Introduction

Cementitious materials are widely used in construction industry. Currently, the mainly used cementitious material is ordinary Portland cement (OPC). However, the utilization of OPC imposes an enormous impact on the environment. The manufacturing of OPC not only consumes significant amount of natural materials and energy but also releases substantial quantity of greenhouse gases. To produce 1 ton of OPC, about 1.5 tons of raw materials is needed and 1 ton of CO2 is released to the atmosphere. Worldwide, the cement industry alone is estimated to be responsible for about 7 % of all CO2 generated [1–4]. Another drawback for OPC is that it may not provide the required properties for specific applications, such as rapid development of mechanical strength and high resistance to chemical attack [5].

In the past two decades, a new type of “cement,” called geopolymer or inorganic polymer, has attracted the attention of many research groups. Geopolymer is a type of cementitious material which is generated from solid aluminosilicates in chemical reaction with an alkaline metal solution at ambient or slightly elevated temperature. The raw material used to produce geopolymer can be obtained from natural sources such as kaolin and volcanic ash or from industry by-products such as fly ash, blast furnace slag, and mine tailings. Geopolymer not only provides performance comparable to OPC in many applications, but has many additional advantages, including rapid curing, excellent fire resistance, high acid resistance, good adherence to aggregates, immobilization of toxic and hazardous materials, and significantly reduced energy usage and greenhouse gas emissions. These characteristics have made geopolymer of great research interest as a potential material for sustainable development [6–15].

As OPC, however, geopolymer exhibits brittle behavior with low tensile strength and is sensitive to cracking [16, 17]. These shortcomings not only impose constraints in structural design, but also affect the long-term durability of structures [16, 18]. To overcome the aforementioned disadvantages, different micro- and macro-fibers have been used to reinforce geopolymer cementitious materials. For example, Zhao et al. [16] used plain woven stainless steel mesh to reinforce geopolymer and showed that the failure mode of the steel fiber-reinforced composite could shift from brittle to ductile. Sun and Wu [19] studied the mechanical behavior of PVA fiber-reinforced fly ash geopolymer by investigating its splitting tensile strength, and demonstrated that 1 % of fibers was the optimum fiber content that can significantly improve the ductility of the composite. He et al. [20] investigated the thermal and mechanical properties of carbon fiber-reinforced geopolymer and found that the mechanical properties of the reinforced composite had great improvement when heated at a temperature from 1100 to 1300 °C. Li and Xu [21, 22] investigated the impact mechanical properties of basalt fiber-reinforced geopolymer concrete using a 100-mm-diameter splitting Hopkinson pressure bar system. They revealed that the addition of basalt fiber can significantly improve the deformation and energy absorption properties of geopolymer concrete. However, these currently studied fibers are all produced by a high energy-consuming process and there is serious concern about what to do with these materials at the end of their life cycle [23, 24].

Growing environmental awareness and the need to ensure sustainability of construction materials have led to efforts to look for alternative fibers to reinforce cementitious materials. Recent years have witnessed an increasing interest in natural fibers because they are abundant, reproducible, and environmentally friendly. Compared with the traditional steel, carbon, and glass reinforcing fibers, natural fibers have many advantages such as low density, high specific strength, low cost, renewability, and CO2 neutral life cycle. So far researchers have studied sisal [25, 26], cotton stalk [27], coconut [28], bamboo [29], jute [27, 30], banana [25, 28, 31], and hemp fibers [32, 33], just list a few, to reinforce cementitious materials and very promising results have been obtained. Pacheco-Torgal and Jalali [24, 34] did an excellent review of the related research. It is noted that the research so far has been focused on OPC-based composites. Very little research has been conducted on utilization of natural fibers to reinforce geopolymer. Teixeira-Pinto et al. [35] studied the utilization of jute fiber to reinforce metakaolin-based geopolymer. The results indicate that raw jute fabric, without any chemical treatment, can be used together with the geopolymer to produce a composite with good mechanical and fire resistance properties. Alomayri et al. [36] studied the physical, mechanical, and fracture behavior of fly ash-based geopolymer reinforced with cotton fibers. The results show that the appropriate addition of cotton fibers can improve the mechanical properties of geopolymer composites. Alzeer and MacKenzie [37] investigated metakaolin-based geopolymer reinforced with unidirectional natural protein-based fibers (carpet and Merino wool). The surface of some of the wool fibers was chemically modified to improve their alkali resistance and reinforcing properties. The results indicated that the presence of wool fibers increased the flexural strength of the composites by approximately 40 % compared to the unreinforced matrix. A recent study, also by Alzeer and MacKenzie [38], on metakaolin-based geopolymer reinforced with natural flax fibers showed that the flexural strength of the fiber-reinforced composites increased with higher fiber content, reaching about 70 MPa at 10 vol.% fiber content.

