Introduction
Biorefinery is the process of integrating renewable biological resources with biotechnologies to obtain a wide range of valuable products, including biofuels, biochemicals and biomaterials (Clark et al., Reference Clark, Budarin, Deswarte, Hardy, Kerton, Hunt, Luque, Macquarrie, Milkowski, Rodriguez, Samuel, Tavener, White and Wilson2006). Fungal biofabrication is a type of biorefinery that uses the capacity of fungi to decompose and recycle nutrients efficiently and sustainably, to convert the components of biomass feedstocks, such as cellulose, hemicellulose, lignin and other organic matter, into materials and products that can replace or supplement those based on energy-intensive processes and the extraction of limited resources (Jones et al., Reference Jones, Mautner, Luenco, Bismarck and John2020). Mycelium of mainly white rot, Basidiomycota filamentous fungi, can grow its complex network of hyphae feeding on carbon- and nitrogen-rich substrates, binding diverse kinds of anthropogenic waste, such as food scraps, agricultural residues and forestry byproducts, into mycelium-based composites (MBCs) (Vandelook et al., Reference Vandelook, Elsacker, Van Wylick, De Laet and Peeters2021). These composites have a low environmental impact compared to synthetic homologs, as their manufacturing requires low-energy processes (e.g., Stelzer et al., Reference Steltzer, Hoberg, Bach, Schmidt, Pfeiffer, Meyer and Finkbeiner2021); they can store carbon from the atmosphere (Livne et al., Reference Livne, Wösten, Pearlmutter and Gal2022), and have high biodegradability (Van Wylick et al., Reference Van Wylick, Elsacker, Yap, Peeters and De Laet2022; Grimm and Wösten, Reference Grimm and Wösten2018). Their physical, mechanical, chemical and biological properties can be tuned by modifying parameters such as fungal species used, type of substrate, growing conditions and processing methods (e.g. Appels et al., Reference Appels, Camere, Montalti, Karana, Jansen, Dijksterhuis, Krijgsheld and Wösten2019; Rigobello and Ayres, Reference Rigobello, Ayres, Satyanarayana and Deshmukh2023, Livne et al., Reference Livne, Pearlmutter, Gal and Wösten2024). MBCs exhibit material characteristics similar to lightweight foams; some of their applications across industries are architectural forms, construction materials, furniture, interior designs, textiles, leather, packaging and funeral urns (e.g., Aiduang et al., Reference Aiduang, Chanthaluck, Kumla, Jatuwong, Srinuanpan, Waroonkun, Oranratmanee, Lumyong and Suwannarach2022).
Architecture, engineering and construction industry (AEC) is considered the world’s largest raw materials consumer (e.g. Zimmann et al., Reference Zimmann, O’ Brien, Hargrave and Morrell2016; Ghaffar et al., Reference Ghaffar, Burman and Braimah2020) accounting for 11% of greenhouse gas (GHG) emissions globally generated in the construction, renovation and end of life phases of a building’s life cycle (Röck et al., Reference Röck, Ruschi Mendes Saade, Balouktsi, Nygaard Rasmussen, Birgisdottir, Frischknecht, Habert, Lützkendorf and Passer2020). Reducing the embodied energy of building materials and the GHGs associated with those, and developing and implementing energy-efficient and low-carbon alternatives, are essential for stabilizing global temperature increases to no more than 1.5 °C (IEA, 2019; de Coninck et al., Reference de Coninck, Revi, Babiker, Bertoldi, Buckeridge, Cartwright, Dong, Ford, Fuss, Hourcade, Ley, Mechler, Newman, Revokatova, Schultz, Steg and Sugiyama2018). Thus, this sector is a promising and emergent area of advancement of MBCs (Ghazvinian and Gürsoy, Reference Ghazvinian and Gürsoy2022a,Reference Ghazvinian and Gürsoyb). Block materials, particle boards, acoustic panels, thermal insulations, claddings, extrudable pastes, surface sheets and films are some examples of applications in this industry (Aiduang et al., Reference Aiduang, Chanthaluck, Kumla, Jatuwong, Srinuanpan, Waroonkun, Oranratmanee, Lumyong and Suwannarach2022). However, they have limitations related to lower mechanical properties and durability with respect to fossil-based materials (e.g., Sydor et al., Reference Sydor, Cofta, Doczekalska and Bonenberg2022). Hybridization strategies are being discussed in recent literature (Womer et al., Reference Womer, Huynh and John2023; Ballen Sierra et al., Reference Ballen Sierra, Mendes-Pereira, Yupanqui García, Queiroga Werkhaizer, de Rezende, Borges Rodrigues, Badotti, Santos de Carvalho Cardoso, Miura da Costa, Uetanabaro, Aguilar and Góes-Neto2023; Elsacker et al., Reference Elsacker, Peeters and De Laet2022; Jones et al., Reference Jones, Mautner, Luenco, Bismarck and John2020) to improve these materials’ performance beyond the parameters of growth and post-processing. These include adding other organic or inorganic components to lignocellulosic biomasses before inoculation of the fungal species. As previous research has shown, adding fibrous reinforcements can significantly contribute to the compressive strength of the MBCs (Ghazvinian and Gursoy, Reference Ghazvinian and Gürsoy2022a,Reference Ghazvinian and Gürsoyb; Womer et al., Reference Womer, Huynh and John2023), a key mechanical feature required for scalable applications in the AEC industry.
In this study, biomasses designed using agricultural wastes are hybridized with fibers sourced from post-consumer textile waste, toward the development of composite materials with low environmental impact that can help decarbonize the construction industry. Four sets of variables were investigated: type of biomass, type of mycelium, deactivation method for the living mycelium and mycelium composite reinforcement with textile waste fibers, for a total of 77 samples (Tables 1 and 2).
