Delft University of Technology

Determination of Chloride Content in Cementitious Materials

From Fundamental Aspects to Application of Ag/AgCl Chloride Sensors
Pargar, Farhad; Koleva, Dessi; van Breugel, Klaas

DOI
10.3390/s17112482
Publication date
2017
Document Version
Final published version
Published in
Sensors

Citation (APA)
Pargar, F., Koleva, D., & van Breugel, K. (2017). Determination of Chloride Content in Cementitious
Materials: From Fundamental Aspects to Application of Ag/AgCl Chloride Sensors. Sensors, 17(11), [2482].
[Link]

Important note
To cite this publication, please use the final published version (if applicable).
Please check the document version above.

Copyright
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent
of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Takedown policy
Please contact us and provide details if you believe this document breaches copyrights.
We will remove access to the work immediately and investigate your claim.

This work is downloaded from Delft University of Technology.

For technical reasons the number of authors shown on this cover page is limited to a maximum of 10.
 sensors
Review
Determination of Chloride Content in Cementitious
Materials: From Fundamental Aspects to Application
of Ag/AgCl Chloride Sensors
Farhad Pargar * ID
, Dessi A. Koleva and Klaas van Breugel
Faculty of Civil Engineering and Geosciences, Department Materials and Environment,
Delft University of Technology, Stevinweg 1, 2628 CN Delft, The Netherlands; [Link]@[Link] (D.A.K.);
[Link]@[Link] (K.v.B.)
* Correspondence: [Link]@[Link]; Tel.: +31-15-27-81742

Received: 20 September 2017; Accepted: 26 October 2017; Published: 29 October 2017

Abstract: This paper reports on the advantages and drawbacks of available test methods for the
determination of chloride content in cementitious materials in general, and the application of
Ag/AgCl chloride sensors in particular. The main factors that affect the reliability of a chloride
sensor are presented. The thermodynamic behaviour of silver in the presence or absence of chloride
ions is described and kinetic restrictions are addressed. The parameters that can affect the activity of
chloride ions in the medium and/or the rate of ion exchange and dissolution/precipitation processes
at the sensor’s surface are also considered. In this regard, the contribution of morphology and
microstructure of the AgCl layer, binding of chloride ions and the compactness of hydration products
around the chloride sensor are highlighted. The important parameters for a reliable sensor’s response
are discussed and the possible causes of inaccuracies are evaluated.

Keywords: Ag/AgCl electrode; chloride sensor; concrete; cement; chloride

1. Introduction
Chloride ions are among the primary causes of steel corrosion in reinforced concrete structures.
Determination of the chloride content in the concrete cover and near the steel reinforcement is needed
for evaluating the risk of corrosion. The drawbacks of conventional techniques for determination
of the chloride content call for novel and reliable techniques. One of the first documented attempts
for non-destructive determination of the chloride content in cementitious materials was reported in
the early 1990s, when Ag/AgCl chloride sensors were embedded in mortar specimens [1]. However,
the literature on the performance of chloride sensors in cementitious materials is still scarce and the
interpretation of sensors’ readings is far from straightforward.
The performance of embedded chloride sensor in cementitious materials depends on the physical
condition of the interfaces (Ag/AgCl/cement paste) at the sensor’s surface as well as the pore solution
composition. In this regard, a schematic presentation of Ag/AgCl sensor in cement paste is depicted
in Figure 1. This figure illustrates some morphological and microstructural features of the AgCl layer
and adjacent cementitious material, both important for the sensor’s response. The pores at the surface
of the AgCl layer provide a pathway for penetration of ions into the layer (Figure 1a). The presence of
different ions in the pore channels of this layer, subsequently affects the sensor’s response (Figure 1b).
When the chloride sensor is embedded in cementitious materials (Figure 1c), the microstructure of
these cementitious materials is also important for the electrochemical reactions relevant to the sensor’s
response (Figure 1d,e). This is mainly in view of pore solution composition and compactness of
hydration products at the interface between sensor and cementitious materials.

Sensors 2017, 17, 2482; doi:10.3390/s17112482 [Link]/journal/sensors

Sensors 2017, 17, 2482 2 of 22
Sensors 2017, 17, 2482 2 of 21

[Link]
Figure Schematicrepresentation
representation of chloride
chloride sensor
sensorembedded
embeddedinina acementitious
cementitiousmatrix.
matrix.

The open circuit potential (OCP) of a chloride sensor is a function of the surface characteristics
The open circuit potential (OCP) of a chloride sensor is a function of the surface characteristics
of the sensor and the rate of electron/ion transfer at the sensor’s surface [2]. In this regard, the
of the sensor and the rate of electron/ion transfer at the sensor’s surface [2]. In this regard,
surrounding medium can affect the thermodynamically plausible reactions at the chloride sensor’s
the surrounding medium can affect the thermodynamically plausible reactions at the chloride sensor’s
surface. For example, the binding of chloride ions to the hydration products, together with the
surface. For example, the binding of chloride ions to the hydration products, together with the relative
relative distribution and compactness of hydration products around the sensor, make the chloride
distribution and compactness
ions less “effective” in the pore of solution
hydration products
than expected around
based the sensor,
on their make the chloride
concentration. Moreover,ionstheless
“effective” inofthe
sensitivity pore solution
a chloride sensor than
to theexpected based
chloride ions on their
depends onconcentration.
the concentration Moreover, the sensitivity
of interfering species,
of such
a chloride sensor to the chloride ions depends on the concentration of interfering
as hydroxide and sulphide ions in the medium. These processes subsequently limit the rate species, such
of as
hydroxide and sulphide ions in the medium. These processes subsequently limit the rate
reactions at the sensor’s surface. A better insight into these parameters is important for understanding of reactions
at the
the chloride
sensor’s sensor’s
surface. response
A better (Figure
insight 1).
intoTherefore,
these parameters
the cementis important
chemistry forandunderstanding
composition ofthe
chloride sensor’s
cementitious response
materials are(Figure
reviewed1). Therefore, the cement
in the following chemistry and composition of cementitious
sections.
materials
In are
viewreviewed in the following
of the factors sections.
outlined above, this paper follows a sequence of steps to determine
chloride
In viewions’
of theand sensors’
factors responses
outlined in this
above, a concrete environment.
paper follows Figure
a sequence of 1steps
can to
bedetermine
consideredchloride
as a
schematic
ions’ presentation
and sensors’ of theindifferent
responses a concreteinfluencing parameters,
environment. Figureinterfaces
1 can beand points of interest
considered with
as a schematic
regard to aof
presentation Ag/AgCl sensorinfluencing
the different performance. This workinterfaces
parameters, encompassesandfive main
points of sections, structured
interest with regardasto a
follows:
Ag/AgCl sensor performance. This work encompasses five main sections, structured as follows:
 In the first section, the main hydration products of cement paste in the bulk matrix and at the
• In the first section, the main hydration products of cement paste in the bulk matrix and at the
sensor–cement paste interface are described. The composition of the pore solution of
sensor–cement paste interface are described. The composition of the pore solution of cementitious
cementitious materials, i.e., the various ions in contact with the sensor’s surface and hence of
materials, i.e., the various ions in contact with the sensor’s surface and hence of importance for
importance for sensors’ readings, is also assessed.
sensors’ readings, is also assessed.
 Next, the classification of chloride ions in cementitious materials is reviewed and the importance
• Next, the classification
of chloride sensors forof chloride
local ions
chloride in cementitious
measurements, materials
specifically atis reviewed
the andreinforcement,
level of the the importance
of is
chloride sensors for local chloride measurements, specifically at the level of the reinforcement,
discussed.
 is discussed.
Section 4 presents the different techniques for the determination of chloride content in concrete
• Section
and the4 presents
advantagestheofdifferent
chloridetechniques for the determination
sensors in particular of chloride
for free chloride content in concrete
determination.
and the advantages of chloride sensors in particular for free chloride determination.
Sensors 2017, 17, 2482 3 of 22

• The next section bridges fundamental considerations and technical background, both related
to the performance, characterization and application of Ag/AgCl sensors. The advantages and
drawbacks of the chloride sensors are discussed. The thermodynamic behaviour of silver and the
electrochemical kinetics of AgCl formation are discussed. Then, the relevance of these parameters
to the reliability of the sensors’ response in cementitious materials is evaluated. The above
framework outlines the structure of this paper, highlighting the principles, related mechanisms
and determining factors for monitoring of chloride ions in cementitious materials.

2. Important Properties of Cementitious Materials in View of Chloride Measurements via

Ag/AgCl Sensors
A cementitious paste is a heterogeneous material with high alkalinity, a pore solution with different
composition, and a pore system with different porosity and pore size distribution (Figure 1d,e) [3].
In the following, some properties of cementitious materials that are important for the response of
chloride sensor are presented. The main hydration products and microstructure of cementitious
materials are described (Section 2.1). Moreover, the concentration of interfering ions, i.e., hydroxide
and sulphide ions (Figure 1b) in the pore solution of Portland cement paste and blast furnace slag
cement paste is presented (Section 2.2). The slag cement paste is considered because it contains a high
amount of water soluble sulphates, and sulphide ions [4].

2.1. Hydration Products and Microstructure of Cement Paste

Hardened Portland cement paste is a porous material containing calcium hydroxide (Ca(OH)2 ,
Portlandite), aluminate compounds, such as monosulphate, hydroxyl-AFm, ettringite (AFt) and
unhydrated cement particles, surrounded by an amorphous hydration product, known as C-S-H
(calcium silicate hydrate) [5].
The heterogeneity of hydration products and the microstructure of cementitious materials are
important factors in view of the reliability and reproducibility of the sensor’s response (Figure 1d,e).
If these factors are not considered, the performance of a chloride sensor will remain uncertain. These
concerns are currently not well understood. However, they can be assessed from the available studies
on the interfacial properties between the cement paste and the aggregate. The microstructure and
hydration products of cement paste in the vicinity of an aggregate differ from those further away
from the aggregate. This distinctive region is termed the interfacial transition zone (ITZ) [5]. When
the cement grains encounter the “wall” of the aggregate, a region of higher porosity and different
composition near the aggregate surface appears [6]. The porosity of ITZ is two to three times higher
than the porosity in the bulk [7]. The wall effect creates a water concentration gradient around the
aggregate, i.e., a w/c ratio gradient. A locally higher w/c ratio with fewer nucleation sites forms larger
and preferentially oriented crystals in contact with aggregates [7]. In this region, ettringite and calcium
hydroxide tend to form larger crystals in oriented layers [5]. A rim of massive calcium hydroxide
can often be observed around the aggregates. This implies that the ITZ properties at the interface
between the sensor and cementitious materials are the principal factor influencing the performance of
the chloride sensor.

2.2. Ions in Pore Solution

As shown in Figure 1b, the composition of the pore solution of cementitious materials represents
the ions that are in contact with the sensor. Hydroxide and sulphide ions are the two main ions of
cement paste that can interfere with the chloride sensor’s response. In the following, the concentration
of these ions in the pore solution of Portland cement and slag cement is reviewed.

