The Effect of Calcium Formate , Sodium Sulfate , and Cement Clinker on Engineering Properties of Fly Ash-Based Cemented Tailings Backfill

-e influence of admixtures on the engineering properties of fly ash-based cemented tailings backfill (CTB) is a topic of significant practical interest, as it affects the backfilling cost and the environmental effect of mining operation. -is paper presents results of an experimental study on the influence of different activators on the engineering properties of the CTB containing fly ash. CTB samples are mixed with different contents of calcium formate, sodium sulfate, and cement clinker (4%, 8%, and 12% by mass of total binder) and cured in a cubic chamber (at 20°C and RH 90± 5%) for 3, 7, and 28 days. Specimen tests were performed to assess the slump height, setting time, leaching water rate, vertical settlement, and strength development. Furthermore, the XRD analyses were conducted on the hydration products of fly ash-based CTB mixtures. -e results show that activators can cause decrease in the slump height, leaching water rate, and vertical settlement of fly ash-based CTBmixtures. However, inclusion of cement clinker ranging from 8%–12% of total binder can reduce the slump height, setting time, leaching water rate, and vertical settlement to an acceptable range. Addition of calcium formate in the fly ash-based CTB caused negligible change in compressive strength. -e compressive strength improved with higher content of sodium sulfate and cement clinker at the age of 28 days. XRD analyses showed considerable intensity counts of C-S-H gel, calcium hydroxide, and ettringite, resulting from the addition of sodium sulfate and cement clinker. -is study also shows that an understanding of the effect of activators on the engineering properties of fly ash-based CTB is crucial for designing a cost-effective and workable CTB with reduced environmental impact.


