See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/334421924

Formation of thin film Ag/AgCl reference electrode by electrochemical and

chemical method

Conference Paper · May 2019

DOI: 10.23919/MIPRO.2019.8756828

CITATION READS

1 1,031

5 authors, including:

Nina Lokar Danilo Vrtacnik

University of Ljubljana University of Ljubljana
2 PUBLICATIONS   1 CITATION    88 PUBLICATIONS   762 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

COMPETE - Chair Of MicroProcess Engineering and TEchnology View project

All content following this page was uploaded by Nina Lokar on 17 July 2019.

The user has requested enhancement of the downloaded file.

 Formation of thin film Ag/AgCl reference
electrode by electrochemical and chemical
method
D. Resnik, B. Pečar, M. Možek, N. Lokar and D. Vrtačnik

University of Ljubljana, Faculty of Electrical Engineering, Laboratory of Microsensor Structures and Electronics,
Tržaška 25, SI-1000 Ljubljana, Slovenia
Corresponding author: [email protected]

Abstract - The formation and analyses of thin film Commonly used laboratory type AgCl reference
Ag/AgCl reference electrode by two different methods were electrodes have liquid junction, i.e. inner electrolyte fill
studied with aim to extend the lifetime and to determine the reservoir and are not convenient for miniaturization. The
potential stability over time. To convert part of thin Ag film same confines also the use of another laboratory reference
into AgCl, chemical formation in FeCl3 of different electrode such as standard calomel electrodes (SCE), often
molarities and electrochemical formation in solution of 3 M used in electrochemistry.
NaCl were employed. Optical and SEM analyses of formed
Ag/AgCl microstructure revealed that both methods result In laboratory environment Ag/AgCl electrode is
in uniform distribution of polycrystalline grains with size usually a silver wire with deposited AgCl with various
between 1-3 micrometers. Potential stability measurements techniques and dipped in fill electrolyte composed of 3.5
were performed by measuring the potential of thin film M KCl, saturated with AgCl. The conductive path from
reference electrode against standard calomel liquid junction electrode to bulk electrolyte is via frit glass, which enables
electrode. When measuring potential difference in high only ionic transport to/from reference electrode, without
molarity electrolyte (3 M NaCl), lifetime of thin film mixing of the filling and bulk electrolyte. Effusion, i.e.
electrodes was found to be limited to few hours, though in time dependent transfer of inner fill electrolyte toward
weak solutions it can be maintained up to several days. bulk electrolyte solution, which is actually a part of
Preliminary results showed good potential stability (<1%
conductive process is a drawback for liquid junction
change per hour). It was shown that additional passivation
electrodes, since fill electrolyte has to be replenished
of AgCl layer by a thin PDMS layer significantly increased
the lifetime of the electrode.
periodically [3,4]. When miniature electrochemical cell
setup is required, the very inconvenient reservoirs for
Keywords – thin film reference electrode; Ag/AgCl; liquid junction, filled with electrolyte or gel is required,
chemical and electrochemical formation, potential stability which inevitably includes solving of fluidic problems like
hermeticity and frequent inner electrolyte replenishment
in miniature compartments. Screen printed thick film
I. INTRODUCTION
AgCl electrodes are somewhere in between, are cost
Thin film microelectrodes are very convenient for the effective, exhibit lower accuracy and are mainly used for
realization of miniaturized electrochemical cell. By disposable electrode systems [5].
implementing microfabrication techniques, planar metal
electrodes with well defined patterns can be incorporated Solid-state reference thin film electrodes without any
within microfluidic platform, leading toward micrototal internal reference electrolyte are preferable solution when
analysis systems. In the miniaturization process toward it comes to miniaturization, however, there are still several
e.g. biosensors, detecting on the basis of electrochemical issues remained unsolved to reach the reliability of
response, three electrode configuration is commonly previous. Beside the limited lifetime due to very thin
employed. Here, working electrode (WE), where the AgCl layer, they are susceptible to changing activity of its
electrochemical processes are taking place is commonly a primary ion, e.g.Cl- and also to the interference with redox
thin film Au, covered by e.g. biological samples, counter couples present in the bulk electrolyte. [6].
electrode (CE) is usually made of thin film Pt, while the The objective of this study was to compare the two
reference electrode (RE) is usually a thin film Ag methods of forming thin film Ag/AgCl reference electrode
converted to Ag/AgCl reference electrode. In many ways, and to determine which are the dominant parameters
in metrological applications the reference potential of during formation that affect the electrode lifetime and its
Ag/AgCl electrodes defines the measurement uncertainty potential stability.
[1]. Requirements for reference electrode are to allow high
