Cement and Concrete Research 73 (2015) 1–16

Contents lists available at ScienceDirect

Cement and Concrete Research

journal homepage: http://ees.elsevier.com/CEMCON/default.asp

Degradation mechanisms of natural fiber in the matrix of

cement composites
Jianqiang Wei ⁎, Christian Meyer
Department of Civil Engineering and Engineering Mechanics, Columbia University, NY, 10027, USA

a r t i c l e i n f o a b s t r a c t

Article history: The degradation mechanisms of natural fiber in the alkaline and mineral-rich environment of cement matrix are
Received 14 April 2014 investigated. Cement hydration is presented to be a crucial factor in understanding fiber degradation behavior by
Accepted 18 February 2015 designing a contrast test to embed sisal fibers in pure and metakaolin modified cement matrices. In addition to
Available online 11 March 2015
durability of sisal fiber-reinforced cement composites determined by means of flexural properties, degradation
degree of the embedded fibers is directly evaluated by proposing a novel separation approach. The results indi-
Keywords:
Degradation (C)
cate that, by reducing alkalinity of pore solution, metakaolin effectively mitigates the deterioration of natural
Durability (C) fiber. By combining results of thermogravimetric analysis and microstructure, the alkali degradation process of
Blended cement (D) natural fiber, which consists of hydrolysis of lignin and hemicellulose, stripping of cellulose microfibrils and de-
Fiber reinforcement (E) terioration of amorphous regions in cellulose chains, is visually presented. Two new concepts of mineralization
mechanism, calcium hydroxide (CH)-mineralization and self-mineralization, are also proposed and quantitatively
characterized.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction resistance in alkaline environments, which means it loses strength

The increasing environmental concern and awareness of industrial
pollution have forced the construction and manufacturing industries
to search for innovative materials that are reliable, sustainable and can
replace conventional synthetic fibers as reinforcement of structural ma-
terials. Natural fibers, such as sisal, jute, cotton, flax, hemp, and kenaf,
have already been considered as potential alternatives, given their envi-
ronmental friendliness and ready availability in fibrous form and the
fact that they can be extracted from plant leaves at low cost [1]. Natural
fiber reinforced composites also offer environmental advantages such
as reduced dependence on non-renewable energy/material sources,
lower pollutant emissions, lower greenhouse gas emissions, enhanced
energy recovery, and biodegradability [2]. Cement and concrete products
are notable for their weakness in tension and low toughness which gives
rise to early cracking under impact loads or thermal shock, unless they
are reinforced. Fiber reinforcement can be used to mitigate these issues
[3]. Therefore, in the last two decades, considerable effort has been
directed towards the use of various natural fibers, which are available
in abundance in tropical and sub-tropical countries, as reinforcement of
cement composites for producing cost-effective construction materials
[4].
However, in spite of natural fibers' advantages, their usefulness in
cement based materials is limited by the relatively low degradation
⁎ Corresponding author at: 500 W. 120th St., Room 706D Mudd, New York, NY, 10027,
USA. Tel.: +1 347 421 1465.
E-mail addresses: jw2938@yahoo.com, jw2938@columbia.edu (J. Wei).
when used as reinforcement of a cement matrix exposed to aggressive
environmental conditions [5,6]. The low durability caused by degrada-
tion of natural fiber in cement's alkaline-mineral environment, which
depletes the reinforcing effect of fiber, is often regarded as a challenge
need be avoided. This has been the central issue that needs to be solved
before promoting the widespread application of natural fiber in various
composites.
Studies on the mechanical performance of natural fiber reinforced
cement based composite [7–18], the durability of natural fiber-cement
composite in different aging conditions [19–25], and treatment
methods for improving its durability [3,26–29] have been performed
in recent years. However, the investigation of the degradation mecha-
nisms of natural fiber in the complex alkaline and mineral-rich environ-
ment of cement matrices, as characterized by hydration process,
crystallization, setting and hardening, was reported rarely. Only a few
publications addressed the failure of natural fiber in cement composites.
Gram [30] and Filho [19] described the dissolution of lignin and hemi-
cellulose in a cement pore solution and alkaline hydrolysis of cellulose
molecules, which leads to the degradation of molecular chains and the
reduction in the degree of polymerization and the tensile strength
were the main aging mechanisms of natural fibers in cement based
composites. Besides, the crystallization of lime in the lumen and middle
lamellae also decreased the fiber's flexibility and strength [19,31]. Mohr
et al. [21] proposed a three-part progressive degradation mechanism
during wetting and drying cycles, the third part of which is the fiber
embrittlement due to fiber cell wall mineralization, which was
indicated by the lack of recovered toughness. Melo Filho et al. [32]

http://dx.doi.org/10.1016/j.cemconres.2015.02.019
0008-8846/© 2015 Elsevier Ltd. All rights reserved.
2 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16

investigated two fiber degradation mechanisms in Portland cement adsorption of calcium and hydroxyl ions. However, the results ob-
composites: fiber mineralization and degradation of cellulose, hemi- tained so far have been disappointing, primarily because the intrinsic
cellulose and lignin due to the existence of calcium hydroxide and degradation kinetics and mechanisms of natural fiber and its three

Cellulose
Secondary cell wall-S3 Lignin
Hemicellulose
Secondary cell wall-S2

Secondary cell wall-S1

Primary cell wall Cellulose

Hemicellulose
Middle lamella Pectin

Hemicellulose

Lignin

Cellulose

1,4- -D-polyglucose

Fig. 1. Microstructure, schematic diagram and molecular structures of natural fiber cell wall.
 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16 3

main components, cellulose, hemicellulose and lignin, in solid Table 1

phase and pore solution of cement have not been successfully Chemical and mineralogical compositions and readily soluble compounds of PC and MK.

investigated. Chemical and mineralogical compositions

As shown in Fig. 1, there are three distinct layers from outside to in- XRF-analysis [wt.%] Mineralogical phase composition of cement
side of a natural fibril: (i) the middle lamella containing pectin, hemicel- [wt.%]
MK PC
lulose and lignin; (ii) the primary cell wall containing hemicellulose and
cellulose; and (iii) the secondary cell (S1, S2 and S3) walls consisting CaO 0.0707 63.7 C3Sa 52.4
SiO2 51.8 12.9 C2Sb 16.0
mainly of cellulose. One single natural fiber is composed of several of c
Fe2O3 4.15 7.97 C3A 10.2
these elementary fibrils (typically 10–40), bound together by the pectin SO3 0.105 5.25 C4AF d
8.6
substances in the middle lamella [33,34]. Cellulose is a linear polymer Al2O3 42.4 4.14 Arcanite 2.5
made of glucose subunits linked by β-1,4 bonds, and the basic repeating MgO – 3.50 Anhydrite 2.3
unit is cellobiose [35]. Cellulose's stiffness and durability are supported SrO 0.0397 0.915 Hemihydrate 2.0
K2O 0.218 0.907 Syngenite 2.0
by their formation of rigid, insoluble micro-fibrils chains, integrated by
TiO2 1.07 0.279 Dolomite 1.2
the numerous intra- and intermolecular hydrogen bonds. The high ZnO – 0.251 Calcite 0.5
tensile strength of cellulose enables plant cells to withstand osmotic ZrO2 0.0884 0.119 Other 2.3
pressure and is responsible for the resistance of plants to mechanical Cl 0.0457 – Readily soluble alkalis of cement [wt.%]
stress [35]. The distinct cell wall structure endows the excellent me- Compton 0.86 0.98 Na2O 0.07
Rayleigh 1.24 1.19 K2O 0.85
chanical properties of natural fiber, with a low density, while it also
a
leads to the poor durability of the natural fibers in the matrix of cement, b
Tricalcium silicate.
Dicalcium silicate.
due to the low degradation resistance of lignin and hemicellulose in c
Tricalcium aluminate.
high alkali environments [19,36]. d
Tetra-calcium aluminoferrite.
In this paper, the durability of sisal fiber-reinforced cement compos-
ites exposed to wetting and drying cycling was investigated. In contrast
to previous works reported in the literature, a novel separation ap- 2.2. Mixture proportion and composite production
proach is proposed, which makes it possible to accurately determine
fiber's degradation rate by directly studying composition changes, and A water/cement (w/c) ratio of 0.4 and a binder-to-sand ratio of 1:1
mechanical and physical properties of the embedded sisal fibers. There- were used for all mortar mixtures. The moisture content and water
fore, an intuitive understanding of degradation mechanisms for sisal absorption capacity of the sisal fibers were considered when determin-
fiber in the cement matrix was developed. Degradation degree of sisal ing the water–solid ratio. The cement pastes were mixed with gradually
fiber was determined by analyzing surface morphology, crystallinity added sand in a mechanical mortar mixer at 60 rpm for 2 min followed
index, and thermal and uniaxial tensile characteristics of the fibers. by 1 min rest and 3 min of further mixing at 120 rpm. The workability
The alkali hydrolysis process of sisal fiber, which consists of four steps, of the mortar was adjusted when necessary using ADVA 408 super-
was presented according to the feature of natural fiber's distinct cell plasticizer up to 3 wt.% of binder. ADVA 408 is a kind of polycarboxylate
wall structure and its degree of degradation in each aging stage. superplasticizer admixture, which does not affect cement paste me-
The mineralization of cell wall, which has been presented in previous chanical strength at the early age of cement hydration [38].
studies, was redefined by proposed two novel concepts: CH- For each blend at each testing point, three 200 mm × 50 mm × 12 mm
mineralization and self-mineralization. Both of the two mineralization beams with 2% volume fraction of sisal fiber were cast for four-point flex-
mechanisms were parameterized by combing the results of tensile test- ural texts. A mortar layer of 2 mm thickness was placed at the bottom of
ing of the embedded fibers with precipitation amount of CH and crystal- the mold followed by a layer of long unidirectional aligned fibers. The
lization properties of cell wall, respectively. second layer of mortar of 3–4 mm thickness was placed, and a second
layer of sisal fiber was aligned distributed, followed by the third layer
of mortar to fill the mold. For compressive strength tests, 50 mm ×
2. Materials and methods 50 mm × 50 mm cubic samples were cast in steel molds. A vibration

2.1. Characterization of raw materials

0.1 1 10 100 1000
Sisal fiber produced from Tanzania was provided by Bast Fibers LLC 100
of Creskill, New Jersey. The cement used in this paper is ASTM Type I 6 Cement
MK 90
Portland cement (PC). Except the pure cement samples, for comparison,
Relative frequency (vol.%)