Summary of investigated parameters
Summary of specimens’ dried densities and compressive properties
a Samples B1-0-p-s were only tested to 15%.
In particular, one objective was to assess whether MBCs could be designed to achieve the properties of low-strength concrete, with a compressive strength
$ \ge $
1000 kPa for sidewalk applications. In four common types of urban concrete sidewalks, the largest contributor (∼75%) to global warming potential (kg CO2 eq.) was found to be the use of cement (Oliver-Solà et al., Reference Oliver-Solà, Josa, Rieradevall and Gabarrell2009), therefore we hoped to achieve a minimum strength of 1000 kPa with MBCs.
Methods
Selection of agricultural waste and textile waste
In this section, we discuss information related to the selected local agricultural waste and local textile waste. The choice was driven by local availability and the desire to incorporate nutshells, with higher chances of reaching the target compressive strength of low-strength concrete with respect to hemp hurds (the substrate in the “status quo” products by Ecovative Design). Rupture energies of various cultivars of nuts are listed in Du and Tan (Reference Du and Tan2021), with the best performer being macadamia (not local to California), and with almond shells and walnut shells being very promising (and widely available in Central Valley California). Additionally, walnut shells were recently used as a replacement for aggregate in concrete, with Beskopylny et al. (Reference Beskopylny, Stel’makh, Shcherban’, Mailyan, Meskhi, Shilov, Chernil’nik and El’shaeva2023) finding “An increase into strength characteristics up to 3.5%, as well as the maximum ratio of strength to density of concrete, was observed at a walnut-shell dosage of 5%.”
Almond shells
Almond trees (Prunus dulcis) are the world’s most abundant nut trees (Debevc et al., Reference Debevc, Weldekidan, Snowdon, Vivekanandhan, Wood, Misra and Mohanty2022). The California almond industry owns 80% of the global market share: it produces an estimated 1.36 billion kilograms of almonds, and considerable waste, in particular 0.86 billion kilograms of almond shells (Almond Board of California, 2021). Almond shells are typically used for animal bedding and mulching (70%), dairy feed (10%), energy feedstock (10%) and soil amendment (10%) (Lewis et al., Reference Lewis, Huang, Pohl and Ryan2020). Almond shells are made of hemicellulose, cellulose and lignin, all types of polysaccharides, and have a high carbon content, 47% (Debevc et al., Reference Debevc, Weldekidan, Snowdon, Vivekanandhan, Wood, Misra and Mohanty2022). Therefore, using almond shells in biomass for mycelium composites is not only a valorization strategy for this abundant agricultural by-product, but also a strategy for carbon sequestration.
Cover crops
Fava bean (Vicia faba L.) is one of the world’s oldest crops, having been cultivated for over 10,000 years in the Mediterranean basin (Caracuta et al., Reference Caracuta, Barzilai, Khalaily, Milevski, Paz, Vardi, Regev and Boaretto2015). They are the third most important legume worldwide after soybeans and peas and are currently cultivated in 58 countries (Bangar and Dhull, Reference Bangar and Dhull2022). They are a food staple and protein source in the diets of inhabitants of the Mediterranean basin, Middle East, Ethiopia and China. They are also used for animal feed (Jensen et al., Reference Jensen, Peoples and Hauggaard-Nielsen2010). The health benefits are numerous: extracts and individual chemical compounds from different parts of these plants possess antioxidant, hypoglycemic, hypolipidemic effects, antihemolytic activity, hepatoprotective and anti-Parkinson disease activity (Alam and Najam, Reference Alam, Najam, Bangar and Dhull2022). The beans also have quinine-like anti-malaric compounds, protecting against Plasmodium falciparum malaria (Golenser et al., Reference Golenser, Miller, Spira, Navok and Chevion1983), a historically common disease in the Mediterranean basin. Moreover, fava plants are very beneficial cover crops and nitrogen fixers: they capture N2 from the air and convert it into ammonia (NH3) through their root nodules. Their biomass can be used as green manure, replenishing nitrogen and carbon in the soil, stimulating microbial growth, decreasing disease, pest and weed build-up and enhancing soil’s capacity to hold moisture (Jensen et al., Reference Jensen, Peoples and Hauggaard-Nielsen2010). A recent study has documented their ability to remediate soil co-contaminated by cadmium and lead, considered the most toxic heavy metals in soil (Tang et al., Reference Tang, Hamid, Zehra, Sahitro, He, Hussain, Kumar Gurajala and Yang2019). The use of cover crops such as fava beans improves climate resilience, according to the US Department of Agriculture (USDA, 2023).
Walnut shells
Walnut trees (Juglandaceae family) are an important nut trees both globally and locally in Northern California. Domingos et al. (Reference Domingos, Ferreira, Cruz-Lopes and Esteves2022) estimate world production in 2022 at around 4 million tons of walnuts, of which approximately 56% of the weight is the walnut shell. This means that approximately 2.24 million tons of walnut shell waste is produced globally each year, with 1.5 million tons (shells only) projected for this current year 2023–2024 (Fructidor.com). Although some small percentages of walnut shells are burned for energy or converted into biofuels or pyrolyzed to produce useful chemicals, most of the local walnut-shell agricultural waste stands as an abundant resource needing a functional use. One author is a direct witness to mountains of shells piled up behind the processing plant of a large walnut handler in Yolo County, where the University of California, Davis is located, that produces over 9920 tons of walnut shells each year (Cogdell, Reference Cogdell2023).
