Application of rice husk biochar and thermally treated low silica rice husk ash to improve physical properties of cement mortar

Application of rice husk biochar and thermally treated low silica rice husk ash to improve physical properties of cement mortar

Journal Pre-proofs Application of rice husk biochar and thermally treated low silica rice husk ash to improve physical properties of cement mortar Shr...

4MB Sizes 0 Downloads 19 Views

Journal Pre-proofs Application of rice husk biochar and thermally treated low silica rice husk ash to improve physical properties of cement mortar Shravan Muthukrishnan, Souradeep Gupta, Harn Wei Kua PII: DOI: Reference:

S0167-8442(19)30059-X https://doi.org/10.1016/j.tafmec.2019.102376 TAFMEC 102376

To appear in:

Theoretical and Applied Fracture Mechanics

Received Date: Revised Date: Accepted Date:

3 April 2019 29 September 2019 29 September 2019

Please cite this article as: S. Muthukrishnan, S. Gupta, H.W. Kua, Application of rice husk biochar and thermally treated low silica rice husk ash to improve physical properties of cement mortar, Theoretical and Applied Fracture Mechanics (2019), doi: https://doi.org/10.1016/j.tafmec.2019.102376

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Elsevier Ltd. All rights reserved.

APPLICATION OF RICE HUSK BIOCHAR AND THERMALLY TREATED LOW SILICA RICE HUSK ASH TO IMPROVE PHYSICAL PROPERTIES OF CEMENT MORTAR Shravan Muthukrishnan; Souradeep Gupta; Harn Wei Kua1 Department of Building, School of Design and Environment 4 Architecture Drive, National University of Singapore, (S) 117566 Abstract Rice husk ash, an industrial by-product from boilers for energy generation and furnace for parboiling rice, has been explored as admixture to replace part of cement in structural mortar. However, due to uncontrolled burning process, the produced rice husk ash (iRHA) has low amorphous silica and contain unburnt husk particles, which affects durability of mortar containing iRHA. This study investigates thermal treatment of iRHA (TRHA) to produce ash with improved physical and chemical properties, which can then be used to reduce cement content in mortar by 20 % (by weight). Furthermore, combination of rice husk biochar (RHB) and iRHA, where RHB is used to replace 10% and 40% by weight of iRHA, is used to improve mechanical and durability properties of iRHA-RHB mortar. The performance was compared with control (without RHA) and mortar containing RHA produced under controlled laboratory condition (LabRHA). Experimental results showed that addition of TRHA increased the strength of mortar by 20% and 34% at early stage (after 7 days) and matured age (after 120 days) compared to mortar with iRHA respectively. Although strength development was similar to control, TRHA-mortar showed lower autogenous shrinkage and water permeability, indicative of its improved durability as building material. Addition of RHB, as 40% replacement of iRHA, offers improvement in long-term (120-day) compressive strength and water

1

Corresponding author: [email protected]

1

tightness by 17% and 23% respectively compared to iRHA; it also has 12% more watertightness than the control. Strength afforded by the carbon content of RHB particles and its reservoir effect eliminated autogenous shrinkage over the 6-week monitoring period. Among the tested admixtures, LabRHA offered maximum improvement in water-tightness, leading to 23% lower capillary absorption compared to control. Findings from this study suggest that thermal treatment and application of RHB can valorise industrial grade RHA as admixture in cementitious building materials, which will also reduce the need for landfilling and sequester carbon in the built environment. Keywords: Rice husk ash; biochar; cement; durability; carbon sequestration 1.Introduction The reduction in the use of natural resources and mitigation of greenhouse gas (GHG) emissions associated with concrete manufacturing and construction processes is a pressing need in the construction industry [1]. The annual global production of concrete is currently 10 billion tonnes, and it is predicted to rise to 18 billion tonnes by 2050 due to increase in population and demand for more building and infrastructure facilities [2, 3]. Although cement constitutes 20-40% of the total volume of concrete or cementitious mortar, processing and manufacturing of cement is associated with significant CO2 emission (about 0.95 tonnes for 1 tonne of cement), which contributes to approximately 7% of global CO2 emission [4]. Therefore, if the demand for cement for mortar or concrete can be reduced by using alternate materials from waste, its associated GHG emission can also be reduced [1, 2, 5]. Application of supplementary material from waste biomass in cementitious composites has attracted academic and commercial interest due to its potential to recycle and valorize organic waste; these waste may actually pose environmental hazard during 2

incineration or landfilling [1, 2, 6-8]. One such waste is rice husk, which is a major agricultural by-products produced during de-husking operation of paddy rice. It is estimated that the total mass of rice husk generated is about 20% of the 500 billion tons of paddy produced globally [9]. Disposal of rice husk can be a challenge, because its siliceous content is resistance to natural decomposition [10, 11]. However, burning of rice husk yields 20-30% ash that is rich in amorphous silica, which is generally found to improve fresh and hardened properties of concrete [12-14]. Burning method and production process is one of the key parameters influencing properties of RHA blended cement. Open heap burning of rice husk at 300-400 ˚C close to the source in rural areas are commonly practiced, which is an economical way of disposing waste rice husk in many developing countries [15, 16]. Due to relatively low density, RHA particles are easily air-borne, which leads to accumulation of particulate matters in the environment. Moreover, uncontrolled burning below 500˚C results in incomplete ignition, leading to presence of unburnt carbon and reduced pozzolanic activity of the produced ash [17]. Al-Khalaf and Yousif [17] investigated compressive strength development of cement mortar by replacing 27% weight of cement by rice husk ash produced at temperature of 450 ˚C, 500 ˚C, 550 ˚C, 600 ˚C, 700 ˚C and 850 ˚C respectively. The combustion time was fixed at 2 hours for all the production temperature values. It is interesting to note that although pozzolanicity of ash increased with production temperature, there was no significant difference in strength of RHA blended mortar with change in production temperature [17]. It might be due to presence of elemental carbon in low temperature ash, that impart toughening effect on cement composites [6, 18] although pozzolanic reactivity may be relatively low [19]. De Sensale [20] explored durability of concrete by replacing cement by 5%,10% and 15% (by weight) with RHA produced by uncontrolled burning (RRHA) and controlled 3

laboratory (CRHA) burning respectively. Results showed that both types of RHA reduced chloride diffusivity and chloride penetration by 10-20% at 15% replacement level. Addition of rice husk ash at 15% also increases resistance to degradation by hydrochloric acid, indicated by 30% lower mass loss compared to control. This was attributed to reduced permeability by filler effect and pozzolanic action of rice husk ash, which leads to precipitation of binder gel by reacting with weaker calcium hydroxide in the cementitious matrix [21]. Chindaprasirt and Rukzon [22] reported that replacing 10-20 wt.% of cement by RHA led to slight improvement (2-3%) in compressive strength at 7-day and 120 day age of mortar. Higher percentage (40%) replacement led to 20% reduction in strength due to excessive dilution effect and increase in porosity which offsets the benefit of improved hydration by addition of RHA [22]. An alternative method of valorizing rice husk is through conversion into biochar by pyrolysis at temperature 450-550˚C. Zeidabadi et al.[23] reported that replacement of 5% cement by rice husk biochar (RHB) in concrete led to approximately 12% improvement in compressive strength compared to control, attributed to filler effect of RHB particles at early age. Pyrolysis of rice husk fixes stable carbon in the physical structure of the biochar, which is sequestered in the cementitious material, instead of being emitted to the environment through incineration or decomposition in the landfill. Overall, existing studies suggest that RHA prepared under controlled conditions (for example, in laboratory facility with controlled heating rate and temperature) offers higher improvement compared to RHA prepared from uncontrolled burning. However, preparation under controlled environment will require relatively high financial investment on facilities. An alternative way to valorize rice husk from uncontrolled burning is through further thermal treatment (treated RHA, or TRHA) before adding it to cement composites. This step 4

is recommended to reduce the unburnt carbon in RHA, which leads to higher relative amount of silica on the surface compared to those directly obtained by uncontrolled burning. Application of TRHA as cement replacement material in cement mortar is relatively unexplored. Rice husk biochar (RHB) can be used in combination with RHA to potentially improve physical properties. Besides acting as filler, the high porosity of biochar contributes to retention of water initially, which can be later released for sustaining hydration in cementitious composites [24]. Furthermore, the biochar particles can provide the filler effect that strengthens the mortar containing it. Therefore, a blend of RHA and RHB may improve mechanical properties, shrinkage and permeability of cement mortar through filler effect, enhanced curing and pozzolanic action. Therefore, this study aims to investigate the influence of TRHA, and blend of RHA and RHB respectively on hydration, strength, ductility, shrinkage and permeability of cement mortar. 2. Materials and methods 2.1 Cement and sand used The fine aggregate used in the study is natural sand with size gradation conforming to ASTM C33 [25]. Maximum size, fineness modulus and specific gravity of the sand are 4 mm, 2.54 and 2.65 respectively. ASTM Type I 52.5N ordinary Portland cement (OPC) conforming to specifications stated in ASTM C150 [26] is used in this study.