Sweet sorghum has become an important crop for research and development given its potential as a feedstock for large-scale bioethanol production. Sweet sorghum is especially suitable to be grown in arid areas because compared with corn it requires less fertilizer, water, and pesticides and is cheaper to grow [39–42]. After the juice is extracted from sweet sorghum stalks during the process of ethanol production, a large amount of bagasse (residue) is left behind. It is a great challenge to handle the significant amount of bagasse. Since the bagasse contains a large amount of fibers, 37 % acid detergent fiber (ADF) which is insoluble fiber within the plant cell wall and is composed of cellulose and lignin, 56 % neutral detergent fiber (NDF) which is composed of ADF plus insoluble hemicelluloses, and 3.8 % protein based on the measurements by Ottman [39], it has a great potential to be used as a reinforcing fiber of cementitious materials.

This paper studies the feasibility of utilizing sweet sorghum fiber to reinforce geopolymer cementitious material. Specifically, the unit weight of fly ash-based geopolymer specimens containing different contents of sweet sorghum fibers was measured. Unconfined compression, splitting tensile, and flexural tests were conducted to evaluate the effect of inclusion of sweet sorghum fiber on the compressive, tensile, and flexural strength of geopolymer paste. Based on the splitting tensile tests, the post-peak toughness was also evaluated. In addition, scanning electron microscopy (SEM) imaging was performed to study the distribution of sweet sorghum fibers in the geopolymer paste in order to better understand how they reinforce the geopolymer cementitious material.

Experimental study

Materials

The sweet sorghum bagasse was provided by the Campus Agriculture Center (CAC), University of Arizona. The sorghum bagasse was harvested in October 2010. Leaves and husks were stripped from the fresh stalks, then juice was squeezed from the stalks for ethanol production, and bagasse was produced.

Class F fly ash and sodium hydroxide solution were used to produce the geopolymer paste. The paste specimens were used so that the reinforcement effect of sweet sorghum fibers could be better studied with no need to consider the effect of the presence of aggregate. The fly ash was provided by Salt River Materials Group (SRMG) in Phoenix, Arizona. The fly ash contains about 57 % (weight percentage) SiO2, 29 % Al2O3, and 6.1 % CaO and has 71 % particles passing #325 (44 μm) sieve. The specific gravity of the fly ash is 1.97. The sodium hydroxide (NaOH) solution was prepared by dissolving NaOH flakes in deionized water. The NaOH flakes were obtained from Alfa Aesar in Ward Hill, Massachusetts.

