Cement and Concrete Research 70 (2015) 9–20

Contents lists available at ScienceDirect

Cement and Concrete Research

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

Hydrophobic concrete using waste paper sludge ash

Hong S. Wong ⁎, Robert Barakat, Abdulla Alhilali, Mohamed Saleh, Christopher R. Cheeseman
Department of Civil and Environmental Engineering, Imperial College London, SW7 2AZ, UK

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

Article history: The feasibility of using a low-cost super hydrophobic powder as water-resisting admixture or water-repellent
Received 29 October 2014 surface coating for concrete has been investigated. The powder was produced from paper sludge ash (PSA), a
Accepted 6 January 2015 by-product from the manufacture of recycled paper. The effect of hydrophobic PSA on workability, strength
Available online 16 January 2015
and transport properties, including sorptivity, water absorption, diffusivity, permeability and electrical conduc-
tivity is reported. Samples were prepared at water/cement ratio of 0.38, cured up to 28 days and conditioned
Keywords:
Durability (C)
at 50 °C to constant mass prior to testing. It was found that replacing Portland cement with 12% hydrophobic
Transport properties (C) PSA reduced water absorption, sorptivity and conductivity by 84%, 86% and 85% respectively, with no major det-
Admixture (D) rimental effects on hydration, strength and density. When used as a surface coating, the hydrophobic PSA re-
Concrete (E) duced both absorption and sorptivity by 85–99% depending on the adhesive used. Samples surface coated with
Hydrophobic paper sludge ash hydrophobic PSA showed excellent water repelling and self-cleaning characteristics.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction advantage over surface protection because successful application is

not vulnerable to deterioration, does not require regular maintenance
Ingress of water is the main cause of all the major physical and and can be used in structures where surface protection is difficult to
chemical degradation processes affecting concrete structures. Water apply.
may be the agent causing distress or it may allow the penetration of ag- A range of integral water-resisting admixtures are commercially
gressive species that cause damage. Furthermore, excessive water pen- available. Many claims have been made on their effectiveness, including
etration compromises the functionality of certain structures, e.g. the provision of permanent reduction in water transport, superior resis-
basements, reservoirs and containment structures. Commonly used ap- tance to deterioration processes and increased service life. However, a
proaches to achieve durable watertight concrete involve increasing the recent review by the Concrete Society [8] found a distinct lack of inde-
density by using a low water–cement ratio, using supplementary ce- pendent data to substantiate these claims. Most of the available data is
mentitious materials and limiting crack widths using reinforcement or from manufacturers and much of the existing peer-reviewed research
using surface protection technologies [1–4]. However, it is very difficult has examined generic materials rather than proprietary products. Con-
to eliminate water ingress because concrete is inherently porous and stituents in these products are usually unknown and the mechanisms of
micro-cracked, it is often exposed to harsh environments and it is pre- how these admixtures work are also often not reported. This is especial-
pared using variable site practices. ly true where the product is a mix of several components. Some water-
An alternative approach to limit water ingress is through the use of resisting admixtures contain workability aids such as plasticizers that
integral water-resisting admixtures. These exist in solid or liquid form reduce the water/cement (w/c) ratio and produce a more dense con-
and can be classified broadly as: 1) materials that alter the surface ten- crete. Therefore, caution must be exercised when assessing these prod-
sion or surface energy within pores and cracks to increase liquid contact ucts because the improved performance may be due in part to the
angle and thereby resistance to absorption (i.e. water repellents, change in w/c ratio [8]. Some products appear to reduce water transport
hydrophobers) or 2) materials that increase the resistance to water pen- in low-grade porous concrete (high w/c, limited curing), but have limit-
etration under pressure by pore blocking resulting from deposition of ed effect on normal grade concrete [1,10]. Therefore these admixtures
solids via chemical reaction or removal from suspension (i.e. densifiers, seem to do little to improve the performance of concrete beyond what
permeability reducers, crystalline admixtures) [5–9]. Hydrophobic can already be achieved by good mix design and construction.
agents typically consist of fatty acids, vegetable oils, animal fats, wax Paper sludge ash (PSA) is a waste generated by the paper recycling
emulsions, hydrocarbons, silanes and siloxanes and some proprietary industry. It is produced when dewatered waste paper sludge, a by-
admixtures that contain combinations of these. This approach has the product of the de-inking and re-pulping of paper, is combusted to re-
duce waste volume and to produce energy. The combustion process is
⁎ Corresponding author. Tel.: +44 20 7594 5956. regulated in the EU and typically uses fluidised bed combustion at
E-mail address: [email protected] (H.S. Wong). ~850 °C to 1100 °C. The UK produces ~ 125,000 tonnes/year of PSA, of

http://dx.doi.org/10.1016/j.cemconres.2015.01.005
0008-8846/© 2015 Elsevier Ltd. All rights reserved.
10 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20

