Accepted Manuscript

Rice husk ash and spent diatomaceous earth as a source of silica to fabricate a
geopolymeric binary binder

Johanna M. Mejía, Ruby Mejía de Gutiérrez, Carlos Montes

PII: S0959-6526(16)00092-5
DOI: 10.1016/j.jclepro.2016.01.057
Reference: JCLP 6645

To appear in: Journal of Cleaner Production

Received Date: 3 August 2015

Revised Date: 28 November 2015
Accepted Date: 23 January 2016

Please cite this article as: Mejía JM, de Gutiérrez RM, Montes C, Rice husk ash and spent
diatomaceous earth as a source of silica to fabricate a geopolymeric binary binder, Journal of Cleaner
Production (2016), doi: 10.1016/j.jclepro.2016.01.057.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.
 ACCEPTED MANUSCRIPT

1 4964
2 RICE HUSK ASH AND SPENT DIATOMACEOUS EARTH AS A SOURCE OF SILICA
3 TO FABRICATE A GEOPOLYMERIC BINARY BINDER
4 *Johanna M. Mejía, Ruby Mejía de Gutiérreza, Carlos Montes b.
a
5 Composites Materials Group (CENM), Universidad del Valle, Colombia

PT
b
6 Institute for Micromanufacturing, Louisiana Tech University, United States of America
7 *[email protected]

RI
8
9 Abstract

SC
10 The goal of this research is to propose the use of industrial by-products as solid sources of silica
11 in replacement of commercial sodium silicate (SS) for the manufacture of geopolymeric binary
12 binders. The result of this approach is a binder with good mechanical properties and
13 environmental benefits, which can be used in civil engineering applications such as the

U
14 manufacturing of masonry elements.
AN
15 A Colombian fly ash (FA) and a commercial metakaolin (MK) were used as geopolymer
16 precursors. Rice husk ash (RHA), spent diatomaceous earth (SDE) and commercial sodium
17 silicate were used as sources of silica to adjust the SiO2/Al2O3 molar ratio. These silica sources
18 were mixed with a sodium hydroxide (NaOH) solution. Three alkaline solutions were used to
M

19 activate the binary system FA/MK; SN (sodium silicate and sodium hydroxide), RN (rice husk
20 ash and sodium hydroxide) and DN (spent diatomaceous earth and sodium hydroxide). A FA/MK
21 proportion of 70/30 by weight, and the respective SiO2/Al2O3 and Na2O/SiO2 molar ratios of 4.4
D

22 and 0.2 were kept constant for all experiments. Mechanical and microstructural analysis were
23 conducted on the three systems, and compressive strength was measured at 28, 180 and 360 days.
TE

24 The reaction products were characterized by using X-Ray Diffraction (XRD) and Scanning
25 Electron Microscopy (SEM).
26 Results show that it was feasible to utilize RHA and SDE as a source of silica for
EP

27 geopolymerization. The use of these alternative silica sources delayed the initial and final setting
28 time at 60°C of a FA/MK-based binary alkali activated geopolymer system. Regardless of the
29 source of silica used, the mechanical strength of the resulting geopolymer binders did not show a
30 significant change over time. The FA/MK-SN system exhibited a final compressive strength of
C

31 75 MPa at 360 days, and at the same age FA/MK-DN and FA/MK –RN both reached final
32 compressive strength values of 38 MPa.
AC

33
34 Key words: silica source, metakaolin, fly ash, geopolymer, diatomaceous earth, rice husk ash.
35
36
37
38

1
 ACCEPTED MANUSCRIPT

39
40 1. Introduction
41 The coal combustion process generates waste materials such as fly ash and bottom ash. In
42 Colombia, the main thermoelectric plants and some industrial boilers generate approximately 600
43 kt / y of fly ash, whose quality does not permit its use as an active addition in concrete. This
44 requires disposal in landfills and generate environmental problems due to their mismanagement.

PT
45 The high content of alumina and silica make the fly ash a suitable material for being used as raw
46 material in the production of simple and binary geopolymers (Andini et al. 2008, Chindaprasirt et
47 al. 2009, Mejia et al. 2013, Mejia et al. 2015). The geopolymerization process helps to reduce the

RI
48 environmental impact, given the lower energy consumption, saving space required for the
49 disposal of waste, low processing temperatures, at the same time develop materials with good
50 mechanical and thermal properties with a broad spectrum of applications.