2.2.1. Portland Cement

The aqueous electrolyte in the spaces not filled by solid hydration products is known as the pore
solution [8]. The chloride sensor in cementitious material, similar to embedded reinforcement, is in
 Sensors 2017, 17, 2482 4 of 21
Sensors 2017, 17, 2482 4 of 22

different cement phases and solubility of these products. Once cement is mixed with water and starts
to dissolve,
contact with every element
the pore that The
solution. is present in the cement
composition can solution
of the pore also be found in the
depends on form of dissolved
the hydration of
ions in the
different pore phases
cement solution.
and solubility of these products. Once cement is mixed with water and starts
The pore
to dissolve, everysolution of Portland
element that is presentcement paste can
in the cement is mainly composed
also be found in theofform
alkaline hydroxides
of dissolved ions
in(mainly
the poreKOH and NaOH). Other species, such as SO4 and Ca , are present in much lower
solution. 2− 2+

concentrations [9–12]. of Portland cement paste is mainly composed of alkaline hydroxides (mainly
The pore solution
The change
KOH and NaOH). Other in thespecies,
alkalinity
suchofasthe 2− and
SOpore solution during
Ca2+ , are cement
present hydration
in much depends on the
lower concentrations w/c
[9–12].
4
ratioThe
andchange
cementintype [13]. Alkalis
the alkalinity of in
thePortland cement
pore solution (K2O and
during Na2hydration
cement O) play a significant
depends on role
thefor the
w/c
high pH range of the pore solution (13.2 to 13.8) [14]. The hydroxide concentration
ratio and cement type [13]. Alkalis in Portland cement (K2 O and Na2 O) play a significant role for the in cement paste
withpH
high CEM I 42.5N
range and
of the w/csolution
pore = 0.5 increases
(13.2 tofrom
13.8)150 mM
[14]. The athydroxide
7 h to 540 mM after 29 days
concentration [13]. paste
in cement
with CEM I 42.5N and w/c = 0.5 increases from 150 mM at 7 h to 540 mM after 29 days [13].
2.2.2. Blast Furnace Slag Cement
2.2.2. Blast Furnace Slag Cement
The alkalinity of the pore solution in slag cement paste (CEM III/B, w/c = 0.5, pH = 13.1) is lower
thanThe
thatalkalinity
of Portland ofcement
the pore solution
paste (CEM in slag= 0.5,
I, w/c cement
pH =paste (CEM
13.9) [3]. For III/B,
slag cement,w/c =the 0.5,concentration
pH = 13.1)
isoflower than
alkaline ionsthat
(Naof+ Portland
and K+) and cement paste ions
hydroxide (CEM I, w/c =during
decreases 0.5, pH the= slag
13.9)hydration
[3]. For slag cement,
process. This
the + +
concentration
results of alkalineof
in a pH reduction ions
the(Na
poreand K ) and
solution forhydroxide
slag cement ionspaste,
decreasesto aduring
value thelowerslagthan
hydration
that of
Portland
process. cement
This paste
results in a[3,15].
pH reduction of the pore solution for slag cement paste, to a value lower than
that ofThe pore solution
Portland cement composition
paste [3,15]. of slag cement paste is to a large extent similar to that of Portland
cement
The paste. The difference
pore solution is a remarkable
composition of slag cementamount
pasteofissulphide
to a largeionsextent (S2−similar
) in slagtocement paste [4].
that of Portland
The sulphide
cement paste. Thecontent in slag is about
difference 1%, while
a remarkable in Portland
amount cement
of sulphide ions 2− ) inthan
it is(Slower slag0.1%
cement[16].paste
After[4].
10
h ofsulphide
The hydration, sulphide
content ionsiswere
in slag about detected
1%, whilein the pore solution
in Portland cement of blended
it is lower cement paste[16].
than 0.1% containing
After
10slag
h of(CEM III/B 42.5
hydration, L) [4].ions
sulphide Thewere
concentration
detected inof thesulphide ions increased
pore solution of blended tocement
5 mM after
paste seven days,
containing
and(CEM
slag then stabilized
III/B 42.5at L)8–11 mMconcentration
[4]. The [4]. The concentration
of sulphide of hydroxide
ions increasedand to sulphide
5 mM afterions seven
is important
days,
for the
and thenperformance
stabilized atof8–11 an Ag/AgCl
mM [4]. Thechloride sensor (Figure
concentration 1b), as will
of hydroxide and besulphide
reviewedions andisdiscussed
importantin
Section
for 6.
the performance of an Ag/AgCl chloride sensor (Figure 1b), as will be reviewed and discussed in
Section 6.
3. Chloride Ions in Concrete
3. Chloride Ions in Concrete
3.1. Classification of Chloride Ions
3.1. Classification of Chloride Ions
The ingress of chloride ions in concrete takes place via the pore system and can be affected by
The ingress
the binding of chloride
of chloride ions ions
to theinhydration
concrete takes
[Link]
viawell-known
the pore system and can of
classification bechloride
affectedions
by
the binding of chloride ions to the hydration products. The well-known classification
in cement-based materials is that of chemically bound chloride, physically bound chloride and free of chloride ions
inchloride
cement-based
ions. The materials
chlorideisions
that can
of chemically bound chloride,
be bound chemically physicallylike
in compounds bound chloride
Friedel’s saltand free
(calcium
chloride ions. The chloride ions can be bound chemically in compounds
chloroaluminate hydrate) or adsorbed physically at the surface of cement hydration products, such like Friedel’s salt (calcium
chloroaluminate
as C-S-H, Friedel’s hydrate) or adsorbed
salt and physically
Portlandite. at the surface
The chloride ions inof cement
the bulk porehydration
solutionproducts, suchfree
are called as
C-S-H, Friedel’s salt and Portlandite. The chloride ions in the bulk pore solution
chloride (Figure 2). The binding of chloride ions to the cement hydration products continues up to a are called free chloride
(Figure
level at2).which
The binding of chloride
an equilibrium ions to bound
between the cement hydration
(chemically andproducts
physically) continues
and freeup to a level atis
chlorides
which an equilibrium between bound (chemically and physically) and free chlorides is reached.
reached.

[Link]
Figure Schematicof
ofmain
mainhydration
hydrationproducts
productsofofcement
cementpaste
pasteand
andtheir
theircontribution
contributiontotothe
thebinding
binding
ofofchloride
chlorideions.
ions.

[Link].[17]
Glass [17]classified
classified
thethe chloride
chloride ions
ions in the
in the concrete
concrete porepore solution
solution into into
four four categories—
categories—free,
free, loosely bound, bound and strongly bound chlorides. The amount of
loosely bound, bound and strongly bound chlorides. The amount of free chloride can be measured free chloride can be
measured by the “equilibrium method”. In this method, concrete samples are stored in a solution of
Sensors 2017, 17, 2482 5 of 22

by the “equilibrium method”. In this method, concrete samples are stored in a solution of known
chloride concentration until equilibrium is reached between the external solution and the concrete
pore solution. For example, the time required for a cement paste sample with 1 cm thickness to reach
equilibrium can extend up to a year. In equilibrium condition, the free chloride in the concrete sample
is equal to the chloride concentration in the external solution. The chloride is “loosely bound” when it
can be extracted from the concrete samples under high pressure (in the range of hundreds of MPa).
The physically bound chloride can be considered as loosely bound chloride. The term “bound chloride”
refers to the chloride that cannot be extracted from the concrete samples under pressure. This holds
for the chemically bound chloride. The traditional method for determination of total chloride content
in a concrete sample is “acid-soluble chloride extraction” [18–20]. Extraction of total chloride content
using this method may not end with release of all the bound chlorides. Therefore, the total chloride
content of the sample can be underestimated [21]. The part of bound chloride that is not released
by the acid-soluble extraction method is known as “strongly bound chloride”. Although most of the
phases containing chloride compounds dissolve in the acidic medium of the test [18–20], the release of
“strongly bound” chloride may require a higher acidity.

3.2. Free Chloride

There are several techniques for the evaluation of the free chloride content. Different test procedures,
applied in these techniques, result in different outcomes for the free chloride content. For instance,
chemical extraction of chloride ions from a concrete sample gives a significantly higher amount of
chloride than extraction under high pressure [22]. Determination of free chloride content is discussed
in Section 4.2 with respect to chloride sensors’ response.
Section 3.3 reviews main aspects of the classification of bound chlorides in a cementitious matrix.
These are important in view of the general debates about the determination of chloride content and
also the sensors’ application. The available techniques for the determination of chloride content in
cementitious materials are described in Section 4.

3.3. Bound Chloride

3.3.1. Chemical Binding of Chloride

Friedel’s salt is the main reaction product of chemical binding of chloride ions in concrete. It forms
due to the reaction between the chloride ions and hydration products of C3 A, e.g., hydroxyl-AFm,
monosulphate and AFt phases [23]. It is assumed that all aluminate hydrates transform to Friedel’s
salt with increasing chloride concentration in the pore solution [24]. In general, the tendency of
sulphate ions (SO4 2− ) to bind in hydration products is higher than that of chloride and hydroxide
ions, i.e., SO4 2− > Cl− >> OH− [25]. However, the concentration of sulphate ions (SO4 2− ) in the pore
solution of mature (28 days) cement paste is low (Table 1). Therefore, chloride ions can react with the
AFm phases to form Friedel’s salt. Even the sulphate-containing hydration products (monosulfate and
Aft) convert to Friedel’s salt if the chloride concentration in the pore solution is sufficiently high [26].

Table 1. Concentration of ions in the pore solution of hydrated Portland cement paste (w/c = 0.5) after
a 28-day curing period [9].

Concentration (mmol/L)
Na+ K+ Ca2+ Cl− SO4 2− Ionic strength (mmol/L)
Hydrated cement paste 88.9 74.8 4.8 1.4 0.2 178.3

3.3.2. Physical Binding of Chloride

Similar to most other materials, the surface of hydration products becomes electrically charged
when in contact with a polar solution such as water. The ionic binding properties of hydration
Sensors 2017, 17, 2482 6 of 21

3.3.2. Physical Binding of Chloride

Sensors 2017, 17, 2482 6 of 22
Similar to most other materials, the surface of hydration products becomes electrically charged
when in contact with a polar solution such as water. The ionic binding properties of hydration
productsgovern
products governthe the dissolution
dissolution of surface
of surface species
species or adsorption
or adsorption of ionsoffrom
ionsthefrom the solution.
solution. This
This process
process electrically
electrically charges thecharges
surfacetheofsurface of the hydrated
the hydrated phases. phases.
IfIfwe
welook
lookat atthe
theinterface
interfaceof ofaahydration
hydrationproduct,
product,weweseeseean
anaccumulation
accumulationof ofchloride
chlorideionsionsand
andaa
depletionofofcations
depletion cationsnearnearthethe charged
charged surface
surface of hydration
of hydration products
products and and
in theinvicinity
the vicinity of adsorbed
of adsorbed layer
layer
of cationsof cations
(Figure(Figure
3) [27,28].3) In
[27,28].
the bulk In the bulk solution,
solution, electroneutrality
electroneutrality prevails
prevails (Figure 3) (Figure 3) [28].
[28]. Although
Although
the layer ofthe layer of bound
physically physically bound
chloride chloride
into into the
the solution solution
is not more is notamore
than thanitacan
few nm, fewcomprise
nm, it can a
comprise a considerable
considerable amountions
amount of chloride of chloride
[9]. ions [9].
Asmentioned
As mentionedininSection
Section 3.1,3.1,
thethe physically
physically boundbound chloride
chloride (Figure
(Figure 3) can3)becan be “squeezed
“squeezed out”
out” when
when
the cementthe cement
paste is paste is subjected
subjected to high and/or
to high pressure pressure and/orloads
thermal thermal loads
[17,29]. [17,29]. the
Therefore, Therefore,
physicallythe
physically
bound bound
chloride haschloride has been
been classified asclassified
a loosely as a loosely
bound bound
chloride chloride
[17]. [17]. random
The erratic The erratic random
movement
movement
of physically of bound
physically bound
chloride is chloride
continuouslyis continuously
affected byaffected by the surrounding
the surrounding molecules molecules [30].
[30]. In this
In this condition,
condition, the of
the activity activity
chlorideof chloride ions increases
ions increases with increasing
with increasing distance from distance from the
the surface surface of
of hydration
hydration
product product
(Figure (Figure Therefore,
3) [30,31]. 3) [30,31]. Therefore,
physicallyphysically boundcan
bound chloride chloride can be considered
be considered as free
as free chloride
chloride
ions, ions, low
but with but with low activity.
activity.

