Conference PaperPDF Available

Chloride-induced corrosion of steel embedded in alkali-activated materials: state of the art


Abstract and Figures

Alkali-activated materials are a group of alternative binders based on aluminosilicate precursor and an alkali activator. Since precursors are mostly waste materials and by-products, AAMs are consideredto be a valid and more environmental-friendly alternative to Portland cement. Although the AAMs show good performance, there are not enough studies in the literature about their long-term durability performance related to chloride ingress and embedded steel corrosion. The issue of this paper is to highlight the main difference between OPC and the different types of AAMs, and to give a short introduction about the state of art of this embedded steel corrosion.
Content may be subject to copyright.
1University of Zagreb, Faculty of Civil Engineering, Department of Materials, Croaa,
2University of Zagreb, Faculty of Civil Engineering, Department of Materials, Croaa,
3University of Sheeld, Department of Materials Science and Engineering, UK, j.provis@she
Antonino Runci1, Marijana Serdar2, John Provis3
Alkali-acvated materials are a group of alternave binders based on aluminosilicate pre-
cursor and an alkali acvator. Since precursors are mostly waste materials and by-products,
AAMs are consideredto be a valid and more environmental-friendly alternave to Portland
cement. Although the AAMs show good performance, there are not enough studies in the
literature about their long-term durability performance related to chloride ingress and em-
bedded steel corrosion. The issue of this paper is to highlight the main dierence between
OPC and the dierent types of AAMs, and to give a short introducon about the state of art
of this embedded steel corrosion.
Key words: alkali-acvated material, review, durability, steel corrosion, chloride ingress,
chloride binding capacity
Korozija čelika u alkalno-akviranim materijalima
uzrokovana kloridima: pregled stanja područja
Materijali akvirani alkalijama su skupina alternavnih veziva na bazi aluminosilikatnog pre-
kursora i alkalnog akvatora. Oni bi u budućnos mogli bi zastupljeni kao valjana i eko-
loški prihvatljiva alternava portland cementu. Osim što su AAM pokazali dobre rezultate,
u literaturi nema dovoljno studija o dugotrajnoj izvedbi klorida i ugrađene korozije čelika.
Problem ovog rada je osvijetli glavnu razliku između OPC-a i različih pova AAM-ova i da
kratak uvod o stanju ove ugrađene korozije čelika.
Ključne riječi: alkalijski akvirani materijal, pregled, trajnost, korozija čelika, prodor klorida,
sposobnost vezanja klorida
Chloride-induced corrosion of steel embedded
in alkali-acvated materials: state of the art
DOI: hps://
1 Introducon
Alkali-Acvated Material (AAM) is a general name used to indicate a group of alter-
nave binders, obtained by the reacon between an aluminosilicate source, usually
an industrial by-product such as slag from iron and steel producon, coal y ash
from thermoelectric plants, among others, and an alkali acvator which is usually
a concentrated aqueous soluon of alkali hydroxide, silicate, carbonate or sulfate
[1-3]. Since AAMs are based on industrial byproducts and waste materials, they
are considered to reduce the CO2 emission coming from the cement clinkerisaon
process, and also divert industrial byproducts from landlling. Another important
advantage to use byproducts is the preservaon of natural materials which are nor-
mally used in producon of Portland cement clinker. It is for these reasons that
AAM have received a lot of interest from the scienc community as one of the
opons when considering more sustainable alternaves to ordinary Portland Ce-
ment (OPC).
The AAMs can be disnguished into two dierent categories based on the nal
phase assemblage [1-5]:
Low Ca systems – based on the acvaon of a precursor with low Ca content,
such as y ash or metakaolin, where the main reacon product is a three-dimen-
sional alkali-aluminosilicate hydrate (N-A-S-H) type gel;
High Ca systems – based on the acvaon of a precursor with high Ca content,
such as slag, where the main reacon product is calcium-aluminosilicate hydrate
(C-A-S-H) type gel.
The AAMs show many engineering characteriscs comparable to OPC, and pub-
lished studies report higher stability when exposed to elevated temperature [6],
higher resistance against chemical aack [7-9] and potenally beer resistance to
freeze-thaw cycles [10, 11], all compared to OPC. However, being a comparavely
young engineering material, the quanty of available durability data is limited, and
the long-term performance of structures made with AAMs is yet to be determined.