SC
51 According to McLellan et al. (2011) and Yang et al. (2013), geopolymers may have during their
52 life cycle, environmental issues similar to those of Portland cement in terms of CO2 footprint and
53 energy requirements. According to Tempest et al. (2009), the major part of the emissions and
54 energy consumption of geopolymers can be attributed to the sodium silicate and sodium

U
55 hydroxide activators. Sodium silicate, for example, is produced from natural resources (silicon
56 oxide and sodium carbonate) at high temperatures (1300 ° C), so that their production is energy
AN
57 intensive and generates CO2 in the atmosphere. Habert G. et al. (2011) proposed interesting
58 strategies to minimize the adverse effect of sodium silicate in the alkaline activation process. The
59 first of these strategies is to minimize the SiO2/Al2O3 molar ratio in order to reduce the amount of
M

60 sodium silicate added to the system. The second strategy is to use a mixture of ground blast
61 furnace slag and activated clays as raw material, which will also result in a geopolymer with a
62 low SiO2/Al2O3 ratio. As a contribution to improve the ecological properties of geopolymer,
63 Bernal et al. (2015) and (2012), He et al. (2013), ), Mejia et al. (2013), Puertas and Torres-
D

64 Carrasco (2014) proposed to use a supplementary source of silica in alkali activated systems,
65 such as rice husk ash or glass wastes to produce silicate solutions via chemical reaction with
TE

66 NaOH or KOH. Bernal et al. (2011) and Kim et al. (2013), suggested another way to avoid the
67 use of sodium silicate by using CaO and alkali sulfates as a low cost alternative to classic alkali
68 activators.
EP

69
70 According to the mentioned above, this study promote the use of a supplementary sources of
71 silica instead sodium silicate to adjust the SiO2/Al2O3 molar ratio of a geopolymer based on fly
72 ash (FA) and metakaolin (MK). The silica sources namely, spent diatomaceous earth (SDE) and
C

73 a rice husk ash (RHA).

74 Commercial diatomaceous earth (DE) comes from a single-celled algae named diatoms, when
AC

75 diatom algae die the silica shell or skeleton, also known as frustules, collects at the ocean floor,
76 forming fossils. After millions of years, these fossilized cell walls become diatomaceous earth
77 (DE) or diatomite. DE is composed from amorphous nano-silica particles, and consists of 87–
78 91% silicon dioxide (SiO2) and a small amount of alumina (Al2O3) and iron oxide (Fe2O3).
79 Commercial DE because of its characteristics, such as porous structure, high silica content, low
80 density, thermal resistance, is widely used as filter, catalytic support, biological support,
81 functional filler, and adsorbent. (Ibrahim et al. 2012). In this study the DE was used in a brewery
82 company, after it was reused several times, it eventually lost its properties and became a waste
83 named spent diatomaceous earth (SDE).

2
 ACCEPTED MANUSCRIPT

84
85 RHA is an agro-industrial by-product generated during the combustion process of rice husk, this
86 sustainable biomass fuel is used to produce electricity during the rice drying stage in a milling
87 rice, it consists of approximately 40 wt.% cellulose, 30 wt.% lignin, and 20 wt.% silica (He et al.
88 2013).
89
90 Nowadays it has been identified that decreasing the consumption of natural resources and

PT
91 improving the use of wastes and by-products as a raw material for several industrial process will
92 highly contribute to the sustainable development and will benefit the environment by
93 preserving natural resource, air, soil and earth for future generations. Thus the use of wastes and

RI
94 by-products such as RHA, SDE and FA among others as geopolymer feedstocks may help
95 companies to produce high added value materials and low CO2 footprint. On the other hand, is
96 being carrying out a solution for companies concerned to waste managment issues by including a

SC
97 new stage in their production process, recycling and transforming waste and by-products into a
98 new material with a commercial value.
99
100 The performance of silicate solutions prepared from these wastes (RHA and SDE) was compared

U
101 to that of a commercial sodium silicate solution. Three alkaline solutions were produced in the
102 laboratory to alkali activate a binary system based on FA/MK. Setting time, mechanical strength
AN
103 and microstructural development were evaluated. X-ray fluorescence (XRF), powder X-ray
104 diffraction (XRD), and scanning electron microscopy (SEM) were employed as characterization
105 methods.
M

106
107

108
D

109
110 .
TE

111
112
113
114
EP

115
116 2. Experimental Procedure
117
C

118 2.1. Materials

119 A Colombian fly ash (FA) from a thermoelectric power station, and a commercial metakaolin
AC

120 (MK) (Metamax) were used as geopolymeric precursors. The binary system was designed using a
121 FA/MK proportion of 70/30, with SiO2/Al2O3 and Na2O/SiO2 molar ratios of 4.4 and 0.2
122 respectively.
123 A commercial sodium silicate (SS) with 32.4 wt.% SiO2, 13.5 wt.% Na2O, and 54.1 wt.% H2O
124 was used in the study. Rice husk ash was obtained directly from a Colombian rice milling
125 factory, where rice husk was used as a source of energy in a combustor. The total silica content
126 was 90.91% and 36% corresponds to amorphous silica. The spent diatomaceous earth (SDE) was
127 obtained from a Colombian brewing company, where it was used as a filter in one of the beer