which 70% is used in low value applications such as land spreading, conditions and produced a hydrophobic powder [22]. Dry milling was
while the remaining 30% is typically sent to landfill [11,12]. The compo- carried out using a 3 litre porcelain ball mill containing 19 mm diameter
sition and properties of PSA vary according to the feedstock (dewatered high density alumina milling media. The milling media to sample ratio
waste paper sludge sent to the fluidised bed combustion unit) and com- was 5:1. Milling breaks down the PSA agglomerates and decreases the
bustion conditions, but it is highly alkaline with pH of 12–13 and is com- median particle size (d50) from ~100 μm to 2–5 μm, but has no signifi-
posed mainly of oxides of calcium, silicon and aluminium. The drive cant effect on mineral composition. The resultant powder had a water
towards resource efficiency and reduced landfill has promoted research contact angle of 153° measured using the sessile drop method (Krüss
into the beneficial reuse of PSA. Easy Drop tensiometer) and as a result is classified as super-
Previous PSA research has focussed on hydraulic properties, pozzo- hydrophobic. It is light grey in colour and has a specific gravity of 2.85.
lanic reactivity and potential use as a supplementary cementitious ma- Portland cement CEM I 52.5N complying with BS EN 197-1:2011 was
terial. For example, Pera and Amrouz [13] showed that calcining paper used as the main binder in all samples, supplied by Hope Cement Ltd.
sludge at 700–750 °C produces a reactive pozzolanic material and this Thames Valley siliceous gravel (b10 mm) and sand (b5 mm) were
has been confirmed by others [14–17]. PSA can have cementitious prop- used as coarse and fine aggregates respectively. The gravel complied
erties; it reacts with water, sets and hardens. However, it has very high with BS EN 12620:2002 + A1 overall grading, and the sand complied
water demand due to high porosity, the developed strength are low for with the BS 882:1992 medium grading. The gravel had a specific gravity
structural applications and the presence of free lime causes problematic of 2.48, moisture content of 1.3% and 24-h absorption of 2%. The sand
expansion [14,18]. The presence of trace metallic aluminium in PSA has had a specific gravity of 2.54, moisture content of 1.3% and 24-h absorp-
also been reported to cause swelling and expansion under alkaline con- tion of 1.8%. Tap water was used as the batch water.
ditions [17]. Shrinkage-induced cracking is also another potential prob-
lem. However, blending PSA with ground granulated blastfurnace slag 2.2. Samples
improves strength development and alleviates some detrimental effects
[14,18]. Fava et al. [19] also observed that a small increase in compres- Three series of samples were prepared: Series I, cement pastes con-
sive strength can be achieved in mortar when 5% by weight of cement taining admixed hydrophobic PSA, Series II, cement pastes surface coat-
was replaced with PSA. ed with hydrophobic PSA, and Series III, concretes containing admixed
Recent research has demonstrated that PSA can be transformed into hydrophobic PSA. The free water/cement (w/c) ratio for all samples
a super-hydrophobic powder using simple, low-cost processing, involv- was fixed at 0.38. A low w/c ratio was selected so that the effectiveness
ing dry milling with stearic acid which acts as a surface functionalising of the hydrophobic PSA can be assessed on a dense concrete. Mix pro-
agent [20]. Subsequent work has optimised the type and quantity of sur- portions for all 18 samples are shown in Table 1.
face functionalising agent using a range of fatty acids and milling condi-
tions [21] and optimal process produced a super-hydrophobic powder 2.2.1. Series I
with a water contact angle of 153° [22]. Hydrophobicity results from Series I consists of pastes containing 0, 2, 4, 8 and 16 wt.% replace-
the micro-particulate texture induced by milling and the formation of ment of cement with hydrophobic PSA. The pastes were prepared by
calcium stearate self-assembling monolayer that coats the fractured dry mixing cement and hydrophobic PSA in a bowl mixer. Batch water
PSA surfaces. was then added and mixed for 3 min. Disc samples were cast in steel
The aim of this study was to evaluate the feasibility of using super- moulds (100Ø × 50 mm) and compacted in two layers using a vibrating
hydrophobic PSA powder as a partial cement replacement to improve table until no significant air bubbles escaped the surface. There was a
the resistance of concrete to water ingress. Initial studies have also ex- noticeable decrease in workability caused by the hydrophobic PSA pow-
amined applying the super-hydrophobic PSA as a surface treatment to der, particularly at 8% addition and higher. Thus, the amount of vibra-
produce a water-repellent and self-cleaning coating. The influence of tion was adjusted to account for this. Compacted samples were
the super-hydrophobic PSA on a range of properties including covered with polyethylene sheet and wet hessian for the first 24 h,
absorption, diffusion, permeation and electrical conduction has been in- then de-moulded and cured in a fog room at 20 °C and 95% ± 5% relative
vestigated and as far as we are aware, no published research has previ- humidity for 3 days. After curing, samples for transport testing were
ously been reported on the use of PSA as an integral water-resisting conditioned in an oven at 50 °C and 7% relative humidity until constant
admixture or surface coating. The novelty of the work is two-fold: mass, which was taken to be achieved when consecutive 24-hour mass
transforming a waste product into a high value super-hydrophobic readings showed less than 0.02% difference. The purpose of this was to
powder with very little processing and applying the hydrophobic pow- minimise the effect of sample moisture content on the measured mass
der to improve the performance of concrete. transport properties so that a meaningful comparison can be made.
The oven dried samples were then cooled overnight to room tempera-
2. Experimental ture in a desiccator containing silica gel prior to transport testing to en-
sure that their moisture content did not increase during cooling.
2.1. Materials Samples were taken out of the desiccator only when required for test-
ing. Checks by weighing showed negligible mass change during cooling
Paper sludge ash (PSA) was obtained from Aylesford Newsprint Ltd., and storage prior to testing.
a major UK newsprint manufacturer producing 400,000 tonnes per year
of recycled newsprint from 500,000 tonnes of waste paper fibre. This 2.2.2. Series II
generates about 70,000 tonnes of PSA every year. The PSA is mainly Series II consists of paste samples that were surface coated with hy-
calcium aluminosilicate with a composition of 61.2% CaO, 21.2% drophobic PSA. Cylindrical samples were prepared as in Series I above,
SiO2, 12.6% Al2O3, 2.8% MgO, 0.9% Fe2O3, 0.4% K2O, 0.3% TiO2, 0.2% except that they were cast in plastic moulds (50Ø × 68 mm) and
SO3, 0.1% P2O5 and 0.1% SrO as determined by X-ray fluorescence spec- contained no admixed PSA. They were cured until the age of 3 days at
troscopy. The main crystalline phases determined by XRD are gehlenite 95% ± 5% RH, 20 °C and then conditioned to constant mass at 50 °C.
(Ca2Al(AlSiO7)), calcite (CaCO3), lime (CaO) and mayenite (Ca12Al4O33). After cooling to room temperature the samples were surface treated
Scanning electron microscopy showed that the PSA is highly porous and by applying a thin layer of adhesive, then covering the surface with hy-
comprises of loose agglomerates of smaller particles. drophobic PSA, followed by a gentle pressure to set the powder into the
PSA was processed to form a hydrophobic powder by dry ball milling adhesive. The coating was left to cure for 24 h and excess hydrophobic
for 8 h using a 4 wt%. addition of stearic acid (Sigma Aldrich, 95% purity) PSA was removed with compressed air. The following adhesives were
as a surface functionalising agent. This was the optimum processing tested: a) wood varnish (Ronseal), b) high-strength polyvinyl acetate
 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20 11

Table 1
Samples and mix proportions.

Sample Cement: kg/m3 Water: kg/m3 Free w/c Hydrophobic Sand: kg/m3 Gravel: kg/m3 Curing: days Notes
PSA

% kg/m3

Series I
P-0% 1434 545 0.38 – – – – 3
P-2% 1419 539 0.38 2 28.4 – – 3
P-4% 1405 534 0.38 4 56.2 – – 3
P-8% 1378 524 0.38 8 110 – – 3
P-16% 1327 504 0.38 16 212 – – 3

Series II
P-0% 1434 545 0.38 – – – – 3 No surface coating
P-0% V 1434 545 0.38 – – – – 3 Coated with varnish only
P-0% V-PSA 1434 545 0.38 – – – – 3 Coated with varnish and PSA
P-0% P 1434 545 0.38 – – – – 3 Coated with PVA adhesive only
P-0% P-PSA 1434 545 0.38 – – – – 3 Coated with PVA adhesive and PSA
P-0% C 1434 545 0.38 – – – – 3 Coated with cyanoacrylate adhesive
P-0% C-PSA 1434 545 0.38 – – – – 3 Coated with cyanoacrylate and PSA
P-0% E 1434 545 0.38 – – – – 3 Coated with epoxy resin only
P-0% E-PSA 1434 545 0.38 – – – – 3 Coated with epoxy resin and PSA

Series III
C-0% 487 185 0.38 – – 569 1057 3, 28
C-4% 478 182 0.38 4 19.1 569 1057 3, 28
C-8% 469 178 0.38 8 37.5 569 1057 3, 28
C-12% 460 175 0.38 12 55.2 569 1057 3, 28

(PVA) adhesive (EverBuild), c) a two-component epoxy resin based ad- strength of hardened concrete was measured in accordance with
hesive (Araldite) and d) cyanoacrylate adhesive (Loctite). For each ad- BS EN 12390-3 [24] on 100 mm cubes at 3 and 28 days using four
hesive, two samples were prepared: one with only the adhesive layer replicate samples and averaged. Samples were sandwiched between
as control, and a second with adhesive and hydrophobic PSA. Prelimi- two 150 × 150 × 25 metal plates to ensure uniform loading at a rate
nary trials where the hydrophobic PSA was simply mixed with the ad- of 0.3 MPa/s. Prior to strength testing, the density of each sample was
hesive and then applied directly onto the surface did not produce a determined by Archimedes's principle in accordance with BS EN
coating with good water-repelling characteristic. Thus, it is important 12390-7:2009 [25].
to apply the adhesive as an interlayer to preserve the hydrophobicity
of the PSA while providing adequate surface binding. 2.4. Degree of hydration

