Introduction
In our contemporary epoch, anthropogenic disturbances to natural ecosystems are pervasive, with human activities and their resultant materials increasingly overshadowing natural cycles globally. As articulated by Schellnhuber in 1999, the global system can be viewed through two primary lenses: the ecosphere (N), encompassing the intricate network of atmospheric, hydrospheric, cryospheric, lithospheric, and biospheric elements, and the human factor (H), which encapsulates all human actions, products, and the metaphysical dimensions of human endeavor [1].
Achieving a sustainable coevolution between the ecosphere and the anthroposphere necessitates a paradigm shift in scientific perspectives and methodologies, including the reinvention of manufacturing processes. The escalating human footprint on the planet demands a conscious integration of natural feedstocks into innovative manufacturing processes. This integration allows for the life cycle of products and materials to be considered from the outset of production. One significant stride towards enhancing the sustainability of additive manufacturing (AM) is the substitution of conventionally sourced feedstocks with surplus natural materials. This approach not only bolsters sustainability but also adds substantial value to materials and products considered “pre-owned.”
Adhering to the principle of prioritizing natural and recycled feedstocks over dedicated resources is crucial for all production processes, especially within the framework of a circular economy. The proactive development of naturally available feedstocks represents an environmentally conscious scientific approach in the realm of material sciences.
Additive Manufacturing: A Sustainable Approach
Additive manufacturing, commonly known as 3D printing, represents a transformative shift in manufacturing technologies. It builds objects by adding material layer by layer, contrasting with traditional subtractive methods that carve away material from a larger block [2]. This layer-by-layer approach is revolutionizing industries by offering unprecedented design flexibility, multi-material processing capabilities, and the potential to create unique material properties and functionalities, including functionally graded materials. Binder Jetting (BJ), a particularly versatile AM technology, utilizes powdered materials as feedstock [3, 4], as illustrated in Fig 1. In BJ, a layer of powder is spread, and a print head selectively deposits a binding liquid, consolidating the powder in specific areas to define the object’s cross-section layer by layer.
Fig 1. Schematic of the Binder Jetting (BJ) process, illustrating layer-by-layer material deposition and selective binder application.
The widespread adoption of BJ stems from its broad compatibility with powdery materials, provided they exhibit sufficient flowability for homogeneous layer spreading. This adaptability has spurred researchers to explore abundant materials like sand [5–7] and recycled materials in powder form [8] as feedstocks, aiming to repurpose waste into valuable new products. While this approach reduces the costs and resources associated with synthesizing initial materials, refining recycled materials into powders suitable for BJ remains essential if they are not directly obtained in an appropriate powdery state [9].
This study explores a strategy that aligns the advancement of AM with the utilization of natural or nature-recycled materials as feedstocks. Specifically, we investigate the powdery residue produced when house borer larvae or drywood termites consume wood—known as frass—as a novel feedstock for 3D printing. While 3D printing of wood chips [10–14] and plastic-wood composites has demonstrated the feasibility of creating objects with wood-like properties, these methods often rely on dedicated feedstocks refined with polymeric additives [15–18]. Utilizing insect frass for 3D printing wood-based materials represents a significant step forward in sustainability by employing naturally occurring and naturally processed materials directly as feedstocks.
The binder system in BJ plays a critical role in consolidating the powder material. It must possess suitable viscosity and surface tension for precise dispensing by commercial printer heads [4, 19, 20], effectively moisten the powder layer, and ideally interact with the powder to create strong interparticle adhesion. However, binder addition typically does not significantly densify the powder, resulting in porous parts that often require post-treatment for densification [21, 22].