3
 ACCEPTED MANUSCRIPT

128 production stages. The total amount of silica in SDE was 91.86 and 43% corresponds to
129 amorphous silica.
130
131 The FA, RHA and SDE were subjected to conditioning treatments. FA, had an initial particle size
132 of 63.9 µm, it was ground in a ball mill during 90 minutes to reduce its particle size to 19.8 µm.
133 The RHA with an initial particle size of 136 µm, first it was ground during 60 minutes in a ball
134 mill, until its particle size was reduced to 28 µm. Then, its particle size was further reduced to 5

PT
135 µm by an attrition mill. The spent diatomaceous earth (SDE) was modified by a calcination
136 process, in order to remove the organic compounds (yeast). Thermal treatment was conducted at
137 400°C for 3 hours.

RI
138
139 Table 1 summarizes the chemical composition, loss of ignition (LOI) and physical properties of
140 the source materials. The chemical analysis of the geopolymer precursors and the two sources of

SC
141 silica was determined by X-ray Florescence (XRF). It is worth noting that the fly ash contains
142 14% of residual carbon, which is usually considered as low quality, since this carbon content is
143 outside ASTM C 618 requirements. However, results from previous studies (Mejia et al. (2015),
144 Mejia et al. 2014, Rivera et al. 2014) have demonstrated that it is feasible the use of this high LOI

U
145 fly ash as a geopolymer precursor.
146
AN
147 Table 1. Chemical analysis of source materials and physical properties
148
M

149 Chemical
compounds FA MK RHA SDE
150 (%)
53.7 52.4 90.93 91.86
D

151 SiO2
Al2O3 21.5 44.3 0.11 3.23
152
TE

Fe2O3 4.5 0.47 0.19 1.56

153 CaO 0.8 0.02 0.36 0.57
154 TiO2 1 1.92 - 0.17
K2O 1.4 0.15 1.97 0.51
EP

155
SO3 0.6 0.02 0.15 0.02
156 MgO 0.6 0.18 0.33 0.29
P2O3 0.5 0.07 0.67 0.45
C

157
Na2O 0.3 0.31 0.02 0.63
158
AC

Sr 0.2 0.01 - -
159 L.O.I 14.8 1.10 5.10 0.68
Physical properties
160
Density
2.21 2.5 2.04 2.06
161 (g/cm3)
Surface área
162 14.8 12.7 11.35 12.99
BET (m2/g)
163 Particle size
19.8 7.76 4.6 24.69
(µm)
164

4
 ACCEPTED MANUSCRIPT

165 Fly ash and metakaolin were used as a geopolymer precursors, since both are aluminosilicate
166 materials with a vitreous phase that allows silica and alumina to be dissolved in alkaline
167 environments. XRD pattern from FA (figure 1) shows more crystalline compounds, such as
168 mullite (2Al2O3∙SiO2; pattern diffraction file (PDF) #00-015-0776), hematite (Fe2O3; PDF# 00-
169 001-1053), and quartz (SiO2; PDF #00-005-0490) in the fly ash compared to the metakaolin. MK
170 has a broad halo between 20 and 35° 2θ, which represents the amorphous phase of the material,
171 and some small crystalline peaks associated to mullite and quartz.

PT
172

RI
U SC
AN
M

173
174 Figure 1. XRD patterns of the raw material. a) Fly ash and b) Metakaolin
175
D

176 2.2. Alkaline solutions preparation and synthesis

177
Three alkaline solutions were used to activate the binary system FA/MK; SN (sodium silicate and
TE

178
179 sodium hydroxide), RN (rice husk ash and sodium hydroxide) and DN (diatomaceous earth and
180 sodium hydroxide). The solutions were prepared by mixing the source of silica with analytical
181 grade sodium hydroxide (99% purity). The liquid/solid ratio was determined depending on the
EP

182 workability of the paste. The ratios were, 0.3 for the paste prepared with SN solution and 0.45
183 for the paste prepared with RN and DN solutions.
184
185 The sodium silicate solution (SN) was mixed with the sodium hydroxide solution, and prepared
C

186 the same day it was used. The RN and DN solutions were prepared by adding and dissolving
187 either rice husk ash (RHA) or spent diatomaceous earth (SDE) in a sodium hydroxide solution.
AC