Figure 3. Schematic diagram of physical binding of chloride ions to the cement hydration product
Figure 3. Schematic diagram of physical binding of chloride ions to the cement hydration product
exposedto
exposed tochloride
chlorideenvironment.
environment. The
The adsorption
adsorption of
of chloride
chloride ions
ionssubsequently
subsequently induces
induces aachloride
chloride
concentration gradient between the bulk solution and the surface of a hydration product.
concentration gradient between the bulk solution and the surface of a hydration product.

The chloride binding mechanisms and microstructure of cementitious materials are also
The chloride binding mechanisms and microstructure of cementitious materials are also important
important for determination of the so-called chloride threshold value. The threshold value, i.e., the
for determination of the so-called chloride threshold value. The threshold value, i.e., the value of the
value of the critical chloride content for the corrosion initiation of reinforcing steel, is still heavily
critical chloride content for the corrosion initiation of reinforcing steel, is still heavily debated. In the
debated. In the next section, several aspects of determination of the chloride threshold value
next section, several aspects of determination of the chloride threshold value are presented.
are presented.
3.4. Chloride Threshold Value
3.4. Chloride Threshold Value
The process of steel corrosion in reinforced concrete can be divided into two phases, initiation
The process of steel corrosion in reinforced concrete can be divided into two phases, initiation
and propagation [23,32–34]. The first stage, i.e., initiation, is generally related to the time needed for a
and propagation [23,32–34]. The first stage, i.e., initiation, is generally related to the time needed for
critical chloride concentration to reach the steel bar and initiate corrosion. The subsequent propagation
a critical chloride concentration to reach the steel bar and initiate corrosion. The subsequent
stage extends up to the time when the corrosion damage is beyond the acceptable limits and cracks
propagation stage extends up to the time when the corrosion damage is beyond the acceptable limits
due to the expansion of corrosion products appearing on the concrete surface.
and cracks due to the expansion of corrosion products appearing on the concrete surface.
An impressive number of studies on the chloride threshold value have been published since the
An impressive number of studies on the chloride threshold value have been published since
1960s [35]. As shown in Table 2, a unique chloride threshold value does not exist. Similarly, a generally
the 1960s [35]. As shown in Table 2, a unique chloride threshold value does not exist. Similarly, a
accepted or standardized method for determination of the critical chloride content does not exist as
generally accepted or standardized method for determination of the critical chloride content does not
well. Even the concept of chloride threshold value has been questioned [36,37]. This controversy is
exist as well. Even the concept of chloride threshold value has been questioned [36,37]. This
mainly due to the variety of influencing factors, such as concrete mixture, environmental factors and
surface condition of the embedded reinforcing steel. Apart from these factors, different techniques and
Sensors 2017, 17, 2482 7 of 22

criteria have been used for both the detection of the time to steel depassivation and measuring the
chloride content close to the reinforcement. The measured chloride content is presented in different
units and related to the concrete or binder weight (Table 2). As the binder content is not always known,
it is sometimes preferred to present the total chloride content as a percentage of the weight of concrete.
The free chloride content is presented in different ways: as a percentage of the binder or concrete
weight, as mole per litre of concrete pore solution (mol/L), or as the ratio of chloride to hydroxide ions
([Cl− ]/[OH− ]) in the pore solution (Table 2).

Table 2. Different expression forms of the critical chloride content and the range of reported chloride
threshold values for steel, embedded in cement-based materials.

Aggressive Species Expressed as Reported Chloride Threshold Value Reference

Total chloride (including % by weight of binder (1.24–3.08), (1–8.34), (0.68–0.97) [38–40]
free chloride) % by weight of concrete (0.03–0.07), (0.03–0.2), (0.06–0.2) (0.06–0.37) [41–44]
% by weight of binder (0.39–1.16), (1-4), (0.5–2) [33,38,39]
% by weight of concrete 1 0.026 [45]
Free chloride
mol/L (0.36–3.22), (0.44–0.65), (0.045–0.55) [46–48]
[Cl− ]/[OH− ] (1.17–3.98), (1.7–20), (3–20) [38,39,49]
1 Expression of critical free chloride content by weight of concrete is scarcely available in the literature.

Some researchers claimed that the chloride threshold value should be expressed as the total
chloride content [34]. Many authors believe, however, that bound chloride does not impose any risk
for corrosion initiation [50]. Still, the chloride threshold value is quantified in terms of either total or
free chloride content (Table 2).
Actually, measuring the free chloride content in the cement paste, mortar or concrete is
complicated. This is mainly due to the limitations of the available techniques for extraction of free
chloride ions from the cementitious materials. As a result, the total chloride content is generally used
for representation of the chloride threshold value.
The different methods for the determination of chloride content also affect the variation in the
reported chloride threshold values. Traditionally, one or more cores from the concrete cover are taken
at the time of corrosion initiation (Figure 4). The sliced cores are analysed for chloride at different
depths from the concrete surface. The chloride content in the slice near the rebar depth is compared
with the chloride threshold value. The measured chloride content in this relatively large sample is not
exactly representative of the local amount of chloride close to reinforcement. The chloride content in
the sample along the rebar–concrete interface is the average amount of chloride over the rebar surface.
To determine the deviation from the average value, it is reasonable to measure the chloride content
in local points at the steel-concrete interface [51]. In this regard, non-destructive in situ techniques
have been developed for the determination of the local chloride content at different places close to
the reinforcement. In the next section, available techniques and main limitations for measuring the
local chloride content are discussed and the significance of developing a chloride sensor for chloride
measurement is explained.
Sensors 2017, 17, 2482 8 of 22
Sensors 2017, 17, 2482 8 of 21

Sensors 2017, 17, 2482 8 of 21

Figure 4.
Figure 4. Chloride
Chloride profile in concrete
profile in concrete and
and determination
determination of
of chloride
chloride threshold
threshold value.
value.
Figure 4. Chloride profile in concrete and determination of chloride threshold value.
4. Techniques for Determination of the Chloride Content in Concrete
4. Techniques for Determination of the Chloride Content in Concrete
4. Techniques for Determination of the Chloride Content in Concrete
4.1. Lab Techniques
4.1. Lab Techniques
4.1. Lab Techniques
The techniques
The techniques for
for determination
determination of of the
the chloride
chloride content
content in in concrete
concrete can
can be
be classified
classified into
into lab
lab
The
techniques and techniques for determination of the chloride content in concrete can be classified into lab
techniques and non-destructive
non-destructivein insitu
situtechniques.
[Link]
Figure5 5gives
gives ananoverview
overview of of
these techniques.
these techniques. In
lab techniques
techniques, and non-destructive
the concrete in situ techniques.
samples/cores are Figure
taken from5 gives
the an overview
structure for of these techniques.
further analysis inIn
the
In lab
lab techniques, theconcrete
techniques, the concretesamples/cores
samples/cores areare taken
taken fromfromthe the structure
structure for further
for further analysis
analysis in thein
laboratory.
the Generally, thethe
lablab
techniques are time-consuming, expensive and cannot be used for
laboratory. Generally, the lab techniques are time-consuming, expensive and cannot be used forfor
laboratory. Generally, techniques are time-consuming, expensive and cannot be used
continuous monitoring
continuous monitoring ofof the
the chloride
chloride content
content in a concrete structure.
structure.
continuous monitoring of the chloride contentininaaconcrete
concrete structure.

Figure 5. An overview of the available techniques for measuring the chloride content in concrete
Figure
Figure [Link]
Anoverview of the
overview available
of the techniques
available for measuring
techniques the chloride
for measuring content incontent
the chloride concretein[29,52–62].
concrete
[29,52–62].
[29,52–62].
TheThemostmost popular
popular and
andtraditional
traditionallablabtechnique
technique for
for chloride determinationinincementitious
chloride determination cementitious
The most
materials
materialsis the popular
leaching
is the and
leaching traditional
method.
method. thislab
InInthis technique
method,
method, for chloride
acid-soluble
acid-soluble determination
chloride
chloride in cementitious
and water-soluble
and water-soluble chloride
chloride
materials
are is
extracted the
from leaching
the method.
powdered In this
concrete method,
samples acid-soluble
into the chloride
solution. The and water-soluble
extracted solution
are extracted from the powdered concrete samples into the solution. The extracted solution is further chloride
is further
areanalysed
extracted
analysed forfor from
chloridethecontent
chloride powdered concrete
contentusing
using samples
different
different into such
methods,
methods, the solution.
such Thetitration
as Volhard
Volhard extracted
titration solution
[19],
[19], is further
potentiometric
potentiometric
analysed for[63]
titration chloride content using
and photometric different
analysis methods,
[64]. In suchthe
most cases, as acid-soluble
Volhard titration [19],
chloride is potentiometric
equivalent to
titration [63] and photometric analysis [64]. In most cases, the acid-soluble chloride is equivalent to
Sensors 2017, 17, 2482 9 of 22

titration [63] and photometric analysis [64]. In most cases, the acid-soluble chloride is equivalent to the
total chloride [18]. Water-soluble chloride does not necessarily represent the free chloride only, as it
is sensitive to the test condition. The extracted chloride content depends on the fineness of sample
powder, the amount of water added to the powder (powder/water ratio), the contact time and the
temperature of the suspension [17,51,65]. In fact, it was claimed that, given enough time and water,
all the chlorides can be extracted from the sample [66]. As a result, the different values of water-soluble
chloride in a concrete sample can partly be attributed to the different test procedure used [67,68].
The traditional method for the determination of free chloride content in concrete specimens in
the lab is extraction of pore water under pressure. The concrete specimens are pressed to extract
the pore water. A few millilitres of pore water are needed for chloride analysis [29]. To obtain the
minimal volume of pore water, a large volume of concrete sample with high w/c ratio, low content
of aggregate and wet moisture state was suggested [69]. This technique is not practical for field
application. Moreover, local chloride gradients cannot be determined with this technique.
The total chloride content can also be measured by X-ray fluorescence, scanning electron
microscopy equipped with energy dispersive spectroscopy and laser breakdown spectroscopy. The first
two techniques need preparation of concrete sample/core before chloride measurement. The advantage
of chloride determination using laser breakdown spectroscopy is the possibility for on-site application
and no need for sample preparation.
Other lab techniques are nuclear magnetic resonance (NMR), prompt gamma neutron activation
analysis and near-infrared, millimetre and micro-wave spectroscopy. These techniques are expensive
and labour-intensive, so their use is limited to appropriately equipped laboratories.