5 80
10 wt.% (MK10) and 30 wt.% (MK30) cements were replaced by
Passing volume (vol.%)

metakaolin (MK) to study the effect of the alkalinity of pore solution on 70

fiber degradation. The chemical compositions of the used cement and 4
60
metakaolin, as determined by X-ray fluorescence (XRF) are shown in
50
Table 1. Portland cement and concretes are essentially alkaline because 3
Portland cement reacts with water to produce solutions which are 40
oversaturated with Ca(OH)2 (CH), but there are also small and variable 2 30
percentages of Na2O and K2O as minor components in cement minerals 20
or as sulfates, which readily form NaOH and KOH solutions as the cement 1
10
aluminate and ferrite phases hydrate to absorb the sulfate ions [37]. In
addition, the mineralogical composition of PC was analyzed by quantita- 0 0
tive XRD, and also shown in Table 1. 0.1 1 10 100 1000
Size (μm)
The particle size distributions (PSD) of PC and MK measured by
means of laser diffraction are shown in Fig. 2. The BET-specific surface
Fig. 2. Particle size distributions of the cement and fly ashes measured by means of laser
area of MK (2.93 m2/g) is higher than that of PC (1.7 m2/g). MK can diffraction. The solid lines show the relative frequency of particles as functions of the
therefore modify the cement matrix in both hydration products and size (noncumulative) and the dashed lines show the volume of the particles smaller
microstructure as a pozzolanic material, and fine filler. than a certain size (cumulative).
4 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16
table was used to consolidate both mortar beams and cubic samples after
filling each layer of mortar. The duration and frequency of vibration were
established according to ACI 544.2R. The specimens were demolded after
24 h and immersed in CH-saturated water at 23 ± 2 °C for 28 days prior
to accelerated aging treatment. Fresh cement pastes prepared for chemi-
cal analysis, such as XRD and TGA, were poured into sealed bottles and
then stored at 23 ± 2 °C until their time to be tested.
As reported in the literatures [3,19–21,26,27,39,40], the durability of
natural fiber-reinforced cements and mortars was mainly determined
by the mechanical properties of fiber-composites, such as the lose of
compressive strength, flexural strength and toughness, which were rea-
sonably well understood and predictable. However, the strength of the
composites is influenced by a number of factors, including the strength
of the matrix and the matrix/fiber interfacial bond, in addition to the
fiber strength [41]. The strength behaviors of the natural fiber itself,
which should be the foremost object of this study have not been inves-
tigated in depth. In contrast to the mechanical properties of the compos-
ites, the strength of the fiber is a direct way to investigate the
degradation degree of natural fiber. It is logical here to study the tensile
properties of sisal fiber embedded in the cement matrix as an essential
approach to evaluate its degradation subjected to aggressive environ-
ments. According to the “strand-in-cement test” proposed by Litherland
[41], an improved method was developed to measure the tensile
strength and physical–chemical properties of natural fiber after being
embedded in a cement matrix. This separation method was introduced
detailedly in author's previous work [42]. As shown in Fig. 3, a small
block of cementitious paste was cast around the middle part of a strand
consisting of 10 sisal fibers. In order to remove the cement block, the
strand inside the paste was wrapped with a wire mesh, which allows
the matrix to contact the surface of the natural fiber. The strand outside
the block was protected by epoxy resin and plasticene was used to fill
and protect the edge of the block from corrosion. Three strands were
prepared for each aging stage. The block around the strand was
removed carefully after certain number of wetting and drying cycles.
The durability of fiber reinforced cement mortar and fiber degrada-
tion in the cement matrices were evaluated on the basis of the flexural
properties of the fiber-mortar beams and the embedded fiber speci-
mens subjected to wetting and drying cycles with a duration of
42 h:15 h immersion in tap water at 70 °C and 25 h drying in an oven
at 70 °C, with 1 h air drying under 23 °C and 10% humidity between
each half-cycle.
2.3. Characterization of pore solution of hydrated cement matrix
The pore solution chemistry is of prime importance for understand-
ing trace element mobility and migration processes in cement [43]. This
can be grossly divided into three stages [44]: (i) dissolution of NaOH and
KOH (with pH around 13.4), (ii) dissolution of Ca(OH)2 (pH falls to
12.5), and (iii) dissolution of calcium silicate hydrate (C-S-H) gels (pH
remains above 12.5). The high alkalinity of the pore solution is well
known to be the main factor accelerating the degradation of cellulose
and lignin. The pore solution of hardened samples was obtained by
7 cm
Wire mesh
Epoxy resin impregnated
sisal fiber strand
Without epoxy resin coating
5 cm
Fig. 3. Sketch of sisal fiber strand in cement specimen.
the suspension method [45] and the pH of the pore solutions, which
was used to calculate the free OH− concentrations, was tested by a pH
electrode.
2.4. Chemical composition and water absorption of sisal fiber
The content of these three compositions determines the mechanical
properties and stiffness of fiber and its degradation rate. The procedure
used to determine the content of cellulose, hemicellulose and lignin ac-
cording to the methods proposed by Xu [46] and Viera [47], illustrated
in Fig. 4, was followed. The lignin content was determined using the
Klason method [48], and the cellulose content was obtained as the
sum of α and β cellulose. Four steps are needed to get sisal fiber from
Agave sisalana: (i) cutting and receiving clean sisal leaves, (ii)
extracting primordial sisal fiber from the leaves, (iii) drying sisal fiber
(air drying or artificial drying), and (iv) carding the fibers to remove
knots and other impurities. Although the fiber has been dried during
this process, the moisture content is still high, due to sisal's molecular
structure, lumen and hygroscopic features. Official methods proposed
by the Association of Official Analytical Chemists (AOAC) [49] were
followed to analyze the moisture content of sisal fiber in quintuplicate:
20 g of fiber was oven dried in a circulating air environment at 110 °C to
constant weight. Then the fibers were removed from each ceramic cru-
cible and cooled in a glass desiccator. The moisture content was deter-
mined by the average of weight loss ratios calculated as follows:
ω ¼ 100%  ðW i −W d Þ=W i ð1Þ
where ω is the moisture content in percent, Wi is the initial weight of
sisal fiber, and Wd is the residual weight of sample after drying and
cooling.
Unlike moisture content, the water absorption of natural fiber has
a direct influence on the reinforcing performance and durability of the
fiber in the cement matrix. Fibers absorb water, accompanied by
volume expansion, from the cement composite during the mixing and
casting process up to a dynamic equilibrium. During the hardening pro-
cess, cement captures some of that water back from fibers to form a high
alkaline pore solution in the space between matrix and fiber caused by
drying shrinkage, which, in turn accelerates the deterioration of natural
fiber.
Each sample of 10.0 g sisal fiber was soaked in tap water, NaOH so-
lution (pH = 13.0) and Ca(OH)2 solution (pH = 12.3) and then placed
in the curing box with temperatures of 23 ± 2 °C, 50 °C and 70 °C and
relative air humidity of 85%. The fibers were weighed every 10 min,
until they reached a moisture absorption balance. The water absorption
was calculated according to Eq. (2):
δ ¼ 100%  ðW b −W i Þ=W i ð2Þ
where δ is the water absorption in percent, Wb is the weight of sample at
the moisture absorption balance, and Wi is the initial weight of sisal
fiber.
PC or blended
cement paste
Plasticene
`