Introduction
Mine tailings is the main waste from mineral processing with an annual output of about 300 million tons in China alone [1].
erefore, safe disposal of tailings has been a big challenge in the mining industry, mainly due to its toxicity and potential risk to human health, ambient ecological environment, etc. e tailings storage facilities are the traditional method to disposal of these tailings, such as dams, embankments, and dry stacking on surface land.However, this surface disposal is associated with a risk of dam failure due to the rapid rising of pore pressure or seismic liquefaction.is can lead to unexpected catastrophic damage to people, equipment, and the environment in a very short time.ere was such an accident on September 7th, 2008, in Shanxi Province, China, resulting in the death of 277 persons.Chinese government has issued much stricter requirements and regulation on surface tailings disposal since then [2].Hence, applications for the tailings storage facilities on the surface became stricter and harder to approve.Consequently, compared with surface tailings disposal, backfilling has become significantly important in tailings disposal, for its potentials on improving ground stability conditions and reducing ecological impact.In addition, the cemented backfilling, such as cemented tailings backfill (CTB), can enhance ore recovery and reduce ore dilution.However, cemented backfilling mainly uses ordinary Portland Cement (OPC) as a binder to acquire strength, which significantly increases the cost of backfilling.e binder cost of cemented backfilling accounts for up to 75% of the total backfilling operation costs.Furthermore, the cement industry is one of the major emitters of CO 2 , which gives rise to global warming.Based on a 2013 report, cement clinker production contributes up to 5% of the total amount of CO 2 emission globally [3].To decrease the CO 2 emissions in relation to cement manufacturing and the binder cost associated with the cemented backfilling, utilization of artificial and natural minerals products for the partial replacement of OPC has aroused the attention of researchers around the world [4].
Fly ash is a potential pozzolanic material that can replace cement in views of its pertinent silica and alumina composition as well as its low water demand.e annual production of fly ash is predicted to be about 580 million tonnes in China, and the amount is estimated to increase in the future [5].Fly ash can improve the engineering properties of concrete when used as a substitute for cement.e fly ashbased concrete cured at high temperature has been found to have good mechanical performance in both short-and longterm tests [6].
e structural property of fly ash-based concrete have also been reported to be similar or better than that of OPC concrete when tested for beams [7], reinforced columns [8,9], bonding [10], and fracture behaviors [11].However, the early strength gain of fly ashbased concrete is reportedly lower than traditional concrete [12][13][14].
e utilization of alkali-activated fly ash as substitute for OPC has drawn the attention of researchers in recent years.Katz found out that the activation of fly ash blended in cements not only depends on the pH of the activating ambiance but also affects by the ratio of the latter to the fly ash [15].Another study revealed that the effect of NaOH activator solution concentration on pore solution alkalinity and subsequent alkali-silica reaction in fly ashbased concrete [16].Shafigh et al. reported that the rate of compressive strength acquisition in fly ash-based concrete at 28 days age is significantly more than that without fly ash [3].
e addition of small size pozzolanic materials (e.g., silica fume) in fly ash-based concrete has also been studied.Barbhuiya et al. incorporated hydrated lime and silica fume into fly ash-based concrete to alter the early strength gain.
e results indicated that the incorporation of silica fume and hydrated lime significantly increases the 28-day strength and improves the sorptivity of concrete [17].Li presented an experimental research on the properties of high-volume fly ash concrete containing nanosilica.e significant strength increases of concrete containing nanosilica powder were observed as early as 3 days curing age [18].Shaikh and Supit studied the influence of ultrafine fly ash on compressive strength and durability behaviors of concrete incorporating high-volume class F fly ash as partial replacement of cement.
eir results show that the inclusion of 8 wt.% ultrafine fly ash significantly increased the early age as well as later age strengths of ordinary concrete with 100% cement [19].Nath and Sarker also reported that the curing of fly ash-based concrete blended with a small proportion of OPC can be shortened instead of using elevated heat [5].
To date, the findings reported in the above studies illustrate that the proper alkali activator and appropriate proportion can significantly alter the mechanical properties of fly ash-based concrete.However, these observations mainly focused on the properties of modified concrete containing fly ash, and the results cannot be directly suitable for CTB. e size and nature of the aggregates as well as the water-to-cement ratio (w/c) have great effect on the fresh and hardened properties of concrete and CTB.In comparison to concrete, the CTB is different from concrete not only in views of the particle size used (the content of <20 micron is more than 20% by mass in CTB) but also on account of its high water-to-cement ratio (the w/c of CTB: >4.0, w/c of concrete: 0.3-0.6).To enhance the mechanical behavior of CTB, experimental researches have investigated the improvement the hydraulic behavior of binder used in CTB by blending OPC with fly ash, granulated slag, and silica fume [20,21].Cihangir et al. reported that CPB samples mixed with acidic blast furnace slag were found to acquire remarkably high compressive strength and workability over 360 days curing time [22].Zheng et al. assessed the coupled effect of limestone powder and water-reducing admixture dosages on the mechanical properties of CTB with coarse copper mine tailings [23].Presently, reports about the effect of activators on the engineering properties of CTB containing fly ash are scarce.erefore, there is a need for further investigations on the effects of activators on the engineering properties of CTB containing fly ash.
is study aims to investigate the properties of alkaliactivated fly ash-based cemented tailings backfill through various experimental tests, which assess the slump height, setting time, leaching water rate, settlement, and compressive strength.
e proposed procedure cannot only mitigate the ecological impact of surface disposal of both mine tailings and fly ash but also can reduce the CO 2 emissions related to cement manufacturing.Above all, it will recycle fly ash as a binder for the cleaner production of mineral resource and reduce the binder cost of cemented backfilling.

Materials.
e materials used in this research include mine tailings, fly ash, cement, activators, and water.

Mine Tailings and Fly Ash.
e tailings used for making CTB specimens in this study were from an iron oreprocessing plant in Anhui Province, China.e grain sizes of the tailings were measured by a Microtra S3500 laser particle size analyzer.e particle size analysis distribution of the tailings is shown in Figure 1, while the physical properties of the tailings are given in Table 1.e chemical compositions of the tailings were determined by the testing method of X-ray fluorescence (XRF) spectrometry, and the result is shown in Table 2.
e fly ash used is produced by the power plant of a coal mine in Sichuan Province, China.Its particle size distribution is also shown in Figure 1, and the physical properties in Table 1.e chemical composition of the fly ash is given in Advances in Materials Science and Engineering its alkalinity coe cient is 0.243 (<1), which indicates that it is an acidic y ash.

Cement Type and Mixing
Water. e cement used in all the CTB mixtures is the 42.5 R ordinary Portland cement (OPC).e chemical composition of the OPC is given in Table 2. Ordinary tap water was used to prepare the designed CTB mixtures.