4.2. Non-Destructive In Situ Techniques

In the past decades attempts were made to develop non-destructive techniques for in situ monitoring
of the free chloride ions and chloride profiles in concrete structures. Concrete is a heterogeneous
material with high alkaline pore solution, different composition and pore systems with various porosity
and pore size distribution [3]. The many influencing factors make the application of in situ techniques
complex and difficult.
In situ techniques for measuring, or indicating, the chloride concentration are: electrical resistivity,
fibre optic sensors, chronopotentiometry and potentiometry methods. The electrical resistivity can be
used not only in field structures, but also in the laboratory on samples taken from the structure [65].
However, resistivity measurements for cement-based material are very sensitive to moisture content,
while not as sensitive to the chloride content. Therefore, alterations in the concrete resistivity measurements
cannot simply be correlated to the chloride content in the pore water.
A fibre optic sensor consists of a fibre with an optical transducer, sensitive to chloride ions.
The lifetime of optical transducers, protection of fibres and the bulky measurement setup are the
limitations of this technique [60].
Chronopotentiometry is a dynamic electrochemical method used for an indicative determination
of the chloride content at the surface of a working electrode. The current stimulus is applied to
the working electrode and the potential response is measured against a pseudo-reference electrode.
The use of this method for chloride measurement in concrete was recently hypothesized [61].
The potentiometry technique is a measurement of open circuit potential (OCP) of an embedded
Ag/AgCl electrode (chloride sensor) against a reference electrode (Figure 6c). This method is
considered the most practical approach for continuous monitoring of chloride content in the concrete
environment. Interpretation of the measured OCP requires detailed information of the environment
at the sensor–concrete interface (Figure 1d,e and Figure 6e,f) and knowledge of characteristics of the
chloride sensor (Figure 1a,b). In this regard, the composition of the pore solution is important for
the potential response of both the chloride sensor and reinforcing steel rod (Figure 6). For example,
the concentration of hydroxide ions can affect the stability of the chloride sensor. The chloride
sensor is generally not stable in an alkaline medium with low chloride concentration [62,70,71]
Sensors 2017, 17, 2482 10 of 21

[62]. In addition, in low alkaline mediums, the interference of hydroxide ions with the sensor
response
Sensors 2017,is17,
insignificant
2482 [72,73]. A cement-rich layer at the steel–concrete interface contributes 10 toofthe
22
formation of a protective layer on the surface of the reinforcement (Figure 6b) [74]. Under certain
circumstances the protective layer can be disrupted, causing active corrosion. The reinforcing steel
(Figure
corrodes 6f). The stability
uniformly of chloride
if the alkalinitysensor
of theincreases
mediumwhen chloride
decreases ions arebelow
to values present pHin 9the pore
[23]. solution
However,
(Figure 6e) [62]. In addition, in low alkaline mediums, the interference of
the hydroxide concentration is not the only factor determining steel activity. As mentioned, chloride hydroxide ions with the
sensor
ions areresponse
well-known is insignificant
corrodents[72,73].
and causeA cement-rich layer at the
steel de-passivation uponsteel–concrete interface
reaching a certain contributes
concentration
to the formation of a protective layer on the surface of the reinforcement (Figure
(threshold). It was suggested that the ratio [Cl ]/[OH ] is a better criterion to evaluate the chloride
− − 6b) [74]. Under certain
circumstances the protective layer can be disrupted, causing active corrosion.
content for corrosion initiation [35,38]. Although the pH of the pore solution is not the only parameter The reinforcing steel
corrodes uniformly
affecting the chlorideifthreshold
the alkalinity
valueof(as the mediumindecreases
discussed Section 3.4),to values below pH
and although 9 [23].
chloride However,
ions, in fact,
the hydroxide concentration is not the only factor determining steel activity.
do not alter the pH of the pore solution, pH affects the chloride ions’ activity towards corrosion As mentioned, chloride
ions are well-known
initiation [38,75,76]. corrodents
It was reportedand cause
that steel de-passivation
the effect of [Cl−]/[OH upon reaching
−] ratio a certain
on the corrosionconcentration
initiation
(threshold). It was suggested that the ratio [Cl − ]/[OH− ] is a better criterion to evaluate the chloride
should be considered with respect to the other influential factors, such as the physical condition of
content for corrosion
the steel/concrete initiation
interface [35,38].
[67,68]. Although
Therefore, the pH[Cl
a higher of −the pore
]/[OH solution
−] ratio is not
at the theofonly
level parameter
reinforcement
affecting
does not the chloriderepresent
necessarily thresholdavalue
higher (asrisk
discussed in Section
of corrosion 3.4), and
initiation [77].although chloride ions, in fact,
do not alter the pH of the pore solution, pH affects
In the previous sections, the main properties of cementitious the chloride ions’ materials
activity towards
for thecorrosion
sensor’s
initiation [38,75,76]. It was reported that the effect of [Cl − ]/[OH− ] ratio on the corrosion initiation
measurement were discussed. The obtained information revealed the importance of microstructure
should be considered
of cementitious withat
materials respect to the other
the interface withinfluential
the sensorfactors, such assensor
for a reliable the physical
reading condition
(Figure of the
1d,e).
steel/concrete interface [67,68]. Therefore, a higher [Cl − ]/[OH− ] ratio at the level of reinforcement
In the next section, the importance of the chloride sensor’s characteristics in the presence of
does not necessarily
interfering represent
ions is discussed a higher
(Figure risk of corrosion initiation [77].
1a,b).

Figure 6. Schematic representation of chloride sensor and steel rods in cementitious materials. The
Figure 6. Schematic representation of chloride sensor and steel rods in cementitious materials. The OCP
OCP of chloride sensor can be related to the chloride content in the medium, while the OCP of steel
of chloride sensor can be related to the chloride content in the medium, while the OCP of steel rod
rod represents the electrochemical state of the steel in concrete.
represents the electrochemical state of the steel in concrete.

5. Working Principles of the Ag/AgCl Electrode

In the previous sections, the main properties of cementitious materials for the sensor’s measurement
were Although
[Link] for therevealed
obtained information Ag/AgClthechloride sensor
importance of seems to be simple
microstructure (Figures 1c
of cementitious
materials at the interface with the sensor for a reliable sensor reading (Figure 1d,e). In the next state
and 6c), the interpretation of the measurements requires knowledge of the electrochemical of
section,
the importance
the chloride sensor itself
of the and the
chloride interaction
sensor’s of the sensor
characteristics with
in the the environment.
presence of interfering ions is discussed
Similar
(Figure 1a,b).to all electrochemical phenomena, the response of an Ag/AgCl chloride sensor follows
the laws of thermodynamics. Therefore, an overview of the thermodynamically plausible reactions
5.
andWorking Principles
different oxidationof the Ag/AgCl
states of AgElectrode
in highly alkaline solutions will be presented. A clear
understanding of the sensor’s response
Although the instrumentation for the requires
Ag/AgClknowledge
chloride of the kinetics
sensor seems toofbethe reactions
simple at the
(Figures 1c
and 6c), the interpretation of the measurements requires knowledge of the electrochemical state of the
chloride sensor itself and the interaction of the sensor with the environment.
Sensors 2017, 17, 2482 11 of 22

Similar
Sensors to2482
2017, 17, all electrochemical phenomena, the response of an Ag/AgCl chloride sensor follows the
11 of 21
laws of thermodynamics. Therefore, an overview of the thermodynamically plausible reactions and
sensor’s oxidation
different surface. As a result,
states of Agkinetic parameters
in highly alkaline and relevant
solutions willconstraints
be [Link] alsounderstanding
clear discussed in the of
nextsensor’s
the [Link] requires knowledge of the kinetics of the reactions at the sensor’s surface. As a
result,Silver
kineticin parameters
the form ofand an Ag/AgCl electrode iswill
relevant constraints widely
be alsoused in industrial
discussed in the technologies,
next section. medical
instruments and as a reference electrode in electrochemistry. Some
Silver in the form of an Ag/AgCl electrode is widely used in industrial technologies, state-of-the-art reports
medical on
Ag/AgCl electrodes
instruments date backelectrode
and as a reference to 1900 in [78]. However, the Some
electrochemistry. relation between thereports
state-of-the-art AgCl structure,
on Ag/AgCl the
resistivity of the AgCl layer and the transport process of silver and chloride ions
electrodes date back to 1900 [78]. However, the relation between the AgCl structure, the resistivity of in the layer has not
been
the sufficiently
AgCl layer anddescribed [79].process
the transport Consequently, debates
of silver and are still
chloride ions going on ashas
in the layer to how
not been to interpret the
sufficiently
sensor readings.
described [79]. Consequently, debates are still going on as to how to interpret the sensor readings.
One of
One of thethe characteristics
characteristics of of an
an Ag/AgCl
Ag/AgClelectrode
electrodeisisitsits(supposedly)
(supposedly) rapid
rapid response.
response. WhenWhen no
kinetic
no restrictions
kinetic restrictionsapply,
apply,thethedynamic
dynamic equilibrium
equilibrium between
between metallic
metallicsilver (Ag
silver 0
(Ag ) and
0 ) andAgAgcancan
+ + be
established
be established in in
a short
a shortperiod.
[Link] makes
feature makesthe the
noble nobleAg Ag metal
metalprone
prone to “corrosion”
to “corrosion” in
environments with aggressive ions, such as chloride ions. The
in environments with aggressive ions, such as chloride ions. The reaction rate depends on thereaction rate depends on the
electrochemicalstate
electrochemical stateof
ofthe
theelectrode
electrodein inthe
the aqueous
aqueous medium.
medium. The The electrochemical
electrochemical oxidation
oxidation of of silver
silver
in chloride-containing solutions, e.g., HCl solution, results in the formation of
in chloride-containing solutions, e.g., HCl solution, results in the formation of a silver chloride layera silver chloride layer
on
on the silver substrate (Figure 7) [79]. The mechanism of silver chloride formation
the silver substrate (Figure 7) [79]. The mechanism of silver chloride formation (Figure 7) is discussed (Figure 7) is
discussed
in in Section 6.1.
Section 6.1.

Figure 7. The
Figure 7. The electrochemical
electrochemical oxidation
oxidation of
of Ag
Ag in
in HCl
HClsolution.
solution.

The electrochemical
The electrochemical response
responseof ofthe
theAg/AgCl
Ag/AgCl electrode in aincertain
electrode environment
a certain environment is different from
is different
the Ag
from the metal
Ag metal alone. The The
alone. change
changein in
activity of ofchloride
activity chlorideionsions(concentration)
(concentration) in the medium
in the medium
subsequently affects the equilibrium at the Ag/AgCl interface. However, the electrode
subsequently affects the equilibrium at the Ag/AgCl interface. However, the electrode returns rapidly returns rapidly
to the equilibrium potential after a small transient perturbation. The principle of Ag/AgCl
to the equilibrium potential after a small transient perturbation. The principle of Ag/AgCl electrode electrode
response is
response is based
based onontwotwoequilibriums:
equilibriums: the
the electrochemical
electrochemical equilibrium
equilibrium involving
involving thethe formation
formation of of
interfacial potential, and the solubility equilibrium between the Ag cation and its sparsely
interfacial potential, and the solubility equilibrium between the Ag cation and its sparsely soluble salt soluble
salt (AgCl)
(AgCl) [80]. [80].
This This is further
is further explained
explained in theinfollowing.
the following.
The electrochemical
The electrochemicalequilibrium
equilibriumat atthe
thesilver
silversurface
surfacecancanbebedescribed
describedbybyEquation
Equation(1): (1):
Ag ↔ Ag+ + e−. (1)
Ag ↔ Ag+ + e− . (1)
At 25 °C, the solubility product (Ksp) of AgCl is Ksp = 1.8 × 10−1° [81].
The
At ◦ C, the solubility
25solubility equilibrium is in (K
product accordance toisEquation × 10−10 [81].
Ksp = 1.8(2):
sp ) of AgCl
The solubility equilibrium is in accordance to Equation (2):
AgCl ↔ Ag+ + Cl−; (2)
combining the above two reactions, the AgCl ↔ Ag+ + Cl− ; equilibrium reaction can be written as):
oxidation–reduction (2)
AgCl + e− ↔ Ag + Cl−, (3)
in which solid AgCl is deposited at a potential near the thermodynamic reversible potential for the
Ag/AgCl electrode in the chloride-containing medium. In this description, the silver chloride is the
“core” of Ag/AgCl electrode (chloride sensor), controlling its selectivity for the chloride ions. The
Sensors 2017, 17, 2482 12 of 22

combining the above two reactions, the oxidation–reduction equilibrium reaction can be written as:

