03-05-2012, 01:45 PM
Self healing in polymers and polymer composites. Concepts, realization and outlook: A review
[attachment=21164]
Abstract.
Formation of microcracks is a critical problem in polymers and polymer composites during their service in structural
applications. Development and coalescence of microcracks would bring about catastrophic failure of the materials and
then reduce their lifetimes. Therefore, early sensing, diagnosis and repair of microcracks become necessary for removing
the latent perils. In this context, the materials possessing self-healing function are ideal for long-term operation. Self-repairing
polymers and polymer composites have attracted increasing research interests. Attempts have been made to develop
solutions in this field. The present article reviews state-of-art of the achievements on the topic. According to the ways of
healing, the smart materials are classified into two categories: (i) intrinsic self-healing ones that are able to heal cracks by
the polymers themselves, and (ii) extrinsic in which healing agent has to be pre-embedded. The advances in this field show
that selection and optimization of proper repair mechanisms are prerequisites for high healing efficiency. It is a challenging
job to either invent new polymers with inherent crack repair capability or integrate existing materials with novel healing
system.
Keywords: smart polymers, polymer composites, self-healing, cracks
eXPRESS Polymer Letters Vol.2, No.4 (2008) 238–250
Available online at www.expresspolymlett.com
DOI: 10.3144/expresspolymlett.2008.29
cification into fibrous bone and lamellar bone.
Clearly, the natural healing in living bodies
depends on rapid transportation of repair substance
to the injured part and reconstruction of the tissues.
Having been inspired by these findings, continuous
efforts are now being made to mimic natural materials
and to integrate self-healing capability into
polymers and polymer composites. The progress
has opened an era of new intelligent materials.
On the whole, researches in this field are still in the
infancy. More and more scientists and companies
are interested in different aspects of the topic. Innovative
measures and new knowledge of the related
mechanisms are constantly emerging. Therefore, it
might be the right time to review the attempts carried
out so far in different laboratories in the world.
According to the ways of healing, self-healing
polymers and polymer composites can be classified
into two categories: (i) intrinsic ones that are able to
heal cracks by the polymers themselves, and (ii)
extrinsic in which healing agent has to be preembedded.
2. Intrinsic self-healing
The so-called intrinsic self-healing polymers and
polymer composites are based on specific performance
of the polymers and polymeric matrices that
enables crack healing under certain stimulation
(mostly heating). Autonomic healing without external
intervention is not available in these materials
for the time being. As viewed from the predominant
molecular mechanisms involved in the healing
processes, the reported achievements consist of two
modes: (i) physical interactions, and (ii) chemical
interactions.
2.1. Self-healing based on physical interactions
Compared to the case of thermosetting polymers,
crack healing in thermoplastic polymers received
more attention at an earlier time. Wool and coworkers
systematically studied the theory involved
[7, 8]. They pointed out that the healing process
goes through five phases: (i) surface rearrangement,
which affects initial diffusion function and
topological feature; (ii) surface approach, related to
healing patterns; (iii) wetting, (iv) diffusion, the
main factor that controls recovery of mechanical
properties, and (v) randomization, ensuring disappearance
of cracking interface. In addition, Kim
and Wool [9] proposed a microscopic model for the
last two phases on basis of reptation model that
describes longitudinal chain diffusion responsible
for crack healing.
Accordingly, Jud and Kaush [10] tested crack-healing
behavior in a series of poly(methyl methacrylate)
(PMMA) and poly(methyl methacrylate-comethyl
ethylacrylate) (MMA-MEA copolymer)
samples of different molecular weights and degrees
of copolymerization. They induced crack healing
by heating samples above the glass transition temperature
under slight pressure. It was found that full
resistance was regained during short term loading
experiments. The establishment of mechanical
strength should result from interdiffusion of chains
and formation of entanglements for the glassy polymer
[11]. Wool [12] further suggested that the
recovery of fracture stress is proportional to t1/4
(where t is the period of heating treatment). Jud
et al. [13] also performed re-healing and welding of
glassy polymers (PMMA and styrene-acrylonitrile
copolymer (SAN)) at temperatures above the glass
transition temperatures, and found that the fracture
toughness, Kli, in the interface increased with contact
time, t, as Kli ∝ t1/4 as predicted by the diffusion
model.