Activators.
ree types of activators-calcium formate, sodium sulfate, and cement clinker-were mixed as additives to improve the engineering properties of the CTB mixtures at di erent curing ages.Tables 1 and 2 show the physical properties and chemical composition of cement clinker, respectively.From Table 1, it is clear that the cement clinker has large quantities of ne particles and t particle size accounting for 50 wt.% in cement clinker is less than 13.69 μm.

Mix Proportions and Specimen Preparation.
e CTB mixtures were proportioned to reveal the e ect of activators on the CTB samples containing y ash.Mix variables included the amount of y ash as a replacement of OPC as well as the types and the amounts of activators.Fly ash was added by 48% of total binder according to the actual utilization in the iron mine.e binder-to-tailing ratio and solid content are set to 1/8 and 72% by mass, respectively.All of these mixtures blended with a constant amount of y ash were activated with three levels of di erent types and amounts of activators.e calcium formate, sodium sulfate, and cement clinker were added as 4%, 8%, and 12% of total binder by mass, respectively.An additional specimen was used as a control mixture without y ash.e detailed mixture proportions of CTB samples are presented in Table 3.
e tailings, binder, y ash, and water with di erent proportions were prepared and thoroughly homogenized for about 12 minutes to produce the desired CTB mixtures.All mixtures have similar slump values between 220 and 250 mm.After mixing, the prepared CTB was poured into plastic cylindrical moulds that were 100 mm in height and 50 mm in diameter.en, the well-prepared specimens were sealed to avoid the water loss and cured in a curing box of constant temperature (20 °C) and humidity (relative  Advances in Materials Science and Engineering humidity of 90 ± 5%).CTB Specimens were separated from the mould after 24 h to ensure that it was not too soft.Finally, all CTB specimens were put into the curing box and cured continuously up to 3, 7, and 28 days.

Testing of Fresh CTB Mixtures.
Slump height is a key fluidity index of fresh CTB mixtures.To determine this, three portions a fresh CTB mixture under uniform stirring were poured into an inverted conical slump cylinder (100 mm top diameter, 200 mm bottom diameter, and 300 mm of the height).All slump tests are carried out in accordance with ASTM C143 standard [24].
en, the cylinder was lifted vertically upward; meanwhile, the fresh CTB mixture slumped under the action of self-weight.e slump height was calculated by the following equation (1): where H s is the slump height of CTB, in mm, H c is the height of the inverted conical slump cylinder ( 30 mm), and H m is the center point height of CTB mixture after slump, in mm.Setting times of the fresh CTB mixtures were determined in accordance with ASTM C191-08 [25].e testing was conducted at a temperature of 21-23 °C.
e pastes were prepared by mixing the tailings, binder, and activators manually in a 500 ml beaker and tested for setting time by using a standard Vicat apparatus.
Bleeding rate is one of the main concerns of CTB technology, because the leached water can separate out the fine particle of cement added in cemented tailings slurry.Good fresh CTB slurry should behave very low leaching water.e leaching water was drained out rapidly with a straw and weighed after the slurry was mixed with binder, water, and tailings.e slurry was weighed every 20 minutes until the weight is constant and showed no further leaching water.
e leaching water rate of slurry was determined according to the following equation ( 2): where m 0 is the initial weight of water prepared for the cemented tailings slurry and m w is the mount of the leached water.
Water leaching from the slurry results in volume shrinkage.
e shrinkage of the slurry induced vertical settlement of the slurry and generation of cracks in the upper layer of the slurry.e vertical settlement was measured in terms of settlement ratio in a column mould by a laser ranger apparatus.e settlement ratio represents the value of total amount of settlement (h s ) to the initial height (h 0 ), which may cause unfilled mined-out area in underground stope, as shown in Figure 2. e settlement ratio (β) is calculated as follows:

Mechanical Tests.
e unconfined compressive strength (UCS) test was performed on all CTB specimens by a loading apparatus in accordance with ASTM C 39-96 [26].
e load was applied at a displacement ration 0.5 mm per second and monitored until failure.All test data were recorded and saved by a computer acquisition system.