Different studies have determined different ratios in walnut shells for their chemical composition of cellulose, hemicellulose, lignin and other extractive substances. The differences likely stem from the testing of different varieties grown in different regional conditions, different soil compositions, and different sun exposure or spacing between the trees in planting (Domingos et al., Reference Domingos, Ferreira, Cruz-Lopes and Esteves2022). Walnut shell composition was found to be approximately 35% lignin, 30% cellulose and 25% hemicellulose, with the other 10% being extractives (dichloromethane, ethanol, and water) (Domingos et al., Reference Domingos, Ferreira, Cruz-Lopes and Esteves2022). Within these extractives, there are some substances with antimicrobial and antioxidant properties present in the walnut shell powder or in walnut shell pyrolysis (Jahanban-Esfahan et al., Reference Jahanban-Esfahlan and Amarowicz2018; Li et al., Reference Li, Liu, Tan, Zhang and Yue2016).
Spent grains from breweries
Brewer’s spent grains (BSG, sometimes referred to as malt bagasse) are a byproduct from beer manufacturing consisting primarily of different species of barley, but also other grains such as corn, rice, wheat, rye or sorghum, depending on which varieties of beer are being produced (de Paula et al., Reference de Paula, Machado Latorres and Guimarães Martins2023). High in protein and fiber (hemicellulose, cellulose, and lignin) and owing to their low cost, BSGs often become a food waste fed to livestock or revalorized in food production. Annually, about 36.4 million tons of BSGs are produced globally, averaging 20 kg/hL of beer manufactured (Nyhan et al., Reference Nyhan, Sahin, Schmitz, Siegel and Arendt2023). This abundance and its nutritional value make it an excellent component for fungal substrates. The spent grain sourced in March 2022 from Ruhstaller Brewery and Farm (Dixon, CA) consisted of mixed grains; it was dried in an oven at 120 °C for 3 h and then frozen for use as an additive in mycelium substrate composition.
Textile waste
Textile waste constitutes one of the biggest problems in the textile and fashion Industry: in 2018 in the USA, around 17 million tons of textile waste were generated, from which only 14.7% was recycled; 18.9% was incinerated for energy recovery, and the rest (66.4%) was discarded into landfills (US EPA, 2020). Worsening the problem, 70% of the world’s textile consumption corresponds to synthetic fibers, such as nylon, acrylic, polyester, spandex, and polypropylene, which are polymers derived from petroleum extraction, and that may take several hundred years to break down in the environment (Echeverria et al., Reference Echeverria, Handoko, Pahlevani and Sahajwalla2019). Textile waste can be classified according to where it is generated in the product’s life cycle: post-industrial textile waste is produced mainly in the process of cutting and making textile products, pre-consumer textile waste appears in the process of delivery of textile products to the consumer; post-consumer textile waste is generated in the process of use and final disposal (Kamble and Kumar Behera, Reference Kamble and Kumar Behera2021).
The main textile waste streams in the Northern California region are post-consumer wastes coming from carpets, mattresses, and clothes at the end of their lifecycle. Although there are several collection, reuse, and recycling strategies for them, they convert into about 630,000 tons of carpet, 265,000 tons of mattresses and 1,200,000 tons of used textiles reached the landfills in 2020, accounting for 5.3% of total municipal solid waste (CalRecycle, 2020). Discarding these materials when they still have their energy potential creates environmental and health issues and huge economic losses.
Looking for both diversion and revalorization of waste streams, textile waste fibers were involved in this study as reinforcements of the MBCs; the focus of the tests performed was to understand their contribution to the mechanical properties.
Mycelium species selection
We selected two mycelium types, Pleurotus ostreatus and Ganoderma lucidum, investigated by Haneef et al. (Reference Haneef, Ceseracciu, Canale, Bayer, Heredia-Guerrero and Athanassiou2017), who assessed their mechanical properties when they were grown on agar plates in a cellulose-rich or a sugar-rich substrate. We also avoided known pathogenic species, not only due to their environmental impact (see for example review by van den Brandhof and Wösten, Reference van den Brandhof and Wösten2022), but also because of the lack of continuous access to a biosafety cabinet/fumehood throughout this study. While Ganoderma lucidum is known to produce chlamydospores, which can potentially produce pathogens in an environment, Pleurotus ostreatus does not, to current knowledge; in either case, we aim for full inactivation of the living mycelium through our processes (Elsacker et al., Reference Elsacker, Zhang and Dade-Robertson2023; Rigobello, Reference Rigobello2023). Several authors have also reported how the fungal properties varied depending on the growth substrate (e.g., Appels et al., Reference Appels, Camere, Montalti, Karana, Jansen, Dijksterhuis, Krijgsheld and Wösten2019, plus review papers by Elsacker et al., Reference Elsacker, Vandelook, Van Wylick, Ruytinx, De Laet and Peeters2020, and Jones et al., Reference Jones, Mautner, Luenco, Bismarck and John2020). The selected species are commercially available in the region, and are known for being edible and medicinal: Pleurotus ostreatus is the second most cultivated mushroom species in the world (Sánchez, Reference Sánchez2010), while Ganoderma lucidum has been known in Asian traditional medicine for thousands of years (Boh et al., Reference Boh, Berovic, Zhang and Zhi-Bin2007). The mycelia were easy to grow on the selected grain for our grain spawn preparation, discussed below, and had a reasonable colonization time in ambient temperature (less than 4 weeks for fully colonized grain spawn).