2.2.3. Series III The degree of hydration of concrete was estimated by measuring the
Series III consists of concrete samples containing 0, 4, 8 and 12 wt.% non-evaporable water content of equivalent paste samples after 3 and
replacement of cement with hydrophobic PSA. Total aggregate fraction 28 days curing. Approximately 50 g from each sample was crushed,
for all mixes was maintained at 65 vol.% and sand-to-total aggregate dried at 105 °C to constant mass to remove evaporable water, and heat-
mass ratio was 0.35. Batch water was adjusted to account for aggregate ed to 1050 °C for 3 h. The non-evaporable water content was taken as
absorption so that the target free w/c ratio was achieved. An entrapped the mass loss between 105 °C and 1050 °C, corrected for the loss on ig-
air content of 1 vol.% was assumed in the mix design based on our pre- nition of the cement and PSA. The degree of hydration was then calcu-
vious measurements of similar concrete mixes. The mix design was car- lated as the ratio of the non-evaporable water content per gram of
ried out by absolute volume. Batching was done in a 30-litre capacity cement to the amount at complete hydration, which was assumed to
pan mixer by first dry mixing cement and aggregates for 60 s, followed be equal to 0.23 g/g for ordinary Portland cement based on the classical
by the hydrophobic PSA powder for 45 s. Finally, water was added and work of Powers and Brownyard [26].
mixed for a further 3 min. Cubes (100 mm) and discs (100Ø × 50 mm)
were cast in steel moulds and vibrated in three layers until no signifi- 2.5. Visual assessment of hydrophobicity
cant release of air bubbles was observed. Samples were then placed
on a polythene sheet, covered with damp hessian and a further poly- The susceptibility of the samples to wetting was assessed in a simple
thene sheet for the first 24 h. Samples were then de-moulded, weighed, qualitative manner by placing droplets of water on the surface and visu-
wrapped in cling film and sealed in polythene bags and cured at 20 °C ally inspecting them over time. In another experiment, drops of black
for 3 and 28 days. The cubes were used for density and compressive ink were placed on the sample. The sample was then tilted to allow
strength testing. Discs were conditioned at 50 °C, 7% RH until mass equi- the ink to roll off. Photographs were taken before and after to compare
librium following the same approach as in Series I for transport testing. any differences in resistance to staining. The amount of staining seen on
the sample was used as a qualitative indicator of self-cleaning and abil-
2.3. Workability, density and compressive strength ity to repel liquids.

Workability of freshly mixed paste and concrete was determined 2.6. Sorptivity, O2 diffusivity and O2 permeability
using the slump test in accordance to BS EN 12350-2 [23] for concrete.
For paste, the procedure was modified by using a mini slump cone The main focus of the research was to evaluate the effect of the
which had a height of 75 mm and diameters of 40 mm (top) and super-hydrophobic PSA on the ability of cement-based materials to ab-
90 mm (bottom). The cone was filled in three layers, compacted with sorb and transmit water by capillary suction. This was carried out by
a tamping rod 25 times per layer and then carefully lifted. The average placing disc samples (top trowelled face) on two Perspex rods in a
subsidence from the top of the cone was measured. Compressive tray of deionised water to a depth of 2–5 mm above the exposed surface
12 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20

at room temperature (21 ± 1 °C). Prior to that, the circumference of area fraction. Disc samples were split using a diamond rotary saw. The
each sample was sealed with two layers of waterproof tape to prevent cross-section (~100 × 50 mm) was then surface treated to increase con-
side absorption and to ensure unidirectional flow. The amount of trast and facilitate segmentation of the air voids. This involved grinding
water absorbed with time was measured to the nearest 0.01 g by with silicon carbide paper starting from a coarse grit of 80, followed by a
weighing at regular intervals up to 72 h. The water level in the tray grit of 120 and 500 at 70 rpm for 2 min per grit size. The flat ground sur-
was maintained at 2–5 mm above the exposed surface throughout. face was then cleaned with acetone in an ultrasonic bath to remove dust
The tray was covered to prevent sample drying. Care was taken to and slurry from the voids and then dried at 40 °C for 2 h. Two thin coat-
avoid condensation falling onto the sample. Sorptivity (g/m2min0.5) ings of black ink were then applied in perpendicular directions using a
was calculated according to the classical unsaturated flow theory [27] permanent marker and allowed to dry. A white talcum powder was
by plotting the absorbed water per unit flow area against square-root then spread on the surface and gently pressed into voids with a flat
of time, and determining the slope of the best-fit line. Six to ten mass glass slide. Excess talcum powder was wiped away with the edge of a
readings taken during the first 24 h were used to draw the best-fit razor blade. Samples were then scanned with a flatbed scanner at
line. The coefficient of regression of the least squares fit was always 1200 dpi resolution (20 μm pixel spacing). Scanning was carried out
greater than 0.98. in greyscale with the same settings applied to all samples. Images
Concrete samples from Series III were also tested for oxygen diffu- were cropped to exclude any damaged edges or large irregular voids
sion, oxygen permeation and electrical resistivity. The diffusion test caused by aggregate pull-outs. Finally, the air voids were segmented
was carried out by placing samples in a cell with the top trowelled sur- by referring to the brightness histogram, which shows two characteris-
face exposed to a stream of oxygen and the bottom cast face to a stream tic peaks representing the talcum powder-filled voids and the sur-
of nitrogen at equal pressure. The sample was fitted into a silicone rub- rounding black painted background. A threshold was set at the mid-
ber ring placed in a diffusion cell and the circumference of the sample point value between peaks to produce a binary image, from which the
was sealed by applying a lateral confining pressure of 0.57 MPa to pre- area fraction of the segmented voids was calculated.
vent side leakage. The gasses were allowed to counter diffuse through
the sample and the oxygen concentration in the nitrogen stream was 3. Results
monitored using a zirconia analyser. When a steady-state condition
was reached, the oxygen concentration and flow rates were recorded 3.1. Workability, density, compressive strength and degree of hydration
to calculate intrinsic diffusivity (m2/s) according to Fick's first law of
diffusion. Fig. 1 shows the variation in slump with the addition of the super-
Oxygen permeability was determined by placing the sample in a cell hydrophobic PSA powder and it can be seen that workability decreased
similar to the one used for diffusion, but with the top face exposed to significantly with increasing PSA content. Both paste and concrete
pressurised oxygen and the bottom face to atmospheric pressure. A mixes became considerably stiffer as the amount of PSA increased.
pressure gradient was generated by applying oxygen at input pressures There appears to be a strong linear relationship between slump and
of 0.05, 0.15 and 0.25 MPa above atmospheric pressure. At each applied PSA content, with a coefficient of regression of 0.99 for Series I and
pressure, flow was allowed to stabilise and the steady-state outflow rate 0.95 for Series III. Slump decreased at a greater rate for the concrete
was measured using bubble flow meters. The apparent permeability at samples. Nevertheless, the samples were compactable and did not
each pressure gradient was calculated according to Darcy's law for com- show any signs of honeycombing on the cast surface.
pressible fluids and the intrinsic permeability (m2) was determined by Results for density, compressive strength and degree of hydration
applying Klinkenberg's correction for gas slippage. Transport tests are presented in Table 2. Note that the standard error was calculated
were carried out in four replicates per mix. Detailed descriptions of by dividing the standard deviation (σ) by the square root of number
the test procedures are available elsewhere [28]. of replicates (n). All three properties were not significantly affected
within the range of PSA content studied. However, there was a notice-
2.7. Absorption and electrical conductivity able decrease in density at 12% PSA. Samples cured for 28 days exhibited
greater strength and degree of hydration than samples cured for 3 days.
On completion of the 72-h capillary rise test, concrete samples were There were significant increases in strength and degree of hydration be-
fully immersed in deionised water and placed in a vacuum desiccator tween 3 and 28 days and the hydration and strength values achieved
for 2 h. The vacuum was then released to allow ambient air pressure appear to be typical for the mix.
to force water in. Subsequently, samples left fully immersed in water
for a further 40 days were surface dried and weighed. Absorption was
calculated as the mass of water absorbed from the start of the pre-
conditioned state (50 °C oven dried) divided by initial mass, expressed
as a percentage.
Subsequent electrical conduction testing was carried out by
clamping samples between two brass plates connected to an LCR
data-bridge. A generous amount of a salt-free electrode gel was applied
to ensure good electrical contact between the sample and electrodes.
Electrical resistance was then measured at an alternating current fre-
quency of 1 kHz to reduce polarisation effects. The readings typically
stabilise within 1 min of connection. Three readings were taken per
sample at three consecutive minutes and averaged. Electrical conductiv-
ity was then calculated from the resistance and the sample dimensions.
Four replicate samples were tested per mix.