1000
Concentration of OH- (mMol/L)
100
PC
MK30
10
0.01 0.1 1 10
Time(h)
Fig. 4. Concentration of OH− for pore solutions of PC and MK30.
2.5. X-ray diffraction
The X-ray diffraction (XRD) analysis was carried out on ground
powders with an X2 Scintag X-Ray diffractometer in a θ–2θ configu-
ration using CuKα source (λ = 1.54 Å) at − 40 kV and 35 mA. The
sisal samples were scanned in step mode with a step size of 0.02°
(2θ) in the angular range of 5–25°. The percentage of crystallinities
using the peak height method [50] (%Cr1 ) and deconvolution
method [51–54] (%Cr2), crystallinity index (Cr.I.), and crystallite
size (L) were calculated from the X-ray diffraction spectra to character-
ize the crystallinity fraction of sisal fibers after aging (presented in
Appendix A).
2.6. Mechanical properties
2.6.1. Flexural behavior of sisal fiber-reinforced cement composites
The 28-day flexural properties of sisal fiber-reinforced mortar were
determined and served as reference. After 28 days curing, the sample
was treated in wetting and drying cycles. Against 28-day properties,
the strength and toughness of specimens exposed up to 30 wetting-
drying cycles were compared. Four-point bending tests were carried
out on an INSTRON 5984 34 k Universal Testing Machine at a crosshead
rate of 0.2 mm/min to investigate the flexural behavior of sisal fiber re-
inforced cement beams. Three specimens with a span of 150 mm were
tested for each mix. The displacements at mid span were determined
using LVDTs. Then the reinforcing effect of sisal fiber was investigated
by determining the first crack strength (σf), post-crack flexural strength
(σp), and post-cracking toughness (Tp). These were determined as
follows:
(i) σf: According to ASTM C1018 [55], the first crack load Ff is
defined as “the point on the load-deflection curve at which the
form of the curve first becomes nonlinear”. Thus the first-
cracking strength is given by Eq. (7):
2
σ f ¼ F f L=bh
where L is the lower fulcrum span (150 mm),and b and h are the
width and thickness of the specimen, respectively.
(ii) σp: Determined from the maximum load (Fmax) of bending test
after the first crack event by using Eq.(7).
(iii) Tp: The post-cracking toughness is defined as the area under the
load-deflection curve after cracking.
J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16 5
2.6.2. Tensile strength of single fiber and Weibull distribution
The uniaxial tensile properties of a single raw and embedded fiber at
each aging stage were tested using a micro-force Instron 5948 testing
system according to ASTM D 3822. A gauge length of 20 mm was
employed with a fixed loading rate of 0.2 mm/min. The ultimate load
and elongation at break of the sisal fibers were recorded. SEM was
used to determine the diameter of each fiber to calculate the cross sec-
tional area and to determine the stress–strain curves. Twenty single
fibers were tested for each group. Because of the non-homogeneous na-
ture of sisal fiber, especially of those exposed to alkaline attack and min-
eralization in cement, the data of the tensile tests exhibited considerable
scattered. Therefore, the method of two-parameter Weibull distribution
model was followed to analyze the tensile test results (see Appendix B).
100 1000 10000 2.7. Microstructure and thermal analysis
The microstructural analysis of sisal fibers and the composite frac-
ture surfaces were performed on a Hitachi 4700 scanning electron mi-
croscope (SEM) under an accelerating voltage ranging from 10 kV to
20 kV. The sample used for imaging was slices taken from specimens
subjected to the flexural test. The slices of fracture surfaces were
polished and then made conductive by sputtering deposition of a thin
layer of gold.
Components of the immersed sisal fiber were determined by using a
Perkin Elmer Pyris1 Thermogravimetric Analyzer at a heating rate of
20 °C/min from room temperature to 700 °C in N2-atmosphere.
3. Results and discussion
3.1. Pore solution of cement matrix
The development of PC pore solution chemistry, such as the pH
value, ion concentration as a function of hydration time, and the effect
of supplementary cementitious materials on pore solution chemistry,
has been a major focal point of research for several years as an index
of cement hydration. Here the alkalinity of pore solutions was investi-
gated as a critical factor leading to the alkaline hydrolysis of amorphous
components in natural fiber. According to the elemental concentration
levels and their effects on alkalinity and mineralization, the concentra-
tion of OH− of pore solution gained from neat PC and PC blended with
30 wt.% MK was tested, as shown in Fig. 4, to investigate the effect of
pore solution alkalinity on sisal fiber's degradation rate.
Because the pore solution of cement is saturated in Ca(OH)2 during
cement hydration, the alkalinity is mainly dominated by KOH and
NaOH and the corresponding sulfates [56]. From Fig. 4, it can be found
that in neat PC, caused by the consumption of pore water and gypsum
through reaction with C3A [57], dissolution of alkali sulfates (release of
sodium and potassium), and the depletion of the calcium sulfates [58],
the concentration of OH− increased rapidly after 10 h. These changes
of concentrations in pore solution at the early age have interactions
and correlate with the silicate reaction and the second aluminate reac-
tion. Correlating with the changes of OH− concentrations, the pH of
PC stays at a level of 13.1 ± 0.1 for the first 10 h, then rises up to
13.7 ± 0.06, and increases slightly with hydration for a long time. The
development of concentrations in pore solution was adjusted by replac-
ing 30 wt.% cement by MK. The initial concentrations of OH− are re-
duced due to the dilution effect through replacing cement by MK and
ð7Þ its high content of Al2O3, which contributes to the first aluminate reac-
tion. The concentration of OH− is reduced by mixing MK and continues
to decrease with the hydration time, not only because of less readily
soluble alkalis in the matrix caused by the dilution effect, but also due
to the binding of alkalis in the additional amounts of C-S-H generated
by the pozzolanic reaction [59]. The pH of pore solution in MK-30 blends
was 12.9 ± 0.1 during the first 8 h, and then decreased to 12.7 ± 0.08
after the deceleration and stabilization period of hydration.
6 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16
3.2. Chemical composition and moisture content of natural fiber
As shown in Table 2, the chemical composition of sisal fiber includes
pectin, extractive, lignin, hemicellulose, and cellulose contents, among
which cellulose is the major component and the one that provides
strength and stability. Due to the special structure of vegetable fiber as
shown in Fig. 1, lignin and pectin are the first two components which
encounter degradation in aggressive environments. The degradation
of lignin, pectin and hemicellulose determines the separation degree
of single cellulose fibers. Therefore, the contents of these three compo-
nents mainly affect the durability of natural fiber in a cement matrix at
early age. Owing to the similar contents of lignin, cellulose, and hemicel-
lulose with other natural fibers, sisal fiber was selected as a representa-
tive to determine the degree of degradation and loss of tensile strength,
as well as the durability of natural fiber-reinforced cement composites.
3.3. Water absorption of sisal fiber
The mass changes in different solutions under the three tempera-
tures, as functions of time are shown in Table 3. The fibers soaked in
tap water and Ca(OH)2 solution yielded the lowest and highest water
absorption ratios, respectively, at all three temperatures. Compared
with that soaked in tap water, the fiber's water absorption ratio in
Ca(OH)2 solution increased by 5%. It also can be seen that the moisture
absorption capacity of fibers rises with increasing temperature. It is well
known that natural fibers encounter more serious attack in alkaline
solution under higher temperature, which promotes the hydrolysis of
amorphous materials in fibers. The difference of the fiber's water ab-
sorption ratio in Ca(OH)2 and NaOH solutions is due to the mineraliza-
tion caused by penetration and precipitation of minerals rich in Ca.
Lignin, hemicelluloses and pectin act as chemical and physical barriers
to the penetration of Ca ions into the fibers [60]. Although non-
cellulose materials' degradation proceeds in both solutions, the mineral-
ization in natural fiber only occurs in Ca(OH)2 solution.
3.4. Degradation of sisal fiber-reinforced cement composites
In order to analyze the degradation of fiber in cement matrix and the
durability of fiber reinforced cement mortar, the investigation of flexur-
al behaviors was performed and the results are shown in Fig. 6. Accord-
ing to Mohr et al. [26], prior to cycling, the MK composites exhibited
lower first crack strength than the control. However, conflicting results
are found in this study. As shown in Fig. 5, the initial first crack strength
of fiber reinforced cement mortar was improved due to the presence of
MK. MK-10 and MK-30 gave 9% and 34% higher strength than that of PC,
respectively. However, the improvement of durability cannot be detect-
ed from the first crack strength of MK-10 and MK-30. After 5 cycles, the
first crack strength of PC increased slightly, and then decreased with ad-
ditional wetting and drying cycles. The increase in density of the matrix
around the fibers, and the fiber embrittlement due to migration of hy-
dration products from the matrix to the fiber cell walls could be the
main factors contributing to the increase of the first crack strength
and reduction of toughness, by reducing the flexibility and deformation
capacity of the fibers. Stronger, more rigid and brittle fibers, and im-
proved bond strength, are known to lead to composites with higher
strength and lower ductility [19]. Over the range of wet/dry cycles in-
vestigated here, the decrease of first crack strength for MK-10 and
MK-30 is similar as that of the control. First crack strength depends
Table 2
Chemical composition of natural fiber.
Sisal Component (%)
Lignin Holocellulose Cellulose
12.3 81.9 68.2
Table 3
Water absorption ratio of sisal fiber.
Temperature (°C) Water absorption ratio, δ (%)
Tap water Ca(OH)2 solution NaOH solution
23 153.0 160.7 156.8
50 155.2 165.1 158.7
70 159.7 169.3 160.2
mainly on characteristics that govern matrix strength and is minimally
dependent on fiber parameters such as type, size and amount [61].
Therefore, in addition to the first crack strength, the degradation of
fiber in cement should be analyzed by investigating properties calculat-
ed from multiple cracking stage and softening and failure stage.
The initial peak strength of sisal fiber reinforced mortar, before wet-
ting and drying exposure, was improved by 21% and 36% for samples
with 10 wt.% and 30 wt.% metakaolin, respectively. The changes in
peak strength values for all three groups were more obvious than
those of the first crack strength, and sisal fiber reinforced neat cement
mortar gave the sharpest drop. After 30 cycles, the peak strength of PC
dropped to 35% of the initial strength, whereas the decrease of peak
strength for MK-10 was 23%. Compared with PC, the MK-30 suffered
much smaller changes, which means that the decrease of alkalinity of
pore solution effectively arrested the deterioration of sisal fiber.
Fiber reinforcement in cement based composites results not only in a
large increase in bending strength, but also in a considerable increase in
energy absorption capability. After cracking of the matrix, the cracks
cannot propagate without stretching and debonding of the fibers. As a
result, a large additional amount of energy is absorbed before ultimate
failure of the specimen [62]. Toughness is one of the most important
performance indicators of energy absorption capability for fiber-
reinforced cement composites. Post-cracking toughness results are
shown in Fig. 5, which illustrate that flexural toughness is most sensitive
to wetting and drying cycles. The substitution of 10 wt.% and 30 wt.%
MK in cement led to 23% and 48% improvement in initial post-
cracking toughness, respectively. Under progressive cycling, sisal fiber
reinforced PC mortar encountered the most significant drop in post-
crack toughness. The cement substitution by MK not only improved
the initial post-cracking toughness of SFRCC, but also effectively delayed
the degradation of the composite when subjected to progressive
cycling. After 30 cycles, indicated by the residual toughness, the durabil-
ity of fiber reinforced mortar was improved by 69% and 83%, respective-
ly, in MK10 and MK30. From these results, it can be concluded that the
reinforcing effect of sisal fiber for cement composites is mainly in the lat-
ter two stages of load-deflection behavior, and the matrix with declined
alkalinity enhances the energy absorption capability in the softening and
failure stage.
3.5. Degradation of the embedded sisal fiber
3.5.1. Tensile strength
Fig. 6 shows the stress–strain curves obtained by uniaxial tensile
testing of selected sisal fibers. It can be seen that the tensile strength
of the embedded fiber in MK30 after 5 wetting and drying cycles is
slightly higher than that of a raw fiber. According to Gu [63], this in-
crease may be caused by the removal of non-cellulosic materials and
impurities, which remain dispersed in interfibrillar regions of natural
Hemicellulose Extractive Pectin Moisture & other
13.7 2.4 1.3 3.2
 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16 7

12 12
decreases and then gradually increases with increasing characteristic
σf σp Τp strength, which indicates that the data of a single fiber tensile test are
10 10 concentrated when the tensile strength is high or low, and apparently

Prot-crak toughness (MPa)

scattered for strengths inbetween. The distinct cell wall structure of
Flexural strength (MPa)