It is worth noting that whereas craze healing occurs
at temperature above and below the glass transition
temperature [14], crack healing happens only at or
above the glass transition temperature [15]. In order
to reduce the effective glass transition temperature
of PMMA, Lin et al. [16] and Wang et al. [17]
treated PMMA with methanol and ethanol, respectively.
They reduced the glass transition temperature
to a range of 40~60°C, and found that there
were two distinctive stages for crack healing: the
first one corresponding to the progressive healing
due to wetting, while the second related to diffusion
enhancement of the quality of healing behavior.
Besides simple heating induced healing, thermomechanical
healing is valid for some specific polymers,
like poly(ethylene-co-methacrylic acid)
(EMAA) copolymers [18]. EMAA films prove to
be able to heal upon ballistic puncture and sawing
damages. This occurs through a heat generating
frictional process, which heats the polymer to the
viscoelastic melt state and provides the ability to
rebond and repair damage. In contrast, low speed
239
Yuan et al. – eXPRESS Polymer Letters Vol.2, No.4 (2008) 238–250
friction event fails to produce sufficient thermal
energy favorable to healing. As a result, thermomechanical
healing is not active in the material.
Unlike thermoplastics, heating induced healing of
thermosetting polymers depends on crosslinking of
unreacted groups. Healing of epoxy, for instance,
has to proceed above the glass transition temperature
[19]. Then, the molecules at the cracking surfaces
would interdiffuse and the residual functional
groups react with each other. A 50% recovery of
impact strength can thus be obtained [20]. During
the repair study of vinyl ester resin, Raghavan and
Wool reported critical strain energy release rate,
GIC, for the interfaces after crack healing (i.e.
annealing above the glass transition temperature) is
1.7% of the virgin value. Lower crosslink density
favors the repair effect [21].
Thermoplastic/thermosetting semi-interpenetrating
network is factually a material associated with
repeatable self-healing ability. The group of Jones
introduced a soluble linear polymer to a thermosetting
epoxy resin [22–24]. The selected thermoplastic
is poly(bisphenol-A-co-epichlorohydrin), which
is highly compatible with the matrix diglycidyl
ether of bisphenol-A based resin. Upon heating a
fractured resin system, the thermoplastic material
would mobilize and diffuse through the thermosetting
matrix, with some chains bridging closed
cracks and thereby facilitating healing. When this
healable resin was compounded with crossply glass
fiber, effective healing of composites transverse
cracks and delamination has been demonstrated.
The requirements for such thermal diffusion of a
healing agent were summarized as follows [23]. (i)
The healing agent should be reversibly bonded (e.g.
through hydrogen bonding) to the crosslinked network
of the cured resin below the minimum healing
temperature to limit its effect on thermomechanical
properties. (ii) The healing agent should become
mobile above this minimum healing temperature so
that it can diffuse across a hairline crack, such as a
transverse crack, to provide a recovery in strength.
(iii) The addition of the linear chain molecule
should not significantly reduce the thermomechanical
properties of the resin matrix.
2.2. Self-healing based on chemical interactions
In fact, cracks and strength decay might be caused
by structural changes of atoms or molecules, like
chain scission. Therefore, inverse reaction, i.e.
recombination of the broken molecules, should be
one of the repairing strategies. Such method does
not focus on cracks healing but on ‘nanoscopic’
deterioration. One example is polycarbonate (PC)
synthesized by ester exchange method. The PCs
were treated in a steam pressure cabin at 120°C
prior to the repair [25]. As a result, molecular
weight of the PCs dropped by about 88 to 90%.
After drying them in a vacuum cabin, the repairing
treatment was done in an oven at 130°C with N2
atmosphere under reduced pressure. The reduced
tensile strength due to the deterioration treatment
can thus be gradually recovered. The repairing
mechanism was considered as the following procedures.