Available knowledge on degradaon mechanisms from OPC cannot be directly
transferred to alkali-acvated systems, due to the dierence in reacon products,
pore soluon chemistry, and microstructure of the matrix. Addionally, what makes
it more challenging is that any generalizaon in the case of AAMs is praccally im-
possible since all the parameters of the system dier for each AAM, depending on
the type and chemistry of precursor material and the type and amounts of acva-
tors used [1,4,5,11]. Among the degradaon mechanisms that seek aenon is the
de-passivaon of steel reinforcement embedded in alkali-acvated concrete.
The most common causes of reinforcement corrosion are localized de-passivaon
of the steel surface due to chloride ingress, and/or more general de-passivaon due
Chloride-induced corrosion of steel embedded in alkali-acvated materials: state of the art
to acidicaon of the pore soluon as a result of carbonaon of the cement paste
[13]. In either of the two cases, behaviour of steel when embedded in concrete can
be divided into three phases:
passivaon phase during which passive lm is formed on the surface of the
acvaon phase or corrosion iniaon - during which passive lm losses the
stability and nally breaks down, and
degradaon phase or corrosion propagaon - during which corrosion products
start to precipitate on the surface of the steel, leading to concrete cracking.
In all of these three phases chemical and physical properes of concrete surround-
ing the steel play a crucial role. It can therefore be expected that once steel is em-
bedded in AAM, all three phases will occur with some parcularies as compared
to OPC.
2 Passivaon phase – Formaon of passive lm
The passive lm is the external layer formed on the surface of reinforcing steel aer
contact with an alkaline environment, like cement. It is a protecve lm formed at
the beginning of the embedding of steel, but the quality and the stability of that
lm depend on the exposure duraon and the chemical composion of the pas-
sivang soluon. In a classical OPC system in which black steel reinforcement is
embedded, the passive lm has a bi-layer structure:
inner layer Fe2+-rich layer oxy-hydroxide (1-3 nm)
outer layer Fe3+-rich hydroxide lm (5-10 nm) [17].
The behaviour of steel in AAMs is more complex because the passivaon process is
regulated by the redox potenal of the steel and oxygen availability, pH and chem-
istry of the surrounding environment: the dierent precursor material and acvator
of AAM change this equilibrium. The most important factor that aects the stability
of passive lm is the pH and, consequently, the chemical composion of pore solu-
on. The pore soluon of a system is a funcon of the composion of the blend; for
that reason, the chemical characteriscs of SCMs (Supplementary Cement Materi-
al) (i.e. chloride binding capacity, alkalinity) signicantly aect the corrosion rates
of steel in concrete.
2.1 pH value
The pH is the main factor inuencing the stability of the passive lm on the steel.
Gouda et al. [22] found that using dierent NaOH and Ca(OH)2 in soluons with
dierent pH the passive lm is stable unl pH 12.1-11.75. The leaching of OH- from
AAS and, probably, also the accelerated carbonaon could reduce the pH of AAS
faster than OPC [24]. Shi et al. [25] showed that AAS has a higher pH at early ages,
but aer one year of exposure to chloride environment pH value decreases faster
than in the case of OPC. Similar was observed in the study by Ma et al. [27]. The pH
value decreased sharply near the surface zone of AAS due to the outward diusion
from AAS into the chloride soluon. In OPC, Ca(OH)2 would dissolve in the pore
soluon to buer its alkalinity; this explains the reason for a constant pH value in
the OPC concrete at dierent depths. However, Ca(OH)2 is not one of the reacon
products of AAS, and for that reason the loss of alkalinity is more severe in the case
of the AAS concretes, parcularly in the near surface zone, as can be seen in Figure
The hydroxyl ion concentraon of OPC is typically in the range of 0.53 M to 0.71
M. Sco and Alexander [20] showed the changing concentraon of OH- for AAMs
blends and OPC aer 90 days of maturaon; in the system based on OPC it increases by
2.5-2.8 mes, the blends contained y ash, silica fume or slag show a more constant
value or a decrease that is approximately 5 mes lower than OPC. This reducon
in OH- concentraon inhibits the capacity of the systems to preserve the stability
of the passive lm on the steel surface. Babaee and Castel [28] displayed that pH
of AAM is lower than OPC; pH value of AAM was around 11.5 at the level of the re-
inforcing bar compared to the widely accepted pH of about 12.5 for uncarbonated
2.2 Sulde content
It is not only the pH value of the soluon that has the paramount inuence on the
stability of the passive lm, it is also the chemical composion of pore soluon [23].
