INTERNATIONAL MATERIALS REVIEWS

2023, VOL. 68, NO. 5, 556–603

https://doi.org/10.1080/09506608.2022.2145747

FULL CRITICAL REVIEW

A review of the efficiency of self-healing concrete technologies for durable and

sustainable concrete under realistic conditions
Vanessa Cappellesso a,b, Davide di Summa a,c, Pardis Pourhaji d, Niranjan Prabhu Kannikachalam a,c,
Kiran Dabral e, Liberato Ferrara c, Maria Cruz Alonso d, Esteban Camacho e, Elke Gruyaert b and
Nele De Belie a
a
Magnel-Vandepitte Laboratory, Department of Structural Engineering and Building Materials, Faculty of Engineering and Architecture,
Ghent University, Ghent, Belgium; bDepartment of Civil Engineering, Materials and Constructions, KU Leuven, Ghent, Belgium;
c
Department of Civil and Environmental Engineering, Politecnico di Milano, Milan, Italy; dConsejo Superior de Investigaciones Científicas,
Instituto Eduardo Torroja de Ciencias de la Construcción (CSIC-IETcc), Madrid, Spain; eRDC Research and Development Concretes, València,
Spain

ABSTRACT ARTICLE HISTORY

Self-healing is recognized as a promising technique for increasing the durability of concrete Received 14 July 2022
structures by healing cracks, thereby reducing the need for maintenance activities over the Accepted 21 October 2022
service life and decreasing the environmental impact. Various self-healing technologies have
KEYWORDS
been applied to a wide range of cementitious materials, and the performance has generally Self-healing concrete; crack;
been assessed under ‘ideal’ laboratory conditions. Performance tests under ideal conditions, durability; sustainability;
tailored to the self-healing mechanism, can demonstrate the self-healing potential. However, aggressive environment;
there is an urgent need to prove the robustness and reliability of self-healing under realistic degradation; durability
simulated conditions and in real applications before entering the market. This review indicators; service life
focuses on the influence of cracks on degradation phenomena in reinforced concrete
structures, the efficiency of different healing agents in various realistic (aggressive)
scenarios, test methods for evaluating self-healing efficiency, and provides a pathway for
integrating self-healing performance into a life-cycle encompassing durability-based design.

Introduction conditions, including mechanical and environmental

actions, tensile stresses may result in the formation
Concrete, the most widely used construction material, of cracks. Cracks impair the functionality/serviceabil-
is intrinsically weak in tension and is quasi-brittle. In ity and durability of RC structures by lowering their
many cases, this results in the need to couple it with stiffness and creating a preferential pathway for the
reinforcement, generally steel. When Reinforced Con- penetration of harmful agents, which will cause degra-
crete (RC) structures are exposed to service dation of the concrete matrix and the reinforcement.

CONTACT Nele De Belie [email protected] Magnel-Vandepitte Laboratory, Department of Structural Engineering and Building Materials,
Faculty of Engineering and Architecture, Ghent University, Tech Lane Ghent Science Park, Campus A, Technologiepark Zwijnaarde 60, Ghent 9052, Belgium
© 2023 Institute of Materials, Minerals and Mining and ASM International Published by Taylor & Francis on behalf of the Institute and ASM International
 INTERNATIONAL MATERIALS REVIEWS 557

Figure 1. The increasing number of publications on ‘self-healing concrete’ or ‘self-healing’ in combination with ‘cement’ or ‘paste’
or ‘mortar’ or ‘concrete’ or ‘cementitious’ (Web of Science, 13 April 2022) over the last decades showing that the development and
performance assessment of self-healing concrete is a topical subject within the research community.

Cracking of concrete can hardly be avoided (unless [12], Resilient Materials for Life [13], EndurCrete
pre-stressing is applied), but through correct mix pro- [14], ReSHEALience [15] and SMARTINCS [16].
portioning and structural design, in combination with When discussing self-healing, a distinction has to
adequate construction and quality control, effective be made between autogenous healing, which relies
control of crack width and spacing can be achieved. on the conventional constituents of the cementitious
It is thus of utmost importance to study the cracking matrix, and autonomous or autonomic healing,
phenomenon, which will govern the kinetics of which uses unconventional engineered additions to
material deterioration and the decay of structural per- provide the self-healing function [17]. The most
formance. This knowledge is essential in planning the important mechanisms that contribute to autogenous
frequency and nature of the required maintenance healing when a crack is formed and exposed to water
activities. Especially for structures in extremely aggres- are the chemical processes of continued hydration of
sive environments, rapid and durable repair is needed, yet un-hydrated binder particles near the crack walls,
the execution of which should consider the accessibil- and precipitation of calcium carbonate crystals on
ity of the degraded area, together with the total repair the crack faces as a result of the reactions between
costs [1]. the calcium ions Ca2+ in the concrete matrix and the
In this regard, technologies to heal cracks autono- carbonate ions CO2− 3 in the water or carbonic acid
mously and hence increase the durability of concrete formed from carbon dioxide CO2 in the air entering
structures and their service life have gained increasing the crack. Autogenous healing requires the presence
attention in recent years (Figure 1). The ‘functionali- of water and is more effective when crack widths are
zation’ of cementitious materials with self-healing restricted. Therefore, methods to limit crack width
properties was first proposed in the 1990s by Dry (fibre addition), provide water (e.g. addition of super-
[2–4] it was only in 2001 with the article published absorbent polymers), or enhance hydration or crystal-
by White et al. [5] in Nature about self-healing in poly- lization (mineral additions or crystalline admixtures)
mer-based materials, that the interest of both the aca- will be called stimulated or improved autogenous
demic research community and public funding healing.
authorities started being steered towards the topic. Autonomous healing includes the application of
Four years after this publication in Nature, the encapsulated minerals and polymeric healing agents,
Dutch government funded a research programme on as well as (encapsulated or granulated) micro-organ-
self-healing materials [6]. Then, other projects with isms that induce precipitation of CaCO3 as a result
the main focus on self-healing in cementitious of their metabolism and the nucleation properties
materials emerged, in which leading universities in and zeta-potential of the cell wall (‘bacteria-based
Europe were and still are involved, including SIM-pro- self-healing concrete’). For encapsulation, microen-
gramme SHE [7], HEALCON [8], Materials for Life capsulation, macroencapsulation, and vascular sys-
[9], CAPDESIGN [10], LORCENIS [11], SARCOS tems have been applied, and this will dictate the size
558 V. CAPPELLESSO ET AL.

of the damage that can be healed, the repeatability of effects that environmental conditions can have on
healing, and the recovery rate [17]. In these systems, the performance of each healing agent, care must be
the trigger for self-healing is usually the formation of taken when comparing results from different studies,
the crack itself, by which the capsules are ruptured since degrees of healing can show fluctuations accord-
and release their content. ing to environmental conditions [30]. Some research
Over the past years, several state-of-the-art papers studies considered in-situ applications in order to
have been published, focusing on the different autoge- show that the self-healing techniques developed at lab-
nous and autonomous technologies and their effec- oratory scale could be extended to real situations
tiveness [17], the test methods to assess them [18], [3,18,34–40]. However, further research should focus
the mix design of self-healing concrete [19] and the on the long-term durability of the healed structures
models to predict, also from a design perspective, [17,41], considering more realistic environmental
the outcomes of the phenomenon, including data-dri- conditions [22,30,31,42–44] including phenomena
ven models derived through Artificial Intelligence (AI) such as freeze and thaw cycles, salt crystallization
algorithms [17,18,20,21]. Literature surveys provide pressures, corrosion of reinforcement, long- and
ample confirmation of autonomous healing of cracks short-term actions including cyclic and dynamic
triggered by the crack formation or external stimuli. ones, which can call for repeated activation of the heal-
However, it has to be recognized that much of this evi- ing functionality at the same or different crack
dence was acquired in ideal conditions, often ambient locations. Moreover, assessing the durability of
laboratory environments and that, despite the contin- materials after damage and consequent healing is
ued efforts, there is still a lack of knowledge regarding essential to validate full-scale structural use [31], in a
the behaviour of self-healing cementitious materials in framework in which large-scale self-healing concrete
real structural service scenarios, including extreme applications are rare and generally show insufficient
conditions or non-steady and non-stress free cracks. [45,46] or even unproven efficiency [35,36,47].
This knowledge is of paramount importance for asses- Since the use of self-healing technologies in
sing the durability and the robustness of the perform- reinforced concrete structures results in longer main-
ance of this new technology [22]. tenance-free service lifespans, they have to be evalu-
The definition of an ideal condition can vary ated from a life-cycle perspective which encompasses
depending on the healing agent, as exemplified below. not only technical but also environmental, economic
and social aspects [42,48]. In this regard, Life Cycle
. Crystalline admixtures (CA) show optimal per- Assessment (LCA) can help to better highlight the
formance under continuous water immersion or, advantages of these cement-based materials with
though to a somewhat lesser extent, under wet/ advanced functionalities, evaluating all the possible
dry cycles [23–28]. inputs, outputs and potential environmental benefits
. Encapsulated polyurethane has the ability to poly- over the complete life cycle. Previous studies regarding
merize rapidly in the presence of humidity [29– sustainable concrete solutions have mainly addressed
31]. On the other hand, in unsaturated specimens, ‘cradle-to-grave’ analysis [49–51]. Only a few have
the release of a healing agent can be encouraged analysed self-healing cementitious composites, some-
with the aid of the capillary action by the crack. times using a ‘cradle-to-gate’ approach, taking into
Temperature changes lead to a variation in polymer account all the impacts up to the production stage
viscosity, thus affecting the release process and the [52–54]. This is due to the lack of data regarding the
effectiveness of crack filling [32]. behaviour of these materials when exposed to real ser-
. Bacteria need water for their metabolic activities vice scenarios and to the lack of environmental inven-
and to provoke calcium carbonate precipitation. tory information regarding the self-healing agents.
Aerobic bacteria additionally need oxygen, and Therefore, a wide cradle-to-grave or a cradle-to-cradle
spores may not germinate in the deeper part of LCA analysis is needed to convince the users
cracks where available oxygen is limited. Chemical (designers, contractors, owners) and penetrate the
and enzymatic reactions proceed faster at a higher market.
temperature, up to a certain limit [33]. This article represents one of the milestone works
of the ITN-MSCA project SMARTINCS funded by
From these examples, it is clear that, in order to the European Commission, with as main objective to
make those self-healing technologies reliable and review the state-of-the-art on the durability and sus-
robust for implementation in real structures, efforts tainability of self-healing cementitious materials.
to prove the functionality even in adverse conditions, This review is conducted with the following purposes:
are required [34]. Furthermore, it is important to (1) elucidate how cracking influences the transport of
know how to select the best healing agent as a function aggressive agents and hence the durability of struc-
of the intended structural service scenario and tures in the main aggressive scenarios; (2) review
environmental conditions. Due to the variety of how self-healing mechanisms can increase the
 INTERNATIONAL MATERIALS REVIEWS 559

durability for each aggressive scenario; (3) give a short penetration in the direction orthogonal to the crack
overview of methods to evaluate self-healing walls is also of importance, the latter being affected
efficiency; (4) suggest a pathway for incorporating by the evolution of the width of the crack along its
self-healing performance into a life-cycle spanning depth and by crack tortuosity.
durability-based design. In addition, during the service life of a structure,
cracks are affected by the loading and environmental
exposure scenario, which leads to evolving character-
‘Structural durability’: influence of the
istics, mainly in terms of crack opening. In fact, struc-
cracked state on degradation phenomena in
tural elements in real-life service conditions
reinforced concrete structures
experience sustained loads or cyclic actions, which
The crack formation has to be considered as normal result in through-crack stress states, which may
for any reinforced concrete structure under service cause the cracks to be ‘active’, meaning unsteady
loads (e.g. Eurocode 2 [55] and fib Model Code 2020 along time. This is of utmost importance also with
[56]). However, it has to be controlled and limited reference to the long-term performance of healing
to an extent which does not compromise the service- materials, since the majority of the studies available
ability and durability of the structure. As such, a so far have investigated the performance of healing
description of the crack ‘geometrical’ parameters in specimens that are not subjected to any kind of
which affect the kinetics of the ingress of aggressive load and hence under intrinsically static crack con-
substances and which can be entered into the model- ditions. To obtain a reliable prediction of the service
ling of material and structural durability is needed. life, the complex interaction between physical, chemi-
These parameters include crack density, orientation, cal and mechanical effects on concrete structures has
tortuosity, width, and effective depth, which, together to be considered. Moreover, different degradation
with their self-healing characteristics [57], may all mechanisms can influence each other, and combined
affect the durability of the structure. In international effects can be more harmful [58,59].
design codes, widely accepted methods are proposed Table 1 presents a scheme of the main aggressive
for the calculation of crack width and spacing in scenarios with related degradation processes, affected
ordinary reinforced concrete structures under static properties of the reinforced concrete and the related
loads or restrained deformations. Similar method- indicators. It has been compiled considering the
ologies scarcely exist or are still under development main exposure classes, as reported in EN 206 [60],
in the case of cracks due to environmental degradation which discern among: no risk of corrosion or attack,
or very early age cracks (e.g. due to plastic shrinkage) carbonation-induced corrosion, chloride-induced cor-
or when advanced cement-based materials, such as rosion, freeze–thaw attack, and aggressive chemical
fibre-reinforced cementitious composites, are con- environments. For each exposure class, prescriptions
cerned. Most of all, validation through assessment in are provided related to maximum w/c ratio, minimum
real structures for the abovementioned cases is strength class, minimum cement content, minimum
lacking. air content (for freeze/thaw attack) and other specific
The main parameters employed to characterize and requirements (e.g. related to types of cement to be
design the cracked state of a reinforced concrete used). Specific prescriptions for chemical attack from
element are the crack spacing and the crack width. natural soils and groundwater are based on limiting
Structural design codes, including fib Model Code values for indicators related to the aggressive medium,
2020, Eurocode 2 and ACI 318, propose formulations including pH, contents of carbon dioxide, sulphates,
to calculate the crack widths as a function of the ammonium, magnesium, and soil acidity.
reinforcement bar diameter and reinforcement area It can be observed that codes currently provide a
ratio, concrete cover, concrete tensile strength, steel- rather prescriptive approach to durability, referring
concrete bond strength and the stress in the reinforce- to the crack width as a durability parameter only in
ment induced by the relevant combination of actions. a few exposure conditions. This is because the effects
Code formulations allow to predict the maximum of cracks and crack width on the kinetics of the degra-
width of the crack, e.g. at the intrados for a beam dation mechanisms and phenomena are not available
element in flexure; however, it is important to con- in the literature for each exposure scenario. For
sider that the crack width in such elements varies example, the effects of cracks in structures exposed
along its depth. It is apparent that the inherent 3D fea- to aggressive chemical environments (excluding car-
tures of the crack are relevant concerning self-healing bonation, chlorides and marine environments) have
effectiveness (the volume of the crack has to be filled barely been studied. Degradation by sulphates and
and not only the part of the crack close to the crack chlorides, or combined attack as occurring in a marine
mouth). Moreover, besides the permeation of aggres- environment, has received more attention.
sive agents in the direction of the crack, which is For the whole set of environmental exposure
mainly affected by the maximum crack width, classes, the concrete strength, also related to the w/c
560 V. CAPPELLESSO ET AL.

Table 1. Degradation process and durability indicators for each aggressive scenario.
Specific durability Concrete
Degradation indicators (even in SH General durability quality
Main aggressive scenarios Type of attack process in RC Effect on RC concrete) indicators indicators
Chloride attack Chemical Salt precipitation ↑Porosity Chloride Penetration Water accessible Strength
attack Corrosion ↑Diffusivity Corrosion Potential porosity w/c ratio
Pitting ↑Potential Diffusion Cement
Cracking gradient coefficient content
↓Resistivity Gas/water Cover depth
↑Acidification permeability Chloride
↓pH Portlandite content
Carbonation Chemical Rebar ↕Porosity Carbonation rate content Porosity
attack depassivation ↕Permeability Corrosion Potential Capillary
Corrosion ↕Diffusivity absorption
Cracking ↑Potential Electrical
gradient resistivity
↓Resistivity
↑Acidification
↓pH
Sulphate attack Chemical/ Expansion ↑Porosity Volume changes
Physical Internal tension ↑Permeability Mass loss
attack Mechanical ↑Diffusivity Sulphate content
softening ↑Internal
Cracking damage
Mass loss ↑Potential
Flaking gradient
Scaling ↓Resistivity
↑Acidification
↓pH
Combined attack Chemical/ Expansion ↑Porosity Chloride Penetration
(marine environment) Physical Internal tension ↑Permeability Corrosion Potential
attack Mass loss ↑Diffusivity Volume changes
Cracking ↑Internal Mass loss
Corrosion damage
↑Potential
gradient
↓Resistivity
↑Acidification
↓pH
Acid attack Chemical/ Leaching of ↑Porosity Portlandite content
Physical hydrated ↑Permeability Mass loss
attack products ↑Diffusivity Volume change
Cracking ↑Internal
damage
↑Potential
gradient
↓Resistivity
↑Acidification
↓pH
Freeze–thaw attack with Chemical/ Scaling ↑Porosity Mass loss
de-icing salts Physical Internal tension ↑Permeability Chloride Penetration
attack Cracking ↑Diffusivity Corrosion Potential
Corrosion ↑Internal
damage
↑Potential
gradient
↓Resistivity
↑Acidification
↓pH
Freeze–thaw attack Physical attack Internal tension ↑Porosity Ultrasonic velocity
without de-icing salts Cracking ↑Permeability Acoustic emission
↑Internal Modulus of elasticity
damage
Water pressure Physical attack Leaching of ↑Porosity Water permeability
hydration ↑Permeability under pressure
products ↑Internal
damage
High/low temperature Physical attack Shrinkage ↑Permeability
and temperature Cracking ↑Internal
variation damage
↓Resistivity
Notes: SH: Self-healing, RC: reinforced concrete, the arrows mean: ↑ increase, ↓ decrease and ↕ increase or decrease.

ratio [61], which contributes to porosity, resistivity, un-cracked state (e.g. high strength concrete is likely
diffusivity, among other factors, is assumed as a to provide higher durability, as result of its lower por-
universal concrete performance indicator and hence osity). Furthermore, the initial calcium hydroxide
also as an indicator of durability, at least in its content is an indirect durability indicator in
 INTERNATIONAL MATERIALS REVIEWS 561

cementitious materials. It is related to the availability for the determination of the time to chloride-induced
of leachable material and the buffering capacity corrosion: (1) the time for the transport of chloride
against carbonation. In addition, the cement content, ions penetrating from the environment through the
concrete cover, and chloride content are widely used concrete cover to reach the rebar level, and (2) the
indicators [62]. Diffusivity, permeability and porosity time for accumulation of chloride at the rebar level
are extremely important to indicate the durability of up to a critical content (chloride threshold) which
concrete structures as they are correlated to transport allows active corrosion to start. From then onwards,
and diffusion phenomena. pitting corrosion propagates. This model concept
In general, water and gas can carry harmful sub- has been quantified and incorporated in the standards
stances that cause concrete deterioration and/or for the service life calculation of concrete structures,
rebar corrosion. Those fluids have three main trans- such as fib Bulletin 34 [66], fib Bulletin 53 [67], Euro-
port mechanisms: diffusion, permeation and sorption, code 2 [55], EHE-08 [68], ENV 206 [60], Spanish
but the main ones are diffusion and permeation. Structural Code [69]. Standards refer to a chloride
Among these mechanisms, permeation is the one threshold expressed as % of total chloride by weight
that is usually associated with the durability perform- of the binder, assuming that bound chlorides could
ance of reinforced concrete [63], defining the ease with be released, as can be the case upon carbonation. In
which a fluid flows through a porous material under a reality, the threshold value of chloride for corrosion
pressure gradient [63]. Permeability is influenced by initiation has been found to show a large scatter
the size, distribution and continuity of the pores [70,71]. It depends on several factors characterizing
[64], and it is accelerated by the presence of cracks the steel-concrete interface, as those mentioned in
which are considered to be a pathway to aggressive Angst et al. [72,73]. Construction codes generally con-
agents. Nevertheless, when harmful substances are sider the limit of 0.4% total chloride by cement weight.
present as salts or CO2, the diffusion mechanism In contrast, in the recent Spanish Structural Code [69],
plays the primary role in view of durability. The salt different values are included as chloride thresholds
concentration at the surface and the diffusion coeffi- depending on environmental exposure. Moreover,
cient of the concrete define the rate of ingress. The the rate of chloride ingress is affected by the quality
diffusion coefficient is highly dependent on the pore of the concrete cover, chloride concentration and
structure, mainly affected by the w/c ratio, cement exposure conditions [74–76].
and pozzolan content, curing conditions and presence The presence of cracks in the concrete cover pro-
of cracks. In this respect, the benefits of self-healing vides easier access for the chloride ions to the deeper
concrete are related to their ability to restore the values parts of the concrete [77,78] (Figure 2), causing cor-
of the aforementioned indicators to those typical for rosion to initiate in a shorter period [77]. The chloride
un-cracked concrete [42]. ingress will increase with the increase of micro-crack
Investigations regarding self-healing in many of the density independently of the concrete characteristics
aforesaid scenarios can hardly be found in the litera- [79]. Moreover, the chloride attack could be aggra-
ture. Because of the lack of information, the current vated by salt crystallization pressures in the crack
review will not discuss acid attack. Nevertheless, it is [80]. In chloride-rich environments, the crack width
included in Table 1 as it should be considered as one limit should be more restrictive than for environments
of the main aggressive scenarios. The following sec- free of chlorides since chloride ions have a small size
tions provide a comprehensive review regarding the and high diffusion capacity [81,82]. The crack geome-
current knowledge on the effects of cracks on the try, including width, depth and tortuosity, affects the
degradation mechanisms and the behaviour of self- transport of chloride ions [83–85]. The wider and dee-
healing concrete in those scenarios. per the crack, the earlier the reinforcement corrosion
will occur [83]. Maes [86] stated that 10 µm is the criti-
cal crack width, above which chlorides penetrate more
Chloride attack
rapidly through the crack than in the un-cracked zone,
Marine structures, bridge decks, road slabs and park- whereas Van den Heede et al. [87] suggested that for a
ing structures are the most common concrete con- crack width lower than 50 μm, chloride penetration
structions being susceptible to corrosion induced by through the crack would be negligible. Russo et al.
chlorides. Chloride ions originate from various [85] found that chloride penetration increased at
environmental sources to which the structure is least 1.5 times for concrete with different character-
exposed during its service life, including seawater, istics when increasing the crack width from 5 to
marine airborne salts and de-icing salts. A thermodyn- 70 µm and crack depth to 40 mm. Djerbi et al. [88]
amic service life model was proposed by Tuutti in the found that the chloride diffusion coefficient for crack
80´s [65] for reinforcement corrosion that highlights widths less than 80 μm was higher for conventional
two phases: initiation and propagation. In the concrete than for high-performance concrete (HPC)
initiation phase, two sub-periods have been derived due to its higher porosity. However, for crack widths
562 V. CAPPELLESSO ET AL.

Figure 2. Cracks in concrete provide easy access for chloride ions. The line on the left indicates the crack location. Chloride ingress
can be visualized by spraying with 0.1 M AgNO3 resulting in white AgCl precipitation – the line over the sample denotes the dis-
colouration front (A). The chloride element map shows the concentration gradients by electron probe micro analysis (EPMA) (B).
[78] Copyright 2019, Elsevier.

larger than 80 μm, the diffusion coefficient did not concentration, age and curing conditions on the car-
depend on the crack width or material characteristics. bonation mechanism, kinetics and microstructural
Kušter Marić et al. [89] measured the chloride content alterations in cementitious systems, including those
at different locations of a concrete bridge in service. containing supplementary cementitious materials
They presented 100 µm as a crack width limit beyond (SCM). Moreover, the effect of carbonation on the
which depassivation time was significantly reduced, transport properties and porosity of concrete with
independent of the cover thickness. SCMs was studied [94].
Francois and Maso [77] noticed a higher chloride Cracks can favour the CO2 penetration through the
diffusion in the tensile zone than in the compression concrete cover, thus triggering an earlier start of the
zone of beams after flexural loading due to the damage reinforcement corrosion. Alahmad et al. [95] studied
at the grain/paste interface. This was confirmed by for mortars the effect of cracks with openings ranging
Gowripalan et al. [90], who related the difference to from 9 to 400 μm on the ability of CO2 to diffuse along
cracks in the tension zone and a decrease of porosity the crack. After accelerated carbonation (50% CO2)
in the compression zone [90]. There is an evident for 65 days, at 23°C and RH 65%, they found that
lack of information regarding the effect of cracks on crack opening greatly impacted the carbonation. For
the chloride threshold for reinforcement corrosion crack widths higher than 60 μm, the carbonation
onset. Pettersson et al. [91] found that for a high depth perpendicular to the crack was similar to carbo-
crack width/cover ratio, the chloride threshold nation at an outer surface, indicating that CO2 pene-
decreased in the seawater splash zone for high- trates freely in the crack. However, for cracks below
strength concrete containing silica fume. 41 μm, the carbonation perpendicular to the crack
progressively decreased. For crack widths of 9 μm, car-
bon dioxide diffusion through the crack was stopped,
Carbonation
which was attributed to self-healing by carbonation
The alkaline nature of the concrete facilitates its inter- products. Van Mullem et al. [96] showed that carbona-
action with environmental CO2 decreasing the pH of tion along realistic cracks in mortars was significantly
the cover. When carbon dioxide from the air comes affected by the crack tortuosity once crack widths went
in contact with the concrete, first, it has to be dissolved below 100 µm. Han et al. [97] claimed similar results
in the pore water to form HCO–3 ions. The dissolved in concrete after accelerated carbonation at 20%
CO2 reacts with portlandite (CH) and C-S-H gel to CO2, at 20 ± 2°C and RH 70 ± 5%. For crack widths
produce calcium carbonate and decrease the pH of bigger than 100 μm, the carbonation depth perpen-
the pore solution. That results in depassivation of dicular to the crack walls was similar as for the con-
the reinforcement and enables rebar corrosion if crete surface (Figure 3). Bogas et al. [98] found for
enough moisture is available in the environment. structural lightweight aggregate concrete (LWAC)
The rate at which this process occurs depends mainly that the crack opening (100–300 μm) was less relevant
on concrete and environmental characteristics [92,93]. in comparison with regular weight concrete since por-
A recently published literature review by RILEM TC ous lightweight aggregates anyhow led to an increase
281-CCC presents the effect of concrete character- in the carbonation rate. Carević and Ignjatović [99]
istics, relative humidity, temperature, CO2 claimed that the presence of small cracks (50 µm) in
 INTERNATIONAL MATERIALS REVIEWS 563

Figure 3. Relative carbonation content after 10 years in cracks with different crack widths using the virtual COMSOL Multiphysics
model proposed by Saetta [97,100]. Copyright 2016, Elsevier.