2.8. Air content

The air content of concrete samples from Series III were obtained by
scanning the sample with a flat-bed scanner to produce a digital image
of the cross-section and applying image analysis to measure the void Fig. 1. Effect of admixed hydrophobic PSA on slump.
 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20 13

Table 2 images were taken 30 s after placement of water drops on pre-

Density, compressive strength and degree of hydration of concrete containing admixed conditioned samples. Increased PSA content resulted in less wetting
hydrophobic PSA. Results for density and strength are average of four replicates and stan-
pffiffiffi and hydrophobicity can be observed. No significant difference was ob-
dard errors (=σ= n) are shown in brackets.
served between samples cured for 3 and 28-days. Further trials pro-
PSA (%) Density (kg/m3) Compressive strength Degree of duced a hydrophobic surface on the control mix (C-0%) either by
(MPa) hydration
forming a layer of PSA powder on the top surface of freshly compacted
3 days 28 days 3 days 28 days 3 days 28 days concrete, or by placing a layer of PSA evenly in the base of the mould
0 2368 (1.6) 2354 (6.4) 47.5 (0.5) 64.5 (0.9) 0.57 0.67 and pouring fresh concrete on top. The latter produced a more positive
4 2359 (3.8) 2354 (4.3) 47.1 (0.4) 64.0 (0.6) 0.58 0.67 result (Fig. 4f) and it was observed that water beads would run off an in-
8 2358 (1.8) 2362 (2.9) 50.0 (0.5) 65.0 (0.3) 0.58 0.68 clined surface without wetting the sample. Sprinkling PSA powder on
12 2337 (1.5) 2338 (3.0) 49.4 (0.9) 61.7 (0.6) 0.59 0.68
top of fresh concrete was less effective (Fig. 4e), probably because it
did not form a continuous surface coating.

3.2. Visual assessment of hydrophobicity

3.3. Absorption and sorptivity of pastes
Fig. 2 shows the results of the ink-staining tests. Fig. 2a shows the
samples immediately after black ink was dropped on the surface. Note Fig. 5a shows the cumulative water absorption plotted against
that there is a slight depression at the centre of each sample due to square-root of elapsed time for paste samples containing admixed
the plastic mould used during casting. All samples containing super- hydrophobic PSA (Series I). The data show typical behaviour of a rapid
hydrophobic PSA coating exhibited interesting behaviour in which the initial water uptake that gradually decreases with time. Samples
ink droplets form beads on the surface that immediately roll to the cen- containing increased additions of PSA absorbed significantly less
tre, merging into one large bead and settling at the lowest point. This is water by capillary action. The control sample showed a 12.4% increase
shown more clearly in Fig. 3. However, the control samples and samples of initial mass due to absorption and reached near saturation after
with adhesive only coatings did not exhibit this behaviour. Fig. 2b 72 h. In comparison, the paste containing 8% PSA had a 3.1% gain in
shows the effect of tilting the sample to allow the ink to roll off the sur- mass and the sample containing 16% PSA only increased in weight by
face. It can be seen that the control sample stained immediately because 1.6% (Fig. 5c).
the ink was rapidly absorbed. Samples with adhesive only coatings Fig. 5b shows the corresponding values for PSA coated samples
displayed varying amounts of staining. Samples with adhesive and (Series II). The amount of water absorbed was greatly reduced as a re-
PSA coatings were highly hydrophobic and the ink bead simply rolled sult of the surface coating. The mass of the coated samples increased
off with virtually no staining. These samples therefore exhibited a self- by an average of 0.82% compared to 12.4% for the non-coated sample.
cleaning property. The control samples with the adhesive only coating were also very ef-
Fig. 4a to d shows the effect of admixed hydrophobic PSA on the sur- fective at reducing water absorption. However the absorption of sam-
face wetting characteristics of concrete samples. For consistency, the ples with adhesive and hydrophobic PSA coating were consistently

a)

b)
Fig. 2. Visual assessment of hydrophobicity by ink staining. Samples have a diameter of 50 mm.
14 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20

Fig. 3. Close up photo showing a single large ink bead forming on the centre of samples from Series II coated with hydrophobic PSA: a) P-0% C-PSA, b) P-0% P-PSA.

lower (by 10–26%), indicating that the super-hydrophobic PSA and ad- Replacement of 2% PSA produced a 62% reduction in sorptivity com-
hesive layer provided the best performance (Fig. 5c). pared to the control. The corresponding decreases in sorptivity are
The sorptivity coefficients for all samples from Series I & II are shown 72% (for 4% PSA), 83% (for 8% PSA) and 92% (for 16% PSA). The most ef-
in Fig. 5c. The sorptivity indicates the rate at which water is absorbed via fective coatings are those based on the super-hydrophobic PSA and ei-
capillary action and this reduced as the percentage of PSA increased. ther cyanoacrylate or epoxy resin, both of which reduced absorption

a) C-0% b) C-4%

c) C-8% d) C-12%

e) C-0% with surface coated hydrophobic PSA, f) C-0% with surface coated hydrophobic PSA,
produced by sifting a layer of PSA powder on the produced by placing fresh concrete on top of a layer
surface of fresh concrete. of PSA powder in the base of mould.