AgCl + e− ↔ Ag + Cl− , (3)

in which solid AgCl is deposited at a potential near the thermodynamic reversible potential for the
Ag/AgCl electrode in the chloride-containing medium. In this description, the silver chloride is
the “core” of Ag/AgCl electrode (chloride sensor), controlling its selectivity for the chloride ions.
Sensors 2017, 17, 2482 12 of 21
The relation between the half-cell potential of the chloride sensor and the chloride activity is expressed
by the Nernstbetween
relation equation the[62]:
half-cell potential of the chloride sensor and the chloride activity is expressed
by the Nernst equation [62]:
RT
EAg/AgCl = E00Ag/AgCl − 2.303 RTnF lg[ aCl− ], (4)
E Ag/AgCl =E Ag/AgCl  2.303 lg[aCl ] , (4)
nF
where EAg/AgCl is the measured electrode potential [V], E◦ Ag/AgCl is the standard electrode potential
where EAg/AgCl is the measured electrode
− is the potential [V], E°Ag/AgCl is the standard electrode potential of
of the Ag/AgCl electrode [V], acl activity of the chloride ions [mol ·dm−3 ] in the vicinity of
the Ag/AgCl electrode [V], acl− is the activity
− 1 of1 the chloride ions [mol∙dm−3] in the vicinity
− − of the
the electrode, R is the gas constant [J·mol ·K ], F is the Faraday constant [C·mol 1 ] and T is the
electrode, R is the gas constant [J∙mol−1∙K−1], F is the Faraday constant [C∙mol−1] and T is the absolute
absolute temperature
temperature (K). (K).
The chloride activity
The chloride is linked
activity to thetochloride
is linked ions concentration
the chloride by theby
ions concentration activity coefficient
the activity (γ) [82–84]:
coefficient
(γ) [82–84]:
aCl− = CCl− · γCl− (5)
acl =CCl .γCl (5)
Therefore, by measuring the potential of the Ag/AgCl electrode and using the Nernst equation,
Therefore, by measuring the potential of the Ag/AgCl electrode and using the Nernst equation,
the chloride activity, and subsequently the chloride concentration in the solution can be calculated.
the chloride activity, and subsequently the chloride concentration in the solution can be calculated.
6. Discussion
6. Discussion
6.1. Electrochemical Kinetics of AgCl Layer Formation
6.1. Electrochemical Kinetics of AgCl Layer Formation
In most studies,
In most anodization
studies, anodizationofofsilver
silverin
inchloride-containing solutions
chloride-containing solutions waswasthe the method
method usedused
for for
the oxidation of silver
the oxidation andand
of silver formation
formationofofAgCl
AgCl on
on the silversubstrate
the silver substrate [85,86].
[85,86]. TheThe
main main differences
differences
among the anodization
among regimes
the anodization are are
regimes the the
current density
current andand
density thethe
duration
duration of of
anodization.
[Link]
somecases,
cases, theAgCl
the anodized anodized
layerAgCl
was layer was additionally
additionally dipped indipped
an AgClin an AgCl
melt melt to achieve
to achieve a more astable
more AgCl
stable layer.
AgCl
It haslayer.
been reported that after nucleation, the growth of an AgCl layer on the silver substrate
proceeds It has the
with beenformation
reported that after nucleation,
of small patches of the growthand
rounded of an AgCl layer
smooth surfaceon the
with silver substrate
no sign of crystal
proceeds with the formation of small patches of rounded and smooth surface with no sign of crystal
orientation [87]. At this stage, the size of the AgCl particles is less than 0.5 µm [88]. However, transition
orientation [87]. At this stage, the size of the AgCl particles is less than 0.5 μm [88]. However,
to a multilayer brings dense and fine AgCl particles to the surface (Figure 8a). Anodization at high
transition to a multilayer brings dense and fine AgCl particles to the surface (Figure 8a). Anodization
current density
at high changes
current thechanges
density morphology of the AgCl
the morphology top AgCl
of the layertop
to alayer
so-called mosaic appearance
to a so-called mosaic
(Figure 8b).
appearance (Figure 8b).

Figure
Figure 8. Surface
8. Surface morphology
morphology ofofananAgCl
AgCllayer
layer prepared
prepared by
byone-hour
one-houranodization at different
anodization current
at different current
densities.
densities.

The galvanostatic growth of a AgCl multilayer is accompanied by an increase in the

overpotential due to the ohmic resistance of the AgCl layer itself [89,90]. The AgCl itself is non-
conductive, so the effective conductivity of such a layer depends on the pores (microchannels)
between the AgCl grains (Figure 7a). Therefore, the growth of the AgCl layer is a function of the ionic
Sensors 2017, 17, 2482 13 of 22

The galvanostatic growth of a AgCl multilayer is accompanied by an increase in the overpotential

due to the ohmic resistance of the AgCl layer itself [89,90]. The AgCl itself is non-conductive, so the
effective conductivity of such a layer depends on the pores (microchannels) between the AgCl grains
(Figure 7a). Therefore, the growth of the AgCl layer is a function of the ionic conductivity of the
microchannels [91]. Both low and high field conduction mechanisms are reported for the growth of an
AgCl layer [79]. When the AgCl layer is porous and the overpotential is low, the ionic conductivity
takes place through microchannels (extrinsic conductivity). This is known as low field conduction.
At high overpotential, when the relation between the current density and the potential is
exponential, microchannels may become blocked. As a result, ions migrate through the solid AgCl
particles (intrinsic conductivity). This is known as high field conduction [92]. The importance of the
above mechanisms for the application of the Ag/AgCl electrode as a chloride sensor is related to their
subsequent influence on the AgCl layer morphology and microstructure.
The observed potential difference among the Ag/AgCl chloride sensors in an alkaline medium
with the same chloride concentration was attributed to the sensor preparation method [93,94]. However,
no evidence in support of this assertion was provided. For instance, the OCPs of two differently
prepared Ag/AgCl chloride sensors were measured in Ca(OH)2 solution with 0.002 M to 0.86 M NaCl.
The observed OCP of the sensors were different in the entire range of the chloride concentrations.
The effect was higher at lower chloride concentration, e.g., 60 mV at 0.002 M chloride concentration
versus 40 mV at 0.86 M chloride concentration [95]. In a similar study, two differently prepared
chloride sensors were embedded in mortar samples. A difference of 100 mV in the initial OCP of the
chloride sensors was reported [93]. These studies did not provide further evidence for the observed
discrepancies between the sensor readings. Thus, the influence of AgCl layer properties on the
reproducibility and reliability of the sensor’s measurements was not further discussed [93,94].
Within anodization, the mechanism of ionic conductivity governs the kinetics for silver dissolution
and AgCl formation. The exchange in this mechanism from extrinsic to intrinsic (see paragraphs 3 and
4 of this section) can subsequently affect the AgCl layer morphology and microstructure. The alteration
in the physical properties of the AgCl layer (porosity, thickness, morphology, etc.) can partly affect the
stability and reliability of the sensor’s measurements in cementitious materials.

6.2. Thermodynamic Behaviour of Silver

AgCl is a “corrosion product” of silver in a chloride-containing environment. The electrochemical
state of silver, similar to all electrochemical phenomena, follows laws of thermodynamics. A full
understanding and prediction of the corrosion process undoubtedly requires a consideration of
thermodynamic principles. Following these principles, the standard Gibbs free energy (∆G) for a
given compound is the energy for the formation of 1 mole of that compound from its constituents.
A negative ∆G is an indication of a spontaneous reaction.
Table 3 gives the standard Gibbs energies of formation of the species and phases in the
silver–chloride–water system [96]. The positive Gibbs energy of silver ions (Ag+ ) and AgO is an
indication of non-spontaneous reactions. The Gibbs energy of the ionic silver compounds, such as
Ag(OH)2 − , AgCl2 − and AgCl4 3− is more negative than the crystalline phases (c), such as AgCl (c)
and Ag2 O (c). Consequently, the formation of the former compounds is more product-favoured than
the latter ones. The spontaneous corrosion product of silver in a chloride-containing medium is
AgCl (e.g., AgCl with Gibbs energy = −110 KJ/mol), rather than silver oxide (e.g., Ag2 O with Gibbs
energy = −11 KJ/mol).
Kinetic restrictions exist in all thermodynamically possible reactions. The actual occurrence,
direction and rate of thermodynamically possible reactions are governed by electrochemical kinetics.
For example, the thermodynamic data indicates the formation of ionic silver compounds or AgCl as
the favoured reaction product on silver. However, the rate of formation of these compounds depends
on the concentration of chloride and hydroxide ions in the medium. This process is also reflected in
the response of Ag/AgCl sensor in an alkaline medium with low chloride content. In this condition,
Sensors 2017, 17, 2482 14 of 22

the alkalinity of the medium and transformation rate of silver chloride to silver oxide at the sensor’s
surface governs the sensor’s response. The influence of this transformation rate on the sensor’s
response is generally not discussed in the reported studies.

Table 3. Gibbs free energies for the formation of species in silver–halide–water systems at 25 ◦ C [96].

Species Gibbs Free Energy (KJ/mol) Species Gibbs Free Energy (KJ/mol)
Ag+ 77 Ag(OH)2 − −260
Cl− −131 AgCl (c) −110
Ag2 O (c) −11 AgCl (aq) −54
AgO (c) 15 AgCl2 − −216
AgOH −80 AgCl4 3− −478
(aq): aqueous phase—(c): crystalline phase.

The Pourbaix (potential—pH) diagram is based on thermodynamics, so it can only “map out”
thermodynamic stability of metal species and compounds at various combinations of equilibrium
potential and pH in an aqueous medium under standard conditions. Brief thermodynamic principles
for the behaviour of silver in water in the presence and absence of chloride ions are discussed in
what follows.
The Pourbaix diagram for the silver–chloride–water system [97] shows that the chloride ion
is a strong oxidizing agent for Ag(s) and crystalline AgCl (AgCl (c)) is the thermodynamically
favourable oxidation product. It is reported that the increase in the concentration of chloride ions
causes dissolution of AgCl (c) and formation of ionic silver chloride, such as AgCl2 − or AgCl4 3− [98].
In chloride-containing alkaline medium, the silver oxide compounds can thermodynamically exist.
Silver chloride can counteract silver oxide formation, depending on the chloride concentration and pH
of the environment. All these processes can be related to the performance of the Ag/AgCl chloride
sensor in cementitious materials. The dissolution of AgCl (c) and formation of silver chloride complexes
(AgCln (1−n) ) or formation of silver oxide may deviate the response of the chloride sensor from the
Nernstian behaviour. Therefore, the potential measured as EAg/AgCl will no longer satisfy Equation (4)
or be valid as such.

