moghaddam, Sci. Rep. 2020, 10, 1. [151] L. C. Duarte, C. L. Chagas, L. E. Ribeiro, W. K. Coltro, Sensor.
[116] D. J. Guckenberger, T. E. De Groot, A. M. Wan, D. J. Beebe, Actuat. B- Chem. 2017, 251, 427.
E. W. Young, Lab Chip 2015, 15, 2364. [152] R. D. Sochol, E. Sweet, C. C. Glick, S. Y. Wu, C. Yang, M. Restaino,
[117] J. Adkins, K. Boehle, C. Henry, Electrophoresis 2015, 36, 1811. L. Lin, Microelectron. Eng. 2018, 189, 52.
[118] J. Ma, A. S. Gago, N. Alonso-Vante, J. Electrochem. Soc. 2013, 160, [153] R. Pol, F. Céspedes, D. Gabriel, M. Baeza, Sensor. Actuat. B- Chem.
F859. 2019, 290, 364.
[119] O. Scialdone, A. Galia, S. Sabatino, D. Mira, C. Amatore, [154] R. M. Cardoso, C. Kalinke, R. G. Rocha, P. L. dos Santos,
ChemElectroChem 2015, 2, 684. D. P. Rocha, P. R. Oliveira, B. C. Janegitz, J. A. Bonacin,
[120] C. K. Chua, M. Pumera, Electroanalysis 2013, 25, 945. E. M. Richter, R. A. Munoz, Anal. Chim. Acta 2020, 1118, 73.
[121] X. Lu, D. Y. C. Leung, H. Wang, J. Xuan, Appl. Energy 2017, 194, [155] A. I. Shallan, P. Smejkal, M. Corban, R. M. Guijt, M. C. Breadmore,
549. Anal. Chem. 2014, 86, 3124.
[122] M. Medina-Sánchez, S. Miserere, S. Marín, G. Aragay, A. Merkoçi, [156] B. C. Gross, J. L. Erkal, S. Y. Lockwood, C. Chen, D. M. Spence,
Lab Chip 2012, 12, 2000. Anal. Chem. 2014, 86, 3240.
[123] H. Becker, C. Gärtner, Anal. Bioanal. Chem. 2008, 390, 89. [157] J. L. Erkal, A. Selimovic, B. C. Gross, S. Y. Lockwood, E. L. Walton,
[124] O. Scialdone, A. Galia, S. Sabatino, G. M. Vaiana, D. Agro, S. McNamara, R. S. Martin, D. M. Spence, Lab Chip 2014, 14, 2023.
A. Busacca, C. Amatore, ChemElectroChem 2014, 1, 116. [158] P. Panjan, V. Virtanen, A. M. Sesay, Analyst 2018, 143, 3926.
[125] A. Dector, F. M. Cuevas-Muñiz, M. Guerra-Balcázar, L. A. Godínez, [159] C.-H. Yang, C.-W. Chen, Y.-K. Lin, Y.-C. Yeh, C.-C. Hsu, Y.-J. Fan,
J. Ledesma-García, L. G. Arriaga, Int. J. Hydrogen Energy 2013, 38, I.-S. Yu, J.-Z. Chen, J. Electrochem. Soc. 2017, 164, B534.
12617. [160] E. A. Hayter, A. D. Castiaux, R. S. Martin, Anal. Methods 2020, 12,
[126] H. Pramanik, A. K. Rathoure, Int. J. Hydrogen Energy 2017, 42, 5340. 2046.
[127] C. Renault, M. J. Anderson, R. M. Crooks, J. Am. Chem. Soc. 2014, [161] J. A. Díaz-Real, J. Ma, N. Alonso-Vante, Appl. Catal. B. 2016, 198, 471.
136, 4616. [162] T. Hisatomi, J. Kubota, K. Domen, Chem. Soc. Rev. 2014, 43, 7520.
[128] M. Gaines, M. J. Gonzalez-Guerrero, K. Uchida, F. A. Gomez, [163] J. A. Diaz-Real, T. Holm, N. Alonso-Vante, Current Trends and
Electrophoresis 2018, 39, 2131. Future Developments on (Bio–) Membranes (Eds: A. Basile,
[129] A. D. Davydov, V. M. Volgin, Russ. J. Electrochem. 2016, 52, 806. T. W. Napporn), Elsevier, Amsterdam 2020, pp. 255–289, Ch. 9.
[130] T. Saito, Y. Tsujimoto, Y. Miyamoto, N. Okamoto, K. Kondo, Jpn. [164] C. Byrne, G. Subramanian, S. C. Pillai, J. Environ. Chem. Eng. 2018,
J. Appl. Phys. 2015, 54, 5. 6, 3531.
[131] A. Mohammadzadeh, A. Fox-Robichaud, P. R. Selvaganapathy, [165] F. E. Osterloh, Chem. Mater. 2008, 20, 35.
Sensors 2020, 20, 3. [166] A. Kudo, Y. Miseki, Chem. Soc. Rev. 2009, 38, 253.
[132] Y. Li, C. Sella, F. Lemaître, M. GuilleCollignon, L. Thouin, [167] M. E. Schiavello, in Photoelectrochemistry, Photocatalysis and
C. Amatore, Electroanalysis 2013, 25, 895. Photoreactors Fundamentals and Developments, 1st ed., Springer,
[133] A. Farzbod, H. Moon, Biosens. Bioelectron. 2018, 106, 37. The Netherlands 1985.
[134] A. M. Gaikwad, J. W. Gallaway, D. Desai, D. A. Steingart, [168] C. Wang, Z. H. Sheng, J. Ouyang, J. J. Xu, H. Y. Chen, X. H. Xia,
J. Electrochem. Soc. 2011, 158, A154. ChemPhysChem 2012, 13, 762.
[169] N. Wang, X. Zhang, B. Chen, W. Song, N. Y. Chan, H. L. W. Chan, [180] N. L. Ritzert, W. Li, C. Tan, G. G. Rodríguez-Calero, J. Rodríguez-