Biomass and sample preparation
For biomass B1, aged dry almond shells from California were mixed with oak pellets (Mushroom Media Online), pulverized fava bean stalks from a local garden, and used coffee grounds from a household. For biomass B2, walnut shells from Yolo County large walnut handler and Ruhstaller Farm & Brewery in Solano County (California) were procured. They were pulverized in a Vitamix blender to an average particle size of 2 mm2 and mixed with oak pellets, spent brewery grain, and gypsum.
Each biomass was split into three parts to be conditioned into 0%, 25% and 50% fiber volume ratio (Vf) of reinforcement. For the samples containing textile waste, shredded textile fibers from blue jeans, of lengths between 8 and 37 mm, were added in the required proportion and homogenized in the Vitamix blender for 30 seconds at low speed. The blue jeans had the following fiber composition: 78% cotton, 20% polyester and 2% elastane. All samples (with and without textile waste) were put into autoclavable growing bags with a filter patch. For samples using biomass B1, water was added to reach 60% of water content in the total mass prior to inoculation. For samples using biomass B2, water was added to reach 85% of water content in the total mass for 0% fiber volume (Vf) ratio condition, and 100% of water content in 25% and 50% Vf ratio conditions. Each bag was then sealed with sterilization tape, and sterilized in an Instant Pot pressure cooker at a working temperature of 115–118 °C and a working pressure of 0.70–0.80 bars (10.2–11.6 psi) for 1 hour. The use of an Instant Pot as a low-cost and legitimate autoclave replacement is shown by Swenson et al. (Reference Swenson, Stacy, Gaylor, Ushijima, Philmus, Cozy, Videau and Videau2018). After cooling to room temperature, the content was poured into a bowl sanitized with 70% isopropyl alcohol and inoculated using grain spawn mycelium (10% of the wet weight of the total biomass to be colonized). The in-house grain spawn was prepared respectively with liquid cultures of Pleurotus ostreatus or Ganoderma lucidum purchased from Root Mushroom Farm (Amazon.com), and whole oats used for horses from a local animal feed store.
The preparation of the in-house grain spawn is discussed briefly here. The whole oats were soaked in water overnight. After the chaff was removed, the oats were dried on a towel for 30 minutes. The oats were then mixed with pulverized agricultural gypsum (3% of the grain mass), pulverized chalk (a quantity equal to 1/4 of the gypsum quantity) and coffee (2 teaspoons per 1 liter), as per Stamets and Chilton (Reference Stamets and Chilton1983). The oat mixture was then placed in mason jars with lids having a filtered port, and sterilized in an Instant Pot for one hour. Upon cooling, the oat mixture was inoculated by the commercial liquid cultures. Colonization and growth times of the grain spawn depended on the mycelium species and ambient temperatures.
The inoculated biomasses were distributed evenly by hand into previously sanitized silicone ice tray molds (Mossime Store, Amazon.com) with 6 cubic compartments of 50.8 × 50.8 × 50.8 mm3 each, selected according to the geometry requirements of ASTM C109/C109M (Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-in or [50 mm] Cube Specimens). We needed the direct comparison of our results with those of 2 in. (50.8 mm) cube samples made with concrete. ASTM C165 (Standard Test Method for Measuring Compressive Properties of Thermal Insulations) also mentions a 4 in2 square or circular cross-section of test specimens, compatible with ASTM C109/C109M. The silicone molds allowed ease of extraction of the samples. Each filled silicone tray was put into a growing bag and sealed (Figure 1).
Synthesis of the process of preparation of the samples and sample sets achieved. B1 = almond shells, fava cover crop; B2 = walnut shells, spent brewery grains; “0, 25, 50’” = no textile waste, 25% fiber volume ratio, 50% fiber volume ratio; “p” = Pleurotus ostreatus, “g” = Ganoderma lucidum; “s” = saline solution, “d” = dehydrator.
All samples (except those with Ganoderma and the final set B2-0-p-s, discussed later) were put into a controlled incubation chamber (Binder FP56) to guarantee a constant 25 °C during growth. The Ganoderma samples and the B2-0-p-s samples were cured in a loosely covered cardboard box in ambient conditions on the top shelf of a home office in Davis, California, with temperature and humidity measured by hygrothermal sensors (Mocreo ST3) outside the bags and varying between 22 and 30 °C and 35% relative humidity (RH) and 70% RH during the typical 2–3 weeks colonization time. In a more recent experiment in the same home office during a 14 days, 18 hours period (April 25, 2024 to May 10, 2024), the same type of hygrothermal sensor was placed inside a filtered bag containing a 1.2 kg sample of B1 biomass inoculated by P. ostreatus using the same procedures described above; temperature and relative humidity ranges were respectively (22.7 ± 1.81) °C and (95.5 ± 2.40) RH%.
Within a week, the specimens were taken out of the silicone molds, flipped 180 degrees, and left to grow for another 7–14 days inside the filtered bags and in the controlled incubation chamber or air-conditioned home office (Figure 2a and b). When the visible surfaces appeared completely colonized (Figure 2c), the cubes were removed from their growth bag and processed to deactivate the mycelium and prevent fruiting. Two methods were used: either dehydration in an Excalibur dehydrator at 35 °C for 24 h (Figure 2d), or wrapping the samples in a paper towel soaked in an aqueous solution of 10% NaCl for 20 min. Chang et al. (2019) documented reduced fruiting when mycelium is exposed to LiCl, which is however a hazardous chemical based on the 2012 OSHA Hazard Communication Standard (ThermoFisher Scientific, 2021a). On the other hand, Basidiomycetes fungi, to which Pleurotus and Ganoderma belong, have shown “outstanding” intolerance and fruiting inhibition (Tresner and Hayes, 1971) when exposed to solutions of NaCl, which is non-hazardous (according to the same OSHA standard, see ThermoFisher Scientific, 2021b). This fruiting inhibition was also documented in detail by Sung et al. (Reference Sung, Sul, Kong, Yoo, Cheong and Chun2006), for samples from the Pleurotus family. For safety and accessibility reasons, NaCl solutions were preferred for our study, in particular solutions of 5% by weight and 10% by weight. Stunted fruiting was observed in Pleurotus composite samples treated in 5% solutions, while the occurrence of fruiting was typically limited in samples treated in NaCl 10% solutions. After NaCl exposure, the mycelium composite samples were dried in ambient conditions and stored in ambient conditions. Their weights and dimensions were measured prior to testing of their compression properties. The 10% NaCl treatment was applied to B1 and B2 biomasses colonized by Pleurotus (respectively B1-0-p-s and B2-0-p-s), and to all the samples colonized by Ganoderma (with biomass B1 and biomass B2), for a total of 42 samples out of 77 samples. We report that the B2-0-p-s set was cured at room temperature due to the unavailability of the incubation chamber at the time.