exchange current densities, to be highly nonpolarizable,
II. MATERIALS AND METHODS
meaning that its role in electrochemical cell is to provide
stable potential during measurements and also to allow for The substrate material used in experimental work were
control of WE [2], where all the kinetics occur. thermally oxidized high resistivity (500 ohmcm) Si
wafers, four inch in diameter, with dc sputtered Ag/Ni/Ti
thin film layers (1000 nm/350 nm/20 nm). Here, Ti layer Ag
acts as adhesive prime layer between SiO2 and
subsequently deposited Ni and Ag layers. Ni buffer layer
was implemented according to its beneficial effect on the
durability of AgCl electrode presumably due to
suppressed Ti and O diffusion toward Ag, causing
relatively pore free Ag/AgCl surface as shown by Kim
[7]. Finally, wafers were thermally annealed and
subsequently diced in 20x7 mm2 samples. 3M NaCl
Thin
Prior to use, individual samples were cleaned in 1 M film Pt
HNO3 solution for 10 min, thoroughly rinsed by DI water
and dried by nitrogen. This was found mandatory in order AgCl
to commence the process of electrode formation uniformly
across the entire Ag electrode surface. Two approaches 10 mm
were utilized in transformation of Ag into AgCl. Ref. AgCl
With chemical method, samples were chlorinated by Figure 2. Experimental setup with thin film AgCl, liquid fill SCE and
vertical dipping to approximately two thirds into prepared Pt CE in NaCl electrolyte.
FeCl3 solutions of defined molarity (molarity range 50
mM to 1 M) for a certain period of time (5 min to 60 min) multimeter/data acquisition system. The experimental
and then rinsed thoroughly by DI water and dried with N2. electrochemical cell setup where forming and measuring
was performed is shown in Fig. 2.
Samples for electrochemical formation were cleaned
by the same method and then subdued to a procedure of III. RESULTS AND DISCUSSION
electrochemical oxidation of Ag in an electrolyte solution.
During formation, samples were dipped vertically to a A. Chemical formation
defined depth in order to know the exact exposed area to Realization of solid state Ag/AgCl reference electrode
calculate the current density during formation. Selected involves chemical oxidation of top Ag thin film layer in
forming electrolyte in our case was 3 M NaCl solution. the presence of chloride ions to form polycrystalline
The counter electrode used in anodization process was Pt AgCl, but still leaving beneath an electronically
wire coil. The setup is presented in Fig. 1a.Measurement conductive Ag part. Ferric chloride FeCl3 was used as a
of electric current during formation period was performed source of chloride ions. The following reaction takes place
indirectly by measuring the voltage drop on resistor, and in the chlorination of Ag:
the current density that actually dictates the
electrooxidation of Ag to AgCl was taken as the main
parameter of formation. Current densities between 0.1 and Ag + FeCl3 → AgCl + FeCl2 (1)
2 mA/cm2 were adjusted in the preliminary experiments.
The formation time was varied between 5 min and 60 min.
To study the properties of AgCl formed under different
Lower current densities would require prolonged conditions, samples were prepared by varying the molarity
formation times to obtain sufficiently thick AgCl layer, of ferric chloride and by varying the time of soaking in the
while higher current densities proved inappropriate since solution. Beside optical microscopy and SEM
they produced AgCl prone to rapid reaction with room characterization of AgCl microstructure, i.e. grain size and
moisture and further reduction to Ag. After formation, its lateral distribution, the most adequate method to
thin film AgCl electrodes were characterized by means of evaluate efficiency and appropriateness of formed
measuring their potential in bulk electrolyte against Ag/AgCl electrode is to directly measure the potential
commercial SCE electrode. Cell potential between the difference and its time variations against commercial
electrodes was measured with Keithley 2700
reference electrode (Fig. 1b). In our case standard calomel
DC
electrode was used as a reference electrode. Generally,
Power Supply reference electrodes in electrochemical cells are
performing in a more than decade lower bulk electrolyte
concentrations and last therefore longer, however, to
V
VR R A
determine the lifetime of our thin film electrodes, highly
concentrated electrolyte was applied in our case to shorten
the measuring time. All potential measurements presented
Ag bellow were done in 3 M NaCl bulk electrolyte solution.
3M NaCl (xM NaCl or KCl) Fig. 3 shows the time dependent behaviour of potential of
fabricated electrodes vs. SCE electrode for four different
a) Pt Ag b) AgCl Ref: SCE molarities of FeCl3 and same forming time. It can be
interpreted in terms of increasing the thickness of formed
AgCl with increased chloride ion concentration.
Figure 1. a) electrochemical AgCl forming setup, b) thin film AgCl
potential stability measuring setup.
 0.050 Initial current density in the range between 0.1 and 0.5
t= 10 min 0.05M FeCl3 mA/cm2 was adjusted to silver thin film layer for each
0.045 0.1M FeCl3 sample separately and monitored during AgCl formation
0.5M FeCl3
0.040
time.
1M FeCl3
E SCE vs AgCl [V]