Pretreatment of sweet sorghum fiber

A major concern for natural fiber-reinforced cement composites is durability. The alkaline pore water in the cementitious composite could dissolve the lignin and hemicellulose of the natural fiber, and weaken the individual fiber cell [43]. Another problem is the poor interface quality between the fiber and the matrix of the cementitious material. The performance of the composite depends not only on those individual components, but also on their interfacial compatibility [44]. Surface modification of the natural fiber is essential to achieve good performance of the composites. Different methods have been proposed to pretreat the natural fiber used for reinforcement in order to enhance the durability of the composite and improve the interfacial condition. Sealing the matrix pores, impregnation of natural fiber with blocking agents or water repellent agents, and partial replacement of cement with low alkaline binders, like silica fume, fly ash, and metakaolin, provided some promising results [43, 45–48]. The embrittlement tendency of the composites could also be slowed down using the aforementioned methods, although could not be completely avoided. Pulping, either from chemical or mechanical process, is another fiber pretreatment method that can effectively improve the durability of the composite and the adhesion between the fiber and the matrix [25, 49]. To enhance the adhesion between the two components, alkaline solutions are often applied to pretreat the natural fibers. The alkali-treatment gives rise to fiber fibrillation by breaking down fiber bundle into smaller fibers thereby improving the effective surface area contacting with the matrix. It also improves the cohesion between the fiber and the matrix by removing the surface debris and irregularities [44]. Alkali-treatment affects both the fiber strength and the fiber–matrix adhesion in a positive way [50, 51]. Li et al. [32, 33] found that the pretreatment of hemp fibers in NaOH solution at 120 °C for 40 min favored the increase of compressive and flexural strengths as well as flexural toughness of the hemp fiber-reinforced cement composite. Sedan et al. [51] also found that the flexural strength of hemp fiber-reinforced cement composite could be significantly improved by pretreating the hemp fiber with NaOH solution. Gomes et al. [52] reported that alkali-treated natural fiber-reinforced composites showed two to three times increase in fracture strain compared to untreated fiber-reinforced composites, without significant loss in strength. Van de Weyenberg et al. [53] studied the effect of alkali-treatment of flax fiber on the mechanical properties of the composite and claimed that alkalization of flax fibers was a simple and effective method to enhance the adhesion between the fiber and the matrix.

In this research, the received sweet sorghum bagasse (see Fig. 1a) was pretreated using an alkaline solution as follows. The alkali method was selected for pretreating the sweet sorghum fiber based on the compatibility of the alkaline pretreatment solution and the alkaline environment of geopolymer.

Fig. 1
figure 1

Sweet sorghum bagasse: a as received; and b after treatment

Full size image
  1. (1)

    Remove the soft inner portion of the bagasse (referred to as “depithing”).

  2. (2)

    Cut the bagasse into smaller than 5 cm lengths.

  3. (3)

    Dissolve NaOH flakes in water to produce NaOH solution of 2 M concentration.

  4. (4)

    Submerge the bagasse in the NaOH solution for 12 h.

  5. (5)

    Wash the bagasse thoroughly until the slippery texture is removed.

  6. (6)

    Dry the washed bagasse in a 90 °C oven for 24 h.

  7. (7)

    Grind the dried bagasse to pass #20 (840 μm) sieve screen.

Figure 1b shows the bagasse after treatment.

Specimen preparation

First, the fly ash and the treated sweet sorghum fibers at a specified weight percentage were dry-mixed to ensure uniform fiber distribution. The fiber content of 1, 2, and 3 % by weight of fly ash were used to evaluate the effect of fiber on the mechanical properties of geopolymer, and 0 % was also considered as a control. Then the NaOH solution at a concentration of 10 M was slowly added and the mixing was continued for about 10 min to ensure sufficient dissolution of silica and alumina in the alkaline solution. The NaOH solution to fly ash ratio was kept constant at 0.36 for all the specimens, giving the molar ratios of SiO2/Al2O3 = 3.34, Na2O/SiO2 = 0.135, and H2O/Na2O = 11.1. The NaOH solution was prepared by dissolving NaOH flakes in deionized water and stirring for about 10 min. Considering the generated heat, enough time was allowed for the NaOH solution to cool down before it was used. Some additional water was added to the geopolymer paste with 2 and 3 % of fibers to improve the workability. The resulted geopolymer–fiber paste was then placed in cylindrical Plexiglas molds of 35 mm inner diameter and 70 mm length to make unconfined compression and splitting tensile test specimens, and in wood molds of 360 mm × 60 mm × 25 mm to make flexural test specimens. The mold was shaken by a vibrator during the casting to release the trapped air bubbles. Then, the mold was capped and placed in oven for curing at 60 °C. The specimens were demolded after 3 h and placed back in the oven for prolonged curing of 7 days before being tested.