Sample growth and deactivation: a) growth after 6 days; b) flipped samples; c) samples after 16 days of growth; d) dehydrated samples e) saline-deactivated samples.
Testing of compression properties
The majority of the specimens were tested for compressive strength and modulus in an Instron 5965 screw-driven machine with a 1 kN load cell calibrated on a regular basis. MBCs are heterogeneous materials that have not been standardized; we used guidelines from several ASTM standards. There appears not to be a unanimous agreement in the scientific community on the testing standard for compressive properties of MBCs.
ASTM C165 is mentioned for the compressive properties of commercial grow-it-yourself mycelium hemp-based samples by Ecovative Design (https://grow.bio/pages/spec-sheet). Elsacker (Reference Elsacker2021) and Rigobello and Ayres (Reference Rigobello and Ayres2022) used ASTM D1037 (Standard Test Methods for Evaluating Properties of Wood-Base Fiber and Particle Panel Materials). In the review by Jones et al. (Reference Jones, Mautner, Luenco, Bismarck and John2020), the use of ASTM C578 (Standard Specification for Rigid, Cellular Polystyrene Thermal Insulation) is reported, which provides guidelines for several material properties, and recommends either ASTM C165 or ASTM D1621 (Standard Test Method for Compressive Properties of Rigid Cellular Plastics) for the measurement of compressive properties. Yang et al. (Reference Yang, Zhang, Still, White and Amstislavski2017) used ASTM D2166-16 (Standard Test Method for Unconfined Compressive Strength of Cohesive Soil). ASTM C165, D1037, D1621 and D2166 have different requirements for sample geometries, testing rates and loading fixtures, which could impact a direct comparison of results. The minimum number of samples in ASTM C165 is four per condition, and we satisfied this requirement. The only loading fixture for compression tests that was available to us for the selected testing machine consisted of two platens with a circular cross-section with a 50.8 mm diameter and an area of 2207 mm2. ASTM C109/109M requires uniform contact of the samples with the platens. ASTM C165 and ASTM D1621 call for the platens to be larger than the samples cross-sectional area in order to achieve a uniform applied load on the samples. Upon drying, our MBC samples had various levels of shrinkage along the height of the samples, leading to cuboid shapes with some variability along the height. In Table 2, we report the cuboid areas (average, standard deviation and medians) for all sets; we used median cross-sections for the computation of the compressive properties.
Testing occurred at a displacement rate of 2 mm/min (based on the range provided by ASTM C165), with time, displacement and load data acquired every 0.5 s. We note that ASTM D1621 has a higher displacement rate (12.5 mm/min), and that ASTM C165 states that higher testing speeds “usually result in higher compressive resistance values,” consistent with the rate-dependent nature of polymers such as foams and mycelium. The first endpoint of our tests was selected to be 15% deformation (machine displacement/(initial height)), to compare our results with those published by Ecovative Design, namely 124 kPa compressive strength and 1138 kPa compressive modulus. One set of samples (B1-0-p-s) was only tested in this condition. To further investigate the compressibility of the materials, the other samples were all re-tested to 30% deformation, either on the same day or over different days based on machine availability. The set of 6 samples (B2-0-g1-s) reached almost the 1kN load capacity of the testing machine under 15% compression, so the samples were tested on a different machine under 30% compression (a 100 kN MTS 322.21 with wide platens, regularly used for concrete testing). Compressive resistance is defined as the compressive load per unit of the original area (the median area of each set was used). If the sample fails, the compressive resistance is called compressive strength. The compressive modulus of elasticity E, or compressive stiffness when the samples are loaded in axial compression, was computed while the material was still linear elastic, and according to the ASTM C165 formula:
$E = {{{load/\left( {unit\;area} \right)}}\over{{deformation/\left( {original\;thickness} \right)}}}$
. We selected a deformation of 2 mm where the linearity criterion was met for all the samples, using a least squares fit.
The data was assessed for normality using two common tests, Kolmogorov-Smirnov and Lilliefors, at a confidence level of 95%, using the commercial software MATLAB (MathWorks, MA, USA). All the data did not pass the Kolmogorov–Smirnov test, but did pass the Lilliefors test. Hence, the data was instead analyzed with boxplots, a well-established representation tool of data affected by scatter. Boxplots allow straightforward non-parametric statistical inferences. They were computed with the commercial software GraphPad Prism (Dotmatics, MA, USA).
Results
We present and discuss hereby the results, with an assessment of the limitations of this first study. Photos of the tests are shown in Figure 3, with plots of normalized compression resistance and compression modulus in Figure 4, and results in Table 2.
testing compressive behavior: Column a) specimen B1-50-g-s ready to test; column b) specimen (top) B2-25-p-d and (bottom) B2-0-g1-s at the end of the 30% test; column c) specimens B1-0-p-d collapsing during their tests; column d) specimen B2-50-p-d during testing and after test; column e) specimen B1-0-g-s during test and after test.