6.3. Considerations with Regard to Chloride Ions’ Activity and pH

It is known that the presence of AgCl(c) is essential for the Nernstian behaviour of the sensor.
The stability of silver chloride is a function of silver and chloride activity at the sensor’s surface as
well as the pH of the medium. For example, the stability of AgCl(c) is limited to a certain range
of chloride activity (10−6 < αcl <1) [99]. Trespassing the limit, products other than silver chloride
form, which undeniably affect the performance of the sensor. The silver oxide complex ions, or silver
chloride complexes, are not necessarily attracted to the sensor’s surface [100]. The ions can leave the
sensor’s surface and diffuse into the environment (e.g., within the cementitious materials). In this
regard, the properties of cementitious materials, such as porosity, are important. This process can
subsequently shift the sensor’s OCP towards values that cannot be expected from the Nernst equation
for an Ag/AgCl electrode. Therefore, the AgCl layer properties and the Ag/AgCl interface may
influence the stability and reliability of the sensor’s response. This should be taken into account
when long-lasting performance and ease of application of chloride sensors in cementitious materials
are desired.

6.4. Reliability of Sensor Response in Cementitious Materials

The performance of a chloride sensor in cementitious materials depends on the thermodynamic
stability of the AgCl layer at the sensor’s surface. In the presence of chloride ions in the solution, AgCl
is the dominant reaction product (Figure 9a). However, the interference of hydroxide and sulphide
ions in the solution can limit the otherwise plausible reaction of AgCl formation on the sensor’s surface
Sensors 2017, 17, 2482 15 of 22

(Figure 9b). Therefore, the rate of AgCl transformation into Ag2 O or Ag2 S controls the variation in the
sensor’s OCP. This process essentially makes the potential of the Ag/AgCl sensor deviate from the
expected Nernstian
Sensors 2017, 17, 2482 response (as would be expected from Equation (4)). 15 of 21

Figure
Figure 9. Schematic
9. Schematic ofof Ag/AgCl/pore solution
Ag/AgCl/pore solution interface
interfaceinincementitious
cementitiousmaterials.
materials.(a)(a)
InInthe
thepresence
presence
of chloride ions, the surface of the sensor mainly consists of AgCl particles.
of chloride ions, the surface of the sensor mainly consists AgCl particles. The binding of The binding of chloride
chloride
ions
ions to to thehydration
the hydration products
products together
togetherwith
withthe microstructure
the microstructure of cementitious
of cementitiousmaterials affect affect
materials the
theactivity
activityofofthe
thechloride
chlorideions
ionsatatthe
thesensor’s
sensor’ssurface.
surface;(b)(b)InInthe
theabsence
absenceofofchloride
chlorideions,ions,gradual
gradual
dissolution
dissolution of AgCl
of AgCl layer
layer is accompanied
is accompanied by formation
by formation of Ag of2 O,
AgAg 2O,0 Ag 0 or Ag2S (in the case of slag
or Ag 2 S (in the case of slag cement
cement concrete) [101–103]. The formation of these products
concrete) [101–103]. The formation of these products depends on the concentration depends on the concentration
of hydroxide of and
sulphide ions in the medium and the microstructure of cementitious materials at thematerials
hydroxide and sulphide ions in the medium and the microstructure of cementitious at the
interfacial zone
interfacial
with zone with the sensor.
the sensor.

In a solution with high pH and low chloride concentration (Figure 9b), AgCl can be partly or
In a solution
completely with high
converted pH and
into silver lowcompounds
oxide chloride concentration (Figureetc.).
(Ag2O, AgO, AgOH, 9b), Therefore,
AgCl can the
be partly
sensor or
completely converted into silver oxide compounds (Ag
gradually acts as a pH sensor rather than a chloride sensor.2 O, AgO, AgOH, etc.). Therefore, the sensor
In this condition, the activity of silver
gradually actsvicinity
ions in the as a pHof sensor rather
the sensor than a chloride
is determined [Link]
by exchange In this condition, the activity
and formation of Agof
2O,silver
as
shown in Figure 9b. However, the dissolved silver chloride and/or the formed silver oxide ions
(Ksp (AgOH) = 2 × 10−8 at 25 °C [104]) do not necessarily adhere to the sensor’s surface [100]. These
ions can leave the sensor’s surface and diffuse into the cementitious material.
Sensors 2017, 17, 2482 16 of 22

ions in the vicinity of the sensor is determined by exchange equilibrium and formation of Ag2 O,
as shown in Figure 9b. However, the dissolved silver chloride and/or the formed silver oxide ions
(Ksp (AgOH) = 2 × 10−8 at 25 ◦ C [104]) do not necessarily adhere to the sensor’s surface [100]. These
ions can leave the sensor’s surface and diffuse into the cementitious material.
Thermodynamically, the formation of Ag2 S with lower solubility (Ksp (Ag2 S) = 1.6 × 10−49 at
◦
25 C) is favoured when a sufficient amount of sulphide ions is in the medium, e.g., in concrete with
slag cement. It has been stated that the sulphide content of higher than 6 mM can affect the chloride
sensor’s response in a model solution [103]. Such a small content of sulphide ions can significantly
change the sensor’s response. The extent to which Ag2 S contributes to the product layer at the sensor’s
surface determines the cathodic potential shift of the sensor (Figure 9b) [105]. The interference of
sulphide ions with the chloride sensor is more serious than that of hydroxide ions [101,106]. This is
mainly because of very small solubility product of silver sulphide (as above mentioned). It is strongly
suggested that the sulphide ions in the pore solution of slag cement may affect the chloride sensor’s
measurements [103]. This assertion cannot be generalized for the application of the chloride sensor in
concrete with medium/low slag cement (e.g., CEM III/A). In such a medium, the kinetics of reactions
of sulphide ions with the AgCl layer at the sensor’s surface should be considered.
The AgCl layer gradually dissolves in chloride-free solutions, even in the absence of interfering
ions (hydroxide and sulphide ions). The gradual dissolution of the AgCl layer can be fatal for the
sensor performance [107]. The dissolved ions can gradually diffuse into the medium, exposing the Ag
substrate to direct contact with the environment. In this condition, the exact behaviour of the Ag is
difficult or even impossible to predict, when the sensor is under “mixed potential control”. The OCP of
the Ag metal is unstable, irreproducible and not theoretically meaningful. The measurements depend
on adventitious impurities in the medium (Figure 9b) [108]. Therefore, the potential response of silver
metal in cementitious materials cannot represent the alteration in the concentration of a certain ion,
i.e., chloride in the medium.
All the above considerations with regards to (electro) chemical reactions and actual sensors’
response depend on the availability of ions (chloride ions and interfering ones) and ion transport,
or limitations in the surrounding medium, i.e., pore water and pore network.
The importance of the microstructure of cementitious materials involves the influence of different
hydration products and their compactness at the sensor’s surface. In the presence of large pores
at the sensor–cement paste interface (Figure 1d,e), a considerable number of chloride ions are free.
Ongoing hydration of cement densifies the interface between the sensor and the cementitious material.
Densification of the interface results in the presence of fine pores of nm size between the sensor and
hydration products. The physical binding of chloride ions to the surface of hydration products lowers
the activity of chloride ions at the sensor’s surface. The lower activity of chloride ions will be reflected
in the sensor’s response. The presence of preferentially oriented crystalline phases at the interface
(Section 2.1) limits the dissolution/precipitation processes at the sensor’s surface. This will be reflected
in the sensor’s response. Consequently, the properties of the sensor and surrounding cementitious
materials can affect the thermodynamically plausible reactions at the sensor’s surface. This is due
to the limitations of electron/ion transport at the sensor surface and/or limitations caused by the
resistive properties of the surrounding medium. The AgCl layer properties, different cement hydration
products and their compactness at the interface between sensor and bulk matrix are the major factors
that limit the sensor performance. If the morphology and microstructure of the AgCl layer, binding
of chloride ions to the hydration products and the pore size distribution around the sensor are not
considered, the readings of chloride sensor are difficult to interpret. Thus, a link between the inherent
properties of the sensor and the relevant exposure environment has to be made. To overcome the
lack of information on the performance of chloride sensor in cementitious materials, further research
is needed.
Sensors 2017, 17, 2482 17 of 22

7. Conclusions
The available techniques for determination of the free chloride content in cementitious materials
are prone to several sources of errors. The potentiometric technique, using an Ag/AgCl chloride
sensor, is one of the most promising methods for non-destructive in situ free chloride measurement.
The deviation of the sensor’s OCP from Nernstian behaviour has been considered as the main criterion
for evaluation of the sensor performance. The causes of inaccurate sensor measurement are often
a concern. The available studies superficially commented on the short- /long-term stability of the
chloride sensor, paying little attention to the causes of inaccuracies. As a result, the identification of
the underlying sources of inaccuracies is a prerequisite for the practical application of the sensor.
A reliable and reproducible chloride sensor measurement requires knowledge of the environmental
exposure condition and the inherent properties of the sensor. In this regard, the following should
be considered:

1. The interfacial properties between the silver substrate and the AgCl layer, i.e., morphology of the
AgCl particles and the conductivity of the layer.
2. The presence of interfering hydroxide and sulphide ions, i.e., the pore water composition.
3. The relative distribution and compactness of cement hydration products at the sensor’s surface,
i.e., the occurrence of “screening” effects.

The physical properties of the AgCl layer (porosity, thickness, morphology, etc.) affect the rate of
ion exchange and dissolution/precipitation processes at the sensor’s surface. Instability of Ag/AgCl
chloride sensor is typically due to dissolution of the AgCl layer and/or the formation of other silver
compounds. As a result, mixed potential subsequently develops at the sensor’s surface. The AgCl layer
is more stable when the chloride concentration is high, while at low concentrations, products other
than silver chloride form. The presence of other silver compounds obviously affects the performance
of the sensor. If the aim is to determine the chloride content close to the reinforcement at the time of
corrosion of reinforcement with high accuracy, the sensor’s response should be reliable in a wide range
of chloride concentration, i.e., 0.045–3.22 mol/L (Table 2). As the source of chloride ions is mainly from
the external environment, the time required for chloride ions to penetrate into the concrete and reach
at least 0.045 mol/L at the sensor surface can be very long. Consequently, the sensor will be exposed
to an alkaline medium free of chloride ions or with a very low chloride content for a relatively long
period of time. In such a case, the accuracy of the chloride sensor cannot be simply judged on the
thermodynamic principles, depending on the reaction kinetics at the sensor’s surface.
The microstructure of cementitious materials is the resistive component that can limit the ion
transport at the sensor surface and affect the thermodynamic plausible reaction of the chloride sensor.
Cementitious materials also serve as a diffusion barrier for an embedded chloride sensor, slowing down
the rate of AgCl dissolution. This feature can be beneficial for the long-term performance of the chloride
sensor. The dissolution of the AgCl layer can be fatal for the sensor performance. The dissolved AgCl
particles can gradually diffuse into the medium, exposing the Ag substrate to direct contact with the
environment. Moreover, the presence of hydroxide and/or sulphide ions in cementitious materials can
interfere with the Nernstian behaviour of the sensor. The significance of different cement hydration
products and their compactness around the sensor is from the subsequent influence on the activity
of the chloride ions. The alteration in chloride ions activity is subsequently reflected in the sensor’s
response. Therefore, recognition of the imposed limitations from these parameters is important for
reliable chloride sensor measurement.

Conflicts of Interest: The authors declare no conflict of interest.