Microstructural Tests.
All samples were dried in an electric blast oven (temperature: 50 °C, time: 6 h) before microstructural tests to remove the free water until mass stabilization.A part of the dried sample was taken out and crushed by a mortar.en, the hydration products types that formed in the CTB specimens were investigated by a D/Max 2500 X-ray diffraction (XRD).In addition, small pieces were obtained from the dried samples and sprayed with Au before scanning electron microscopy (SEM) observations.e microstructure morphology characteristics of the CTB specimens were investigated by a Quanta 250 FEG SEM with a solution of 1.0 nm and an accelerating voltage of 30 kV. e XRD and SEM tests were performed on the CTB samples CC4%, CC12%, and control cured for 28 days.

Effect of Activators on Slump Height.
e tested slump heights of all the fly ash-based and non-fly ash CTB mixtures are shown in Figure 3.As illustrated in the graph, the .is fact may be attributed to the higher amount of ne particles in cement clinker and the lling e ect induced by dispersing the voids among tailings particles, resulting in the inhibition of particles movement.However, the slump height of all CTB mixtures were found to be more than 180 mm, which is signi cantly higher than the threshold level that is required to produce the desired uid slurry for back lling in the mining industry [27].

E ect of Activators on Setting Time.
e results of the initial setting and nal setting time are shown in Figure 4.
e setting times of CTB mixtures were observed to be shorter than that of the control which does not contain y ash.e incorporation of 4%, 8%, and 12% calcium formate reduced the initial and nal setting times of CTB mixtures by 1.7, 2.6, and 2.6 hours and 2.0, 3.3, and 3.2 hours, respectively.But the setting time increment for CF4%, CF8%, and CF12% mixtures were negligible in comparison with control mixture.e probable reason is that the calcium formate, being an organic admixture, is nonabsorbent.Slight decreases of the initial and nal setting times were observed for the mixtures with sodium sulfate (e.g., SS4%, SS8%, and SS12%).e initial and nal setting times for SS4%, SS8%, and SS12% mixtures were 25, 23.1, and 20.4 hours and 29.2, 26.2, and 22.0 hours, respectively.e addition of sodium sulfate relatively a ects the setting times by accelerating both initial and nal setting times with the increasing proportion of the content.e reason may be due to the presence of the strong cation in the mixture (i.e., Na + ) that a ects the solubility of a less strong cation (Ca 2+ ) and anions of CTB mixture.However, from Figure 4, the reduction in the setting time of the mixtures with cement clinker are more pronounced than that containing equal amount of other two admixtures, and the e ect of cement clinker is found severe compared to that of control mixture.e initial and nal setting times were reduced, respectively, to 11.0, 9.9, and 6.8 hours and 14.0, 12.1, and 7.9 hours for CC4%, CC8%, and CC12%.A general decreasing trend in both initial and nal setting times with the increase of the clinker can be observed.
e probable cause is that clinker has larger ne particles and requires more water, and hence the setting time is shortened.
e Ca(OH) 2 forms in the presence of water (CaO + H 2 O ⟶ Ca(OH) 2 ), and this accelerates the hydration process.Furthermore, the results establish that the three types of activator admixture blended in CTB mixtures are found to improve the setting time.e magnitude of the setting time of the fresh CTB is in a descending order as follows: calcium formate > sodium sulfate > cement clinker.Cement clinker is more e ective on the decreases of the initial and nal settling time of CTB mixture.us, the required setting time for CTB mixture can be achieved by varying the clinker content of the y ash-based binder.

E ect of Activators on the Bleeding Rate of CPB.
Using equation ( 1), the bleeding rates of di erent fresh CTB mixtures were determined, and the results are plotted against time as shown in Figure 5.It illustrates that the curves follow a similar development pattern with time.e bleeding rate of di erent mixtures sharply increases to a certain value and then gradually remains constant with time.
e mixtures blended with calcium and sodium sulfate have higher bleeding rate than that with cement clinker.e increasing period of bleeding could be prolonged to Advances in Materials Science and Engineering 150 minutes in mixtures with calcium and sodium sulfate, but the value only reached approximate 80 minutes in CC4%, CC8%, and CC12%.It is also found that the relative amplitude of bleeding peak increases with the content of activator, especially for the mixtures with cement clinker.It suggests that the CTB mixtures containing cement clinker have shorter separation phase in hardening transition process and better water-absorption than the other two activators.us, it can be inferred that fresh CTB mixtures with high content of activator may generate low amount water and contribute to the formation of CTB with good uniformity.
As discussed above, the type and content of activators signi cantly a ect the bleeding rate of CTB mixtures.e nal leaching rates of CTB mixtures are summarized and compared as shown in Figure 5(d).e nal bleeding rate mainly ranges from 3.5% to 18.5%.By reference to the concrete standard, a segregation index of maximum 10% is considered as the limit for a concrete mixture to exhibit good resistance to segregation [28].Based on this perspective, the CTB mixture blended with cement clinker by 8%∼12% of total binder shall be considered as a suitable activator material to alter the bleeding characteristic of CTB blended with y ash.