5

2.2 Preparation of rice husk ash and rice husk biochar The rice husk (feedstock) used in this study is the by-product of milling operations of paddy sourced from Malaysian paddy fields. Based on processing of the rice husk, three types of rice husk ash (RHA) were produced – industrial RHA (iRHA), treated RHA (TRHA) and laboratory RHA (LabRHA, which was produced in laboratory under controlled conditions). 2.2.1 Laboratory RHA (LabRHA) The rice husk was washed thoroughly under tap water to remove any impurities in the form of sand or agricultural waste particles from the surface. The washed rice husk was then airdried (28-32 ˚C) for 24 hours. Thermal conversion of rice husk to RHA was carried out in a muffle furnace with a forced ventilation system to evacuate the gases and organic vapours during combustion of the feedstock. The heating rate was maintained at 8 ˚C/min until the steady state temperature of 500˚C was reached from initial temperature of 28˚C. The combustion was carried out for 2 hours at 500 ˚C. Grinding of the produced ash was carried out in an electrical mixer-grinder before addition in cement mortar. Fig. 1 (a), (b) and (c) shows the raw rice husk (feedstock) and rice husk ash produced after combustion and subsequent grinding respectively.

(a)

(b)

(c)

Fig. 1 (a) Washed and dried rice husk before controlled burning. (b) Rice husk ash obtained from burning rice husk (c) Rice husk powder after grinding 6

2.2.2 Rice husk ash from uncontrolled burning and treated RHA Rice husk produced by uncontrolled burning (industrial grade RHA, or iRHA) was procured from a local industry, which uses open burning of rice husk in heap to produce ash for agricultural applications. Fig. 2(a) shows a sample of iRHA, which contains traces of unburnt particles of rice husk. The RHA is black, unlike the one produced under laboratory conditions, which is due to the trapping of carbon under a surface melted layer of silica in presence of potassium oxide impurity during uncontrolled burning operation that prevent uniform distribution of heat and air circulation [27]. The iRHA was thermally treated at 425 ˚C for 2 hours at heating rate similar to that of LabRHA to produce the third type of ash, TRHA. The produced ash was subject to grinding in a mixer-blender. It can be observed from fig. 2(b) that further thermal treatment of iRHA leads to change in color from dark grey (or black) to light grey. It is due to the more complete oxidation of the trapped carbon and removal of potassium oxide coating resulting in exposure of silica ash on the surface [27].

(a)

(b)

Fig.2 (a) Industrial grade RHA before thermal treatment (b) Treated RHA after grinding

7

2.2.3 Rice husk biochar Rice husk biochar (RHB) was produced by pyrolysing washed rice husk in a hermetically sealed furnace at 500˚C, at a heating rate of 1˚C/second. The produced RHB was then removed from the furnace and sealed with metal plate to reduce oxidation. The cooled RHB was grinded using the same method as in case of RHA. 2.3 Characterization of produced rice husk ash and rice husk biochar 2.3.1 Morphology and particle size distribution The morphology of produced RHA and RHB was investigated by scanning electron microscope, operated at 15 kV accelerating potential. Particle size distribution (PSD) was determined by polarized intensity differential scattering (PIDS) using laser diffraction particle size analyser. 3-4 gram of ground RHA and RHB samples were dispersed in water by mechanical stirring and laser beam was passed through the solution. The machine software (Beckman Coulter, LS13 320)) calculated the size distribution (output) as volume equivalent of sphere diameter using the laser diffraction angle, which is influenced by particle size. 2.3.2 Elemental analysis and chemical composition of produced RHA and RHB Elemental carbon content and percentage of other inorganic elements including silicon, magnesium, calcium, potassium and aluminium in RHA and RHB were estimated from CHNS analysis and inductive coupled plasma-optical emission spectroscopy (ICP-OES). The experimental method is provided in supplementary information. The oxide composition of RHB and RHA was determined by X-ray fluorescence spectroscopy.

8

2.3.3. Crystallinity, specific surface area and porosity of RHA and RHB Crystallinity of produced RHA and RHB samples was studied using X-ray diffraction (XRD), conducted using a computer-controlled diffractometer (Shimadzu XRD-6100/7000). Before testing, the test samples were dried and grinded to below 100 µm size. The diffractometer was operated at 40 kV and 30 mA. Scanning was conducted in the angular (2θ) range of 10˚ to 60˚ at scan steps of 0.02˚. Specific surface area (SSA) and pore volume of RHA and RHB were determined by multipoint N2 adsorption using Brunauer-Emmett-Teller (BET) method. Dried powder samples were degassed over 8 hours at 105˚C prior to BET. Pore volume was calculated using BarrettJoyner-Halenda (BJH) theory from desorption isotherm data. Mercury intrusion porosimetry (MIP, AutoPore IV 9500) was used to estimate the total porosity and average pore diameter of RHA and RHB respectively. This method was chosen because of its suitability to estimate dimension of pores over a wider size range (typically 0.01 µm to 200 µm) compared to N2-BET [28, 29]. Intrusion pressure was increased from 0.003 MPa to 130 MPa and the volume of intruded mercury was measured for each pressure increment. 2.4 Mix design, mixing and curing procedure A total of six mixes were investigated in this study (table 1). These mixes were designed to understand the effect of partial replacement of cement with different proportions and combinations of RHA and RHB. The level of replacement was kept constant at 20% by weight of cement, based on the findings from literature that report up to 20% by mass of cement replacement by RHA can offer improvement in strength and permeability

9

properties of cement composites [22, 30, 31]. Ratio of mass of cementitious material to sand was kept at constant 1:2 for all mixes, while water-to-cementitious material ratio was maintained at 0.40. Polycarboxylate-based superplasticizer (SP) was used to maintain flowability of the fresh mortar pastes. The flow spread was measured in accordance to ASTM C1437 [32]. The dosage of SP varied between 0.88 – 0.96% by weight of binder (cement + RHA or RHB) to maintain similar flowability as that of control mortar (140 ± 5 mm). The dosage of SP was higher in mortar composites with RHA or RHB, compared to that of control in general. This is due to the angular morphology and porous nature of RHA and RHB, which tend to reduce the workability of the fresh paste by resisting flow and absorbing a portion of the free mixing water.

10

Table 2. Mix design of the mortar samples investigated in this study (Numbers in bracket indicate kg/m3 of each component in mortar)

Mix codes

Control iRHA20 iRHA18 RHB2 iRHA12 RHB8 TRHA20 LabRHA 20

Mix descriptions

Cement (g) / (kg/m3)

Sand (g) / (kg/m3)

RHA (g) / (kg/m3) iRHA

LabRHA

TRHA

SP (% wt. of RHB(g) Water (g) cementitiou / / (kg/m3) s material) / (kg/m3) (kg/m3)

Plain mortar without RHA/RHB

100 (932.68)

200 (1,865.36)

_

_

_

_

40 (373.07)

0.85 (7.92)

Mortar with 20% cement replaced by industrial RHA from open burning Mortar with 20% cement replaced by combination of 18% industrial RHA and 2% RHB Mortar with 20% cement replaced by combination of 12% industrial RHA and 8% RHB Mortar with 20% cement replaced by 20% treated RHA

80 (596.12)

200 (1,490.29)

20 (149.03)

_

_

_

40 (298.06)

0.95 (6.33)

80 (599.84)

200 (1,499.60)

18 (134.96)

_

_

2 (15.00)

40 (299.92)

0.88 (6.37)

80 (611.30)

200 (1,528.25)

12 (91.70)

_

_

8 (61.13)

40 (305.65)

0.96 (6.50)

80 (596.12)

200 (1,490.29)

_

_

20 (172.03)

_

40 (298.06)

0.89 (6.33)

Mortar with 20% cement replaced by 20% laboratory produced RHA under controlled conditions

80 (596.12)

200 (1,490.29)

_

20 (149.03)

_

_

40 (298.06)