Fig. 4. Photographs showing the effect of hydrophobic PSA on the wetting characteristics of concrete samples. For consistency, images were taken 30 s after placement of water droplets on
pre-conditioned samples.
 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20 15

a) b)

c)

Fig. 5. Cumulative water uptake plots, sorptivity and absorption for cement pastes containing admixed PSA (Series I) or surface coated PSA (Series II).

a) b)

Fig. 6. Cumulative water absorption plots for concretes containing admixed PSA.
16 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20

and sorptivity by more than 99% compared to the control. The least ef- super-hydrophobic PSA can withstand significant amount of hydrostat-
fective adhesive coating was the PVA (85% reduction in absorption ic pressure.
and sorptivity).
3.5. Air content

3.4. Transport properties of concretes An assumption was made that all samples were well compacted de-
spite the variability in slump. This was based on visual assessment of the
Fig. 6 a and b shows the cumulative water uptake for 3-day and external surface quality. However, there is a possibility that this as-
28-day cured concrete samples from Series III plotted against sumption is incorrect and that poor compaction may have influenced
square-root of elapsed time. Water uptake is initially rapid and transport properties. This raises the question of whether the observed
gradually decreases with time. The control was nearly saturated at trends are indeed due to the effect of the super-hydrophobic PSA. There-
the end of the experiment (72 h). Absorption decreased with PSA fore, it was necessary to carry out further testing to determine if samples
content. At 4% PSA, absorption was reduced by over 55% and 12% were affected by poor compaction and to what extent. Representative
PSA reduced absorption by 83–84%. The difference between 8% and binary images showing the segmented air voids and calculated air con-
12% PSA was negligible, suggesting that 8% may represent an optimal tent on the cross-section of concrete samples are presented in Fig. 8. It
addition. Note that the water uptake in concrete samples is approx- was found that air content only varied between 1.2% and 3.9%, but
imately half that of the pastes (Fig. 5a). This is expected because of there was a clear indication of an increase with PSA content with a re-
the diluting effect of the non-sorptive aggregates present in the con- gression coefficient of 0.99. This correlates well with the results from
crete samples. Nevertheless, the reductions in water uptake due to the slump test. Therefore, samples containing super-hydrophobic PSA
PSA were similar for pastes and concretes. are not as well compacted as the control samples. This increased air con-
Table 3 shows the averaged sorptivity, diffusivity, permeability, elec- tent is expected to influence the measured mass transport properties
trical conductivity and absorption for concretes from Series III. In Fig. 7, and the significance of this will be discussed later.
the transport coefficients are normalised by the respective control value
and plotted against PSA content to highlight trends. Note that the error 3.6. Effect of very high PSA dosage
bars indicating variability within replicates are small and not visible in
most cases. The results show that sorptivity, absorption and electrical A further trial was completed in which a paste sample was prepared
conductivity decreased with increase in PSA content. Replacement at containing 50 wt.% super-hydrophobic PSA. Initially, the paste was pre-
4% PSA decreased sorptivity by 50%, while replacement at 12% PSA de- pared at w/c ratio of 0.38, but the mix was too dry. The water content
creased sorptivity by 83–86% relative to the control. The decrease in was gradually increased up to w/c ratio of 0.70 but the mix remained
electrical conductivity ranged between 26% and 85%. The effect of PSA dry. On closer examination, it was observed that many small agglomer-
on oxygen diffusivity and oxygen permeability was less consistent. ates containing trapped mix water had formed. The water can be re-
The lowest diffusivity and permeability occurred at 8% PSA for both cur- leased if the agglomerates are squeezed, but the water agglomerates
ing ages, but there was some increase at 12% PSA. Nevertheless, the would subsequently reform. The hardened paste formed was very po-
change in diffusivity and permeability are small compared to sorptivity rous, lightweight with a density of about a third of the reference sample
and resistivity. and weak. Fig. 9a shows a cross section of the paste containing 50% PSA
At the end of the 72-hour capillary rise test, the water absorption of at w/c 0.7. Due to the large porosity and pore sizes, the sample does not
samples containing PSA were 58–84% lower than the control (Table 3). exhibit hydrophobicity on its surface. In contrast, it had a high capacity
As an additional test, all samples were fully immersed in water and for absorption and showed a 27% mass increase in the first hour, rising
placed under vacuum for 2 h. They were then left immersed for another to 37% after 24 h.
40 days and weighed to recalculate absorption. Results show that all When a drop of water is placed and allowed to roll on a layer of
samples experienced further mass gain. However, the measured ab- super-hydrophobic PSA, the powder attaches on the droplet and
sorption for the concrete samples containing super-hydrophobic PSA forms a film that eventually coats the entire surface of the droplet. The
remained much smaller than the control. In all cases, the super- water is contained and effectively trapped within the PSA film. This ef-
hydrophobic PSA remained effective under conditions of vacuum satu- fect is shown in Fig. 9b. The droplet can be transferred onto a piece of
ration and long-term immersion. Increasing PSA content enhanced the paper without wetting it. However, breaking the PSA film releases the
ability of concrete to resist water ingress. This suggests that the admixed trapped water, wetting the paper. This interaction can be explained by

Table 3
pffiffiffi
Transport properties of concretes from Series III pre-conditioned at 50 °C. Results are average of four replicates and standard errors (=σ= n) are shown in brackets.

a) 3-day cured

Sample Sorptivity O2 diffusivity O2 permeability Conductivity Water absorption (%)

(g/m2·min0.5) (×10−8 m2/s) (×10−17 m2) (10−3 S/m)
After 72 h capillary rise After 2-h vacuum + 40 day immersion

C-0% 153.4 (4.8) 9.05 (0.09) 11.4 (0.68) 5.09 (0.33) 5.11 (0.06) 6.07 (0.08)
C-4% 76.5 (1.6) 9.27 (0.09) 10.1 (0.93) 3.76 (0.07) 2.13 (0.03 5.18 (0.14)
C-8% 30.7 (1.7) 7.40 (0.13) 9.92 (0.73) 3.65 (0.02) 0.95 (0.06) 4.32 (0.22)
C-12% 21.4 (0.9) 7.63 (0.14) 15.5 (0.61) 1.53 (0.13) 0.81 (0.03) 3.29 (0.06)

b) 28-day cured

Sample Sorptivity O2 diffusivity O2 permeability Conductivity Water absorption (%)

(g/m2·min0.5) (×10−8 m2/s) (×10−17 m2) (10−3 S/m)
After 72 h capillary rise After 2-h vacuum + 40 day immersion

C-0% 151.4 (4.0) 8.53 (0.12) 15.7 (1.41) 6.03 (0.05) 4.91 (0.11) 5.76 (0.05)
C-4% 77.1 (3.4) 8.44 (0.15) 12.6 (1.29) 4.33 (0.28) 2.13 (0.11) 4.51 (0.15)
C-8% 29.3 (1.2) 6.38 (0.09) 12.3 (0.55) 0.90 (0.06) 0.80 (0.04) 2.62 (0.05)
C-12% 26.0 (0.5) 7.14 (0.11) 17.6 (1.06) 1.17 (0.11) 0.82 (0.02) 2.87 (0.05)
 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20 17

a) b)

Fig. 7. Influence of admixed hydrophobic PSA on sorptivity, diffusivity, permeability, conductivity and accessible porosity of concrete from Series III. Results are normalised to the control
sample (0% PSA). Note that “Absorption*” was calculated after 72-h capillary rise, while “Absorption**” was calculated after a further 2-h vacuum saturation + 40 day immersion.

