Valorization of bone waste: effect of animal bone origin on hydroxyapatite structure and properties

Abstract

This study demonstrates the valorization of bone waste from different animal sources as a sustainable approach to produce high-value hydroxyapatite (HA) powders, supporting circular economy principles. The findings provide a scientific basis for selecting bone waste sources depending on desired material properties, promoting resource-efficient recovery and reuse of biowaste. Three different types of bone—bovine, ostrich, and porcine—were selected for this research to compare species-dependent differences in HA derived from animal sources. Bovine bone served as a common reference, ostrich bone represented a non-mammalian source, and porcine bone was chosen for its close structural similarity to human bone. The HA powders were characterized in terms of particle size, specific surface area, crystallite size, phase composition, and porosity. X-ray diffraction (XRD) analysis revealed variations in crystallite size with calcination temperature. Mechanical testing revealed that bovine-derived HA exhibited the highest compressive strength (17 MPa) and porcine-derived HA showed the highest hardness (0.5 GPa). These findings highlight the significant influence of the bone source on the microstructural and physicochemical properties of HA, providing a foundation for selecting optimal HA sources for targeted applications. With the results obtained in this paper, it is possible to select the animal of origin of the bones to be used based on the desired characteristics of the powder to be developed.

Introduction

The global meat and food processing industries generate over 130 million metric tons of waste animal bones (WABs) annually (Nassar et al. 2021), creating pressing environmental challenges. As the world population continues to grow, the accumulation of waste bones leads to challenges in terms of landfill usage and potential health risks. In the context of a circular economy, valorizing WABs into high-value materials offers a sustainable route for waste minimization and resource recovery. The utilization of waste bones is governed by the European Union Directives, which classify different types of waste into specific categories. The meat-bone meal and all elements of the body of animals suspected of a possibility of bovine spongiform encephalopathy disease belong to this category, and the only way of utilization is combustion at a temperature of a minimum of 850 °C (Staroń et al. 2010).

Compositionally, animal bones consist of organic components (primarily collagen and other proteins) accounting for 30–35% of their composition, while the inorganic components make up 65–70% (Malla et al. 2020). WABs are a natural apatite-rich material, predominantly composed of HA with a chemical formula of Ca₁₀(PO₄)₆(OH)₂. While chemical synthesis of HA remains costly and resource-intensive (Mohd Pu’ad et al. 2020), thermal treatment of WABs offers an environmentally friendly alternative, producing high-purity HA through pyrolysis at ~ 950 °C (Barakat et al. 2009; Khoo et al. 2015).

Wet chemical methods such as electrochemical deposition, chemical precipitation, and sol–gel techniques generally yield nanosized HA particles with high surface area and good bioactivity but may result in lower crystallinity due to moderate processing temperatures. These methods also involve considerable chemical consumption, particularly of calcium and phosphate precursors, and may require subsequent calcination to improve crystallinity. Hydrothermal and sonochemical methods produce well-crystallized, stoichiometric HA with controlled morphology; however, they typically demand higher energy input because of elevated pressure or the use of specialized equipment.

In contrast, dry and high-temperature routes, such as solid-state reaction and combustion synthesis, result in highly crystalline HA with fewer impurities but tend to produce larger, agglomerated particles. Although chemical consumption in these processes is relatively low, the energy requirements are substantial.

Different chemical methods of HA synthesis are presented in Table 1.

Table 1.