E ect of Activators on Vertical Settlement.
e leaching of water from CTB mixtures results in volume shrinkage, and this usually induces vertical settlement.Figure 6 shows the evolution of vertical settlement ratio of CTB mixtures with di erent activators over time.It illustrates that the settlement ratio of CTB mixtures increases as time elapses.e maximum settlement ratio was observed in the rst 50 to 100 minutes after mixing.e settlement ratio development of the samples prepared with calcium formate and sodium sulfate is found to be in a constant growth after about 100 minutes, while the CTB mixtures with same content of clinker reach to the stable phase much earlier.Variation of the amount of activator also a ects the settlement ratio of the CTB mixtures.Comparing the results shown in Figure 6(d), it can be seen that the magnitude of the nal settlement of the fresh CTB is in a descending order as follows: calcium formate > sodium sulfate > cement clinker.
e CTB mixtures with content of calcium formate and sodium sulfate show a much higher settlement than that made of cement clinker.It can be further observed that the nal settlement generally decreases with the increase of admixtures.For instance, the nal settlement for the CC8% and CC12% mixtures is noted to be 4% and 1.5%, respectively.e generation of the observed di erence in CTB mixtures can be explained by the fact that the cement clinker has high content of calcium oxide and therefore reacts rapidly with water.e generated calcium hydroxide gel can ll the space and voids between the tailings particles, contributing to the compacting of mixtures and causing less shrinkage.e results also show that the bleeding and shrinkage have a similar trend as the time elapsed, comparing results of Figures 5 and 6.

E ect of Activators on UCS.
e e ect of di erent activators on UCS is apparent by comparing control mixture with other mixtures, as shown in Figure 7. e UCS of all CTB specimens, regardless of the type and content of the activators, increases with curing ages, which can be attributed to the improvement of the degree of hydration reaction in the CTB with the extension of curing time.To enable comparisons between the specimens with di erent activators, the UCS of CTB mixture without y ash (control specimen) was chosen as a baseline level for reference in

6
Advances in Materials Science and Engineering Figure 7.At curing age of 3 days, the CTB specimens blended with calcium formate have the lowest UCS.In regard to CTB specimens blended with sodium sulfate, the UCS is significantly a ected by the activator content.e UCS increased by 30% with an addition of 12% sodium sulfate as compared to the reference control specimen value.e main reason is that after hydrolysis of sodium sulfate, the content of SO 4 2− in the CTB mixtures increased obviously.e Ca 2+ in the CTB mixtures was more likely to participate in the hydration reaction to form ettringite (AFt) under the action of SO 4 2− , which facilitates the strength gain of the CTB specimens [29].Meanwhile, the structure of y ash was destroyed by OH − in the CTB mixtures, which improves the activity degree of the y ash and enhances its hydration reaction.e improvements noticed on UCS are minor in CTB specimens with cement clinker.Moreover, for the specimens with calcium formate and cement clinker, the inuence of the mixed activator content is negligible.Similar trend is obtained in comparison to the UCS of di erent specimens at curing age of 7 days, and the UCS values of   8 Advances in Materials Science and Engineering experimentally supported by the results of the XRD analysis, as shown in Figure 8. From the gures, it can be noticed that the intensity of C-S-H, calcium hydroxide (CH), and ettringite (AFt) are comparatively higher in the CTB samples containing 12% sodium sulfate and 12% cement clinker than that in the y ash-free CTB sample. is indicates that more C-S-H gel, AFt, and CH are formed in the samples containing activators. is increased formation of the hydration products has a positive e ect on the strength gain of CTB mixtures with activator since C-S-H gel and AFt are the main bonding phases in hardened cement [30].e formation of the hydration products also is con rmed by SEM observation, as shown in Figure 9.It can be seen from Figures 9(a) and 9(b) that a large amount of hydration products such as AFt and C-S-H gel are formed inside the CTB specimens with sodium sulfate content of 12%, cement clinker content of 12%, and cured for 28 days (CC4%) and ll in the microscopic pores of the CTB specimens, which are bene cial to the CTB specimens gaining higher UCS.Under the same magni cation, the quantity of hydration products such as AFt and C-S-H gel of the CTB specimen blended without y ash (Control) and cured for 28 days is signicantly reduced, crystalline particle sizes are smaller, and the internal microscopic structure are relatively loose, as shown in Figure 9(c).
As mentioned above, the type and content of activators signi cantly a ect the UCS of CTB specimens.erefore, it can be concluded that the e ect of calcium formate on UCS is negligible.Remarkably, the CTB specimens containing sodium sulfate and cement clinker have better strength enhancement than non-y ash mixture.erefore, from this standpoint, the CTB specimens blended with sodium sulfate and cement clinker can be considered as suitable activator materials to alter the UCS of y ash-based CTB.