Firstly the carbonate bond was cut by
hydrolysis, and then the concentration of the phenoxy
end increased after deterioration. The (–OH)
end-group on the chain was substituted by sodium
ion. The (–ONa) end might attack a carbonate bond
at the end of one of the other chains, leading to
recombination of these two chains with the elimination
of the phenol from PC (Figure 1). The
repairing reaction was accelerated by weak alkaline,
such as sodium carbonate. It suggested that
two conditions are required for the PC to re-combine
the polymer chains. One is the chemical structure
of the chain end and the other is the catalyst
(Na2CO3) for acceleration of the reaction.
240
Yuan et al. – eXPRESS Polymer Letters Vol.2, No.4 (2008) 238–250
Figure 1. Hydrolysis and recombination reaction of PCs
with the catalyst of NaCO3
Another example is poly(phenylene ether) (PPE) in
which the repairing agent was regenerated by oxygen
[26]. The polymer chain of the PPE was cut by
a deterioration factor (such as heat, light, and external
mechanical force) to produce a radical on the
end of the scission chain. Subsequently, a hydrogen
donor stabilized the radical. The catalyst existing in
the system, Cu (II), would react with each end of
the scission chains to form a complex. Then, the
chains combined by eliminating two protons from
the ends, and the copper changed from Cu (II) to
Cu (I). Afterward, two Cu (I) reacted with an oxygen
molecule to be oxidized to Cu (II), and an oxygen
ion reacted with two protons to form a water
molecule that evaporated from the specimen.
The above examples show that PC or PPE might be
probably designed as a self-repairing material by
means of the reversible reaction. The deterioration
is expected to be minimized if the recovery rate is
the same as the deterioration rate. However, the
systems in these studies are not sufficient for construction
of real self-repairing composites because
the recovery of the broken molecules needs higher
temperature and other rigorous conditions. A much
more effective catalyst should be found, which is
able to active the recombination of degraded
oligomers at room temperature.
Thermally reversible crosslinking behavior has
been known for quite a while. Wudl et al. combined
this with the concept of ‘self-healing’ in making
healable polymers [27, 28]. They synthesized
highly cross-linked polymeric materials with multifuran
and multi-maleimide via Diels-Alder (DA)
reaction. At temperatures above 120°C, the ‘intermonomer’
linkages disconnect (corresponding to
retro-DA reaction) but then reconnect upon cooling
(i.e. DA reaction). This process is fully reversible
and can be used to restore fractured parts of the
polymers. The polymers are transparent and possess
mechanical properties comparable to commercial
epoxy and unsaturated polyester. In principle,
an infinite number of crack healing is available
without the aid of additional catalysts, monomers
and special surface treatment.
In a latter work by Liu and Hsieh [29], Wudl’s
approach was modified. The multifunctional furan
and maleimide compounds were prepared in simple
routes, using epoxy compounds as precursors. The
furan and maleimide monomers could be therefore
considered as epoxy-based compounds, so as to
incorporate the advantage characteristics of epoxy
resins, including solvent and chemical resistance,
thermal and electrical characters, and good adherence,
to their corresponding cured polymers.
Besides, Liu and Chen prepared polyamides possessing
furan pendent groups (PA-F) from reacting
furfuryl amine with maleimide containing
polyamides (PA-MI) via a Michael addition reaction
[30]. Thermally reversible cross-linked polyamides
were obtained from PA-MI and PA-F
polyamides by means of DA and retro-DA reactions.
The thermally reversible cross-linked polyamides
also exhibited a self-repairing property as
well as the ability of mechanical property recovery.
To quantify the degree of structural restoration
after damages have been repaired, characterization
of healing efficiency is necessary but no specific
testing standard is available now. Different testing
procedures sometimes give different results [23].
When Wudl’s group measured healing efficiency
of their thermally reversible crosslinked polymers,
fracture toughness from compact tension (CT) tests
was used [27]. Values for the original and healed
fracture toughness were determined by the propagation
of the starter crack along the middle plane of
the specimen at the critical load. In consideration of
the difficulties in (i) precise registration of the fracture
surface and (ii) protection of pre-notching,
Plaisted and Nemat-Nasser [31] applied double
cleavage drilled compression (DCDC) to evaluate
mending efficiency of the reversibly cross-linked
polymer based on Diels-Alder cycloaddition. The
testing geometry allowed for controlled incremental
crack growth so that the cracked sample
remained in one piece after the test, improving ability
to realign the fracture surfaces prior to healing.