In Fig.5 it can be observed that current density

0.035
decreased rapidly during initial period when the reaction
0.030 starts (20-30%) and only ca. 5-10% during rest 30 min
period of formation.
0.025
Decrease of current density was gradually after initial
0.020 drop and most probably corresponds to increase of ionic
resistance due to polycrystalline AgCl formation. During
0.015
0 20 40 60 80 100 120 140 160 180
AgCl growth, the interface with Ag layer beneath has to
be reached by chlorine ions via grain boundaries or
Measuring time [min]
formed microchannels [8]. The process of anodization and
Figure 3. Time stability of potential vs. molarity of FeCl3 during growth of crystals is in the initial stage very fast and
formation. intensive what is typical for diffusion limited processes.
Formed samples were evaluated by measuring the
0.040 potential of electrochemically formed Ag/AgCl vs. SCE
0.5M FeCl3 t=5 min reference electrode. Before each measurement, initial
t=10 min
0.035 t=20 min stabilization time (5-10 min) is always mandatory in order
t=30 min that electrode wetting process is accomplished. Fig. 6
E SCE vs AgCl [V]

0.030
shows the representative results of measurements by
comparing samples formed at 0.2 and 0.5 mA/cm2. The
duration of stable potential is proportional to thickness of
0.025 formed AgCl, i.e. current density magnitude during
formation being the dominant parameter. Three fold
0.020 increase of lifetime time can be observed at criterion of 2
mV tolerance.
0.015
0 50 100 150 200 250 300
0.8 J=0.1 mA/cm2
Measuring time [min] J=02 mA/cm2
Figure 4. Time stability of potential vs. formation time in 0.5 M FeCl3. J=0.3 mA/cm2
0.6 J=0.4 mA/cm2
J [mA/cm2]

Diluted FeCl3 resulted in slower formation rate and J=0.5 mA/cm2

thus in thinner AgCl layer which is consequently
0.4
dissolved in a shorter period of measurement. The optical
inspection could not reveal notable molarity dependent
difference of microstructure such as grain size or porosity; 0.2
therefore it is believed that formed AgCl thickness is
mainly responsible for poor lifetime. However, high
molarity FeCl3 (1 M) revealed that AgCl growth does not 0.0
follow proportionally the increase of molarity as was the 0 5 10 15 20 25 30
case between 0.05 M and 0.5 M, which needs further Formation time [min]
study. As shown, highest lifetime was achieved in 0.5 M Figure 5. Exchange current density during AgCl formation.
FeCl3 and was taken for further experiments to determine
formation time dependence on Ag/AgCl electrode 0.06
lifetime. Results are shown in Fig. 4. The lifetime is
J=0.2 mA/cm2
increasing proportionally with soaking time between 5 and J=0.5 mA/cm2
20 min, though prolonged time of 30 min showed lower 0.05
lifetime. The lifetime criterion was defined as the time
E SCE vs AgCl [V]