Timber-Derived Feedstocks: Frass as a Novel Material
There is a growing interest in employing eco-friendly and recycled materials in 3D printing, particularly in Binder Jetting [11–16, 23, 24]. BJ’s versatility allows it to utilize virtually any material that can be reduced to powder of an appropriate particle size. Particle size is crucial for ensuring adequate flowability for defect-free layer deposition; excessively fine powders exhibit poor flowability, while overly coarse powders compromise part precision [25]. Wood particles, often byproducts of wood machining like sawdust, or deliberately processed wood, are commonly used. Many studies refine wood particles for 3D printing by mixing them with polymers or other binding phases [15, 17, 26–29]. However, timber can also be naturally processed into printable powders by wood-feeding insects. This research explores the use of small particles from the feeding byproducts of the European house borer (EHB) Hylotrupes bajulus and drywood termite Incisitermes marginipennis as raw materials for 3D printing.
Insects that infest construction timber possess unique anatomical and physiological adaptations suited to this environment. Wood is a heterogeneous, porous matrix composed primarily of cellulose and lignin, with varying proportions between heartwood and sapwood, and within the early and late wood of annual rings. This variability leads to localized differences in strength and nutrient distribution for insects. Wood-boring beetle larvae and drywood termites possess strong mandibles capable of abrading all wood components. Cellulose, the primary hydrocarbon source, is typically digested with the assistance of cellulase-producing microorganisms. However, the scarcity of nitrogen-containing elements in wood is a limiting factor. Consequently, wood-feeding insects consume significantly more wood than required for development to maximize nitrogen intake. H. bajulus larvae, for example, extensively tunnel through sapwood, leaving behind frass-filled galleries. Frass consists of loosely chopped wood particles (debris) and densely packed feces, the latter being cylindrical formations of semi-digested cellulose-lignin conglomerates. Insect-induced damage to timber can pose significant structural risks, necessitating pest control measures. Institutions like BAM rear pest populations to evaluate control strategies. The frass byproducts from these rearings, typically discarded, present a unique opportunity. Modified by insect digestive systems, this initially non-uniform wood is transformed into a homogeneous cellulose-lignin mixture, potentially suitable for technical applications like 3D printing without further processing. This study evaluates frass as a 3D printing feedstock, comparing drywood termite frass, characterized by uniform, six-sided pellets, to the sawdust-like, irregularly shaped frass of house borers. Unlike drywood termites, house borer larvae only partially digest abraded wood, resulting in frass containing both debris and more compact feces.
Materials and Methods
European house borer (EHB) larvae were cultivated under controlled conditions at 28 ± 2°C and 75 ± 5% relative humidity. Newly hatched larvae were manually introduced into small blocks of pine (Pinus sylvestris) sapwood (1.5 x 2.5 x 5 cm3) enriched with peptone and yeast. This nutrient enrichment was achieved by impregnating the sapwood with a 1% peptone and 0.3% yeast aqueous solution under low pressure (100 to 200 mbar for 30 minutes) to accelerate larval development. Two larvae were placed per wood block and allowed to feed for approximately six months before being transferred to pure pine sapwood blocks (3 x 4 x 5.5 cm3) without nutrient enrichment. As larvae fed within the wood, they produced debris and feces, collectively known as frass, which accumulated in the tunnels. Debris comprises undigested wood particles resulting from mandibular abrasion, generally bypassing the larvae during movement. Feces are densely packed into cylindrical pellets after passing through the larval hindgut. As wood consumption progressed, frass (debris and feces) gradually expelled and collected in larger quantities.
EHB frass was sieved using a vibratory sieve shaker (Analysette 3 spartan, Fritsch, Germany) for 30 minutes at a 1 mm amplitude. A particle size fraction of 45 μm–100 μm, representing 17% of the total frass (with 57% > 100 μm and 26% < 45 μm), was used for 3D printing (see Fig 2). Despite their flake-like shape, the frass particles exhibited considerable flowability. The Hausner Ratio (HR), a measure of flowability [30], was calculated using Eq 1:
(1)
where HR is the Hausner ratio, ρBulk is the freely settled bulk density, and ρTap is the tapped bulk density after reaching a plateau, both in g/cm3. Tapped density was determined following ISO 787–11 using a STAV 2003 equipment (J. Engelsmann AG, Germany). With ρBulk = 0.14 g/cm3 and ρTap = 0.18 g/cm3, EHB frass particles had an HR of 1.25, indicating fair flowability [31].