reinforced concrete leads to a twice higher carbona- can be diffusion or permeation when water contami-
tion depth after exposing to a CO2 concentration of nated with sulphates penetrates the concrete. Nehdi
2% for 28 days. The average carbonation depth et al. [104] commented that studying only chemical
increased when the crack width increased. attacks, as is generally done in experimental research,
There is an evident lack of knowledge regarding does not represent the situation in most concrete
corrosion initiation risk due to the crack presence structures with some area exposed to the air.
and the carbonation effect. Marcos-Meson et al. [82] Sulphates are mostly found as sodium sulphate
exposed steel fibre-reinforced concrete (SFRC) speci- (Na2SO4) or magnesium sulphate (MgSO4) but can
mens with 50 to 300 µm cracks to wet–dry cycles also have calcium (Ca) or iron (Fe) as the cation, lead-
involving both chloride and carbon dioxide for one ing to a different attack mechanism and kinetics [110].
and two years. Steel fibres directly exposed at the For Na2SO4, secondary gypsum and ettringite precipi-
crack were corroded since the pH at the crack walls tate with a volume expansion which can lead to an
decreased to about 9 due to carbonation. However, intense cracking process because of internal stresses
the impact on the corrosion of reinforcement was lim- [111], leading to increased permeation of the sulphate
ited due to the precipitation of calcite and hydration solution. While tricalcium aluminate (C3A) and cal-
products partially sealing the cracks, thus slowing cium hydroxide (CH) are attacked, the calcium silicate
down the further carbonation process. hydrate (C-S-H) is kept intact. However, considering
MgSO4 attack, thaumasite and brucite can be formed,
causing more impact on the mechanical resistance
Sulphate attack
[110]. At first, the brucite layer slows down the sul-
Each salt induces a different level of deterioration phate attack, which slowly continues by diffusion
related to the solubility of its reaction products, the across the surface layer [111]. As the attack progresses,
reaction kinetics and the nature of the crystallization expansion results in cracking of the brucite layer,
pressure of solid products, with sulphate being the further decalcification of the C-S-H and ultimately
most harmful to concrete [101]. Examples of struc- conversion to the non-cementitious magnesium sili-
tures which can be subjected to sulphate attack include cate hydrate (M-S-H) [111].
basement walls, piers, abutments, and tunnels, which Idiart et al. [112] progressed in a model prediction
are in contact with ground water, soil, seawater, indus- considering the presence of cracks and further degra-
trial effluents, sewage water or waste water [102,103]. dation by sulphate attack. They included diffusion
Depending on the sulphate exposure condition, the through the cracked concrete based on the study by
attack is categorized as a chemical or physical Djerbi et al. [88], defining a linear relation between
phenomenon [104,105]. Total immersion ends in the crack width and the diffusion coefficient in cracks
chemical attack, while physical attack prevails when up to 100 µm. Cracks larger than 100 µm had diffusion
surfaces are (temporarily) exposed to drying con- similar to diffusivity in free solution. Liu et al. [113]
ditions [106]. The physical salt attack is connected analysed fatigue loading, showing that the created
with continuous surface scaling and flaking of con- cracks accelerate the process of deterioration by
crete due to salt crystallization in the pores adjacent sulphates.
to the drying surface [107]. Overall, the major factors influencing the sulphate
On the other hand, a chemical salt attack results in attack are the available amount of sulphate ions, per-
a direct reaction with the hydrated cement paste that meability of concrete, quantity of C3A, type of binder,
leads to mechanical softening, expansion, cracking, pH, temperature and type of cation [86,105,114]. The
spalling and disintegration [108,109]. Overall, these first stage of the attack involves the formation of
processes are related to the entry mechanism that expansive products leading to the second stage
564 V. CAPPELLESSO ET AL.

where cracks appear and where crack formation exposure time. In the first stage, chloride diffusion
advances proportional to the rate of expansion was barely affected by the presence of sulphates. In
reactions until complete deterioration [114,115], the second stage, the chloride diffusion decreased
which can be worse if combined with a physical attack with the sulphate content since voluminous sul-
[104]. The presence of cracks accelerates the sulphate phate-related reaction products obstructed the pore
ingress and is the main transfer path in the later stage network. In the final stage, chloride diffusion in the
[105]. presence of sulphates increased since the continued
growth of the expansive products leads to crack for-
mation. Zhao et al. [121] confirmed that a combined
Combined attack (including marine
sulphate-chloride attack aggravates the damage
environment)
because an increasingly severe cracking process
Academic research tends to focus on isolated effects of opens paths for chloride ion penetration. Further-
a single aggressive agent, and only a few studies have more, chloride binding decreases when sulphate con-
evaluated combined attack actions. Still, the topic tent increases, even worse for higher temperatures
has been getting more and more attention over the when the diffusivity increases [122,123].
last years. Most real situations are likely to be charac- Sun et al. [124] examined the steel corrosion and
terized by combined attack of different aggressive cracking process in cement paste exposed to combined
species involving different interacting degradation sulphate-chloride attack by microtomography (µCT)
mechanisms. For example, Meng et al. [116] con- analysis. The cracks formed due to the sulphate attack
sidered the combined effect of sulphate attack and occurred before the cracks formed by corrosion;
freeze–thaw actions. They identified that sodium sul- moreover, the chloride ions mitigated the sulphate
phate accelerates concrete damage under freezing attack delaying the development of cracks. Wu et al.
and thawing compared to tap water. Li et al. [117] [125] found that the presence of chlorides can mitigate
also reported that concrete deterioration is aggravated the effect of magnesium sulphate attack and delay the
when freezing-thawing is combined with sulphate deterioration in the early stage, but it can promote the
attack. Gong et al. [118] investigated the expansion migration of sulphates inside the concrete in the later
and mechanical deterioration of concrete damaged stage. Once the formation of expansive products has
by coupled alkali-silica reactions (ASR) and freeze– resulted in cracks, these trigger more severe chlor-
thaw cycles (FT). They verified that the silica gel ide-induced corrosion of steel reinforcement [89].
formed by ASR could partially restore the damages Kušter Marić et al. [89] verified that cracks larger
due to the FT showing a positive effect on the mechan- than 100 µm decrease the time for depassivation sig-
ical properties. Liu et al. [119] conducted a numerical nificantly, despite the thickness of the cover zone.
analysis for combined load-induced cracking and Whereas in un-cracked concrete with a sufficient
chloride transport. They identified that longer cracks cover, a service life of 100 years is considered by the
facilitate the chloride ingress with higher concen- European standards, in this study, the cracked con-
trations at the crack tip. crete in the marine environment had its service life
Among all combined attack scenarios, the one with reduced to one year (chloride threshold exceeded at
the greatest coverage is the marine environment, the reinforcement). A large-scale experimental pro-
where the main salts, sodium chloride and magnesium gramme on concrete with blended cement during 18
sulphate, act combined. The deterioration of concrete years considering submerged, tidal and atmospheric
in the marine environment is related to the saturation zones [126] showed that the expansive reaction pro-
degree (submerged, tidal or atmospheric zone), affect- ducts (brucite, ettringite, calcite) could fill cracks and
ing the attack mechanism (chemical/physical) and the pores, preventing further chloride ingress at first
transport of the aggressive agents. On the other hand, instance. However, as the attack persisted, the decal-
whereas the seawater changes considerably in salinity cification of C-S-H increased the total porosity result-
in different areas, the type of ions and their pro- ing in further degradation.
portions are similar [120]. Montes et al. [127] indicate that corrosion caused
Maes [86] considered combined attack by simu- by marine exposure tests was aggravated by increasing
lated seawater and investigated the effect of chlorides the crack width from 250 to 500 µm. Otieno et al.
on sulphate attack as well as the effect of sulphates [128] noticed by a long-term experimental pro-
on chloride attack. In the case of magnesium sulphate gramme with cracked concrete subjected to a tidal
attack, a large dependence on the environmental zone that the cover thickness can prevent corrosion
temperature and binder type was observed, while in narrow cracks. Nonetheless, for wider crack widths
sodium sulphate attack mainly was related to binding of 400 and 700 µm, the cover thickness did not have a
with the C3A hydration products. When considering considerable effect. Lopez-Calvo et al. [129] revealed a
chloride ingress and the risk of reinforcement cor- clear increase of corrosion potential with crack width
rosion, the influence of sulphates was affected by the in a simulated marine tidal condition. However, with a
 INTERNATIONAL MATERIALS REVIEWS 565

higher cover thickness, the limits on crack widths spacing factor between air voids depends on the freez-
could be increased: for a cover of 25 mm, corrosion ing condition, i.e. the rate of freezing, the minimum
was noticed for cracks wider than 90 µm, whereas temperature and the duration at the minimum temp-
for a cover of 45 mm, this was only the case for cracks erature. The second factor mentioned by Wang et al.
above 180 µm. Du et al. [130] detected by mimicking a [137] is the reduction in water absorption (2), which
marine exposure that 700 days were needed to cause means a lower permeability [133,141], and hence a
corrosion in cracks with a width of 100 µm; on the reduced scaling of the surface layer [142]. Though
other hand, while this was reduced to 460 days for scaling is superficial and does not compromise mech-
cracks of 300 µm. anical integrity, the weakened surface layer becomes
Studies on advanced cement-based materials are susceptible to further ingress of aggressive agents
more scarce. Lv et al. [131] verified the chloride ion affecting the durability [143,144]. Crack control (3)
transport in microcracked ultra-high performance is the last factor that can contribute to the FT resist-
concrete (UHPC) with steel fibres in a marine ance and which might be improved by using fibres
environment. They observed that even cracks around and nanoparticles. Zeng et al. [145] proved that steel
50 µm wide formed in the protective layer over the fibres increased the FT resistance by bridging cracks
steel fibres become rich in chloride ions. Taking that and controlling their width, hence decreasing the per-
into account, the corrosion products on the steel meability in the cracked state. Wu et al. [146] applied
fibres can produce further stress internally, resulting cellulose/PVA hydrogels to reduce the free water con-
in tiny cracks in the surrounding matrix. tent, avoid crack propagation and hence decrease the
effects of the FT.
The presence of de-icing salts on the surface of con-
Freeze–thaw attack with or without de-icing
crete structures at freezing temperatures causes severe
salts
scaling damage, intensified by the further salt solution
In freezing conditions, concrete will crack when the penetration into the pores [63,93,143,147,148]. Salts
microstructure cannot support the stress caused by change the damage development due to crystallization
water expansion, and stress release upon cracking and melting temperature variation [149]. Ma et al.
will open additional space for water which will expand [150] illustrated three types of de-icing salts, ethylene
again during freezing [132]. There are two types of glycol, NaCl and calcium magnesium acetate how
deterioration due to freezing and thawing (FT): spalling started when reaching the critical superficial
internal cracking, which is the main problem in the crack density of 85–88%. Figure 4 presents how the
absence of chlorides and surface scaling, which is cracking behaviour progresses in concrete without
related to the presence of de-icing salts [133]. The initial cracks and with a preliminary cracking process
internal damage is controlled by the degree of pore [151]. The presence of cracks accelerates the water
saturation, which depends on the bulk moisture absorption, increasing the concrete saturation. Then
uptake [134,135]. Shields et al. [136] varied degrees the freezing and expansion process induces tensile
of saturation under FT cycles and detected higher stress. After several FT cycles, the width and quantity
crack volumes with increased saturation levels. Besides of the cracks increase until the worst-case scenario of
the saturation, Peng et al. [132] identified that the scaling is reached. Li et al. [152] made a numerical
temperature distribution greatly influences FT’s study analysing chloride penetration in cracks induced
effect. Overall, three factors play an essential role, by freezing and thawing. The authors concluded that
namely material permeability, which is related to the the cracks affect the chloride ingress, which is further
composition and the presence of cracks; the spacing increased with the number of FT cycles as a result of
factor between the capillary pores and air voids; and the crack propagation. Cracks below 50 µm had a
the rate of freezing [133]. Recently, Wang et al. small effect, while cracks between 50 and 100 µm
[137] reviewed the three main essential factors to increased the chloride penetration significantly.
achieve a better performance in FT conditions: (1)
improving the internal pore system, (2) reducing the
Water pressure
water absorption of the concrete, and (3) controlling
the cracking process. Dams, water tanks, reservoirs, swimming pools, and
The first, namely the internal pore system (1), is structures in deep water such as port and harbour
enhanced mainly by the use of an air-entraining installations, undersea tunnels, offshore oil platforms,
agent but also by the use of supplementary cementi- and caissons are examples of structures that should
tious materials and nanoparticles [137], beyond the sustain hydrostatic pressure loads. The hydrostatic
reduction of the water/cement ratio [138,139]. A pressure causing water pressure inside cracks might
proper air void distribution is of significant impor- reduce the ultimate resistance of the structure signifi-
tance to alleviate the expansion by the freezing water cantly [153]. Huge hydraulic loads can accelerate the
[140]. Şahin et al. [139] claimed that the adequate propagation of cracks in concrete structures and
566 V. CAPPELLESSO ET AL.

Figure 4. The degradation caused by the internal tension leads to cracking propagation in concrete without damage (top) com-
pared to concrete having prior cracks (bottom) under freezing and thawing cycles [151]. Copyright 2020, Elsevier.

increase the damage caused by hydraulic fracture of 0.3–0.9 MPa on the other side of the wall. Brzo-
[154]. With this concern, concrete has to be adopted zowsk and Horszczaruk [157] clarified the influence
with a higher capacity to control crack propagation of hydrostatic pressure on the adhesion of repair
and permeability, even under sustained loadings. It materials.
should be emphasized that for immersed structures Water pressure has an important role in crack propa-
(deep or shallow), the durability (in normal con- gation. Wang et al. [58] developed a numerical model
ditions) is mostly not a significant issue, predomi- considering crack propagation in dams under static
nately because of the low O2 levels. Hence corrosion and seismic loads, showing that the water pressure
takes a very long time to initiate and a very long within the crack contributes to an increasing crack
time to propagate. What is of major importance length. Cui et al. [158,159] found that the concrete com-
though is the stability of the cracks and the structural pressive strength and bulk modulus were reduced with
integrity, especially in deep water installations. High- an increase in hydrostatic pressure between 30 and
performance fibre-reinforced cementitious compo- 500 MPa, due to the increased volume of microcracks
sites (HPFRCCs) can be the ideal candidates in these in the aggregate-mortar interfacial transition zone
scenarios, mainly because of their ability to spread (ITZ). Yi et al. [160] verified for cracked concrete up
otherwise localized damage into narrow and tightly to 2 MPa water pressure that the water transport
spaced multiple cracks [155]. increased with crack width and hydraulic pressure.
Moreover, water under pressure can foster the pen- Cracks of 30 and 50 µm had similar flow rate, while a
etration of aggressive agents, and permeability is the crack width of 100 µm increased the flow rate signifi-
most important indicator of durability. For example, cantly. Indeed, Edvardsen [161] has shown that the
in structures in deep seawater, the diffusion mechan- flow rate scales with the third power of the crack width.
ism of chlorides becomes less important than pen-
etration under water pressure. In underground
High/low environmental temperatures and
structures, including foundations, the water pressure
temperature variations
is affected by the surrounding soil and the high
water level that may intensify sulphate attack. Xiao The temperature of the environment can vary signifi-
et al. [156] considered a mimicked seawall where cantly, leading to the expansion and contraction of
both sides of the structure were affected by water structural elements, which can affect their durability.
pressure. In this case, the chloride ingress at the sea- Cold weather during winter times can result in nega-
water side was prevented because of the water pressure tive temperatures for long periods. However, freezing
 INTERNATIONAL MATERIALS REVIEWS 567

conditions were already treated in section ‘Freeze– crystalline admixture (CA) or supplementary cementi-
thaw attack with or without de-icing salts’ and will tious materials (SCM), it is well known that there is a
not be discussed here. Also, fire damage related to need for water. Either high humidity or preferably
spalling caused by the development of internal steam wet–dry cycles or immersion in water, are needed
pressure, dehydration and phase-changes of hydration for developing the best performance [17,23–
products will not be treated. Still, structures subject to 26,166,167]. Generally, autogenous healing will be
solar radiation can reach high temperatures on the limited to relatively small crack openings [17,18,42].
exposed surface (around 60°C) on hot days. In sand- There is evidence of crack healing between 10 and
wich panels, the difference between the internal and 100 μm, sometimes up to 200 μm, only in the presence
external temperature can reach more than 20°C of water [17].
[162] as a function of the thermal diffusivity and con- If in real situations, the structure is not in contact
ductivity of concrete. These depend essentially on the with rain or other sources of liquid water, these mech-
water content in the concrete, the type of aggregate, anisms may become less effective or even useless [22].
and the proportion of aggregates in the mixture Some agents, such as superabsorbent polymers (SAP),
[163], whereas the effects of the presence of cracks can contribute to the capture of moisture from the air
can be considered as negligible [164]. and hence serve as an activator for autogenous healing
The volume variations resulting from temperature [168]. The crack sealing provided by the immediate
cycles and the temperature gradients inside a struc- swelling of SAP particles in contact with moisture,
tural element may generate tensile stresses in the and the activation of autogenous healing by releasing
matrix resulting in cracks, because of the likely pres- this moisture to unhydrated binder particles near the
ence of external and internal restraints, including crack wall, can contribute to the self-healing mechan-
reinforcement. Structures must hence be designed ism. The effect of the humidity level is reflected in the
with details allowing those movements, such as expan- effectiveness of crack closure. As shown by Snoeck
sion joints, and employing concrete that can effec- et al. [169], in environments with a relative humidity
tively control the consequences of such higher than 60%, only samples with SAPs showed
deformations, including, for instance, fibre reinforce- visual closure of the cracks, indicating that SAP par-
ment to control the crack opening [165]. ticles have hygroscopic capacity and can attract moist-
ure from the environment. Samples containing SAPs
and exposed to wet/dry cycles could close cracks com-
Self-healing under aggressive
pletely up to about 140 μm, while cracks larger than
environmental conditions
200 μm showed only partial healing [169]. To guaran-
Most of the laboratory studies dealing with self-heal- tee healing efficiency, it is necessary to keep the crack
ing of concrete do consider ‘ideal’ conditions regard- opening below the threshold of the healing agent.
ing the healing regime. And while the definition of Crack width reduction through fibre reinforcement
the ‘ideal’ condition is obviously dependent on the will enable SAPs to seal off the matrix from pen-
healing technology considered, as will be explained etration of aggressive agents even in environments
further, all ‘ideal’ conditions have a common denomi- with less available water [17,169,170].
nator: (1) cracks are stress-free upon healing and (2) The behaviour and triggers required for healing of
young specimens are subjected to a single cracking/ concrete with CA are similar to those required for
healing event. This does not match with reality, so it autogenous healing. Thus, it has been highlighted
will be extremely important to extend the work from that the use of CA as a healing agent is limited to struc-
‘ideal’ conditions to ‘real’ conditions and prove the tures that have water availability [26,171,172]. Struc-
performance of self-healing concrete in real situations tures such as water tanks and reservoirs, submerged
in order to allow self-healing to penetrate the market structures or structures with abundant water avail-
further. Starting from healing under ‘ideal’ conditions, ability can be healed with CA. Cementitious matrices
in the next sections, we will gradually move away from with CA have a higher pH and release more Ca2+,
it, considering the stress state through the crack and favouring the precipitation of calcium carbonate and
the stability of the crack over time as well as the stab- thus the crack closure [23]. Borg et al. [26] reported
ility of the healing functionality over time. In the end, that under ideal conditions of submersion, cracks
healing under different aggressive conditions is between 100–150 μm closed in 56 days, cracks of
addressed. 150–300 μm closed in 84 days, while complete closure
was not achievable for cracks above 300 μm. Without
water contact, CA could not fully close the cracks
Ideal conditions and water availability
under 300 μm in 42 days, even with high relative
Ideal conditions vary according to the self-healing humidity between 95 and 100% [25,172]. Though,
mechanism considered. For autogenous self-healing some CA demonstrated the capacity to catalyse the
or improved autogenous healing by including healing reactions even by exploiting the atmospheric
568 V. CAPPELLESSO ET AL.

humidity [24]: a regular concrete, upon exposure to Absorption of bacterial spores into expanded clay par-
humid climates (RH > 65%), showed a healing ticles guaranteed the microbiologically induced calcite
capacity equal to the autogenous healing exhibited precipitation (MICP), but resulted in a reduction of
by the reference without CA addition when perma- mechanical resistance [182,183]. A new granulation
nently submerged in water. technology that helps to protect nitrate-reducing bac-
The use of fibres in engineered cementitious com- teria called ACDC (active compact denitrifying core)
posites (ECC) [173–176] as well as in ultra-high-per- revealed that these granules are compatible with con-
formance fibre-reinforced concrete (UHPFRC) [177– crete and allow the bacteria to survive in mortar.
179] or other cementitious matrices brings great When added as 1% w/w cement, after 28 days of
benefits for self-healing. First of all, these materials water immersion, they induced complete crack closure
benefit from their high autogenous healing capacity up to 500 μm crack width and a 74% reduction of
due to their composition (high binder content, high water absorption. The crack filling was noticed in
content of supplementary cementitious materials fea- the internal volume of the crack and not just near
turing pozzolanic or late cementitious action, and the crack mouth [41,48]. Mortars with 0.5% ACDC
low w/b ratio). Moreover, the fibres contribute to w/w cement showed healing of cracks up to 400 μm
redistributing the stresses after cracking and are for 95%, which led to an 80% decrease in water per-
instrumental in generating cracks with a smaller meability [184]. A similar method of granules contain-
width, so the damage, otherwise localized into a single ing ureolytic bacteria called CERUP (Cyclic EnRiched
wider crack, is spread into a larger number of more Ureolytic Powder) or MUC (Mixed Ureolytic Culture)
tightly spaced and narrow cracks. These cracks have achieved closure of cracks with a width of 450 µm after
a greater potential for closing, which promotes recov- 28 days immersed healing regime [185]. Among the
ery of both durability and mechanical properties, since published works, the largest local crack closure from
the ratio of available healing agent to crack volume for bacterial activity in ideal conditions was 970 μm,
smaller cracks is greater than in wide cracks. After ten about four times that of the non-bacteria control series
wet–dry cycles (submersion in water at 20°C for 24 h [186]. Generally, cracks around 350–500 µm can be
and drying in laboratory air at 21 ± 1°C, 50 ± 5% RH closed by means of bacterial activity under ideal con-
for 24 h), Li and Yang [173] reported that cracks of ditions, as presented by a lot of researchers
up to 50 µm could be fully healed autogenously with [41,48,183,185,187,188].
complete restoration of mechanical capacity and pre- Other autonomous mechanisms, such as micro-
vention of the transport of fluids. Guan et al. [176] and macro-capsules containing polymeric healing
analysed an ECC with limestone powder, noting agents, may be independent of water availability; how-
increased self-healing because of crack control by the ever, one-component polyurethane precursors need
fibres and contribution from the limestone powder humidity to polymerize [36,162]. Any damage capable
by forming nucleation sites accelerating the hydration of disrupting the encapsulation system triggers the
product formation, and increasing calcium carbonate healing. It makes the method more reliable in terms
precipitation. Hung and Su [174] showed that healing of closure due to the occurrence of the cracking
allowed ECCs to regain strength and stiffness effec- event, though it depends on a series of other variables
tively. Ferrara et al. [177,178] highlighted that for such as the probability of cracks intercepting the cap-
High-Performance Fibre-Reinforced Cementitious sules, and hence their distribution, the effectiveness of
Composites (HPFRCCs), the presence of water from the healing agent in coming out of the capsule, type of
high air humidity or wet and dry cycles enabled the shell and thickness, etc.
autogenous healing, even after periods as long as two
years.
Healing of cracks considering the crack stress
A bacterial self-healing concrete has requirements
state and stability of the crack over time
which depend on the characteristics of the microor-
ganisms that are used. The aerobic bacteria proposed In the broadest sense, the durability of a (construc-
by Jonkers et al. [180], for example, need in addition tion) material has to be regarded as its ability to retain
to water also the presence of O2, which can be limited its intended performance over time under the antici-
deeper down in the cracks. Alternative methods can be pated structural service scenario. This does include
used in case of a lack of O2, such as using nitrate-redu- not only the (aggressive) environmental conditions
cing bacteria together with nitrate as an alternative as discussed above but also a dedicated and careful
electron acceptor that enables bacterial CaCO3 pre- consideration of the stress state in the structural
cipitation without O2 being available [41,48,181]. Pro- element. Hence, the presence of ‘through crack’ stress
tective means for these organisms are necessary to states has to be considered as well as their effects on
guarantee survival within the hostile environment of the reliability and repeatability of the healing capacity.
a cement matrix. Several technologies for the protec- In most reinforced concrete structural elements,
tion of bacterial cells or spores have been used. tensile through crack stress states hold after the
 INTERNATIONAL MATERIALS REVIEWS 569

formation of a crack because of the cohesive nature of

concrete fracture and the interaction with the
reinforcement. These tensile through crack stress
states can be either sustained throughout the intended
lifespan or cycle between a minimum and a maximum
value, with the amplitude of the variation and its fre-
quency likely affecting the reliability and repeatability
of the healing performance. On the other hand, and
specifically in precast structural elements which
during their service life go through different transient
situations, tensile stress states in previously cracked
regions may turn into compressive ones, with the
same possibility of being sustained or cycling within Figure 5. Effect of loading and reloading for self-healing con-
a certain range as mentioned above. crete with PU (top) or WRA (bottom) [162].
Through-crack compressive stresses positively
affect healing since the stress partially closes the after healing also depends on the elasticity of the heal-
crack and helps the crack-surface binding by the heal- ing agent. If flexible enough, the healing agent can
ing products. This condition might arise in precast comply with the crack movement and the crack can
columns or tunnel segments, which could crack be kept sealed, so repeated healing is unnecessary; if
during transport while experiencing axial or circum- not flexible enough, repeated healing should be guar-
ferential compressive stress states in their final in ser- anteed [18]. Previous research has shown the effect of
vice configuration. Ferrara et al. [189,190] investigated reloading by temperature variation after healing [162].
the effects of through-crack compressive stress on the For cracks treated with water-repellent agents (WRA),
healing capacity of notched beam specimens of differ- the absorption coefficient remained low even when the
ent fibre-reinforced concrete mixes, either autogenous cracks grew upon reloading because the crack faces
or stimulated by crystalline admixture and latex poly- were still impregnated with WRA. At the same time,
mer, confirming the statements described above. On polyurethanes (PU) lost their bond with the crack
the other hand, through-crack tensile stresses have surface, resulting in increased water absorption
an adverse effect on healing. Özbay et al. [191] inves- (Figure 5).
tigated the effects of sustained flexural loading in In the case of healing agents providing the healing
Engineered Cementitious Composites (ECCs), which functionalities through expansive reactions, it has to
is likely to widen the pre-induced cracks, on the effec- be guaranteed that the volume variations induced by
tiveness of the healing process. A slight decrease in these reactions do not exceed the tensile deformation
strength recovery was found in specimens experien- capacity of the concrete [42], otherwise, more cracks
cing sustained flexural load compared to unloaded will likely appear. As an illustration, Sisomphon
specimens [192]. The healing mechanism can become et al. [171] showed the excessive ettringite formation
ineffective if the crack is active, so further experiencing in concrete with expansive self-healing agents causing
cyclic through-crack stress states. micro-cracking due to expansion.
For example, continuous opening and closing of Repeatability may be a concern in case of autoge-
cracks can occur in concrete bridge elements under nous healing since the necessary healing materials
traffic loads or restrained elements, even simply are consumed over time by normal hydration of the
experiencing thermal (day/night) cycles. The concern matrix or by crack filling. Stimulated healing methods,
regarding the performance of self-healing materials in such as the use of crystalline admixtures, have been
these conditions is due to the constant movement that found to be effective in facilitating long-term repeated
can damage the healing products or detach them from healing for up to one year with re-loading after 1, 3
the crack walls [31]. Depending on the self-healing and 6 months, followed by healing regimes of 1 or 2
mechanism, it may also be challenging for the healing months under continuous water immersion by
products to form under these dynamic conditions. Cuenca et al. [27]. This has been attributed to osmotic
The mechanical properties of the healing agents are migration of the active crystalline molecules through-
very relevant in this case since they influence the reoc- out the concrete towards the freshly opened/reopened
currence of damage [42]. If the material formed crack surfaces, where they can be consumed by the
during healing is less resistant than the cementitious healing reactions. Similar results were obtained by
matrix, cracks tend to reoccur at the same site as the Lo Monte and Ferrara [179] for autogenous healing
initial damage. If healing materials with greater resist- of UHPC stimulated by crystalline admixtures. In
ance are applied, it ensures that posterior damage the same framework, Prabhu et al. [193] developed a
occurs at another point, ensuring the availability of methodology to assess the effects of stimulated autoge-
the healing agent. For active cracks, the performance nous healing on the fatigue life of UHPCs. After pre-
570 V. CAPPELLESSO ET AL.