0.91 (6.33)

11

Mortar for casting were prepared by mixing in a mechanical Hobart mixer with three positive speed control - low (60 RPM), medium (124 RPM) and high (255 RPM). Dry ingredients (cement, sand and RHA or RHB) were mixed at medium speed (124 RPM) for one minute. Part of SP was added to the mixing water, which was then poured slowly and uniformly in the dry mix and the mixing was continued for two more minutes. During this time, the rest of the SP was added using a dropper and the mixing was continued for two more minutes at medium speed. The side of the mixing bowl were scrapped, and mixing was continued for 1 minute at high speed (255 RPM) and 30 seconds at medium speed. The fresh mortar then was cast in three layers into respective molds on a vibration table. Vibration was carried out after casting each layer to attain sufficient compactness. After casting, the samples were covered with plastic sheet to minimize loss of water by evaporation. Samples for strength and permeability test were cured in a fog room (100% RH and 28±2˚C). 2.5 Test methods for blended cement mortar 2.5.1 Hydration kinetics Addition of supplementary materials has significant influence on hydration kinetics and total hydration of blended cement pastes. The cumulative heat generated during isothermal calorimetry can be used to compare the hydration across mixes of different compositions [33]. Control paste was prepared by adding 20 g cement with 8 g of water (W/C = 0.40). In case of blended pastes, 20% cement was replaced by RHA (or blend of RHA and RHB), while the total water-to- cementitious material (cement +RHA/RHB) was maintained at 0.40. 10-11 g of control and blended pastes were transferred to sealed glass ampoules put in the calorimeter. Heat evolved due to hydration was measured in the form of electrical signals

12

(mW) for every 300 second until 7 days from mixing. The test was run at 30˚C to simulate the outdoor temperature in the tropical region. 2.5.2 Compressive strength, fracture toughness and flexural strength Compressive strength at 7, 28 and 120 days were determined according to the ASTM C109 [34] using a load-controlled set up at loading rate of 1,250 N/s. Another set of samples were subject to displacement-controlled compression test to study the stiffness and ductility post failure of the blended cement mortars. Images of failure pattern at incremental displacement were captured using a high-resolution digital camera. The displacement rate was maintained at 2 mm/min, while the vertical displacement for each load increment was obtained by using three linear variable displacement transducers fixed between upper and bottom plates of the compression testing machine. Fracture toughness was determined from the area under the plastic zone in stress-strain curve following the method used by Eldin and Senouci [35]. The plastic energy absorbed per unit volume (E1) was computed by subtracting elastic energy released per unit volume (E2) during fracture from the total energy per unit volume (E). In case of tough material, most of the energy generated during fracture is plastic energy while for brittle material, elastic energy comprises most of the total energy at fracture. The total energy (E) was computed from area under the best-fit polynomial equation fitted to describe the experimental stress-strain data [35]. Flexural strength was tested using 40 mm x 40 mm x 160 mm prism samples according to ASTM C348 [36]. All the tests were conducted in triplicates (n =3).

13

2.5.3 Water permeability Permeability was measured by unidirectional capillary absorption by 50 mm mortar cubes over a period of 7 days. 28-day old samples were dried in a ventilated oven at 105 ˚C over 24 hours before water absorption test. The top-face of the cube was immersed in water up to depth of 2.5 -3 mm according to ASTM C1403 [37]. Capillary absorption was calculated from the mass gain of the samples at specified time periods. 2.5.4 Drying shrinkage and sealed shrinkage Drying shrinkage is the deformation of cementitious samples due to exchange of moisture with the external environment. In contrast, sealed shrinkage is due to the progress self-desiccation from water consumption (for hydration) within the cementitious matrix. Six mortar bars (25 x 25 x 285 mm steel mould) per mix were cast and sealed with teflon sheets. The hardened mortar was demoulded after 24 hours. Three of the mortar bars were sealed with aluminium sheet to prevent loss of moisture during the shrinkage monitoring period. The other three bars were left unsealed to measure the strain due to drying shrinkage. First shrinkage reading was taken at 24 hours from casting. During the measurement period, all the mortar samples were stored at constant temperature and humidity (30˚C, 65% RH). 3. Results and discussion 3.1 Characterization of produced rice husk and rice husk biochar 3.1.1 Elemental analysis and chemical composition Table 2 presents the elemental composition of iRHA, TRHA, LabRHA and RHB. RHB and iRHA are found to have highest percentage of elemental carbon (39.67% and 41.01%

14

respectively). Cook et al.[19] reported that carbon content in excess of 30% may reduce pozzolanicity of rice husk ash. Further treatment of iRHA under controlled conditions leads to reduction of elemental carbon to 15.22%, with an increase in silicon in TRHA. This is attributed to the further oxidation of carbon fixed in the structure of iRHA and unburnt rice husk particles (feedstock). However, Si in the form of amorphous silica (SiO2) determines the pozzolanicity of material. Table 2. Elemental composition of rice husk ashes and biochar used in this study Mix

C

H

N

S

Ca

Mg

Si

Al

Fe

iRHA

39.67

2.85

1.14

ND

0.20

0.12

16.57

0.24

0.19

TRHA

15.22

0.94

0.67

<0.5

0.50

0.25

30.61

0.52

0.37

LabRHA

< 0.50

< 0.50 ND

ND

0.33

0.19

37.95

< 0.01

0.08

RHB

41.01

2.12

ND

0.13

0.11

19.24

ND

0.06

0.68

Table 3 shows the chemical composition of RHB and three types of RHA used in this study. The trend is similar to that of elemental composition – LabRHA shows significantly higher SiO2 content compared to iRHA, TRHA and RHB respectively. According to ASTM C618 [38], siliceous mineral admixture can be classified as class F pozzolan if the total amount of SiO2 , Al2O3 and Fe2O3 is more than 70%. In this case, it is found that the total amount of these compounds is 90.28% and 60.59% for LabRHA and TRHA respectively. This implies that labRHA can be classified as class F pozzolan. The content of sulphur trioxide (SO3) in all RHA and RHB

15

samples is less than 4%, which, according to ASTM C618 [38], is the maximum allowable limit for application as pozzolanic admixture in cementitious material. iRHA and RHB show similar chemical composition with low SiO2 content of 15-17% and the significantly higher CO2 content (80-82%) compared to TRHA and LabRHA. In case of RHB, pyrolysis leads to fixation of carbon in the structure of biochar by decomposition of cellulosic compounds and lignin. However, in case of iRHA, uncontrolled burning leads to fast disassociation of potassium oxide (K2O), an impurity sourced from fertilizers, into molten layer of elemental potassium on the surface, which entraps carbon and accelerates crystallization of SiO2 [27]. It is evident from the relatively low fraction of K2O in iRHA compared to LabRHA and TRHA respectively. However, although LabRHA and TRHA contains relatively high SiO2 (%), the Si needs to be amorphous to contribute to pozzolanic action in RHA-cement composites. This is further discussed based on the XRD spectra of RHA and RHB in the next section. Table 3. Chemical composition (%) of produced RHA and RHB Chemical iRHA composition (%) SiO2 17.54

TRHA

LabRHA

RHB

57.71

89.89

15.77

CO2

80.80

36.01

3.84

82.57

Al2O3

0.26

0.54

0.11

0.03

Fe2O3

0.24

0.34

0.28

0.05

SO3

0.15

0.29

0.40

0.08

MgO

0.04

0.09

0.13

0.04

CaO

0.25

0.59

1.05

0.19

Na2O

_

0.01

_

_

16

K2O

0.28

1.62

2.92

0.88

3.1.2 X-ray diffraction analysis and morphology XRD diffractogram of the produced RHAs and RHB are shown in Fig. 3. A qualitative assessment of crystallinity can be performed by comparing the peaks around 2θ = 22˚ (Fig. 3) with that of narrow and high intensity peak of quartz (crystalline silica) at same diffraction angle reported by Jamil et al.[31]. The broad peaks formed at 2θ = 22˚ indicate presence of amorphous silica in the RHA and RHB samples. Some peaks can also be observed at 2θ = 26.76˚ on top of the broad amorphous background for iRHA and TRHA respectively. This corresponds to one of the major peaks for quartz (SiO2)[31], suggesting that iRHA and TRHA are partially crystalline in nature. Similar observation was reported by De Sensale [20] and Rêgo et al. [39] that RHA produced from uncontrolled burning show partial crystallinity (presence of cristobalite) compared to that produced under controlled conditions. This is attributed to crystallization of amorphous silica due to insufficient residence time (flash burning), ineffective control of burning temperature and non-uniform distribution of oxygen during the conversion of rice husk to ash [40, 41].