the PSA particles having both hydrophobic and hydrophilic properties, will have induced microcracking as shown in Wu et al. [31]. Therefore,
with the latter attaching to water leaving the hydrophobic surface fac- the performance of the super-hydrophobic PSA used in this research
ing outward in air. This allows water droplets to be coated with a hydro- was evaluated under a rigorous and demanding set of test conditions.
phobic PSA film. The increase in air content due to the super-hydrophobic PSA could
A similar phenomenon is likely to occur when mixing concrete with influence transport properties and so the significance of this must be ex-
PSA. A portion of the mix water is ‘locked in’ by the hydrophobic powder amined. Air voids can increase or decrease transport properties, with
and prevented from wetting the ingredients. This increases with PSA the net effect dependent on the transport mechanism under consider-
content. As a consequence, the amount of free water available is re- ation and the moisture content of the air voids. Air voids facilitate gas-
duced causing the marked fall in workability (Fig. 1) and degree of com- eous transport because they are interconnected via capillaries in the
paction (Fig. 8) observed. At very large hydrophobic PSA contents, the cement paste. However they also reduce the capacity for water absorp-
powder stabilises and entrains air bubbles formed during mixing. This tion because they generate lower suction compared to capillary pores.
produces a highly porous and lightweight but weak material as shown Empty air voids also act as non-conductive inclusions and so reduce
in Fig. 9a. An air entraining effect has also been observed using other hy- electrical conductivity. For example, it has been shown that air contents
drophobic admixtures [8,29,30]. of 2–4% can increase gaseous diffusivity and permeability by up to a fac-
tor of two relative to the control [32]. A reduction in sorptivity and con-
4. Discussion ductivity is only noticeable at air contents greater than 4% (Fig. 3 of Ref.
[32]). Since all the samples tested in this study had air contents less than
The mass transport properties in cementitious materials are influ- 4%, it is unlikely that the observed decrease in sorptivity and conductiv-
enced by a host of factors including w/c ratio, binder type, admixtures, ity with increasing PSA content (Fig. 7) is caused by air voids. Further-
aggregate content, type and duration of curing, degree of hydration more, the magnitude of the decrease in sorptivity and conductivity
and the test method employed. These factors should be taken into con- measured is far greater than what can be attributable to the effect of
sideration when assessing and comparing the effectiveness of the air voids [32].
super-hydrophobic PSA against other options for reducing water trans- Most concrete structures in service are unsaturated and so ingress of
port. Transport properties are also greatly influenced by the moisture water and water-borne aggressive agents is largely controlled by capil-
content and distribution in the sample at the time of testing. Therefore, lary absorption. This is the major transport process relevant to many ex-
it is important that samples are conditioned for an extended period of posure environments and degradation mechanisms and so its reduction
time prior to testing. By the end of the conditioning, samples will have is the key to enhancing durability. The addition of super-hydrophobic
reached a fairly consistent low moisture content so variations in mea- PSA was able to produce large decreases in both the rate (sorptivity)
sured transport cannot be attributed to sample moisture condition. as well as the amount of water absorbed, but without having any nega-
Oven-drying at 50 °C is a relatively extreme drying condition and it tive effects on strength and other properties. When used as partial

Fig. 8. Binary images showing segmented air voids from cross-sections of concretes from Series III. Air content increased with addition of PSA. Field of view of the images is ~45 × 45 mm.
18 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20

Fig. 9. a) Cross-section of paste with 50% admixed hydrophobic PSA. Sample was very porous, lightweight and weak. b) Droplet 1 placed on PSA, while droplet 2 was rolled and coated with
a thin film of PSA.

cement replacement at 12%, the hydrophobic PSA decreased sorptivity Medeiros and Helene [35] tested six commercial surface treatments
and absorption by up to 86% and 84% respectively. When used as a sur- including water repellents (silane/siloxane), acrylic, polyurethane coat-
face coating, it decreased both sorptivity and absorption by up to 99%. ing and double systems (water repellent + acrylic coating). The base
Furthermore, electrical conductivity was decreased by up to 85%. This concrete had a w/c ratio of 0.52 and was wet cured for 91 days, and
is because electrical conduction occurs through ions in the pore solu- then dried at 60 °C to constant mass prior to surface treatment and test-
tion, so the measured conductivity is dependent on the water content. ing. The coatings reduced sorptivity by 73–98%. In another study,
This explains why conductivity follows the trend of sorptivity and ab- Christodoulou et al. [36] tested concrete cores from a 43-year old motor-
sorption (Fig. 7). way bridge that were surface treated with a commercial silane-based
A w/c ratio of 0.38 was used in this study so that the performance of product. The surface treatment had 12–20 years of service and the con-
the super-hydrophobic PSA could be assessed on a dense concrete. This crete mix composition was unknown. Their results showed that the
is significant as it has been known that some integral water-resisting treated cores had lower sorptivity by 8–73% relative to the control
admixtures are only effective on porous matrixes [1,10]. Most available cores. When the control cores were treated with a new silane coating,
studies on water-resisting admixtures and surface treatments have sorptivity decreased by up to 88%.
been carried out on samples with relatively high w/c ratio, typically at Comparing the percentage reduction in water absorption or
~0.5 or higher. Nevertheless, it would be interesting to compare results. sorptivity in the literature with the result on the performance of the
Aldred et al. [10] tested a proprietary hydrophobic admixture on con- super-hydrophobic PSA suggests that it is on par with all, if not better
cretes at w/c ratio of 0.6 and 0.4, cured in a fog room up to 28 days, than most of the products tested. Perhaps another way to assess
then oven-dried at 40 °C prior to testing. They observed that the admix- the super-hydrophobic PSA is to compare the measured sorptivity
ture decreased water absorption by ~67–78% relative to the control. The against those of well cured high-performance concretes with low w/c
reduction in water transport was more pronounced in concretes of ratio. The data from studies where samples were conditioned at 50 °C
higher w/c ratio, as well as under conditions of limited curing and great- prior to testing shows that sorptivity ranged from 35–85 g/m2·min0.5
er drying. Justnes [29] reported the use of seven vegetable oils as hydro- for concretes with w/c ratio 0.3–0.4 that were wet cured for up to one
phobic admixtures at dosages of 0.5–1.5% cement weight. Tests year. This includes data from high-performance concretes containing
were carried out on mortars at w/c ratio of 0.5 that were cured for microsilica (7.5–10%) or metakaolin (10–20%) [37–40]. Comparing
28 days, then dried at 50 °C. It was observed that capillary absorption this to our data (Table 3, Fig. 5c) suggests that the super-hydrophobic
after 3 day immersion was reduced by 15–66% depending on the type PSA is able to decrease sorptivity beyond the level that can be achieved
of admixture used. Lanzón and García-Ruiz [30] tested the effect of six by low w/c ratio, prolonged curing or incorporation of reactive
waterproofing admixtures: calcium stearate, zinc stearate, sodium ole- pozzolans.
ate, silicones and hydrophobic polymer at dosages up to 2 wt.% of sam- These findings are promising as the use of super-hydrophobic PSA
ple. It was observed that absorption after 90-minute immersion was would lead to an overall improvement in concrete durability. The
reduced by 57–86% at the highest dosage. super-hydrophobic PSA decreases water absorption, but has no influ-
de Vries and Polder [33] tested nine commercially available hydro- ence on gaseous transport and therefore the rate of evaporation. There-
phobic surface treatments including silanes and silane/siloxanes in var- fore, it is expected that the moisture content of concrete will decrease
ious solvents. The concretes had w/c ratio of 0.5 and were cured for over time, leading to dryer interior and increased resistance to many de-
3 days, dried for 6 weeks at 20 °C, 65% RH and then surface treated terioration mechanisms that require water as the reaction medium. The
and tested. It was observed that the surface treatments reduced 24- benefits associated with restricting water absorption include decreasing
hour water absorption by 60–98%. In another study, Almusallam et al. ingress of water borne aggressive agents such as chlorides and sul-
[34] evaluated ten commercial surface coatings that consisted of acrylic, phates, reducing the amount of freezable water, limiting reaction rates
polymer emulsion, epoxy resin, polyurethane and chlorinated rubber and reducing leaching. The decrease in electrical conductivity is benefi-
coatings. The base concrete samples (0.45 w/c ratio) were moist cured cial from the point of view of reducing steel corrosion rates. The use of
for 14 days and then dried at 70 °C for 24 h prior to surface coating super-hydrophobic PSA allows a reduction of cement content in con-
and testing. They observed a wide variation in the performance of the crete while maintaining or improving performance. This is beneficial
coatings, even those within the same generic type. Overall, the surface in the context of sustainability, which is often considered in relation to
coatings decreased water absorption by 28–96% and decreased the embodied energy/carbon of cements and whole-life impacts of con-
sorptivity by 4–94% relative to the control. crete structures.
 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20 19

