Bio-mediated soil improvement of loose sand with fungus

Aswin Lim, Petra Cahaya Atmaja, Siska Rustiani

Department of Civil Engineering, Universitas Katolik Parahyangan, Bandung, 40141, Indonesia

Keywords:Bio-mediated soil improvement Loose sand Rhizopus oligosporus

A B S T R A C T This article presents an innovative method of bio-mediated soil improvement for increasing the shear strength of loose sand. The improvement is realized by mixing the loose sand with the inoculum of Rhizopus oligosporus,a kind of fungus widely used in food industry for making Indonesian tempeh.The objective of this article is to investigate the performance and mechanism of mixing tempeh inoculum as a binding agent of loose sand particles. The inoculum dosage, water content of loose sand, and curing time were examined for identifying the increment of unconfined compressive strength (qu) of the samples. The results showed that qu of the treated samples increased when the inoculum dosage was elevated.It shows that 5.24%inoculum could yield 68 kPa of qu,and 5%water content and 3 d curing time produced the maximum qu. Moreover, the mechanism of hypha and mycelium in binding the soil particles was clearly observed using a digital microscope and scanning electron microscope.

1. Introduction

Unified soil classification system (USCS) defines sand as a coarse-grained soil with diameter between 0.074 mm and 4.75 mm.Loose sand has a low shear strength and it is susceptible to liquefaction during an earthquake when saturated. One means to mitigate liquefaction is to increase the shear strength of loose sand.Some methods of soil improvements have been proposed, such as soil cement wall(Boulanger et al.,2018),and deep mixing method(Porbaha et al.,1999; Namikawa et al., 2007). This method is very effective but may contaminate groundwater due to potential pollution of the chemical solutions (Fan et al., 2018). Hence,application of soil-cement should be carefully considered.For this,some environmentally friendly methods were developed, such as microbe induced calcite precipitation (MICP) method (Stocks-Fisher et al.,1999; DeJong et al., 2006, 2010) and enzyme induced carbonate precipitation (EICP) method (Yasuhara et al., 2012;Hamdan and Kavazanjian, 2016). These methods introduce the bacteria into soils, which could induce calcite precipitation and cement the soil.They rely upon the hydrolysis of urea as catalyzed by the enzyme urease(Khodadadi et al.,2017).Also,application of these methods to field is of still undergoing research(Van Paassen,2011; Gomez et al., 2015). In addition, another method, so-called microbially induced desaturation and precipitation (MIDP), was also proposed (Rebata-Landa and Santamarina, 2012; He et al.,2013; He and Chu, 2014). It employs dissimilatory reduction of nitrogen (denitrification) to induce both desaturation (through biogas generation) and calcium carbonate precipitation. The challenge of this method is to ensure the uniformity of treatment(Khodadadi et al., 2017). Hoang et al. (2019) introduced bacterial enzyme induced calcite precipitation (BEICP) method to biostabilize both non-plastic sand and low plasticity silty-sand soils.The main difference from the aforementioned EICP method is that the BEICP method used bacterial-derived urease whereas the EICP method used either plant or commercial urease.In addition,it was demonstrated that the BEICP method could yield higher rates of catalytic urease activity,a wide range of applications of non-plastic and low plasticity silty-sand soils, and the ability to retain higher levels of soil permeability.

Nowadays, MICP method has been combined with introducing fiber for the ductile improvement of bio-cemented soils. Polyvinyl alcohol(PVA) fiber, polypropylene (PP) fiber, and basalt-fiber have been applied for this purpose(Choi et al.,2016;Li et al.,2016;Xiao et al.,2019).According to Xiao et al.(2019),basalt-fiber-reinforced bio-cemented sand can produce a similar increment of unconfined compressive strength and splitting tensile strength compared with PVA fiber-reinforced bio-cemented sand. In addition, the PP fiberreinforced bio-cemented sand yielded lower splitting tensile strength compared with basalt-fiber-reinforced bio-cemented sand.