Measurements of unit weight

Before conducting the mechanical (unconfined compression, splitting tensile, and flexural) tests, the unit weight of the geopolymer paste specimens containing different contents of sweet sorghum fibers was determined by weighing and sizing the cured cylindrical specimens.

Unconfined compression tests

Unconfined compression tests were performed on the 7-day-cured cylindrical specimens with an ELE Tri Flex 2 loading machine at a constant loading rate of 0.1 mm/min following ASTM C39. The tests were performed to measure the unconfined compressive strength (UCS) of geopolymer specimens containing different contents of sweet sorghum fibers. For each condition, three specimens were tested and the average of the measured UCS values was used for the analysis. Before conducting the compression test, the end surfaces of the specimens were polished to make sure that they are accurately flat and parallel. In addition, the end surfaces were lubricated with WD-40 to minimize the friction between the specimen and the steel platens.

Splitting tensile tests

The splitting tensile tests were conducted to measure the tensile strength of the geopolymer paste specimens containing different contents of sweet sorghum fibers following ASTM C496. The same specimen size as in the unconfined compression tests was used. Based on the test, the splitting tensile strength was determined as follows:

$$ \sigma_{\text{t}} = \frac{2P}{\pi DL} $$
(1)

where σt is the splitting tensile strength (MPa); P is the maximum load on the specimen (N); D is the diameter of the specimen (mm); and L is the length of the specimen (mm). Again, for each condition, three specimens were tested and the average of the measured values was used.

Using the force–displacement curves obtained from the splitting tensile tests, the effect of sweet sorghum fibers on the toughness of geopolymer was also evaluated. In this paper, the toughness is simply defied as the post-peak toughness or the area under the force–displacement curve beyond the peak (see Fig. 2).

Fig. 2
figure 2

Definition of post-peak toughness

Full size image

Flexural tests

A three-point bending setup with a loading span of 320 mm was used to measure the flexural strength of the sweet sorghum fiber-reinforced geopolymer following ASTM C1609. The flexural strength was calculated using the following equation:

$$ \sigma_{f} = \frac{{3P_{m} L}}{{2bd^{2} }} $$
(2)

where σ f is the flexural strength (MPa); P m is the maximum load (N); L is the loading span of the specimen (mm); b is the width of the specimen (mm); and d is the thickness of the specimen (mm). For each condition, three specimens were tested.

Scanning electron microscopy (SEM) characterization

To better understand the effects of sweet sorghum fibers on the microstructure and mechanical performance of the geopolymer–fiber composites, SEM imaging characterization was also performed. The SEM imaging was performed in SE conventional mode using the FEI INSPEC-S50/Thermo-Fisher Noran 6 microscope. The freshly failed surfaces from the splitting tensile tests, without polishing to keep the fractured surface “un-contaminated,” were used for the SEM imaging. All specimens were coated with gold before SEM imaging and the operating voltage of the microscope was 15 kV.

Results and discussion

Unit weight

The effect of fiber content on the unit weight of sweet sorghum fiber-reinforced geopolymer is shown in Fig. 3. As expected, the unit weight decreased as the fiber content increased, from 15.4 kN/m3 at 0 % fiber content to 14.4 kN/m3 at 3 % fiber content. Since the fiber content was low, the values of unit weight were still within the range of reported values for fly ash-based geopolymer in the literature, 14.5–17.1 [54] and 11.8–15.7 kN/m3 [55].