17

TRHA

LabRHA

RHB

Intensity (a.u.)

iRHA

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

2θ, Cu-Kα

Fig. 3. X-ray diffraction spectra of RHA and RHB Fig. 4 shows the particle morphology and surface porosity of the produced RHA and RHB. Fig 4(a) shows that the rice husk particles have rough surface with surface protrusions aligned in regular pattern. It can be observed from Fig. 4 (b) that the iRHA particles have irregular shapes with relatively smooth surface after grinding. However, unburnt particles of rice husk could be detected in iRHA (Fig. 4(c)), which was also observed in the bulk material (Fig. 4(a)). These unburnt particles get mixed in with the rice husk ash (iRHA) during grinding; therefore, iRHA is a mix of ash particles and the unburnt fraction of the feedstock. TRHA and LabRHA show similar particle morphology with some angular particles and macro-pores on the surface (Fig. 4(d), (e) and (f)). Macro-pores are present on the surface of RHB (Fig. 4(h)), which are inherited from the physical structure of the rice husks and formed by the release of volatiles during the thermal decomposition of rice husks.

18

(b)

(a) Surface protrusions

(c)

(d)

19

(f)

(e)

(h)

(g)

Macro-pores on RHB surface

20

Fig. 4. SEM micrograph of (a) rice husk particles (feedstock) (b) industrial grade RHA (iRHA) (c) Unburnt particles of rice husk in iRHA (d) Treated RHA (TRHA) (e) Angular structure of TRHA after grinding (f) morphology of LabRHA (g) morphology of rice husk biochar (h) Macro-pores on surface of RHB 3.1.3 Particle size distribution and pore characteristics Fig. 5 shows similar particle size distribution of the different types of RHA (iRHA, TRHA and LabRHA) and RHB after grinding; the d50 was 2-3 µm and d90 6-7 µm. These particles were all finer than the cement used in this study. TRHA has a significant increase in specific surface area compared to iRHA (Table 4). The micro-pore surface area of TRHA is 56.91 m2/g, which suggests that 33% of the total BET surface area is contributed by the micro-porous network. Due to the formation of molten potassium layer on the surface during uncontrolled burning, release of volatiles and organic matters is incomplete, which leads to low micro-porosity in iRHA. Further thermal treatment leads to opening up of micro-pores by decomposing the molten layer and unburnt fraction of rice husk, which results in the 11-fold increase in micro-pore volume of TRHA compared to iRHA. RHA produced under laboratory condition has higher surface area compared to iRHA, but the total BET surface area is lower compared to TRHA; the external surface area of LabRHA found from BET analysis is 67.80 m2/g. In the case of RHB, the BET surface area is lower than that of RHA, although the micro-pore volume is similar to that of LabRHA and iRHA. RHB also shows significantly higher porosity (from mercury intrusion results) compared to that of iRHA and TRHA respectively, which is largely contributed by macro- and meso-pores.

21

iRHA Lab RHA RHB

100%

Cumulative Finer (%)

80%

60%

40%

20%

0% 1

10 Grain size (μm)

100

Fig. 5. Particle size distribution of RHA, RHB and cement used in the study Table 4. Summary of pore characteristics and surface area from BET and MIP analysis Samples

BET-N2 specific surface area (m2/g)

Micro-pore volume (cm3/g)

Total porosity Permeability (%) (MIP) (Darcy)

Average pore diameter (nm) (MIP)

iRHA

0.53

0.0168

30.29

17.04

134.40

TRHA

172.86

0.1878

29.03

17.32

175.10

LabRHA

68.86

0.0179

13.00

19.68

112.60

RHB

9.00

0.0164

57.29

2.022

143.50

22

3.2 Compressive strength and specific strength of RHA-mortar and RHB-mortar composites Fig. 6(a) and (b) compares the compressive strength and specific strength of mortar composites with RHA and RHB with that of control mix at 7-day, 28-day and 120-day age. Specific strength is a measure of structural efficiency, quantified as ratio of compressive strength and hardened density of cementitious mixes [42, 43]. Relatively low density for higher (or similar) strength compared to control can potentially reduce self-weight of concrete structures, which can reduce foundation size and construction cost. At 7-day, mortar containing iRHA, LabRHA and iRHA-RHB have lower compressive strength and specific strength than control mortar (p-value < 0.050, Table S1 supplementary material). TRHA20 mix showed similar strength as that of control, and a 24% improvement in strength compared to iRHA20 after 7 days. This shows that TRHA offers better development of early strength of mortar compared to iRHA and LabRHA. Compressive strength is influenced by the hydration of the binder phase, effective water-cement ratio and filler effect of the admixtures added to the cementitious matrix. Isothermal calorimetry results (Fig. 7) shows that filler effect (due to the fine RHA and RHB particles) and densification of mortar (due to the pores in the RHA and RHB absorbing and storing some of the free water in the curing mixes) result in the strength development of these mortar samples being comparable to that of the control, even though the cumulative heat of hydration is lower than control after 7 days. Although TRHA has lower relative SiO2 than LabRHA, TRHA20 has 13% and 7% higher strength than LabRHA20 mortar at 28-day and 120-day mark respectively; due to low average density of TRHA20 (2,165 ± 4 kg/m3) compared to LabRHA20 (2,196 ± 18 kg/m3), the former’s specific strength is also higher. These phenomena can be explained by TRHA containing more 23

carbon compared to LabRHA (table 2). High toughness and tensile strength of carbonaceous micro-fillers were reported to impart toughness and contribute to strength development of cementitious composites [6, 44, 45]. This observation is consistent with the results for iRHA18RHB2 and iRHA12RHB8, where increase in dosage of RHB by 6% and reduction in iRHA content lead to increase in compressive strength by 9.38%, 16.50% and 13% after 7, 28 and 120 days respectively. Furthermore, due to high porosity, RHB absorbs part of mixing water and thus densifying the mix in the surrounding; part of this absorbed water diffuses back into the matrix when more water is needed to sustain secondary hydration (that is, the reservoir effect), thus contributing to strength development [24, 46] reduces strength and this can be attributed to the excessive pores present in the biochar, this phenomenon may not occur in the presence of iRHA; as shown in fig.5, the average RHA particle size is much smaller than that of cement particle’s, and a portion of these RHA particles are almost as large as some macro-pores in the biochar. This increases the likelihood of the hydration products from the pozzolanic reaction in the RHA particles to either occur inside macro-pores or cover the pores surfaces. As a result, this will reduce the negative effect due to increased number of pores when the dosage of biochar is increased. iRHA contains significant amount of unburnt rice husk particles that are sources of hemicellulose [47]. Due to its amorphous nature and sugar content (for example, glucose and fructose), dissolution of hemicellulose takes place in highly alkaline cementitious matrix, which can retard hydration and consequently reduce early strength development [48, 49]. Pozzolanic action plays an important role in influencing the rate of strength gain over the long-term curing period of cementitious materials. Strength gain in control mortar is 12.80% from 7-day to 28-day age, but there was only a marginal increase of 3.70% from 28-day to

24

120-day mark. In comparison, LabRHA20 and TRHA20 show 13.50% and 13.70% increase in strength from 7-day to 28-day mark respectivey; they also have strength gain of 21% and 13.73% respectively from 7-day to 120-day mark. This phenomenon suggests that the contribution to strength by pozzolanic action of amorphous silica in RHA, via the reaction of CH to produce secondary C-S-H gel, is prominent only after around 28 days and this finding is in line with earlier observations [50, 51]. Furthermore, higher strength gain from 28-day to 120-day mark in LabRHA (21%) than that of TRHA (13.73%) can then be attributed to higher fraction of amorphous SiO2 (Table 3), which increases the pozzolanic reactivity of the former.