Moreover, application of biopolymers to improve the shear strength of loose sand particles has been evaluated by Khatami and O'Kelly (2013). They used agar and six modified starches with a range of concentrations to increase the cohesion intercept and stiffness of the treated sand. The cohesion intercept could reach around 240 kPa, but the effective internal friction angle (φ′) dropped from 32.3°(untreated sample) to 17.6°. Smitha and Sachan(2016) also demonstrated the success of using agar biopolymer to strengthen the loose saturated Sabarmati sand by making use of the gel-forming capability of the agar biopolymer. The cohesion value could increase to around 86 kPa,and the internal friction angle also tends to increase from 27°(untreated sample)to around 30°.Chang and Cho (2012) concluded that β-1,3/1,6-glucan biopolymer could be used to strengthening Korean residual soil (clayey soil). The results showed that the compressive strength of the treated sample increased four times that of the untreated sample. Furthermore,Chang et al. (2015) used thermo-gelation biopolymers to strengthen not only clayey soil but also sandy soil. From those studies, it suggests that application of biopolymers has great potential for future development.

This paper proposes an innovative method for increasing the shear strength of loose sand using fungus.Indeed,fungus has been widely used for biodegradation, especially the white-root fungus.The main mechanism of biodegradation applied by the white-root fungus is the lignin degradation system of enzymes (Leonowicz et al., 1999). The idea of this research came from the process of making Indonesian tempeh.The soybeans are bounded each other by the fungus, i.e. Rhizopus oligosporus. Rhizopus oligosporus grows effectively at 30°C-40°C.In the process of making tempeh,Rhizopus oligosporus is used for its enzymes in the fermentation process such as amylolytic, lipolytic and proteolytic. The soybeans are bound together by the hyphae in less than 4 d due to their aggressive growth(Ashoka et al.,2002).Fungus requires substrates such as cellulose or another carbon source as a source of energy(Adenipekun and Rasheedah, 2012).

In addition, hypha is tubular and branching fungal structures that can be seen in plain view when it has formed a collection of hypha or called mycelium. Rhizopus oligosporus has longer hypha than other Rhizopus.The branching mycelium is composed of three types of hyphae: stolons, rhizoids, and sporangiophores that are usually unbranching.

It was also expected that the hypha and mycelium of Rhizopus oligosporus could grow and bind the loose sand particles. The objective of this article is to investigate the performance and mechanism of mixing inoculum of Rhizopus oligosporus as a binding agent of loose sand particles. The inoculum dosage, the water content of loose sand,and the curing time were examined for identifying the increment of the unconfined compressive strength(qu)of the samples.Finally,the mechanism of hypha and mycelium in binding the soil particles was understood by making use of a digital microscope and a scanning electron microscope (SEM).

2. Methods

Padang loose sand was selected as the soil sample, considering many liquefactions occurring during the 2009 Padang earthquake where predominantly loose sands are found (Tohari et al., 2011;Hakam and Suhelmidawati, 2013). Table 1 lists the index properties of Padang loose sand and Fig.1 shows the gradation curve of Padang loose sand based on sieve analysis(ASTM D6913/D6913M-17, 2017). According to USCS, Padang loose sand is categorized as poorly graded sand (symbolized as SP). The percentage of sand particles was 98.55%.X-ray fluorescence(XRF)was also conducted to determine the elemental composition of the soil sample. The results of XRF are summarized in Table 2. As listed in Table 2, themajor composition was silicon dioxide (SiO2), a constituent of quartz mineral.It should be noted that quartz is the most abundant mineral which could be found in sand (Mitchell and Soga, 2005).

Table 1 Index properties of Padang loose sand.

Fig.1. Gradation curve of Padang loose sand.

In addition,the fungus starter,so-called tempeh inoculum,was obtained from the commercial market.The composition of tempeh inoculum is rice flour mixed with the spores of Rhizopus oligosporus. The brand of tempeh inoculum used in the experiments is Raprima made in Indonesia.