Figure 1. pH proles determined at the end of the chloride ponding exposure regime [27]
Chloride-induced corrosion of steel embedded in alkali-acvated materials: state of the art
In the case of low-Ca systems, Mundra et al. [14] showed that despite the dier-
ent chemical composion of precursor materials, this system has a broadly similar
passivaon process to that of OPC systems. However, the high-Ca systems based
on blast furnace slag, or the high-volume blends of slag with Portland cement, can
sll have a convenonal passivaon preocess or show a dierent behavior mostly
due to the sulde content in the slag. The general range of sulde concentraons in
pore soluons for slag replacement levels up to 90 % in PC blends appear to be in
the range of 110 to 300 mg/l in addion to thiosulfate (S2O3
2−), found by Glasser et
al. [21], and these concentraons are similar in AAS cements [12]. At the beginning
the suldes are oxidised to sulfate by the oxygen available in the system, thus de-
pleng the oxygen concentraon at the steel, creang a reducing environment and
leading to a more negave value of corrosion potenal [18, 35], as seen in Figure 2
a). The reducing environment would favor the formaon of Fe(II) rather than Fe(III),
which forms the α-Fe2O3 responsible for passivaon [20]. Instead of an iron oxide
passive lm, the precipitaon of mackinawite (Fe1+XS) on the steel surface occurs
[19, 24]. According to this the passivaon process would be inhibited. Sll some
studies have shown that corrosion rate generally decreased with the increase of the
concentraon of sulde due to the reducing environment provided by sulde, Fig-
ure 2 b) [27]. Since the steel is embedded in an alkaline soluon devoid of oxygen,
this absence at the steel surface inhibits the cathodic reacon and the corrosion is
controlled [26].
Figure 2. a) Evoluon of the potenal values as a funcon of me, for steel embedded in AAS (BFS),
OPC + slag (WPS) and Portland cement (WPC) mortars immersed in water [35], and b) Rela-
onship between the corrosion rate and the sulphide concentraon [27]
3 Acvaon phase – Corrosion iniaon
The steel surface works as a mixed electrode composed of anodes and cathodes
electrically connected through the body of the steel itself, upon which coupled
anodic and cathodic reacons take place. Concrete pore water funcons as an aque-
ous medium, i.e. a complex electrolyte [13, 29]. The anodic reacon is an oxidaon
process that depends on the pH and the presence of aggressive anions, such as Cl-.
The cathodic reacon is a reducon process that depends on the O2- availability and
the pH [13]. There are numerous parameters aecng me of corrosion iniaon
and hereaer only two main inuences (crical chloride content and chloride bind-
ing) will be considered in detail.
3.1 Crical chloride content
The presence of chloride ions will contribute to the breakdown of the passive layer,
while other anions such as OH are responsible for its stability and have inhibing
properes. There is believed to be a point at which the concentraon of aggressive
ions overcomes the inhibing ions and ‘corrosion’ can iniate, and this is called the
crical chloride content (Ccrit) [15].
Babaee and Castel [31] analysed in depth the Cl- threshold value for dierent AAM.
According to the authors the Ccrit of AAM based on y ash is 0.19-0.69 (% by binder
mass), compared to the 0.2-0.4 % by mass of binder for OPC reinforced concrete
[16]. For slag-dominated samples, when increasing slag content the Ccrit decreases
independently of the alkali content or the sodium silicate acvator modulus, which
is indicang a less-developed passive lm in these binders that can be broken down
comparavely easily. The lower stability of the passive layer can be aributed to
the oxidaon of sulde anions, which was explained in previous chapter. The AAMs
with higher alkaline acvator have higher chloride thresholds due to a beer dis-
soluon of the precursors that generated a more homogeneous binder. As a result,
a less permeable passive layer formed around the bars which took longer to break
down. Another inuenal factor could be the presence of more OH- ions that re-
acted with the iron caons to form the iron hydroxide layer around the bars [31].
3.2 Chloride binding capacity
The value of total chlorides does not always coincide with the free chloride (Cl- in
pore soluon) because during the corrosion process the chloride can move through
the pore soluon or/and bind with the hydraon products of the system; it is si-
multaneously present in the pore soluon and in the cement matrix. The amount
of bound chloride depends on the percentage chloride binding capacity (Pcb) of the
Chloride-induced corrosion of steel embedded in alkali-acvated materials: state of the art
In OPC, the Cl may be bounded through physical bonds or chemical bonds. The rst
bond depends on the amount C-S-H, amount of acvator, w/b rao and amount
of aggregates and can be released into the pore soluon on diluted condions.