[attachment=21164]
Abstract.
Formation of microcracks is a critical problem in polymers and polymer composites during their service in structural
applications. Development and coalescence of microcracks would bring about catastrophic failure of the materials and
then reduce their lifetimes. Therefore, early sensing, diagnosis and repair of microcracks become necessary for removing
the latent perils. In this context, the materials possessing self-healing function are ideal for long-term operation. Self-repairing
polymers and polymer composites have attracted increasing research interests. Attempts have been made to develop
solutions in this field. The present article reviews state-of-art of the achievements on the topic. According to the ways of
healing, the smart materials are classified into two categories: (i) intrinsic self-healing ones that are able to heal cracks by
the polymers themselves, and (ii) extrinsic in which healing agent has to be pre-embedded. The advances in this field show
that selection and optimization of proper repair mechanisms are prerequisites for high healing efficiency. It is a challenging
job to either invent new polymers with inherent crack repair capability or integrate existing materials with novel healing
system.
Keywords: smart polymers, polymer composites, self-healing, cracks
eXPRESS Polymer Letters Vol.2, No.4 (2008) 238–250
Available online at www.expresspolymlett.com
DOI: 10.3144/expresspolymlett.2008.29
cification into fibrous bone and lamellar bone.
Clearly, the natural healing in living bodies
depends on rapid transportation of repair substance
to the injured part and reconstruction of the tissues.
Having been inspired by these findings, continuous
efforts are now being made to mimic natural materials
and to integrate self-healing capability into
polymers and polymer composites. The progress
has opened an era of new intelligent materials.
On the whole, researches in this field are still in the
infancy. More and more scientists and companies
are interested in different aspects of the topic. Innovative
measures and new knowledge of the related
mechanisms are constantly emerging. Therefore, it
might be the right time to review the attempts carried
out so far in different laboratories in the world.
According to the ways of healing, self-healing
polymers and polymer composites can be classified
into two categories: (i) intrinsic ones that are able to
heal cracks by the polymers themselves, and (ii)
extrinsic in which healing agent has to be preembedded.
2. Intrinsic self-healing
The so-called intrinsic self-healing polymers and
polymer composites are based on specific performance
of the polymers and polymeric matrices that
enables crack healing under certain stimulation
(mostly heating). Autonomic healing without external
intervention is not available in these materials
for the time being. As viewed from the predominant
molecular mechanisms involved in the healing
processes, the reported achievements consist of two
modes: (i) physical interactions, and (ii) chemical
interactions.
2.1. Self-healing based on physical interactions
Compared to the case of thermosetting polymers,
crack healing in thermoplastic polymers received
more attention at an earlier time. Wool and coworkers
systematically studied the theory involved
[7, 8]. They pointed out that the healing process
goes through five phases: (i) surface rearrangement,
which affects initial diffusion function and
topological feature; (ii) surface approach, related to
healing patterns; (iii) wetting, (iv) diffusion, the
main factor that controls recovery of mechanical
properties, and (v) randomization, ensuring disappearance
of cracking interface. In addition, Kim
and Wool [9] proposed a microscopic model for the
last two phases on basis of reptation model that
describes longitudinal chain diffusion responsible
for crack healing.
Accordingly, Jud and Kaush [10] tested crack-healing
behavior in a series of poly(methyl methacrylate)
(PMMA) and poly(methyl methacrylate-comethyl
ethylacrylate) (MMA-MEA copolymer)
samples of different molecular weights and degrees
of copolymerization. They induced crack healing
by heating samples above the glass transition temperature
under slight pressure. It was found that full
resistance was regained during short term loading
experiments. The establishment of mechanical
strength should result from interdiffusion of chains
and formation of entanglements for the glassy polymer
[11]. Wool [12] further suggested that the
recovery of fracture stress is proportional to t1/4
(where t is the period of heating treatment). Jud
et al. [13] also performed re-healing and welding of
glassy polymers (PMMA and styrene-acrylonitrile
copolymer (SAN)) at temperatures above the glass
transition temperatures, and found that the fracture
toughness, Kli, in the interface increased with contact
time, t, as Kli ∝ t1/4 as predicted by the diffusion
model.