188 The last two solutions were kept in constant stirring during 24 hours. (Figure 2)
189

5
 ACCEPTED MANUSCRIPT

PT
190

RI
191
192 Figure 2. Alkaline solutions made from alternative silica sources a) RN solution and b) DN
193 solution
194 Geopolymer pastes were prepared by mixing the solid precursors (FA/MK 70/30 by weight) with

SC
195 the alkaline solutions SN or RN or DN for 10 minutes in a Hobart Mixer. The slurry was cast
196 inside cubic molds (2x2x2 cm). The specimens were oven cured for 24 hours at 60°C and 90% of
197 relative humidity. After curing, the samples were kept at ambient temperature until they reached

U
198 the testing age. Table 2 shows the three geopolymer systems with their respective work
199 conditions: alkaline solutions, molar (SiO2/Al2O3 and NaO2/SiO2) and liquid/solid (l/s) ratios.
AN
200 Table 2. Geopolymer systems, alkaline solutions, molar and liquid/solid ratios
System Alkaline SiO2/Al2O3 NaO2/SiO2 l/s
solution
M

FA/MK-SN SS+NaOH 4.4 0.2 0.35

solution
FA/MK-RN RHA+NaOH 4.4 0.2 0.42
D

solution
FA/MK-DN SDE+NaOH 4.4 0.2 0.42
TE

solution
201
202 2.3. Techniques
EP

203 Setting time was determined using a Vicat needle as it is described in ASTM C191. Compressive
204 strength testing was conducted at the ages of 28, 180 and 360 days, using a compression testing
205 instrument (Universal INSTRON 3369), at a displacement speed of 1 mm/min. Density,
C

206 absorption and voids were measured as described in ASTM C642.

207 The microstructural characterization of the binary alkaline activated systems was carried out
AC

208 using instrumental techniques such as Scanning Electron Microscope (SEM), using a JEOL JSM-
209 6490LV microscope under a high vacuum (3x10-6 torr), fitted with an OXFORD
210 INTRUMENTS 7573 INCA PentaFETx3 detector. The samples were covered with gold with a
211 Denton VacuumDesk IV unit in order to increase the resolution of the microphotographs. X-ray
212 diffraction (XRD), was conducted on a PanAnalytical X´Pert MRD diffractometer using a K∝
213 radiation, a copper X-ray tube, in a range 2θ of 5 to 60°.
214
215 3. Experimental results and discussion
216

6
 ACCEPTED MANUSCRIPT

217 3.1. Alternative sources of silica characterization

218 Rice husk ash and spent diatomaceous earth (as shown in Table 1) are mainly composed of silica.
219 The X-ray diffraction (XRD) shows the pattern of RHA and SDE (Figure 3) it exhibits an
220 intense peak of tridymite (American Mineralogist Cristal Structure Database AMCSD, 0000531)
221 and cristobalite (AMCSD, 0001629), both SiO2 polymorphs formed during the combustion
222 process of the RHA. The absence of a halo between 15-30º 2θ, indicates the lack of amorphous

PT
223 phase. On the other hand SDE has a halo between 15-30º 2θ which represents the amorphous
224 phase of the material. Also, a few intense peaks were found, which belong to crystalline phases
225 such as hematite (Fe2O3; PDF# 00-001-1053), quartz (SiO2; PDF #00-005-0490) and cristobalite
226 (AMCSD, 0001629).

RI
U SC
AN
M

227
228
D

229 Figure 3. XRD patterns of Rice husk ash (RHA), and Spent diatomaceous earth (SDE)
230
TE

231 Figure 4 shows the particle morphology of the silica source (RHA and SDE). RHA acquired an irregular
232 shape after the milling process, but the original rice husk ash sample had a honey-comb morphology with
233 a cellular porous surface. On the other hand, different shapes characteristic of the diatomaceous shape can
234 be identified in Figure 2b. Most of the particles look like tubes with small holes on the surface, which is
EP

235 the reason why SDE is used as a filter.

236
C
AC

237
238
239 Figure 4. SEM images- a) rice husk ash and d) spent diatomaceous earth
240 3.2.Geopolymeric binder characterization
7
 ACCEPTED MANUSCRIPT

241
242 3.2.1. Setting time
243 Figure 5 shows the initial and final setting time for each geopolymeric binder cured at 60ºC.
244 Perna I, and Hanzlícek T., (2015), mention that additives and the mixing method partially modify
245 the geopolymer properties and the setting time, thus in the present study it was found that also the
246 reactivity of the silica source and the amount of water in the system also plays an important role.