The second one depends on the amount of C3A and C4AF from which are formed
the Friedel’s salt (3CaOAl2O3CaCl210H2O) and other AFm phases. The AAMs show
dierent behavior due the chemical and mineralogical composions. The absenc-
es of C3A and C4AF hinder generally the precipitaon of Friedel’s salt [31, 26], but
chloride could bind in the AFms and other layered phases that do form [32]. Ac-
cording to Ke et al. [32] hydrotalcite-like (Mg-Al) phases and strätlingite (AFm struc-
ture) can take up chloride from highly alkaline soluons with dierent inial [Cl]/
[OH] raos. In hydrotalcite-like phases the Cl- is mainly adsorbed (~90 %) and the
amounon-exchanged is just 10 %, while the strätlingite has less dominant surface
adsorpon, and lace substuon of chloride also takes place [31, 32]. Khan and
Kayali [33] demonstrated that the blast furnace slag inuences more than OPC the
binding capacity: increasing the amount of slag the free chloride content is reduced
and Pcb increased, this suggests that the Cl- bound in AFm or hydrotalcite is higher
than absorbed in C-S-H. Maes et al. [30] explained that increasing the amount of
slag in an OPC blend the Friedel’s salt formaon increases because it is linked to the
Al2O3 content. The chloride binding ability gradually decreases with increasing the
quanty of sulfate in blended systems because SO4
2- has high anity for the inter-
layer of hydrotalcite, more than Cl-.
In low-Ca AAM systems, the N-A-S-H favors more the adsorpon of Cl- than C-(A)-S-H.
Alkali binding increases as calcium content decreases with y ash addion, hence
reducing the alkalinity in the pore soluon. High alkali binding capacity is also at-
tributed to the surface charge of the N-A-S-H, which aracts more ions from the
surrounding pore soluon [29, 27]. Lee and Lee [34] found a correlaon between
the aluminosilicate gel and the chloride-binding capacity in AAMs; increasing the
aluminosilicate gel with the depth, the Pcb increases; therefore, the increase of
bounded Cl with the depth is linked with the aluminosilicate gel content in AAMs;
the chloride-binding capacity increases with an increase in the y ash content and
with a decrease in the slag content because the addion of y ash results in great-
er physical absorpon of chloride compared to the C-(A)-S-H gel due to the high-
er surface area of the aluminosilicate gel in the alkali-acvated y ash, Figure 3.a)
The NaOH concentraon has an important role during the chloride binding because
more chloride ions were bound in geopolymer with higher NaOH concentraon
than with lower NaOH concentraon because the physical binding of chloride ions
increases due the greater geo-polymerisaon, Figure 3 b) [29].
Figure 3. a) Bound chloride-to-total chloride rao ( %) and penetraon depth (mm) over the N-A-S-H
gel amount ( %) in alkali-acvated y ash-slag samples [34]; and b) The eect of sodium hy-
droxide (NaOH) concentraon on chloride binding capacity of y ash-based AAM concrete
4 Conclusion
The alkali-acvated material systems have dierent chemical and mineralogical
composion compared to OPC, which strongly inuences all phases of corrosion
process. In the case of low-Ca AAM, the higher amount of N-A-S-H gel rather than
C-S-H favors chloride adsorpon, but the absence of C3A and C4AF hinder the for-
maon of AFm phases that can give chemical binding. In the case of high-Ca AAM,
such as AAS systems, a high content of sulde generates reducve environment in
which thinner and potenally less durable passive lm is preferenally formed. In
such systems, the main reacon product is C-A-S-H, and the high Ca and Al contents
favor the AFm precipitaon and the chemical chloride binding. Although both types
of AAMs showed binding of chlorides, the main dierence is in the gel phases. The
ner structure of C-A-S-H reduces the chloride penetraon but also the physical
binding, while the porous structure of N-A-S-H increases the penetraon and the
adsorpon of chlorides.
The dierent chemical composion of AAMs makes it dicult to use the OPC stand-
ards as a reference. The variability is high for each type of system and depends
on the type of precursor and the specic blends made. Establishing and validang
durability and corrosion tesng methods for alkali-acvated concretes remains the
major obstacle to their commercial adopon in demanding structural applicaons,
and ulmately their acceptance in naonal and internaonal regulatory standards
for structural concrete. Finally, even though AAM aract a lot of aenon due to the
potenal CO2 saving, calculaons on CO2 emissions are more complex than merely
considering the volume of byproducts substung cement. Some of the materials
used as precursors must be milled which also consumes a lot of energy and emits
some CO2. Addional challenge lays in the fact that most of these materials are cur-
Chloride-induced corrosion of steel embedded in alkali-acvated materials: state of the art
rently considered as waste, which makes their transport and industrial applicaon
dicult. The properes and use of these materials in the future must be regulated
by internaonal standard, such as EN 450 for y ash.