It is worth noting that whereas craze healing occurs
at temperature above and below the glass transition
temperature [14], crack healing happens only at or
above the glass transition temperature [15]. In order
to reduce the effective glass transition temperature
of PMMA, Lin et al. [16] and Wang et al. [17]
treated PMMA with methanol and ethanol, respectively.
They reduced the glass transition temperature
to a range of 40~60°C, and found that there
were two distinctive stages for crack healing: the
first one corresponding to the progressive healing
due to wetting, while the second related to diffusion
enhancement of the quality of healing behavior.
Besides simple heating induced healing, thermomechanical
healing is valid for some specific polymers,
like poly(ethylene-co-methacrylic acid)
(EMAA) copolymers [18]. EMAA films prove to
be able to heal upon ballistic puncture and sawing
damages. This occurs through a heat generating
frictional process, which heats the polymer to the
viscoelastic melt state and provides the ability to
rebond and repair damage. In contrast, low speed
239
Yuan et al. – eXPRESS Polymer Letters Vol.2, No.4 (2008) 238–250
friction event fails to produce sufficient thermal
energy favorable to healing. As a result, thermomechanical
healing is not active in the material.
Unlike thermoplastics, heating induced healing of
thermosetting polymers depends on crosslinking of
unreacted groups. Healing of epoxy, for instance,
has to proceed above the glass transition temperature
[19]. Then, the molecules at the cracking surfaces
would interdiffuse and the residual functional
groups react with each other. A 50% recovery of
impact strength can thus be obtained [20]. During
the repair study of vinyl ester resin, Raghavan and
Wool reported critical strain energy release rate,
GIC, for the interfaces after crack healing (i.e.
annealing above the glass transition temperature) is
1.7% of the virgin value. Lower crosslink density
favors the repair effect [21].
Thermoplastic/thermosetting semi-interpenetrating
network is factually a material associated with
repeatable self-healing ability. The group of Jones
introduced a soluble linear polymer to a thermosetting
epoxy resin [22–24]. The selected thermoplastic
is poly(bisphenol-A-co-epichlorohydrin), which
is highly compatible with the matrix diglycidyl
ether of bisphenol-A based resin. Upon heating a
fractured resin system, the thermoplastic material
would mobilize and diffuse through the thermosetting
matrix, with some chains bridging closed
cracks and thereby facilitating healing. When this
healable resin was compounded with crossply glass
fiber, effective healing of composites transverse
cracks and delamination has been demonstrated.
The requirements for such thermal diffusion of a
healing agent were summarized as follows [23]. (i)
The healing agent should be reversibly bonded (e.g.
through hydrogen bonding) to the crosslinked network
of the cured resin below the minimum healing
temperature to limit its effect on thermomechanical
properties. (ii) The healing agent should become
mobile above this minimum healing temperature so
that it can diffuse across a hairline crack, such as a
transverse crack, to provide a recovery in strength.
(iii) The addition of the linear chain molecule
should not significantly reduce the thermomechanical
properties of the resin matrix.
2.2. Self-healing based on chemical interactions
In fact, cracks and strength decay might be caused
by structural changes of atoms or molecules, like
chain scission. Therefore, inverse reaction, i.e.
recombination of the broken molecules, should be
one of the repairing strategies. Such method does
not focus on cracks healing but on ‘nanoscopic’
deterioration. One example is polycarbonate (PC)
synthesized by ester exchange method. The PCs
were treated in a steam pressure cabin at 120°C
prior to the repair [25]. As a result, molecular
weight of the PCs dropped by about 88 to 90%.
After drying them in a vacuum cabin, the repairing
treatment was done in an oven at 130°C with N2
atmosphere under reduced pressure. The reduced
tensile strength due to the deterioration treatment
can thus be gradually recovered. The repairing
mechanism was considered as the following procedures.
Firstly the carbonate bond was cut by
hydrolysis, and then the concentration of the phenoxy
end increased after deterioration. The (–OH)
end-group on the chain was substituted by sodium
ion. The (–ONa) end might attack a carbonate bond
at the end of one of the other chains, leading to
recombination of these two chains with the elimination
of the phenol from PC (Figure 1). The
repairing reaction was accelerated by weak alkaline,
such as sodium carbonate. It suggested that
two conditions are required for the PC to re-combine
the polymer chains. One is the chemical structure
of the chain end and the other is the catalyst
(Na2CO3) for acceleration of the reaction.