cracking the specimens and subjecting them to up to conditions [48,186], whereas few studies analysed
700,000 cycles between 10 and 80% of the pre-cracking the stability of self-healing mechanisms in real con-
load, specimens were healed for up to three months in ditions [28,32,59,86,198–203]. Actually, a self-healing
water and then further reloaded for 100,000 cycles mechanism should be tailored to the scenario in
after each scheduled healing period. With crack seal- which it will be employed. Knowing the limitations
ing of at least 80% after one month and up to 100% of each healing agent, it is possible to determine
after three months, the stimulated autogenous healing which methods and solutions can contribute to the
was instrumental for flexural stiffness recovery. It survival of the agents. Some agents need protection
reduced the crack opening rate by as much as 18 to survive in extreme environments, for example, bac-
times along the fatigue cycles, improving significantly teria or reactive minerals, which must remain intact
the structure’s fatigue life. and without chemical or physical activity until damage
Feiteira et al. [194] assessed the performance of occurs in the concrete. Survival is the main point to
three encapsulated polyurethane (PU) healing agents guarantee functioning over time since cracks may
with different viscosities under cyclic loading by appear in concrete structures even after several
reloading healed specimens in 3-point bending tests. years, upon which the healing agent should still be
The authors found that the lower viscosity PU (200 active to induce the self-healing operation.
mPa.s) could heal cracks up to 300 µm and reduced Regarding (stimulated) autogenous healing, Ferrara
the water uptake to the level of uncracked specimens. et al. [177,178,204] reported that autogenous healing
After reloading, the strain capacity supported by the in HPFRCCs continued to be active even in specimens
healing agent was between 50% and 100%; hence, pre-cracked at about one year after casting and then
when increasing the crack width above 100% strain, exposed to the healing scenario. Cuenca et al. [27] ver-
the healing failed due to debonding of the PU from ified that the effect of the crystalline admixture,
the crack wall. Nji and Li [195] developed a self-heal- employed as stimulator of the autogenous healing
ing composite with the ability to heal repeatedly the capacity in regular strength concrete, persisted up to
damages by dispersing thermoplastic particles (copo- one year upon repeated cycles of cracking and healing.
lyester) in a shape memory polymer matrix. Despite Regarding autonomous self-healing, extreme con-
a healing efficiency of around 65% of the peak load ditions can greatly influence the stability of healing
being reached, it was only possible to apply the agents. Polymeric healing agents are highly sensitive
method when the samples were kept at 150°C for to high sunlight-induced temperatures, which can
20 min to activate the mechanism. Moreover, the frac- cause premature hardening inside capsules or vari-
ture occurred after five cycles of loading. Botusharova ation in their viscosity [32,42]. Chemical reactions
[196] observed that a bacterium (Sporosarcina ureae) between the precursor and water are other phenomena
is able to go through cycles of self-healing multiple observed by Gruyaert et al. [203] and resulted in cur-
times in cemented sand columns without its activity ing of the precursor within the capsule. Moreover, the
being significantly affected. It was desirable not to pro- permeability of different types of capsule walls, or
voke a repetition in the healing process at the same multilayer capsules, to resist the effects of water or
damage site due to the local consumption of nutrients. oxygen present in the matrix should receive more
However, if there is a guarantee of survival in the research attention. This is important since permeable
matrix and the damage occurs later in another capsule walls can allow chemical interactions [205],
location where spores with activation capacity and leading to premature curing the healing agent and
nutritional availability still exist, it is expected that inactivating its self-healing property [206]. In the vas-
there will be MICP. Justo-Reinoso et al. [197] cular-based approach, the continuity of the tube may
confirmed this by testing bacteria-based mortar be disturbed when the agent hardens inside a broken
samples and obtaining healing at later ages when part of the tube, thus impairing its usability over
new cracks form, but no repeated healing where the time after rupture [42]. To solve these possible inter-
original crack was. ruptions in the tubes, previous works suggested
using an interconnected three-dimensional hollow
network or ‘multi-walled’ tubes consisting of a brittle
Stability of the healing functionality over time
outer tube and an inner flexible, permeable tube to
The use of self-healing technologies in concrete is preserve its continuity [207].
aimed at guaranteeing an increase in the durability Bacteria are susceptible in their enzymatic and
of structures, extending their service life. These metabolic activity to changes in pH and temperature,
benefits can only be achieved when there is a guaran- the presence of salts or acids, or cold climates. The sur-
tee of functionality stability over time in real con- vival of bacteria inserted in cement matrices is a con-
ditions, representing the service scenario. Many cern [48,208] and deserves more intense research
studies have focused on the stability of the healing efforts. However, there are indications that spores
agent, considering only ideal environmental can survive in extreme conditions for a long time
 INTERNATIONAL MATERIALS REVIEWS 571

[197,200,209]. Studies on microbial concrete are cracking, through-crack stress states, non-steadiness
mainly limited to short curing periods and crack for- of cracks, and crack width [17,18,26]. Age of cracking
mation at young ages (28 days or less) [183,186,188]. and repeatability of the cracking and healing scenario
Limiting survival factors for the use of bacteria as heal- relate to the stability and repeatability of the healing
ing agents are related to high pH values in concrete efficiency over time (section ‘Stability of the healing
(pH ≈ 13), dense matrix (narrow pores), and unsuita- functionality over time’).
ble humidity conditions [17]. In order to guarantee the In addition to this guarantee, there is a need to
survival of these microorganisms within the cementi- guide the users on which healing agents should be
tious matrix, protective measures can be taken, such as selected to obtain the best performance in each poss-
encapsulation [186,200] or biofilm formation by the ible real application situation. If self-healing efficiency
organisms themselves [48,185,210]. The microbially is guaranteed, durability can be significantly improved
induced CaCO3 precipitation (MICP) using self-pro- even in contact with aggressive substances. Table 2
tected cultures such as activated compact denitrifying gives an overview (non-exhaustive) of self-healing
core (ACDC) granules is not only effective for crack agents studied in main aggressive scenarios so far.
healing at an early age but also closes cracks in mature They are subdivided according to the mechanism as
specimens of 6 months [48]. Erşan et al. [48] noticed (1) autogenous or stimulated autogenous healing, (2)
that with the ACDC, three days after the occurrence autonomous healing, and (3) bacteria-based self-heal-
of the damage, the MICP process started. Wiktor ing concrete [17]. In the table, tests on self-healing
and Jonkers [183] used bacterial spores (organic agents in paste, mortar and concrete are all considered
acids pathway) and nutrients embedded into without distinction. Still, it should be noted that
expanded clay particles and mixed in mortar speci- upscaling from mortar to concrete while maintaining
mens. They were influential in closing cracks up to the healing efficiency is a significant challenge.
460 µm formed at 56 days of age and healed for 100 When the concentration of the healing additive is
days in water immersion. The authors noticed that kept constant relative to the cement weight, the
bacterial crack healing started 20 days after water move from mortar to concrete results in a significant
immersion, but it is essential to mention that the dilution of the additives. However, when keeping the
time for healing depends on several factors related to same dosage in proportion to the total volume, an
the healing conditions, such as the availability of unacceptable strength decrease and high costs due to
water, bacteria-cell concentration and supply of nutri- high healing agent dosage may result [17]. It is clear
ents. Justo-Reinoso et al. [197], using a similar encap- that most of the research in aggressive environments
sulation process, verified crack healing in new cracks has focused on (stimulated) autogenous healing and
formed in 22-month-old samples after 8 weeks in par- exposure to chlorides and marine environment. So a
tially submersed healing conditions. The needed level noticeable knowledge gap exists regarding the per-
for factors such as pH, water activity, osmolarity, oxy- formance of different types of self-healing materials
gen availability and temperature depends on the in several aggressive scenarios.
characteristics of each microorganism. For Sporosar-
cina ureae, a spore-forming, aerobic, ureolytic bacter- Chloride attack
ium, encapsulated spores remained viable for at least Research on the performance of self-healing concrete
six months, and multiple cycles of sporulation, in chloride-rich environment has mainly considered
encapsulation, and regeneration were possible in the capacity of the cracks to heal. Recently, some
ideal (aqueous) conditions [200]. On the other hand, studies also dealt with improving the durability indi-
for Sporosarcina pasteurii, a highly ureolytic, aerobic cators, such as the apparent chloride diffusion coeffi-
bacterium, there was no guarantee of long-term survi- cient. However, studies that quantify the effect of
val [200]. self-healing on corrosion initiation of the reinforce-
ment are scarce.
Autogenous sealing of micro-cracks was observed
Healing efficiency under different
by Gowripalan et al. [90] in chloride-polluted environ-
environmental conditions
ments. Darquennes et al. [216] performed chloride
There is much evidence from studies that high self- migration tests on self-healed mortars and claimed
healing efficiency can be obtained; however, most of that, although blast furnace slag as SCM had a great
those studies consider the potential for self-healing impact on the self-healing process, crack width played
under ideal conditions [23,26,211,212], while only a a more significant role than material characteristics.
few consider a real or harsh environment Borg et al. [26] studied the sealing of cracks in mortars,
[26,31,44,213]. In order to market self-healing tech- including pulverized fly ash (PFA), silica fume (SF)
nologies, the efficiency of healing agents must be guar- and CA, when the specimens were exposed to con-
anteed under the structural service scenarios, taking tinuous immersion and wet/dry cycles in distilled
into account variables such as mix design, time of water, 33 and 165 g L–1 sodium chloride solution.
572 V. CAPPELLESSO ET AL.

Table 2. Overview of the research on different self-healing mechanisms in aggressive scenarios.

Self-healing mechanism
Main aggressive scenarios (Stimulated) autogenous healing Autonomous healing Bacteria-based healing
Chloride attack PC [90,214] Polyurethane [78,224–228] ACDC bacteria [210,229]
SAP [215] Silver alginate hydrogel capsules
SCMs [43]
. Blast furnace slag [216] Macrocapsules filled with
. Fly ash [26,217] water-repellent agent [46]
. Silica fume [26]
Crystalline admixture [26,28,44,166,218–221]
MgO expansive agent [218]
Fibres
. UHPFRC [219,222]
. ECC [217,223]

Carbonation Calcium nitrate [230] Bacillus cohnii bacteria [230]

Sulphate attack SCMs Isophorone diisocyanate
. Fly ash [231] microcapsules [233]
. Blast furnace slag [231]
Fibres [231]
. ECC [232]

Combined attack (marine PC [234] Sporosarcina halophila

environment) SCMs bacteria [248,249]
. Silica fume [235]
. Blast furnace slag [213,236]
. Fly ash [235]
Organic chelation agents [237–243]
Crystalline admixture [59,106,203,221,244–246]
Granules of Ca(OH)2 and Na2SO4[236]
Cellulose nanocrystals [59,203,244,247]
Fibres
. ECC [232]
. HPFRC [59,203,244,247]
. Alumina nanofibers [59,203,244,247]
. Cellulose nano-fibrils [59,203,244]

Freeze–thaw attack with de- SCMs Water-repellent agent [253] Bacillus cohnii bacteria [254]
icing salts . Blast furnace slag [250] Sodium silicate [253]
. Fly ash [250]
Fibres
. ECC [251,252]

Crystalline admixture [250]

Freeze–thaw attack without Crystalline admixture [237] Toluene-di-isocyanate
de-icing salts Fibres microcapsules [256]

. ECC [252,255]

Water pressure PC [160] ACDC bacteria [184]

SAP [257]
Crystalline admixture [25,172]
Fibres
. ECC [155,258]

High/low temperature and PC [259,260] Double-walled microcapsules [261] MICP [196]

temperature variation (PMMA-MA) capsules filled with Bacillus cohnii bacteria
MgO [262] [268]
Toluene-di-isocyanate Psychrotrophic species
microcapsules [263] (Psy39) [268]
Polyurethane [32,162,264] Bacillus halmapalus
Sodium silicate [201,205,265] bacteria [269,270]
Water-repellent agent [162,264]
Shape memory polymers [195,266]
Epoxy resin [267]
Notes: PC: Portland Cement; SAP: Superabsorbent Polymers; SCMs: Supplementary Cementing Materials; MgO: Magnesium Oxide; UHPFRC: Ultra-High-Per-
formance Fibre-Reinforced Concrete; ECC: Engineered Cementitious Composite; ACDC: Activated Compact Denitrifying Core; MICP: Microbially Induced
Calcium Carbonate Precipitation.

The results were limited to measuring crack sealing sealing crack widths between 150 and 300 μm. CA per-
capacity by image analysis of micrographs, with no formed very well in most exposure conditions and for
tests to quantify the improvement of durability prop- crack widths up to 300 μm. Both in submerged and
erties. Using PFA up to 20% improved the sealing of wet/dry conditions, there were better self-healing
cracks below 150 μm. The addition of 15% SF allowed results in chloride solution than in water [26].
 INTERNATIONAL MATERIALS REVIEWS 573

The ability of crystalline admixture (CA) for self- Concerning the role of superabsorbent polymers
healing in chloride environment has been investigated (SAP) in self-healing, Van Mullem et al. [215] verified
more recently [26,44,220]. Closure of cracks smaller the chloride ingress in mortar samples with a crack
than 300 µm was reached when submerged in chloride width of 150 μm with and without SAPs. The healing
solution. For wet/dry cycles, cracks of the same width regime of wet/dry cyles was performed for 28 days
were also closed, but they took longer [26,44]. Due to before submersion of the specimens in a chloride sol-
its high hydrophilic character, CA improved the seal- ution with 150 g L–1 of sodium chloride. The use of
ing of cracks up to 300 μm in every exposure con- SAP stimulated autogenous healing, and there was a
dition. The higher the chloride content, the faster considerable reduction in chloride ingress in the first
and better the crack healing with CA occurred. More week of exposure. However, healed SAP specimens
recently, Cuenca et al. [219] studied the effect of auto- had a similar chloride ingress as healed reference
genous healing and CA as a healing agent on the specimens after exposure to aggressive media for five
chloride ingress of cracked ultra-high performance weeks.
fibre-reinforced concrete (UHPFRC). Cracks with a As mentioned above, studies on the actual effect of
width of 100 ± 50 μm were generated, and the cracked self-healing on corrosion of reinforcement in chlor-
samples were immersed in 3% NaCl solution for 1, 3, 6 ide-containing environments are rare. Van Belleghem
and 12 months. The results showed that the healing of et al. [227] performed corrosion monitoring on con-
the cracks postponed the ingress of chloride perpen- crete prisms 120 × 120 × 500 mm3 containing a steel
dicular to the crack and the healing efficiency of CA reinforcement configuration with spatially separated
increased with longer exposure time. Abro et al. anode and cathode, allowing measurement of the
[220] studied the effect of a CA system consisting of macro-cell corrosion current. A flexural crack of
Na2CO3 and organic Ca2+ together with zeolite (called 300 µm was self-healed with macro-encapsulated
SH-A) and the same CA with calcium sulfoaluminate polyurethane and exposed to wet–dry cycles with a
(CSA) as an expansive agent and bentonite as a swel- 33 g L–1 NaCl solution. Figure 6 shows corrosion
ling agent (called SH-B). They adapted the ASTM initiation in cracked samples after 6 to 10 weeks of
C1202 to apply different voltages and calculated exposure, while in self-healing samples with low vis-
migration coefficients in mortar with 100 and cosity polyurethane (PU_LV) corrosion did not
300 μm crack widths. The specimens were, after initiate over the 26 weeks of testing.
crack creation, subjected to two healing regimes sub- Corrosion inhibition by nitrite produced by nitrate-
merged in water for 28 and 56 days. For SH-A, the reducing bacteria as self-healing agents can only be
components react with each other to produce proven by actual corrosion tests. Erşan et al. [229]
CaCO3 and fill the cracks, while for SH-B, the CSA were the first to prove the inhibitory effect of the
incorporated as an expansive admixture creates ettrin- microbially produced nitrite on steel plates immersed
gite and bentonite swells by absorbing water to fill the in a chloride-containing solution. Then, ACDC
crack space. Specimens with 100 μm wide cracks after microbial culture (0.5% wt/wt cement), was added to
56 days of healing had almost the same diffusion a cementitious matrix together with Ca(NO3)2 (3%
coefficient as the uncracked specimens, regardless of wt/wt cement) and Ca(HCOO)2 (2% wt/wt cement)
the mixture. For wider crack widths, concrete contain- [210]. Mortar specimens contained 300 µm artificial
ing SH-B showed a greater reduction in coefficient
than SH-A.
Other studies considering self-healing agents with
polymethyl methacrylate or glass capsules as shells
filled with water-repellent agent revealed that the
resistance of cracked concrete against chloride could
be increased [46]. Xiong et al. [43] presented a novel
capsule-based self-recovery system with a chloride
ion trigger, which is functionalized via a smart
response to chloride ions. A chloride concentration
of 0.1 wt-% is already sufficient to disintegrate the cap-
sules of silver alginate hydrogel, releasing the activated
core materials for self-recovery [43]. Van Belleghem
et al. [225] used macro-encapsulated polyurethane to
heal the cracks in concrete exposed to 165 g L–1 of
Figure 6. Electrochemical corrosion parameters of uncracked
sodium chloride solution. The chloride concentration
samples (UNCR), cracked samples (CR), samples healed with
of the healed specimens at the anodic surface of the high viscosity polyurethane (PU_HV) and samples healed
reinforcement was reduced by about 75% compared with low viscosity polyurethane (PU_LV) [227]. Copyright
with cracked specimens without polyurethane [227]. 2018, Elsevier.
574 V. CAPPELLESSO ET AL.

Figure 7. Open circuit potential (OCP) versus standard hydrogen electrode (SHE) for uncracked and cracked plain mortar and
self-healing mortar with denitrifying ACDC. The dotted line shows the limit OCP [210]. Copyright 2018, Elsevier.

cracks, running up to a central steel rebar. The open efficient than for carbonated mortar since there were
circuit potential (OCP) was measured for 120 days more calcium ions available. An encapsulated source
and compared with the limit OCP value (–250 mV of calcium ions aided the self-healing of carbonated
vs standard hydrogen electrode), while the cracked mortars. In carbonated bacteria-based mortars, cal-
surfaces were immersed in 0.5 M Cl– solution. For cium carbonate precipitated on the surface of the bac-
the cracked plain mortar (Figure 7), rebar corrosion terial biofilm, and this led to the closing of the crack.
initiated after 16 days of exposure to Cl– solution
(OCP drop below –250 mV), whereas the ACDC-con-
taining specimens gave proof of corrosion inhibition Sulphate attack
by the produced nitrite during the first 28 days when Only a few studies related to self-healing processes in
the cracks had not yet been completely healed. The sulphate-containing environments can be found [231–
measured mass losses of the steel reinforcement 233], probably due to the fact that sulphate attack
pointed at similar rebar protection as for uncracked mainly occurs combined with chlorides in marine
mortar. environments [86]. He et al. [245] focused on an in-
house developed crystalline admixture and reported
an increased sulphate resistance of uncracked speci-
Carbonation mens considering wet/dry cycles with 5% of sodium
Typically, autogenous healing of cracks in neutral sulphate solution. Dobrovolski et al. [273] studied
environments is based on two main mechanisms, 1) internal sulphate attack by using pyrite, which could
further hydration of unhydrated cement particles seal- release sulphate ions upon oxidation, and investigated
ing the crack and 2) interaction of the hydrated the effect of using CA. Both did, however, not analyse
cement paste phases (calcium hydroxide and C-S-H) the effect for cracked or self-healed samples. Neverthe-
with the CO2 dissolved in water, forming calcium car- less, using CA combined with sulphate can contribute
bonate that seals the crack. These reactions occur to gypsum, ettringite and aragonite crystals formed
when the concrete is under immersion or in high around the cracks. Liu et al. [232] induced multiple
moisture environments [21,271,272]. Most autoge- tensile cracks up to around 150 µm in ECC. The speci-
nous healing processes lead to the precipitation of cal- mens were then immersed directly in a solution con-
cium carbonate in the crack; however, the concrete taining 5% of sodium sulphate. The authors reported
matrix generally has not suffered from the carbonation that ECC tends to heal faster and more completely
in these cases. in sulphate solutions than in water, closing the cracks
There are a limited number of studies related to the after 120 days, probably due to the ettringite/gypsum
self-healing of carbonated concrete or the effect of self- formation within the cracks [232].
healing on concrete carbonation. Alahmad et al. [95] Hung and Hung [231] considered a sodium sul-
showed no carbonation along cracks less than 9 µm phate solution as a medium to promote autogenous
wide due to the autogenous healing of concrete lead- healing in a type of innovative fibre-reinforced
ing to the crack closure. Tan et al. [230] studied the material with cement replaced by ground granulated
effect of carbonation on crack self-healing. The blast furnace slag (GGBS) or fly ash (FA). Cracks
authors searched for two different ways of making between 60 and 120 μm showed a two times higher
more calcium ions available for the healing process closure in sulphate solution than in water after 28
in mortars: 1) adding calcium directly in the form of days. GGBS concrete had a higher crack reduction
a solution of calcium nitrate to the mix design, and ratio than FA concrete for crack widths below 100
2) adding encapsulated calcium in combination with μm, and the presence of fibres allowed to regain the
a bacteria-based self-healing system. They showed strength and stiffness due to the precipitation of the
that self-healing of non-carbonated mortar was more healing product on the fibres. The authors verified
 INTERNATIONAL MATERIALS REVIEWS 575

the formation of calcium carbonate, calcium silicate two years of exposure, considering cracks up to
hydrates, and ettringite mainly deposited within the 50 μm opening. The inclusion of nano-additions
cracks. resulted in a calcium-rich protective layer, also affect-
Regarding encapsulation systems, Du et al. [233] ing the nature of the crack healing precipitates (less
considered microcapsules with isophorone diisocya- presence of ettringite in the inner part of the cracks
nate (IPDI) as the healing agent and paraffin wax, because of lower penetration of sulphates).
polyethylene wax, and nano-silica as shells to promote In marine conditions, the deterioration of concrete
sulphate resistance by healing cracks formed due to is also caused by multi-ion attack, the main ions being
sulphate attack. In theory, the capsules should break chlorides (Cl-) and sulphates (SO2–4 ), both competing
during crack propagation, releasing the healing agent for binding with the C3A of the cement [86]. Accord-
and filling the cracks. However, the evidence was ing to several authors, autogenous healing is improved
insufficient to prove that this happened, although an under seawater immersion, especially when cement is
increase in sulphate resistance in the mortar was partly replaced by blast-furnace slag (BFS) and fly ash
noticed. [274–276]. On the other hand, Palin et al. [277]
showed that after 56 days, BFS specimens could heal
Combined attack (including marine cracks completely when the widths were up to 104
environment) µm in seawater and 408 µm in freshwater. Cracks
Most of the self-healing studies that contemplate com- also healed in PC specimens when their openings
bined attack consider performance in marine environ- were 592 and 168 µm in seawater and freshwater,
ment (sulphate-chloride attack). However, other respectively. The researchers claimed that the healing
combined attack mechanisms can also be found in process in these cracked specimens was related to
the literature as the effect of pressurized water and sus- the availability of calcium hydroxide in mortars and
tained load on engineered cementitious composite specific ions in seawater. Due to the presence of mag-
(ECC) [258]. In [258], the authors verified that cracks nesium ions in the seawater, the combined action may
of 100 µm healed entirely after ten days under con- contribute to brucite precipitation, causing additional
tinuous water pressure and sustained load, blocking sealing of the cracks [213]. Besides, Mohammed et al.
the leakage, and the main products formed were [81] observed that products like ettringite, calcite and
CaCO3 and C-S-H/CH. brucite sealed small microcracks when exposed to a
Recently, Cuenca et al. [59,203] and Lo Monte and marine environment. Danner et al. [235] investigated
Ferrara [244] have investigated the crack self-healing the mineralogy of the self-healing products formed in
capacity of UHPC made with different types of cement cracked marine concrete after 25 years of exposure.
(CEM I and CEM III) with CA, in case also associated They confirmed the closure of cracks with widths
with specific nano-additions (alumina nanofibres and smaller than 200 µm. While calcite formed in the
cellulose nanocrystals and nanofibrils), when exposed outer part of the crack (0–5 mm in depth) followed
to natural geothermal water containing both chlorides by a brucite layer (5–30 mm depth), deeper into the
and sulphates. Both permanent immersion and crack, only ettringite was observed. Xue [221] studied
exposure to wet/dry cycles were considered, and two types of CA in concrete exposed to seawater and
effects of re-cracking after scheduled healing times, observed that the self-healing performance is strongly
as long as six months, were investigated. Single loca- related to the type of CA combined with the environ-
lized cracks opened up to 150 μm, and multiple crack- mental condition. The author produced cracks with a
ing states with single cracks not wider than 100 μm width of 60 to 70 μm in mortars to be analysed under
were studied. While the self-healing performance of immersion and wet–dry cycles. The wet–dry cycles
UHPC made with CEM III was generally better than increase the chloride ingress due to the capillary suc-
with CEM I, the differences tended to fade upon mul- tion compared to the immersion. He also considered
tiple re-cracking and longer exposure times. Effects of using MgO-cement to collaborate with autogenous
nanoparticles were evident in enhancing the crack- healing in seawater and chloride solution. The sea-
width control capacity of the material, guaranteeing water reduces the chloride ingress compared to a
single cracks not wider than a few tens of microns pure chloride solution because of the brucite precipi-
for the same level of deformation, and hence guaran- tation in the cracks [246]. He et al. [106] applied a
teeing a better healing performance [247]. For the non-commercial crystalline admixture in mortar
same UHPCs and aggressive scenario, Giménez et al. exposed to seawater and could heal completely cracks
[247] have developed a tailored test set-up to simulate up to 350 µm, while in freshwater, the maximum crack
interaction with continuously flowing water as it width healed was 240 µm. They verified brucite and
occurs, e.g. in cooling towers in geothermal power calcium carbonate as healing products in the cracks.
plants. They found that while the continuously The healing agent used in seawater conditions must
flowing water may exacerbate the effects of previously guarantee efficiency because of the combined attack.
induced cracking, it resulted in partial healing after In the case of bacteria-based self-healing concrete, a
576 V. CAPPELLESSO ET AL.