The influence of the admixed super-hydrophobic PSA on the hydra- • A decrease in workability was observed, but no major detrimental ef-
tion kinetics of cement and microstructural development is not fully un- fects on compressive strength and hydration occurred for cement re-
derstood and research is on-going in this area. Nevertheless, current placements of up to 8%. At 12% PSA, a slight reduction in strength and
results suggest that replacement of 8% by weight of cement with the density was observed due to an increase in air content. At very high
super-hydrophobic PSA has no detrimental effect on the degree of hy- replacements (50%), the super-hydrophobic PSA showed an air
dration, compressive strength and density (Table 2). At 12% PSA, a slight entraining effect which produced a lightweight, porous but weak ma-
reduction in strength and density was observed. However, there was no terial.
retardation in degree of hydration suggesting that the decreases in • When used as a surface coating, the super-hydrophobic PSA was able
strength and density are due the higher air content as shown by to reduce both absorption and sorptivity by 85–99%, depending on the
image analysis (Fig. 8). We envisage that this could be overcome adhesive used. Samples with surface coated PSA showed excellent
through the use of compatible superplasticizers to compensate for the water repelling and self-cleaning characteristics.
loss of workability. Thus, a greater replacement of cement by super- • Comparison with data from the literature suggests that the perfor-
hydrophobic PSA may be possible. mance of the super-hydrophobic PSA is similar/better than many
A fundamental question is the mechanism by which the admixed water-resisting admixtures and surface treatments. The super-
super-hydrophobic PSA influences transport properties. There are sev- hydrophobic PSA is also able to decrease sorptivity beyond the level
eral possibilities. The PSA may be reactive, forming additional hydration that can be achieved by a low w/c ratio, prolonged curing or incorpo-
products that fill void spaces and increase the density of the microstruc- ration of reactive pozzolans.
ture. At 12% cement replacement, the volume of PSA is substantial,
about 2% of the sample and therefore it may also act as pore-blocking
inert filler. If these are the main mechanisms, the PSA would cause a re- The study concludes that super-hydrophobic PSA has significant po-
duction in the volume, size and connectivity of the pore structure. The tential as an admixture or surface coating to enhance the resistance of
results suggest that this is unlikely because there were little changes concrete to water ingress, and therefore against a range of deterioration
in the measured strength, oxygen diffusivity or oxygen permeability. A mechanisms. This offers a high-value application route for a waste ma-
more plausible mechanism is that the PSA arranges itself within the mi- terial that can potentially enhance the durability and sustainability of
crostructure or bonds with the hydration products in a way that pro- concrete structures.
duces hydrophobic surfaces within the pore structure. Thus, the
volume and structure of the capillary pores are not so much affected, Acknowledgement
but the surface characteristics, in particular the solid–liquid contact
angle, are altered. This will decrease capillary suction that affects liquid The original research to produce the hydrophobic PSA was funded
transport, but not gasses, which is consistent with the results reported. by an EPSRC (voucher number 11330227) Industrial Case Award with
More research is required to determine the long-term effects of the support from Aylesford Newsprint Ltd., UPM and Smithers Pira. We
admixed super-hydrophobic PSA and surface coating on the properties would also like to thank Andrew Morris and Marcus Yio for their help
of concrete. Further work is needed to establish its performance on a with the laboratory work.
larger range of concretes with different w/c ratio, curing and condition
regimes, and to establish its effectiveness in porous and cracked sys- References
tems. Further studies should be carried out to verify the influence it [1] Concrete Society, Enhancing reinforced concrete durability: guidance on selecting
has on chloride ingress and binding, carbonation, time to damage initi- measures for minimising the risk of corrosion of reinforcement in concrete, Techni-
ation and deterioration rates. It would also be interesting to complete cal Report, 61, 2004 (Camberley).
[2] Building Research Establishment, Concrete in aggressive ground, Special Digest, 1,
more rigorous testing on the resistance to externally applied water BRE, Garston, 2005.
pressure. Research on optimising the surface coating process and under- [3] ACI Committee 201, Guide to Durable Concrete, ACI 201.2R-08, American Concrete
standing its long-term durability is also required. Institute, Farmington Hills, 2008.
[4] BS 8500-1, Concrete — complementary British standard to BS EN 206-1, Part 1:
Methods for Specifying and Guidance for the Specifier, British Standards Institution,
5. Conclusions London, 2006.
[5] R. Rixom, N. Mailvaganam, Chemical Admixtures for Concrete, Third edition E & FN
Spon, London, 1999.
This study examined the influence of a super-hydrophobic powder
[6] R.M. Edmeades, P.C. Hewlett, Cement admixtures, in: P.C. Hewlett (Ed.), Lea's Chem-
derived from waste paper sludge ash (PSA) on a range of properties of istry of Cement and Concrete, Fourth editionButterworth-Heinemann, 2003,
pastes and concretes at water/cement ratio of 0.38 and cured for up to pp. 841–905.
28 days. The aim was to determine the feasibility of the super- [7] ACI Committee 212, Chemical Admixtures for Concrete, ACI 212.3R-10, American
Concrete Institute, Farmington Hills, 2010.
hydrophobic PSA as an integral water-resisting admixture or as water- [8] Concrete Society, The influence of integral water-resisting admixtures on the dura-
repellent surface coating for concrete. The following conclusions can bility of concrete, CS, 174, 2013 (Camberley).
be drawn: [9] BS EN 934-2, Admixtures for concrete, mortar and grout. Concrete admixtures, Def-
initions, Requirements, Conformity, Marking and Labelling, British Standards Insti-
tution, London, 2009.
• Super-hydrophobic PSA is very effective at reducing both the amount [10] J.M. Aldred, S. Swaddiwudhipong, S.L. Lee, T.H. Wee, The effect of initial moisture
and rate of capillary water absorption. Partial replacement of cement content on water transport in concrete containing a hydrophobic admixture, Mag.
Concrete Res. 53 (2) (2001) 127–134.
with 12% super-hydrophobic PSA decreased water absorption in con- [11] WRAP, Paper Sludge Ash: A Technical Report on the Production and Use of Paper
crete by 83–84% and decreased sorptivity by 83–86% relative to the Sludge Ash, Environment Agency, UK, 2007.
control concrete prepared at the same water/cement ratio, curing [12] A.M. Dunster, Paper sludge and paper sludge ash in Portland cement manufacture,
MinRes Case Study, Building Research Establishment, Garston, 2007.
age and conditioning regime. Electrical conductivity was decreased [13] J. Pera, A. Amrouz, Development of highly reactive metakaolin from paper sludge,
by up to 85% because of the reduced internal moisture content. Adv. Cem. Based Mater. 7 (2) (1998) 49–56.
These effects increased with an increase in PSA content. [14] J. Bai, A. Chaipanich, J.M. Kinuthia, M. O'Farrell, B.B. Sabir, S. Wild, M.H. Lewis, Com-
pressive strength and hydration of wastepaper sludge ash-ground granulated blast
• Gaseous diffusion and permeation were not significantly affected by
furnace slag blended pastes, Cem. Concr. Res. 33 (2003) 1189–1202.
the super-hydrophobic PSA. [15] P. Banfill, M. Frias, Rheology and conduction calorimetry of cement modified with
• Concretes containing super-hydrophobic PSA showed significantly calcined paper sludge, Cem. Concr. Res. 37 (2007) 184–190.
lower absorption even after vacuum saturation and 40-day immer- [16] M. Frias, M.I. Sanchez de Rojas, O. Rodriguez, R. Garcia Jimenez, R. Vigil de la Villa,
Characterisation of calcined paper sludge as an environmentally friendly source of
sion in water. This indicates that the super-hydrophobic PSA can resist metakaolin for manufacture of cementitious materials, Adv. Cem. Res. 20 (1)
hydrostatic pressure to some extent. (2008) 23–30.
20 H.S. Wong et al. / Cement and Concrete Research 70 (2015) 9–20