This incepve study was nancially supported by DuRSAAM project (hp://www.
[1] Provis J.L., Geopolymers and other alkali acvated materials: Why, how, and what?
Materials and Structures 47: 11–25 (2014).
[2] Provis J.L., van Deventer J.S.J., Alkali-acvated materials: State-of-the-Art Report,
RILEM TC 224-AAM. Springer/RILEM, Berlin (2014).
[3] Provis J.L., Palomo A., Shi C., Advances in understanding alkali-acvated materials.
Cement and Concrete Research 78:110-125 (2015).
[4] Li C.,Sun H., Li L., A review: The comparison between alkali-acvated slag (Si+Ca) and
metakaolin (Si+Al) cements. Cem Concr Res. 40: 1341–1349 (2010).
[5] Mundra S., Bernal Lopez S., Criado M., Hlaváček P., Ebell G., Reinemann S., Gluth G.,
Provis J.L., Steel corrosion in reinforced alkali-acvated materials. RILEM Technical
Leers, 2: 33-39 (2017).
[6] Kupwade-Pal K., Soto F., Kunjumon A., Allouche E.N., Mainardi D.S., Mul-
scale modeling and experimental invesgaons of geopolymeric gels at elevated
temperatures, Comput. Struct. 122: 164-277 (2013).
[7] Fernandez-Jimenez A., García-Lodeiro I., Palomo A., Durability of alkali-acvated y ash
cemenous materials, J. Mater. Sci. 42:3055–3065 (2006).
[8] Bakharev T., Resistance of geopolymer materials to acid aack, Cem. Concr. Res. 35:
658–670 (2005).
[9] Bakharev T., Durability of geopolymer materials in sodium and magnesium sulfate
soluons, Cem. Concr. Res. 35: 1233–1246 (2005).
[10] Fu Y., Cai L., Wu Y., Freeze–thaw cycle test and damage mechanics models of alkali-
acvated slag concrete, Constr. Build. Mater. 25: 3144–3148 25 (2011).
[11] Zhang Y., Sun W., Fly ash based geopolymer concrete, Indian Concr. J. 80: 20–24 (2006).
[12] Gruskovnjak A., Lothenbach B., Winnefeld F., Figi R., Ko S.-C., Adler M., Mäder U.,
Hydraon mechanisms of super sulphated slag cement, Cement and Concrete Research
38: 983–992 (2008).
[13] Shamsad A, Reinforcement corrosion in concrete structures, its monitoring and service
life predicon––a review, Cem Concr. Compos., 24: 459-471 (2003).
[14] Mundra S., Criado M., Bernal S.A., Provis J.L., Chloride-induced corrosion of steel
rebars in simulated pore soluons of alkali-acvated concretes, Cem. Con. Res., volume
100:385-397 (2017).
[15] Angst U, Elsener B., Larsen C.K., Vennesland, Ø., Crical chloride content in reinforced
concrete — A review, Cement and Concrete Research 39, 1122-1138 (2009).
[16] European Standard EN 206, Concrete — Part 1: Specicaon, Performance, Producon
and Conformity, European Commiee for Standardisaon, 2011.
[17] Angst U., Geiker A., Mee R., Michel A., Gehlen C., Wong H., Isgor O., Elsener B.,
Hansson C.M., François R., Hornbostel K., Polder R., Alonso M.C., Sanchez M., Correia
M.J., Criado M., Sagüés A., Buenfeld N., The steel–concrete interface, Materials and
Structures 50: 143 (2017).
[18] Sco A., Alexander M.G., Eect of supplementary cemenous materials (binder type)
on the pore soluon chemistry and the corrosion of steel in alkaline environments.
Cem. Concr. Res. 89: 45–55 (2016).
[19] Shoesmith D.W., Bailey M.G., Ikeda B., Electrochemical formaon of mackinawite in
alkaline sulphide soluons. Electrochim Acta 23: 1329–1339 (1978).
[20] Sco A., Alexander M.G., Eect of supplementary cemenous materials (binder type)
on the pore soluon chemistry and the corrosion of steel in alkaline environments,
Cement and Concrete Research 89: 45-55 (2016).
[21] Glasser F., Luke K., Angus M., Modicaon of cement pore uid composions by
pozzolanic addives, Cement and Concrete Research 18: 165–178 (1988).
[22] Gouda V., Corrosion and corrosion inhibion of reinforcing steel I: immersed in alkaline
soluons, Brish Corrosion Journal5 198–203(1970).