Summary of HA synthesis methods highlighting sustainability metrics

Method	Sub method	Description	Energy consumption	Chemical consumption	Ref
Wet	Electrochemical	Deposition via current in solution	Moderate (depends on current and time)	High (electrolytes, Ca/P salts)	(Montero et al. 2006; Djošić et al. 2008; Nur et al. 2014)
	Chemical precipitation	Precipitation from Ca and P solutions	Low to moderate	High (e.g., Ca(NO₃)₂, (NH₄)₂HPO₄)	(Yelten-Yilmaz and Yilmaz 2018)
	Hydrothermal	Reaction in autoclave at high P/T	High (energy for pressure and heat)	Moderate	(Ortiz et al. 2017; Mohd Pu’ad et al. 2020)
	Sonochemical	Uses ultrasound for mixing and reaction	Moderate (ultrasound equipment)	Moderate	(Mohebali and Shokuhfar 2011)
	Sol–Gel	Transition from sol to gel state	Moderate (drying and calcination)	High (solvents like ethanol, alkoxides)	(Haider et al. 2017)
Dry	Solid state	High-temp reaction of powders	Very high (T > 1000 °C)	Moderate	(Haider et al. 2017)
	Mechanochemical	Ball milling reaction without solvents	Moderate (long milling times)	Low	(Nasiri-Tabrizi et al. 2009; Mohd Pu’ad et al. 2019)
High temperature	Combustion	Self-sustaining exothermic reaction	High (initial ignition energy)	Moderate (fuels, oxidizers)	(Haider et al. 2017; Mohd Pu’ad et al. 2019)
	Pyrolysis	Thermal decomposition of animal bone	Moderate to High (850–950 °C)	None	(Cho and Rhee 2012; Mohd Pu’ad et al. 2019)

Recent attention has focused on the use of HA from animal bone waste in biomedical applications, but its production also demonstrates a promising model for sustainable material recovery. In addition to offsetting synthetic HA production, this method diverts significant biowaste from traditional disposal routes such as landfilling or combustion, thereby contributing to waste prevention goals (Zhao et al. 2021; Fendi et al. 2024).

While bone grafting in orthopedics and dentistry remains a main application area for HA (Zhao et al. 2021), in recent years, HA finds its application in other biomedical fields such as a coating material for metallic implants and a component in drug delivery systems (Lara-Ochoa et al. 2021). However, HA nanomaterials also find their application in electrochemistry, particularly in sensors production because of their large surface area. For example, Hertati et al. demonstrated the use of a HA-modified glassy carbon electrode for the determination of cysteine (Vladislavić et al. 2021)․ General information about different fields of HA application is presented in Table 2.

Table 2.

General fields of HA application

Applications	Role	Ref
Bone Grafting/Bone Substitutes	Bone tissue repair and reconstruction due to its osteoconductive, biocompatible, and non-inflammatory properties	(Rstakyan et al. 2023; Fendi et al. 2024; Mkhitaryan et al. 2025a, 2025b)
Biomedical Coatings	HA coatings on medical devices to improve biocompatibility and reduce wear	(Goodman et al. 2013; Prakash and Uddin 2017)
Biosensors	HA’s conductivity and catalytic properties find applications in electrochemical sensors and energy storage devices	(Hartati et al. 2022; Adusei et al. 2023)
Drug Delivery Systems	HA nanoparticles are explored as drug delivery carriers due to their controlled release capabilities	(Moslemi et al. 2015; Mondal et al. 2018)
Cosmetic/Aesthetic Applications	HA is employed in dermal fillers and implants in cosmetic procedures to enhance facial features	(Carella et al. 2021)
Water Treatment	The adsorption properties of HA make it useful in water filtration systems to remove contaminants	(Koliyabandara et al. 2022)
Corrosion Protection	HA coatings provide corrosion resistance to metal surfaces	(Sonmez et al. 2012; Raj and Mumjitha 2015)

While bovine bones are commonly studied, comparative assessments of HA derived from different animal sources remain scarce. Understanding the influence of animal origin on HA yield, purity, and ceramic behavior is essential for developing efficient, scalable, and adaptable recovery strategies across diverse waste streams.

The objective of this study is to develop HA powders from bone waste of different animal origins (cow, porcine, and ostrich) and to evaluate their potential as sustainable materials for circular resource utilization. A key aim is to compare the physicochemical and ceramic characteristics of the resulting HA powders based on the animal source. The powders developed in this work are primarily intended for use in biomedical coatings and bone regeneration systems, though their characteristics may also make them suitable for other advanced applications like water treatment or corrosion protection. Additionally, the study investigates the ceramic behavior of these powders to determine whether the origin significantly affects their performance. Particular emphasis is placed on identifying optimal processing conditions for the efficient production of bone-derived HA.