Figure 8. Crack closure by image analysis of self-healing organic chelation agents in cement paste with crack widths of 400 µm
immersed in seawater [240], with permission from ASCE.

bacterium must be chosen suitable to survive in sea- hydroxide or sodium sulphate, enveloped in PVA
water. Palin et al. [269] developed a bacteria-based film, as an activator for self-healing in mortars
self-healing cementitious composite for application exposed to seawater. The authors identified complete
in low-temperature (8°C) marine environments, closure of cracks smaller than 300 µm after 60 days
showing almost complete healing of cracks; 95% heal- of exposure when using calcium hydroxide as an acti-
ing capacity for cracks up to 400 μm in width and 93% vator; sodium sulphate was ineffective. Liu et al.
for 600 μm wide cracks [269]. Erşan et al. [210] ana- [239,240] proposed using organic chelation agents to
lysed the tolerance of a nitrate-reducing ACDC efficiently block cracks in a marine environment.
microbial culture to osmotic stress in the marine The mechanism involves an innovative way to
environment. The major influential parameter rearrange the crystals in the cement matrix related to
appeared to be the ionic strength (μ) of the solution the growth process and shape by means of their chela-
rather than a change in specific ion concentrations. tion activities [237,238,241]. Liu et al. [240] prepared
In typical seawater (μ ∼ 0.7 and [Cl−] ∼ 0.5 M) results cement pastes with 1.5% vs cement of tetrasodium
revealed that ACDC culture showed adequate resist- ethylenediaminetetraacetic acid (EDTA-4Na), trietha-
ance [210]. Khan et al. [248] used spores of Sporosar- nolamine (TEA), or sodium hexametaphosphate
cina halophila bacteria, calcium lactate and expanded (SHMP) and exposed them to seawater (submerged
perlite aggregate as the carrier in mortar subjected to condition/wet–dry cycles). The TEA agent could
submerged and tidal marine environment for 90 heal completely cracks with a width of 400 µm after
days. The presence of bacteria and availability of two days under immersion in seawater due to an
water and oxygen within the crack enhanced the ara- increase in the OH- concentration, inducing the
gonite and brucite formation along the entire crack rapid formation of brucite (Figure 8). For cracks of
depth, with a crack closure ratio of 50% considering 800 µm, 56 days of exposure were necessary to reach
cracks up to 800 µm. a crack closure of 97%. Using 1% of TEA (by the
The improvement of autogenous healing due to the mass of cement) closure of 400 µm wide cracks was
use of blast furnace slag (BFS) is well known [274– obtained 20 times faster than for a control paste
276]. However, the reaction usually takes time, and [239]. While for autogenous healing in seawater cracks
in realistic aggressive environments, it might not pre- up to 150 µm could close for both submerged and
vent the aggressive agent from entering before the wet–dry cycles, cracks of 400 µm reached only 90%
healing occurs. In this respect, the use of alkali-activa- of closure in wet–dry cycles, and 60% under sub-
tors combined with BFS can bring benefits to improve merged condition [234]. Wu et al. [242] considered
healing. Kim et al. [236] tested granulated calcium to include 5% by volume of CaO–NaAlO2 in cement
 INTERNATIONAL MATERIALS REVIEWS 577

paste and exposed the samples to seawater for acceler- samples (binder containing 5% silica fume) to FT
ating the healing process by binding chemically the cycles to cause microcracking (1–10 µm width cracks).
ions present in this solution. They could heal cracks After damage, they completely submerged the samples
of 400 µm wide in 1 d of seawater immersion. How- in water where the specimens could heal autoge-
ever, there is no evidence how these additions can nously; however, the samples could not recover their
affect a concrete structure in the long-term with pristine performance according to resonance fre-
regard to performance and durability. quency and compressive strength tests. In another
research also regarding autogenous healing, concrete
Freeze–thaw attack with or without de-icing salts that lost 50% of the dynamic modulus due to frost
In order to increase the durability of structures sub- damage could almost completely recover after sub-
jected to freeze–thaw (FT) and reduce internal sequent storage in water, somewhat varying with con-
damage, the most appropriate way is to reduce the crete composition and degree of deterioration [278].
possible moisture uptake. FT tests on cracked and However, in real situations, it is unlikely that struc-
self-healed specimens showed clearly that the deterio- tures will have ideal curing to guarantee the healing
ration was governed by the capacity of water pen- after the damage. Ma et al. [150] analysed concrete
etration through cracks [278]. Self-healing specimens subjected to FT cycles with de-icing salts
mechanisms have been studied to a limited extent, according to the standard GB/T 50082-2009. Each
but they can contribute to a large degree in guarantee- freeze–thaw cycle lasted for nearly 4 h, the thawing
ing the reduction of water entry through the closure of time being longer than 1 h. The central temperature
cracks. Most research related to this extreme condition of the specimens was controlled subsequently at −18
shows the improvement by reducing matrix per- ± 2 and 5 ± 2°C. After cracking occurred due to the
meability or by methods to reduce the crack propa- FT actions, they noticed no tendency of autogenous
gation [133,135,252,279,280]. Other studies are healing which may be due to the fact that the samples
restricted to the use of products which increase the were continuously kept in cold conditions. Zhu et al.
scaling resistance of the surface layer of concrete struc- [252] noticed that ECC had weakened properties
tures [133,135]. when submitted to FT conditions, and in the presence
SAPs can be added to cementitious materials to of salts, the self-healing degree was worse than in the
increase the FT resistance [279] and induce self-seal- condition with only water. In contrast, Şahmaran
ing and self-healing effects [280]. SAPs create macro- and Li [251] reported that ECC has the capacity to
pores that act in the same way as air voids created by heal sufficiently under freezing and thawing cycles in
an air entraining agent. These macropores also attract the presence of salt solutions to restore nearly the orig-
microcracks formation, stimulating multiple cracks of inal stiffness because of the multiple cracks with a
reduced width and contributing to self-healing [280– narrow width. Hooshmand et al. [255] studied large-
284]. Due to the addition of SAP, a much less pro- scale panels made with ECC (with blast furnace slag
nounced decrease in the dynamic modulus was or fly ash) and PC for seven months in a natural
observed as a result of FT testing without de-icing environment that suffered FT cycles during certain
salt [279]. However, published research does not periods. They noticed the greater presence of C-S-H
include the analysis of cracked and healed SAP-con- compared to calcium carbonate in cracks healed in
taining concrete in this extreme condition but is lim- the ECC made with fly ash compared to the one
ited to the increased frost-thaw resistance of made with slag. The authors believe that this might
uncracked matrices. be the reason for the higher resistance of the healing
Suleiman and Nehdi [260] evaluated autogenous products in the fly ash ECC mixes to the harsh natural
self-healing in mortar specimens under water submer- conditions. In this case, the fly ash led to a better self-
sion at a constant temperature of 19°C and in an healing performance of cracks formed, while the ECC
environmental chamber under cyclic temperature with slag had higher crack formation exceeding the
from −10 to 40°C and relative humidity in the range recommended crack width limit (100 µm) as a result
of 20–90% while monitoring during one year. The of higher shrinkage strains.
complete cycle was twelve days, four days at 20°C Zha et al. [237] analysed a novel CA as an ion che-
and 90% RH, four days at −10°C and 20% RH, and lating agent in mortars after 100 FT cycles. The speci-
four days at 40°C and 60% RH. They did not report mens were subjected to a further healing regime of 28
significant healing in the cyclic temperature environ- days submerged in water after FT. The CA mortars
ment even for 30 µm crack width, while in submerged showed a compressive strength recovery of 51.8%,
conditions, cracks with 300 µm width were closed and cracks with a width of 320 µm could be healed.
entirely. However, one could comment that such Needle-like crystals were observed in the pores and
accelerated tests might not allow sufficient time for micro-cracks due to the freezing, with calcium car-
healing to occur between the low temperature parts bonate, C-S-H and ettringite being identified as the
of the cycles. Jacobsen et al. [134] subjected concrete main components. Wang et al. [250] investigated a
578 V. CAPPELLESSO ET AL.

Figure 9. Set-ups for water permeability test under low/high pressure: (a) Aldea set-up for water permeability [285] Copyright
2011, Elsevier., (b) HEALCON water permeability test [289]. Copyright 2019, Elsevier, (c) Korean water permeability test [294].
Copyright 2017, Elsevier.

non-commercial CA with blast furnace slag or fly ash. the FT resistance. Cappellesso et al. [254] considered a
First, mortar specimens were subjected to 100 FT self-healing bacterial concrete exposed to FT associ-
cycles with 3 wt-% NaCl. Subsequently, the damaged ated with chlorides. The Bacillus cohnii bacteria
specimens were subjected to a healing regime (water increased the performance of the concrete reducing
immersion) for 28 days. The authors verified the heal- the scaling by 90% and chloride ingress by 46%
ing ability after exposure by compressive strength under FT compared to a reference concrete without
tests. The mortar made with CA and blast furnace bacteria. The addition of the bacteria had an air-
slag recovered about 22% of the strength and reduced entraining effect that might explain the improvements
about 27% of harmful pores (larger than 0.1 mm). reached. However, it could not block the chloride
Concerning autonomous self-healing mechanisms, ingress through a pre-defined crack of 170 ± 50 µm
Du et al. [256] used toluene-di-isocyanate (TDI) healed during 28 days of water immersion before
microcapsules in concrete specimens subjected to exposure. The healing products formed by the bacteria
100 FT cycles to verify the ability to heal cracks after before exposure to freezing temperatures were
FT. After exposure, the specimens were inserted in a damaged by scaling due to freezing during the sub-
healing regime of 7 days under water immersion. sequent exposure period. The same authors [253]
The authors believe that the damage caused by crack also considered two types of healing agents (water-
propagation was able to break the capsules to release repellent agent and sodium silicate), injected them in
the TDI within the microcracks. The TDI might cracks of 100 µm wide, and subjected them to FT scal-
react with the water, healing the cracks and improving ing. The sodium silicate after a healing regime of 14

Figure 10. Outcomes of leakage rate versus time in ECC under water pressure: (a) short-term self-healing in hours, (b) long-term
self-healing in days. North represents cracks with a width of 100 µm, while South represents 50 µm wide cracks [155]. Copyright
2021, Elsevier.
 INTERNATIONAL MATERIALS REVIEWS 579

days in a layer of water could prevent the chloride 10(b)), whereas, for the crack with a width of 100 µm
ingress through the crack after 56 FT cycles with (‘North’ in Figure 10(b)), ten days were needed to
3 wt-% NaCl. quash the leakage [155]. SEM-EDX analyses showed
that the cracks of ECC were filled with higher
Water pressure quantities of dense self-healing products (CaCO3
Many studies related to self-healing mechanisms and C-S-H / CH) compared to conventional concrete.
assess the influence of water pressure through water
permeability methods [25,36,285–293,172,184,194, High/low environmental temperatures and
212,213,224,225,257]. Figure 9 shows some examples temperature variations
of test methods used to verify the self-healing per- A few studies have been devoted to assessing the per-
formance under low or high water pressure. However, sistence of the healing capacity under or after exposure
most of these studies aim to quantify the crack closure to a non-standard temperature, which can cause issues
by the reduction in permeability to water under a cer- to the survival of the healing agent, whether it is a min-
tain pressure, which usually is smaller than 0.005 MPa eral, bacterium or an encapsulated polymer.
(0.05 bar), without aiming to assess the ability of the In bacteria-based self-healing concrete, the use in
healing agent to face the hydraulic pressure exerted sporulated form brings benefits for application at
on the healing products [25,172,257]. Gruyaert et al. high temperatures, given their resistance to survive
[257] analysed mortars made with SAP to investigate in these conditions [200]. Botusharova [196] demon-
the resistance of the healed specimens to a high strated that the MICP biosealant used was thermally
water pressure of 0.2 MPa. The authors could verify stable at temperatures up to 840°C. Contrarily, ureoly-
the sealing efficiency by water flow tests on mortar tic activity is mainly maintained at temperatures ran-
made with SAP having 150 μm of crack width and ging from 4°C to 60°C [196]. Palin et al. [269] applied
healed in wet–dry cycles for 28 days. a bacteria-based self-healing cementitious material
CA have been applied in real structures with the containing Bacillus halmapalus spores at a low temp-
aim of reducing the permeability of concrete essen- erature of 8°C submerged in seawater. They obtained
tially in tanks, containment walls in contact with a healing efficiency of 95% for 400 μm crack width
groundwater, tunnels and reservoirs, i.e. structures and 93% for 600 μm crack width after 56 days. Paine
where water is acting with a pressure gradient. Borg [295] called attention to the need to choose bacteria
et al. [26] called attention to the fact that structures fit to the conditions, e.g. types that can withstand
under water pressure present an ideal condition for hot temperatures (thermophilic), low temperatures
the crystalline admixture because of the huge avail- (psychrophilic), and/or presence of salts (halophilic).
ability of water. Nonetheless, research is needed to Skevi et al. [268] focused on the survival of bacteria-
determine the hydraulic pressure threshold that can based self-healing agent at low temperature (7.5 ± 2°
be supported by the formed healing products and C) where a psychrotrophic species (Psy39) and Bacil-
the effect of hydraulic pressure on the continuation lus cohnii were used. Mortars with cracks between
of the cracking process, which can increase the 400–500 µm made at 28 days of age were subjected
damage and leakage in these structures. Regarding to a semi-submerged condition at 7.5 ± 2°C and 20°
autonomous healing, Erşan et al. [184] measured the C during more than 28 days. While Bacillus cohnii
healing performance of biomortar with ACDC under bacteria have better performance at 20°C, the Psy39
0.01 MPa of pressure in a water permeability test. reached the greatest efficiency for both temperatures.
The cracks of 400 µm showed 80% decrease in the Reinhardt and Jooss [259] assessed autogenous self-
flow rate after 28 days of healing in water immersion. healing under water in a high-performance concrete at
ECC has shown high potential to withstand 20, 50 and 80°C, relating it to variation of crack width.
extreme conditions, mainly because it contains A clear tendency for a faster self-healing process in
fibres, contributing to the reduction in the crack case of an increase in temperature could be recognized
widths. Hooshmand et al. [155] verified the efficiency and is presented in Figure 11.
of ECC, pre-cracked in direct tension up to two differ- Van Tittelboom et al. [264] assessed the efficiency
ent crack levels and healing under 0.048 MPa water of encapsulated polyurethane and water-repellent
pressure, in short-term (up to 30 h – Figure 10(a)) agent as healing agents in case of thermal crack for-
and in long-term (up to 10 days – Figure 10(b)). The mation in sandwich panels. Similar research by
authors showed the potential of the investigated Gruyaert et al. [162] in a real-scale test evaluated
high-performance cementitious composites to heal self-healing mechanisms in the presence of thermal
better under water pressure, as compared to conven- cracks of around 50 and 150 μm. Polyurethanes
tional concrete, rightly due to the capacity to spread (PU) and water-repellent agents embedded in glass
the localized damage into multiple cracks. Closure macrocapsules were both able to reduce the per-
by autogenous self-healing in less than a week was meability by healing the thermally induced cracks,
observed for the 50 µm wide cracks (‘South’ in Figure but under reloading, only water-repellent agents
580 V. CAPPELLESSO ET AL.

Figure 11. Decrease of the normalized flow rate because of the self-healed cracks for HPC at a pressure gradient of 1 MPa/m (a)
with various crack widths and a temperature of 20°C, (b) with various temperatures and crack width of 50 μm [259]. Copyright
2003, Elsevier.

retained the crack watertight. The PU lost the bond microcapsules provide enhanced durability at high
with the crack wall increasing water absorption and temperatures compared with single-walled microcap-
showed moreover a high leakage out of the cracks sules preserving a suitable interfacial bonding. Du
due to the decrease in viscosity because of the high et al. [263] presented different shell materials which
temperature ( 5̴ 5°C) [162]. Van Belleghem et al. [32] were analysed under different temperatures (from 10
tested five different types of polyurethanes to verify to 60°C), containing toluene-di-isocyanate as core
the efficiency of self-healing for concrete exposed to material. The authors were concerned regarding the
high temperatures (50°C). They showed that for the thermoplastic materials commonly used to produce
specimens healed with Product E, a certain type of shells, such as paraffin, polyethylene, or polypropylene.
polyurethane, at 50°C, the glass macrocapsules were For instance, paraffin has a melting point of around 60°
completely empty (Figure 12(d)). It is related to the C; thus, at high temperatures, the shell could be
reduction by around 85% of the polyurethane viscosity destroyed before being broken by the cracking process.
when the temperature was increased from 20 to 50°C. Then the healing agent inside the microcapsules may
The excessively low viscosity of the healing agent can not flow out to heal the crack at the right time, or the
impair the healing efficiency as the polymer may leak healing agent efficiency could be affected. Kanellopou-
from the crack [32,162]. This indicates that tailoring of los et al. [205] indicated that microcapsules with
the healing agent’s viscosity to the expected tempera- sodium silicate solution as a core and gelatine-acacia
ture conditions is a key issue in obtaining greater gum as a shell were stable under thermal effects up to
self-healing efficiency. 190°C. Mao et al. [201,265] demonstrated that sodium
Regarding self-healing systems with microcapsules, silicate microcapsule-based self-healing of oil well
Mostavi et al. [261] showed that their double-walled cement could present a good stability at the high

Figure 12. X-ray µCT cross sections showing the crack intersecting an embedded macro-capsule for a specimen healed with Pro-
duct C at (a) 20°C and (b) 50°C; a specimen healed with Product E at (c) 20°C and (d) 50°C; and (e) a specimen with an untreated
crack, with permission from ASCE [32].
 INTERNATIONAL MATERIALS REVIEWS 581

Figure 13. Digital image correlation (DIC) camera snapshots showing closure of unreinforced concrete crack using PET tendons.
Axis indicates strain. (a) Loaded to 0.5 mm CMOD; (b) unloaded; (c) post-activation [266].

temperature of 80°C. Ren et al. [262] applied innovative Test methods to evaluate self-healing
microcapsules with temperature-adaptive properties efficiency
and analyzed how those capsules affect the concrete
The high number of varying test methods to measure
properties when the concrete mixing is developed at
the durability of healed mortar/concrete makes it
normal temperature (20°C) and high temperature
difficult to compare and interpret data by different
(60°C). The capsules were made by polymethylmetha-
research groups. The standardization of methods for
crylate-methacrylate (PMMA-MA) shells and mag-
crack creation, curing procedures, and methods to
nesium oxide core. The authors verified 3% higher
evaluate self-healing efficiency, would contribute to
healing efficiency considering the high temperature
obtaining reliable data and would allow for incorpor-
due to the fact the PMMA-MA shell has a glass tran-
ating self-healing concepts into durability-based pre-
sition temperature of 40°C, so less premature rupture
dictive models and design approaches.
of the shell appears when mixing at 60°C.
A previous extensive literature review performed by
Thermal effects can also be applied to help heal
Ferrara et al. [18] shows that there are multiple essen-
cracks. Shape memory polymers (SMP) act as tendons
tial variables that influence laboratory experiments to
which contract upon heating to restore the original
assess self-healing concrete, such as the age of crack-
shape, creating a prestress in the element, closing the
ing, crack width, presence of water, the status of the
crack mechanically and, as such, enhancing the heal-
crack (static vs active cracks), and repeatability.
ing action [195,266]. Li and Neetles [296] compared
Beyond that, to measure durability, the performed
two types of SMP, reaching almost 100% of recovery
tests depend on the main attack mechanism(s),
by thermomechanical analysis and closing completely
which are related to the aggressive scenarios, and to
the crack with no specification of the width. Nji and Li
the main transport mechanism of the aggressive sub-
[195] considered an SMP with 6% of copolyester par-
stances. Often due to the combined action of aggres-
ticles obtaining repeatability of 5 heating cycles. Teall
sive agents or the variety of ingress mechanisms,
et al. [266] made use of an SMP (polyethylene tereph-
evaluating an isolated parameter will not bring
thalate [PET] filament) to close cracks of 750 µm by
sufficient information regarding the durability per-
electrical heating activation and reached a crack clo-
formance. The determination of the effectiveness of
sure of 85% in unreinforced beams (Figure 13), and
self-healing inside the cracks is not straightforward,
26–39% considering reinforced beams. They believe
but confidence can be raised with suitable test
that the reinforcement can cause interference in the
methods.
electrical activation. Nishiwaki et al. [267] produced
In order to avoid duplication, the authors refer to
a self-healing system for concrete that incorporates a
the state-of-the-art report prepared by SARCOS
heating pipe device filled with a repair agent, which
COST Action CA15202 to learn more about the exper-
was a one-component epoxy resin. With an increase
imental characterization of the self-healing capacity of
in temperature around the crack, the organic film
cement-based materials and its effects on material per-
(ethylene vinyl acetate [EVA] polymer) that composes
formance [18]. For the sake of completeness of the
the pipe melts at 93°C allowing the healing agent to
current paper, an overview is given (Table 3) of tests
release within the crack.
 582
Table 3. An overview of methods (non-exhaustive) to measure the performance of healing agents with a focus on durability.
Influential factors related to
execution and interpretation of
Test method Purpose Advantages the test Limitations Some references (non-exhaustive)

V. CAPPELLESSO ET AL.
Crack closing evaluation
Optical microscopy . Assessing crack closing at . Non-destructive . Water on the surface can . No guarantee about the closure of the [1,25,277,297,298,26,27,36,171,172,188,225,239]
the surface . Simple working principle and complicate the crack in deeper parts and the liquid
. Quantification of the easily implementable interpretation of the images tightness
change in crack width or . Monitoring crack closure in time . No volumetric crack closing ratio can be
area and space (along the crack) obtained
possible

Fluorescent . Assessing self-healing in . Good visualization of crack . Matrix fragility . Destructive testing [23,36,235]
microscopy and thin depth healing over the complete depth . Sample preparation . Upon extracting the sample, damage can
sections be caused

Scanning electron . Production of high- . Backscattered electrons (BSE): . Sample preparation . Small samples and test area [23,24,26,168,172,188,204,235,297,299,300–309,310]
microscope (SEM) resolution images differences in the atomic number (polished for BSE, unpolished . Many images need to be studied for
. Confirming the presence of can give information on phases for topographic imaging) representative results
healing products . Secondary electrons:
. Providing information topographic information
about the chemical
composition when
combined with EDX (see
below)

Energy-dispersive . Chemical identification of . Self-healing products chemical . Selected location: elemental . Small samples and test area
X-rays (EDX) the self-healing products identification mapping and line scans
. Qualitative and quantitative
analysis

X-ray computed . Defining damage, failure, . Non-destructive testing . Quality of data analysis . Small sample sizes [168,184,188,297,311]
microtomography porosity: volume of the . Evidence of the damage . Restricted to pastes and mortars
(µCT) crack and healing products . Quantification of the crack
. Defining the geometry of closure by self-healing processes
the crack
. Determining the internal
X-ray computed self-healing capabilities . Can be used at the concrete level . Larger equipment [310,312–314]
tomography (CT) . Phase identification and
density

X-ray radiography . Measurement of moisture . Non-destructive testing . Proper image overlap . Difficult to obtain a perfect overlap of the [1,141,292,315–323]
uptake over time . Images at high-spatial resolution images before and after healing
 . Visualization of rebar . Comparison between
corrosion in cracked and specimens in the dry and
healed concrete wet state

Nuclear magnetic . Qualitative and quantitative . Non-destructive testing . Regions filled with water can . Small sample size (nm features) [324,325]
resonance (NMR) approach of water transport . Allowing a detailed study over be confused with crack . Equipment not widely available
time locations

Durability tests
Water permeability . Assessing the flow of water . Different pressures can be . Geometry, tortuosity and . Restricted to saturated elements [25,36,155,172,184,194,212,213,224,225,236,239,257,
test through healed cracks applied width of the cracks 285–293]
. Obtaining the water . Evaluation of the tightness . Pressure of water
permeability coefficient . Measurement of the resistance . High water velocity can
. Assessing the recovery of against aggressive substance remove the healing products
water tightness penetration and causes inaccuracy in the
test results

Gas permeability test . Obtaining the gas . Type of gas Set up with high demands for precise [309,326,327]
permeability coefficient . Pressure of gas measurement control
. Assessing the recovery of
gas tightness

Sorptivity test . Rate of absorption . Much simpler, less time- . Sorptivity is not restricted to . Restricted to dried elements [48,181,194,201,221,252,290,310,322,328–333]
(sorptivity) of water consuming, and does not require the crack but also via the . Interpretation of the results can be difficult
a special device when compared matrix by capillary porosity in the case of partial crack healing, which
to the permeability tests . Tortuosity and crack may increase sorptivity in comparison to
geometry cracked and unhealed samples due to
higher capillary forces

Chloride penetration . Assessing the transport of . Direct evaluation of the self- . Saturation degree of the . Need for accelerated tests instead of [26,78,166,199,224,227,228,254,331]
chlorides and rebar healing performance in chloride- concrete determines the realistic tests, which are time-consuming

INTERNATIONAL MATERIALS REVIEWS

corrosion risk containing environments transport mechanism
. Determination of the Cl . A choice should be made
profile near the crack regarding the test method to
. Determining corrosion evaluate Cl diffusion,
onset, Cl threshold absorption, migration

Sulphate attack . Evaluating the sulphate . Direct evaluation of the self- . Sulphate concentration . Need for accelerated tests due to slow [86,113,231–233,245,334]
degradation healing performance in sulphate- . Specimen size evolution in case of realistic tests
containing environments . pH . Sensitive test methods must be used to
. Temperature estimate the damage when simulating real
non-accelerated conditions

(Continued)

583
Table 3. Continued.

584
Influential factors related to
execution and interpretation of
Test method Purpose Advantages the test Limitations Some references (non-exhaustive)
. Type of binder (amount of

V. CAPPELLESSO ET AL.
C3A)
. Presence of SCMs
. Type of immersion (total,
partial or wet–dry cycles)

Carbonation test . Assessing the transport of . Direct evaluation regarding the . Slow process in natural . Need for accelerated tests to limit test [310]
CO2 and rebar corrosion risk carbonation resistance of self- conditions duration
. Determination of the healing cementitious materials . Risk of high %CO2; artefacts . The concentration of CO2 is suggested to
neutralization profile by high %CO2 be not more than 4%
. Relative humidity

Recovery of mechanical properties

Ultrasonic wave . Detect concrete quality, . Simple working principle and . Variability in the coupling . Qualitatively reliable [24,167,192,252,261,335–339]
propagation internal defects, cracks and easily implementable between the sensor and the . Moisture content, temperature changes,
voids . Immediate results, portable, surface of the specimen presence of reinforcement, and local crack
large penetrating power leads to variability in results bridging can affect results
. Moderate cost . Specimens that are rough, irregular in
shape, very small, thin, or not
homogeneous are difficult to inspect.