PT
247 The FA/MK-SN binder, set faster than the other geopolymeric binders made with an alternative
248 source of silica. Its initial and final setting time was 55 and 70 minutes, respectively. The
249 FA/MK-RN and FA/MK-DN systems showed nearly the same initial and final setting times,

RI
250 with a difference of approximately 25 minutes between both materials. The initial and final
251 setting time for FA/MK-RN 74 and 86 minutes, respectively, and for FA/MK-DN, 100 and 110
252 minutes, respectively.

U SC
AN
M
D

253
254 Figure 5. Setting time measures for FA/MK-SN, FA/MK-RN and FA-MK-DN
TE

255
256 The SiO2/Al2O3 molar ratio of 4.4 and the Na2O/SiO2 molar ratio of 0.2 respectively was kept
257 constant for all systems. When silica from sodium silicate was replaced by either RHA or SDE,
EP

258 the amount of NaOH was increased to adjust the Na2O/SiO2 molar ratio to 0.2. Thus, the
259 hydroxyl ion concentration in the solution increased, which helped dissolve the silica from the
260 alternative sources and to dissolve the aluminum and silica from FA and MK. This chemical
261 reaction took longer time than the chemical reaction that took place in the FA/MK-SN system.
C

262 Regarding to water requirements, FA/MK-SN required a lower liquid/solid ratio than FA/MK-
AC

263 RN and FA/MK-DN. Both FA/MK-RN and FA/MK-DN had a liquid/solid ratio of 0.42, which is
264 likely due to the honeycomb-like morphology and to the porous surface of the RHA and SDE,
265 respectively (Detphan and Chindaprasirt et al. 2009, Tanakorn et al. 2013). This kind of
266 morphology reduces workability and increases water demand, which may have contributed to the
267 delay of the setting time of these geopolymer systems, compared with the initial and final time of
268 the binder made with sodium silicate.
269 3.2.2. Mechanical, microstructure and physical analyses.
270

8
 ACCEPTED MANUSCRIPT

271 The results provided in Figure 6 show the compressive strength of the specimens at 7, 28, 180
272 and 360 days. An interesting observation is that the compressive strength did not change
273 significantly between 28 and 360 days for any of the systems. The highest compressive strength
274 at 360 days was 75 MPa, which was achieved by the geopolymer made with commercial sodium
275 silicate (FA/MK-SN). However, an average compressive strength value of ~ 38 MPa was
276 achieved by FA/MK-RN and FA/MK-DN, which suggests them as suitable materials for civil
277 engineering applications.

PT
278 Geopolymer binders made with RHA (FA/MK-RN) and SDE (FA/MK-DN) exhibited the half of
279 the compressive strength compared to the reference system FA/MK-SN. This results suggest that
280 the source of silica and its amorphous phase has an important influence on the mechanical

RI
281 development. Bernal et al. (2015) state that, the mechanism of activation of the precursor is
282 dominated by the different rates of release of silicate from the different source of silica. For
283 instance, Bernal et al. (2012) used RHA with a high amorphous silica content and silica fume

SC
284 (SF) as alternative source of silica, and they reached comparables mechanical strength values to
285 those synthesized from commercial silicate solutions. Previous studies of geopolymers based on
286 granulated blast furnace slag and fly ash using two types of RHA as silica source, with an

U
287 amorphous silica content of 94.7 and 27.7%, showed differences in compressive strength of
288 approximately 10 MPa and up to 35 MPa compared to activated sodium silicate geopolymer
AN
289 (Mejia et al, 2013)
M
D
TE
EP

290
291 Figure 6. Compressive strength of the different alkaline systems FA/MK-SN, FA/MK.RN and
C

292 FA/MK-DN at 7, 28, 180 and 360 days

AC

293 Both RHA and SDE have crystalline phases that are basically silica compounds such as quartz,
294 trydimite and cristobalite which have low reactivity in alkaline solutions. At early and advanced
295 ages (7-180 days), FA/MK-RN shown similar values of about 26 MPa. After 360 days, the
296 material increased its strength by 10 MPa, reaching a final strength of 36 MPa. This behavior was
297 totally different from that of FA/MK-DN, which only had a small increase in strength from 7 to
298 360 days (4 MPa) but its compressive strength at 7 days was 34 MPa. This result at early ages is
299 related with the higher amorphous silica content of SDE. It is assumed that part of the crystalline
300 silica from RHA dissolves over time to participe into geopolymerization (Duan et al. 2015,
301 Rajamma et al. 2012). Trydimite and cristobalite from RHA appear to be more susceptible to
302 being dissolved by an alkaline solution than quartz and cristobalite from SDE, then quartz is a