Materials and methods

Preparation of HA powder

Bovine, porcine, and ostrich thigh bones were selected as the primary raw materials for this study. Bovine and porcine bones were chosen for their similarity to human bones in terms of weight, density, and size. Ostrich bones were included to investigate HA derived from avian sources. All bones were collected from domestic animal farms.

The bones were initially cleaned with a sharp knife to remove residual muscle tissue and visible impurities. The cleaned samples were thoroughly washed several times with distilled water and then immersed in acetone for 3 h to remove residual fat. The bones were subsequently boiled in a sealed container for 6 h, with the boiling water replaced every 2 h. Following boiling, the bones were cut into 2–4 cm pieces and dried at 250 °C for 4 h using an oven (WGLL-85BE).

The dried bones were then thermally treated in a furnace (Atmosphere Muffle Furnace GCF 1400, Across International) at 700 °C, 1100 °C, and 1350 °C for 5 h with a heating rate of 5 °C/min, resulting in white calcined bones.

The calcined bones were milled using a BKBM-V2 Vertical Planetary Ball Mill in four stages. Initially, milling was performed with a ball-to-powder mass ratio of 7:1 without the addition of calcium stearate. Subsequently, 1 wt% calcium stearate was added, and milling continued with ball-to-powder mass ratios of 7:1 and 10:1 at 240 rpm for 2 h per stage. In the first milling stage, zirconia balls of 20 mm and 15 mm diameter were used, followed by 10 mm balls in the second stage, 8 mm balls in the third stage, and 5 mm balls in the final stage. After each milling stage, the powder was sifted using sieves with mesh sizes of 1000 µm, 105 µm, 74 µm, and 44 µm, respectively.

Processing of powders

The resulting HA powders were mixed with 1 wt% silicon oil using a mortar to serve as a lubricant. The mixture was thoroughly homogenized before being introduced in pressing dies. The samples were axially pressed (TMAX-ZYP) at pressures of 46 MPa, 93 MPa, and 139 MPa to form rectangular prisms (27 mm × 8 mm × 4 mm) and cylinders (11 mm × 11.5 mm). The pressed samples were sintered at 1200 °C, 1230 °C, 1250 °C, and 1270 °C. The heating and cooling rates were 5 °C/min and the holding time at the maximum temperature was 2 h.

Characterization

Microstructural characterization was performed by scanning electron microscopy (SEM, Prisma E, Thermo Fisher Scientific, and Hillsboro, OR, USA and Zeiss, Evo 10). Samples were coated by sputtering with a 20 nm layer of Au to provide enough conductivity to the sample.

The Archimedes densities was measured by densitometer (HR-100AZ, 102/0.1 mg). Also, geometrical density was determined.

Thermogravimetric analysis was performed by devise STA 449 F3 Jupiter with SKIMMER Furnace (NETZSCH, Germany).

Phase characterization was accomplished by X-ray diffraction (Rigaku MiniFlex 600 X-ray diffractometer, 40 mA, 40 kV, CuKα radiation, λ = 0.1542 nm, step size of 0.02°).

Chemical analysis was carried out by X-ray Fluorescence (XRF) using MagiX from Philips Co.

The chemical compound analysis of bones was performed by FTIR Spectrometer (Spectrum Two, PerkinElmer).

Powder particle size analysis (BET Surface Area (m²/g), BJH Adsorption Cumulative Surface Area of pores (m²/g), BJH Adsorption Cumulative Pore Volume (cm³/g), BJH Adsorption Average Pore (nm)) was measured by laser scattering using a Mastersizer 3000 equipment (Malvern Panalytical B.V., Almelo, and The Netherlands).