90

7 days

28 days

(a)

120 days

Compressive strength (Mpa)

80 70 60 50 40 30 20 10 0

Control

iRHA20

iRHA12RHB8 iRHA18RHB2 Lab RHA20

25

TRHA20

45

7 days

28 days

(b)

120 days

Specific strength (MPa.m3. 103/kg )

40 35 30 25 20 15 10 5 0

Control

iRHA20

iRHA12RHB8 iRHA18RHB2

Lab RHA20

TRHA20

Fig.6. (a) Compressive strength and (b) Specific strength of tested mortar composites at 7day, 28-day and 120-day age (a)

4

Control-CP

Heat evolution (mW/g)

iRHA20-CP iRHA12RHB8-CP

3

iRHA18RHB2-CP LabRHA20-CP

2

TRHA20-CP

1

0 0

5

10

15

20

Time (hours)

26

25

30

35

Cumulative heat evolution (J/g)

300

(b)

250 200

Control-CP iRHA20-CP

150

iRHA12RHB8-CP iRHA18RHB2-CP

100

LabRHA20-CP TRHA20-CP

50 0 0

20

40

60

80

100

Time (hours)

120

140

160

Fig.7 (a) Hydration kinetics of cement paste with iRHA, TRHA, LabRHA and RHB; (b) Cumulative heat of hydration in cement paste with RHA and RHB 3.3 Flexural strength Fig. 7 shows the comparison of flexural strength of mortar containing RHA and RHB with that of control at 28-day. Regardless of RHA type, flexural strength is reduced by 16-27% in mortar with RHA compared to control, which is significant at 95% confidence interval (table S1, supplementary material). This indicates that unlike compressive strength, the filler effect and pozzolanic reaction by RHA and RHB have less significant effect on development of flexural strength compared to hydration reaction in plain mortar. Although some studies have reported slight increase in flexural strength of RHA-mortar with 15-20% cement replacement at 28-day age [31, 51], reduction in flexural strength may be expected due to porosity in the RHA and RHB particles that weakens the tensile plane of mortar and facilitates the formation and propagation of cracks.

27

It is interesting to note that TRHA20 mix shows significantly higher flexural strength by 14.80% (p-value = 0.026 < 0.050) compared to iRHA20 mix. The flexural strength of TRHA20 is similar to that of LabRHA20, although the latter has higher hydration (due to stronger pozzolanic reaction) and higher external surface area (67.80 m2/g from BET analysis, section 3.1.3) compared to TRHA (56.91 m2/g from BET analysis). This indicates that the combination of filler effect and internal curing effect (enhanced by the pores) in TRHA offers similar benefit as the stronger pozzolanic effect in LabRHA. The mix of iRHA and RHB (iRHA12RHB8 and iRHA18RHB2) did not show improvement in flexural strength compared to iRHA20 mix (p-value > 0.050). However, it is worth noting that increase in dosage of RHB from 2% to 8% did not change its flexural strength, which is different from the case of compressive strength, where iRHA12RHB8 offered improvement in strength at 28-day and 90-day age compared to iRHA18RHB2. This observation is similar to that of Gupta et al. [7] and Khushnood et al. [44]. 14

Flexural strength (Mpa)

12 10 8 6 4 2 0 Control

iRHA20

iRHA12RHB8 iRHA18RHB2

LabRHA20

TRHA20

Fig. 7. 28-day flexural strength of tested RHA-mortar and RHB-mortar composites 28

3.4 Load-displacement behaviour This section discusses the effect of RHA on load-displacement behaviour of mortar composites under compression. The performance is compared to that of control and mortar with RHB as partial replacement of iRHA (iRHA18RHB2 and iRHA12RHB8). Fig. 8 shows the load-displacement behaviour of the mentioned mortar composites. The peak load, critical displacement and slope pre-failure (up to 40% of peak load, which is generally used to assess stiffness or static elastic modulus of cementitious composites) and post-peak are summarized in table 5. Fig. 9 shows the failure pattern of the mortar samples at 20 seconds after the peak load was reached. It can be observed that iRHA20 shows lowest stiffness among the tested mixes, evident from relatively high critical displacement at peak load and low value of ascending slope (Slope-1). Incorporation of biochar leads to improvement of stiffness. For example, iRHA18RHB2 showed a slight increase in slope-1 value, while a larger improvement (by about 3 times) was observed for iRHA12RHB8. This may be attributed to toughening effect of carbon in the biochar structure and the hydration products from the pozzolanic reaction of RHA either covering or entering the biochar pores, which is reported to increase fracture energy of biochar-mortar and biochar-cement paste by crack branching and crack contouring [45, 52, 53]. Secondly, stiffness of the iRHA20 mix can be adversely affected by the presence of unburnt rice husk particles and low pozzolanic action. Based on XRD results (Fig. 3), RHB is expected to have higher pozzolanic properties compared to iRHA because of the presence of amorphous silica and higher surface area, whereas part of the silica in iRHA is in crystalline form. Therefore, incorporation of RHB would be more effective in converting calcium 29

hydroxide to binder gel (C-S-H), which would result in higher pore-filling and impart stiffness to the composite. Similar trend can be observed for TRHA20 and LabRHA20, which showed higher slope-1 value and lower critical displacement than iRHA20. Improved pozzolanicity in TRHA and LRHA, due to higher silica content and surface area (tables 3 and 4) compared to iRHA, resulted in significantly higher stiffness of mortar. Table 5 shows that the average fracture toughness of iRHA20, LabRHA20 and TRHA20 in the elastic zone are higher than control mortar. In case of mixes TRHA20 and LabRHA20, higher elastic deformation is obtained at comparable peak load compared to control mortar resulting in higher fracture toughness in elastic zone. In the plastic zone, the fracture toughness of mortar with LabRHA and TRHA are lower than control, which is due to the densification of the matrix and the interfacial zones by pozzolanic action. As the interface strength increases to the level of cementitious paste and aggregates, the propagating cracks are less likely to follow tortuous paths around the interfaces [54]. Thus, the densification results in lower tortuosity of cracks during coalescence (just before fracture) and higher brittleness, leading to lower fracture toughness in the plastic zone. Similar trend is reported by Arino and Mobasher [54]- addition of 10 wt.% of ground copper slag increases compressive strength but leads to reduction in area of plastic zone and higher brittleness post-peak due to densification effect of slag particles. In case of iRHA20 mix, substantially lower peak load than control was observed; however, the elastic deformation was substantially higher than control, which is attributed to the voids introduced due to unburnt rice husk particles. These particles also act as fibers absorbing more energy needed for crack widening under compressive stress. However, this mix demonstrate reduction in plastic fracture toughness than control, which is due to low

30

surface area of iRHA (Table 4) and poor bonding between unburnt particles and the matrix. This leads to lower energy absorption during crack propagation and coalescence in the plastic zone. Ductility of composite material can be assessed from the post peak variation of load over displacement. Abrupt reduction in load (higher slope post-peak) indicates reduced ductility (thus increased brittleness) of the composite. iRHA20 is found to have higher ductility, explained by higher slope-2 values compared to other mixes, which is due to presence of unburnt rice husk particles, that act as natural fibre reinforcement. This is visible from the failure pattern of iRHA20 (Fig. 9) where it shows reduced lateral expansion and higher integrity compared to control and TRHA20. In contrast, TRHA20, LabRHA20 and iRHA12RHB8 show higher brittleness at post-peak stage, which is linked to the densification of the mortar micro-structure by pozzolanic action. The fracture mode of cementitious materials densified by precipitation of C-S-H gel tend to be more brittle compared to those containing higher amount of crystalline calcium hydroxide and micro-voids [55]. This finding is also supported by Giaccio et al. [56] and Jauberthie et al. [57], who reported increased brittleness of cement mortar and concrete due to high pozzolanicity and densification by RHA prepared under control conditions.