Fig. 3
figure 3

Unit weight versus fiber content for sweet sorghum-reinforced geopolymer

Full size image

Unconfined compressive strength (UCS)

The effect of fiber content on the UCS of sweet sorghum fiber-reinforced geopolymer is presented in Fig. 4. The UCS slightly decreased with higher content of fibers included. The UCS decreased from 27.7 MPa with no fiber to 25.1, 22.9, and 20.4 MPa at 1, 2, and 3 % of fibers, respectively. The trend is in agreement with the results by other researchers. For example, the research by Al-Oraimi and Serbi [56] showed the decrease of UCS for natural fiber-reinforced concrete. Kriker et al. [57], Li et al. [33], and De Gutiérrez and et al. [58] also reported that the incorporation of natural fiber in cement mortar and concrete caused decrease in UCS. This is in agreement with the general concept that the main function of fiber is not to increase the compressive strength (sometimes it may decrease the compressive strength); but to control the cracking of the reinforced composite by bridging across the cracks and providing post-cracking ductility [59].

Fig. 4
figure 4

Effect of fiber content on the UCS of geopolymer paste

Full size image

Splitting tensile strength

Figure 5 shows the typical splitting tensile test load and displacement curves of geopolymer paste specimens containing different contents of sweet sorghum fibers. The plain geopolymer paste specimens failed suddenly at the peak load, whereas the incorporation of fiber improved the post-peak ductility significantly. The load at the peak increased with the content of sweet sorghum fibers up to 2 % and then decreased.

Fig. 5
figure 5

Splitting tensile test load and displacement curves of geopolymer paste specimens containing different amounts of sweet sorghum fibers

Full size image

Figure 6 shows the tensile strength versus the content of sweet sorghum fibers included in the geopolymer paste. The tensile strength increased about 36 % when 2 % sweet sorghum fibers were utilized. Further increase of the fiber content, however, led to decrease of the tensile strength. Similar trend was also reported by Mansur and Aziz [30] for the jute fiber-reinforced cement paste.

Fig. 6
figure 6

Effect of fiber content on tensile strength of geopolymer paste

Full size image

The enhanced ductility comes from the debonding and pull-out of fibers that bridge across the cracks, which can carry large amount of loads. The fibers transfer loads back to the uncracked parts of the specimen, which permits multiple cracking of the specimen [19]. Figure 7 shows the direct comparison of the failure modes between two specimens during and after the splitting tensile tests, one containing no fibers and the other 1 % sweet sorghum fibers. One can clearly see the brittle failure of the plain specimen and the “ductile” failure of the specimen containing sweet sorghum fibers. The same behavior was also reported for other natural fiber-reinforced cement composites [30]. The pulling out of these fibers during the loading process absorbs energy and thus improves the tensile behavior of geopolymer paste.

Fig. 7
figure 7

Different failure modes of geopolymer paste specimens containing a no fibers; and b 1 % sweet sorghum fibers

Full size image

Figure 8 presents the comparison of the splitting tensile strength (T) versus the UCS of current study with those available in the literature [60–68]. The comparison is limited to fly ash-based geopolymer (paste, mortar, and concrete), with and without fiber reinforcement. The fly ash-based geopolymer has a wide range of UCS values, between approximately 10 and 90 MPa, and splitting tensile strength values, between approximately 0.25 and 7.5 MPa with most of the values ranging from 1 to 5 MPa. It is interesting to note that all the data points for unreinforced geopolymer are on or below the T:UCS = 1:10 line while those for reinforced geopolymer are above the line. This can be more clearly seen in Fig. 9 which shows the T/UCS ratio versus UCS. The fiber reinforcement with sweet sorghum fiber in the current study and metalized plastic waste fiber in [68], effectively increases the T/UCS value.