Dilatometry was performed using a horizontal dilatometer, DIL L75 HX 1600 (LINSEIS, Germany).

Compressive strength of the samples was measured using a universal mechanical testing machine (Instron Legend 5965 testing system, USA), equipped with a 5 kN load cell, with a crosshead speed of 20 mm/min at room temperature.

Microhardness was measured using a FALCON 600G2FA (INNOVATEST Europe BV, Maastricht, Netherlands).

Relative density ${\rho }_r$ was measured by Eq. (1):

$${\rho }_r=\frac{{\rho }_b}{{\rho }_t}\ast 100\%$$ 1

where ${\rho }_b$ is the bulk density of the HA sintered samples, which was calculated using the geometric method, and ${\rho }_t$ stands for the theoretical density of HA (3.16 g/cm³).

Total porosity (P) was determined using an Eq. (2).

$$P=\left(1,-,{\rho }_r\right)\ast 100\%$$ 2

The volume shrinkage of HA samples was calculated by Eq. (3).

$$\text{Volume Shrinkage}\left(\text{\%}\right)=\frac{V_i-V_f}{V_i}\ast 100\%$$ 3

where $V_i$ is the original volume of the samples before sintering, and $V_f$ is the final volume after sintering.

Density, microhardness, and compressive strength measurements were carried out on three samples, and the average values were reported. In this case, the level of uncertainty ± 10%.

Results and discussion

Morphology of raw bones

The SEM micrographs (Fig. 1) reveal the microstructural characteristics of bovine, ostrich, and porcine raw bone samples, highlighting the differences in their vascular architecture. The observed pores correspond to cross-sections of Haversian and Volkmann’s canals, essential components of the bone’s vascular system (Da Fontoura Costa et al. 2006; Kim et al. 2015).

Fig. 1.

Microstructure of bovine (a), ostrich (b), porcine (c) raw bone samples and comparison of geometrical and Archimedes’ densities of bovine, ostrich, and porcine bones (d)

In the bovine bone (Fig. 1a), large, well-defined Haversian canals are evident, indicative of a dense structure typical of large mammalian cortical bone. The presence of microcracks surrounding some of the canals was possibly introduced during sample preparation. The ostrich bone (Fig. 1b) exhibits a different structural pattern, with fewer and less pronounced vascular channels. However, in flightless birds like ostriches, bones tend to be denser compared to other birds. Even so, their bone microstructure often differs from that of mammals, showing fewer Haversian canals and a higher degree of primary bone tissue (Conti et al. 2023). Microscopic observations reveal that the pores in ostrich bone are smaller in diameter and the total pore volume appears reduced, likely due to its thinner cortical layer, which may provide greater mechanical resistance. However, this does not imply the presence of fewer pores overall; rather, the pores are smaller and less interconnected. In contrast, the porcine bone shows a higher density (Fig. 1c, d) of evenly distributed vascular channels. This suggests a more uniform porosity. Due to the low interconnection between the pores, it is not possible to evaluate porosity by nitrogen absorption or mercury porosimetry.

TGA and DSC analysis of raw bones

The TGA and DSC analysis for all three animal bones is represented in Fig. 2. The TGA curves (Fig. 2a) exhibit distinct weight loss stages corresponding to the removal of water, the decomposition of organic matter (collagen and proteins), and the transformation of bone mineral phases. All three bone types show similar decomposition trends, with minor variations in weight loss percentages and temperature ranges.

Fig. 2.

Comparison of the TG (a) and DSC (b) analysis of raw bones

• First Stage (50–200 °C): A minor mass loss (~ 2–5%) is observed in all samples, attributed to the evaporation of water. The DSC curves (Fig. 2b) confirm this event by displaying a broad endothermic peak in the same range, indicating water loss and dehydration. The intensity of this peak varies based on moisture content.