Boxplots reporting mechanical properties normalized to median dry weights prior to testing of a total of 77 tested MBC samples. In each boxplot, the whiskers extend to maximum and minimum value, the box shows the 75th percentile (top line), the median (horizontal line inside the boxplot) and the 25th percentile (bottom line) of the data set. (Top) estimated compressive modulus/(median dry weight) measured at 2 mm (only in Test 1). (Bottom) estimated compressive resistance/(median dry weight) measured for the two sets of compression tests (Test 1, 15% deformation; Test 2, 30% deformation). Samples are labeled using the following code in the format “biomass- textile fiber volume ratio-fungal species-deactivation method.” B1 = almond shells, fava cover crop; B2 = walnut shells, spent brewery grains; “0, 25, 50” = no textile waste, 25% fiber volume ratio, 50% fiber volume ratio; “p” = Pleurotus ostreatus, “g” = Ganoderma lucidum; “s” = saline solution, “d” = dehydrator. “g1” and “g2” are two different sets of G. lucidum used with B2, with different densities (see Table 2).
Material properties and comparison with conventional construction materials
During axial compression, the MBC samples, which were unsupported (as per standards) typically expanded laterally (Figure 3). As a result, the output compressive properties need to be considered estimates, because uniform compressive loading throughout the entirety of the tests could not be obtained. Lateral deformation was observed in all samples at various degrees, due to geometry shrinkage during the mycelium deactivation, and the heterogeneous nature of the samples.
All samples without textile waste, with the exception of B1-0-g-s and B2-0-g1-s, exhibited significant cracking during loading (Figure 3c).
Regarding the compressive modulus, data is only reported for Test 1 (“1” on the x-axis of Figure 4, top) since most samples were not elastic anymore after the first compression test. Samples prepared with the B2 biomass had superior stiffness properties for 0% and 25% Vf textile waste with respect to B1 samples. However, when results are normalized by the median dry sample weight prior to testing, the boxplot (Figure 4) shows statistical equivalence of all the samples.
Regarding the compressive resistance per median dry weight, the samples behaved similarly with respect to each other when compressed to 15% deformation (“1” on the x-axis of Figure 4, bottom). In the second round of tests (“2” on the x-axis of Figure 4, bottom), however, we observe that the samples made with Ganoderma (B1-Vf%-g-s and B2-0-g1-s, all deactivated with saline solution) clearly outperformed all the others. Samples B2-0-g2-s were prepared with higher packing in each cube (70 g prior to colonization versus the 60 g per cube of B2-0-g1-s) and had a considerably lower resistance per weight (see also Table 2).
Table 2 reports the mean and standard deviations of the dry samples’ densities and the estimated non-normalized compressive properties obtained with the formulas from ASTM C165, Young’s modulus in compression/compressive stiffness, measured at 2 mm displacement, and compressive resistance measured at 30% deformation.
The mycelium composites of our study are heterogeneous due to the presence of a living organism colonizing agricultural waste from harvested nut trees and fava bean plants, to which textile waste may be added (textiles have variable behaviors as well). Variability has been observed in many common material systems used in engineering, from structural concrete (with coefficients of variation between 15% and 20%, Mirza et al., Reference Mirza, Hatzinikolas and MacGregor1979), to structural wood (coefficients of variations from 24% to 39% for compression perpendicular to the grain, Wood, Reference Wood1960) to natural and synthetic fibers (reviews by Faruk et al., Reference Faruk, Bledzki, Fink and Sain2012; Singh et al., Reference Singh, Tewari, Zafar, Mavinkere Ragappa and Siengchin2023). The scatter of our results (shown in Figure 4 and in Table 2) could also be due to manufacturing variability, for example, due to the presence of inoculated whole oats randomly located in the samples. We did not homogenize the inoculated oats out of concern about their sterility, however, we will attempt this route in future work. Our work is a novel exploratory study, and we expect challenges in working with nonconventional material such as MBCs and designing future research paths.
Considerable scatter of results was also observed by Yang et al. (Reference Yang, Zhang, Still, White and Amstislavski2017), Rigobello and Ayres (Reference Rigobello and Ayres2022), Ghazvinian and Gürsoy (Reference Ghazvinian and Gürsoy2022b) and Livne et al. (Reference Livne, Pearlmutter, Gal and Wösten2024), who used different testing standards and different biomasses for their mycelium composites.
Our best performer was the B2-0-g1-s group, with an average compressive resistance of 775 kPa (77.5% of low-strength concrete). When normalized with respect to median dry weight, these samples are statistically similar to the B1-Vf%-g-s. In particular, B1-50-g-s has the smallest coefficient of variation (standard deviation/mean) equal to 11%.