Fig. 8
figure 8

Comparison of splitting tensile strength versus UCS relationship of current study with those in the literature

Full size image
Fig. 9
figure 9

Comparison of splitting tensile strength over UCS (T/UCS) ratio versus UCS relationship of current study with those in the literature

Full size image

Flexural strength

The effect of sweet sorghum fiber content on the flexural strength of geopolymer is presented in Fig. 10. It can be seen that the flexural strength of the fiber-reinforced geopolymer shows a similar trend to that of the tensile strength, which increases with higher fiber content up to 2 % and then decreases thereafter. The presence of fiber at the optimum content can effectively carry more tensile load and thus delay the growth of microcracks and increase the flexural strength. However, further increase of the fiber content induces poor workability and fiber agglomeration, resulting in increase of air bubbles entrapped in the composite and nonuniform fiber dispersion. These flaws may lead to stress concentrations and degrade the flexural strength. Similar trend was also reported by Alomayri et al. [36] for fly ash-based geopolymer reinforced with cotton fibers.

Fig. 10
figure 10

Effect of fiber content on flexural strength of geopolymer paste

Full size image

Post-peak toughness

Figure 11 shows the effect of sweet sorghum fiber content on the post-peak toughness of geopolymer paste specimens. It can be seen that the post-peak toughness reaches the highest at fiber content of 2 %. Beyond 2 %, the post-peak toughness decreases gradually, but is still higher than that of the plain geopolymer paste. The debonding, sliding, and pull-out of fibers dissipate significant amount of energy that would otherwise be used to propagate the cracks. Hence, the post-peak toughness of the composite increases considerably due to the incorporation of sweet sorghum fibers.

Fig. 11
figure 11

Effect of fiber content on post-peak toughness of geopolymer paste

Full size image

Figure 12 presents the splitting tensile strength, the flexural strength, and the post-peak toughness in the same figure. One can see that they follow essentially the same trend when the fiber content increases and have the same optimum fiber content of 2 %. This is not surprising because they are all related to the improvement of the tensile behavior of the geopolymer reinforced with sweet sorghum fibers.

Fig. 12
figure 12

Splitting tensile strength, flexural strength, and post-peak toughness of geopolymer paste versus fiber content

Full size image

SEM

Figure 13 shows the representative SEM images for sweet sorghum fiber-reinforced geopolymer specimens tested after 7-days curing. It is evidently noticeable that fiber pull-out and fracture (Fig. 13a, b) are the main mechanisms that lead to the enhanced tensile and flexural strength and ductility, which is in accordance with the description by Savastano Jr. et al. [69] for cement–fiber composites. Closer inspection shows that some fibers have good bonding with the matrix (Fig. 13c), but some do not (Fig. 13d). Fibers with good bonding with the surrounding matrix tend to fracture rather than pull-out at the failed surface [70]. Fibers poorly bonded to the matrix contribute little to the improvement of the tensile and flexural strength, ductility and toughness; sometimes they may act as flaw or crack initiation deteriorating the mechanical performance of the composite. The different bonding conditions may be caused by the non-homogeneity of the composite, the fiber–matrix chemistry, or both. Further research should be conducted to ensure that all fibers are in good contact with the geopolymer matrix so that the effectiveness of sweet sorghum fibers in reinforcing the geopolymer is maximized.

Fig. 13
figure 13

SEM images of the failed surface of a splitting tensile test specimen

Full size image

Conclusions

The feasibility of using alkali pretreated sweet sorghum fiber to improve the mechanical behavior of fly ash-based geopolymer was studied. Based on the results, the following conclusions can be drawn.

  1. (a)

    The unit weight of geopolymer paste decreases with higher sweet sorghum fiber content.

  2. (b)

    The inclusion of sweet sorghum fibers in geopolymer paste slightly decreases the UCS.

  3. (c)

    The tensile and flexural strengths both increase with the content of sweet sorghum fibers up to 2 % and then decrease to be lower than that of the plain geopolymer paste.

  4. (d)

    The post-peak toughness increases significantly with the content of sweet sorghum fibers up to 2 % and then slightly decreases but is still much higher than that of the plain geopolymer paste.

  5. (e)

    There is a clear transition from the brittle failure of the plain geopolymer paste specimen to the “ductile” failure of the geopolymer paste specimen containing sweet sorghum fiber.