Resonance frequency . Determining the rate and . Non-destructive testing . Further continued hydration . Comparative analysis is defined by [311,340]
extent of self-healing . Quantification of the crack can mask the self-healing controlled features such as saturation and
closure efficiency rate of damage
. Sensitive to small variations . Fluctuations in the
measurements due to the
saturation changes

Mechanical recovery . Evaluation of recovery of . Correlate the recovery of . Loading rate . Comparative test needed on sound [18,24,31,171,177,179,186–188,237,309,329,338,341–
test mechanical properties different mechanical properties . Choice of characteristic specimens 346]
(strength, stiffness, strain and points on the loading curve . Measurement of the overall properties of
deformation capacity) with crack and choice of parameters the specimen
closure. used to calculate recovery . Need to distinguish between effects of
. Measures evolution of indexes concrete aging and effects of healing
parameters useful for structural
design

Acoustic Emission . Continuous monitoring of . Breakage of capsules, cracking, . Positioning of sensors . Expensive [36,236,291,310,336,339,347–349]
crack initiation and and healing can be monitored . Defects or early cracks . Defects already present are not detected
propagation . Evaluation of the durability of blocking the wave travel
. Evaluation of healing concrete elements subjected to path for later cracks
. Monitoring release of water different loading types
by SAPs
 INTERNATIONAL MATERIALS REVIEWS 585

and methods for characterization of self-healing design service life based on the ‘deemed to-satisfy’
related to (1) crack closing quantification, (2) recovery and the ‘avoidance of deterioration’ approach, where
of durability properties and (3) recovery of mechanical durability-related exposure conditions are defined
properties, mainly based on [18]. according to 17 exposure classes [60]. Nevertheless,
fib Model Code 2010 [350] recognizes that
A pathway for the incorporation of self- If more refined service life designs are to be under-
healing performance into a life-cycle taken by the use of deterioration modelling, this
spanning durability-based design classification of the environmental load must be
related to quantified parameters, e.g. chloride concen-
The obstacles preventing the large-scale application of trations for marine structures. When publishing this
self-healing technologies in day-to-day construction Model Code, such quantified parameters were not
available in any operational standard. The infor-
practice are currently related, on the one hand, to
mation must therefore be found by measurements
the lack of standardized test methods to assess the on existing structures and in the literature.
self-healing performance and, on the other hand, to
the complete absence, in structural design codes, of It is worth remarking that, even in the case of those
any pathway to fully recognize the effect of self-healing deterioration mechanisms for which models with a
functionalities on material and structural relatively broad international acceptance exist, the
performance. codes consider the concrete as a passive provider of
In a broader sense, this is related to how durability protection to the rebar rather than as an active player
is currently dealt within structural design codes. The in the durability performance achievement. Verifica-
options, as per fib Model Code for Service Life Design tion of performance implies checking that the critical
are: full probabilistic approach; semi-probabilistic front of the deterioration mechanism does not reach,
approach (partial factor design); deemed-to-satisfy over the intended time, a depth equal to the reinforce-
rules; and avoidance of deterioration [66]. ment cover. No information is given on the time evol-
ution of the material properties, and hence no
(1) The full probabilistic approach is based on prob- information is obtainable on the residual level of
abilistic models for deterioration and material ‘safety’, which characterizes the structure at the
resistance. Statistical information on the variabil- intended end of its service life. Moreover, the models
ity (randomness) of the contributing parameters proposed for the deterioration mechanisms explicitly
is required in this approach. These models should considered in the codes, including carbonation-
be validated to give realistic, representative, and induced corrosion, chloride-induced corrosion and
reliable results. The method should also be acid attack, do refer to sound un-cracked concrete, a
based on appropriate test methods with statistical state which hardly represents the real service con-
evaluation. Generally, the normal probabilistic ditions of reinforced concrete structures [55,60].
distribution is used to account for the randomness Though the verification of some durability-related
of the parameters involved. limit states may include a crack width control, no
(2) The partial factor approach uses statistically information is provided on the evolution of such a
derived partial factors and is intended to be a width along the structure’s service life due to inter-
practical, statistically reliable design tool. How- action with material aging, deterioration or self-heal-
ever, the partial factors are derived using the ing phenomena, which can accelerate or slow down
same models as the full probabilistic approach. the advance of the deterioration mechanism’s critical
Such data is obtained from a huge set of examples. front. This is also due to the fact that most of the cur-
(3) The deemed-to-satisfy approach is conceptually rently available design codes have been originally for-
like current prescriptive durability specifications, mulated for normal strength (Portland cement)
based on a selection of design values (e.g. dimen- concrete and then further extended, with some limit-
sioning, material and product selection, execution ations, to the broad category of high-performance
procedures). concretes, which do actually have to include self-heal-
(4) The avoidance of deterioration approach requires ing concrete.
the use of deterioration-resistant materials such A possible pathway to incorporate the benefits of
as stainless steel, or concrete protection systems self-healing into a durability-based structural design
such as coatings, thus limiting or eliminating should, first of all, start from the quantification of a
deterioration of the structure. Maintenance may ‘scenario-dependent’ healable crack width. In this
still be required, such as renewal of coatings respect, an attempt has been recently made to quantify
from time to time. the dependence of the healable crack width on a plur-
ality of variables, including concrete composition,
So far, national and international concrete stan- initial crack width, type and duration of the healing,
dards provide requirements to achieve the desired through artificial neural networks [20]. Similarly, the
586 V. CAPPELLESSO ET AL.

effects of self-healing on the parameters governing the performances like strength, durability and unit weight,
degradation mechanisms should be quantified, e.g. while in Van den Heede et al. [368], extending the
from an AI-based analysis of the literature reviewed assessment by Damineli et al. [370], proposed a FU
in this paper. The incorporation of the ‘self-healing defined as ‘the total amount of binder per m3 concrete
affected’ durability parameters and ‘self-healing necessary to deliver 1 MPa of strength and one year of
delayed’ degradation mechanisms into structural service life’ emphasizing the link between LCA out-
design algorithms could lead to the quantification of puts and SL.
the benefits of using self-healing concrete in terms of System boundaries, as summarized in Figure 14,
extended design service life and reduced maintenance could be: (i) cradle-to-gate (referring to the pro-
[351–353]. duction process), (ii) cradle-to-grave (including,
In addition, life cycle assessment (LCA) can prop- besides the ‘production phase’ also the ‘use’ and
erly help to support and to promote the deployment ‘end-of-life’ stages) or, better, but not yet investigated
of self-healing concrete, assessing all the impacts for self-healing concrete (iii) cradle-to-cradle (which
from mining, production, and use until the possible includes ‘production’, ‘use’, ‘end of life’ and ‘recycling
end-of-life scenario [354,355]. Compared to LCA of the waste’ phases) [371,372]. Wu et al. [373] warned
studies for traditional concrete, the scientific literature to exclude ‘use’ and ‘end of life’ stages according to the
about LCA of self-healing concrete, is limited. As a ISO 24067 standard only if they affect LCA results by
matter of fact, previous LCA studies focused mostly less than 1%. The importance of an appropriate system
on sustainable concrete options like concrete contain- boundary is clarified further by De Schepper et al.
ing incinerator ashes, marble sludge, blast furnace [372], who demonstrated how, choosing an appropri-
slag, recycled aggregates or fly ash [50,356–359]. ate cradle-to-cradle system boundary for LCA of
This was motivated by the fact that cement-based recyclable concrete, the recycling of concrete for clin-
composites are often associated with huge negative ker production can reduce the impacts by 8–15% in
environmental impacts: the production of Portland comparison to traditional concrete.
cement (PC) only, can lead to 2 billion tons/year of As mentioned above, service life and durability are
CO2 emissions (approximately equal to 8-10% of glo- crucial in LCA studies, but, unlike ordinary concrete,
bal anthropogenic emissions) [360–363]. Therefore, there is little literature on this subject. Caruso et al.
considering the consistent annual production of con- [374] and di Summa et al. [375] investigated the
crete, higher than 10 billion tons globally [364–367], effect of the extended durability with a resulting ser-
more sustainable solutions in this field can help to vice life of at least 30% longer on LCA output, devel-
reduce negative environmental impacts related on a oping a cradle-to-grave analysis for innovative
global, regional and local scale [368]. The cited studies cementitious materials exposed to aggressive ambi-
demonstrated that both the cement content and the ent conditions. Here the authors investigated, in the
extended service life had a significant impact on the construction of a basin for geothermal water settling,
overall environmental performance. More specifically, a high-performance fibre-reinforced cementitious
36–43% and 36–38% of carbon footprint and energy composite with high content of slag (50% replace-
consumption, respectively were avoided in the study ment by volume of cement) and 1.5% by volume of
developed by Nath et al. [50] by replacing 40% of steel fibres (to obtain a strain hardening tensile
cement by fly ash and increasing the service life 1.6– behaviour) and autogenous healing stimulated by
1.75 times. Robayo-Salazar et al. [356] used a natural crystalline admixtures (CA). They estimated a
volcanic pozzolan and granulated blast furnace slag, reduction of 71 and 87% for abiotic depletion and
reaching 45% carbon footprint reduction. Therefore, acidification, respectively. Similarly, Van den Heede
LCA can also prove that self-healing concrete can et al. [53] highlighted that the use of healing agents,
achieve significant benefits from an environmental together with PVA microfibres, allowed to save
point of view because of the resulting extended service approximately 80% of global warming potential
life (SL) [225]. (GWP) if compared to Portland cement (PC) con-
The LCA procedure is outlined by ISO 14040:2006, crete due to the extended service life equal to 60
and 14044:2006 standards that identify four major years (35 years longer than PC concrete).
steps: (i) definition of goal and scope (choosing a Similarly, the recent results of di Summa et al. [353]
representative functional unit and considering an are inline with what is mentioned above. More specifi-
appropriate system boundary), (ii) inventory analysis, cally, they demonstrated that in a timeframe of 100
(iii) impact analysis and (iv) interpretation. A proper years of service life, the concrete containing superab-
functional unit (FU) choice, as pointed out by Panesar sorbent polymers (SAPs), favouring both internal cur-
et al. [49] and Van den Heede et al. [368], being the ing and self-healing, performed better than the
basis for estimating input and outputs, is mandatory conventional solution. This was mainly due to the
to obtain reliable LCA results. Gursel et al. [369] high- reduced frequency of the maintenance activities
lighted the necessity to include most of the concrete ensured by the innovative technology, and it allowed
 INTERNATIONAL MATERIALS REVIEWS 587

Figure 14. Possible LCA system boundaries for self-healing concrete.

reductions up to 50% for some impact indicators. Additionally, Hafez et al. [377] underlined that the
Moreover, the study outlined the importance of uncertainties of LCA are also due to the fact that there
using accurate durability parameters to get more are no standards to get LCI data. Therefore, they rec-
reliable LCA and life cycle cost (LCC) outputs. In ommend following the suggestions by Anand et al.
this case, the value taken into account for the apparent [378] to use either primary sources (like environ-
chloride diffusion coefficient, Dapp, had a major effect mental product declarations or laboratory results),
on the time of corrosion initiation of the reinforce- when possible, or secondary data from accredited
ment bars. This largely affected the forecast of the fre- environmental databases like Ecoinvent, Gabi and
quency of maintenance activities with the consequent ELC. In addition to accurate data input, a representa-
environmental and cost implications. tive impact analysis is essential. Jolliet et al. [379]
A cradle-to-gate system boundary was used by Van defined the impact analysis as the touch point between
den Heede et al. [52] for LCA of self-healing concrete LCI results and the corresponding environmental
containing different kinds of superabsorbent poly- impacts that can be divided into two main impact
mers. Here, the authors stressed two critical points, categories:
explaining why some studies are limited to the ‘gate’
stage: (i) the limited knowledge regarding the service (i) the first one, defined as problem-oriented,
life when exposed to a specific environmental con- regroups specific environmental problems into
dition and (ii) the lack of data regarding the life specific midpoint categories, like the impact on
cycle inventory (LCI). Because of the absence of data climate change expressed in kilograms of CO2
related to the healing agent (due to confidentiality equivalents (e.g. CML 2002).
issues of the manufacturer), the authors combined (ii) the second one, defined as damage oriented, tries
data from the Ecoinvent database with information to estimate the actual environmental damage,
reported in the literature. Furthermore, being aware sometimes with high uncertainty (e.g. Ecoindica-
that the development of new materials in the labora- tor 99).
tory (as for most of the healing agents currently
used) is not an energy-efficient process, they adjusted The damage-oriented impact category promotes an
the LCI input by adopting the proposal by Piccinno easier to understand interpretation of LCA, but, as
et al. [376] to go from the laboratory to the industrial described by Benetto et al. [380], the problem-related
scale. The same challenges were faced by di Summa approach is more reliable.
et al. [353], who also adopted a previously mentioned All of the available literature for LCA studies for
cradle-to-gate system boundary in the study. This was self-healing concrete uses a problem-related approach
mainly due to the lack of information regarding the assessing up to 10 indicators in total (as exemplified in
disposal scenario of the innovative composite. Figure 15): global warming (GWP); acidification (AP);
Moreover, even in this case, most of the LCIs have eutrophication (EP); ozone depletion (ODP); photo-
been redrafted according to existing scientific litera- chemical oxidation (POCP); abiotic depletion poten-
ture or to the environmental product declarations tial (ADP); human toxicity potential (HTP);
(EPDs). freshwater aquatic ecotoxicity potential (FAETP);
588 V. CAPPELLESSO ET AL.

Figure 15. Comparison of the reduction of impacts among the available literature about LCA of self-healing concrete. (1) concerns
concrete composites with 1 mass% of SAP and 2 vol% of PP microfiber, taking into account a FU with 100 years of service life [52];
(2) similarly, but for self-healing concrete with polyurethane [225]; (3) concerns concrete with crystalline admixture using a cradle
to gate system boundary [54]; (4) concerns a concrete composite with 1 mass% of SAP and 2 vol% of PVA microfiber assuming 60
years of service life [53]. * data are obtained from ‘Figure 2’ and ‘Figure 4’ of the corresponding papers, respectively, as the exact
value is not reported.

marine aquatic ecotoxicity (MAETP) and terrestrial the case in the study by Van Belleghem et al. [225],
ecotoxicity potential (TETP). All the studies estimate where the same indicators were reduced by more
promising results for self-healing concrete from the than 50% because of the extended durability of con-
environmental point of view. The investigation by crete healed by encapsulated polyurethane.
Rigamonti et al. [54], which used a cradle-to-gate sys- Due to the uncertainties associated with data qual-
tem boundary, cast a light on the better performance ity, some studies highlighted the necessity to process
in terms of environmental impacts of self-healing con- further the results obtained by LCA analysis using a
crete containing a crystalline admixture as an autoge- stochastic modelling approach, known as Monte
nous healing stimulator. In fact, impact indicators like Carlo simulation, to estimate the uncertainty and
AP, ODP, EP and POCP were reduced. This was also communicate the results in a probabilistic way [50].
 INTERNATIONAL MATERIALS REVIEWS 589

Even in Van den Heede et al. [53], considering the Durability indicators, both general and specifically
uncertainties related to maintenance and durability, related to a specific degradation mechanism, should
the authors assumed a standard triangular error distri- be determined to ensure the efficiency of smart con-
bution that confirmed the sustainability of self-healing crete developed with self-healing ability. This is
concrete with SAP (1 mass%/kg cement) and PVA needed to incorporate self-healing into durability-
microfibre (2 vol%). based structural design approaches consistently. In
Finally, regarding the Life Cycle Cost Assessment this respect, it is of the utmost importance to focus
for self-healing concrete, only very few studies on standardizing tests capable of measuring the
assessed the economic advantages of self-healing con- efficiency of self-healing mechanisms and quantifying
crete compared to a conventional solution. In a work their effects on the material mechanical and durability
focused on the use of shape memory polymers to performance indicators. Artificial Intelligence may
favour the autogenous healing, Teall already addressed support this purpose, as currently done in several
the potential costs reduction within a service life of other fields of civil engineering, to ‘extract’ design-
120 years due to the reduced maintenance activities oriented knowledge from widely flourishing literature,
[381]. Inline with this, two other recent studies have the information from which must be screened and
highlighted the influence of the repairing activities guided to the goal.
throughout the anticipated service life T, the cost of It can be surely recognized that the incorporation of
which for both the conventional and the self-healing self-healing into a truly performance-based design is
solution, complying with ISO 16627:2015, has to be likely to bring economic benefits in terms, e.g. of
‘actualized’ through the annual real discount rate r reduction or optimization of concrete covers or
according to the following equation [353,382]. shrinkage reinforcement thanks to crack healing and
reduced values of permeability to aggressive agents.
(T) = 1/(1 + r)T Furthermore, the actual benefits of self-healing, mostly
This highlights the need to further analyse these related to the extension of maintenance-free lifespan
innovative materials from an economic point of and the reduced frequency of maintenance over the
view, the reason why Caruso et al. [383], Gursel intended service life, have to be assessed in the frame-
et al. [369] and Colangelo et al. [358] supported the work of a life-cycle-based approach. Due to the large
idea of assessing thoroughly social and economic per- variability in the nature and costs of the aforemen-
formances to achieve a more holistic sustainable tioned maintenance operations, very few studies
approach combining the environmental implications have assessed the economic benefits of using self-heal-
with the ones that these composites can have in the ing concrete with reference to suitably defined ‘func-
society. tional units’ (structures or structural elements). To
this aim, a closer collaboration between all the players
of the (concrete) construction industry has to be
sought, from researchers to major stakeholders, for a
Conclusions
precise setting of the scope and goals. Efforts should
Self-healing is a promising solution to make concrete range from standardization of test methods to formu-
structures more durable. The aim is that cracks heal lation and validation of design methodologies and life-
without intervention, increasing the service life and cycle assessment criteria to promote the widespread
decreasing resource use, contributing to reducing use of self-healing concrete in practice.
environmental impacts and overall maintenance
costs. Many self-healing mechanisms have been inves-
tigated and are available, but in real conditions, the Disclosure statement
most suitable one should be selected for the specific No potential conflict of interest was reported by the author(s).
situation. Under ideal conditions, the healing
efficiency of different self-healing systems has been
widely studied. Actually, the ‘ideal condition’ depends Funding
on the self-healing mechanism; however, it mainly This project has received funding from
refers to unstressed specimens and young specimens the European Union’s Horizon 2020
Research and Innovation Programme
subjected to a single cracking/healing event, with the under the Marie Sklodowska-Curie
availability of water and at room temperature. The [grant agreement No 860006].
results obtained in this way may not be transferrable
to real structural service scenarios, imposing limit-
ations on their widespread use in current routine con- ORCID
struction practice. Bridging the gap to move from ideal Vanessa Cappellesso http://orcid.org/0000-0002-3886-6884
laboratory conditions to realistic field exposure should Davide di Summa http://orcid.org/0000-0002-2881-0801
now be the focus. Pardis Pourhaji http://orcid.org/0000-0002-8548-8302
590 V. CAPPELLESSO ET AL.

Niranjan Prabhu Kannikachalam http://orcid.org/0000- Innovation Programme under Grant Agreement No

0003-4475-4800 760824; 2018.
Kiran Dabral http://orcid.org/0000-0002-7907-2108 [16] SMARTINCS, Self-Healing – Multifunctional –
Liberato Ferrara http://orcid.org/0000-0002-6826-9917 Advanced Repair Technologies in Cementitious
Maria Cruz Alonso http://orcid.org/0000-0002-4308- Systems, European Union’s Framework Programme
9506 for Research and Innovation Programme under
Esteban Camacho http://orcid.org/0000-0001-9360-3043 Marie Skłodowska-Curie Grant Agreement
Elke Gruyaert http://orcid.org/0000-0003-0117-2544 SMARTINCS No 860006 Priv. Policy; 2019.
Nele De Belie http://orcid.org/0000-0002-0851-6242 Available from: https://smartincs.ugent.be/index.
php/about-us.
[17] De Belie N, Gruyaert E, Al-Tabbaa A, et al. A review
of self-healing concrete for damage management of
References structures. Adv Mater Interfaces. 2018;5:1–28.
[1] Van Belleghem B, De Belie N, Dewanckele J, et al. doi:10.1002/admi.201800074.
Analysis and visualization of water uptake in cracked [18] Ferrara L, Van Mullem T, Alonso MC, et al.
and healed mortar by water absorption tests and Experimental characterization of the self-healing
X-ray radiography. In: Concrete repair, rehabilita- capacity of cement based materials and its effects
tion and retrofitting IV. CRC Press, 2015. p. 12–13. on the material performance: a state of the art report
doi:10.1201/b18972-8. by COST action SARCOS WG2. Constr Build Mater.
[2] Dry C. Smart multiphase composite materials which 2018;167:115–142. doi:10.1016/j.conbuildmat.2018.
repair themselves by a release of liquids which 01.143.
become solids. Proceedings of Symposium of the [19] Hermawan H, Minne P, Serna P, et al.
Smart Structures and Materials; 1994 Oct 10–12; Understanding the impacts of healing agents on the
Orlando (FL). p. 62–70. properties of fresh and hardened self-healing con-
[3] Dry CM. Repair and prevention of damage due to crete: a review. Processes. 2021;9:2206. doi:10.3390/
transverse shrinkage cracks in bridge decks. Smart pr9122206.
Struct Mater. 1999;3671:253–256. doi:10.1117/12. [20] Gupta S, Al-Obaidi S, Ferrara L. Meta-analysis and
348675. machine learning models to optimize the efficiency
[4] Dry C. Procedures developed for self-repair of poly- of self-healing capacity of cementitious material.
meric matrix composite materials. Compos Struct. Materials. 2021;14:4437. doi:10.3390/ma14164437.
1996;35:263–269. [21] Jefferson T, Javierre E, Freeman B, et al. Research
[5] White SR, Sottos NR, Geubelle PH, et al. Autonomic progress on numerical models for self-healing
healing of polymer composites. Nature. cementitious materials. Adv Mater Interfaces.
2001;409:794–797. doi:10.1038/35057232. 2018;5:1701378. doi:10.1002/admi.201701378.
[6] T.D. SHM Program. Self-healing materials. 2005. [22] De Belie N, Van Belleghem B, Erşan YÇ, et al.
[7] SIM-SHE. SIM-program SHE: ‘self-healing Durability of self-healing concrete. MATEC Web
materials’. 8 projects including, Mater. Durable Conf. 2019;289:01003. doi:10.1051/matecconf/
Sustain. Struct. (Devan), SIM-FLANDERS; 2010. 201928901003.
[8] HEALCON. Self-healing concrete to create durable [23] Sisomphon K, Copuroglu O, Koenders EAB. Self-
and sustainable concrete structures. University of healing of surface cracks in mortars with expansive
Gent, Grant Agreement ID 309451; 2013. additive and crystalline additive. Cem Concr
[9] M4L. Materials for life (M4L): biomimetic multi- Compos. 2012;34:566–574. doi:10.1016/j.
scale damage immunity for construction materials, cemconcomp.2012.01.005 .
Cardiff University. EPSRC (Engineering and [24] Ferrara L, Krelani V, Carsana M. A “fracture testing”
Physical Sciences Research Council). EP/K026631/ based approach to assess crack healing of concrete
1; 2013. with and without crystalline admixtures. Constr
[10] CAPDESIGN. Encapsulation of polymeric healing Build Mater. 2014;68:535–551. doi:10.1016/j.
agents in self-healing concrete: capsule design. conbuildmat.2014.07.008 .
Belie, Nele; 2014. [25] Roig-Flores M, Pirritano F, Serna P, et al. Effect of
[11] LORCENIS. LOng Lasting Reinforced Concrete for crystalline admixtures on the self-healing capability
ENergy Infrastructure under severe operating con- of early-age concrete studied by means of per-
ditions. European Union’s Framework Programme meability and crack closing tests. Constr Build
for Research and Innovation Programme under Mater. 2016;114:447–457. doi:10.1016/j.
Grant Agreement No 685445; 2016. conbuildmat.2016.03.196 .
[12] SARCOS. Self-healing As prevention Repair of [26] Borg RP, Cuenca E, Gastaldo Brac EM, et al. Crack
COncrete Structures, COST (European Cooperation sealing capacity in chloride-rich environments of
in Science and Technology). CA152020; 2016. mortars containing different cement substitutes and
[13] RM4L. Resilient Materials 4 Life (RM4L), Cardiff crystalline admixtures. J Sustain Cem Mater.
University. EPSRC – Engineering and Physical 2018;7:141–159. doi:10.1080/21650373.2017.
Sciences Research Council; 2017. 1411297.
[14] EndurCrete. New environmental friendly and dur- [27] Cuenca E, Tejedor A, Ferrara L. A methodology to
able concrete. European Union’s Framework assess crack-sealing effectiveness of crystalline
Programme for Research and Innovation admixtures under repeated cracking-healing cycles.
Programme under Grant Agreement. No 760639; Constr Build Mater. 2018;179:619–632. doi:10.1016/
2018. j.conbuildmat.2018.05.261.
[15] RESHEALIENCE. RESHEALIENCE. European [28] Cuenca E, Rigamonti S, Gastaldo Brac E, et al.
Union’s Framework Programme for Research and Crystalline admixture as healing promoter in
 INTERNATIONAL MATERIALS REVIEWS 591

concrete exposed to chloride-rich environments: [42] Van Tittelboom K, De Belie N. Self-healing in

experimental study. J Mater Civ Eng. cementitious materials – a review. Materials.
2021;33:04020491. doi:10.1061/(ASCE)MT.1943- 2013;6:2182–2217. doi:10.3390/ma6062182.
5533.0003604. [43] Xiong W, Tang J, Zhu G, et al. A novel capsule-based
[29] Ait Ouarabi M, Antonaci P, Boubenider F, et al. self-recovery system with a chloride ion trigger. Sci
Ultrasonic monitoring of the interaction between Rep. 2015;5:10866. doi:10.1038/srep10866.
cement matrix and alkaline silicate solution in self- [44] Cuenca E, Brac EMG, Rigamonti S, et al. Self-healing
healing systems. Materials. 2017;10:46. doi:10.3390/ stimulated by crystalline admixtures in chloride rich
ma10010046. environments: is it possible to extend the structure
[30] Gliozzi AS, Scalerandi M, Anglani G, et al. service life ? ReSHEALience RILEM Work; 2019.
Correlation of elastic and mechanical properties of [45] Paine K, Sharma T, Alazhari M, et al. Application
consolidated granular media during microstructure and performance of bacteria-based self-healing con-
evolution induced by damage and repair. Phys Rev crete Final Conference of RILEM TC 253-MCI:
Mater. 2018;2:013601. doi:10.1103/ Microorganisms-Cementitious Materials Interactions;
PhysRevMaterials.2.013601 . Toulouse, France; 2018. p. 387–394.
[31] Anglani G, Tulliani J-M, Antonaci P. Behaviour of [46] Araújo M, Chatrabhuti S, Gurdebeke S, et al. Poly
pre-cracked self-healing cementitious materials (methyl methacrylate) capsules as an alternative to
under static and cyclic loading. Materials. the proof-of-concept glass capsules used in self-heal-
2020;13:1149. doi:10.3390/MA13051149. ing concrete. Cem Concr Compos. 2018;89:260–271.
[32] Van Belleghem B, Gruyaert E, Van Tittelboom K, doi:10.1016/j.cemconcomp.2018.02.015.
et al. Effect of polyurethane viscosity on self-healing [47] De Belie N, Van Mullem T, Gruyaert E, et al. Self-
efficiency of cementitious materials exposed to high healing concrete applied in a roof plate for an inspec-
temperatures from sun radiation. J Mater Civ Eng. tion pit in the oosterweel link: healing efficiency and
2018;30:04018145. doi:10.1061/(ASCE)MT.1943- preliminary service life estimation. 4th International
5533.0002360. conference on Service Life Design for Infrastructures
[33] Wang J, Erşan YC, Boon N, et al. Application of – 27–30 August 2018; Delft, Netherlands, 2018.
microorganisms in concrete: a promising sustainable p. 429–436.
strategy to improve concrete durability. Appl [48] Erşan YÇ, Gruyaert E, Louis G, et al. Self-protected
Microbiol Biotechnol. 2016;100:2993–3007. doi:10. nitrate reducing culture for intrinsic repair of con-
1007/s00253-016-7370-6. crete cracks. Front Microbiol. 2015;6. doi:10.3389/
[34] Davies R, Teall O, Pilegis M, et al. Large scale appli- fmicb.2015.01228.
cation of self-healing concrete: design, construction, [49] Panesar DK, Seto KE, Churchill CJ. Impact of the
and testing. Front Mater. 2018;5:1–12. doi:10.3389/ selection of functional unit on the life cycle assess-
fmats.2018.00051. ment of green concrete. Int J Life Cycle Assess.
[35] Sierra-Beltran MG, Jonkers HM, Ortiz M. Field 2017;22:1969–1986. doi:10.1007/s11367-017-1284-0.
application of self-healing concrete with natural [50] Nath P, Sarker PK, Biswas WK. Effect of fly ash on
fibres as linings for irrigation canals in the service life, carbon footprint and embodied
Ecuador. Fifth International Conference on Self- energy of high strength concrete in the marine
Healing Materials; Durham, NC, 2015. Vol. 32, environment. Energy Build. 2018;158:1694–1702.
p. 1–4. doi:10.1016/j.enbuild.2017.12.011.
[36] Van Tittelboom K, Wang J, Araújo M, et al. [51] Cadenazzi T, Dotelli G, Rossini M, et al. Cost and
Comparison of different approaches for self-healing environmental analyses of reinforcement alternatives
concrete in a large-scale lab test. Constr Build for a concrete bridge. Struct Infrastruct Eng.
Mater. 2016;107:125–137. doi:10.1016/j. 2020;16:787–802. doi:10.1080/15732479.2019.
conbuildmat.2015.12.186. 1662066.
[37] Al-Tabbaa A, Litina C, Giannaros P, et al. First UK [52] Van den Heede P, Mignon A, Habert G, et al. Cradle-
field application and performance of microcapsule- to-gate life cycle assessment of self-healing engin-
based self-healing concrete. Constr Build Mater. eered cementitious composite with in-house devel-
2019;208:669–685. doi:10.1016/j.conbuildmat.2019. oped (semi-)synthetic superabsorbent polymers.
02.178. Cem Concr Compos. 2018;94:166–180. doi:10.1016/
[38] Selvarajoo T, Davies RE, Gardner DR, et al. j.cemconcomp.2018.08.017.
Characterisation of a vascular self-healing cementi- [53] Van den Heede P, De Belie N, Pittau F, et al. Life
tious material system: flow and curing properties. cycle assessment of self-healing engineered cementi-
Constr Build Mater. 2020;245:118332. doi:10.1016/j. tious composite (SH-ECC) used for the rehabilitation
conbuildmat.2020.118332. of bridges. Life Cycle Analysis and Assessment in
[39] Selvarajoo T, Davies R, Freeman B, et al. Mechanical Civil Engineering: Towards an Integrated Vision
response of a vascular self-healing cementitious Proceedings of the Sixth International Symposium
material system under varying loading conditions. on Life-Cycle Civil Engineering, IALCCE 2018,
Constr Build Mater. 2020;254:119245. doi:10.1016/j. 2019. p. 2269–2275.
conbuildmat.2020.119245. [54] Rigamonti S, Cuenca E, Arrigoni A, et al. Self-healing
[40] Van Mullem T, Gruyaert E, Caspeele R, et al. First concrete vs. conventional waterproofing systems in
large scale application with self-healing concrete in underground structures: a cradle to gate LCA com-
Belgium: analysis of the laboratory control tests. parison with reference to a case study. Life Cycle
Materials. 2020;13:997. doi:10.3390/ma13040997. Analysis and Assessment in Civil Engineering:
[41] Erşan YÇ. Microbial nitrate reduction induced Towards an Integrated Vision Proceedings of the
autonomous self-healing in concrete [PhD thesis]. Sixth International Symposium on Life-Cycle Civil
Belgium: Ghent University; 2016. Engineering, IALCCE 2018, 2019. p. 2277–2284.
592 V. CAPPELLESSO ET AL.