[23] Ghods P., Isgor O.B., McRae G., Miller T., The eect of concrete pore soluon composion
on the quality of passive oxide lm on black steel reinforcement, Cement and Concrete
Composites 31: 2–11(2009).
[24] Moncelli C., Natali M., Balbo A., Chiavari C., Zanoo F., Manzi S., Bignozzi M., A study
on the corrosion of reinforcing bars in alkali-acvated y ash mortars under wet and
dry exposures to chloride soluons, Cement and Concrete Research 87: 53–63 (2016).
[25] Shi J., Ming J., Sun W., Electrochemical behaviour of a novel alloy steel in alkali-acvated
slag mortars, Cement and Concrete Composites 92: 110–124 (2018).
[26] Ismail I., Bernal S.A., Provis J.L., San Nicolas R., Brice D.G., Kilcullen A.R., Hamdan S.,
van Deventer J.S.J., Inuence of y ash on the water and chloride permeability of alkali-
acvated slag mortars and concretes, Construcon and Building Materials 48: 1187–
1201 (2013).
Chloride-induced corrosion of steel embedded in alkali-acvated materials: state of the art
[27] Ma Q., Nanukuan S.V., Basheer P.A.M., Bai Y., Yang C., Chloride transport and the
resulng corrosion of steel bars in alkali acvated slag concretes. Mater. Struct. 49,
3663–3677 (2016).
[28] Babaee M., Castel A., Chloride-induced corrosion of reinforcement in low-calcium y
ash-based geopolymer concrete, Materials and Structures 49:3663–3677 (2016).
[29] Chindaprasirt P., Chalee W., Eect of sodium hydroxide concentraon on chloride
penetraon and steel corrosion of y ash-based geopolymer concrete under marine
site, Construcon and Building Materials 63: 303–310(2014).
[30] Maes M., Gruyaert E., De Belie N., Resistance of concrete with blast-furnace slag against
chlorides, invesgated by comparing chloride proles aer migraon and diusion,
Materials and Structures 46:89–103 (2013).
[31] Babaee M., Castel A., Chloride diusivity, chloride threshold, and corrosion iniaon in
reinforced alkali-acvated mortars: Role of calcium, alkali, and silicate content, Cement
and Concrete Research 111: 56-71 (2018).
[32] Ke X., Bernal S.A., Provis J.L., Uptake of chloride and carbonate by Mg-Al and Ca-Al
layered double hydroxides in simulated pore soluons of alkali-acvated slag cement,
Cement and Concrete Research 100: 1-13 (2017).
[33] Khan M.S.H., Kayali O., Chloride binding ability and the onset corrosion threat on alkali
acvated GGBFS and binary blend pastes, European Journal of Environmental and Civil
Engineering 8: 1023-1039(2018).
[34] Lee N.K., Lee H.K., Inuence of the slag content on the chloride and sulfuric acid
resistances of alkali-acvated y ash/slag paste, Cement and Concrete Composites 72:
168-179 (2016).
[35] Criado M., Provis J., Alkali acvated slag mortars provide high resistance to chloride-
induced corrosion of steel, Froners in Materials 5, #34 (2018).
... In this regard, the result reported in [51] is promising; it shows that the amount of bound chlorine (P cb ) depends ce [60] and is reasonable to counteract the aggressive effects of the anionic residue of strong acids on steel reinforcement in concrete. It could be assumed that the advantage in the formation of these phases would depend both on the composition and content of SSA and the composition (alkalinity) of the cement matrix, the morphology of hydrate new formations, and the transporting properties of the structure of concrete in general. ...