31

180 Control

iRHA20

140

iRHA18RHB2

iRHA12RHB8

120

Lab RHA20

TRHA20

Load (KN)

160

100 80 60 40 20 0 0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

Displacement (mm)

Fig. 8 Load-displacement curve of mortar composites with RHA and RHB Table 5. Consolidated parameters derived from load – displacement curve (Slope-1 is the slope for the ascending part of the load displacement curve; Slope-2 is the slope of descending branch post-peak used to assess ductility of the composite)

Mix

Average peak load (KN)

Average critical displacement (mm)

Slope-1

Slope-2

Fracture toughness (MPa-mm/mm) E1

E

E2

Control

157.96

0.384

463.110

-252.553

0.172

0.393

0.221

iRHA20

90.522

1.226

72.281

-51.386

0.304

0.431

0.127

iRHA18RHB2

66.14

1.014

83.732

-56.776

0.142

0.300

0.158

iRHA12RHB8

110.15

0.568

234.955

-127.387

0.179

0.259

0.080

LabRHA20

157.30

0.588

210.790

-236.66

0.284

0.403

0.119

TRHA20

169.68

0.438

278.955

-229.083

0.212

0.369

0.157

32

Fig. 9. Failure pattern of RHA and RHB mortar composites at 20 seconds after the peak load is reached 3.5 Drying and sealed shrinkage of mortar with RHA and RHB Sealed shrinkage indicates the autogenous deformation of early age mortar samples due to self-desiccation. Control mortar shows high autogenous shrinkage, which is attributed to pore formation due to drawing of capillary water to continue the hydration process [28]. Replacement of cement by 20% LabRHA leads to mitigation of autogenous shrinkage by 20% (compared to control), because of higher effective water-cement ratio. TRHA20 shows slightly lower shrinkage (by 12%) compared to LabRHA20 after the 46-day monitoring period. Autogenous shrinkage of TRHA20 during the first 7-day monitoring period is similar to that of LabRHA20; after that, a noticeably higher shrinkage for LabRHA20 can be observed. Both TRHA and LabRHA are expected to act as fillers during the initial period of hydration, because pozzolanic reaction is slower than the clinker reaction and starts after 5-7 days once sufficient amount of CH is generated [58]. After this period, LabRHA had higher shrinkage, because of more silica reaction [59] and higher pozzolanicity. iRHA20 shows expansion over the initial 7 days of hydration, which indicates that iRHA eliminated the initial self-desiccation of the cement mortar. However, after 12 days, a steep increase in shrinkage of iRHA20 mix is observed, leading to similar shrinkage as LabRHA at 46day mark. Combination of iRHA and RHB completely compensated autogenous shrinkage, as 33

shown from the net expansion of iRHA18RHB2 and iRHA12RHB8 mixes. iRHA12RHB8 and iRHA18RHB2 has the same amount of internal water as control. However, during hydration, a significant part of the water is drawn into the biochar pores as well; this implies that lesser amount of water is now being drawn into the capillary pores in the cementitious matrix. To sustain ongoing hydration in the matrix, water in the macro-pores of biochar is released back into the matrix (via the reservoir effect) [6, 8]. The stiffer carbon microstructure of the biochar implies that shrinkage of this microstructure does not occur even when water is drawn from it. iRHA18RHB2 and iRHA12RHB8 starts shrinking only after the 8-day mark; it is expected that from the 8-day mark onward, most of the water originally stored in the biochar pores has already been released back into the matrix. As a result, water is now being drawn from the capillary pores either into the curing matrix (to sustain hydration). Overall, these results suggest that combination of iRHA and RHB can significantly reduce autogenous shrinkage, and thus prevent early age shrinkage cracking of mortar.

34

Change in volume due to sealed drying (microstrain)

300

(a)

200 100 0 -100

0

5

10

15

20

25

30

35

40

45

50

-200 -300

iRHA20 Control iRHA12RHB8 iRHA18RHB2 LabRHA20 TRHA20

-400 -500 -600

0

Change in volume due to free drying (microstrain)

-100

-300

5

10

15

Days of hydration

20

25

30

35

40

iRHA20

Control

iRHA18RHB2

iRHA12RHB8

LabRHA20

TRHA20

45

50

(b)

-500

-700

-900

-1100

Days of hydration

Fig.10 (a) Sealed shrinkage and (b) Drying shrinkage of mortar composites with RHA and RHB

35

Free drying shrinkage is caused by the loss of water from the curing matrix. While biochar pores absorb a portion of the free water, thus reducing the amount of water that is lost through bleeding and evaporation (and hence the porosity resulted from this loss) and densifying the matrix around them [7], a portion of the water released back by these pores into the matrix via the reservoir effect will still get evaporated or bled away over a long period of time. This phenomenon is more drastic when the number of biochar particles and pores are more. This is shown in Fig. 10(b) in which the volume reduction of iRHA12RHB8 is significantly more drastic than the others. 3.6 Water permeability test Fig. 11 compares the total capillary absorption by mortar with RHA, RHB and RHA-RHB mixes. The capillary absorption profile of cementitious materials can be distinguished into initial (or primary) and secondary absorption, based on the different rates of water uptake. Primary rate of absorption is attributed to rapid absorption of water by micro-pores and small capillary pores in the cementitious matrix in first 24 hours, while the slower secondary absorption (from day 1 onwards) is due to bigger capillary pores and voids in the matrix and interfacial zones (ITZ) [60]. It can be observed that addition of iRHA (iRHA20 mix) leads to increase in capillary absorption by about 30% compared to control. Absorption (%) of iRHA20 mix is 10%, which is the maximum allowable for durable building materials [61]. Although the initial absorption of control and iRHA20 are similar, there is a drastic increase in secondary absorption of iRHA20, which is possibly due to introduction of air voids and weakening of ITZ by unburnt carbon in the rice husk. It is found that replacement by combination of iRHA and RHB leads to reduction in total absorption compared to control (Fig. 11). iRHA18RHB2 showed reduction in rate of 36

primary absorption by 36% and 27% compared to control and iRHA20 mix respectively (Fig. 12), although significant difference in secondary absorption rate was not observed. This finding is in agreement with our previous reported observations [8, 62] that addition of 2 wt.% biochar (by cement) significantly reduces initial absorptivity due to its filler effect, although secondary sorptivity may not be improved, which is caused by the meso-pores and macropores of biochar that may form channels with the existing capillary pores in the cement matrix, thereby increasing secondary rate of water uptake. TRHA has the highest strength of all samples tested, but this is attributed to its high carbon content. On the other hand, LabRHA20 showed the highest reduction in total capillary absorption (%) among the tested samples: 24% (p = 0.021 <0.050), 36% (p =0.003 < 0.050), and 13% (p =0.054) reduction in absorption compared to control, iRHA20 and TRHA20 respectively. Permeability of cementitious matrix is strongly influenced by its porosity, which is dependent on hydration, water-cement ratio and filler effect of admixtures [14, 63]. Higher degree of hydration results in more pore-filling by solid hydration products. Several studies have reported reduction in porosity of RHA-cement composite by pore-refinement and filler effect of fine RHA particles [10, 31, 63]. LabRHA, TRHA and iRHA have similar mean particle size (Fig. 5), which would induce similar level of filler effect in the mortar composites. The mortar mixes with RHA (iRHA, TRHA and LabRHA) are designed with same water-cement ratio. Therefore, the only difference is the higher amorphous SiO2 content in LabRHA (Table 3 and Fig.3), which induces better hydration through pozzolanic generation of C-S-H gel. Due to high solubility, CH is prone to leaching, which increases macro-porosity of cementitious matrix [64]. Therefore, effective pozzolanic conversion of CH to C-S-H in LabRHA20 leads to lower permeability and higher water tightness of mortar compared to that with iRHA and TRHA.

37

0.80

Average absorption (g/cm2)

0.70 0.60 0.50 0.40 0.30

Control

iRHA20

0.20

iRHA12RHB8

iRHA18RHB2

0.10

LabRHA20

TRHA20

0.00 0

20

40

60

80

Time (hours )

100

120

140

160

Fig. 11. Capillary absorption of water by RHA-mortar and RHB-mortar composites

Primary rate of absorption

Secondary rate of absorption

% Absorption (24 hours)

%Absorption (6 days)

12

0.12

10

0.10

8

0.08 6 0.06

Absorption (%)

Rate of Absorption (mm/√min)

0.14

4

0.04

2

0.02 0.00

0 Control

iRHA20

iRHA12RHB8

iRHA18RHB2

LabRHA20

TRHA20

Fig. 12. Primary and secondary rate of absorption by the different mortar mixes

38

4. Innovation, potential application of this research and future work As mentioned above, substantial amount of RHA generated from open burning or industrial operations are discarded to landfill sites. RHA’s high organic content and impurities renders it unsuitable to be used in structural mortar. This study highlights two innovative strategies to improve properties of iRHA-mortar composites – by thermally treating iRHA (into TRHA) and studying a blend of RHA and RHB. To the best of the authors’ knowledge, this is the first study to demonstrate that further thermal treatment of iRHA could offer improvement in mechanical and durability properties of cementitious composites compared to laboratory grade RHA mortar and control mortar. The findings indicate a promising potential of valorizing RHA produced from uncontrolled burning as partial cement replacing material in concrete, which would lead to higher recycling rate and lower carbon footprint associated with RHA-concrete. Furthermore, this investigation is also the first to highlight the beneficial role of RHB blended with iRHA in reducing autogenous shrinkage and improving water tightness of mortar composites. Thus, the findings from the study present a promising potential of applying TRHA and iRHA-RHB as an alternative mortar admixture to LabRHA that is generally more expensive and energy intensive to manufacture. Concrete constructions in warm and dry climate tend to suffer from shrinkage in early stage resulting in micro-cracking that may impair the durability of the structure. The water retention property of RHB would be beneficial to reduce early shrinkage and promote internal curing, which would improve durability of concrete with blend of iRHA and RHB. Lower evolution of hydration heat in TRHA mixes and iRHA-RHB mixes compared to LabRHA mix would reduce the risk of thermal cracks especially in case of mass concreting under warm climatic conditions. Production of laboratory grade RHA (LabRHA) may need substantial investment in equipment, infrastructure and transportation, which may not be possible in 39