Further Discussions.
e results on the slump height, setting time, leaching water rate, settlement, and strength development of the y ash-based CTB sample mixed with di erent content and activators type with di erent curing age, which were presented and discussed above, give useful technical information for optimal choice and design of y ash-based CTB and also have practical implications.It is observed that both the initial and nal setting times of the y ash-based CTB mixture with cement clinker are relatively shorter than that of the corresponding CTB samples containing other activators.Although the CTB mixtures with sodium sulfate have also shorter setting time compared with the CTB mixtures without y ash (e.g., control samples), the nal leaching water rate of mixtures with sodium sulfate could be prolonged to 150 minutes.According to the performance criteria of concrete and CTB, the optimal CTB mixtures should have a desired slump height, a limit leaching rate, settlement (i.e., maximum nal segregation index of 10%), suitable setting time (i.e., minimum initial setting time of 60 min) [31], and high compressive strength (i.e., minimum 28 days compressive strength of 1.0 MPa for CTB) [32].erefore, the y ash-based CTB mixtures with cement clinker replacement level from 8% to 12% of total binder can be suggested as optimal mixtures comparable to that of OPC CTB mixture.Fly ash-based binder modi ed with amount of cement clinker can be a suitable material for CTB at ambient curing condition.is is an energy and costsaving alternative as it reduces not only the potential of the ecological impact of surface disposal of both mine tailings and y ash but also the CO 2 emissions related to cement manufacturing.

Conclusions
is paper presents the experimental results of research work that investigate the properties of y ash-based cemented tailings back ll with various activator types and contents.e following conclusions are made based on the results obtained: (1) Presence of di erent activators shortened the initial and nal setting time of y ash-based CTB mixtures compared to that of traditional CTB mixture.e addition of sodium sulfate reduces the setting time proportionately with the increasing of content.However, the reduction in the setting time of the mixtures with cement clinker is more pronounced than that containing equal content of other two admixtures.
(2) Increase of activators content in the y ash-based CTB mixtures reduces the nal leaching water rate.However, by reference to the concrete standard, if a segregation index of maximum 10% is considered as the limit for a concrete mixture to exhibit good resistance to segregation, the CTB mixture blended

Figure 1 :
Figure 1: Grain size distribution of tailings and y ash used.

Figure 3 :Figure 2 :
Figure 3: Slump heights of y ash-based and non-y ash CTB mixtures.

Figure 4 :
Figure 4: Setting time of CTB mixtures with and without di erent activators.

Figure 7 :
Figure 7: UCS of CTB mixtures with di erent activator types and contents.

Table 2 .
It is mainly composed of SiO 2 , Al 2 O 3 , CaO, and Fe 2 O 3 .e activity coefficient of the fly ash used is 0.68, and 2

Table 1 :
Physical properties of tailings, cement clinker, and y ash used.

Table 2 :
Chemical composition of tailings and y ash used.

Table 3 :
e mix proportions of CTB mixtures.