where the potential exceeded 2 mV differences as

compared to initial value. Each potential runaway also 0.04
corresponds to change of sample colour, i.e. from
brownish to silver bright. It should be noted that variations
in baseline potential (at t=0) can be attributed to molarity 0.03
change due to evaporation during prolonged consecutive 40 min (+-1 mV) 120 min (+-1 mV)
measurements.
0.02
0 20 40 60 80 100 120 140 160 180
B. Electrochemical formation
Measuring time [min]
Electrochemical formation was performed in 3 M
NaCl electrolyte by variation of current density and time Figure 6. Potential stability of samples formed at 0.2 and 0.5 mA/cm2.
of formation according to the setup shown in Fig. 1a.
 It should be noted that separate experiments with 0.10
higher current densities (>1 mA/cm2) revealed that formed PDMS-25 °C
0.09 PDMS-80 °C
AgCl layer is prone to reaction with moisture, forming
acidic product on the surface of the film with pH<2. In 0.08

E SCE vs AgCl [V]

contrast, this was not observed after chemical formation of
AgCl in FeCl3 even in cases of high molarity (1 M). 0.07

Thickness of formed AgCl layer was determined by 0.06

using Faraday’s law modified expression [9], which gives 0.05
AgCl thickness of 0.8 µm when anodizing at J=0.5
mA/cm2 for 10 min. Accordingly, this leaves still around 0.04
420 nm of electronically conductive Ag layer to remain
beneath the AgCl. 0.03
0 10 20 30 40 50 60 70
Measuring time [min]
C. Role of protective layer
The dissolution of AgCl being directly exposed to high Figure 7. Influence of PDMS curing temperature on stability of
potential.
molarity bulk electrolytes is a well known problem in
electrochemistry and is even more evident when thin film
0.040
materials are involved. According to our experimental
AgCl + PDMS
work, also purely chemical dissolution of thin film AgCl AgCl
in 3 M NaCl occurs within few hours of immersion, 0.039
formed in 0.5 M FeCl3
without any other electrode present in the electrolyte.
E SCE vs AgCl [V]
measured in 3M NACl
Adding dissolved Ag(+) ions into electrolyte solution 0.038
somewhat mitigate this effect but might interfere the
measurements. There exist various approaches in the 0.037
literature, how to prevents excessive AgCl dissolution. In ∆=0.7 mV
general, these approaches suggest to cover the AgCl with 0.036
additional protective layer or a combination of layers
(forming dry ionic bridge), that reduce effectively the
0.035
AgCl dissolution and at the same time allow free ionic 0 2 4 6 8 10 12 14 16 18
transport to/from reference electrode. Materials such as
Measuring time [hours]
gel saturated with KCl [6], graphene oxide [10], SAM
layers [11] and several others were proposed and applied Figure 8. Long term stability of potential between PDMS-passivated
in different applications, though with various success. AgCl and bare AgCl electrode.

To enhance the durability of AgCl electrodes by It is well known that the equilibrium potential of the
means of protective, ion conductive layer, electrode in electrochemical cell is defined by the Nernst
polydimethylsiloxane (PDMS) was applied on formed equation. Briefly, in our case AgCl electrode is said to
AgCl surface to obtain a layer cca 100 µm thick. Fig. 7 exhibit Nernstian behavior when potential decreases by
shows the measured potential vs. SCE reference electrode approximately 59 mV per decade of change in chlorine
for the case when incomplete polymerization was ion concentration at 25 °C [5]. To correlate our electrodes
performed (PDMS cured at room temperature for 12 versus theoretical value a potential difference between
hours) and after the sample was additionally cured for 1 h thin film AgCl and SCE reference at 6 different electrolyte
at 80 °C. Due to incomplete polymerization, severe molarities were measured, each over 10 min period (Fig.
fluctuations of potential resulted, while after completing 9). It is noteworthy that the potential stability is more
the PDMS polymerization, the noise has completely affected at decreasing electrolyte concentration.
vanished and the stable potential was established.
Furthermore, the long term measurements for a period of 0.04
3M NaCl
16 hours is shown in Fig. 8 and compared to sample with 0.02 2M
bare AgCl, formed under same conditions. Variation less 0.00
1M
0.5M
than 0.5 mV was achieved over the 16 h period of 0.1M
E SCE vs AgCl [V]