Fig 2. a) European house borer (Hylotrupes bajulus) larva (top) and adult beetle (bottom); b) Sieved frass (45-100 μm fraction) produced by larvae, used for 3D printing.
Drywood termite frass was also evaluated as a 3D printing feedstock (see Fig 3). Termites rely on gut symbionts (fungi, protists, and bacteria) to digest wood, breaking down lignin and cellulose. In contrast to EHB frass, drywood termite frass consists of almost uniformly sized, six-sided pellets with excellent flowability, ideal for layer-wise construction in 3D printing (see Fig 3). These compact pellets are composed of fine fibers and particles. Their high flowability (HR = 1.1, ρBulk = 0.67 g/cm3, ρTap = 0.74 g/cm3) makes them highly suitable for depositing uniformly packed layers in 3D printing.
Fig 3. a) Drywood termites (Incisitermes marginipennis), soldier and worker with frass pellets; b) SEM micrograph of drywood termite frass pellets showing their fibrous composition.
Drywood termite pellets were imaged using 3D X-ray computed tomography (ZEISS Xradia 620 Versa). Imaging was performed at 80 kV and 10 W, with X-ray filtering using a LE1 filter. A 6.8x geometrical magnification (25 mm source-object, 145 mm object-detector distance) and 0.4x optical magnification yielded a 10 μm effective pixel size. Pellets were fixed on adhesive tape rolled up for measurement due to their light weight and electrostatic repulsion (see Fig 4). 3D tomographic data from 801 angular projections were reconstructed using ZEISS reconstructor software and processed with Avizo software (Thermo Fisher Scientific, USA). Length, breadth, and width, defined as ferret diameters, were measured for each 3D object. Length is the longest ferret diameter, width the shortest, and breadth the longest orthogonal diameter to length.
Fig 4. 3D rendered X-ray tomographic dataset of termite pellets (left) and dimensional analysis (right) showing length, breadth, and width distributions of 180 pellets.
Fig 4 shows the 3D rendering of termite pellets, revealing their uniform, ellipsoidal shape and narrow size distribution. The 6-fold symmetry observed in SEM images (Fig 3) is less apparent in the 3D rendering. Dimensional measurements of 180 pellets (graphed in Fig 4) show significant differences between length (average 1050 μm) and breadth/width (average 580 μm), with width and breadth being similar, indicating a rounded shape and an aspect ratio of approximately 1.8. The pellet size is relatively large for BJ. To accommodate this, layer thicknesses of at least 600 μm were tested, with 800 μm proving optimal for smooth layer deposition.
Specimens were 3D printed from EHB and drywood termite frass using a commercial 3D printer (RX-1, Prometal RCT GmbH, Germany). A water-based binder (PM-B-SR2-02, ExOne GmbH, Germany; viscosity 10.7 mPa·s @ 1000s-1) was used. Binder cross-linking occurred after each layer in the printer’s thermal curing station. Printing precision depends on parameters like powder particle size, flowability, layer thickness, and binder saturation. With standard fine powders (< 60 μm), the printer achieves < 100 μm volumetric resolution. Printing parameters were adjusted for frass, including binder saturation, layer thickness, and curing time. Binder saturation, the ratio of binder volume to powder bed porosity, was set at 100% for EHB frass and 166% for termite frass. Optimal layer thicknesses were 100 μm for EHB and 800 μm for termite frass.
Fig 5 displays the printed structure models. A 9 mm3 cube with rectangular struts (Fig 5A) was used to assess printing accuracy with fine ESB powders. A model designed to evaluate fine feature reproduction with 1050 μm particles (Fig 5B) consisted of inner (2 mm gap) and outer (4 mm gap) frames.
Fig 5. Schematic drawings with dimensions of the 3D printed specimens: a) Cubic structure for EHB frass, b) Framed structure for termite frass.