2.1. Sample preparation

The samples were prepared by mixing Padang loose sand with tempeh inoculum manually. The amount of Padang sand wasmaintained at about 117 g, and the dry unit weight of the sample(γd) was around 13.63 kN/m3with relative density (Dr) of about 40%. Then, the soil mixture was sprayed with distilled water until the soil was mixed homogeneously. Finally, the mixed soil was poured into the cylindrical mold,following the pluviation method.The opening of the funnel was 10 mm and the mass flow equalled 11.5 g/s. The dimensions of the mold are 76 mm in height and 38 mm in diameter, following the standard sample sizes of unconfined compression test (ASTM D2166/D2166M-16, 2016).Furthermore, the samples were stored at room temperature(24.6°C-29.3°C). The amounts of distilled water, tempeh inoculum, and the curing time depend on the experimental series and will be discussed in next section. Prior to test, the samples were extruded with a laboratory sample extruder.

Table 2 The results of XRF analysis.

Table 3 List of experimental programs.

2.2. Experimental program

Fig. 2. Comparison of reconstituted loose sand sample with and without tempeh inoculum.

Table 3 summarizes the experimental program. Four series of experiments were conducted in this study with variations in the amounts of tempeh inoculum, distilled water, and curing time. In the first series of experiments, samples were prepared with 5%water content, 3 d curing time, and variation of tempeh inoculum dosage. The objective of Series 1 was to investigate the effect of tempeh inoculum dosage on the increment of unconfined compressive strength(qu)of the mixed soil.The purpose of Series 2 is to investigate the effect of water content on the unconfined compressive strength of treated sand. For Series 2, samples were prepared with constant tempeh inoculum dosage and curing time,i.e.3.93%and 3 d,respectively,and varying water content.Series 3 was conducted by varying the curing time of samples from 1 d to 10 d to examine the performance and mechanism of fungus growth.Series 4 was performed to examine the soaking effect on the unconfined compressive strength of treated samples. It should be noted that the unconfined compressive strength was tested for all of the samples by following ASTM D2166/D2166M-16(2016).Fig.2 shows the comparison of reconstituted loose sand sample with tempeh inoculum(treated sample) and the untreated sample. It is obvious that the treated sample could stand in cylinder shape but the untreated sample only formed a sand dune.

3. Results and discussion

3.1. Effect of tempeh inoculum dosage

Fig. 3. Relationship between unconfined compressive strength and tempeh inoculum dosage.

Fig. 4. Stress-strain curves of treated and duplicate samples.

The result of Series 1 is depicted in Fig.3.It was observed that qucontinuously increased with increase of the tempeh inoculum dosage until 5.24%.At the optimum dosage of tempeh inoculum,i.e.5.24%,qucould reach 68 kPa, which is equivalent to quof medium clay soil(Terzaghi and Peck, 1996). Furthermore, qusignificantly decreased when the tempeh inoculum dosage exceeded 5.24%.It indicates that too much tempeh inoculum is unfavorable for increasing quunder the current mixing ingredient.Fig.4 shows the typical stress-strain curve of the treated sand with the Young's modulus of about 1000 kPa.For checking the consistency of mixing the soil sample, two more experiments were conducted with similar water content (5%), the tempeh inoculum dosage(5.24%),curing time(3 d),and preparation technique (as mentioned in Section 2.1). The stress-strain curves of the duplicate samples are also plotted in Fig.4 and yield close results with the initial sample. Hence, it could be conclusive that sample preparation is consistent.

3.2. Effect of water content

Fig. 5. Relationship between unconfined compressive strength and water content.

Fig. 6. Relationship between unconfined compressive strength and curing time.