Within our constraints, methods and limitations, this work highlights the better performance of the Ganoderma-based samples versus all the Pleurotus-based samples. Future manufacturing improvements will focus on Ganoderma-based samples with biomass B1 and B2. We hypothesize that the type of hyphae (trimitic in the Ganoderma, dimitic in the Pleurotus) may have also synergistically contributed to these results (Porter and Naleway, Reference Porter and Naleway2022; Sydor et al., Reference Sydor, Cofta, Doczekalska and Bonenberg2022). Based on our results, the estimated compressive resistance properties of our MBC samples are approaching those of low-strength concrete for sidewalk applications; we could not quite reach the 1000 kPa target with the current configurations. The compressive strength properties at 30% deformation of our MBCs appear very competitive with respect to those reported in the MBCs’ review literature (e.g. Attias et al., 2023; Jones et al., Reference Jones, Mautner, Luenco, Bismarck and John2020; Girometta et al., Reference Girometta, Picco, Baiguera, Dondi, Babbini, Cartabia, Pellegrini and Savino2019), and with those of some synthetic materials of the residential and commercial construction industries, that have a significant environmental burden (for example, Zheng and Suh, Reference Zheng and Suh2019; for polyurethane, see Kylili et al., 2019, which also includes an assessment of toxicity; for polystyrene versus mycelium, see Caneso et al., Reference Caneso, Canjura and Woodell2018, Enarevba and Haapala, Reference Enarevba and Haapala2023, with the latter including toxicity data). We give here three examples of such commercially available materials: Carlisle SealTite Pro spray foam polyurethane for insulation has a published compressive strength of 31 psi, or 213 kPa, measured with ASTM D1621 (https://www.carlislesfi.com/wp-content/uploads/CSFI-13615-SealTite-PRO-HFO-TDS_05-26-22-1.pdf); Owens Corning Foamular® and Foamular NGX® 150 has a compressive strength
$ \ge $
104 kPa, also measured with ASTM D1621 (http://www.owenscorning.com/dms/10018086); applications of this material are cavity walls, wall sheathing, weather-resistant barrier, perimeter/foundation walls, basement walls; finally, floor underlayment material DMX1-Step 2.0TM has a published compressive strength of 10,000 lbs/sq.ft., i.e. 479 kPa (www.dmx1step.com), in the ballpark of the properties of the B1-Vf%-g-s and B2-0-g1-s sets. One objective of our study was to design materials that could be adopted in the construction industry, starting from the structural properties. While we have not tested for other relevant properties (e.g. thermal, fire and moisture resistance), our results provide a very promising set of compressive properties, which warrants further investigation towards the replacement of some conventional construction materials with MBCs. The low-strength concrete compressive resistance is also within reach with further investigation of the B2-0-g1-s configuration, updated with textile waste or sand.
Incubation and post-processing methods
In our study, we do not notice statistical differences between using an incubation chamber for 2 weeks or using ambient conditions in a home office for 2 weeks, which has a significantly lower carbon footprint.
In the Pleurotus samples with biomass B1 and with no textile waste (B1-0-p-s and B1-0-p-d), the compression resistance does not significantly depend on the two selected deactivation methods. Soaking samples in a saline solution for 20 minutes followed by their drying in ambient conditions is a significantly affordable, accessible process with an insignificant energy footprint with respect to a dehydrator running for 24 hours. We point out the significant larger shrinkage from the original dimensions (50.8 mm cube) of the live MBCs with the saline solution post-processing (Table 2). We have a direct comparison of mechanical properties of Pleurotus samples de-activated with saline solution or with heat, which show to be statistically equivalent. This deactivation approach was not effective with biomass B2 and Pleurotus. We will continue investigating the salinity-based de-activation method, not only to characterize and fine-tune the dimensional changes that occur as a result of it but also to investigate their self-healing ability after mechanical loading. In fact, preliminary testing in our group showed self-healing to occur after mechanical testing, when samples had been processed at low (5%) NaCl concentrations.
Textile reinforcement contribution
For Pleurotus samples, the presence of textile waste shows some improvement in the compressive resistance in the second test (30% compression). In the Ganoderma samples, there is a significant increase of compressive resistance in the second test for all tested textile waste fiber volume ratios. We also note that the textile waste incorporated in the samples was unfortunately not biodegradable, but it provided an upcycle opportunity for textile waste, which has a considerable potential for materials design.
Carbon:nitrogen ratio
In Figure 5, the carbon:nitrogen (C:N) ratios are presented for selected samples of our study, including non-inoculated dry biomasses B1 and B2. The C:N ratio of biomass B1 turned out to be significantly higher than that of biomass B2. Almond shells and walnut shells have similar lignin content (28–30%), but almond shells have a higher percentage (about 6%) of polysaccharides than walnut shells (Queirós et al., Reference Queirós, Cardoso, Lourenço, Ferreira, Miranda, Lourenço and Pereira2020). We surmise that the N content from the proteins present in the brewery spent grains could be more significant than the N content in fava stalks biomass.
Carbon:nitrogen (C:N) ratios of selected samples. “B1 biomass” and “B2 biomass” are the dried ingredients prior to addition of water and grain spawn. “g,” “p” and “tr” indicated samples colonized respectively by Ganoderma lucidum, Pleurotus ostreatus and Trametes versicolor (note: T. versicolor samples from a parallel ongoing project). T1, T2 and T3 are three different samples with 25% textile waste (T1 and T2 come from the same bag). Samples indicated by the labels “active” were tested prior to their deactivation.
Regarding the inoculated biomasses, the literature reports C:N values of 3–51 for mycorrhizal fungi, and 4–62 for saprotrophic fungi (Strickland and Rousk, Reference Strickland and Rousk2010) -Pleurotus, Ganoderma and Trametes versicolor (also tested) are saprotrophic. The textile waste’s contribution to the C:N ratio does not stand out, with the exception of the B2-pT value, which may be an outlier. We observe that inactive (through both methods, heat or saline solution) and active mycelium samples show a similar range of C:N. When active, mycelium efficiently converts the carbon in the accessible biomass into energy and polysaccharides (mostly chitin) needed for 80–90% of its cell walls, while nitrogen and carbon together build the enzymes that will decompose the biomass into more easily digestible nutrients. Experimentally, we noticed that the B1-based mycelium samples appeared to have in general a faster and smoother colonization process than those of the B2-based samples; higher C:N content of the B1 biomass may justify this timeline difference.