• Second Stage (200–350 °C): A significant weight loss (~ 20–30%) occurs due to the thermal degradation of organic components such as collagen. The maximum decomposition temperatures differ slightly among species, with porcine bone showing the most rapid degradation, followed by bovine and ostrich bones.

• Third Stage (350–450 °C): Well-defined effects are visible mainly for bovine and porcine bones. This can be explained by the degree of cross-linking in mammalian collagen, which significantly influences its thermal stability. Studies have shown that increased cross-linking elevates the denaturation temperature of collagen (Schroepfer and Meyer 2017). For instance, research on bovine collagen demonstrated that cross-linking enhances thermal stability, as evidenced by higher denaturation temperatures in DSC thermograms.

• Fourth Stage (600–900 °C): A gradual weight loss (~ 5–10%) is observed, associated with the partial decomposition of carbonate groups in the HA structure. Ostrich bone exhibits a broader endothermic transition around 700–800 °C, suggesting a unique carbonate substitution pattern in HA.

These thermal characteristics are particularly relevant for the processing and application of HA. High thermal stability and well-defined crystallization behavior ensure that the material can withstand sintering or coating processes without structural degradation. Moreover, controlled removal of organic residues enhances purity, which is critical for biomedical use, while stable HA crystallinity contributes to improved mechanical integrity and bioactivity in implant coatings and scaffolds.

XRD analysis of raw and calcined bones

The phase evolution and crystallinity of HA derived from bovine, ostrich, and porcine bones were investigated after calcination at 700 °C, 1100 °C, and 1350 °C (Fig. 3). The XRD pattern of raw bones shows broad and poorly defined peaks, indicating the presence of low-crystalline apatite combined with organic phases. The amorphous background arises from organic components such as collagen and lipids, which mask the HA reflections. After calcination at 700 °C, characteristic HA peaks appear at 2θ ≈ 26°, 32°, 39°, and 49°, corresponding to the (002), (211), (310), and (213) planes of HA (JCPDS 09–0432). The diffraction peaks become more pronounced, confirming the formation of crystalline HA as the dominant phase. With further calcination at 1100 °C, sharper and more intense HA reflections are observed, indicating enhanced crystallinity and grain growth. Bovine and porcine bones exhibit well-developed HA structures, whereas the ostrich-derived HA shows slightly broader peaks, suggesting the persistence of carbonate substitution, which inhibits complete crystallization. Heat treatment at 1350 °C promotes coarsening and partial decomposition of HA into β-tricalcium phosphate (β-TCP), evident from additional peaks at 2θ ≈ 27° and 31° (Ou et al. 2013). The coexistence of HA and β-TCP phases confirms the thermal instability of HA above 1300 °C. Crystallite size analysis (Fig. 4) supports these observations, showing a significant increase in crystallite size between 700 and 1100 °C, followed by a slight reduction at 1350 °C due to partial phase decomposition. Among all samples, bovine HA consistently exhibits the largest crystallite size, while ostrich and porcine HA maintain smaller grain sizes and lower crystallinity.

Fig. 3.

XRD patterns of raw (a) and calcined bones at 700 °C (b), 1100 °C (c), and 1350 °C (d)

Fig. 4.

Crystallite size (± 5%) of HA derived from bovine, ostrich, and porcine bones calcined at 700 °C, 1100 °C, and 1350 °C

Morphology of calcined bones

The SEM images (Fig. 5) of bovine, ostrich, and porcine bones subjected to calcination at 700 °C and 1100 °C reveal the grain growth behavior and its correlation with crystallinity. At 700 °C (Fig. 5a–c), the microstructures exhibit fine grains with significant porosity, indicative of incomplete densification. This is consistent with the XRD results, where the samples still display relatively low crystallinity, as shown in Fig. 4. Ostrich bone (Fig. 5b) presents a more homogeneous grain structure with smaller grains compared to bovine (Fig. 5a) and porcine (Fig. 5c), aligning with the observation that ostrich bone retains broader XRD peaks and slightly lower crystallinity at this temperature. The higher carbonate content in ostrich bone, which decomposes at lower temperatures, likely affects calcination behavior, leading to finer grains.