The result of Series 2 is depicted in Fig.5.It is obvious that quis nil when the sample is dry(zero water content).It indicates that in the absence of water, fungus could not grow. Moreover, for the water content of 30%, quis also nil probably due to the saturation state of the sample. The peak quvalue was observed when the water content was 5%. This result shows that the water content affected quof the mixed soil. It is worth mentioning that during preparation of the tempeh, too much moisture during incubation can circumvent the growing of fungus, because it might breed the wrong types of bacteria. Indeed, fungus and bacteria worked together in tempeh production(Kovac and Raspor,1996).Hence,it seems that the same phenomena also occurred when Padang loose sand was mixed with the tempeh inoculum. Further investigation should be made to clarify this hypothesis.

3.3. Effect of curing time

In this testing series, ten soil samples were tested with 3.93%tempeh inoculum and 5%water content.The curing time varied from 1 d to 10 d.The relationship between quand curing time is plotted in Fig.6.It isobserved thatqureacheda peakatthe curingtime of 3 d and then it dropped significantly.From visual inspection,the color of the fungus changed from white into yellowish when the curing time increased.When the fungus color turned to be yellowish,it indicates that the fungus died. In order to understand this phenomenon, the soil samples and fungus growth were observed with a digital microscope and an SEM.

Fig. 7. Relationship between unconfined compressive strength and soaking time.

Fig. 8. Evolution of fungus growth: (a) day 1, (b) day 2, and (c) day 3.

3.4. Effect of soaking time

Fig. 7 depicts the effect of sample soaking on the unconfined compressive strength of the treated sample(Series 4).The tempeh inoculum dosage, water content and curing time before soaking were 5.24%, 5% and 3 d, respectively. The purpose of 3 d of dry curing is to make sure that fungus grow well before soaking. As discussed in Section 3.2, too much water at sample preparation stage could make fungus do not grow.As shown in Fig.7,it is clear that qudecreased by about 50% when the sample was soaked for 3 d, compared with the sample which was soaked for 1 d. In addition,qutended to be constant after 3 d of soaking and quafter 7 d of soaking yielded a close value.

4. Mechanism of soil improvement by fungus

Fig. 9. Evolution of fungus loss: (a) day 4, (b) day 7, and (c) day 10.

Fig.10. SEM images of sand particles bound by mycelium: (a) 110 times of magnification, and (b) 600 times of magnification.

Observation using a digital microscope was carried out in this experiment to investigate the mechanism of unconfined strength development. The photos were captured from day 1 until day 10 to check the growth and the loss of the fungus.Fig.8 depicts the evolution of fungus growth. In day 1, the hyphae started growing from tempeh inoculum and spread in the surface of the soil.During day 2 and day 3,the mycelium was formed from the colony of fungus and it covered most of the soil surface.From this observation and the data as plotted in Fig. 6, it seems that the increment of quwas obtained because of the soil particles bound by the mycelium. The mycelium functioned like a “rope” that binds all of the soil particles together.Furthermore,Fig.9showstheevolutionof thefungusloss.Asshownin Fig. 9a (day 4), the color of mycelium changed from white into yellowish and some spots of dead fungus were observed.When the curing time increased,the spots of dead fungus increased and yielded the loss of unconfnied compressive strength significantly.Hence,it is apparent that the mechanismof the incrementof quwas caused by the mycelium which binds sand particles.

Fig. 11. Failure envelope of untreated and treated Padang sand samples with 3.93%tempeh inoculum dosage and 3 d curing.

Furthermore,Fig.10 provides the images obtained from an SEM.The purpose of this imaging is to see the microstructure of sand particle-fungus interaction. The sample used in this observation was prepared with 3.93% tempeh inoculum dosage, 5% water content,and 3 d curing time.From the images,it is clearly seen that mycelium covered the entire surface of the sand particles. In addition,the sand particles were also bound by the mycelium.This finding was a proof of the statement earlier where the mechanism of the increment of quwas caused by the mycelium which binds sand particles. When the mycelium started to die, i.e. the color turned to yellowish color(see Fig.9),quof treated soil was dropped significantly. Hence,it was clear that the increase of quwas due to the mycelium bonding among the particles.