8 8
6 6
4 4
0 cycle
2 5 cycles
15 cycles
30 cycles
0 0
PC MK10 MK30 PC MK10 MK30 PC
Fig. 5. Flexure test results. average first crack strength (MPa), average post-crack flexural
strength (Mpa), and average post-cracking toughness (N-m).
fibers. The rearrangements of fibrils and close packing of cellulose
chains in natural fibers are also beneficial to lead to preferable tensile
deformation, better load sharing behaviors and higher tensile strength
and elongation at break in the stretched state [64–66]. However, the
tensile strength of fibers embedded in PC and MK10 is lower than the
control group. All fibers, especially the ones pulled out from PC, encounter
a degradation in strength when subjected to aging condition.
The decreasing strength and gradually increased slope of elastic
region of the stress–strain curves indicate that the mineralization and
alkaline hydrolysis in natural fiber give rise to a higher stiffness and
brittleness. According to author's analysis, the exit of hemicellulose
and lignin in natural fiber is of dual character, on one hand they increase
interfibrillar space and disperse the fibrils, and on the other hand, they
protect the main structural component of the fibers, cellulose, from
the alkaline attack. Without non-cellulose materials as protective coat-
ing, cellulose chains encounter corrosion directly from the pore solution
of high alkalinity. Another phenomenon shown in Fig. 6 is that the deg-
radation degree of sisal fiber decreases with increasing replacement of
PC by MK, which indicates that MK effectively mitigates the degradation
of sisal fiber and improves the durability of the composite by reducing
the alkalinity of the pore solution.
Tensile test results for sisal fibers are difficult to analyze due to wide
scatter, and a statistical approach is essential for assessing their tensile
properties [67]. Calculated from the Weibull plots (Fig. B.1), characteris-
tic strength, average Young's modulus and Weibull modulus of sisal
fibers are given in Table 4. The results show that the Weibull modulus
700
5c-MK30
600
sisal fiber is the main reason for its excellent mechanical properties,
meanwhile it also leads to the low durability in cement matrix and scat-
ted testing data when subjected to aggressive environments. The raw
sisal fibers and the fibers embedded in MK30 with 5 wetting and drying
cycles did not experience severe alkali attack, therefore the testing data
of tensile strength remains concentrated. The fibers embedded in PC
and MK10 partly degraded due to the high alkalinity of the two matri-
2 ces, after 5 and 15 wetting and drying cycles, respectively. Because of
the surface structure of sisal fiber, even in a same matrix, the degrada-
tion degrees of the embedded fibers have significant difference. At the
MK10 MK30 end of aging treatment (30 cycles), in neat PC, the fibers degraded
completely and the tensile strength decreased to a very low level, there-
fore the testing data gradually increased. After 5 wetting and drying cy-
cles, sisal fiber's degradation was mitigated in MK modified cement
blends. Compared with raw fiber, the fibers immersed in neat PC and
MK-10 have 35% and 13% lower tensile strengths, respectively, but a
slightly improved strength is obtained for the fibers embedded in MK-
30. The cause is the removal of hemicellulose and lignin from the fibers
due to the relatively mild alkaline environment of a matrix created by
blending 30 wt.% MK as shown in Fig. 4. Because of the high alkalinity
of pore solution and high CH amount in the PC matrix, the fiber embed-
ded in neat PC matrix yielded the lowest tensile strength and elongation
at break at every step of aging. After 30 wetting and drying cycles, the
tensile strength of fibers immersed in PC decreased by 86%. However,
MK10 and MK30 decrease the fiber's strength by only 66% and 38%, re-
spectively, compared with the raw fiber.
The degradation of sisal fiber is also evidenced by fiber's maintained
deformability (in inverse proportion to Young's modulus) after aging.
As can be seen in Table 4, after 5 cycles of aging, the deformability is re-
duced only slightly, and the tensile strength of fibers embedded in
MK30 is about the same as that of raw fiber. The fibers immersed in a
neat PC matrix yield the highest Young's modulus, due to the alkaline
corrosion. After 30 wetting and drying cycles, Young's modulus of
fiber in PC rises by 29%, however, that of fiber immersed in MK30
increases by only 6%. Both mineralization and alkaline hydrolysis of
non-cellulose materials in fibers cause the increased stiffness and brit-
tleness. The formation of ettringite/monosulfate layers in the lumen
cavity and CH re-precipitation in the center of the lumen [68] signifi-
cantly increase the stiffness of natural fiber, evidenced by the increased
Young's modulus of fibers in Fig. 6 and thus affect the embrittlement of
composites. Due to the alkaline hydrolysis of hemicellulose and lignin,
the content of cellulose crystal chains, which are the framework of nat-
ural fibers, becomes the main component. The increased crystallinity in-
dicates that the lower amount of phenolic polymer and amorphous
components, and as a result, stiffness and embrittlement of natural
fiber were increased.

500
Table 4
Stress (MPa)

400 The Weibull parameters and deformability of sisal fibers in tensile test.

Sample Number Weibull Characteristic R2 Young's

300 raw fiber of cycles modulus, m strength, σ0 (MPa) modulus, E
5cMK10 (GPa)

200 Raw 0 8.74 605.86 0.98934 15.4

5c-PC PC 5 3.60 393.15 0.95462 17.26
30c-MK30
MK10 5 5.13 525.97 0.97302 16.44
100 Mk30 5 8.95 657.73 0.95387 15.42
30c-MK10
PC 15 5.47 233.06 0.97124 18.66
30c-PC
0 MK10 15 3.36 350.67 0.94849 17.15
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 MK30 15 6.23 576.74 0.94594 16.07
Strain PC 30 8.60 83.82 0.95378 19.79
MK10 30 5.50 208.04 0.98462 17.73
MK30 30 6.32 488.53 0.97777 16.40
Fig. 6. Stress–strain curve of tensile test on sisal fibers.
8 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16

3.5.2. X-ray diffraction analysis
As analyzed above, quantifying crystalline properties is a logical ap-
proach to evaluate the effects of mineralization and alkaline hydrolysis
on the properties of natural fibers. Natural fibers are semi-crystalline
materials and the tensile strength and elastic modulus have an evident
relationship with its crystal content. Here, the intensity of the (1 10),
(110), and (002) peaks and the amorphous scatter between 18° and
19° were measured.
Fig. 7 shows the XRD diffractograms of the raw sisal fiber and that
embedded in PC, MK10 and MK30 samples after 5, 15 and 30 wetting
and drying cycles. It can be seen that cellulose has a crystalline nature
with an intensive peak at 2θ ≈ 22.5° corresponding to (002) lattice
plane and a broad peak between 13° and 18° corresponding to (1 10)
and (110) lattice planes. Peaks of ettringite are found in all fibers except
the raw one, while the peaks of CH are only detected in fibers immersed
in a neat PC matrix. Due to the dilution effect and pozzolanic activity of
MK, CH and ettringite are effectively consumed and the concentrations
of OH− and Ca2+ are decreased. Therefore less re-precipitation of these
two hydration products occurs in the lumen of natural fibers. Peaks cen-
tered around 29.2° (2θ), which is the dominant peak in both C-S-H and
calcite [69], were detected in the X-ray patterns of the embedded fibers
by 15 wetting and drying cycles, and become stronger after 30 cycles.
This may be caused by the residual cement hydrates on sisal fiber's
surface. The degradation degree of embedded sisal fibers increases
with increasing wetting and drying cycles, as a result, it becomes more
difficult to separate fibers from cement matrices and more residues
were detected. In addition, MK reacts with calcium hydroxide, decreas-
ing the amount of this product and producing more C-S-H and calcium-
alumino-silicate-hydrate (C-A-S-H) [70], the intense XRD pattern of
which also centered around 29.2° (2θ) [69]. Therefore, the intensity of
12 3 45 6,7 7 3 7 8
these peaks increases with increasing cement replacement with
metakaolin at both 15 and 30 wetting and drying cycles. The increasing
residual mounts of C-S-H and C-A-S-H also indicate the beneficial effect
of MK on interfacial bonding strength between fiber and cement matrix.
This has already been reported in author's previous publication [71].
Fig. 8-a and b shows the Cr.I. and %Cr1 of sisal fiber as functions of MK
content and duration of aging, respectively. It can be seen that the crys-
tallinity index of raw fiber is around 0.71 and increases at different rates
with the aging duration of all three matrices. After 5 cycles of aging,
fibers immersed in MK10 have the highest crystallinity index and per-
centage of crystallinity. After 30 wetting and drying cycles, the crystal-
linity index and crystallinity percentage of fiber immersed in neat PC
were 20 and 13% higher than those of raw fiber, respectively. This
means that the amorphous materials encountered serious attack in
the PC matrix. At every stage of aging, the fiber embedded in MK30
shows the lowest crystallinity properties, and the Cr.I and % Cr1 increase
by only 0.65% and 0.19% from 15 to 30 wetting and drying cycles. The re-
sults agree well with the single fiber's tensile test data and flexural
properties of sisal fiber reinforced cement mortar as analyzed above.
Table 5 shows sisal fibers' percentage of crystallinity (%Cr2) and
crystal size, which is calculated from their corresponding peak position
and the full width of the peak at its half maximum height (FWHM)
using deconvolution method. Curve-fitting and deconvolution process-
es of XRD spectra were carried out to separate diffraction peaks using
Gaussian functions. An important assumption for this analysis is that in-
creased amorphous contribution is the main contributor to peak broad-
ening, however, in addition to crystalline disorder, there are other
intrinsic factors that influence peak broadening, such as crystallite size
and non-uniform strain within the crystal [72]. Therefore, the regularity
of %Cr2 is not significant as Cr.I and %Cr1. However, it still can be seen
that after accelerated aging, immersed fibers show bigger crystallite
size than raw fiber. Fibers pullout from PC and MK30 give the smallest
crystallite size and crystallinity percentage after 30 cycles of aging,
respectively.

3.5.3. Thermal analysis

The results of a thermal analysis (TGA and DTA) of sisal fibers are
n
shown in Fig. 9 and summarized in Table 6. They present similar
thermal degradation behaviors of raw fiber and fibers embedded in
the matrix of PC, MK10 and MK30 after 30 wetting and drying cycles be-
k
tween room temperature and 600 °C. As analyzed in previous work [73],
there are three main stages of weight loss: (i) slight weight loss from 50
to 110 °C due to evaporation of moisture in the fibers; (ii) weight loss for
h decomposition of hemicelluloses and the main part of lignin under tem-
perature between 270 and 350 °C; and (iii) the maximum weight loss
g under higher temperature accompanying the degradation of cellulose
and the rest of lignin. The extensive range of the degradation of the
f lignin in two or three stages [74,75] makes the analysis more difficult.
In addition, the DTG curves of selected samples show that the ranges
e of each stage shift for fibers with different degradation degree. For the
first stage, the original sisal fiber shows a higher moisture content
than that of fibers embedded in the cementitious matrices. The reason
d for the high moisture content of raw sisal fiber was its water-absorption
capacity in the laboratory environment (25.5 °C, CO2 concentration of
c 514 ppm and humidity of 10.5%) in this study. Although the pulled out fi-
bers absorbed water from fresh cement paste when they were immersed
b into the matrix, they gradually lost water due to the high water demand
of the matrix, especially the cement-MK composites. The accelerated
a aging ended with drying, which also accounted for the lower moisture
contents of pulled out fibers than those of raw fibers. The degradation
5 10 15 20 25 30 35 40 45 and damage of cell walls also caused a reduction of water retention.
2θ (degree) Therefore, the fibers, embedded in the PC matrix, encountered the most
serious degradation after accelerated aging treatment and showed a
Fig. 7. XRD pattern of sisal fibers: (a) raw fiber, (b) PC-5 cycles, (c) MK10-5 cycles,
lower moisture content than those of fibers pulled out from MK10 and
(d) MK30-5cycles, (e) PC-15 cycles, (f) MK10-15 cycles, (g) MK30-15 cycles, (h) PC- MK30. For the second stage, the raw fiber and fiber pulled out from
30 cycles, (k) MK10-30 cycles, (n) MK30-30 cycles. MK10 presented evidence of independent weight loss as shown in the
 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16 9

0.86 88
87
0.84
86
0.82 85
84
0.80
83

Cr1 (%)
Cr.I

0.78 82
81
0.76
80
0.74
raw 79 raw
PC PC
0.72 78
MK10 MK10
MK30 77 MK30
0.70
raw 5 cycles 15 cycles 30 cycles raw 5 cycles 15 cycles 30 cycles

Fig. 8. Crystallinity index and percentage of crystallinity of pull-out sisal fiber as functions of MK content and exposure duration.