Fig. 5.

Microstructure of calcined bones at 700 °C: bovine (a), ostrich (b), porcine (c); and at 1100 °C: bovine (d), ostrich (e), porcine (f)

At 1100 °C, the microstructures show significant grain growth with reduced porosity, particularly in bovine and porcine samples, which corresponds to their higher crystallinity values (Fig. 4).

Overall, the calcination temperature strongly influences grain growth, with higher temperatures leading to densification. Based on this result, 700 °C was chosen as the temperature for the calcination of bones.

XRF analysis of bones calcined at 700 °C

XRF analysis was performed to determine the elemental composition of calcined bone powders and to evaluate their suitability as precursors for HA synthesis. Comparing animal-derived bones with human bone allows assessment of how closely these powders mimic the mineral composition of human bone, which is critical for biomedical applications such as bone grafts and implants.

Table 3 presents the XRF results for bovine, porcine, and ostrich bones, alongside literature-reported ranges for human bone. The predominant oxides in all samples are calcium oxide (CaO) and phosphorus pentoxide (P₂O₅), which form the main mineral phases of biological apatite. Bovine bones exhibited the highest CaO content (54%), followed by ostrich (54%) and porcine (53%), while P₂O₅ content was 42%, 40%, and 41%, respectively. These values fall within the ranges reported for human bone (CaO, 52–60%; P₂O₅, 33–41%), and the Ca/P ratios (≈1.67) are consistent with stoichiometric hydroxyapatite.

Table 3.

XRF analysis of bovine, porcine, and ostrich bones calcined at 700 °C

Oxide	Bovine	Porcine	Ostrich
CaO	54	53	54
P2O5	42	41	40
MgO	1.26	1.5	1.2
Na2O	1.05	0.95	1.0
K2O	< 0.05	< 0.10	< 0.10
SiO2	< 0.02	< 0.6	< 0.2
Al2O3	0.1	0.08	0.08
Fe2O3	< 0.10	< 0.10	< 0.10
SO3	—	—	—
Cl (as Cl−)	—	—	—
Cal. Loss	2.9	2.9	3.0

The comparison demonstrates that animal bones provide a mineral composition closely resembling human bone, justifying their use as HA precursors. Additionally, the XRF analysis allows evaluation of the effects of calcination and milling. Calcination at 700 °C removes organic matter and facilitates grinding, increasing crystallite size and reducing reactivity. Milling conditions influence particle size distribution, surface area, and overall reactivity of the powders.

ATR-FTIR analysis of raw and calcined bones

The ATR-FTIR spectra (Fig. 6a–c) of raw and calcined bones at 700 °C and 1350 °C show the same major functional groups (OH¹⁻, CO₃²⁻, and PO₄³⁻), indicating the presence of similar structural groups dominated by HA. In all spectra, calcined samples (Fig. 6b, c) exhibit a sharp band around 3500–3600 cm⁻¹, indicating the presence of structural hydroxyl groups (OH⁻). These bands are less pronounced or absent in raw samples (Fig. 6a), likely due to interference from organic components. The calcined samples of all bones show bands around 1400–1500 cm⁻¹ corresponding to carbonate (CO₃²⁻) stretching vibrations. These peaks are weaker or less defined in raw bones, reflecting the partial incorporation of carbonate in the crystalline HA structure during calcination or after remaining in air. The strong, characteristic peaks for phosphate vibrations are observed in the region of 1000–1100 cm⁻¹ (PO₄³⁻) for both raw and calcined samples across all bone types (Table 4).

Fig. 6.

ATR-FTIR spectra of bovine, ostrich, and porcine bones: raw (a), calcined at 700 °C (b), and 1350 °C (c), and comparison of inorganic content (d)

Table 4.