many developing parts of the world. The findings suggest that thermal treatment of locally generated rice husk ash from uncontrolled burning may offer better compressive strength than control mortar and mortar containing LabRHA, which is a more economical alternative to valorize high volume rice husk and reduce cement demand in concrete constructions. Difference in carbon content and pozzolanicity of the different cementitious conglomerated investigated in this research may substantially influence fracture behaviour of mortar composites. Therefore, further research may be directed towards investigating the fracture behaviour of cementitious composites with LabRHA, TRHA, iRHA and RHB respectively. Notched beam with monitoring of crack mouth opening displacement (CMOD) may be used to investigate the roles of various admixtures on fracture toughness and fracture energy of cementitious composites. Furthermore, future research may investigate combination of TRHA and RHB as supplementary admixture in mortar. Water retention by RHB can source for internal moisture in mortar and enhance pozzolanic reaction of TRHA at later stage. 5. Conclusion The findings from this study shows that thermally iRHA leads to higher silica content and specific surface area, which improve pozzolanicity of RHA as mortar admixture. This is found to improve long term compressive and specific strength (strength-density ratio) of mortar compared to control mix of same water-binder ratio. TRHA at 20% cement replacement offers similar early strength development as control (at 7-day mark), while iRHA and LabRHA showed lower strength development. However, due to its higher carbon content, TRHA has the highest strength after 28 and 120 days.

40

Addition of RHB as partial replacement of iRHA can significantly improve compressive strength by about 17% compared to iRHA. However, flexural strength was found to be adversely affected by RHA and RHA regardless of the production process. iRHA has the highest ductility under compressive loading, although its stiffness is significantly lower than control. Therefore, mortar with iRHA can be considered as a controlled low strength material (CLSM) with reduced cement demand. TRHA significantly improve stiffness compared to iRHA, although the failure mode shows similar level of brittleness as control mortar. Combination of iRHA and RHB almost eliminates autogenous shrinkage over the six weeks monitoring period considered in this study; this is due to the porous biochar particles releasing absorbed water into the matrix to sustain hydration, thus reducing the drawing out of water from the capillary pores. Increase in biochar dosage in combination with iRHA reduces autogenous deformation of mortar. Mortar with iRHA, TRHA and LabRHA demonstrate similar drying shrinkage as control, while increasing the dosage of biochar has been found to exacerbate drying shrinkage. RHA produced under controlled laboratory conditions directly from feedstock (LabRHA) improves water tightness through enhanced pozzolanic reaction. Combination of iRHA-RHB significantly reduce primary rate of water uptake, although the secondary absorption is higher compared to LabRHA; the presence of unburnt feedstock in iRHA is expected to contribute to his phenomenon. Overall, this study suggests that application of TRHA, initially from uncontrolled burning operations, and RHB to replace 20% cement can be a sustainable way to deliver cementitious materials with improved durability properties and similar strength to that of 41

control. This approach will recycle and valorise rice husk and the ash produced as by-product of industrial operations, which will also mitigate the environmental hazards due to disposal of rice husk to landfills. References [1] B.S. Thomas, Green concrete partially comprised of rice husk ash as a supplementary cementitious material–A comprehensive review, Renewable and Sustainable Energy Reviews, 82 (2018) 3913-3923. [2] E. Aprianti, P. Shafigh, S. Bahri, J.N. Farahani, Supplementary cementitious materials origin from agricultural wastes–A review, Construction and Building Materials, 74 (2015) 176-187. [3] C. Meyer, The greening of the concrete industry, Cem. Concr. Compos., 31 (2009) 601-605. [4] D.-Y. Oh, T. Noguchi, R. Kitagaki, W.-J. Park, CO2 emission reduction by reuse of building material waste in the Japanese cement industry, Renewable and Sustainable Energy Reviews, 38 (2014) 796810. [5] B.S. Thomas, R.C. Gupta, Properties of high strength concrete containing scrap tire rubber, Journal of Cleaner Production, 113 (2016) 86-92. [6] S. Gupta, H.W. Kua, Carbonaceous micro-filler for cement: effect of particle size and dosage of biochar on fresh and hardened properties of cement mortar, Sci. Total Environ., 662 (2018) 952-962. [7] S. Gupta, H.W. Kua, S. Dai Pang, Biochar-mortar composite: Manufacturing, evaluation of physical properties and economic viability, Construction and Building Materials, 167 (2018) 874-889. [8] S. Gupta, H.W. Kua, H.J. Koh, Application of biochar from food and wood waste as green admixture for cement mortar, Sci. Total Environ., 619 (2018) 419-435. [9] N. Bhanumathidas, P.K. Mehta, Concrete mixtures made with ternary blended cements containing fly ash and rice-husk ash, Special Publication, 199 (2001) 379-392. [10] R. Madandoust, M.M. Ranjbar, H.A. Moghadam, S.Y. Mousavi, Mechanical properties and durability assessment of rice husk ash concrete, Biosystems engineering, 110 (2011) 144-152. [11] R. Zerbino, G. Giaccio, G. Isaia, Concrete incorporating rice-husk ash without processing, Construction and building materials, 25 (2011) 371-378. [12] H.T. Le, H.-M. Ludwig, Effect of rice husk ash and other mineral admixtures on properties of selfcompacting high performance concrete, Materials & Design, 89 (2016) 156-166. [13] K.-B. Park, S.-J. Kwon, X.-Y. Wang, Analysis of the effects of rice husk ash on the hydration of cementitious materials, Construction and Building Materials, 105 (2016) 196-205. [14] N. Van Tuan, G. Ye, K. Van Breugel, O. Copuroglu, Hydration and microstructure of ultra high performance concrete incorporating rice husk ash, Cem. Concr. Res., 41 (2011) 1104-1111. [15] R. Pode, Potential applications of rice husk ash waste from rice husk biomass power plant, Renewable and Sustainable Energy Reviews, 53 (2016) 1468-1485. [16] R. Singh, M. Srivastava, A. Shukla, Environmental sustainability of bioethanol production from rice straw in India: a review, Renewable and Sustainable Energy Reviews, 54 (2016) 202-216. [17] M.N. Al-Khalaf, H.A. Yousif, Use of rice husk ash in concrete, International Journal of Cement Composites and Lightweight Concrete, 6 (1984) 241-248. [18] A. Akhtar, A.K. Sarmah, Novel biochar-concrete composites: Manufacturing, characterization and evaluation of the mechanical properties, Science of the total environment, 616 (2018) 408-416. [19] C. JD, Concrete technologyand design. Cement replacement material, SurreyUniversity Press, London, 1986. [20] G.R. De Sensale, Effect of rice-husk ash on durability of cementitious materials, Cem. Concr. Compos., 32 (2010) 718-725. [21] Q. Yu, K. Sawayama, S. Sugita, M. Shoya, Y. Isojima, The reaction between rice husk ash and Ca (OH) 2 solution and the nature of its product, Cem. Concr. Res., 29 (1999) 37-43. 42