Results and Discussion
Binder Jetting 3D Printing of EHB Frass
A cubic structure was printed based on the model in Fig 5A (see Fig 6A). The printed structure closely replicated the model, except for a convex bottom plane, indicating binder oversaturation. Oversaturation also reduced capillary diameters; designed 1 mm square cross-sections measured approximately 500 μm. Dimensional tolerance was around 200 μm.
Fig 6. Images of manufactured specimens from European house borer (EHB) frass: a) Photograph of a printed cubic sample; b) Light microscope image showing wooden chops and macroscopic channels, highlighting the porous nature of the printed material.
Due to EHB frass’s low packing density, 3D-printed parts exhibited low mechanical strength and were unsuitable in their as-printed state. However, their good detail quality makes them potential preforms for further processing, such as infiltration to enhance mechanical strength. Frass particles were loosely packed during layer deposition, and binder addition did not significantly increase densification. Fig 6B shows the as-printed part’s fluffy, porous structure.
Binder Jetting 3D Printing of Drywood Termite Frass
Particle dimensions from Fig 4 indicate an average 580 μm breadth/width and 1050 μm length, resulting in an aspect ratio of ~1.8. 3D analysis confirmed termite pellets’ uniform, ellipsoidal shape and narrow size distribution. Drywood termite frass exhibits excellent flowability and achieves higher packing densities than EHB frass. This higher density suggests potential for load-bearing applications of printed parts without extensive post-processing.
However, compacted porous particles like frass pellets present a challenge in BJ: binder is drawn into the pellets’ fine pore network by capillary forces. Pellet pores must be saturated before sufficient binder becomes available for inter-pellet gluing. Smaller intra-pellet capillaries exert stronger binder uptake forces than larger inter-pellet capillaries. Consequently, significant binder is consumed within pellets before effective interparticle bonding occurs. To compensate for pellet binder uptake, a 166% binder saturation was used. No pellet swelling or structural integrity loss was observed upon binder uptake. Layer deposition onto initial printed cross-sections was challenging due to binder-saturated pellet adhesion to the recoater. This was resolved by applying gas flow through the powder bed, stabilizing deposited material, a technique recently introduced for poorly flowing powders [32]. Fig 7 shows parts printed from termite frass based on models in Fig 5.
Fig 7. a) Binder jetting powder bed filled with drywood termite frass; b), c) Part printed according to the framed model in Fig 5B, showing successful reproduction of larger features; d) Printed part similar to the cubic model shown in Fig 5A.
To assess fine feature reproduction with 1050 μm particles, the model from Fig 5B was printed (see Fig 7C). The inner square frame (2 mm gap) was poorly reproduced, while the outer frame (4 mm gap) was reasonably well resolved. Dimensional tolerance was approximately 1000 μm. Although termite frass pellets are about an order of magnitude larger than typical 3D printing powders, coarser powders can be advantageous for printing larger wooden objects like furniture [16], enabling thicker layers and faster build rates. Structural details are reproducible at millimeter scales but not below one millimeter.
Conclusions
This study explored drywood termite (Incisitermes marginipennis) and European house borer (EHB) (Hylotrupes bajulus) processed wood as 3D printing feedstocks, aligning with a strategy of utilizing naturally available materials for sustainable material sciences. The processability of frass from these insects, feeding on construction timber, varied significantly. EHB frass exhibited a flaky structure and poor packing density, whereas termite frass comprised uniform pellets with excellent packing. Despite these differences, both could be spread into homogeneous thin layers for Binder Jetting 3D printing. Termite frass pellets, with dimensions of 580 μm (short axis) and 1050 μm (long axis), are about ten times larger than typical 3D printing powders. For wooden objects like furniture, coarser powders can facilitate thicker layers and faster build rates. Finer, sawdust-like EHB frass particles are better suited for intricate structures. Neither feedstock is ideal for “ready-to-use” structures due to low mechanical strength, but printed structures serve as useful preforms for further processing, particularly infiltration to enhance strength. Termite frass, with its superior flowability and packing density, is especially promising for such applications.