DTG curves. For fibers pulled out from MK30 and PC matrices, the second in Fig. 9-b. As the amounts of the components reduced with decreasing
stage merged with the thermal decomposition peak of cellulose. The crystallite, the DTG peaks shifted to lower temperatures. In stage 3, the
weight losses of the raw and embedded fibers in stage 2 indicate that raw fiber highest weightless and the weight losses of the embedded
hemicellulose and lignin of natural fiber are sensitive to the alkalinity fibers are proportional to the cement replacement level by MK. Com-
and CH amount of the cement matrix. These two amorphous components pared with the effect on hemicellulose and lignin, MK's effectiveness
encountered the most serious degradation in a PC matrix, while the MK10 in slowing down the degradation of cellulose in natural fiber was slight-
and MK30, with reduced alkalinity of pore solution and less portlandite in ly lower. This was caused by the specific structure of natural fibers and
solid phase, gave slightly arrested corrosion to sisal fiber. the complex mechanisms of degradation. Lignin and hemicellulose are
As depicted in Fig. 9, the raw sisal fiber and the fiber embedded more sensitive to the alkalinity pore solution and are in direct contact
in MK30 experienced pronounced degradation processes starting at with the hydration products of the matrix, therefore, their degradation
around 300 °C, while for fibers pulled out from the MK10 and PC matri- is more impressionable than that of cellulose. The pore solution can
ces slightly less pronounced degradation processes occurred at 280 °C infiltrate the cell wall of the fiber and cellulose encounters alkali hydro-
and 260 °C, respectively. The main decomposition step occurred in the lysis even though the lignin and hemicellulose are only partially
range of 260 °C to 490 °C for the raw fiber; 270 °C to 460 °C for fiber im- decomposed. In addition, the precipitation of calcium hydroxide among
mersed in MK30; 255 °C to 455 °C and 250 °C to 410 °C for fiber im- cellulose microfibrils, through the gap of lignin and hemicellulose, also
mersed in MK10 and PC, respectively. In this stage the cleavage of the accelerates the embrittlement and decomposition of cellulose.
glycosidic linkages of cellulose reduces its polymerization degree, lead- The continued loss of mass at higher temperatures is due to the ther-
ing to the formation of CO2, H2O and a variety of hydrocarbon deriva- mal decomposition of residual cellulose and lignin of fiber. However, the
tives [76]. The decomposition profiles indicate slight thermal stability significant differences between the raw fiber and the embedded fibers
differences for the samples. The DTG peaks were centered at 421 °C, in the temperature range of 500 to 600 °C indicate the existence of pre-
411 °C, 407 °C and 385 °C for the original fiber and fibers pulled out cipitation of portlandite phase in cell walls of sisal fiber. The mass loss of
from the matrices of PC, MK10 and MK30, respectively, as presented raw fiber after the main decomposition step was smallest, while the
fibers embedded in MK30, MK10 and PC matrices experienced a weight
loss, which is inversely proportional to the MK amount in cement
Table 5 matrix. Assuming that the weight loss for residual components of each
Calculated and observed crystalline parameters of sisal fiber. sample is the same between 500 and 600 °C, the extra weight loss of
Curing or exposure Samples Peak positiona FWHM (β) L (nm) %Cr2 the embedded fiber after the main decomposition step can be consid-
duration (2θ) (2θ) ered to serve as the amount of portlandite precipitated in cell walls.
– Raw fiber 16.81 4.89 1.62 41.20 The fiber embedded in the pure PC matrix yielded portlandite amount
22.31 3.37 2.38 of 10.9%. However, MK30 (2.8%) and MK10 (3.4%) indicate a consider-
5 cycles PC 16.47 4.88 1.63 36.64 able inhibiting effect on the mineralization of fiber. This also confirms
22.53 2.66 3.01
the important roles of lignin and hemicellulose as protective barriers
MK10 16.59 4.28 1.85 42.12
22.48 3.19 2.51 to the infiltration of pore solution and precipitation of portlandite in
MK30 16.43 4.38 1.81 47.56 cell walls of natural fiber.
22.47 3.15 2.54
15 cycles PC 16.46 4.37 1.82 39.62 3.5.4. Microstructure analysis
22.56 2.64 3.04
A microstructure analysis was carried out of the fracture surfaces of
MK10 16.53 4.60 1.73 42.47
22.61 2.49 3.22 sisal fiber reinforced cement samples to investigate the effect of aggres-
MK30 16.21 4.04 1.97 45.28 sive environments on degradation of sisal fiber. Fig. 10-a and b shows
22.63 2.53 3.17 micrographs of sisal fiber in a cement matrix after 7 and 28 day curing,
30 cycles PC 16.47 4.58 1.73 49.72
respectively, without accelerated aging. It can be seen that the sisal fiber
22.59 2.30 3.48
MK10 16.08 4.14 1.92 50.22 embedded in PC, at an early age, still remained intact without indication
22.61 2.38 3.36 of mineralization on the surface or defect caused by corrosion. In
MK30 16.17 4.14 1.92 45.02 contrast, after 28 days of hydration, a space was found around the
22.67 2.40 3.34 fiber due to the drying shrinkage of both cement matrix and the fiber.
a
Full width of the peak at its half maximum height. The fiber still kept its initial appearance, except for small signs of
10 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16

100
0.0
90
80 -0.2

Deriv. Weight (%/°C)

Weight loss (%)

70 -0.4
60
-0.6
50
40 -0.8
Raw fiber Raw fiber
30 30C-PC 30C-PC
-1.0
30C-MK10 30C-MK10
20 30C-MK30 30C-MK30
-1.2
10
100 200 300 400 500 600 100 200 300 400 500 600
Temperature (°C) Temperature (°C)

a b
Fig. 9. TGA and DTG curves of original sisal fiber and fibers pulled out from the matrix of PC, MK10 and MK30 after 30 wetting and drying cycles.

mineralization on the fiber surface. The crack on the fiber surface and
stripping of single cellulose micro-fibrils shown in Fig. 10-c and d
were the result of fiber degradation in PC after 15 cycles of aging. This
was caused by the alkaline hydrolysis of lignin and hemicellulose due
to the deposition of cement hydration products, especially calcium
hydroxide, as part of the alkaline degradation process. As shown in
Fig. 10-e, f and g, hemicelluloses and lignin were decomposed almost
completely after 30 cycles in the aggressive environment. The sisal
fiber's surface was fully decomposed and the lumen was saturated
with the hydration products of cement, which caused the hydrolysis
of cellulose through peeling-off reaction, chemical stopping reaction,
physical stopping reaction, and alkaline hydrolysis in degradation
mechanisms. The three main components of natural fiber, lignin, hemi-
cellulose and cellulose, were decomposed progressively from the surface
to the inside. As a result, the fibers lost their reinforcement capacity in ce-
ment based materials. Fig. 10-h, i and j shows the sisal fiber pulled out
from a MK30 sample after 30 cycles of wetting and drying, from which
two remarkable observations could be made, in contrast to the neat PC
sample (Fig. 10-g): (i) the stripping space between fiber and matrix
was reduced by mixing 30 wt.% MK due to the decreased autogenous
shrinkage of the cement matrix and less drying shrinkage of sisal fiber
in the low alkaline pore solution due to the pozzolanic reaction and dilu-
tion effect of MK; and (ii) the fiber still retained its initial morphology
with a good integrity, except for a slight defect on the surface, which
was also due to the low alkalinity of the pore solution and low content
of portlandite in the solid phase of the matrix because of the pozzolanic
reaction of MK in cement. This is also the reason for the reduction of flex-
ural strength of sisal fiber reinforced PC and the retention of flexural be-
havior of MK30 specimens after accelerated aging, as shown in Fig. 5.
4. Discussion
degradation of the polysaccharides. As shown in Fig. 1, hemicelluloses do
not serve as load-bearing component in the cell wall of natural fiber, but
they bind and fix the cellulose micro-fibrils to make the fiber more stable.
Lignin and hemicellulose have one characteristic in common: they are
both amorphous and sensitive to the alkaline environment of the cement
matrix, which means that they are easy to undergo the alkaline hydroly-
sis. Cellulose molecular chains are ordered into strands of approximately
40 cellulose microfibrils, which are not homogeneous, consisting of
crystalline and amorphous regions. The crystalline regions of cellulose
are much more stable than the amorphous region in alkali, and provide
strength of cellulose fiber. Generally, the degradation of cellulose can be
simply generalized as the disconnection of discrete cellulose nano-
crystals caused by decomposition of reducing ends (feature original C4-
OH) in amorphous regions.
Fig. 11 is a diagrammatic sketch showing what we believe to be the
alkaline degradation of natural fiber in a cement matrix, due to the high
alkalinity of cement solid phase and pore solution. There are four steps:
(i) Degradation of lignin and part of hemicellulose, which leads to the
exposure of holocellulose in the pore solution and solid phase of the ma-
trix; (ii) In this step, degradation occurs mainly on hemicellulose, which
causes the decrease of integrity and stability of the cell wall of natural
fiber; (iii) After the degradation of lignin, hemicellulose and intra-
molecular hydrogen bonding, there remains no binding for cellulose
micro-fibrils, and as a result cellulose fibers disperse in the pore solution
of the matrix, which in turn, accelerates the degradation of cellulose;
(iv) The last step is the failure of cellulose micro-fibrils, which is caused
by alkaline hydrolysis of amorphous regions containing non-reducing
end, and leads to the complete degradation of natural fiber. As the deg-
radation proceeds, the hydration products, such as C-S-H and soluble
portlandite, gradually infiltrate into the cell wall, which in turn leads
to mineralization and embrittlement of the natural fiber.