measurement in 3M NaCl electrolyte, which is an order of -0.02

0.02M
magnitude improvement over bare Ag/AgCl samples -0.04
shown also in Fig. 3-4 and Fig. 6. PDMS thermal curing -0.06
to attain sufficient degree of polymerization has to be
-0.08
accomplished to obtain stable behavior. Not only lifetime
but as well baseline of potential and variations over period -0.10
of 16 hours remained within expectations. However, after -0.12
this period, chemical attack on PDMS and lack of
-0.14
adhesion were optically observed, resulting in potential 0 2 4 6 8 10
instability. In contrast to very low lifetime observed in
preliminary study in highly concentrated NaCl, we have Measuring time [min]
successfully proved that the application of cover layers Figure 9. Baseline potential and time stability at various molarities.
extremely increased the life time as shown below.
 log NaCl molarity [M] since it is formed on metallic Ag substrate (having lattice
0.01 0.1 1 10 constant a=4.079 Å).
40
Beside the abovementioned surface topography
20
difference, both methods produce comparable porosity of
0 microstructure and crystallites of comparable grains size
being between 1 and 3µm. Moreover, by measuring the
E SCE vs AgCl [V]

-20
potential of two typical Ag/AgCl electrodes, prepared by
-40
the two methods and shown in Fig. 12, only a minor
-60 slope= 61 mV/decade difference could be observed regarding the time stability
-80 or baseline potential (<0.5 mV). Latter could be easily
-100
attributed to change of solution molarity or temperature
variation. Initial transient, observed in Fig. 12 is due to
-120
insufficient stabilization time, i.e. caused by immersion of
-140 dry electrodes into the electrolyte. These two samples
0 1 2 3 were without protective PDMS layer.
NaCl molarity [M]
0.0365
Figure 10. Dependence of potential versus molarity of NaCl electrolyte. chemically formed
0.0360 electrochemically formed
By taking into account values at t=10 min, the plot in measured in 3M NaCl
Fig. 10 was drawn, which shows the relationship between

E SCE vs AgCl [V]