This paper proposes a technique to prevent the corrosion of steel reinforcement in concrete based on slag cement (SC) activated by Na(K) salts of strong acids (SSA) in the composition of by-pass cement kiln dust (BP). The technique implies using additional modifiers in the form of the Portland cement CEM I 42,5 R and the calcium-aluminate admixture (CAA) С3А∙6H2O. It is shown that adding the Portland cement contributes to enhancing the intensifying influence of BP on the SC hydration, accompanied by an increase in the strength of artificial stone. This effect is predetermined by the formation of hydrosilicates in hydration products with an increased crystallization degree in the form of CSH(I) and C2SH(A). Modifying SC with CAA ensures the intensive formation of low-soluble AFm phases in the composition of hydration products, aimed at reliable binding the SSA anions (Cl-, SO42-) that are aggressive to steel reinforcement. The study result has established the possibility to produce SC, activated by SSA, when using BP, the Portland cement, and CAA. Mathematical methods to plan the experiment were applied to produce an SC composition of "granulated blast furnace slag – BP – Portland cement – CAA", characterized by a strength class of 42.5 and a molar ratio of Cl-/OH- in a porous solution not exceeding 0.6. The resulting properties predetermine the feasibility of using SC in steel-reinforced concrete. The relevance of this work is due to the modern trends in the development of the construction industry. The introduction of cement that contains mineral additives, in particular granulated blast furnace slag, contributes to improving the environment by reducing СО2 emission. The use of such cement as a base of steel-reinforced concrete ensures the increase in their functionality and durability
Full-text available
The pore solutions of alkali-activated slag cements and Portland-based cements are very different in terms of their chemical and redox characteristics, particularly due to the high alkalinity and high sulfide content of alkali-activated slag cement. Therefore, differences in corrosion mechanisms of steel elements embedded in these cements could be expected, with important implications for the durability of reinforced concrete elements. This study assesses the corrosion behavior of steel embedded in alkali-activated blast furnace slag (BFS) mortars exposed to alkaline solution, alkaline chloride-rich solution, water, and standard laboratory conditions, using electrochemical techniques. White Portland cement (WPC) mortars and blended cement mortars (WPC and BFS) were also tested for comparative purposes. The steel elements embedded in immersed alkali-activated slag mortars presented very negative redox potentials and high apparent corrosion current values; the presence of sulfide reduced the redox potential, and the oxidation of the reduced sulfur-containing species within the cement itself gave an electrochemical signal that classical electrochemical tests for reinforced concrete durability would interpret as being due to steel corrosion processes. However, the actual observed resistance to chloride-induced corrosion was very high, as measured by extraction and characterization of the steel at the end of a 9-month exposure period, whereas the steel embedded in WPC mortars was significantly damaged under the same conditions.
Full-text available
The development of alkali-activated materials (AAMs) as an alternative to Portland cement (PC) has seen significant progress in the past decades. However, there still remains significant uncertainty regarding their long term performance when used in steel-reinforced structures. The durability of AAMs in such applications depends strongly on the corrosion behaviour of the embedded steel reinforcement, and the experimental data in the literature are limited and in some cases inconsistent. This letter elucidates the role of the chemistry of AAMs on the mechanisms governing passivation and chloride-induced corrosion of the steel reinforcement, to bring a better understanding of the durability of AAM structures exposed to chloride. The corrosion of the steel reinforcement in AAMs differs significantly from observations in PC; the onset of pitting (or the chloride ‘threshold’ value) depends strongly on the alkalinity, and the redox environment, of these binders. Classifications or standards used to assess the severity of steel corrosion in PC appear not to be directly applicable to AAMs due to important differences in pore solution chemistry and phase assemblage.
Full-text available
This is a State of the Art Report resulting from the work of RILEM Technical Committee 224-AAM in the period 2007-2013. The Report summarises research to date in the area of alkali-activated binders and concretes, with a particular focus on the following areas: binder design and characterisation, durability testing, commercialisation, standardisation, and providing a historical context for this rapidly-growing research field.
Full-text available
The passivation and chloride-induced depassivation of steel rebars immersed in varying alkaline environments (0.80 M, 1.12 M and 1.36 M NaOH solutions), simulating the pore solutions of low-Ca alkali-activated concretes, were investigated using a range of electrochemical techniques. The passive film on the steel rebars was complex in chemical makeup, composed of Fe–hydroxides, oxy-hydroxides and oxides. An increased degree of passivation of the rebars was observed when exposed to solutions with higher hydroxide concentrations. The critical chloride level ([Cl−]/[OH−] ratio) required to induce depassivation of steel was strongly dependent on the alkalinity of the pore solution, and was found to be 0.90, 1.70 and 2.40 for 0.80 M, 1.12 M and 1.36 M NaOH solutions, respectively. These values all correspond to a constant value of [Cl−]/[OH−]3 = 1.25, which is a novel relationship to predict the onset of pitting, interlinking chloride concentration and the solubility of the passive film.
Full-text available
Chloride ingress and carbonation are major causes of degradation of reinforced concrete. To enable prediction of chloride ingress, and thus to improve the durability of structural alkali-activated slag cement (AAS) based concretes, it is necessary to understand the ionic interactions taking place between chlorides, carbonates, and the individual solid phases which comprise AAS. This study focused on two layered double hydroxides (LDH) representing those typically identified as reaction products in AAS: an Mg-Al hydrotalcite-like phase, and an AFm structure (strätlingite), in simulated AAS pore solutions. Surface adsorption and interlayer ion-exchange of chlorides occurred in both LDH phases; however, chloride uptake in hydrotalcite-group structures is governed by surface adsorption, while strätlingite shows the formation of a hydrocalumite-like phase and ion exchange. For both Ca-Al and Mg-Al LDHs, decreased chloride uptakes were observed from solutions with increased [CO3²⁻]/[OH⁻] ratios, due to the formation of carbonate-containing hydrotalcite and decomposition of AFm phases, respectively.