Supporting Information
S1 Fig.
https://doi.org/10.1371/journal.pone.0246511.s001
(JPG)
S2 Fig.
https://doi.org/10.1371/journal.pone.0246511.s002
(JPG)
S3 Fig.
https://doi.org/10.1371/journal.pone.0246511.s003
(JPG)
S4 Fig.
https://doi.org/10.1371/journal.pone.0246511.s004
(JPG)
S5 Fig.
https://doi.org/10.1371/journal.pone.0246511.s005
(PNG)
S6 Fig.
https://doi.org/10.1371/journal.pone.0246511.s006
(JPG)
S7 Fig.
https://doi.org/10.1371/journal.pone.0246511.s007
(TIF)
S8 Fig.
https://doi.org/10.1371/journal.pone.0246511.s008
(JPG)
S9 Fig.
https://doi.org/10.1371/journal.pone.0246511.s009
(JPG)
S1 File.
https://doi.org/10.1371/journal.pone.0246511.s010
(XLSX)
S2 File.
https://doi.org/10.1371/journal.pone.0246511.s011
(XLS)
S3 File.
https://doi.org/10.1371/journal.pone.0246511.s012
(XLSX)
References
[1] Schellnhuber HJ. Earth system analysis and the second Copernican revolution. Nature. 1999;402(6761 Suppl):C19-23.
[2] Kruth JP. Material incress manufacturing by rapid prototyping techniques. CIRP Annals-Manufacturing Technology. 1991;40(2):603–14.
[3] Deckard CR. Method and apparatus for production of articles by selective sintering of powders. Google Patents; 1989.
[4] Sachs EM, Cima MJ, Cornie JA. Three-dimensional printing: rapid tooling and prototypes directly from a CAD model. CIRP Annals-Manufacturing Technology. 1990;39(1):201–4.
[5] Colombo P, Lee JW, Pantano CG,ρακήτς EM. 3D printing of porous silica structures with hierarchical porosity. Journal of the American Ceramic Society. 2018;101(11):510–9.
[6] Travitzky N, Bonet A, Dermeik B, Scheithauer U, Schwarzer N. Additive Manufacturing of Ceramics: State of the Art and Perspectives. Journal of the American Ceramic Society. 2016;99(3):963–85.
[7] Zocca A, Colombo P, Tucci F, Esposito L. Additive Manufacturing of Ceramics: Materials, Processes and Applications. Materials. 2015;8(10):6979–7026. doi: 10.3390/ma8105351.
[8] Lackner M, Ionescu M. Additive manufacturing of sustainable construction materials with enhanced properties using waste and recycled materials. Journal of Cleaner Production. 2019;217:575–89.
[9] Gašperin L, Nagode M, Vrbek Krajnc M. Sustainable binder jetting of foundry cores using quartz waste sand. Journal of Cleaner Production. 2020;257:120527.
[10] Bader TK, Winkler CW, Gierl-Mayer C, Danner M, Griesser T, Keplinger T, et al. 3D Printing of Bioinspired Cellular Materials Based on Wood. Advanced Materials Technologies. 2018;3(1):1700254.
[11] Campia U, Große-Wöhrmannb B, Smitsb P, Adamopoulosc S, Serranoa E, Wågberga L, et al. 3D printing of lightweight cellular structures based on microfibrillated cellulose and nanoclay. Materials & Design. 2019;180:107948.
[12] Compston P, Cheknane A, Bhattacharjee D, Melchels FP, Blaker JJ, Cartmell SH, et al. Additive manufacture of biodegradable tissue engineering scaffolds from hydroxyethyl cellulose. Biofabrication. 2017;9(2):025015. doi: 10.1088/1758-5090/aa6b8a.
[13] Nam BK, Hopkinson N, Tuck C, Lee JH, Quinn JP, Love AJ. Additive manufacture of high performance cellulose structures. Additive Manufacturing. 2016;12:203–12.