ATR-FTIR band assignments and functional group identification for bone samples

Sample type	Calcination temperature (°C)	Main FTIR bands (cm⁻1)	Assigned functional groups	Observations
Bovine	Raw	1035, 566, 1410	PO43⁻, CO32⁻	Clear phosphate and carbonate peaks; organic signals present
	700	3570, 1030, 603	OH⁻, PO43⁻	Strong phosphate, removal of organic phase
	1350	3572, 1040, 608	OH⁻, PO43⁻	Partial β-TCP formation, good crystallinity
Ostrich	Raw	1032, 1420	PO43⁻, CO32⁻	Higher carbonate presence, broader peaks
	700	3570, 1415, 1045	OH⁻, CO32⁻, PO43⁻	More pronounced carbonate peaks
	1350	3573, 1040, 608	OH⁻, PO43⁻	Partial β-TCP, reduced HA intensity
Porcine	Raw	1035, 1410	PO43⁻, CO32⁻	Stronger phosphate, compact structure
	700	3572, 1040	OH⁻, PO43⁻	Well-defined HA bands
	1350	3570, 1045	OH⁻, PO43⁻	Stable HA, minor β-TCP phase

The ostrich bone (Fig. 6b red curve) shows more pronounced carbonate (CO₃²⁻) peaks after calcination compared to the other two bones. In Fig. c, slightly different vibration modes are visible, which can be explained by the presence of a secondary phase of β-TCP in the samples sintered at 1350 °C.

Figure 6d represents the comparison of inorganic content of bovine, ostrich, and porcine raw bones based on TGA and calcination results. Bovine and porcine bones exhibit the highest inorganic content, indicating a higher proportion of HA relative to organic matter. Ostrich bones, adapted for a lightweight skeletal structure to support flightless movement, tend to have lower mineral density and a relatively higher organic content. Additionally, the mineral phase in ostrich bones has a higher carbonate substitution, influencing thermal decomposition behavior. The increased carbonate content in HA makes it more susceptible to decomposition at lower temperatures, resulting in a slightly lower final inorganic content after calcination (Barralet et al. 2002)․

Milling of calcined bones

The initial milling of bovine, porcine, and ostrich calcined bones was performed without the use of any additives. However, this approach resulted in the sticking of HA particles to the inner surfaces of the milling jar, significantly obstructing the process and reducing the yield of the final milled powder (Table 5 (Powder 1)).

Table 5.

Bone milling parameters for powder preparation

Powder	Composition (wt%)	Milling cycle (direct/quiet/reverse, min)	Speed (rpm)	Ball-to-powder (wt)	Ball size (mm)	Sieve size (µm)	Sieved powder yield (%)
							Bovine	Porcine	Ostrich
1	HA 100	60/10/60	240	7:1	20/15	1000	91	89	88
					10	105	80	76	77
					8	74	92	73	75
					5	44	80	78	79
2	HA: CaSt 99: 1				20/15	1000	100	98	97
					10	105	90	89	88
					8	74	95	85	86
					5	44	90	88	85
3				10:1	20/15	1000	100	95	100
					10	105	93	92	97
					8	74	97	88	97
					5	44	90	90	95

To address this challenge, calcium stearate (CaSt) was incorporated into the milling process. CaSt acts as a milling aid, specifically functioning as a lubricant and dispersant. Its hydrophobic properties reduce surface energy, which helps minimize particle-particle interactions, thereby reducing the tendency of particles to agglomerate and stick to the jar walls (Deepika and Chen 2014). The milling was performed with ball-to-powder mass ratios of 7:1 and 10:1 (Table 5 (Powders 2 and 3)). After each milling cycle, the powders were sequentially sieved through 1000, 105, 74, and 44 µm meshes to separate coarse particles. Following each sieving step, milling was continued using smaller balls to achieve finer particle sizes. The highest yield was obtained for the 10:1 ball-to-powder ratio in the presence of CaSt (Powder 3).