References
1. Molina, M. Zerstörungsfreie Erfassung der Gelösten Chloride im Beton. Ph.D. Thesis, ETH Zurich, Zürich,
Switzerland, 1993. (In Germany)
Sensors 2017, 17, 2482 18 of 22

2. Bard, A.J.; Rubinstein, I. Electroanalytical Chemistry: A Series of Advances; Marcel Dekker, Inc.: New York, NY,
USA, 1996; Volume 22.
3. Hewlett, P.C. Lea’s Chemistry of Cement and Concrete, 4th ed.; Butterworth-Heinemann, Elsevier Ltd.: Oxford,
UK, 2003.
4. Lothenbach, B.; Le Saout, G.; Haha, M.B.; Figi, R.; Wieland, E. Hydration of a low-alkali CEM III/B–SiO2
cement (LAC). Cem. Concr. Res. 2012, 42, 410–423. [CrossRef]
5. Mehta, P.K.; Monterio, P.J.M. Concrete, Microstructure, Properties and Materials, 3rd ed.; McGraw-Hill
Companies: New York, NY, USA, 2006.
6. Van Breugel, K. Simulation of Hydration and Formation of Structure in Hardening Cement-Based Materials.
Ph.D. Thesis, Delft University of Technology, Delft, The Netherlands, 1991.
7. Ollivier, J.P.; Maso, J.C.; Bourdette, B. Interfacial transition zone in concrete. Adv. Cem. Based Mater. 1995, 2,
30–38. [CrossRef]
8. Swamy, R.N. The Alkali-Silica Reaction in Concrete; The Blackie Publishing Group: Glasgow, UK, 1992.
9. Elakneswaran, Y.; Nawa, T.; Kurumisawa, K. Electrokinetic potential of hydrated cement in relation to
adsorption of chlorides. Cem. Concr. Res. 2009, 39, 340–344. [CrossRef]
10. Longuet, P.; Burglen, L.; Zelwer, A. La phase liquide du ciment hydrate. Rev. Mater. Constr. 1973, 676, 35–41.
11. Barneyback, R.S.; Diamond, S. Expression and analysis of pore fluids from hardened cement pastes and
mortars. Cem. Concr. Res. 1981, 11, 279–285. [CrossRef]
12. Page, C.L.; Vennesland, Ø. Pore solution composition and chloride binding capacity of silica-fume cement
pastes. Matér. Constr. 1983, 16, 19–25. [CrossRef]
13. Lothenbach, B.; Winnefeld, F. Thermodynamic modelling of the hydration of Portland cement.
Cem. Concr. Res. 2006, 36, 209–226. [CrossRef]
14. Diamond, S. ASR—Another Look at Mechanisms. In Proceedings of the Eighth International Conference on
Alkali-Aggregate Reaction in Concrete, Kyoto, Japan, 17–20 July 1989; pp. 83–94.
15. Baron, J.; Ollivier, J.P. La durabilité des Bétons; Presse de l’École Nationale des Ponts et chaussées: Paris, France,
1992. (In French)
16. Garcia, V.; Francois, R.; Carcasses, M.; Gegout, P. Potential measurement to determine the chloride threshold
concentration that initiates corrosion of reinforcing steel bar in slag concretes. Mater. Struct. 2014, 47,
1483–1499. [CrossRef]
17. Glass, G.K.; Wang, Y.; Buenfeld, N.R. An investigation of experimental methods used to determine free and
total chloride contents. Cem. Concr. Res. 1996, 26, 1443–1449. [CrossRef]
18. Analysis of total chloride content in concrete. Available online: [Link]
[Link] (Accessed on 28 October 2017).
19. ASTM C1152/C1152M–04(2012)e1. Standard Test Method for Acid-Soluble Chloride in Mortar and Concrete;
ASTM International: Philadelphia, PA, USA, 2003.
20. Nordtest. Concrete, Hardened: Chloride Content by Volhard Titration; NT Build 208; Nordtest: Espoo,
Finland, 1996.
21. Dhir, R.K.; Jones, M.R.; Ahmed, H.E.H. Determination of total and soluble chlorides in concrete. Cem. Concr.
Res. 1990, 20, 579–590. [CrossRef]
22. Arya, C. Assessment of Chlorides in Concrete. Ph.D. Thesis, University of London, London, UK, 1988.
23. Broomfield, J.P. Corrosion of Steel in Concrete, 2nd ed.; Taylor & Francis: London, UK, 2007.
24. Birnin-Yauri, U.A.; Glasser, F.P. Friedel’s salt, Ca2 Al(OH)6 (Cl,OH)·2H2 O: Its solid solutions and their role in
chloride binding. Cem. Concr. Res. 1998, 28, 1713–1723. [CrossRef]
25. Maslehuddin, M.; Page, C.L. Rasheeduzzafar Temperature effect on the pore solution chemistry in
contaminated cements. Mag. Concr. Res. 1997, 49, 5–14. [CrossRef]
26. Zibara, H. Binding of External Chloride by Cement Pastes. Ph.D. Thesis, Department of Building Materials,
University of Toronto, Toronto, ON, Canada, 2001.
27. Chatterji, S.; Kawamura, M. Electrical double layer, ion transport and reactions in hardened cement paste.
Cem. Concr. Res. 1992, 22, 774–782. [CrossRef]
28. Elakneswaran, Y. Multi-Ionic Transport in Cementitious Materials with Ion-Cement Hydrates Interactions.
Ph.D. Thesis, Hokkaido University, Solid Waste, Resources and Geoenviromental Engineering, Sapporo,
Japan, 2009.
Sensors 2017, 17, 2482 19 of 22

29. Tritthart, J. Chloride binding in Cement II. The influence of the hydroxide concentration in the pore solution
of hardened cement paste on chloride binding. Cem. Concr. Res. 1989, 19, 683–691. [CrossRef]
30. Birdi, K.A.S. Handbook of Surface and Colloid Chemistry; CRC Press: Boca Raton, FL, USA, 2015.
31. Friedmann, H.; Amiri, O.; Ait-Mokhtar, A. Physical modeling of the electrical double layer effects on
multispecies ions transport in cement-based materials. Cem. Concr. Res. 2008, 38, 1394–1400. [CrossRef]
32. Maage, M.; Helland, S.; Poulsen, E.; Vennesland, Ø.; Carlsen, J.E. Service life prediction of existing concrete
structures exposed to marine environment. ACI Mater. J. 1996, 93, 602–608.
33. Hausmann, D.A. Steel corrosion in concrete. How does it occur? Mater. Prot. 1967, 6, 19–23.
34. Melchers, R.E.; Li, C.Q. Reinforcement corrosion initiation and activation times in concrete structures exposed
to severe marine environments. Cem. Concr. Res. 2009, 39, 1068–1076. [CrossRef]
35. Melchers, R.E. Developing realistic deterioration models. In Life-Cycle of Engineering Systems: Emphasis on
Sustainable Civil Infrastructure; CRC Press: Boca Raton, FL, USA, 2016; pp. 28–32.
36. Alonso, C.; Andrade, C.; Castellote, M.; Castro, P. Chloride threshold values to depassivate reinforcing bars
embedded in a standardized OPC mortar. Cem. Concr. Res. 2000, 30, 1047–1055. [CrossRef]
37. Alonso, C.; Castellote, M.; Andrade, C. Chloride threshold dependence of pitting potential of reinforcements.
Electrochim. Acta 2002, 47, 3469–3481. [CrossRef]
38. Oh, B.H.; Jang, S.Y.; Shin, Y.S. Experimental investigation of the threshold chloride concentration for corrosion
initiation in reinforced concrete structures. Mag. Concr. Res. 2003, 55, 117–124. [CrossRef]
39. Anon. Condition Investigation Manual for Concrete Facade Panels: BY 42; Concrete Association of Finland:
Helsinki, Finland, 2002; p. 178. (In Finnish)
40. Stratfull, R.F.; Jurkovich, W.J.; Spellman, D.L. Corrosion Testing of Bridge Decks; Highway Research Record 539;
Transportation Research Board: Washington, DC, USA, 1975.
41. Ferreira, R.M. Probability-Based Durability Analysis of Concrete Structures in Marine Environment.
Ph.D. Thesis, University of Minho, Braga, Portugal, 2004.
42. Maes, M.; Gruyaert, E.; de Belie, N. Resistance of concrete with blast-furnace slag against chlorides.
investigated by comparing chloride profiles after migration and diffusion. Mater. Struct. 2013, 46, 89–103.
[CrossRef]
43. Schiessl, P.; Raupach, M. Corrosion of Reinforcement in Concrete; Elsevier Applied Science: London, UK, 1990.
44. Daily, S.F. Understanding Corrosion and Cathodic Protection of Reinforced Concrete Structures; Corrpro Companies,
Incorporated: Medina, OH, USA, 1999.
45. Pettersson, K. Corrosion Threshold Value and Corrosion Rate in Reinforced Concrete; CBI Report 2; Swedish
Cement and Concrete Research Institute: Stockholm, Sweden, 1992.
46. Elsener, B.; Zimmermann, L.; Fluckiger, D.; Burchler, D.; Bohni, H. Chloride penetration- non-destructive
determination of the free chloride content in mortar and concrete. In Proceedings of the RILEM International
Workshop on Chloride Penetration into Concrete, Paris, France, 15–18 October 1995.
47. Zimmermann, L.; Elsener, B.; Bohni, H. Critical factors for the initiation of rebar corrosion. In Proceedings of
the EUROCORR ’99, European Federation of Corrosion, Aachen, Germany, 30 August–2 September 1999.
48. Lambert, P.; Page, C.L.; Vassie, P.R. Investigations of reinforcement corrosion. 2. Electrochemical monitoring
of steel in chloride-contaminated concrete. Mater. Struct. 1991, 24, 351–358. [CrossRef]
49. Tuutti, K. Corrosion of Steel in Concrete; Swedish Cement and Concrete Research Institute: Stockholm,
Sweden, 1982.
50. Mohammed, T.U.; Hamada, H. Relationship between free chloride and total chloride contents in concrete.
Cem. Concr. Res. 2003, 33, 1487–1490. [CrossRef]
51. Silva, N.; Tang, L.; Lindqvist, J.E.; Boubitsas, D. Chloride profiles along the concrete-steel interface. Int. J.
Struct. Eng. 2013, 4, 100–112. [CrossRef]
52. Arya, C.; Buenfeld, N.R.; Newman, J.B. Assessment of simple methods of determining the free chloride ion
content of cement paste. Cem. Concr. Res. 1987, 17, 907–918. [CrossRef]
53. Proverbio, E.; Carassiti, F. Evaluation of chloride content in concrete by X-ray fluorescence. Cem. Concr. Res.
1997, 27, 1213–1223. [CrossRef]
54. Skoglund, P. Chloride Transport and Reinforcement Corrosion in the Vicinity of the Transition Zone between
Substrate and Repair Concrete. Licentiate Thesis, KEYI, Stockholm, Sweden, 2006.
55. Wilsch, G.; Weritz, F.; Schaurich, D.; Wiggenhauser, H. Determination of chloride content in concrete
structures with laser-induced breakdown spectroscopy. Constr. Build. Mater. 2005, 19, 724–730. [CrossRef]
Sensors 2017, 17, 2482 20 of 22