240
Yuan et al. – eXPRESS Polymer Letters Vol.2, No.4 (2008) 238–250
Figure 1. Hydrolysis and recombination reaction of PCs
with the catalyst of NaCO3
Another example is poly(phenylene ether) (PPE) in
which the repairing agent was regenerated by oxygen
[26]. The polymer chain of the PPE was cut by
a deterioration factor (such as heat, light, and external
mechanical force) to produce a radical on the
end of the scission chain. Subsequently, a hydrogen
donor stabilized the radical. The catalyst existing in
the system, Cu (II), would react with each end of
the scission chains to form a complex. Then, the
chains combined by eliminating two protons from
the ends, and the copper changed from Cu (II) to
Cu (I). Afterward, two Cu (I) reacted with an oxygen
molecule to be oxidized to Cu (II), and an oxygen
ion reacted with two protons to form a water
molecule that evaporated from the specimen.
The above examples show that PC or PPE might be
probably designed as a self-repairing material by
means of the reversible reaction. The deterioration
is expected to be minimized if the recovery rate is
the same as the deterioration rate. However, the
systems in these studies are not sufficient for construction
of real self-repairing composites because
the recovery of the broken molecules needs higher
temperature and other rigorous conditions. A much
more effective catalyst should be found, which is
able to active the recombination of degraded
oligomers at room temperature.
Thermally reversible crosslinking behavior has
been known for quite a while. Wudl et al. combined
this with the concept of ‘self-healing’ in making
healable polymers [27, 28]. They synthesized
highly cross-linked polymeric materials with multifuran
and multi-maleimide via Diels-Alder (DA)
reaction. At temperatures above 120°C, the ‘intermonomer’
linkages disconnect (corresponding to
retro-DA reaction) but then reconnect upon cooling
(i.e. DA reaction). This process is fully reversible
and can be used to restore fractured parts of the
polymers. The polymers are transparent and possess
mechanical properties comparable to commercial
epoxy and unsaturated polyester. In principle,
an infinite number of crack healing is available
without the aid of additional catalysts, monomers
and special surface treatment.
In a latter work by Liu and Hsieh [29], Wudl’s
approach was modified. The multifunctional furan
and maleimide compounds were prepared in simple
routes, using epoxy compounds as precursors. The
furan and maleimide monomers could be therefore
considered as epoxy-based compounds, so as to
incorporate the advantage characteristics of epoxy
resins, including solvent and chemical resistance,
thermal and electrical characters, and good adherence,
to their corresponding cured polymers.
Besides, Liu and Chen prepared polyamides possessing
furan pendent groups (PA-F) from reacting
furfuryl amine with maleimide containing
polyamides (PA-MI) via a Michael addition reaction
[30]. Thermally reversible cross-linked polyamides
were obtained from PA-MI and PA-F
polyamides by means of DA and retro-DA reactions.
The thermally reversible cross-linked polyamides
also exhibited a self-repairing property as
well as the ability of mechanical property recovery.
To quantify the degree of structural restoration
after damages have been repaired, characterization
of healing efficiency is necessary but no specific
testing standard is available now. Different testing
procedures sometimes give different results [23].
When Wudl’s group measured healing efficiency
of their thermally reversible crosslinked polymers,
fracture toughness from compact tension (CT) tests
was used [27]. Values for the original and healed
fracture toughness were determined by the propagation
of the starter crack along the middle plane of
the specimen at the critical load. In consideration of
the difficulties in (i) precise registration of the fracture
surface and (ii) protection of pre-notching,
Plaisted and Nemat-Nasser [31] applied double
cleavage drilled compression (DCDC) to evaluate
mending efficiency of the reversibly cross-linked
polymer based on Diels-Alder cycloaddition. The
testing geometry allowed for controlled incremental
crack growth so that the cracked sample
remained in one piece after the test, improving ability
to realign the fracture surfaces prior to healing.