[55] Eurocode 2, Eurocode 2: design of concrete struc- marine environments. Cem Concr Compos.
tures – part 1-1: general rules and rules for buildings; 2008;30:113–121. doi:10.1016/j.cemconcomp.2007.
2011. 09.005.
[56] fib international federation, fib Model Code; 2020. [76] Alonso C, Andrade C, Rodriguez J, et al. Factors con-
[57] Neville A. Autogenous healing – a concrete miracle? trolling cracking of concrete affected by reinforce-
Concr Int. 2002;24:76–82. ment corrosion. Mater Struct. 1998;31:435–441.
[58] Wang Y-J, Wu Z-M, Qu F-M, et al. Numerical inves- [77] Francois R, Maso JC. Effect of damage in reinforced
tigation on crack propagation process of concrete concrete on carbonation or chloride penetration.
gravity dams under static and dynamic loads with Cem Concr Res. 1988;18:961–970. doi:10.1016/
in-crack reservoir pressure. Theor Appl Fract 0008-8846(88)90033-6.
Mech. 2022;117:103221. doi:10.1016/j.tafmec.2021. [78] Van Belleghem B, Villagrán Zaccardi Y, Van den
103221. Heede P, et al. Evaluation and comparison of tra-
[59] Cuenca E, Mezzena A, Ferrara L. Synergy between crys- ditional methods and electron probe micro analysis
talline admixtures and nano-constituents in enhancing (EPMA) to determine the chloride ingress perpen-
autogenous healing capacity of cementitious compo- dicular to cracks in self-healing concrete. Constr
sites under cracking and healing cycles in aggressive Build Mater. 2019;227:116789. doi:10.1016/j.
waters. Constr Build Mater. 2021;266:121447. doi:10. conbuildmat.2019.116789.
1016/j.conbuildmat.2020.121447. [79] Konin A, Franfois R, Arliguie G. Penetration of
[60] European Committee for Standardization. BS EN chlorides in relation to the microcracking state into
206:2013+A1:2016 concrete. Specification, perform- reinforced ordinary and high strength concrete.
ance, production and conformity. Brussels: Mater Struct. 1998;31:310–316.
European Committee for Standardization; 2016. [80] Pease BJ. Influence of concrete cracking on ingress
[61] Beushausen H, Luco LF. Performance-based specifi- and reinforcement corrosion. Lyngby (Denmark):
cations and conctrol of concrete durability: state-of- University of Denmark (UK); 2010.
art report TC 230-PSC; 2015. [81] Mohammed TU, Otsuki N, Hamada H. Corrosion of
[62] Vichot A, Ollivier J-P. La durabilité des bétons: Bases steel bars in cracked concrete under marine environ-
scientifiques pour la formulation de bétons durables ment. J Mater Civ Eng. 2003;15:460–469. doi:10.
dans leur environnement; 2008. 1061/(ASCE)0899-1561(2003)15:5(460).
[63] ACI Committee 201, ACI 201.2R-16. Guide to [82] Marcos-Meson V, Geiker M, Fischer G, et al.
Durable Concrete, 2016. Durability of cracked SFRC exposed to wet-dry cycles
[64] Panesar DK. Supplementary cementing materials. In: of chlorides and carbon dioxide – multiscale deterio-
Developments in the formulation and reinforcement ration phenomena. Cem Concr Res.
of concrete. Elsevier; 2019. p. 55–85. doi:10.1016/ 2020;135:106120. doi:10.1016/j.cemconres.2020.
B978-0-08-102616-8.00003-4. 106120.
[65] Tuutti K. Corrosion of steel in concrete. Stockholm: [83] Win PP, Watanabe M, Machida A. Penetration
Swedish Cement and Concrete Research Institute; profile of chloride ion in cracked reinforced concrete.
1982. Cem Concr Res. 2004;34:1073–1079. doi:10.1016/j.
[66] fib Bulletin 34. Model code for service life design, fib. cemconres.2003.11.020.
The International Federation for Structural [84] Sangoju B, Gettu R, Bharatkumar BH, et al. Chloride-
Concrete; 2006. doi:10.35789/fib.BULL.0034. induced corrosion of steel in cracked OPC and PPC
[67] FIB Bulleting 53. Structural concrete, textbook on concretes: experimental study. J Mater Civ Eng.
behavior, design and performance, 2nd ed. Volume 2011;23:1057–1066. doi:10.1061/(ASCE)MT.1943-
3. 2009. 5533.0000260.
[68] EHE-08. Code on structural concrete. 2008. [85] Russo N, Gastaldi M, Marras P, et al. Effects of load-
[69] Spanish structural code. BOE 190. Spanish structural induced micro-cracks on chloride penetration resist-
code. 2021. ance in different types of concrete. Mater Struct.
[70] Alonso MC, Sanchez M. Analysis of the variability of 2020;53:143. doi:10.1617/s11527-020-01580-y.
chloride threshold values in the literature. Mater [86] Maes M. Combined effects of chlorides and
Corros. 2009;60:631–637. doi:10.1002/maco. sulphates on cracked and self-healing concrete in
200905296. marine environments. Ghent: Ghent University;
[71] Angst U, Elsener B, Larsen CK, et al. Critical chloride 2015.
content in reinforced concrete – a review. Cem [87] Van den Heede P, Maes M, De Belie N. Influence of
Concr Res. 2009;39:1122–1138. doi:10.1016/j. active crack width control on the chloride pen-
cemconres.2009.08.006. etration resistance and global warming potential of
[72] Angst UM, Geiker MR, Michel A, et al. The steel– slabs made with fly ash+silica fume concrete.
concrete interface. Mater Struct. 2017;50:143. Constr Build Mater. 2014;67:74–80. doi:10.1016/j.
doi:10.1617/s11527-017-1010-1. conbuildmat.2013.10.032.
[73] Angst UM, Geiker MR, Alonso MC, et al. The effect [88] Djerbi A, Bonnet S, Khelidj A, et al. Influence of tra-
of the steel–concrete interface on chloride-induced versing crack on chloride diffusion into concrete.
corrosion initiation in concrete: a critical review by Cem Concr Res. 2008;38:877–883. doi:10.1016/j.
RILEM TC 262-SCI. Mater Struct. 2019;52:88. cemconres.2007.10.007.
doi:10.1617/s11527-019-1387-0. [89] Kušter Marić M, Ožbolt J, Balabanić G. Reinforced
[74] Bertolini L, Elsener B, Pedeferri P, et al. Corrosion of concrete bridge exposed to extreme maritime
steel in concrete. Wiley; 2003. doi:10.1002/ environmental conditions and mechanical damage:
3527603379. measurements and numerical simulation. Eng
[75] Song H-W, Lee C-H, Ann KY. Factors influencing Struct. 2020;205:110078. doi:10.1016/j.engstruct.
chloride transport in concrete structures exposed to 2019.110078.
 INTERNATIONAL MATERIALS REVIEWS 593

[90] Gowripalan N, Sirivivatnanon V, Lim CC. Chloride [104] Nehdi ML, Suleiman AR, Soliman AM. Investigation
diffusivity of concrete cracked in flexure. Cem of concrete exposed to dual sulfate attack. Cem
Concr Res. 2000;30:725–730. doi:10.1016/S0008- Concr Res. 2014;64:42–53. doi:10.1016/j.cemconres.
8846(00)00216-7. 2014.06.002.
[91] Pettersson K, Jorgenson O, FidjestolI P. The effect of [105] Zhao G, Li J, Shi M, et al. Degradation mechanisms of
cracks on reinforcement corrosion in high-perform- cast-in-situ concrete subjected to internal-external
ance concrete in a marine environment. Spec Publ. combined sulfate attack. Constr Build Mater.
1996;163:185–200. 2020;248:118683. doi:10.1016/j.conbuildmat.2020.
[92] Bertolini L, Elsener B, Pedeferri P, et al. Related titles 118683.
from Wiley-VCH DECHEMA corrosion handbook : [106] He P, Yu J, Wang R, et al. Effect of ion chelator on
corrosive agents and their interaction with materials; pore structure, mechanical property and self-healing
2004. capability of seawater exposed mortar. Constr Build
[93] Farnam Y, Esmaeeli HS, Zavattieri PD, et al. Mater. 2020;246:118480. doi:10.1016/j.conbuildmat.
Incorporating phase change materials in concrete 2020.118480.
pavement to melt snow and ice. Cem Concr [107] Bassuoni MT, Rahman MM. Response of concrete to
Compos. 2017;84:134–145. doi:10.1016/j. accelerated physical salt attack exposure. Cem Concr
cemconcomp.2017.09.002. Res. 2016;79:395–408. doi:10.1016/j.cemconres.2015.
[94] von Greve-Dierfeld S, Lothenbach B, Vollpracht A, 02.006.
et al. Understanding the carbonation of concrete [108] Mehta PJM, Monteiro PK. Concrete: microstructure,
with supplementary cementitious materials: a critical properties, and materials. 4th ed. New York:
review by RILEM TC 281-CCC. Mater Struct. McGraw-Hill Education; 2014.
2020;53:136. doi:10.1617/s11527-020-01558-w. [109] Soive A, Tran VQ. External sulfate attack of cemen-
[95] Alahmad S, Toumi A, Verdier J, et al. Effect of crack titious materials: new insights gained through
opening on carbon dioxide penetration in cracked numerical modeling including dissolution/precipi-
mortar samples. Mater Struct. 2009;42:559–566. tation kinetics and surface complexation. Cem
doi:10.1617/s11527-008-9402-x. Concr Compos. 2017;83:263–272. doi:10.1016/j.
[96] Van Mullem T, De Meyst L, Handoyo JP, et al. cemconcomp.2017.07.024.
Influence of crack geometry and crack width on car- [110] Santhanam M, Cohen MD, Olek J. Sulfate attack
bonation of high-volume fly ash (HVFA) mortar. In: research – whither now? Cem Concr Res.
International RILEM Conference on Ambitioning a 2001;31:845–851. doi:10.1016/S0008-8846
Sustainable Future for Built Environment: (01)00510-5.
Comprehensive Strategies for Unprecedented [111] Santhanam M, Cohen MD, Olek J. Mechanism of sul-
Challenge; Guimarães, Portugal; 2020. p. 1–8. fate attack: a fresh look. Cem Concr Res.
[97] Han J, Liu W, Wang S, et al. Effects of crack and ITZ 2003;33:341–346. doi:10.1016/S0008-8846
and aggregate on carbonation penetration based on (02)00958-4.
3D micro X-ray CT microstructure evolution. [112] Idiart AE, López CM, Carol I. Chemo-mechanical
Constr Build Mater. 2016;128:256–271. doi:10.1016/ analysis of concrete cracking and degradation due
j.conbuildmat.2016.10.062. to external sulfate attack: a meso-scale model. Cem
[98] Bogas JA, Carriço A, Pontes J. Influence of cracking Concr Compos. 2011;33:411–423. doi:10.1016/j.
on the capillary absorption and carbonation of struc- cemconcomp.2010.12.001.
tural lightweight aggregate concrete. Cem Concr [113] Liu F, You Z, Diab A, et al. External sulfate attack on
Compos. 2019;104:103382. doi:10.1016/j. concrete under combined effects of flexural fatigue
cemconcomp.2019.103382. loading and drying-wetting cycles. Constr Build
[99] Carević V, Ignjatović I. Influence of loading cracks Mater. 2020;249:118224. doi:10.1016/j.conbuildmat.
on the carbonation resistance of RC elements. 2020.118224.
Constr Build Mater. 2019;227:116583. doi:10.1016/j. [114] Santhanam M, Cohen MD, Olek J. Mechanism of sul-
conbuildmat.2019.07.309. fate attack: a fresh look. Cem Concr Res. 2002;32:915–
[100] Saetta AV, Schrefler BA, Vitaliani RV. The carbona- 921. doi:10.1016/S0008-8846(02)00724-X.
tion of concrete and the mechanism of moisture, heat [115] Qiao X, Chen J. Correlation of propagation rate of
and carbon dioxide flow through porous materials. corrosive crack in concrete under sulfate attack and
Cem Concr Res. 1993;23:761–772. doi:10.1016/ growth rate of delayed ettringite. Eng Fract Mech.
0008-8846(93)90030-D. 2019;209:333–343. doi:10.1016/j.engfracmech.2019.
[101] Haynes H, O’Neill R, Neff M, et al. Salt weathering 01.036.
distress on concrete exposed to sodium sulfate [116] Meng C, Li W, Cai L, et al. Experimental research on
environment. ACI Mater J. 2008;105:35–43. durability of high-performance synthetic fibers
Available from: https://www.worldcat.org/title/aci- reinforced concrete: resistance to sulfate attack and
materials-journal/oclc/13846872. freezing-thawing. Constr Build Mater.
[102] Van Tittelboom K, De Belie N. A critical review on 2020;262:120055. doi:10.1016/j.conbuildmat.2020.
test methods for evaluating the resistance of concrete 120055.
against sulfate attack. International RILEM TC211- [117] Li Y, Wang R, Li S, et al. Resistance of recycled aggre-
PAE Final Conference on Concrete in Aggressive gate concrete containing low- and high-volume fly
Aqueous Environments: Performance, Testing and ash against the combined action of freeze–thaw
Modeling, Rilem Publications - PRO 63, Bagneux, cycles and sulfate attack. Constr Build Mater.
France, 2009. p. 298–306. 2018;166:23–34. doi:10.1016/j.conbuildmat.2018.01.
[103] Alexander M, Bertron A, De Belie N. Performance of 084.
cement-based materials in aggressive aqueous [118] Gong F, Takahashi Y, Segawa I, et al. Mechanical
environments; 2013. doi:10.1007/978-94-007-5413-3. properties of concrete with smeared cracking by
594 V. CAPPELLESSO ET AL.

alkali-silica reaction and freeze-thaw cycles. Cem [131] Lv L-S, Wang J-Y, Xiao R-C, et al. Chloride ion trans-
Concr Compos. 2020;111:103623. doi:10.1016/j. port properties in microcracked ultra-high perform-
cemconcomp.2020.103623. ance concrete in the marine environment. Constr
[119] Liu Q, Hu Z, Wang X, et al. Numerical study on Build Mater. 2021;291:123310. doi:10.1016/j.
cracking and its effect on chloride transport in con- conbuildmat.2021.123310.
crete subjected to external load. Constr Build [132] Peng R, Qiu W, Teng F. Investigation on seawater
Mater. 2022;325:126797. doi:10.1016/j.conbuildmat. freeze-thaw damage deterioration of marine concrete
2022.126797. structures in cold regions from multi-scale. Ocean
[120] Yi Y, Zhu D, Guo S, et al. A review on the deterio- Eng. 2022;248:110867. doi:10.1016/j.oceaneng.2022.
ration and approaches to enhance the durability of 110867.
concrete in the marine environment. Cem Concr [133] Basheer L, Cleland DJ. Freeze–thaw resistance of
Compos. 2020;113:103695. doi:10.1016/j. concretes treated with pore liners. Constr Build
cemconcomp.2020.103695. Mater. 2006;20:990–998. doi:10.1016/j.conbuildmat.
[121] Zhao G, Li J, Shi M, et al. Degradation of cast-in-situ 2005.06.010.
concrete subjected to sulphate-chloride combined [134] Jacobsen S, Marchand J, Hormain H. SEM obser-
attack. Constr Build Mater. 2020;241:117995. vations of the microstructure of frost deteriorated
doi:10.1016/j.conbuildmat.2019.117995. and self-healed concretes. Cem Concr Res.
[122] Maes M, Gruyaert E, De Belie N. Resistance of con- 1995;25:1781–1790.
crete against combined attack of chlorides and sul- [135] Liu Z, Hansen W. Effect of hydrophobic surface
phate. Trondheim (Norway): International treatment on freeze-thaw durability of concrete.
Congress on Durability of Concrete; 2012. p. 14. Cem Concr Compos. 2016;69:49–60. doi:10.1016/j.
[123] Wang P, Mo R, Zhou X, et al. A chemo-thermo- cemconcomp.2016.03.001.
damage-transport model for concrete subjected to [136] Shields Y, Garboczi E, Weiss J, et al. Freeze-thaw
combined chloride-sulfate attack considering the crack determination in cementitious materials using
effect of calcium leaching. Constr Build Mater. 3D X-ray computed tomography and acoustic emis-
2021;306:124918. doi:10.1016/j.conbuildmat.2021. sion. Cem Concr Compos. 2018;89:120–129. doi:10.
124918. 1016/j.cemconcomp.2018.03.004.
[124] Sun H, Jiang C, Cao K, et al. Monitoring of steel cor- [137] Wang R, Hu Z, Li Y, et al. Review on the deterio-
rosion and cracking in cement paste exposed to com- ration and approaches to enhance the durability of
bined sulfate–chloride attack with X-ray concrete in the freeze–thaw environment. Constr
microtomography. Constr Build Mater. Build Mater. 2022;321:126371. doi:10.1016/j.
2021;302:124345. doi:10.1016/j.conbuildmat.2021. conbuildmat.2022.126371.
124345. [138] Nili M, Ramezanianpour AA, Sobhani J. Evaluation
[125] Wu J, Wei J, Huang H, et al. Effect of multiple ions on of the effects of silica fume and air-entrainment on
the degradation in concrete subjected to sulfate deicer salt scaling resistance of concrete pavements:
attack. Constr Build Mater. 2020;259:119846. microstructural study and modeling. Constr Build
doi:10.1016/j.conbuildmat.2020.119846. Mater. 2021;308:124972. doi:10.1016/j.conbuildmat.
[126] Ramezanianpour AA, Pourkhorshidi AR, Sobhani J, 2021.124972.
et al. Durability of concrete containing blended [139] Şahin Y, Akkaya Y, Taşdemir MA. Effects of freez-
cements in harsh marine environments: 18 years ing conditions on the frost resistance and micro-
exposure study. Constr Build Mater. structure of concrete. Constr Build Mater.
2021;299:123863. doi:10.1016/j.conbuildmat.2021. 2021;270:121458. doi:10.1016/j.conbuildmat.2020.
123863. 121458.
[127] Montes P, Bremner TW, Lister DH. Influence of cal- [140] Zhang P, Liu G, Pang C, et al. Influence of pore
cium nitrite inhibitor and crack width on corrosion structures on the frost resistance of concrete. Mag
of steel in high performance concrete subjected to a Concr Res. 2017;69:271–279. doi:10.1680/jmacr.15.
simulated marine environment. Cem Concr 00471.
Compos. 2004;26:243–253. doi:10.1016/S0958-9465 [141] Zhang P, Wittmann FH, Vogel M, et al. Influence of
(03)00043-X. freeze-thaw cycles on capillary absorption and chlor-
[128] Otieno M, Beushausen H, Alexander M. Chloride- ide penetration into concrete. Cem Concr Res.
induced corrosion of steel in cracked concrete – 2017;100:60–67. doi:10.1016/j.cemconres.2017.05.
part I: experimental studies under accelerated and 018.
natural marine environments. Cem Concr Res. [142] Pigeon M, Marchand J, Pleau R. Frost resistant con-
2016;79:373–385. doi:10.1016/j.cemconres.2015.08. crete. Constr Build Mater. 1996;10:339–348. doi:10.
009. 1016/0950-0618(95)00067-4.
[129] Lopez-Calvo HZ, Montes-García P, Jiménez-Quero [143] Valenza JJ, Scherer GW. A review of salt scaling:
VG, et al. Influence of crack width, cover depth and I. Phenomenology. Cem Concr Res. 2007;37:1007–
concrete quality on corrosion of steel in HPC con- 1021. doi:10.1016/j.cemconres.2007.03.005.
taining corrosion inhibiting admixtures and fly ash. [144] Strand MJ, Fridh K. Test method for de-icing salt-
Cem Concr Compos. 2018;88:200–210. doi:10.1016/ frost scaling in high-performance concrete.
j.cemconcomp.2018.01.016. MethodsX. 2018;5:1299–1310. doi:10.1016/j.mex.
[130] Du F, Jin Z, She W, et al. Chloride ions migration and 2018.10.007.
induced reinforcement corrosion in concrete with [145] Zeng W, Ding Y, Zhang Y, et al. Effect of steel fiber
cracks: a comparative study of current acceleration on the crack permeability evolution and crack surface
and natural marine exposure. Constr Build Mater. topography of concrete subjected to freeze-thaw
2020;263:120099. doi:10.1016/j.conbuildmat.2020. damage. Cem Concr Res. 2020;138:106230. doi:10.
120099. 1016/j.cemconres.2020.106230.
 INTERNATIONAL MATERIALS REVIEWS 595

[146] Wu K, Han H, Xu L, et al. The improvement of freez- Mater. 2011;25:2576–2583. doi:10.1016/j.

ing–thawing resistance of concrete by cellulose/poly- conbuildmat.2010.11.107.
vinyl alcohol hydrogel. Constr Build Mater. [161] Edvardsen C. Water permeability and autogenous
2021;291:123274. doi:10.1016/j.conbuildmat.2021. healing of cracks in concrete. ACI Mater J.
123274. 1999;96:448–454. doi:10.14359/645.
[147] Farnam Y, Dick S, Wiese A, et al. The influence of [162] Gruyaert E, van Tittelboom K, De Backer P, et al.
calcium chloride deicing salt on phase changes and Self-healing of thermal cracks in sandwich panels.
damage development in cementitious materials. ICSHM 2013: 4th International Conference on Self-
Cem Concr Compos. 2015;64:1–15. doi:10.1016/j. Healing Materials; Ghent, Belgium; 2013. p.196–200.
cemconcomp.2015.09.006. [163] Schneider U. Concrete at high temperatures – a gen-
[148] Al-Kheetan MJ, Rahman MM, Ghaffar SH, et al. eral review. Fire Saf J. 1988;13:55–68. doi:10.1016/
Comprehensive investigation of the long-term per- 0379-7112(88)90033-1.
formance of internally integrated concrete pavement [164] Ervine A, Gillie M, Stratfor TJ, et al. Thermal diffu-
with sodium acetate. Results Eng. 2020;6:100110. sivity of tensile cracked concrete. Proceedings of
doi:10.1016/j.rineng.2020.100110. Applications of Structural Fire Engineering, Prage,
[149] Esmaeeli HS, Farnam Y, Bentz DP, et al. Numerical 2011. p. 91–102.
simulation of the freeze–thaw behavior of mortar [165] Kwon SC, Dutta PK, Kim YH, et al. Comparison of
containing deicing salt solution. Mater Struct. the fatigue behaviors of FRP bridge decks and
2017;50:96. doi:10.1617/s11527-016-0964-8. reinforced concrete conventional decks under
[150] Ma H, Yu H, Da B, et al. Study on failure mechanism extreme environmental conditions. KSME Int J.
of concrete subjected to freeze-thaw condition in air- 2003;17:1–10. doi:10.1007/BF02984280.
port deicers. Constr Build Mater. 2021;313:125202. [166] Cuenca E, Cislaghi G, Puricelli M, et al. Influence of
doi:10.1016/j.conbuildmat.2021.125202. self-healing stimulated via crystalline admixtures on
[151] Zhang W, Pi Y, Kong W, et al. Influence of damage chloride penetration. American Concrete Institute,
degree on the degradation of concrete under freez- ACI Special Publication, DSCS2018, Moscow, 2018.
ing-thawing cycles. Constr Build Mater. p. 1–10.
2020;260:119903. doi:10.1016/j.conbuildmat.2020. [167] Cappellesso VG, dos S. Petry N, Longhi MA, et al.
119903. Reduction of concrete permeability using admixtures
[152] Li L, Liu Q, Tang L, et al. Chloride penetration in or surface treatments. J Build Pathol Rehabil.
freeze–thaw induced cracking concrete: a numerical 2022;7:38. doi:10.1007/s41024-022-00176-z.
study. Constr Build Mater. 2021;302:124291. doi:10. [168] Snoeck D, Dewanckele J, Cnudde V, et al. X-ray com-
1016/j.conbuildmat.2021.124291. puted microtomography to study autogenous healing
[153] Bhattacharjee SS, Léger P. Fracture response of grav- of cementitious materials promoted by superabsor-
ity dams due to rise of reservoir elevation. J Struct bent polymers. Cem Concr Compos. 2016;65:83–93.
Eng. 1995;121:1298–1305. doi:10.1061/(ASCE)0733- doi:10.1016/j.cemconcomp.2015.10.016.
9445(1995)121:9(1298). [169] Snoeck D, Van Tittelboom K, Steuperaert S, et al.
[154] Wang Y, Hu S, Fan X, et al. Effect of water pressure Self-healing cementitious materials by the combi-
on fracture parameters of concrete. Constr Build nation of microfibres and superabsorbent polymers.
Mater. 2019;199:613–623. doi:10.1016/j. J Intell Mater Syst Struct. 2014;25:13–24. doi:10.
conbuildmat.2018.12.075. 1177/1045389X12438623.
[155] Hooshmand A, Kianoush R, Siad H, et al. [170] Snoeck D, Velasco LF, Mignon A, et al. The effects of
Investigation of leakage and self-healing of direct superabsorbent polymers on the microstructure of
tension cracks under sustained loading and high- cementitious materials studied by means of sorption
water pressure. Constr Build Mater. experiments. Cem Concr Res. 2015;77:26–35. doi:10.
2021;267:120879. doi:10.1016/j.conbuildmat.2020. 1016/j.cemconres.2015.06.013.
120879. [171] Sisomphon K, Copuroglu O, Koenders EAB. Effect of
[156] Xiao L, Chen D, Jiang M, et al. Experimental and exposure conditions on self healing behavior of strain
numerical analysis of chloride transport in finite con- hardening cementitious composites incorporating
crete under reverse water pressure. Constr Build various cementitious materials. Constr Build Mater.
Mater. 2021;304:124576. doi:10.1016/j.conbuildmat. 2013;42:217–224. doi:10.1016/j.conbuildmat.2013.
2021.124576. 01.012.
[157] Brzozowski P, Horszczaruk E. Influence of surface [172] Roig-Flores M, Moscato S, Serna P, et al. Self-healing
preparation on adhesion of underwater repair con- capability of concrete with crystalline admixtures in
cretes under hydrostatic pressure. Constr Build different environments. Constr Build Mater.
Mater. 2021;310:125153. doi:10.1016/j.conbuildmat. 2015;86:1–11. doi:10.1016/j.conbuildmat.2015.03.
2021.125153. 091.
[158] Cui J, Hao H, Shi Y, et al. Experimental study of con- [173] Li VC, Yang EH. Self healing in concrete materials.
crete damage under high hydrostatic pressure. Cem In: S van der Zwaag, editor. Springer series in
Concr Res. 2017;100:140–152. doi:10.1016/j. materials science and technology. Dordrecht:
cemconres.2017.06.005. Springer Netherlands; 2007. p. 161–193.
[159] Cui J, Hao H, Shi Y. Study of concrete damage mech- [174] Hung C-C, Su Y-F. Medium-term self-healing evalu-
anism under hydrostatic pressure by numerical ation of engineered cementitious composites with
simulations. Constr Build Mater. 2018;160:440–449. varying amounts of fly ash and exposure durations.
doi:10.1016/j.conbuildmat.2017.11.083. Constr Build Mater. 2016;118:194–203. doi:10.1016/
[160] Yi S-T, Hyun T-Y, Kim J-K. The effects of hydraulic j.conbuildmat.2016.05.021.
pressure and crack width on water permeability of [175] Qiu J, Tan HS, Yang E-H. Coupled effects of crack
penetration crack-induced concrete. Constr Build width, slag content, and conditioning alkalinity on
596 V. CAPPELLESSO ET AL.