[14] Neiva J, Teixeira-Pinto A, Silva L, Cruz P, Gaspar F, Vilela C, et al. 3D printing lignocellulosic biomass: effect of binder type on mechanical and physical properties. Green Chemistry. 2020;22(9):2893–904.
[15] Gardiner G, Hague R, Ford S. Process optimisation of binder jetting of wood-polymer composites using design of experiments. Virtual and Physical Prototyping. 2017;12(3):213–22.
[16] Konnerth J, Lisjak D, Gorsche M, Moser B, Veigel S, Keplinger T, et al. 3D-printing of Cellulosic Materials. Macromolecular Materials and Engineering. 2018;303(1):1700387.
[17] Qing H, Gao J, Zhang Y, Zhou W, Wu D. Fabrication and properties of wood flour/polylactic acid composites with 3D printing technology. Journal of Materials Science. 2017;52(24):14489–501.
[18] Van den Bulcke A, Van Erps J, Stevens L, Van Puyvelde P, Cardon L. Fused deposition modeling with wood fiber/polypropylene composites. Polymer Composites. 2014;35(3):500–7.
[19] Kumar S, Pegoretti F, Bhowmik S, Simonetti E, Boccaccini AR. Development and characterization of poly (vinyl alcohol) based bioinks for 3D bioprinting: blending with bioactive glass nanoparticles. Materials Science and Engineering: C. 2019;104:109894.
[20] Vora P, Tan J, Zhou W, Zhang Y. Printability and properties of polyvinyl alcohol (PVA) hydrogels for 3D bioprinting. Materials Science and Engineering: C. 2019;103:109778.
[21] Godlinski D, Mielniczek M, Zmudzki J, Fima K, Jachowicz T, Kurzydlowski KJ. Sintering atmosphere influence on microstructure and properties of 3D printed 316L stainless steel. Archives of Metallurgy and Materials. 2018;63(2):659–64.
[22] Singh S, Ramakrishna S. Binder jetting additive manufacturing of porous titanium structures: materials and processes. Materials & Design. 2017;117:198–210.
[23] Kariz M, Sernek M, Obućina M, Kutnar A. Sustainable composites of wood and recycled polypropylene for 3D printing applications. Journal of Cleaner Production. 2019;212:691–8.
[24] Yang B, Zhang Y, Zhang M, Yao F, Zhang Z, Yu L, et al. 3D printing of high-performance cellulose nanocrystal reinforced nanocomposites with a dual-polymer binder system. Journal of Materials Chemistry A. 2018;6(43):21353–62.
[25] Bourell DL, Beaman JJ, Chee Kai C, Leu MC. Additive manufacturing and 3D Printing-A Science and Technology Perspective. Manufacturing Letters. 2017;11:33–5.
[26] Augier S, Tagliavini L, Bellini F, Formentini F, Galletti P, Iovenzi E, et al. Development of a sawdust composite material for fused deposition modeling (FDM). Materials & Design. 2017;114:147–54.
[27] Hossain MSA, Chowdhury MAR, Fazil-E- Rabbi F, Das R, Paul SC, Saha MC. Development of 3D printable composite filament using sawdust and waste high-density polyethylene. Journal of Materials Research and Technology. 2020;9(6):14067–78.
[28] Jiang Y, Yao Y, Fei B, Lu J. Development of a wood flour-filled polypropylene composite filament for fused deposition modeling. Materials & Design. 2017;113:333–42.
[29] Sain M,াফ J, Islam MS, রড J, Farouqui MY. Sustainable Wood Composites for Interior Applications. In: Faruk O, Jawaid M, editors. Sustainable Biobased Composites for Lightweight Structural Applications. Woodhead Publishing; 2013. p. 357–86.
[30] Carr RL. Evaluating flow properties of solids. Chemical Engineering. 1965;72(3):163–8.
[31] Khlebus G, Markl M, Enz J, Wudy K, Klarner F, Wanner A, et al. Gas-assisted recoating in powder-based additive manufacturing. Additive Manufacturing. 2020;36:101390.