0.0355
measured potential and concentration of Cl- ions.
0.0350
For our thin film AgCl electrode, Nernstian behavior
was confirmed by obtaining a slope of 61 mV/decade as 0.0345
compared to theoretical value of 59 mV/decade of Cl- ∆=0.3 mV
concentration in bulk electrolyte. Results presented in Fig. 0.0340
10 are for chemically formed sample (0.5 M FeCl3, t=20
min), but very similar was obtained also for 0.0335
electrochemically formed samples. 0 10 20 30 40 50 60
Measuring time [min]
D. Comparison between chemically and Figure 12. Comparison of chemically and electrochemically formed
electrochemically formed AgCl thin film AgCl reference electrode.
Optical and SEM analyses of formed Ag/AgCl were
performed in order to compare the microstructure of both IV. CONCLUSION
preparation method and to determine the grain uniformity
and grain size. Fig. 11 shows typical SEM micrographs of Thin film Ag/AgCl reference electrodes were prepared
AgCl microstructure obtained by the presented two by chemical formation in FeCl3 at different molarities and
methods. by electrochemical formation at different current densities
in solution of 3 M NaCl with aim to extend the lifetime
and improve the potential stability over time.
Optical and SEM analyses of formed Ag/AgCl
microstructure revealed that both methods result in
uniform distribution of polycrystalline grains with size
between 1-3 micrometers. Electrode potential stability
against commercial SCE liquid junction electrode showed
<1% change per hour. Lifetime of as formed Ag/AgCl
thin film electrodes in high molarity electrolyte (3 M
Figure 11. Polycrystalline grains of AgCl formed by chemical and by NaCl) was found to be limited to a few hours, however,
electrochemical method (3 M NaCl, t=10 min, J=1.5 mA/cm2). passivation layer of PDMS improved this significantly
(>16 hours).
In contrast to chemically formed samples, AgCl
microstructure of electrochemically prepared samples Nernstian behaviour of electrode was confirmed by
showed more pronounced and nonuniform outward obtaining a slope of 61 mV/decade as compared to the
protrusions. This could be due to a large expansion and theoretical value of 59 mV/decade of Cl- concentration in
mechanical stress evolved through electrochemical bulk electrolyte. Treatment in ferric chloride was found
formation involving higher charge transfer. Both methods more convenient and as experiments showed was by no
generally produce AgCl crystallites via the oxidation means inferior to electrochemical anodization method
reaction of Ag with Cl ions though the process might not regarding the potential stability or electrode lifetime.
be exactly the same. The growth inevitably causes volume Baseline potential vs. SCE was independent on the
expansion during AgCl formation at the interface. The method of preparation, when measured by as-formed
AgCl microstructure (having lattice constant a=5.54 Å), Ag/AgCl and also when measured after applying a PDMS
which is actually a salt, is subjected to compressive stress passivation on the top of Ag/Cl layer.
 ACKNOWLEDGMENT variations of thick-film screen printed silver/silver chloride
potentiometric reference electrodes,” Sensors and Actuators A:
This work was supported by Slovene Research Physical, 197, pp. 1-8, 2013.
Agency (Grant No P2-0244) and by the project [6] I-Y. Huang, R.-S. Huang, L.-H. Lo, “Improvement of integrated
BioPharm.SI. Ag/AgCl thin-film electrodes by KCl-gel coating for ISFET
applications,” Sensors and Actuators B, 94, pp. 53-64, 2003.
[7] H.R. Kim, Y.D. Kim, K.I. Kim, J.H. Shim, H.Nam, B.K. Kang,
REFERENCES “Enhancement of physical and chemical properties of thin film
[1] P. J. Brewer and R. J. C. Brown, “Effect of Structural Design of Ag/AgCl reference electrode using a Ni buffer layer”,Sensors and
Silver/Silver Chloride Electrodes on Stability and Response Time Actuators B 97, pp. 348–354, 2004.
and the Implications for Improved Accuracy in pH Measurement,” [8] H. Ha, .J. Payer, “The effect of silver chloride formation on the
Sensors, 9, pp. 118-130, 2009. kinetics of silver dissolution in chloride solution,” Electrochimica
[2] P. Khullar, J. V. Badilla, and R. G. Kelly, “The Use of a Sintered Acta, 56, pp. 2781–2791, 2011.
Ag/AgCl Electrode as Both Reference and Counter Electrode for [9] F. Pargar, H. Kolev, D. and K. van Breugel, “Microstructure,
Electrochemical Measurements in Thin Film Electrolytes,” surface chemistry and electrochemical response of Ag|AgCl
Electrochemistry Letters,10, C31-C33, 2015. sensors in alkaline media,” J. Mater. Sci., 53,:pp. 7527–7550,
[3] H. Suzuki and H. Shiroishi, “Microfabricated Liquid Junction 2018.
Ag/AgCl Reference Electrode and Its Application to a One-Chip [10] T. Y. Kim, S. A. Hong and S. Yang, “A Solid-State Thin-Film
Potentiometric Sensor,” Anal. Chem., 71 (22), pp 5069– Ag/AgCl Reference Electrode Coated with Graphene Oxide and
5075,1999. Its Use in a pH Sensor,” Sensors, 15(3), pp. 6469-6482, 2015.
[4] I. Shitanda, H. Kiryu, M. Itagaki, “Improvement in the long-term [11] T. Rahman and T Ichiki, “Fabrication and Characterization of a
stability of screen-printed planar type solid-state Ag/AgCl Stabilized Thin Film Ag/AgCl Reference Electrode Modified with
reference electrode by introducing poly(dimethylsiloxane) liquid Self-Assembled Monolayer of Alkane Thiol Chains for Rapid
junction,” Electrochimica Acta, 58,pp. 528–531, 2011. Biosensing Applications,” Sensors, 17(10), pp. 2326, 2017.
[5] M. Glanc, M. Sophocleous, J. Atkinson, and E. Garcia-
Breijo,”The effect on performance of fabrication parameter

View publication stats