[22] P. Chindaprasirt, S. Rukzon, Strength, porosity and corrosion resistance of ternary blend Portland cement, rice husk ash and fly ash mortar, Construction and Building Materials, 22 (2008) 1601-1606. [23] Z.A. Zeidabadi, S. Bakhtiari, H. Abbaslou, A.R. Ghanizadeh, Synthesis, characterization and evaluation of biochar from agricultural waste biomass for use in building materials, Construction and Building Materials, 181 (2018) 301-308. [24] S. Gupta, H.W. Kua, Effect of water entrainment by pre-soaked biochar particles on strength and permeability of cement mortar, Construction and Building Materials, 159 (2018) 107-125. [25] ASTM, ASTM C33 / C33M - 16e1 : Standard Specification for Concrete Aggregates, in, ASTM International, West Conshohocken, Pennsylvania, United States, 2016. [26] ASTM, ASTM C150 / C150M - 16e1: Standard Specification for Portland Cement, in, ASTM International, West Conshohocken, PA,USA, 2016. [27] R. Krishnarao, J. Subrahmanyam, T.J. Kumar, Studies on the formation of black particles in rice husk silica ash, J. Eur. Ceram. Soc., 21 (2001) 99-104. [28] S. Mindess, J.F. Young, D. Darwin, Concrete, Prentice Hall, Upper Saddle River, NJ,U.S.A, 2003. [29] S. Gupta, H.W. Kua, Carbonaceous micro-filler for cement: Effect of particle size and dosage of biochar on fresh and hardened properties of cement mortar, Sci. Total Environ., 662 (2019) 952-962. [30] P. Chindaprasirt, S. Rukzon, V. Sirivivatnanon, Resistance to chloride penetration of blended Portland cement mortar containing palm oil fuel ash, rice husk ash and fly ash, Construction and Building Materials, 22 (2008) 932-938. [31] M. Jamil, M. Khan, M. Karim, A. Kaish, M. Zain, Physical and chemical contributions of Rice Husk Ash on the properties of mortar, Construction and Building Materials, 128 (2016) 185-198. [32] ASTM, ASTM C1437: Standard Test Method for Flow of Hydraulic Cement Mortar, in, American Society for Testing and Materials International, West Conshohocken, Pennsylvania, United States, 2015. [33] A.-M. Poppe, G. De Schutter, Cement hydration in the presence of high filler contents, Cem. Concr. Res., 35 (2005) 2290-2299. [34] ASTM, ASTM C109 / C109M: Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-in. or [50-mm] Cube Specimens), in, ASTM International, West Conshohocken, Pennsylvania, United States, 2016. [35] N.N. Eldin, A.B. Senouci, Rubber-tire particles as concrete aggregate, J. Mater. Civ. Eng., 5 (1993) 478-496. [36] ASTM, ASTM C348: Standard test method for flexural strength of hydraulic cement mortars, in, American Society for Testing and Materials International, West Conshohocken, Pennsylvania, United States, 2008. [37] ASTM, C1403: Standard Test Method for Rate of Water Absorption of Masonry Mortars, (2015). [38] ASTM, Standard Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete, in, ASTM, 2019. [39] J.H.S. Rego, A.A. Nepomuceno, E.P. Figueiredo, N.P. Hasparyk, Microstructure of cement pastes with residual rice husk ash of low amorphous silica content, Construction and Building Materials, 2015 (2015) 56-68. [40] G.C. Cordeiro, R.D. Toledo Filho, E.d.M.R. Fairbairn, Use of ultrafine rice husk ash with high-carbon content as pozzolan in high performance concrete, Mater. Struct., 42 (2009) 983-992. [41] M. Hamad, I. Khattab, Effect of the combustion process on the structure of rice hull silica, Thermochim. Acta, 48 (1981) 343-349. [42] Y.-W. Choi, D.-J. Moon, J.-S. Chung, S.-K. Cho, Effects of waste PET bottles aggregate on the properties of concrete, Cem. Concr. Res., 35 (2005) 776-781. [43] J.A. Rossignolo, M.V. Agnesini, Mechanical properties of polymer-modified lightweight aggregate concrete, Cem. Concr. Res., 32 (2002) 329-334. [44] R.A. Khushnood, S. Ahmad, L. Restuccia, C. Spoto, P. Jagdale, J.-M. Tulliani, G.A. Ferro, Carbonized nano/microparticles for enhanced mechanical properties and electromagnetic interference shielding of cementitious materials, Frontiers of Structural and Civil Engineering, 10 (2016) 209-213. 43

[45] L. Restuccia, G.A. Ferro, Promising low cost carbon-based materials to improve strength and toughness in cement composites, Construction and Building Materials, 126 (2016) 1034-1043. [46] W.C. Choi, H.D. Yun, J.Y. Lee, Mechanical properties of mortar containing bio-char from pyrolysis, Journal of the Korea institute for structural maintenance and inspection, 16 (2012) 67-74. [47] N. Johar, I. Ahmad, A. Dufresne, Extraction, preparation and characterization of cellulose fibres and nanocrystals from rice husk, Industrial Crops and Products, 37 (2012) 93-99. [48] D. Miller, A. Moslemi, Wood-cement composites: effect of model compounds on hydration characteristics and tensile strength, Wood and fiber science, 23 (2007) 472-482. [49] K. Kochova, K. Schollbach, F. Gauvin, H. Brouwers, Effect of saccharides on the hydration of ordinary Portland cement, Construction and Building Materials, 150 (2017) 268-275. [50] M.Jamil;, M.N.N. Khan;, M.R. Karim;, A.B.M.A. Kaish;, M.F.M. Zain, Physical and chemical contributions of Rice Husk Ash on the properties of mortar, Construction and Building Materials, 128 (2016) 185-198. [51] H.K. Venkatanarayanan, P.R. Rangaraju, Effect of grinding of low-carbon rice husk ash on the microstructure and performance properties of blended cement concrete, Cem. Concr. Compos., 55 (2015) 348-363. [52] L. Restuccia, G.A. Ferro, Nano-particles from food waste: A 'Green'Feature for Traditional Building Materials., in: 9th International Conference on Fracture Mechanics of Concrete and Concrete Structures., Berkeley, United States 2016, pp. 1-6. [53] S. Ahmad, J.M. Tulliani, G.A. Ferro, R.A. Khushnood, L. Restuccia, P. Jagdale, Crack path and fracture surface modifications in cement composites, Frattura ed Integritá Strutturale, (2015). [54] A.M. Arino, B. Mobasher, Effect of ground copper slag on strength and toughness of cementitious mixes, ACI Mater. J., 96 (1999) 68-73. [55] N. Zhang, R. Shahsavari, Balancing strength and toughness of calcium-silicate-hydrate via random nanovoids and particle inclusions: Atomistic modeling and statistical analysis, Journal of the Mechanics and Physics of Solids, 96 (2016) 204-222. [56] G. Giaccio, G.R. de Sensale, R. Zerbino, Failure mechanism of normal and high-strength concrete with rice-husk ash, Cem. Concr. Compos., 29 (2007) 566-574. [57] R. Jauberthie, F. Rendell, S. Tamba, I.K. Cissé, Properties of cement—rice husk mixture, Construction and building Materials, 17 (2003) 239-243. [58] A. Fraay, J. Bijen, Y. De Haan, The reaction of fly ash in concrete a critical examination, Cem. Concr. Res., 19 (1989) 235-246. [59] G.R. de Sensale, A.B. Ribeiro, A. Gonçalves, Effects of RHA on autogenous shrinkage of Portland cement pastes, Cem. Concr. Compos., 30 (2008) 892-897. [60] Z. Yang, W.J. Weiss, J. Olek, Water transport in concrete damaged by tensile loading and freeze– thaw cycling, J. Mater. Civ. Eng., 18 (2006) 424-434. [61] C.B. Cheah, M. Ramli, The implementation of wood waste ash as a partial cement replacement material in the production of structural grade concrete and mortar: An overview, Resources, Conservation and Recycling, 55 (2011) 669-685. [62] S. Gupta, H.W. Kua, C.Y. Low, Use of biochar as carbon sequestering additive in cement mortar, Cem. Concr. Compos., 87 (2018) 110-129. [63] M.-H. Zhang, V.M. Malhotra, High-performance concrete incorporating rice husk ash as a supplementary cementing material, ACI Mater. J., 93 (1996) 629-636. [64] C. Carde, R. Francois, Modelling the loss of strength and porosity increase due to the leaching of cement pastes, Cem. Concr. Compos., 21 (1999) 181-188.

44

Highlights:     

Rice husk is thermally processed into different forms of rice husk ash (RHA) and rice husk biochar (RHB). RHA from uncontrolled burning improves ductility under compressive loading, thus indicating its potential to be a controlled low strength material (CLSM). Addition of RHB as partial replacement of industrial grade rice husk ash significantly improves strength due to internal curing effect. Combination of RHA and RHB eliminates autogenous shrinkage over the 42-day period of study. 20% replacement of cement with RHA and RHB improved durability and strength.

45

Dear sir, madam, Consideration of revised manuscript for review and publication We are writing to submit the revised manuscript for consideration in your esteemed journal. We declare that the content of this manuscript has neither been published nor concurrently been considered by other journal(s). There is also no conflict of interest involving any of the co-authors. Thank you for your attention. Regards, Dr. Harn Wei Kua, Mr. Shravan Muthukhrisnan and Dr. Souradeep Gupta

46