Based on the observed mechanism,it seems that the mycelium would contribute to increase of soil cohesion (c) to a certain extent due to its “rope” effect, but the internal friction angle (φ)might remain unchanged. In order to investigate this hypothesis,two sets of the direct shear tests were performed, which consist of the untreated and treated sands. Fig. 11 depicts the failure envelope of untreated and treated Padang sand samples with 3.93% tempeh inoculum dosage, 5% water content and 3 d curing.For the untreated sand, the internal friction angle and cohesion were 27° and 0, respectively. It should be noted that the sample was also prepared following Section 2.1 where the relative density of the soil sample was maintained at about 40%.The obtained internal friction angle and cohesion are reasonable for the sand in a loose state.Furthermore,the φ and c values for the treated sand were 25° and 24 kPa, respectively. The φ value decreased by 2°and the c value increased to 24 kPa. From this result, it could be concluded that the mycelium contributed to increasing of the soil cohesion (c) to a certain extent due to its “rope” effect. As a consequence, the internal friction angle would decrease slightly.This phenomenon was also reported by Smitha and Sachan(2016)where they used agar biopolymer to improve the shear strength of Sabarmati sand. The shear stress and horizontal displacement relationships of untreated and treated sand samples at different normal stresses are shown in Fig.12. The failure point (defined as peak shear stress) was observed at smaller horizontal displacement levels for lower applied normal stress. The shear stress and horizontal displacement relationship of treated sand sample indicates a brittle response, especially at high normal stress(σ = 200 kPa). Also, the Young's modulus of treated sand sample is higher than that of untreated sample.

Fig.12. Typical shear stress versus horizontal displacement relationships of untreated and treated Padang sand samples with 3.93%tempeh inoculum dosage and 3 d curing:(a)s=50 kPa, (b) s = 100 kPa, and (c) s = 200 kPa.

5. Discussions

At first,it was assumed that fungus has the potential to be used for liquefaction problem because they could increase the shear strength of loose sand,hence Padang loose sand was directly used for this purpose. However, according to experimental results, it seems that growing fungus on loose sand is inappropriate for solving the liquefaction problem because their lifetimes are short.This challenge needs to be tackled in further study,for example,to examine the necessity to supply nutrition for fungus to extend their lifetimes. In addition, variation of water content with different tempeh inoculum dosages is also worth studying to further explore the performance of this method. Another possible application based on this study is applying fungus as a surficial treatment for wind-caused erosion. Even though application of fungus in geotechnical engineering is not mature and needs investigation, it is promising as an alternative option for soil improvement.

6. Conclusions

In this study,a laboratory test of bio-mediated soil improvement of loose sand using fungus was performed to assess the performance and mechanism of mixing tempeh inoculum as a binding agent of loose sand particles. Several conclusions may be drawn:

(1) The fungus could grow in loose sand with water content in the range of 3%-30%, where 5% water content yielded the maximum unconfined compressive strength with the tempeh inoculum dosage of 5.24%.

(2) The optimum tempeh inoculum dosage was 5.24%and could yield 68 kPa of unconfined compressive strength. The quvalue significantly decreased when the tempeh inoculum dosage exceeded 5.24%, indicating that too much tempeh inoculum dosage was not helpful for increasing quvalue under the current mixing ingredient.

(3) The increase of quin the soil sample was due to the mycelium bonding among the particles.Based on the direct shear test,the mycelium contributes to increasing the soil cohesion to a certain extent due to its“rope”effect.As a consequence,the internal friction angle would decrease slightly.

(4) Because mycelium is an organism,the lifetime is limited.The quvalue of the treated sample would decrease after 3 d of curing. With an observation using a digital microscope, it was clearly seen that the mycelium changed color from whiteto yellowish,which indicates that the mycelium died.When the mycelium died, the quvalue would drop significantly.

Declaration of Competing Interest

We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.

Acknowledgments

The authors acknowledge the research funding provided by Ikatan Alumni Teknik Sipil (IATS) Unpar and Universitas Katolik Parahyangan. Also, the authors would like to thank the reviewers who provide invaluable comments and suggestions for improving the content and quality of this article.