4.1. Alkaline degradation (hydrolysis) 4.2. Mineralization

Lignin functions as cuticle or glue in the cell wall of natural fiber Except for the alkali hydrolysis of the three main components, cell
giving its structure and protecting the fiber against microbial or chemical wall mineralization is also an important degradation mechanism
which leads to the fiber embrittlement and the reduction of strength
and strain capacity. Caused by the migration of hydration products to
Table 6
Weight loss of sisal fibers in TGA test. lumens and to the middle lamella, and volume variation in these fibers
due to their high water absorption, fiber mineralization is sensitive to
Samples Weight loss (%)
the CH amount of the matrix [20]. In contrast to the reviews presented
Stage 1 Stage 2 Stage 3 Stage 4 in articles [19–21], two mineralization mechanisms are proposed in
Raw fiber 3.1 19.7 52.8 3.0 this study: CH-mineralization and self-mineralization. Caused by the
PC-30 cycles 1.0 9.0 28.8 13.9 migration of hydration products (calcium hydroxide) to lumens and
MK10-30 cycles 1.6 12.6 31.8 6.4 to the middle lamella, and volume variation, CH-mineralization indi-
MK30-30 cycles 1.6 15.4 38.3 5.8
cates that the CH content of the matrix and Ca2+ concentration in the
 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16 11

a b

c d

e f

g h

i j

Fig. 10. SEM micrographs of the fracture surface of the aged sample with sisal fiber: (a) 7 days, (b) 28 days, (c–d) PC after 15 cycles, (e–g) PC-30 cycles and (h) MK30-30 cycles.
12 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16
Alkaline degradation Alkaline degradation
of lignin of hemicellulose
Fig. 11. Diagrammatic sketch of natural fiber's alkaline degradation process.
pore solution contributes to the degradation of natural fiber. However,
self-mineralization demonstrates the hydrolysis rate of amorphous
components (lignin and hemicellulose), therefore the contribution of
alkalinity of the matrix to the fiber degradation can be determined. It is
well known that the precipitation of hydration products, likely calcium
hydroxide, within the fiber cell wall structure is the main reason for cell
wall mineralization [21].
As shown in Fig. 6, sisal fibers suffered brittle fracture without
any plastic deformation, therefore, it is logical to characterize the
deformability of sisal fiber using the tensile modulus. Combining the
results of TGA and tensile test of the embedded sisal fibers, Fig. 12-a
shows the effect of precipitated CH amount on Young's modulus of
fiber. It can be seen that the precipitation of CH in the cell wall signifi-
cantly increases the stiffness of sisal fiber. The crystallization of CH
through nucleation and growth processes in the cell wall not only seri-
ously damage the bonding between the various components but also
corrode cellulose micro-fibrils, thereby the strength was significantly
reduced. Unlike CH-mineralization, slight hydrolysis of lignin and hemi-
cellulose has no effect on Young's modulus of sisal fiber. However, when
Cr.I exceeds 0.80, the deformability of the fiber suffers a significant
reduction, which is evidenced by the increased Young's modulus as
shown in Fig. 12-b. This indicates that Young's modulus of natural
fiber mainly depends on the cellulose, which is the primary structural
component of cell walls. However, lignin within the fiber cell wall acts
20
19
Young's modulus (GPa)
18
17
16
15
0 2 4 6 8 10
Amount of CH precipitated in fiber (%)
a b
Fig. 12. Effect of CH precipitation (a) and Cr.I (b) on Young's modulus of sisal fiber.
Stripping Alkaline degradation
of cellulose of cellulose
as a barrier to pore solution ingress/egress and by its very presence
prevents precipitation of mineral phases within the fiber cell wall,
thereby minimizing fiber mineralization [77]. In order to restrain
the degradation of natural fiber in the cement matrix, both amor-
phous components (lignin and hemicellulose) and cellulose should
be protected by reducing the OH− concentration of pore solution in
the matrix.
5. Conclusions
The presented work analyzed the degradation mechanisms of sisal
fiber in the alkaline and mineral-rich environment of cement matrices.
In order to systematically investigate the alkali attack of natural fiber
caused by the pore solution of cement matrix, metakaolin was selected
to partially replace the Portland cement. The deterioration of sisal fiber-
reinforced cement composites and degradation rate of the embedded
sisal fiber were determined by means of flexural behavior of composites
and crystallinity, and thermal and tensile properties of fibers, respec-
tively. This allows the qualitative description of the degradation mech-
anisms of natural reinforcement in cement based materials. Specifically,
the following observations were made.
(1) The chemical composition of sisal fiber is made up of pectin,
extractive, lignin, hemicellulose, and cellulose, among which
20
19
Young's modulus (GPa)
18
17
16
15
12 0.70 0.72 0.74 0.76 0.78 0.80 0.82 0.84 0.86
Cr.I of fiber
 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16
cellulose is the main structural component in cell walls. Com-
pared to fibers that were soaked in tap water, the water absorp-
tion ratio of fibers immersed in Ca(OH)2 solution is higher, and
the moisture absorption capacity of fibers increases with increas-
ing temperature.
(2) By replacing 30 wt.% cement by metakaolin, the durability of sisal
fiber reinforced cement based composites subjected to wetting
and drying cycles is effectively improved. Therefore, controlling
the pH value of the pore solution at a relatively low level is an ef-
fective way to mitigate the degradation of natural reinforce-
ments.
(3) Compared with raw fiber, the tensile strength of fiber im-
mersed in PC suffers the most severe reduction evidenced by
the severe decomposition of the amorphous phases (lignin
and hemicellulose) and the increased crystallinity. Lignin
and hemicellulose are amorphous and easily undergo the alka-
line hydrolysis. Due to the protective barrier role of lignin and
hemicellulose, precipitation of portlandite in cell walls is effec-
tively arrested for the fibers to remain higher amounts of lignin
and hemicellulose.
(4) The alkali degradation of cellulose mainly depends on the reduc-
ing end in amorphous regions and can be simply generalized as
the disconnection of discrete cellulose nano-crystals caused by
decomposition of reducing end through four processes: peeling-
off reaction, chemical stopping reaction, physical stopping reac-
tion, and alkaline hydrolysis.
(5) Four interactional alternate steps for the degradation of natural
fibers in a cement matrix are summarized: (i) degradation of
lignin and part of hemicellulose leading to the exposure of
holocellulose; (ii) degradation of hemicellulose causing the de-
crease of integrity and stability of cell walls; (iii) degradation of
intra-molecular hydrogen bonding leading to the dispersion of
cellulose microfibrils; and (iv) alkaline hydrolysis of amorphous
regions containing non-reducing, and as a result, complete deg-
radation of cellulose micro-fibrils. As the degradation proceeds,
the hydration products, such as soluble C-S-H and portlandite,
gradually infiltrate into the cell wall, which in turn led to miner-
alization and embrittlement of natural fiber.
(6) Cell wall mineralization is found to be an important degrada-
tion mechanism which leads to the fiber embrittlement and
the reduction of strength and strain capacity. In contrast to
the reviews presented in the literature, two mineralization
mechanisms are proposed in this study: CH-mineralization
and self-mineralization. The crystallization of CH through nu-
cleation and growth processes in cell walls not only seriously
damage the bonding between the various components but also
corrode cellulose micro-fibrils, which leads to increased brittle-
ness and reduction of strength. The self-mineralization indicates
that the slight hydrolysis of lignin and hemicellulose has no sig-
nificant effect on Young's modulus of sisal fiber, however, when
Cr.I exceeds 0.80, the deformability of fiber is reduced consider-
ably. This demonstrates the barrier role played by lignin and
hemicellulose by preventing cell walls from precipitation of
mineral phases.
The results demonstrate that natural fibers undergo serious degra-
dation in the cement matrix due to the mineralization of cell wall,
peeling-off reaction and alkaline hydrolysis of amorphous components
(lignin, hemicellulose) of fiber and the amorphous regions of cellulose.
The effect of alkalinity of cement pore solution on the decomposition
of cellulose can be used to predict the failure of natural fiber. Because
of the uncertainties of the kinetics data and the lack of parameterization
of mineralization, further investigation of the natural fiber degradation
in cement matrices is necessary. The effects of temperature, concentra-
tion of Ca2 +, water and cement ratio, as well as different aggressive
13
aging conditions on mineralization, peeling-off reaction, and alkaline
hydrolysis of lignin and hemicellulose, especially the cellulose, need to
be studied in more detail.
Acknowledgment
The Erbilgin family is acknowledged for funding this work and for
providing the author with the opportunity to pursue this study through
the Erbilgin Family Fellowship at Columbia University. The authors
would like to express their sincere thanks to Dr. Liming Li of Columbia
University's Carleton Laboratory for his assistance and cooperation
throughout this study. Mr. Hagh Mckee from Bast Fibers LLC, Creskill,
New Jersey, is thanked for the donation of sisal fiber.
Appendix A
The Segal empirical percentage of crystallinity (%Cr1) was performed
using the following relation [78]:
%Cr 1 ¼ I002 =ðI002 þ Iam Þ  100% ðA:1Þ
where I002 is the maximum intensity of diffraction of the (002) lattice
peak at 2θ = 22.5°, which represents both crystalline and amorphous
materials. And Iam is the intensity of diffraction contributing to the
amorphous fraction, which is taken at a 2θ angle between 18° and 19°
where the intensity is a minimum [79–81].
The percentage of crystallinity (%Cr2) was determined by the amor-
phous peak areas under the diffraction X-ray plot and was calculated
based on Eq. (A.2) [54]:
%Cr 2 ¼ ðAc =At Þ  100% ðA:2Þ
where Ac is the area below the diffraction peak of the (002) plane
(peak at 2θ = 22.5°) and the (101) and (10ī) planes (peak at
13–18°), and At is the area below the whole region of the XRD spectra
representing the amorphous [25] and crystalline material in the
samples.
The crystallinity index (Cr.I.), which expresses the relative degree of
crystallinity [50], was evaluated by using the following equation [66,79,
82,83]:
Cr:I ¼ ðI002 −I am Þ=I002 ðA:3Þ
where I002 and Iam have the same meaning as in Eq. (A.3).
The crystallite size in the (002) plane of sisal fiber was calculated by
using Scherrer's formula [66,82,84]:
L ¼ kλ=ðβ cosθÞ ðA:4Þ
where k = 0.89 is Scherrer's constant; β is the peak's full width half
maxima (FWHM); λ is the wavelength of X-ray beam (1.54 Å) and θ is
the angle of diffraction.
Appendix B
The probability function of this double-parameter distribution is
given by:
m−1
m
f ðxÞ ¼ mðxÞ exp −x ðB:1Þ
where f(x) is the frequency distribution of the variable x, and m is the
Weibull modulus, which indicates the scatter of test data: as m gets
larger, the distribution narrows.
14 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16

By defining the variable x as σ/σ0 and integrating both sides, the where σ is the measured failure stress and σ0 is a normalizing
failure probability P can be calculated using Eq. (B.2): parameter and defined as characteristic strength. The failure probability
    function also can be expressed by
Z∞
σ σ
P ¼ 1− f d ðB:2Þ
σ0 σ0  m
σ=σ 0 P ¼ 1−exp −ðσ=σ 0 Þ : ðB:3Þ

1 Raw fiber 1 PC-5c

0 0

lnln[1/(1-P)]
lnln[1/(1-P)]

-1 0.2
-1 0.4

Regular residual

Regular residual
0.1 0.2
0.0 0.0

-2 -0.1
-2 -0.2
-0.2 -0.4
-0.3 -0.6
6.1 6.2 6.3 6.4 6.5
-3 Independent Variable -3 5.4 5.6 5.8
Independent Variable
6.0 6.2

6.1 6.2 6.3 6.4 6.5 6.6 6.7 5.2 5.4 5.6 5.8 6.0 6.2 6.4 6.6
lnσ lnσ

1 MK10-5c 1 MK30-5c

0 0

lnln[1/(1-P)]
lnln[1/(1-P)]