autogenous healing of engineered cementitious com- results from on-going Brazil-Italy cooperation. Spec
posites. Cem Concr Compos. 2016;73:203–212. Publ. 2015;305:11.1–11.10.
doi:10.1016/j.cemconcomp.2016.07.013. [190] Ferrara L, Albertini I, Gettu R, et al. Self healing of
[176] Guan X, Zhang C, Li Y, et al. Effect of exposure con- cement based materials engineered through crystal-
ditions on self-healing behavior of engineered line admixtures: experimental results from a multina-
cementitious composite incorporating limestone tional university network. Spec Publ. 2015;305:13.1–
powder. Cem Concr Compos. 2020;114:103808. 13.10.
doi:10.1016/j.cemconcomp.2020.103808. [191] Özbay E, Sahmaran M, Yücel HE, et al. Effect of sus-
[177] Ferrara L, Krelani V, Moretti F. Autogenous healing tained flexural loading on self-healing of engineered
on the recovery of mechanical performance of high cementitious composites. J Adv Concr Technol.
performance fibre reinforced cementitious compo- 2013;11:167–179. doi:10.3151/jact.11.167.
sites (HPFRCCs): part 2 – correlation between heal- [192] Yildirim G, Alyousif A, Şahmaran M, et al. Assessing
ing of mechanical performance and crack sealing. the self-healing capability of cementitious composites
Cem Concr Compos. 2016;73:299–315. doi:10.1016/ under increasing sustained loading. Adv Cem Res.
j.cemconcomp.2016.08.003. 2015;27:581–592. doi:10.1680/adcr.14.00111.
[178] Ferrara L, Krelani V, Moretti F, et al. Effects of auto- [193] Kannikachalam NP, Clerque Vela AD, Ocampo
genous healing on the recovery of mechanical per- Pacheco Y, et al. Fatigue behavior and effect of stimu-
formance of high performance fibre reinforced lated autogenous self-healing in ultra high-perform-
cementitious composites (HPFRCCs): part 1. Cem ance concrete. Proceedings of the 14th fib PhD
Concr Compos. 2017;83:76–100. doi:10.1016/j. Symposium in Civil Engineering; Sept 5–7, Rome;
cemconcomp.2017.07.010. 2022.
[179] Lo Monte F, Ferrara L. Self-healing characterization [194] Feiteira J, Gruyaert E, De Belie N. Self-healing of
of UHPFRCC with crystalline admixture: experimen- moving cracks in concrete by means of encapsulated
tal assessment via multi-test/multi-parameter polymer precursors. Constr Build Mater.
approach. Constr Build Mater. 2021;283:122579. 2016;102:671–678. doi:10.1016/j.conbuildmat.2015.
doi:10.1016/j.conbuildmat.2021.122579. 10.192.
[180] Jonkers HM, Thijssen A. Bacteria mediated of con- [195] Nji J, Li G. A biomimic shape memory polymer based
crete structures. Proceedings of the Second self-healing particulate composite. Polymer.
International Symposium on Service Life Design for 2010;51:6021–6029. doi:10.1016/j.polymer.2010.10.
Infrastructures; 4–6 Oct. 2010, Delft, Netherlands, 021.
2010. p. 833–840. [196] Botusharova SP. Self-healing geotechnical structures
[181] Wang JY, De Belie N, Verstraete W. via microbial action [PhD thesis]. Cardiff School of
Diatomaceous earth as a protective vehicle for Engineering, 2017.
bacteria applied for self-healing concrete. J Ind [197] Justo-Reinoso I, Reeksting BJ, Heath A, et al.
Microbiol Biotechnol. 2012;39:567–577. doi:10. Evaluation of cyclic healing potential of bacteria-
1007/s10295-011-1037-1. based self-healing cementitious composites.
[182] Jonkers HM. Bacteria-based self-healing concrete. Sustainability. 2022;14:6845. doi:10.3390/
Heron. 2011;56:1–12. su14116845.
[183] Wiktor V, Jonkers HM. Quantification of crack-heal- [198] In C-W, Holland RB, Kim J-Y, et al. Monitoring and
ing in novel bacteria-based self-healing concrete. evaluation of self-healing in concrete using diffuse
Cem Concr Compos. 2011;33:763–770. doi:10.1016/ ultrasound. NDT E Int. 2013;57:36–44. doi:10.1016/
j.cemconcomp.2011.03.012. j.ndteint.2013.03.005.
[184] Erşan YC, Palin D, Yengec Tasdemir SB, et al. [199] Maes M, Snoeck D, De Belie N. Chloride penetration
Volume fraction, thickness, and permeability of the in cracked mortar and the influence of autogenous
sealing layer in microbial self-healing concrete con- crack healing. Constr Build Mater. 2016;115:114–
taining biogranules. Front Built Environ. 2018;4:1– 124. doi:10.1016/j.conbuildmat.2016.03.180.
11. doi:10.3389/fbuil.2018.00070. [200] Botusharova S, Gardner D, Harbottle M.
[185] da Silva FB, De Belie N, Boon N, et al. Production of Augmenting microbially induced carbonate precipi-
non-axenic ureolytic spores for self-healing concrete tation of soil with the capability to self-heal. J
applications. Constr Build Mater. 2015;93:1034– Geotech Geoenvironmental Eng. 2020;146:1–13.
1041. doi:10.1016/j.conbuildmat.2015.05.049. doi:10.1061/(ASCE)GT.1943-5606.0002214.
[186] Wang JY, Soens H, Verstraete W, et al. Self-healing [201] Mao W, Litina C, Al-Tabbaa A. Development and
concrete by use of microencapsulated bacterial application of novel sodium silicate microcapsule-
spores. Cem Concr Res. 2014;56:139–152. doi:10. based self-healing oil well cement. Materials.
1016/j.cemconres.2013.11.009. 2020;13:1–18. doi:10.3390/ma13020456.
[187] Wang JY, Snoeck D, Van Vlierberghe S, et al. [202] Criado M, Gimenez M, Menéndez E, et al. Durability
Application of hydrogel encapsulated carbonate pre- performance of uncracked and cracked nanoaddi-
cipitating bacteria for approaching a realistic self- tioned UHPFRCs in a dynamic leaching system.
healing in concrete. Constr Build Mater. University of Sheffield (Ed.), 74th RILEM Annual
2014;68:110–119. doi:10.1016/j.conbuildmat.2014. Week and 40th Cement and Concrete Science
06.018. Conference, Sheffield, 2020. doi:10.5281/zenodo.
[188] Wang J, Dewanckele J, Cnudde V, et al. X-ray com- 4289921.
puted tomography proof of bacterial-based self-heal- [203] Cuenca E, D’Ambrosio L, Lizunov D, et al.
ing in concrete. Cem Concr Compos. 2014;53:289– Mechanical properties and self-healing capacity of
304. doi:10.1016/j.cemconcomp.2014.07.014. ultra high performance fibre reinforced concrete
[189] Ferrara L, Ferreira SR, Krelani V, et al. Natural fibres with alumina nano-fibres: tailoring ultra high dura-
As promoters of autogeneous healing in HPFRCCS: bility concrete for aggressive exposure scenarios.
 INTERNATIONAL MATERIALS REVIEWS 597

Cem Concr Compos. 2021;118:103956. doi:10.1016/j. expansive agent. Cem Concr Res. 2021;139:106252.
cemconcomp.2021.103956. doi:10.1016/j.cemconres.2020.106252.
[204] Ferrara L, Krelani V, Moretti F. On the use of crystal- [219] Cuenca E, Lo Monte F, Moro M, et al. Effects of auto-
line admixtures in cement based construction genous and stimulated self-healing on durability and
materials: from porosity reducers to promoters of mechanical performance of UHPFRC: validation of
self healing. Smart Mater Struct. 2016;25:17. doi:10. tailored test method through multi-performance
1088/0964-1726/25/8/084002. healing-induced recovery indices. Sustainability.
[205] Kanellopoulos A, Giannaros P, Palmer D, et al. 2021;13:11386. doi:10.3390/su132011386.
Polymeric microcapsules with switchable mechanical [220] Abro FuR, Buller AS, Lee K-M, et al. Using the
properties for self-healing concrete: synthesis, steady-state chloride migration test to evaluate the
characterisation and proof of concept. Smart Mater self-healing capacity of cracked mortars containing
Struct. 2017;26:045025. doi:10.1088/1361-665X/ crystalline, expansive, and swelling admixtures.
aa516c. Materials. 2019;12:1865. doi:10.3390/ma12111865.
[206] Gruyaert E, Van Tittelboom K, Sucaet J, et al. [221] Xue C. Performance and mechanisms of stimulated
Capsules with evolving brittleness to resist the prep- self-healing in cement-based composites exposed to
aration of self-healing concrete. Mater Construcción. saline environments. Cem Concr Compos.
2016;66:e092. doi:10.3989/mc.2016.07115. 2022;129. doi:doi:10.1016/j.cemconcomp.2022.
[207] Joseph C. Experimental and numerical study of the 104470.
fracture and self-healing of cementitious materials [222] Matos AM, Chaves Figueiredo S, Nunes S, et al.
[PhD thesis]. Cardiff University, 2008. Available Durability of an UHPFRC under mechanical and
from: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl. chloride loads. Constr Build Mater.
ethos.584409. 2021;311:125223. doi:10.1016/j.conbuildmat.2021.
[208] Espitia Nery ME, Corredor Pulido DE, Castaño 125223.
Oliveros PA, et al. Mechanisms of encapsulation of [223] Li M, Li VC. Cracking and healing of engineered
bacteria in self-healing concrete: review. DYNA. cementitious composites under chloride environ-
2019;86:17–22. doi:10.15446/dyna.v86n210.75343. ment. ACI Mater J. 2011;108:333–340. doi:10.
[209] Harwood CR, Cutting SM. Molecular biological 14359/51682499.
methods for bacillus. Chichester: John Wiley and [224] Maes M, Van Tittelboom K, De Belie N. The
Sons Ltd; 1990. efficiency of self-healing cementitious materials by
[210] Erşan YÇ, Van Tittelboom K, Boon N, et al. Nitrite means of encapsulated polyurethane in chloride con-
producing bacteria inhibit reinforcement bar cor- taining environments. Constr Build Mater.
rosion in cementitious materials. Sci Rep. 2014;71:528–537. doi:10.1016/j.conbuildmat.2014.
2018;8:14092. doi:10.1038/s41598-018-32463-6. 08.053.
[211] Joseph C, Jefferson AD, Isaacs B, et al. Experimental [225] Van Belleghem B, Van den Heede P, Van Tittelboom
investigation of adhesive-based self-healing of K, et al. Quantification of the service life extension
cementitious materials. Mag Concr Res. and environmental benefit of chloride exposed self-
2010;62:831–843. doi:10.1680/macr.2010.62.11.831. healing concrete. Materials. 2016;10:5. doi:10.3390/
[212] Palin D, Mo Y, Wiktor V, et al. An improved test for ma10010005.
generating rapid, accurate, and reliable crack per- [226] Van Belleghem B, Van den Heedeand P, De Belie N.
meability data for cementitious materials. Int J Civ Resistance to chloride penetration of self-healing
Eng. 2019;17:645–652. doi:10.1007/s40999-018- concrete with encapsulated polyurethane. 4th
0326-8. International Conference on Sustainable
[213] Palin D, Jonkers HM, Wiktor V. Autogenous healing Construction Materials and Technologies, 2016.
of sea-water exposed mortar: quantification through p. 1291–1300. doi:10.18552/2016/SCMT4D118.
a simple and rapid permeability test. Cem Concr [227] Van Belleghem B, Kessler S, Van den Heede P, et al.
Res. 2016;84:1–7. doi:10.1016/j.cemconres.2016.02. Chloride induced reinforcement corrosion behavior
011. in self-healing concrete with encapsulated poly-
[214] Şahmaran M. Effect of flexure induced transverse urethane. Cem Concr Res. 2018;113:130–139.
crack and self-healing on chloride diffusivity of doi:10.1016/j.cemconres.2018.07.009.
reinforced mortar. J Mater Sci. 2007;42:9131–9136. [228] van den Heede P, van Belleghem B, Araújo MA, et al.
doi:10.1007/s10853-007-1932-z. Screening of different encapsulated polymer-based
[215] Van Mullem T, Caspeele R, De Belie N. The influence healing agents for chloride exposed self-healing con-
of SAPs on chloride ingress in cracked concrete. crete using chloride migration tests. Key Eng Mater.
MATEC Web Conf. 2019;289:08007. doi:10.1051/ 2018;761:152–158. doi:10.4028/www.scientific.net/
matecconf/201928908007. KEM.761.152.
[216] Darquennes A, Olivier K, Benboudjema F, et al. Self- [229] Erşan YÇ, Verbruggen H, De Graeve I, et al. Nitrate
healing at early-age, a way to improve the chloride reducing CaCO3 precipitating bacteria survive in
resistance of blast-furnace slag cementitious mortar and inhibit steel corrosion. Cem Concr
materials. Constr Build Mater. 2016;113:1017–1028. Res. 2016;83:19–30. doi:10.1016/j.cemconres.2016.
doi:10.1016/j.conbuildmat.2016.03.087. 01.009.
[217] Şahmaran M, Li VC. Durability properties of micro- [230] Tan L, Reeksting B, Ferrandiz-Mas V, et al. Effect of
cracked ECC containing high volumes fly ash. Cem carbonation on bacteria-based self-healing of cemen-
Concr Res. 2009;39:1033–1043. doi:10.1016/j. titious composites. Constr Build Mater.
cemconres.2009.07.009. 2020;257:119501. doi:10.1016/j.conbuildmat.2020.
[218] Xue C, Li W, Luo Z, et al. Effect of chloride ingress on 119501.
self-healing recovery of smart cementitious compo- [231] Hung C-C, Hung H-H. Potential of sodium sulfate
site incorporating crystalline admixture and MgO solution for promoting the crack-healing
598 V. CAPPELLESSO ET AL.

performance for strain-hardening cementitious com- [245] He P, Yu J, Wan Y, et al. Effect of ion chelator on
posites. Cem Concr Compos. 2020;106:103461. microstructure and properties of cement-based
doi:10.1016/j.cemconcomp.2019.103461. materials under sulfate dry-wet cycle attack. Constr
[232] Liu H, Zhang Q, Gu C, et al. Self-healing of micro- Build Mater. 2020;257:119527. doi:10.1016/j.
cracks in engineered cementitious composites conbuildmat.2020.119527.
under sulfate and chloride environment. Constr [246] Xue C. Cracking and autogenous self-healing on the
Build Mater. 2017;153:948–956. doi:10.1016/j. performance of fiber-reinforced MgO-cement com-
conbuildmat.2017.07.126. posites in seawater and NaCl solutions. Constr
[233] Du W, Lin R, Liu Q. Investigation of isophorone dii- Build Mater. 2022;326:126870. doi:10.1016/j.
socyanate microcapsules to improve self-healing conbuildmat.2022.126870.
properties and sulfate resistance of concrete. Constr [247] Giménez M, Alonso MC, Menéndez E, et al.
Build Mater. 2021;300:124438. doi:10.1016/j. Durability of UHPFRC functionalised with nanoad-
conbuildmat.2021.124438. ditives due to synergies in the action of sulphate
[234] Liu H, Huang H, Wu X, et al. Effects of external and chloride in cracked and uncracked states.
multi-ions and wet-dry cycles in a marine environ- Mater Construcción. 2021;71:e264. doi:10.3989/mc.
ment on autogenous self-healing of cracks in cement 2021.14021.
paste. Cem Concr Res. 2019;120:198–206. doi:10. [248] Khan MBE, Shen L, Dias-da-Costa D. Self-healing
1016/j.cemconres.2019.03.014. behaviour of bio-concrete in submerged and tidal
[235] Danner T, Hjorth Jakobsen U, Geiker MR. marine environments. Constr Build Mater.
Mineralogical sequence of self-healing products in 2021;277:122332. doi:10.1016/j.conbuildmat.2021.
cracked marine concrete. Minerals. 2019;9:284. 122332.
doi:10.3390/min9050284. [249] Ehsan Khan MB, Shen L, Dias-da-Costa D. Crack
[236] Kim HG, Atta-ur-Rehman, Qudoos A, et al. Self- healing performance of bacteria-based mortar
healing performance of GGBFS based cementitious under sustained tensile loading in marine environ-
mortar with granulated activators exposed to a sea- ment. Cem Concr Compos. 2021;120:104055.
water environment. Constr Build Mater. doi:10.1016/j.cemconcomp.2021.104055.
2018;188:569–582. doi:10.1016/j.conbuildmat.2018. [250] Wang R, Yu J, He P, et al. Investigation of ion chela-
08.092. tor and mineral admixtures improving salt-frost
[237] Zha Y, Yu J, Wang R, et al. Effect of ion chelating resistance of cement-based materials. Constr Build
agent on self-healing performance of cement-based Mater. 2019;227:116670. doi:10.1016/j.conbuildmat.
materials. Constr Build Mater. 2018;190:308–316. 2019.08.051.
doi:10.1016/j.conbuildmat.2018.09.115. [251] Şahmaran M, Li VC. De-icing salt scaling resistance
[238] Li G, Huang X, Lin J, et al. Activated chemicals of of mechanically loaded engineered cementitious
cementitious capillary crystalline waterproofing composites. Cem Concr Res. 2007;37:1035–1046.
materials and their self-healing behaviour. Constr doi:10.1016/j.cemconres.2007.04.001.
Build Mater. 2019;200:36–45. doi:10.1016/j. [252] Zhu Y, Yang Y, Yao Y. Autogenous self-healing of
conbuildmat.2018.12.093. engineered cementitious composites under freeze–
[239] Liu H, Huang H, Wu X, et al. Promotion on self-heal- thaw cycles. Constr Build Mater. 2012;34:522–530.
ing of cracked cement paste by triethanolamine in a doi:10.1016/j.conbuildmat.2012.03.001.
marine environment. Constr Build Mater. [253] Cappellesso VG, Van Mullem T, Gruyaert E, et al.
2020;242:118148. doi:10.1016/j.conbuildmat.2020. Repair of concrete in environments with chlorides
118148. or subjected to freeze-thaw scaling. 75th RILEM
[240] Liu H, Huang H, Wu X, et al. Interactions between Week, Merida, 2021.
organic chelation agents and ions in seawater for [254] Cappellesso VG, Van Mullem T, Gruyaert E, et al.
accelerating self-healing of cracks in cement paste. J Self-healing bacterial concrete exposed to freezing
Mater Civ Eng. 2021;33:04021036. doi:10.1061/ and thawing associated with chlorides. Proceedings
(ASCE)MT.1943-5533.0003639. of Resilient Materials 4 Life 2020. Cardiff: Cardiff
[241] Hu M, Guo J, Xu Y, et al. Influence of triethanola- University; 2021. p. 241–246. doi:10.5281/zenodo.
mine on reactivity of hydrated matrix in sodium sili- 6324948.
cate self-healing system and the mechanism. Constr [255] Hooshmand A, Kianoush R, Siad H, et al. Volumetric
Build Mater. 2018;185:445–452. doi:10.1016/j. strain behaviour and self-healing of large scale engin-
conbuildmat.2018.07.040. eered cementitious composite and normal concrete
[242] Wu X, Huang H, Liu H, et al. A new self-healing panels under natural conditions. Constr Build
agent for accelerating the healing kinetics while sim- Mater. 2021;308:125078. doi:10.1016/j.conbuildmat.
ultaneously binding seawater ions in cracked cement 2021.125078.
paste. Mater Lett. 2021;283:128884. doi:10.1016/j. [256] Du W, Liu Q, Lin R. Effects of toluene-di-isocyanate
matlet.2020.128884. microcapsules on the frost resistance and self-repair-
[243] Wu X, Huang H, Liu H, et al. Reactions of self-heal- ing capability of concrete under freeze-thaw cycles. J
ing agents and the chemical binding of aggressive Build Eng. 2021;44:102880. doi:10.1016/j.jobe.2021.
ions in sea water: thermodynamics and kinetics. 102880.
Cem Concr Res. 2021;145:106450. doi:10.1016/j. [257] Gruyaert E, Debbaut B, Snoeck D, et al. Self-healing
cemconres.2021.106450. mortar with pH-sensitive superabsorbent polymers:
[244] Lo Monte F, Ferrara L. Tensile behaviour identifi- testing of the sealing efficiency by water flow tests.
cation in ultra-high performance fibre reinforced Smart Mater Struct. 2016;25:084007. doi:10.1088/
cementitious composites: indirect tension tests and 0964-1726/25/8/084007.
back analysis of flexural test results. Mater Struct. [258] Hooshmand A, Kianoush R, Lachemi M, et al.
2020;53:145. doi:10.1617/s11527-020-01576-8. Combined effect of pressurized water and
 INTERNATIONAL MATERIALS REVIEWS 599

sustained load on self-healing of [273] Dobrovolski MEG, Munhoz GS, Pereira E, et al.
engineered cementitious composite panels. J Mater Effect of crystalline admixture and polypropylene
Civ Eng. 2022;34. doi:10.1061/(ASCE)MT.1943- microfiber on the internal sulfate attack in Portland
5533.0004104. cement composites due to pyrite oxidation. Constr
[259] Reinhardt H-W, Jooss M. Permeability and self- Build Mater. 2021;308. doi:doi:10.1016/j.
healing of cracked concrete as a function of conbuildmat.2021.125018.
temperature and crack width. Cem Concr Res. [274] Van Tittelboom K, Gruyaert E, Rahier H, et al.
2003;33:981–985. doi:10.1016/S0008-8846(02)010 Influence of mix composition on the extent of auto-
99-2. genous crack healing by continued hydration or cal-
[260] Suleiman AR, Nehdi ML. Effect of environmental cium carbonate formation. Constr Build Mater.
exposure on autogenous self-healing of cracked 2012;37:349–359. doi:10.1016/j.conbuildmat.2012.
cement-based materials. Cem Concr Res. 07.026.
2018;111:197–208. doi:10.1016/j.cemconres.2018.05. [275] Huang H, Ye G, Damidot D. Effect of blast furnace
009. slag on self-healing of microcracks in cementitious
[261] Mostavi E, Asadi S, Hassan MM, et al. Evaluation of materials. Cem Concr Res. 2014;60:68–82. doi:10.
self-healing mechanisms in concrete with double- 1016/j.cemconres.2014.03.010.
walled sodium silicate microcapsules. J Mater Civ [276] Olivier K, Darquennes A, Benboudjema F, et al.
Eng. 2015;27:04015035. doi:10.1061/(ASCE)MT. Early-age self-healing of cementitious materials con-
1943-5533.0001314. taining ground granulated blast-furnace slag under
[262] Ren J, Wang X, Li D, et al. Performance of tempera- water curing. J Adv Concr Technol. 2016;14:717–
ture adaptive microcapsules in self-healing cementi- 727. doi:10.3151/jact.14.717.
tious materials under different mixing [277] Palin D, Wiktor V, Jonkers HM. Autogenous healing
temperatures. Constr Build Mater. 2021;299:124254. of marine exposed concrete: characterization and
doi:10.1016/j.conbuildmat.2021.124254. quantification through visual crack closure. Cem
[263] Du W, Yu J, Gu S, et al. Effect of temperatures on Concr Res. 2015;73:17–24. doi:10.1016/j.cemconres.
self-healing capabilities of concrete with different 2015.02.021.
shell composition microcapsules containing [278] Jacobsen S, Sellevold EJ. Self healing of high strength
toluene-di-isocyanate. Constr Build Mater. concrete after deterioration by freeze/thaw. Cem
2020;247:118575. doi:10.1016/j.conbuildmat.2020. Concr Res. 1996;26:55–62. doi:10.1016/0008-8846
118575. (95)00179-4.
[264] Van Tittelboom K, Gruyaert E, De Backer P, et al. [279] Mechtcherine V, Schröfl C, Wyrzykowski M, et al.
Self-repair of thermal cracks in concrete sandwich Effect of superabsorbent polymers (SAP) on the
panels. Struct Concr. 2015;16:273–288. doi:10.1002/ freeze–thaw resistance of concrete: results of a
suco.201400055. RILEM interlaboratory study. Mater Struct.
[265] Mao W, Litina C, Al-Tabbaa A. The use of self-heal- 2017;50:14. doi:10.1617/s11527-016-0868-7.
ing sodium silicate microcapsules in oil well cement [280] Mignon A, Snoeck D, Dubruel P, et al. Crack mitiga-
under high temperature wellbore environment. tion in concrete: superabsorbent polymers as key to
LORCENIS, Ghent, 2019. success? Materials. 2017;10:237. doi:10.3390/
[266] Teall O, Pilegis M, Davies R, et al. A shape memory ma10030237.
polymer concrete crack closure system activated by [281] Yao Y, Zhu Y, Yang Y. Incorporation superabsorbent
electrical current. Smart Mater Struct. polymer (SAP) particles as controlling pre-existing
2018;27:075016. doi:10.1088/1361-665X/aac28a. flaws to improve the performance of engineered
[267] Nishiwaki T, Mihashi H, Jang B, et al. Development cementitious composites (ECC). Constr Build
of self-healing system for concrete with selective Mater. 2012;28:139–145. doi:10.1016/j.conbuildmat.
heating around crack. J Adv Concr Technol. 2011.08.032.
2006;4:267–275. [282] Lee HXD, Wong HS, Buenfeld NR. Potential of
[268] Skevi L, Reeksting B, Gebhard S, et al. Bacteria-based superabsorbent polymer for self-sealing cracks in
self-healing concrete at low temperature. In: concrete. Adv Appl Ceram. 2010;109:296–302.
Maddalena MW-S R, editor. M4L2020 International doi:10.1179/174367609X459559.
Conference Proceedings, Cardiff University, 2021. [283] Justs J, Wyrzykowski M, Bajare D, et al. Internal cur-
p. 248–252. ing by superabsorbent polymers in ultra-high per-
[269] Palin D, Wiktor V, Jonkers H. A bacteria-based self- formance concrete. Cem Concr Res. 2015;76:82–90.
healing cementitious composite for application in doi:10.1016/j.cemconres.2015.05.005.
low-temperature marine environments. [284] Snoeck D, De Belie N. Repeated autogenous healing
Biomimetics. 2017;2:13. doi:10.3390/ in strain-hardening cementitious composites by
biomimetics2030013. using superabsorbent polymers. J Mater Civ Eng.
[270] Palin D, Wiktor V, Jonkers HM. A bacteria-based 2016;28:04015086. doi:10.1061/(ASCE)MT.1943-
bead for possible self-healing marine concrete appli- 5533.0001360.
cations. Smart Mater Struct. 2016;25:084008. doi:10. [285] Van Tittelboom K, De Belie N, De Muynck W, et al.
1088/0964-1726/25/8/084008. Use of bacteria to repair cracks in concrete. Cem
[271] Ivanov FM, Polyakov BI. Self-healing and durability Concr Res. 2010;40:157–166. doi:10.1016/j.
of hydraulic concrete. Hydrotechnical Constr. cemconres.2009.08.025.
1974;8:844–849. doi:10.1007/BF02380442. [286] Van Tittelboom K, De Belie N, Van Loo D, et al. Self-
[272] de Rooij M, Van Tittelboom K, Schlangen E, et al. healing efficiency of cementitious materials contain-
Self-Healing phenomena in cement-based materials. ing tubular capsules filled with healing agent. Cem
Dordrecht: Springer Netherlands; 2013. doi:10. Concr Compos. 2011;33:497–505. doi:10.1016/j.
1007/978-94-007-6624-2. cemconcomp.2011.01.004.
600 V. CAPPELLESSO ET AL.