[17] P. Segui, J.E. Aubert, B. Husson, M. Measson, Characterisation of wastepaper sludge [30] M. Lanzón, P.A. García-Ruiz, Evaluation of capillary water absorption in rendering
ash for its valorisation as a component of hydraulic binders, Appl. Clay Sci. 57 mortars made with powdered waterproofing additives, Constr. Build. Mater. 23
(2012) 79–85. (10) (2009) 3287–3291.
[18] E. Mozaffari, J.M. Kinuthia, J. Bai, S. Wild, An investigation into the strength develop- [31] Z. Wu, H.S. Wong, N.R. Buenfeld, Influence of drying-induced microcracking and re-
ment of wastepaper sludge ash blended with ground granulated blastfurnace slag, lated size effects on mass transport properties of concrete, Cem. Concr. Res. 68
Cem. Concr. Res. 39 (2009) 942–949. (2015) (2015) 35–48.
[19] G. Fava, M.L. Ruello, V. Corinaldesi, Paper mill sludge ash as supplementary cemen- [32] H.S. Wong, A.M. Pappas, R.W. Zimmerman, N.R. Buenfeld, Effect of entrained air
titious material, J. Mater. Civil Eng. 23 (6) (2011) 772–776. voids on the microstructure and mass transport properties of concrete, Cem.
[20] C. Spathi, Resource Efficient Reuse Applications for Paper Sludge AshPhD thesis Im- Concr. Res. 41 (2011) 1067–1077.
perial College, London, 2013. [33] J. de Vries, R.B. Polder, Hydrophobic treatment of concrete, Constr. Build. Mater. 11
[21] N. Young, Development of Hydrophobic Coatings From Paper Sludge AshMSc thesis (4) (1997) 259–265.
Imperial College, London, 2013. [34] A.A. Almusallam, F.M. Khan, S.U. Dulaijan, O.S.B. Al-Amoudi, Effectiveness of surface
[22] C. Spathi, N. Young, J.Y.Y. Heng, L.J.M. Vandeperre, C.R. Cheeseman, A simple method coatings in improving concrete durability, Cement Concr. Compos. 25 (2003) 473–481.
for the preparation of super-hydrophobic powder using paper sludge ash, Mater. [35] M.H.F. Medeiros, P. Helene, Surface treatment of reinforced concrete in marine en-
Lett. 142 (2015) 80–83. vironment: influence on chloride diffusion coefficient and capillary water absorp-
[23] BS EN 12350-2, Testing fresh concrete, Part 2: Slump Test, British Standards Institu- tion, Constr. Build. Mater. 23 (2009) 1476–1484.
tion, London, 2009. [36] C. Christodoulou, C.I. Goodier, S.A. Austin, J. Webb, G.K. Glass, Long-term perfor-
[24] BS EN 12390-3, Testing hardened concrete, Part 3: Compressive Strength of Test mance of surface impregnation of reinforced concrete structure with silane, Constr.
Samples, British Standards Institution, London, 2009. Build. Mater. 48 (2013) 708–716.
[25] BS EN 12390-7, Testing hardened concrete, Part 7: Density of Hardened Concrete, [37] H.A. Razak, H.K. Chai, H.S. Wong, Near surface characteristics of concrete containing
British Standards Institution, London, 2009. supplementary cementing materials, Cement Concr. Compos. 26 (7) (2004)
[26] T.C. Powers, T.L. Brownyard, Studies of the physical properties of hardened Portland 883–887.
cement paste, Research Laboratory of Portland Cement Association, Skokie, IL, USA, [38] E. Güneyisi, K. Mermedas, Comparative study on strength, sorptivity and chloride
Bulletin 22, reprinted from Journal of the American Concrete Institute (Proceed- ingress characteristics of air-cured and water-cured concretes modified with
ings), 43 (1947)1947. 101–132, 249–336, 469–504, 549–602, 669–712, 845–880, metakaolin, Mater. Struct. 40 (2007) 1161–1171.
933–992. [39] G.D. Ransinchung, B. Kumar, V. Kumar, Assessment of water absorption and chloride
[27] C. Hall, Water movement in porous building materials—I. Unsaturated flow theory ion penetration of pavement quality concrete admixed with wollastonite and
and its applications, Build. Environ. 12 (1977) 117–125. microsilica, Constr. Build. Mater. 23 (2) (2009) 1168–1177.
[28] H.S. Wong, N.R. Buenfeld, J. Hill, A.W. Harris, Mass transport properties of mature [40] H.S. Wong, M. Zobel, N.R. Buenfeld, R.W. Zimmerman, Influence of the interfacial
wasteform grouts, Adv. Cem. Res. 19 (2007) 35–46. transition zone and microcracking on the diffusivity, permeability and sorptivity
[29] H. Justnes, Low water permeability through hydrophobicity, COIN Project Report, 1, of cement-based materials after drying, Mag. Concr. Res. 61 (8) (2009) 571–689.
SINTEF Building and Infrastructure, Norway, 2008.