9
 ACCEPTED MANUSCRIPT

303 non-reactive compound and does not contribute to the chemical reactions of geopolymer (Autef
304 et al. 2012).
305 The XRD spectra in Figure 7 show crystalline compounds in agreement to Autef et al. 2012,
306 which states that all geopolymeric binders display peaks due to crystalline phases coming from
307 the raw material (FA, MK, RHA and SDE), such as mullite (2Al2O3∙SiO2; pattern diffraction file
308 (PDF) #00-015-0776), quartz (SiO2; PDF #00-005-0490), cristobalite (AMCSD, 0001629) and

PT
309 tridymite (AMCSD, 0000531). Both quartz polymorphs (trydimite and cristobalite) appeared in
310 FA/MK-RN and just cristobalite in FA/MK-DN. An arising small hump is observed around 2θ=
311 22º and 2θ=33º for all geopolymeric systems at 28 (figure 7 a) and 360 (figure 7 b) days,
312 indicating the presence of an amorphous phase. It is also worth noting that the binary system

RI
313 synthesized from sodium silicate as silica source presents a sharp increase of that hump, which
314 point out, the higher is the reactive silica in the silica source, the amorphous gel formation is
315 increased.

U SC
AN
M
D
TE

316
317 Figure 7. XRD patterns for geopolymer systems. a) 28 days and b) 360 days
EP

318 Figure 8 shows SEM images of the three geopolymer matrices at 28 days. FA/MK-SN made with
319 commercial sodium silicate has a dense and homogeneous matrix (Figure 8 a), which may have
320 contributed to the high mechanical strength shown in figure 5. As it was explained before, the
321 sodium silicate solution dissolves rapidly, bonding SiO2 and Al2O3 species from FA and MK that
C

322 were previously attacked by OH- ions from NaOH. Porosity is rapidly filled with gel as soon as
323 the liquid phase is able to reach the precursor particles (Al Bakri et al. 2011). Some embedded
AC

324 un-reacted and partially reacted fly ash particles were also found on the surface. It is expected
325 that after 360 days SiO2 and Al2O3 species from the surface continue to dissolve and link into the
326 geopolymer network, improving the compressive strength at advanced ages as was shown in
327 Figure 5.

10
 ACCEPTED MANUSCRIPT

PT
RI
328 Figure 8. SEM images at 28. a) FA/MK-SN, b) FA/MK-RN and c) FA/MK-DN
329 He et al. (2013) and Kusbiantoro et al. (2012) stated that the incomplete dissolution of RHA and
330 SDE results in a lower density of the Si–O–Si bonds in the geopolymer, affecting gel formation,

SC
331 material porosity and, mechanical properties. FA/MK-RN and FA/MK-DN both have a porous
332 and inhomogeneous microstructure, as it is shown in Figures 8b and 8c. The compressive
333 strength of ~36 MPa reached by the mixtures made with alternative silica, compared with 75

U
334 MPa from FA/MK-SN suggests that, there was a lower formation of reaction products.
335
AN
336 To complement the results above, density, absorption and voids were determined for all three
337 systems, and are shown in Table 3. The values obtained are consistent with compressive strength
338 and SEM micrographs. FA/MK-RN and FA/MK-DN exhibited similar microstructure and
339 mechanical strength development. Both of them have lower density than FA/MK-SN, therefore
M

340 greater absorption and voids percentage than FA/MK-SN. Considering the physical properties
341 shown in Table 3, low bulk density and high percentage of voids specifically in those materials
342 made from alternatives source of silica (RHA and SDE) are considered lightweight geopolymers.
D

343 This classification is consistent with other studies where SDE and RHA were used (Detphan and
344 Chindaprasirt et al. 2009, Tanakorn et al. 2013). The values obtained in this work range between
1320 and 1550 kg/m3 and are similar than those reported by Cioffi et al (2003) (1200-1600
TE

345
346 kg/m3) and lower than those found by Andini et al (2008) (1480-1740 kg/m3) for geopolymers
347 based on MK or FA respectively.
348
EP

349 Table 3. Density, absorption and voids

Bulk Density
System Absorption (%) Voids (%)
(kg/m3)
C

FA/MK-SN 1550 15.52 25.26

AC

FA/MK-RN 1370 23.48 33.03

FA/MK-DN 1320 26.77 35.91

350
351 4. Conclusions
352
353 From the results of the research, it can be concluded that rice husk ash and spent diatomaceous
354 earth offer a feasible alternative to conventional sodium silicate as a reactive silica source. The