[287] Van Mullem T. Report test results Korean water per- Constr Build Mater. 2008;22:1981–1989. doi:10.
meability, 2018, 1–12. 1016/j.conbuildmat.2007.07.003.
[288] Ma H, Qian S, Zhang Z. Effect of self-healing on [301] Ahn T-H, Kishi T. Crack self-healing behavior of
water permeability and mechanical property of med- cementitious composites incorporating various min-
ium-early-strength engineered cementitious compo- eral admixtures. J Adv Concr Technol. 2010;8:171–
sites. Constr Build Mater. 2014;68:92–101. doi:10. 186. doi:10.3151/jact.8.171.
1016/j.conbuildmat.2014.05.065. [302] Jiang Z, Li W, Yuan Z, et al. Self-healing of cracks in
[289] Van Mullem T, Van Tittelboom K, Gruyaert E, et al. concrete with various crystalline mineral additives in
Development of an improved cracking method to underground environment. J Wuhan Univ Technol
reduce the variability in testing the healing efficiency Mater Sci Ed. 2014;29:938–944. doi:10.1007/s11595-
of self-healing mortar containing encapsulated poly- 014-1024-2.
mers. MATEC Web Conf. 2018;199:02017. doi:10. [303] Sherir MAA, Hossain KMA, Lachemi M. Self-healing
1051/matecconf/201819902017. and expansion characteristics of cementitious com-
[290] Van Mullem T, Gruyaert E, Debbaut B, et al. Novel posites with high volume fly ash and MgO-type
active crack width control technique to reduce the expansive agent. Constr Build Mater. 2016;127:80–
variation on water permeability results for self-heal- 92. doi:10.1016/j.conbuildmat.2016.09.125.
ing concrete. Constr Build Mater. 2019;203:541– [304] Žižková N, Nevřivová L, Lédl M. Durability of
551. doi:10.1016/j.conbuildmat.2019.01.105. cement based mortars containing crystalline addi-
[291] Van Tittelboom K, Tsangouri E, Van Hemelrijck D, tives. Defect Diffus Forum. 2018;382:246–253.
et al. The efficiency of self-healing concrete using doi:10.4028/www.scientific.net/DDF.382.246.
alternative manufacturing procedures and more rea- [305] Al-Kheetan MJ, Rahman MM, Chamberlain DA.
listic crack patterns. Cem Concr Compos. Influence of early water exposure on modified cemen-
2015;57:142–152. doi:10.1016/j.cemconcomp.2014. titious coating. Constr Build Mater. 2017;141:64–71.
12.002. doi:10.1016/j.conbuildmat.2017.02.159.
[292] Van den Heede P, Van Belleghem B, Alderete N, et al. [306] Coppola L, Coffetti D, Crotti E. Innovative carboxylic
Neutron radiography based visualization and profil- acid waterproofing admixture for self-sealing water-
ing of water uptake in (Un)cracked and autono- tight concretes. Constr Build Mater. 2018;171:817–
mously healed Cementitious materials. Materials. 824. doi:10.1016/j.conbuildmat.2018.03.201.
2016;9:311. doi:10.3390/ma9050311. [307] Byoungsun P, Young CC. Investigating a new
[293] De Nardi C, Bullo S, Ferrara L, et al. Effectiveness of method to assess the self-healing performance of har-
crystalline admixtures and lime/cement coated gran- dened cement pastes containing supplementary
ules in engineered self-healing capacity of lime mor- cementitious materials and crystalline admixtures. J
tars. Mater Struct. 2017;50:191. doi:10.1617/s11527- Mater Res Technol. 2019;8:6058–6073. doi:10.1016/
017-1053-3. j.jmrt.2019.09.080.
[294] Shin KJ, Bae W, Choi S-W, et al. Parameters influen- [308] Al-Kheetan MJ, Ghaffar SH, Madyan OA, et al.
cing water permeability coefficient of cracked con- Development of low absorption and high-resistant
crete specimens. Constr Build Mater. 2017;151:907– sodium acetate concrete for severe environmental
915. doi:10.1016/j.conbuildmat.2017.06.093. conditions. Constr Build Mater. 2020;230:117057.
[295] Paine K. Bacteria-based self-healing concrete: effects doi:10.1016/j.conbuildmat.2019.117057.
of environment, exposure and crack size. In: Wiktor [309] Yang Z, Hollar J, He X, et al. A self-healing cementi-
AB (editor), RILEM Conference on Microorganisms tious composite using oil core/silica gel shell micro-
Cementitious Materials Interactions, Delft, capsules. Cem Concr Compos. 2011;33:506–512.
Netherlands. Paris: RILEM publications S.A.R.L; doi:10.1016/j.cemconcomp.2011.01.010.
2016: p. 16. Available from: https://purehost.bath. [310] García Calvo JL, Pérez G, Carballosa P, et al.
ac.uk/ws/portalfiles/portal/146678313/Paine_delft_ Development of ultra-high performance concretes
conference_revised_archive.pdf. with self-healing micro/nano-additions. Constr
[296] Li G, Nettles D. Thermomechanical characterization Build Mater. 2017;138:306–315. doi:10.1016/j.
of a shape memory polymer based self-repairing syn- conbuildmat.2017.02.015.
tactic foam. Polymer. 2010;51:755–762. doi:10.1016/j. [311] Zhang D, Wu H, Li VC, et al. Autogenous healing of
polymer.2009.12.002. engineered cementitious composites (ECC) based on
[297] Fan S, Li M. X-ray computed microtomography of MgO-fly ash binary system activated by carbonation
threedimensional microcracks and self-healing in curing. Constr Build Mater. 2020;238:117672. doi:10.
engineered cementitious composites. Smart Mater 1016/j.conbuildmat.2019.117672.
Struct. 2015;24. doi:10.1088/0964-1726/24/1/015021. [312] Schlangen E. Crack development in concrete, part 1:
[298] Cappellesso V, Silva D, Petry N, et al. Self-healing fracture experiments and CT-scan observations. Key
approach on early age cracked concretes with smart Eng Mater. 2008;385–387:69–72. doi:10.4028/www.
admixture. Durable Concrete Infrastructure under scientific.net/KEM.385-387.69.
Severe Conditions – Smart Admixtures, Self- [313] Suzuki T, Ogata H, Takada R, et al. Use of acoustic
Responsiveness Nano-Additions, Ghent Magnel emission and X-ray computed tomography for
Lab. Concr. Res., 2019: pp. 89–92. damage evaluation of freeze-thawed concrete.
[299] Huang H, Ye G. Application of sodium silicate sol- Constr Build Mater. 2010;24:2347–2352. doi:10.
ution as self-healing agent in cementitious materials 1016/j.conbuildmat.2010.05.005.
Outline • Introduction • Materials and experiments [314] Darma IS, Sugiyama T, Promentilla MAB.
• Results and discussion Problems Research objec- Application of X-Ray CT to study diffusivity in
tive:1–9; 2011. cracked concrete through the observation of tracer
[300] Yang K, Song J, Ashour A, et al. Properties of transport. J Adv Concr Technol. 2013;11:266–281.
cementless mortars activated by sodium silicate. doi:10.3151/jact.11.266.
 INTERNATIONAL MATERIALS REVIEWS 601

[315] Roels S, Carmeliet J. Analysis of moisture flow in por- [328] Alghamri R, Kanellopoulos A, Al-Tabbaa A.
ous materials using microfocus X-ray radiography. Impregnation and encapsulation of lightweight
Int J Heat Mass Transf. 2006;49:4762–4772. doi:10. aggregates for self-healing concrete. Constr Build
1016/j.ijheatmasstransfer.2006.06.035. Mater. 2016;124:910–921. doi:10.1016/j.
[316] Cnudde V, Dierick M, Vlassenbroeck J, et al. High- conbuildmat.2016.07.143.
speed neutron radiography for monitoring the [329] Qureshi TS, Kanellopoulos A, Al-Tabbaa A.
water absorption by capillarity in porous materials. Encapsulation of expansive powder minerals within
Nucl Instruments Methods Phys Res Sect B Beam a concentric glass capsule system for self-healing
Interact Mater Atoms. 2008;266:155–163. doi:10. concrete. Constr Build Mater. 2016;121:629–643.
1016/j.nimb.2007.10.030. doi:10.1016/j.conbuildmat.2016.06.030.
[317] Zhang P, Wittmann FH, Zhao T, et al. Neutron ima- [330] Tziviloglou H, Wiktor E, Wang V, et al. Evaluation of
ging of water penetration into cracked steel experimental methodology to assess the sealing
reinforced concrete. Phys B Condens Matter. efficiency of bacteria based self-healing concrete:
2010;405:1866–1871. doi:10.1016/j.physb.2010.01. round robin test. In: Wiktor V, Jonkers H, Bertron
065. A (editors), Proceedings RILEM Conference on
[318] Derluyn H, Griffa M, Mannes D, et al. Characterizing Microorganisms-Cementitious Materials
saline uptake and salt distributions in porous lime- Interactions. RILEM Proceedings. PRO102, Delft,
stone with neutron radiography and X-ray micro- 2016. p. 43.
tomography. J Build Phys. 2013;36:353–374. doi:10. [331] Zhang Z, Qian S, Ma H. Investigating mechanical
1177/1744259112473947. properties and self-healing behavior of micro-
[319] Van Tittelboom K, Snoeck D, Vontobel P, et al. Use cracked ECC with different volume of fly ash.
of neutron radiography and tomography to visualize Constr Build Mater. 2014;52:17–23. doi:10.1016/j.
the autonomous crack sealing efficiency in cementi- conbuildmat.2013.11.001.
tious materials. Mater Struct. 2013;46:105–121. [332] Cuenca L, Mezzena E, and Ferrara A. Synergy
doi:10.1617/s11527-012-9887-1. between crystalline admixtures and nano-constitu-
[320] Dewanckele J, De Kock T, Fronteau G, et al. Neutron ents in enhancing autogenous healing capacity of
radiography and X-ray computed tomography for cementitious composites under cracking and healing
quantifying weathering and water uptake processes cycles in aggressive waters. Constr Build Mater. 2020.
inside porous limestone used as building material. [333] Shields Y, Van Mullem T, De Belie N, et al. An inves-
Mater Charact. 2014;88:86–99. doi:10.1016/j. tigation of suitable healing agents for vascular-based
matchar.2013.12.007. self-healing in cementitious materials. Sustainability.
[321] Van Belleghem B, Montoya R, Dewanckele J, et al. 2021;13:12948. doi:10.3390/su132312948.
Capillary water absorption in cracked and uncracked [334] Gruyaert E. Effect of blast-furnace slag as cement
mortar – a comparison between experimental study replacement on hydration, microstructure, strength
and finite element analysis. Constr Build Mater. and durability of concrete. Belgium: Ghent
2016;110:154–162. doi:10.1016/j.conbuildmat.2016. University; 2010.
02.027. [335] Liu S, Bundur ZB, Zhu J, et al. Evaluation of self-heal-
[322] Snoeck D, Steuperaert S, Van Tittelboom K, et al. ing of internal cracks in biomimetic mortar using
Visualization of water penetration in cementitious coda wave interferometry. Cem Concr Res.
materials with superabsorbent polymers by means 2016;83:70–78. doi:10.1016/j.cemconres.2016.01.006.
of neutron radiography. Cem Concr Res. [336] Gruyaert E, Feiteira J, De Belie N, et al. Non-destruc-
2012;42:1113–1121. doi:10.1016/j.cemconres.2012. tive testing techniques to evaluate the healing
05.005. efficiency of self-healing concrete at lab-scale – KU
[323] Van Tittelboom K, Van Belleghem B, Boone MN, Leuven. 6th International Conference on Emerging
et al. X-ray radiography to visualize the rebar-cemen- Technologies in Nondestructive Testing,
titious matrix interface and judge the delay in cor- ETNDT2016. Brussels: CRC Press/Balkema; 2016.
rosion through self-repair by encapsulated [337] Xu J, Yao W. Multiscale mechanical quantification of
polyurethane. Adv Mater Interfaces. self-healing concrete incorporating non-ureolytic
2018;5:1701021. doi:10.1002/admi.201701021. bacteria-based healing agent. Cem Concr Res.
[324] Huang H, Ye G, Pel L. New insights into autogenous 2014;64:1–10. doi:10.1016/j.cemconres.2014.06.003.
self-healing in cement paste based on nuclear mag- [338] De Nardi C, Bullo S, Ferrara L, et al. Effectiveness of
netic resonance (NMR) tests. Mater Struct. crystalline admixtures and lime/cement coated gran-
2016;49:2509–2524. doi:10.1617/s11527-015-0664-9. ules in engineered self-healing capacity of lime mor-
[325] Snoeck D, Pel L, De Belie N. Autogenous healing in tars. Mater Struct. 2017;50:191. doi:10.1617/s11527-
cementitious materials with superabsorbent poly- 017-1053-3.
mers quantified by means of NMR. Sci Rep. [339] Lefever G, Snoeck D, De Belie N, et al. The contri-
2020;10:642. doi:10.1038/s41598-020-57555-0. bution of elastic wave NDT to the characterization
[326] Mechtcherine V, Lieboldt M. Permeation of water of modern cementitious media. Sensors.
and gases through cracked textile reinforced con- 2020;20:2959. doi:10.3390/s20102959.
crete. Cem Concr Compos. 2011;33:725–734. [340] Herbert EN, Li VC. Self-healing of Engineered
doi:10.1016/j.cemconcomp.2011.04.001. Cementitious Composites in the Natural
[327] RILEM TC116-PCD. Recommendation of TC 116- Environment. 2012. p. 155–162. doi:10.1007/978-
PCD: tests for gas permeability of concrete – precon- 94-007-2436-5_19.
ditioning of concrete test specimens for the measure- [341] Williams SL, Sakib N, Kirisits MJ, et al. Flexural
ment of gas permeability and capillary absorption of strength recovery induced by vegetative bacteria
water – measurement of the gas permeability of con- added to mortar. ACI Mater J. 2016;113:523–531.
crete by the RILEM – CE; 1999. doi:10.14359/51688831.
602 V. CAPPELLESSO ET AL.

[342] De Nardi C, Cecchi A, Ferrara L, et al. Effect of [356] Robayo-Salazar R, Mejía-Arcila J, Mejía de Gutiérrez
age and level of damage on the autogenous heal- R, et al. Life cycle assessment (LCA) of an alkali-acti-
ing of lime mortars. Compos Part B Eng. vated binary concrete based on natural volcanic poz-
2017;124:144–157. doi:10.1016/j.compositesb.201 zolan: a comparative analysis to OPC concrete.
7.05.041. Constr Build Mater. 2018;176:103–111. doi:10.1016/
[343] Wang J, Mignon A, Snoeck D, et al. Application of j.conbuildmat.2018.05.017.
modified-alginate encapsulated carbonate producing [357] Colangelo F, Forcina A, Farina I, et al. Life cycle
bacteria in concrete: a promising strategy for crack assessment (LCA) of different kinds of concrete con-
self-healing. Front Microbiol. 2015;6:1088. doi:10. taining waste for sustainable construction. Buildings.
3389/fmicb.2015.01088. 2018;8. doi:10.3390/buildings8050070.
[344] Kan LL, Shi HS, Sakulich AR, et al. Self-healing [358] Colangelo F, Petrillo A, Cioffi R, et al. Life cycle
characterization of engineered cementitious compo- assessment of recycled concretes: a case study in
site materials. ACI Mater J. 2010;107:617–624. southern Italy. Sci Total Environ. 2018;615:1506–
doi:10.14359/51664049. 1517. doi:10.1016/j.scitotenv.2017.09.107.
[345] Yang Y, Lepech MD, Yang EH, et al. Autogenous [359] Kurda R, Silvestre JD, de Brito J. Life cycle assess-
healing of engineered cementitious composites ment of concrete made with high volume of recycled
under wet-dry cycles. Cem Concr Res. 2009;39:382– concrete aggregates and fly ash. Resour Conserv
390. doi:10.1016/j.cemconres.2009.01.013. Recycl. 2018;139:407–417. doi:10.1016/j.resconrec.
[346] Cuenca E, Ferrara L. Fracture toughness parameters 2018.07.004.
to assess crack healing capacity of fiber reinforced [360] Szabó L, Hidalgo I, Ciscar JC, et al. CO2 emission
concrete under repeated cracking-healing cycles. trading within the European Union and Annex B
Theor Appl Fract Mech. 2020;106:102468. doi:10. countries: the cement industry case. Energy Policy.
1016/j.tafmec.2019.102468. 2006;34:72–87. doi:10.1016/j.enpol.2004.06.003.
[347] Granger S, Loukili A, Pijaudier-Cabot G, et al. [361] McLellan BC, Williams RP, Lay J, et al. Costs and car-
Experimental characterization of the self-healing of bon emissions for geopolymer pastes in comparison
cracks in an ultra high performance cementitious to ordinary Portland cement. J Clean Prod.
material: mechanical tests and acoustic emission 2011;19:1080–1090. doi:10.1016/j.jclepro.2011.02.
analysis. Cem Concr Res. 2007;37:519–527. doi:10. 010.
1016/j.cemconres.2006.12.005. [362] Turner LK, Collins FG. Carbon dioxide equivalent
[348] Van Tittelboom K, De Belie N, Lehmann F, et al. Use (CO 2-e) emissions: a comparison between geopoly-
of acoustic emission analysis to evaluate the self-heal- mer and OPC cement concrete. Constr Build Mater.
ing capability of concrete. In: Nondestructive testing 2013;43:125–130. doi:10.1016/j.conbuildmat.2013.
of materials and structures. Dordrecht: Springer 01.023.
Netherlands, 2013. p. 51–57. doi:10.1007/978-94- [363] Ouellet-Plamondon C, Habert G. Life cycle assess-
007-0723-8_7. ment (LCA) of alkali-activated cements and con-
[349] Van Tittelboom K, De Belie N, Lehmann F, et al. cretes; 2015. doi:10.1533/9781782422884.5.663.
Acoustic emission analysis for the quantification of [364] Babu VS, Mullick AK, Jain KK, et al. Mechanical
autonomous crack healing in concrete. Constr properties of high strength concrete with processed
Build Mater. 2012;28:333–341. doi:10.1016/j. recycled aggregate – influence of mixing techniques.
conbuildmat.2011.08.079. Indian Concr J. 2014;88:42–56. Available from:
[350] fib international Federation. fib Model Code, 2010. https://www.scopus.com/inward/record.uri?eid=2-
[351] Al-Obaidi S, Bamonte P, Luchini M, et al. Durability- s2.0-84908012436&partnerID=40&md5=
based design of structures made with ultra-high-per- 88c64e22932e3b72c9bd59f862e8d44a.
formance/ultra-high-durability concrete in extremely [365] Senaratne S, Gerace D, Mirza O, et al. The costs and
aggressive scenarios: application to a geothermal benefits of combining recycled aggregate with steel
water basin case study. Infrastructures. 2020;5:102. fibres as a sustainable, structural material. J Clean
doi:10.3390/infrastructures5110102. Prod. 2016;112:2318–2327. doi:10.1016/j.jclepro.
[352] Al-Obaidi S, Bamonte P, Animato F, et al. Innovative 2015.10.041.
design concept of cooling water tanks/basins in [366] Kline J, Kline C. Cement and CO2: what is happen-
geothermal power plants using ultra-high-perform- ing. IEEE Trans Ind Appl. 2015;51:1289–1294.
ance fiber-reinforced concrete with enhanced dura- doi:10.1109/TIA.2014.2339396.
bility. Sustainability. 2021;13:9826. doi:10.3390/ [367] Hache E, Simoën M, Seck GS, et al. The impact of
su13179826. future power generation on cement demand: an
[353] di Summa D, Tenório Filho JR, Snoeck D, et al. international and regional assessment based on cli-
Environmental and economic sustainability of crack mate scenarios. Int Econ. 2020;163:114–133. doi:10.
mitigation in reinforced concrete with superabsor- 1016/j.inteco.2020.05.002.
bent polymers (SAPs). J Clean Prod. [368] Van Den Heede P, De Belie N. Environmental impact
2022;358:131998. doi:10.1016/j.jclepro.2022.131998. and life cycle assessment (LCA) of traditional and
[354] Anastasiou EK, Liapis A, Papayianni I. Comparative “green” concretes: literature review and theoretical
life cycle assessment of concrete road pavements calculations. Cem Concr Compos. 2012;34:431–442.
using industrial by-products as alternative materials. doi:10.1016/j.cemconcomp.2012.01.004.
Resour Conserv Recycl. 2015;101:1–8. doi:10.1016/j. [369] Petek Gursel A, Masanet E, Horvath A, et al. Life-
resconrec.2015.05.009. cycle inventory analysis of concrete production: a
[355] Eberhardt LCM, Birgisdóttir H, Birkved M. Life cycle critical review. Cem Concr Compos. 2014;51:38–48.
assessment of a Danish office building designed for doi:10.1016/j.cemconcomp.2014.03.005.
disassembly. Build Res Inf. 2019;47:666–680. doi:10. [370] Damineli BL, Kemeid FM, Aguiar PS, et al.
1080/09613218.2018.1517458. Measuring the eco-efficiency of cement use. Cem
 INTERNATIONAL MATERIALS REVIEWS 603

Concr Compos. 2010;32:555–562. doi:10.1016/j. Clean Prod. 2016;135:1085–1097. doi:10.1016/j.

cemconcomp.2010.07.009. jclepro.2016.06.164.
[371] Hafez H, Cheung WM, Nagaratnam B, et al. A pro- [377] Hafez H, Kurda R, Cheung WM, et al. A systematic
posed performance based approach for life cycle review of the discrepancies in life cycle assessments
assessment of reinforced blended cement concrete. of green concrete. Appl Sci. 2019;9. doi:10.3390/
Proceedings of the 5th Sustainable Construction app9224803.
Materials and Technology; 2019. [378] Anand CK, Amor B. Recent developments, future chal-
[372] De Schepper M, Van den Heede P, Van Driessche I, lenges and new research directions in LCA of buildings:
et al. Life cycle assessment of completely recyclable a critical review. Renew Sustain Energy Rev.
concrete. Materials. 2014;7:6010–6027. doi:10.3390/ 2017;67:408–416. doi:10.1016/j.rser.2016.09.058.
ma7086010. [379] Jolliet O, Margni M, Charles R, et al. IMPACT 2002+:
[373] Wu P, Xia B, Zhao X. The importance of use and a new life cycle impact assessment methodology. Int J
end-of-life phases to the life cycle greenhouse gas Life Cycle Assess. 2003;8:324–330. doi:10.1007/
(GHG) emissions of concrete – a review. Renew BF02978505.
Sustain Energy Rev. 2014;37:360–369. doi:10.1016/j. [380] Benetto E, Rousseaux P, Blondin J. Life cycle assess-
rser.2014.04.070. ment of coal by-products based electric power pro-
[374] Caruso MC, Pascale C, Camacho E, et al. Life cycle duction scenarios. Fuel. 2004;83:957–970. doi:10.
assessment on the use of ultra high performance 1016/S0016-2361(03)00258-8.
fibre reinforced concretes with enhanced durability [381] Teall O. Crack closure and enhanced autogenous
for structures in extremely aggressive environments: healing of structural concrete using shape memory
case study analyses. RILEM Bookseries, 2020. p. 121– polymers. Cardiff: Cardiff University; 2016.
125. doi:10.1007/978-3-030-43332-1_24. [382] Caruso M, Pinho R, Bianchi F, et al. A life cycle fra-
[375] di Summa D, Parpanesi M, Ferrara L, et al. Life cycle mework for the identification of optimal building
assessment (LCA) and life cycle cost (LCC) analysis renovation strategies considering economic and
as crucial part of a holistic approach to the design environmental impacts. Sustainability.
of structures with advanced cement based materials. 2020;12:10221. doi:10.3390/su122310221.
Proceedings of International Conference on [383] Caruso MC, Pascale C, Camacho E, et al.
Regeneration and Conservation of Structures, Sept.; Comparative environmental and social life cycle
2022. assessments of off-shore aquaculture rafts made in
[376] Piccinno F, Hischier R, Seeger S, et al. From labora- ultra-high performance concrete (UHPC). Int J Life
tory to industrial scale: a scale-up framework for Cycle Assess. 2022;27:281–300. doi:10.1007/s11367-
chemical processes in life cycle assessment studies. J 021-02017-6.