When the mycelium is inactive (dead), it contains mostly cell wall material (particularly polysaccharides), lipids and proteins, an “important source of both C and N” (Brabcová et al., Reference Brabcová, Nováková, Davidová and Baldrian2016). Since the inactive samples for the C:N tests were not exposed to composting microorganisms prior to being sent to the laboratory, it is plausible that the ratios are similar. We also compare the dead B1-p data point with the active B1-g1, and B1-g2 data points. Trametes is another polypore fungus like Ganoderma, with trimitic hyphae (Cui et al., Reference Cui, Li, Ji, Zhou, Song, Si, Yang and Dai2019); hence, they contain more chitin than Pleurotus’ monomitic hyphae, and thus carbon content.
Future work will continue to assess C:N ratios, to further quantify the ability of these MBCs to sequester carbon.
Conclusions
In this exploratory research, we demonstrated a viable use of biorefinery technologies to reduce, divert, and revalorize local wastes into MBCs with compressive strengths comparable to the properties of selected fully synthetic materials used in the residential and commercial construction industries. One set of 6 samples reached approximately 77.5% of the compressive resistance of low-strength concrete when under 30% deformation, without the incorporation of textile waste; we are not aware of MBC results in the literature with a comparable performance. Therefore, we will be investigating this material combination further. The addition of textile waste fibers as a hybridization strategy contributed to increased compressive resistance, prevented loss of integrity, and produced distributed cracking in higher deformation tests (30%) compared to samples without reinforcement. We also propose short-time exposure to saline solution in ambient conditions, in place of heating, to prevent fruiting. Scatter in the results points at local anisotropy of the samples, as work for example by Yang et al. (Reference Yang, Zhang, Still, White and Amstislavski2017), Rigobello and Ayres (Reference Rigobello and Ayres2022), Ghazvinian and Gürsoy (Reference Ghazvinian and Gürsoy2022b), Livne et al. (Reference Livne, Pearlmutter, Gal and Wösten2024) suggests; therefore, future work will explore improvements in the manufacturing process to reduce this scatter, understanding and improving the interfacing of waste fibers and agricultural feedstocks with the mycelium in the growth face, and testing other properties of the composites achieved. We consider one limitation of our study to be the lack of uniform loading that is expected in standardized tests of conventional materials but could not be achieved by the majority of our MBCs (the exception being set B2-0-g1-s tested on wide platens to 30% deformation).
Nonetheless, MBCs are an example of a paradigm shift and a biotechnology research direction in the area of materials design from assembled materials to grown materials (Camere and Karana, Reference Camere and Karana2018), which can lead to a less extractive and more regenerative economy. Our research demonstrated this through the biofabrication of novel composite materials made entirely from local waste streams (textile and agricultural waste) in collaboration with living organisms. Fungal biorefineries play a crucial role in transitioning to a circular economy where material flows can mimic natural systems’ energetic bioconversion at the end of lifespan, creating opportunities for materials design, waste diversion and revalorization into more sustainable materials.
Data availability statement
All data is given in the main document, particularly Table 2.
Acknowledgments
The authors thank the support of Professors Beth Ferguson (Design), Laura Bogar and David Rizzo (Plant Biology), Michele “Mike” Barbato (Civil and Environmental Engineering, Co-PI of both grants received by V.L.S.), the Davis Mycology Society, Tanya Garcia-Nolen (research engineer and manager of the J.D. Wheat Veterinary Orthopedic Research Laboratory at UC Davis School of Veterinary Medicine). Last but not least, the authors thank for their assistance in manufacturing and testing samples undergraduate student researchers Shivani Torres-Lal, Dominic Soufl, Nicholas Gallo and Zeyyad Borham (Advanced Composites Research, Engineering and Science, ACRES, Laboratory), high-school students Bhavya Mamnani and Lyka Raymundo (Summer 2023 interns in the ACRES Laboratory), Ph.D. student Hyunjune Gill, for the translation of the article by Sung et al. (Reference Sung, Sul, Kong, Yoo, Cheong and Chun2006), and postdoctoral fellow Dr. Nitin Kumar (advised by Dr. Michele “Mike” Barbato, Civil and Environmental Engineering), for his assistance with the testing of the B2-0-g1-s samples.
Author contributions
Alejandra Ruiz: conceptualization (equal), data curation (supporting), funding acquisition (equal), investigation (equal), methodology (equal), project administration (equal), software (supporting), visualization (equal), writing -original draft (equal). Christina Cogdell: conceptualization, funding acquisition, investigation, methodology, resources, supervision, writing -review& editing (supporting all). Jacques Mak: investigation, visualization, writing -review & editing (supporting all). Allison Rowe: investigation, writing -review & editing (supporting all). Shuhao Wan: investigation, methodology, writing -review & editing (supporting all). Valeria La Saponara: conceptualization (equal), data curation (lead), formal analysis (lead), funding acquisition (equal), investigation (equal), methodology (equal), project administration (equal), resources (equal), supervision (equal), writing –original draft (equal).
Financial support
This work was supported by the Laboratory Directed Research and Development program at Sandia National Laboratories, a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia LLC, a wholly owned subsidiary of Honeywell International Inc. for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525 (grant to V.L.S.). It was also supported by funding provided by UC Davis (The Green Initiative Fund to A.R., grant F22-106, The College of Engineering 2022/2023 Next Level Research seed grant to V.L.S.).
Competing interests
We report no conflicts of interest that we are aware of.
Ethics statement
No human subjects/animals were involved in this study.
Open access