-1 -1 0.2
Regular residual

Regular residual
0.2
0.0
0.0
-0.2
-2 -0.2 -2 -0.4

-0.4 -0.6
-0.8
-3 5.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5
Independent Variable -3 6.3 6.4
Independent Variable
6.5 6.6

5.7 5.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.2 6.3 6.4 6.5 6.6 6.7

1 PC-15c 1 MK10-15c

0 0
lnln[1/(1-P)]
lnln[1/(1-P)]

-1 0.4 -1 0.4
Regular residual

0.3
Regular residual

0.2 0.2
0.1 0.0
0.0
-2 -0.1 -2 -0.2
-0.2 -0.4
-0.3
-0.4 -0.6

-3 5.0 5.1 5.2 5.3 5.4 5.5 5.6 5.7

Independent Variable
-3 5.2 5.4 5.6
Independent Variable
5.8 6.0 6.2

4.9 5.0 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4

1 MK30-15c PC-30c
1

0 0
lnln[1/(1-P)]
lnln[1/(1-P)]

-1 0.4
-1
Regular residual

0.2 0.2
Regular residual

0.0
-0.2 0.0

-2 -0.4
-0.6
-2 -0.2

-0.8 -0.4
-1.0
6.0 6.1 6.2 6.3 6.4 6.5 4.1 4.2 4.3 4.4 4.5 4.6
-3 Independent Variable -3 Independent Variable

6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.7 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8
lnσ lnσ

1 MK10-30c 1 MK30-30c

0 0
lnln[1/(1-P)]
lnln[1/(1-P)]

-1 0.2 -1 0.3
Regular residual

Regular residual

0.1
0.2
0.0
0.1
-0.1
-2 -0.2
-2 0.0
-0.1
-0.3
-0.4 -0.2
5.7 5.8 5.9 6.0 6.1 6.2 6.3
-3 4.9 5.0 5.1 5.2
Independent Variable
5.3 5.4 5.5
-3 Independent Variable

4.8 4.9 5.0 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.7 5.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5 6.6
lnσ lnσ