56. De J Cano, F.; Bremner, T.W.; McGregor, R.P.; Balcom, B.J. Magnetic resonance imaging of 1 H, 23 Na, and 35 Cl
penetration in Portland cement mortar. Cem. Concr. Res. 2002, 32, 1067–1070. [CrossRef]
57. Tripathi, S.R.; Inoue, H.; Hasegawa, T.; Kawase, K. Non-destructive inspection of chloride ion in concrete
structures using attenuated total reflection of millimeter waves. J. Infrared Millim. Terahertz Waves 2013, 34,
181–186. [CrossRef]
58. Livingston, R.A.; Al-Sheikhly, M.; Mohamed, A.B. Numerical simulation of the PGNA signal from chlorine
diffusion gradients in concrete. Appl. Radiat. Isot. 2010, 68, 679–682. [CrossRef] [PubMed]
59. McPolin, D.; Basheer, P.A.M.; Long, A.E.; Grattan, K.T.V.; Sun, T. Obtaining progressive chloride profiles in
cementitious materials. Constr. Build. Mater. 2005, 19, 666–673. [CrossRef]
60. Laferriere, F.; Inaudi, D.; Kronenberg, P.; Smith, I.F. A new system for early chloride detection in concrete.
Smart Mater. Struct. 2008, 17, 045017. [CrossRef]
61. Abbas, Y. In-Situ Measurement of Chloride ion Concentration in Concrete. Ph.D. Thesis, University of
Twente, Enschede, The Netherlands, 2015.
62. Atkins, C.P.; Scantlebury, J.D.; Nedwell, P.J.; Blatch, S.P. Monitoring chloride concentrations in hardened
cement pastes using ion selective electrodes. Cem. Concr. Res. 1996, 26, 319–324. [CrossRef]
63. ASTM C114–15. Standard Test Methods for Chemical Analysis of Hydraulic Cement; American Society for Testing
and Materials (ASTM): Philadelphia, PA, USA, 2000.
64. Maierhofer, C.; Reinhardt, H.W.; Dobmann, G. (Eds.) Non-Destructive Evaluation of Reinforced Concrete
Structures: Non-Destructive Testing Methods; Elsevier: Amsterdam, The Netherlands, 2010.
65. Torres-Luque, M.; Bastidas-Arteaga, E.; Schoefs, F.; Sanchez-Silva, M.; Osma, J.F. Non-destructive methods
for measuring chloride ingress into concrete: State-of-the-art and future challenges. Constr. Build. Mater.
2014, 68, 68–81. [CrossRef]
66. Haque, M.N.; Kayali, O.A. Free and water soluble chloride in concrete. Cem. Concr. Res. 1995, 25, 531–542.
[CrossRef]
67. ASTM C1218/C1218M–17. ASTM C1218. Standard Test Method for Water-Soluble Chloride in Mortar and
Concrete; American Society for Testing and Materials (ASTM): Philadelphia, PA, USA, 2015.
68. Analysis of water soluble chloride content in concrete. Available online: [Link]
publis/[Link] (Accessed on 28 October 2017).
69. Arya, C.; Newman, J.B. An Assessment of Four Methods of Determining the Free Chloride Content of
Concrete. Mater. Struct. 1990, 23, 319–330. [CrossRef]
70. Jin, M.; Jiang, L.; Zhu, Q. Monitoring chloride ion penetration in concrete with different mineral admixtures
based on embedded chloride ion selective electrodes. Constr. Build. Mater. 2017, 143, 1–15. [CrossRef]
71. Climent-Llorca, M.A.; Viqueira-Perez, E.; Lopez-Atalaya, M.M. Embeddable Ag/AgCl sensors for in-situ
monitoring chloride contents in concrete. Cem. Concr. Res. 1996, 26, 1157–1161. [CrossRef]
72. De Vera, G.; Climent, M.A.; Anton, C.; Hidalgo, A.; Andrade, C. Determination of the selectivity coefficient
of a chloride ion selective electrode in alkaline media simulating the cement paste pore solution. J. Electroanal.
Chem. 2010, 639, 43–49. [CrossRef]
73. Climent, M.A.; Anton, C.; de Vera, G.; Hidalgo, A.; Andrade, C. The interference of OH-ions in the
potentiometric determination of free Cl-in cement paste pore solution. In Proceedings of the International
RILEM Conference on Advances in Construction Materials through Science and Engineering, Hong Kong,
China, 5–7 September 2011; RILEM Publications: Bagneux, France, 2011.
74. Page, C.L. Mechanism of corrosion protection in reinforced concrete marine structures. Nature 1975, 258,
514–515. [CrossRef]
75. Sandberg, P. Chloride initiated reinforcement corrosion in marine concrete. In Report TVBM 1015; Division of
Building Materials, LTH, Lund University: Lund, Sweden, 1998.
76. Ghods, P. Multi-Scale Investigation of the Formation and Breakdown of Passive Films on Carbon Steel Rebar
in Concrete. Ph.D. Thesis, Carleton University, Ottawa, ON, Canada, 2010.
77. Page, C.; Havdahl, J. Electrochemical monitoring of corrosion of steel in microsilica cement pastes.
Mater. Struct. 1985, 18, 41–47. [CrossRef]
78. Carmody, W.R. A study of the silver chloride electrode. J. Am. Chem. Soc. 1929, 51, 2901–2904. [CrossRef]
79. Ha, H.; Payer, J. The effect of silver chloride formation on the kinetics of silver dissolution in chloride
solution. Electrochim. Acta 2011, 56, 2781–2791. [CrossRef] [PubMed]
80. Janata, J. Principles of Chemical Sensors; Springer Book: New York, NY, USA, 2009.
Sensors 2017, 17, 2482 21 of 22

81. Polk, B.J.; Stelzenmuller, A.; Mijares, G.; MacCrehan, W.; Gaitan, M. Ag/AgCl microelectrodes with improved
stability for microfluidics. Sens. Actuators B Chem. 2006, 114, 239–247. [CrossRef]
82. Dobos, D. Electrochemical Data: A Handbook for Electrochemists in Industry and Universities; Elsevier Science
and Technology: Amsterdam, The Netherlands, 1975.
83. De Vera, G.; Hidalgo, A.; Climent, M.A.; Andrade, C.; Alonso, C. Chloride-Ion Activities in Simplified
Synthetic Concrete Pore Solutions: The Effect of the Accompanying Ions. J. Am. Ceram. Soc. 2000, 83, 640–644.
[CrossRef]
84. Hidalgo, A.; Vera, G.; Climent, M.A.; Andrade, C.; Alonso, C. Measurements of chloride activity coefficients
in real Portland cement paste pore solutions. J. Am. Ceram. Soc. 2001, 84, 3008–3012. [CrossRef]
85. Jaya, S.; Rao, T.P.; Rao, G.P. Mono- and multilayer formation studied of silver chloride on silver electrodes
from chloride-containing solutions. J. Appl. Electrochem. 1987, 17, 635–640. [CrossRef]
86. Jovic, B.M.; Jovic, V.D.; Drazic, D.M. Kinetics of chloride ion adsorption and the mechanism of AgCl layer
formation on the (111), (100) and (110) faces of silver. J. Electroanal. Chem. 1995, 399, 197. [CrossRef]
87. Katan, T.; Szpak, S.; Douglas, N.B. Silver/silver chloride electrodes: Surface morphology on charging and
discharging. J. Electrochem. Soc. 1974, 121, 757–764. [CrossRef]
88. Jin, X.; Lu, J.; Liu, P.; Tong, H. The electrochemical formation and reduction of a thick AgCl deposition layer
on a silver substrate. J. Electroanal. Chem. 2003, 542, 85–96. [CrossRef]
89. Lal, H.; Thirsk, H.R.; Jones, W.F.K.W. A study of the behavior of polarized electrodes, Part I—The silver/silver
halide system. Trans. Faraday Soc. 1951, 47, 70–77. [CrossRef]
90. Birss, V.I.; Smith, C.K. The anodic behavior of silver in chloride solution—I. The formation and reduction of
thin silver chloride films. Electrochim. Acta 1987, 32, 259–268. [CrossRef]
91. Stoica, D.; Brewer, P.J.; Brown, R.J.; Fisicaro, P. Influence of fabrication procedure on the electrochemical
performance of Ag/AgCl reference electrodes. Electrochim. Acta 2011, 56, 10009–10015. [CrossRef]
92. Beck, T.R.; Rice, D.E. Conductivity of anodic silver chloride during formation. J. Electrochem. Soc. 1984, 131,
89–93. [CrossRef]
93. Elsener, B.; Zimmermann, L.; Bohni, H. Non-destructive determination of the free chloride content in cement
based materials. Mater. Corros. 2003, 54, 440–446. [CrossRef]
94. Duffo, G.S.; Farina, S.B.; Giordano, C.M. Characterization of solid embeddable reference electrodes for
corrosion monitoring in reinforced concrete structures. Electrochim. Acta 2009, 54, 2010–2020. [CrossRef]
95. Montemor, M.F.; Alves, J.H.; Simoes, A.M.; Fernandes, J.C.S.; Lourenço, Z.; Costa, A.J.S.; Appleton, A.J.;
Ferreira, M.G.S. Multiprobe chloride sensor for in situ monitoring of reinforced concrete structures.
Cem. Concr. Compos. 2006, 28, 233–236. [CrossRef]
96. Bard, A.J.; Parsons, R.; Jordan, J. Standard Potentials in Aqueous Solution; Marcel Dekker: New York, NY, USA,
1985; pp. 294–312.
97. Giles, R.D. The anodic behaviour of silver single crystal electrodes in concentrated chloride solutions.
J. Electroanal. Chem. Interfacial Electrochem. 1970, 27, 11–19. [CrossRef]
98. Winston, R. Uhlig’s Corrosion Handbook; Wiley: New York, NY, USA, 2000.
99. Welham, N.J.; Kelsall, G.H.; Diaz, M.A. Thermodynamics of Ag-Cl-H2 O, Ag-Br-H2 O and Ag-I-H2 O systems
at 298 K. J. Electroanal. Chem. 1993, 361, 39–41. [CrossRef]
100. Angst, U.; Elsener, B.; Larsen, C.K.; Vennesland, O. Potentiometric determination of the chloride ion activity
in cement based materials. J. Appl. Electrochem. 2010, 40, 561–573. [CrossRef]
101. Birss, V.I.; Wright, G.A. The kinetics of the anodic formation and reduction of phase silver sulfide films on
silver in aqueous sulfide solutions. Electrochim. Acta 1981, 26, 1809–1817. [CrossRef]
102. Graedel, T.E. Corrosion mechanisms for silver exposed to the atmosphere. J. Electrochem. Soc. 1992, 139,
1963–1970. [CrossRef]
103. Femenias, Y.S.; Angst, U.; Caruso, F.; Elsener, B. Ag/AgCl ion-selective electrodes in neutral and alkaline
environments containing interfering ions. Mater. Struct. 2016, 49, 2637–2651. [CrossRef]
104. Moody, G.J.; Rigdon, L.P.; Meisenheimer, R.G.; Frazer, J.W. Selectivity parameters of homogeneous solid-state
chloride ion-selective electrodes and the surface, morphology of silver chloride-silver sulphide discs under
simulated interference conditions. Analyst 1981, 106, 547–556. [CrossRef]
105. Suzuki, H.; Hiratsuka, A.; Sasaki, S.; Karube, I. Problems associated with the thin-film Ag/AgCl reference
electrode and a novel structure with improved durability. Sens. Actuators B Chem. 1998, 46, 104–113.
[CrossRef]
Sensors 2017, 17, 2482 22 of 22

106. Altunbulduk, T.; Kocker, H.M.; Frenzel, W. Studies on the elimination of sulfide interference in the
potentiometric determination of chloride using ion selective electrodes in a flow injection system. Fresenius’ J.
Anal. Chem. 1995, 351, 593–598. [CrossRef]
107. Raynauld, J.P.; Laviolette, J.R. The silver-silver chloride electrode: A possible generator of offset voltages
and currents. J. Neurosci. Methods 1987, 19, 249–255. [CrossRef]
108. Janz, G.J.; Ives, D.J. Silver, silver chloride electrodes. Ann. N. Y. Acad. Sci. 1968, 148, 210–221. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license ([Link]