11
 ACCEPTED MANUSCRIPT

355 change in source of silica, from sodium silicate to rice husk ash and spent diatomaceous earth,
356 influences the setting time, the mechanical behavior and physical properties. The setting time
357 results show a possibility of using rice husk ash and spent diatomaceous earth as an additive to
358 extend the setting time of the geopolymer based on FA/MK for a particular application. The
359 alternative systems FA/MK-RN and FA/MK-DN had a compressive strength at 28 days of up to
360 27 and 36 MPa and at 360 days of 36 and 38 MPa respectively, suggesting that the material can
361 be a cementitious binder comparable with Portland cement. Concerning to the physical

PT
362 properties, the porous particles from the rice husk ash and spent diatomaceous earth helped to
363 decrease the bulk density and to increase the porosity of the geopolymer. These properties
364 suggest the use of the binder as a lightweight material with high mechanical strength (~38 MPa)

RI
365 for the construction industry.
366
367 In terms of linking knowledge to action, challenges to bring this technology to market include a

SC
368 comprehensive study of the variability of rice husk ash and spent diatomaceous earth source
369 materials and its effect in the final properties of the geopolymer pastes, mortars and concrete
370 produced, in order to establish of tolerance limits on their chemical and physical properties to
371 insure a consistent product quality. Another challenge includes the conditioning treatment that is

U
372 necessary to make on the rice husk ash a spent diatomaceous earth before it is utilized as a
AN
373 source material for geopolymers, since feasibility and cost studies are required before large scale
374 production.
375
M

376 5. Acknowledgments
377
378 This study was funded under the Colombian Institute for the Development of Science,
D

379 Technology, and Innovation COLCIENCIAS Project Hybricement (Contract N° 0638-2013).

380 Authors would like to thanks to Louisiana Tech University, in which part of the experimental
TE

381 work was carried out as part of a research internship. Authors also thank GENSA Thermo
382 electrical Company for providing the fly ash, La Esmeralda Colombian milling rice and
383 SABMiller Colombian Brewery for providing the rice husk ash and spent diatomaceous earth,
384 respectively.
EP

385
386
387
388 6. References
C

389 Andini, S., Cioffi, R., Colangelo, F., Grieco, F., Montagnaro, F., Santoro, L., 2008. Coal fly ash
390 as raw material for the manufacture of geopolymer-based products. Waste Manag. 28, 416–42.
AC

391 Al Bakri, A. M., Kamarudin, H., Bnhussain , M., Khairul, I., Mastura, W.I.W., 2011.
392 Mechanism and Chemical Reaction of Fly Ash Geopolymer Cement- A Review. J Asian Sci Res.
393 1 (5), 247-253.
394 Autef, A., Joussein, E., Gasgnier, G., Rossignol, S., 2012. Role of the silica source on the
395 geopolymerization rate. J. of Non-Crystalline Solids. 358, 2886–2893.
396 Bernal, S,A., Rodríguez, E,D., Mejía de Gutiérrez, R., Provis, J,L., 2015. Performance at high
397 temperature of alkali-activated slag pastes produced with silica fume and rice husk ash based
398 activators, Mater Const. 65, 318, e049.

12
 ACCEPTED MANUSCRIPT

399 Bernal, S., Herfort, D., Skibsted, J., 2011. Hybrid binders based on alkali sulfate-activated
400 Portland clinker and metakaolin, 13th International Congress on the Chemistry of Cement. 141-
401 147, Madrid, Spain.
402 Bernal, S,A., Rodríguez, E, D., Mejía de Gutiérrez, R ., Provis, J,L., Delvasto, S., 2012.
403 Activation of Metakaolin/Slag Blends Using Alkaline Solutions Based on Chemically Modified
404 Silica Fume and Rice Husk Ash, Waste Biomass Valoriz. 3, 99–108.
405 Chindaprasirt, P., Jaturapitakkul, C., Chalee, W., Rattanasak, U., 2009. Comparative study on the

PT
406 characteristics of fly ash and bottom ash geopolymers. Waste Manag. 29, 539–543.
407 Cioffi, R., Maffucci, L., Santoro, L., 2003. Optimization of geopolymer synthesis by calcination
408 and polycondensation of a kaolinitic residue. Resour Conserv Recycl. 40, 27–38.