Fig. B.1. The Weibull plots of raw fiber and fibers embedded in PC and blends (MK10, MK30) subjected to 5 (5c), 15 (15c), and 30 (30c) wetting and drying cycles.
 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16
By taking the logarithm of both sides twice, it yields:
lnln½1=ð1−P Þ ¼ mlnσ−mlnσ 0 ðB:4Þ
or
ln½−lnð1−P Þ ¼ mlnσ −mlnσ 0 : ðB:5Þ
It can be seen that the relationship between lnln[1 / (1 − P)] and lnσ
is linear and the Weibull modulus and characteristic strength can be de-
termined by plotting the straight line. m is the slope and σ0 is the inter-
cept at lnln[1 / (1 − p)] = 0, which means p = 1 − (1 / e) = 0.63212.
Actually, the failure probability P can be evaluated from the
provability index:
P ¼ R=ðN þ 1Þ ðB:6Þ
where R is the number of fibers broken at stress σR, and N is the total
number of sisal fibers that have been tested (here N = 20).
A Weibull distribution analysis of the data was performed (Fig. B.1)
which showed a good fit with the statistical analysis of experimental
data with R2 ≥ 0.95.
References
[1] F.d.A. Silva, N. Chawla, R.D.d.T. Filho, Tensile behavior of high performance natural
(sisal) fibers, Compos. Sci. Technol. 68 (2008) 3438–3443.
[2] S.V. Joshi, L.T. Drzal, A.K. Mohanty, S. Arora, Are natural fiber composites environ-
mentally superior to glass fiber reinforced composites? Compos. A: Appl. Sci.
Manuf. 35 (2004) 371–376.
[3] P. Soroushian, J.-P. Won, M. Hassan, Durability characteristics of CO2-cured cellulose
fiber reinforced cement composites, Constr. Build. Mater. 34 (2012) 44–53.
[4] G. Ramakrishna, T. Sundararajan, Studies on the durability of natural fibres and the
effect of corroded fibres on the strength of mortar, Cem. Concr. Compos. 27 (2005)
575–582.
[5] M.A. Aziz, P. Paramasivam, S.L. Lee, Prospects for natural fibre reinforced concretes
in construction, Int. J. Cem. Compos. Light. Concr. 3 (1981) 123–132.
[6] H.E. Gram, Durability of Natural Fibers in Concrete, Swedish Cement and Concrete
Research Institute, 1983.
[7] F.d.A. Silva, B. Mobasher, C. Soranakom, R.D.T. Filho, Effect of fiber shape and
morphology on interfacial bond and cracking behaviors of sisal fiber cement
based composites, Cem. Concr. Compos. 33 (2011) 814–823.
[8] F. Pacheco-Torgal, S. Jalali, Cementitious building materials reinforced with vegetable
fibres: a review, Constr. Build. Mater. 25 (2011) 575–581.
[9] R.S.P. Coutts, A review of Australian research into natural fibre cement composites,
Cem. Concr. Compos. 27 (2005) 518–526.
[10] G. Ramakrishna, T. Sundararajan, Impact strength of a few natural fibre reinforced
cement mortar slabs: a comparative study, Cem. Concr. Compos. 27 (2005)
547–553.
[11] A.E.N. Jose Castro, Cement mortar reinforced with natural fibers, ACI Materials
Journal 78 (1981) 69–78.
[12] H. Savastano Jr., S.F. Santos, M. Radonjic, W.O. Soboyejo, Fracture and fatigue of
natural fiber-reinforced cementitious composites, Cem. Concr. Compos. 31 (2009)
232–243.
[13] F. de Andrade Silva, B. Mobasher, R.D.T. Filho, Fatigue behavior of sisal fiber rein-
forced cement composites, Mater. Sci. Eng. A 527 (2010) 5507–5513.
[14] D.G. Swift, R.B.L. Smith, The flexural strength of cement-based composites using low
modulus (sisal) fibres, Composites 10 (1979) 145–148.
[15] F.d.A. Silva, D. Zhu, B. Mobasher, C. Soranakom, R.D. Toledo Filho, High speed tensile
behavior of sisal fiber cement composites, Mater. Sci. Eng. A 527 (2010) 544–552.
[16] F.d.A. Silva, R.D.T. Filho, J.d.A.M. Filho, E.d.M.R. Fairbairn, Physical and mechanical
properties of durable sisal fiber-cement composites, Constr. Build. Mater. 24
(2010) 777–785.
[17] H. Baluch, Y.N. Ziraba, A.K. Azad, Fracture characteristics of sisal fibre reinforced
concrete, Int. J. Cem. Compos. Light. Concr. 9 (1987) 157–168.
[18] F.d.A. Silva, B. Mobasher, R.D.T. Filho, Cracking mechanisms in durable sisal fiber
reinforced cement composites, Cem. Concr. Compos. 31 (2009) 721–730.
[19] R.D. Tolêdo Filho, K. Scrivener, G.L. England, K. Ghavami, Durability of alkali-
sensitive sisal and coconut fibres in cement mortar composites, Cem. Concr.
Compos. 22 (2000) 127–143.
[20] R.D. Toledo Filho, F.d.A. Silva, E.M.R. Fairbairn, J.d.A.M. Filho, Durability of compres-
sion molded sisal fiber reinforced mortar laminates, Constr. Build. Mater. 23 (2009)
2409–2420.
[21] B.J. Mohr, H. Nanko, K.E. Kurtis, Durability of kraft pulp fiber-cement composites to
wet/dry cycling, Cem. Concr. Compos. 27 (2005) 435–448.
[22] M. Sivaraja, N. Kandasamy, M. Velmani, Pillai, Study on durability of natural fibre
concrete composites using mechanical strength and microstructural properties,
Bull. Mater. Sci. 33 (2010) 719–729.
15
[23] T.E. Omoniyi, B.A. Akinyemi, Durability based suitability of bagasse–cement com-
posite for roofing sheets, J. Civ. Eng. Constr. Technol. 3 (2012) 280–290.
[24] D. MARUTHACHALAM, Studies of durability aspects of fiber reinforced concrete, Int.
J. Eng. Sci. Technol. 4 (2012) 620–628.
[25] M. Ramli, W.H. Kwan, N.F. Abas, Strength and durability of coconut-fiber-reinforced
concrete in aggressive environments, Constr. Build. Mater. 38 (2013) 554–566.
[26] B.J. Mohr, J.J. Biernacki, K.E. Kurtis, Supplementary cementitious materials for miti-
gating degradation of kraft pulp fiber-cement composites, Cem. Concr. Res. 37
(2007) 1531–1543.
[27] J. Claramunt, M. Ardanuy, J.A. García-Hortal, R.D.T. Filho, The hornification of vegeta-
ble fibers to improve the durability of cement mortar composites, Cem. Concr.
Compos. 33 (2011) 586–595.
[28] J.-P.W. Parviz Soroushian, Maan Hassan, Sustainable processing of cellulose fiber
cement composites, ACI Mater. J. 110 (2013) 305–314.
[29] R. MacVicar, L.M. Matuana, J.J. Balatinecz, Aging mechanisms in cellulose fiber
reinforced cement composites, Cem. Concr. Compos. 21 (1999) 189–196.
[30] H.E. Gram, Durability of natural fiber in concrete, Research Report, Swedish Cement
and Concrete Research Institute, Stockholm, Sweden, 1983.
[31] S. SM, Alkali resistance of some vegetable fibers and their adhesion with portland
cement, Res. Indones. 15 (1985) 121–126.
[32] J.d.A. Melo Filho, F.d.A. Silva, R.D. Toledo Filho, Degradation kinetics and aging mech-
anisms on sisal fiber cement composite systems, Cem. Concr. Compos. 40 (2013)
30–39.
[33] C. Baley, F. Busnel, Y. Grohens, O. Sire, Influence of chemical treatments on surface
properties and adhesion of flax fibre–polyester resin, Compos. A: Appl. Sci. Manuf.
37 (2006) 1626–1637.
[34] J.M. van Hazendonk, J.C. van der Putten, J.T.F. Keurentjes, A. Prins, A simple experi-
mental method for the measurement of the surface tension of cellulosic fibres and
its relation with chemical composition, Colloids Surf. A Physicochem. Eng. Asp. 81
(1993) 251–261.
[35] S.G. Bergström, H.-E. Gram, Durability of alkali-sensitive fibres in concrete, Int. J.
Cem. Compos. Light. Concr. 6 (1984) 75–80.
[36] S.M. Singh, Alkali resistance of some vegetable fibers and their adhesion with
Portland cement, Res. Indones. 15 (1985) 121–126.
[37] C.D. Lawrence, 4 — The constitution and specification of Portland cements, Lea's
Chemistry of Cement and Concrete, Fourth ed.Butterworth-Heinemann, Oxford,
2003, pp. 131–193.
[38] F. Puertas, H. Santos, M. Palacios, S. Martínez-Ramírez, Polycarboxylate
superplasticiser admixtures: effect on hydration, microstructure and rheological
behaviour in cement pastes, Adv. Cem. Res. 17 (2005) 77–89.
[39] A.B.L. Pimentel, H. Savastano, Durability of cellulose–cement composites modified
by polymer, Eng. Agric. 26 (2006) 344–353.
[40] G.H. Berhane Z, Effect of cement composition on the durability of natural fibres in
mortar, Proc of the Fourth Intl Conf on Durability of Building Materials and
Components Pergamon Press, Singapore, 1987, pp. 203–210.
[41] K.L. Litherland, D.R. Oakley, B.A. Proctor, The use of accelerated ageing procedures to
predict the long term strength of GRC composites, Cem. Concr. Res. 11 (1981) 455–466.
[42] J. Wei, C. Meyer, Degradation rate of natural fiber in cement composites exposed to
various accelerated aging environment conditions, Corros. Sci. 88 (2014) 118–132.
[43] K. Andersson, B. Allard, M. Bengtsson, B. Magnusson, Chemical composition of
cement pore solutions, Cem. Concr. Res. 19 (1989) 327–332.
[44] L.R. Loon, M.A. Glaus, Review of the kinetics of alkaline degradation of cellulose in
view of its relevance for safety assessment of radioactive waste repositories,
J. Environ. Polym. Degr. 5 (1997) 97–109.
[45] J.L.G.A. Hidalgo, M.C. Cruz, L. Fernadez, L. Fernandez, C. Andrade, Testing methodol-
ogy for pH determination of cementitious materials—application to low pH binders
for use in HLNWR, Proc 2nd Workshop on R&D on Low-pH Cement for a Geological
Repository, 2005 (Madrid, Spain).
[46] F. Xu, J.X. Sun, C.F. Liu, R.C. Sun, Comparative study of alkali- and acidic organic
solvent-soluble hemicellulosic polysaccharides from sugarcane bagasse, Carbohydr.
Res. 341 (2006) 253–261.
[47] R.G.P. Viera, G.R. Filho, R.M.N. de Assunção, C.d.S. Meireles, J.G. Vieira, G.S. de
Oliveira, Synthesis and characterization of methylcellulose from sugar cane bagasse
cellulose, Carbohydrate Polymers 67 (2007) 182–189.
[48] B.L. B., Methods in Wood Chemistry, Interscience Publishers, New York, 1967.
[49] H.I. W, Official Methods of Analysis of the Association of Official Analytical Chemists,
AOAC, Washington, 1997.
[50] L. Segal, J.J. Creely, A.E. Martin, C.M. Conrad, An empirical method for estimating the
degree of crystallinity of native cellulose using the X-ray diffractometer, Text. Res. J.
29 (1959) 786–794.
[51] J. He, S. Cui, S.-y. Wang, Preparation and crystalline analysis of high-grade bamboo
dissolving pulp for cellulose acetate, J. Appl. Polym. Sci. 107 (2008) 1029–1038.
[52] E.-L. Hult, T. Iversen, J. Sugiyama, Characterization of the supermolecular structure of
cellulose in wood pulp fibres, Cellulose 10 (2003) 103–110.
[53] C.J. Garvey, I.H. Parker, G.P. Simon, On the interpretation of X-ray diffraction powder
patterns in terms of the nanostructure of cellulose I fibres, Macromol. Chem. Phys.
206 (2005) 1568–1575.
[54] E.M. Fernandes, J.F. Mano, R.L. Reis, Hybrid cork–polymer composites containing
sisal fibre: morphology, effect of the fibre treatment on the mechanical properties
and tensile failure prediction, Compos. Struct. 105 (2013) 153–162.
[55] A. C1018, ASTM C1018-97 Standard Test Method for Flexural Toughness and First-
Crack Strength of Fiber-Reinforced Concrete (Using Beam With Third-Point Loading),
ASTM International, West Conshohocken, PA, 1997.
[56] R.J. van Eijk, H.J.H. Brouwers, Prediction of hydroxyl concentrations in cement pore
water using a numerical cement hydration model, Cem. Concr. Res. 30 (2000)
1801–1806.
16 J. Wei, C. Meyer / Cement and Concrete Research 73 (2015) 1–16
[57] D. Rothstein, J.J. Thomas, B.J. Christensen, H.M. Jennings, Solubility behavior of Ca-S-,
Al-, and Si-bearing solid phases in Portland cement pore solutions as a function of
hydration time, Cem. Concr. Res. 32 (2002) 1663–1671.
[58] F. Deschner, F. Winnefeld, B. Lothenbach, S. Seufert, P. Schwesig, S. Dittrich, F. Goetz-
Neunhoeffer, J. Neubauer, Hydration of Portland cement with high replacement by
siliceous fly ash, Cem. Concr. Res. 42 (2012) 1389–1400.
[59] J. Duchesne, M.A. Be´rube´, Effect of supplementary cementing materials on the
composition of cement hydration products, Adv. Cem. Based Mater. 2 (1995) 43–52.
[60] Gustavo Henrique Denzin Tonoli, Alessandra Etuko Feuzicana de Souza Almeida,
Marcelo Assumpção Pereira-da-Silva, Alexandre Bassa, Danilo Oyakawa, Holmer
Savastano Jr., Surface properties of eucalyptus pulp fibres as reinforcement of
cement-based composites, Holzforschung 64 (2010) 595–601.
[61] C.D. Johnston, Å. Skarendahl, Comparative flexural performance evaluation of steel
fibre-reinforced concretes according to ASTM C1018 shows importance of fibre
parameters, Mater. Struct. 25 (1992) 191–200.
[62] J.K.W., D.O. AL-GHAMDY, E. TONS, Flexural toughness of steel fiber reinforced
concrete, JKAU:Eng. Sci. 6 (1994) 81–97.
[63] H. Gu, Tensile behaviours of the coir fibre and related composites after NaOH treat-
ment, Mater. Des. 30 (2009) 3931–3934.
[64] A.K. Bledzki, J. Gassan, Composites reinforced with cellulose based fibres, Prog.
Polym. Sci. 24 (1999) 221–274.
[65] D. Ray, B.K. Sarkar, Characterization of alkali-treated jute fibers for physical and me-
chanical properties, J. Appl. Polym. Sci. 80 (2001) 1013–1020.
[66] A. Roy, S. Chakraborty, S.P. Kundu, R.K. Basak, S. Basu Majumder, B. Adhikari,
Improvement in mechanical properties of jute fibres through mild alkali treatment
as demonstrated by utilisation of the Weibull distribution model, Bioresour.
Technol. 107 (2012) 222–228.
[67] I.M. De Rosa, J.M. Kenny, D. Puglia, C. Santulli, F. Sarasini, Morphological, thermal
and mechanical characterization of okra (Abelmoschus esculentus) fibres as potential
reinforcement in polymer composites, Compos. Sci. Technol. 70 (2010) 116–122.
[68] G.H.D. Tonoli, U.P. Rodrigues Filho, H. Savastano Jr., J. Bras, M.N. Belgacem, F.A. Rocco
Lahr, Cellulose modified fibres in cement based composites, Compos. A: Appl. Sci.
Manuf. 40 (2009) 2046–2053.
[69] W.A. HUNNICUTT, Characterization of Calcium–Silicate–Hydrate and Calcium-
Alumino-Silicate-Hydrate, University of Illinois, Urbana-Champaign, 2013.
[70] S. Jalali, A. Peyroteo, R.M. Ferreira, From Waste Product to Metakaolin: Beneficial
Impact on Performance Parameters, 2006.
[71] J. Wei, C. Meyer, Sisal fiber-reinforced cement composite with Portland cement
substitution by a combination of metakaolin and nanoclay, J. Mater. Sci. 49 (2014)
7604–7619.
[72] J.O.B. Sunkyu Park, Michael E. Himmel, Philip A. Parilla, David K. Johnson, Cellulose
crystallinity index: measurement techniques and their impact on interpreting
cellulase performance, Biotechnol. Biofuels 3 (2010) 1–10.
[73] J. Wei, C. Meyer, Improving degradation resistance of sisal fiber in concrete through
fiber surface treatment, Appl. Surf. Sci. 289 (2014) 511–523.
[74] E. Jakab, O. Faix, F. Till, Thermal decomposition of milled wood lignins studied by
thermogravimetry/mass spectrometry, J. Anal. Appl. Pyrolysis 40–41 (1997)
171–186.
[75] R. Sun, Q. Lu, X.F. Sun, Physico-chemical and thermal characterization of lignins from
Caligonum monogoliacum and Tamarix spp, Polym. Degrad. Stab. 72 (2001)
229–238.
[76] S. Bourbigot, S. Chlebicki, V. Mamleev, Thermal degradation of cotton under linear
heating, Polym. Degrad. Stab. 78 (2002) 57–62.
[77] B.J. Mohr, J.J. Biernacki, K.E. Kurtis, Microstructural and chemical effects of wet/dry
cycling on pulp fiber-cement composites, Cem. Concr. Res. 36 (2006) 1240–1251.
[78] A.K. Vijay, K. Kaushik, Sushee Kalia, Effect of mercerization and benzoyl peroxide
treatment on morphology, thermal stability and crystallinity of sisal fibers, Int. J.
Text. Sci. 1 (2012) 101–105.
[79] M.F. Dasong Dai, Characteristic and performance of elementary hemp fibre, Mater.
Sci. Appl. 1 (2010) 336–342.
[80] M.B. Roncero, A.L. Torres, J.F. Colom, T. Vidal, The effect of xylanase on lignocellulosic
components during the bleaching of wood pulps, Bioresour. Technol. 96 (2005)
21–30.
[81] L.Y. Mwaikambo, M.P. Ansell, Chemical modification of hemp, sisal, jute, and kapok
fibers by alkalization, J. Appl. Polym. Sci. 84 (2002) 2222–2234.
[82] R.G. Elenga, G.F. Dirras, J. Goma Maniongui, P. Djemia, M.P. Biget, On the microstruc-
ture and physical properties of untreated raffia textilis fiber, Compos. A: Appl. Sci.
Manuf. 40 (2009) 418–422.
[83] M. Le Troedec, D. Sedan, C. Peyratout, J.P. Bonnet, A. Smith, R. Guinebretiere, V.
Gloaguen, P. Krausz, Influence of various chemical treatments on the composition
and structure of hemp fibres, Compos. A: Appl. Sci. Manuf. 39 (2008) 514–522.
[84] A.M. Hindeleh, D.J. Johnson, An empirical estimation of Scherrer parameters for
the evaluation of true crystallite size in fibrous polymers, Polymer 21 (1980)
929–935.