Full-text available
Although the steel–concrete interface (SCI) is widely recognized to influence the durability of reinforced concrete, a systematic overview and detailed documentation of the various aspects of the SCI are lacking. In this paper, we compiled a comprehensive list of possible local characteristics at the SCI and reviewed available information regarding their properties as well as their occurrence in engineering structures and in the laboratory. Given the complexity of the SCI, we suggested a systematic approach to describe it in terms of local characteristics and their physical and chemical properties. It was found that the SCI exhibits significant spatial inhomogeneity along and around as well as perpendicular to the reinforcing steel. The SCI can differ strongly between different engineering structures and also between different members within a structure; particular differences are expected between structures built before and after the 1970/1980s. A single SCI representing all on-site conditions does not exist. Additionally, SCIs in common laboratory-made specimens exhibit significant differences compared to engineering structures. Thus, results from laboratory studies and from practical experience should be applied to engineering structures with caution. Finally, recommendations for further research are made.
Full-text available
This paper presents an experimental study to investigate the chloride binding ability of alkali-activated ground granulated blast furnace slag (GGBFS) and binary blend pastes. Free chloride and total chloride were measured to assess the chloride binding ability. pH measurement was performed to obtain [Cl⁻/OH⁻] ratio to assess the onset corrosion threat. The results showed that bound chloride significantly increased in GGBFS pastes and it gradually increased with GGBFS content in the binary blends. GGBFS paste mixed with deionised water and binary blend pastes showed corrosion risks although their free chloride contents were significantly low. The results further demonstrated that Ordinary Portland cement may be a more appropriate option than NaOH for activating GGBFS in terms of corrosion resistance.
The aim of this study is to investigate systematically the chloride diffusivity and chloride threshold of a wide range of calcium-rich and fly ash-dominated alkali-activated samples in light of their compositional differences. To this end, the effects of various fly ash (FA)-to-slag ratios, of alkali concentrations and of silicate content in the activator were investigated. The electrochemical aspects of the passive samples were also assessed. Results show the prominent role of calcium in the matrix to reduce the chloride diffusivity. While higher alkali concentration increased the porosity and chloride diffusivities in general, lower modulus ratios provided considerably better performance in the FA-dominated samples. Chloride threshold values range between 0.19 (wt% binder mass) for calcium-rich mortars fabricated at low levels of alkalinities and 0.69 for FA-dominated mortars fabricated with highly alkaline activators. Half-cell potential and polarization resistance of alkali-activated samples were in general lower than their Portland cement counterparts.
The present paper aims to investigate the passivation capability and accelerated chloride-induced corrosion behaviour of a novel alloy steel (00Cr10MoV) and a conventional low-carbon steel (20MnSiV) in ordinary Portland cement (OPC) and alkali-activated slag (AAS) mortars. Both steels were embedded in mortars with intact mill scale. Compared with OPC mortar, AAS mortar resulted in the formation of less protective passive film for both 20MnSiV and 00Cr10MoV steels after passivation due to the presence of reducing sulphides. Despite this, the initial negative effect of AAS mortar on 20MnSiV steel can be well compensated by its denser interfacial microstructure after the occurrence of active corrosion induced by chlorides. As for 00Cr10MoV steel in AAS mortar, however, this compensating effect was less pronounced. Moreover, unexpected low passivation capability and corrosion resistance can be confirmed for 00Cr10MoV steel in both OPC and AAS mortars due to the presence of defective and Cr-depleted mill scale.
The pore solution compositions of paste samples produced with Ordinary Portland Cement (PC), slag 25%, 50% and 75%, fly ash 30%, condensed silica fume (SF) 7%, and a ternary blend of 50% PC, 43% slag and 7% SF were determined. Not only are there significant variations in the concentration of the major cations and anions but also, and equally important from the perspective of development of the passivity of steel in solution, in the level of dissolved oxygen and redox potential. Further, the impact of changes in the pore solution chemistry of cement pastes with SCMs on the passivation and corrosion of steel was investigated with mild steel in simulated pore solutions (SPS). Sulphides and thiosulphates, typically found in slag bearing pastes, appeared to reduce the chloride threshold concentration and increase the rate of corrosion in SPS, which has potential implication for the long term performance of reinforced concrete structures.