RI
409 Detphan, S., and Chindaprasirt, P., 2009. Preparation of fly ash and rice husk ash geopolymer. Int
410 J Min, Metall and Mater. 16. (6), 720-726.
411 Duan, P., Yan, C., Zhou, W., Luo, W., Shen, C., 2015. An investigation of the microstructure and

SC
412 durability of a fluidized bed fly ash–metakaolin geopolymer after heat and acid exposure. Mater
413 Des. 74, 125–137.
414 Habert, G., d’Espinose de Lacaillerie, J, B., Roussel, N., 2011. An environmental evaluation of
415 geopolymer based concrete production: reviewing current research trends. J Clean Prod. 19,

U
416 1229-1238.
417 He, J., Jie, Y., Zhang, J., Yu, Y., Zhang, G., 2013. Synthesis and characterization of red mud and
AN
418 rice husk ash-based geopolymer composites. Cem Concr Compos. 37, 108–118.
419 Ibrahim, S.,,and Selim, A, Q., 2012. Heat treatment of natural diatomite. Physicochem Probl
420 Miner Process. 48(2), 413−424.
421 Kim, M,S., Jun, Y., Lee, C., Oh, J,E., 2013. Use of CaO as an activator for producing a price-
M

422 competitive non-cement structural binder using ground granulated blast furnace slag. Cem Concr
423 Res. 54, 208–214.
424 Kusbiantoro, A., Nuruddin, M, F., Shafiq, N., Qazi, S, A., 2012. The effect of microwave
D

425 incinerated rice husk ash on the compressive and bond strength of fly ash based geopolymer
426 concrete. Constr Build Mater. 36, 695–703.
TE

427 McLellan, B, C., Williams, R, P., Laya, J., van Riessen, A., Corder ,G, D., 2011. Costs and
428 carbon emissions for geopolymer pastes in comparison to ordinary portland cement., J Clean
429 Prod. 19, 1080-1090.
Mejía, J,M., Mejía de Gutiérrez, R., Puertas, F., 2013, Rice husk ash as a source of silica in
EP

430
431 alkali-activated fly ash and granulated blast furnace slag systems. Mater Constr. 63, 311, 361-
432 375.
433 Mejía, J, M, Rodríguez, E, D., Mejia de Gutierrez, R., 2014. Potential Utilization of a Low
C

434 Quality-Fly Ash as an Aluminosilicate Source in the Production of Geopolymers. Ingeniería y

435 Universidad. 18 (2), 309-327.
AC

436 Mejía, J, M., Rodríguez, E, D., Mejia de Gutierrez, R., Gallego, N., 2015, Preparation and
437 characterization of a hybrid alkaline binder based on a fly ash with no commercial value. J. Clean
438 Prod. 104, 346-352.
439 Perna, I., Hanzlícek, T., 2015, The setting time of a clay-slag geopolymer matrix: the influence of
440 blast-furnace-slag addition and the mixing method. J. Clean Prod.
441 http://dx.doi.org/10.1016/j.jclepro.2015.05.069
442 Puertas, F.,Torres-Carrasco, M., 2014. Use of glass waste as an activator in the preparation of
443 alkali-activated slag. Mechanical strength and paste characterization. Cem Concr Res. 57, 95–
444 104.

13
 ACCEPTED MANUSCRIPT

445 Rajamma, R., Labrincha, J, A., Ferreira V. M., 2012. Alkali activation of biomass fly ash–
446 metakaolin blends. Fuel. 98, 265–271.
447 Rivera, J., Mejia, J., Mejia de Gutierrez, R., Gordillo, M., 2014. Hybrid cement based on the
448 alkali activation of by-products of coal. J. Constr. 13(2), 31 – 39.
449 Tanakorn, P-N., Chindaprasirt, P., Sata, V., Sinsiri, T., 2013. High calcium fly ash geopolymer
450 containing diatomite as additive. Indian J Eng Mater Sci. 20, 310-318.
451 Tempest, B., Sanusi, O., Gergely, J., Ogunro, V., Weggel, D., 2009. Compressive strength and

PT
452 embodied energy optimization of fly ash based geopolymer concrete. World of Coal Ash
453 Conference 2009 in Lexington, KY: USA, 1-17.
454 Yang, K, H., Song, J, K., Song, K, I., 2013. Assessment of CO2 reduction of alkali-activated

RI
455 concrete. J. Clean Prod. 39, 265 -272.
456

U SC
AN
M
D
TE
C EP
AC

14
 ACCEPTED MANUSCRIPT

Highlights

• Rice husk ash and spent diatomaceous earth were used as a source of silica.
• Rice husk ash and spent diatomaceous earth delay the setting time of the geopolymer.
• Similar compressive strength were achieved in the geopolymer by using RHA and SDE.
•

PT
To use rice husk ash and spent diatomaceous earth reduce the geopolymer bulk density.

RI
U SC
AN
M
D
TE
C EP
AC