Lab Chip 2012, 12, 3983. López, K. Hernández-Burgos, S. Conte, J. J. Parks, D. C. Ralph,
[170] W. Xu, C. Ma, P. W. Bohn, ChemElectroChem 2016, 3, 422. H. D. Abruña, Faraday Discuss. 2014, 172, 27.
[171] N. M. M. Pires, T. Dong, U. Hanke, N. Hoivik, Sensors (Switzerland) [181] C. Renault, J. Roche, M. R. Ciumag, T. Tzedakis, S. Colin,
2014, 14, 15458. K. Serrano, O. Reynes, C. André-Barrès, P. Winterton, J. Appl.
[172] A. Gencoglu, A. R. Minerick, Microfluid. Nanofluid. 2014, 17, 781. Electrochem. 2012, 42, 667.
[173] X. Zhang, C. Chen, J. Li, L. Zhang, E. Wang, Anal. Chem. 2013, 85, 5335. [182] N. M. Contento, P. W. Bohn, Biomicrofluidics 2014, 8, 1.
[174] P. Podešva, X. Liu, P. Neužil, Sens. Actuators, B 2019, 286, 282. [183] K. Pandian, M. Ajanth Praveen, S. Z. Hoque, A. Sudeepthi,
[175] A. Perro, G. Lebourdon, S. Henry, S. Lecomte, L. Servant, S. Marre, A. K. Sen, Biomicrofluidics 2020, 14, 6.
React. Chem. Eng. 2016, 1, 577. [184] C. Christianson, N. N. Goldberg, D. D. Deheyn, S. Cai, M. T. Tolley,
[176] N. I. Andersen, K. Artyushkova, I. Matanović, D. P. Hickey, Sci. Robot. 2018, 3, 1.
S. D. Minteer, P. Atanassov, ChemElectroChem 2019, 6, 246. [185] M. A. Goulet, M. Eikerling, E. Kjeang, Electrochem. Commun. 2015,
[177] G. L. Nelson, A. M. Lines, A. J. Casella, J. M. Bello, S. A. Bryan, 57, 14.
Analyst 2018, 143, 1188. [186] C. A. Goss, D. H. Charych, M. Majda, Anal. Chem. 1990, 63, 85.
[178] A. M. Lines, G. L. Nelson, A. J. Casella, J. M. Bello, S. B. Clark, [187] S. Qian, H. H. Bau, Mech. Res. Commun. 2009, 36, 10.
S. A. Bryan, Anal. Chem. 2018, 90, 2548. [188] G. Wu, P. Tan, X. Wu, L. Peng, H. Cheng, C. F. Wang, W. Chen,
[179] I. A. Ges, K. P. M. Currie, F. Baudenbacher, Biosens. and Bioelectron. Z. Yu, S. Chen, Adv. Funct. Mater. 2017, 27, 1.
2012, 34, 30. [189] Z. Herrasti, F. Martínez, E. Baldrich, Lab Chip 2015, 15, 3269.
Tianyu Li received his B.Sc. degree in materials science at Nanjing University. He completed
his Ph.D. degree in chemistry at the University of Victoria, specializing in the microfluidic
devices for online detection of alcohol electrooxidation with Raman spectroscopy. He is
currently a postdoctoral researcher at the University of Toronto, working on electrochemical
microfluidic biosensors.
Jesus Adrián Díaz-Real received his Ph.D. degree in electrochemistry in 2016 from the Center
for Research and Technological Development in Electrochemistry (CIDETEQ), working in
semiconductors and microfluidic cells. He worked in photoelectrochemistry during his doctoral
stages at Université de Poitiers, and at the University of British Columbia as a postdoctoral fellow.
Currently, he is a researcher in the energy coordination at CIDETEQ, México.
Thomas Holm received his Ph.D. degree in electrochemsitry in 2016 at the Norwegian University
of Science and Technology, specializing in method development in electrochemistry for
microfluidic detection. He spent 4 years as a postdoctoral researcher at University of Victoria
and University of British Columbia in Canada, and he is currently a researcher at the Institute for
Energy Technology in Kjeller, Norway.

Li 2021

Uploaded by

Copyright:

Available Formats

Li 2021

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Li 2021

Uploaded by

Copyright:

Available Formats

Review

Design of Electrochemical Microfluidic Detectors: A Review

microfluidic devices have become an inte-

Table 1. Common shapes of microfluidic channels.

Shape Purpose Application(s) Typical Reference(s)

method. These problems can be further enhanced due to the 5. Summary

4.4. Electrode Stability and Regeneration

In microfluidic cells, the electrode itself and its stability deter-

You might also like