The influence of ball diameter on powder refinement was evaluated. Smaller milling balls (e.g., 5 mm) promoted a finer particle size distribution due to higher collision frequency, while larger balls (e.g., 10 mm) facilitated more effective breaking of larger agglomerates. The optimized combination of mixed ball sizes improved homogenization and resulted in a narrower particle size distribution, which was later reflected in the improved packing density and sinterability of the HA powders. The ease of milling bones from different sources is influenced by their microstructure and mineral content. Porcine bones are generally more challenging to mill due to their high density and compact microstructure. In contrast, bovine bones exhibit a moderate balance of density and porosity, which makes them easier to mill compared to porcine bones. Ostrich bones, on the other hand, are the easiest to mill due to their unique structure and composition.

Powder structure, particle size distribution, and morphology

XRD analysis of powders

From XRD pattern (Fig. 7), it is obvious that the milling process appears to reduce the crystallinity of the bone powders, making it more amorphous. Bone powder sieved under 1000 µm (Fig. 3b) shows sharper and more intense diffraction peaks, indicating a higher degree of crystallinity. After milling and sieving to 44 µm (Fig. 7), the overall peak intensity decreases, suggesting a partial loss of long-range atomic order. The peaks in the 44-µm sample appear broader compared to the 1000-µm sample. This broadening is typically associated with a reduction in crystallite size (Fig. 7) and an increase in microstrain, which often occurs due to mechanical milling.

Fig. 7.

XRD patterns of 44-µm sieved bone powders and comparison of crystallinity

Particle size distribution, specific surface area, and morphological analysis of 44 µm sieved powders

The particle size distributions of bovine, ostrich, and porcine bone powders sifted with 44 µm are presented in Fig. 8a–c; it is clear that in all cases, the powders have a bimodal characteristic.

Fig. 8.

Particle size distribution of 44-µm sieved bone powders: bovine (a), ostrich (b), porcine (c); and corresponding microstructure of: bovine (d), ostrich (e), porcine (f)

Ostrich bone powder clearly exhibits the finest particle size distribution, with the peak of the frequency curve located in the smallest diameter range. In contrast, porcine bone powder shows the largest particle sizes. This may be due to its higher structural resistance, as it results in a broader particle size distribution under the same milling conditions. SEM images reveal irregularly shaped particles with sharp edges, indicating brittle fracture during grinding (Fig. 8d–f).

From Table 6, it is noticeable that the porcine sample has the largest particle size and the highest BET surface area, which may seem contradictory. However, this can be explained by its higher internal porosity, as indicated by the greater pore surface area and pore volume.

Table 6.

Particle size, surface area and porous structure characteristics of HA powders derived from bovine, ostrich, and porcine bones

Sample name	Particle Size (µm)	BET Surface Area (m2/g)	BJH Adsorption Cumulative Surface Area of pores (m2/g)	BJH Adsorption Cumulative Pore Volume (cm3/g)	BJH Adsorption Average Pore (nm)
Bovine	1.9	4.0	3.5	0.018	21
Ostrich	1.5	4.1	3.7	0.022	24
Porcine	3.3	5.1	4.6	0.023	20

The ostrich sample has the smallest particle size maintaining a relatively high BET surface area and the largest average pore size. This indicates that ostrich-derived HA consists of finer particles with well-developed porosity, which can be beneficial for applications requiring enhanced surface interactions. The bovine sample exhibits intermediate characteristics, with a moderate particle size and lower surface area compared to the ostrich and porcine samples.

Pressing and sintering

Following the milling process, the obtained HA powders sieved with 44 µm (Powder 3) were subjected to pressing to form dense, uniform compacts. Initial pressing trials were conducted at pressures of 46 MPa, 93 MPa, and 139 MPa, resulting in significant issues such as delamination and the formation of structural defects within the compacts, especially at lower pressures. This delamination likely occurred due to insufficient particle bonding and uneven distribution of compressive forces.

To address these issues, silicone oil was used as a lubricant, providing a low-friction environment that minimized particle agglomeration and allowed for better particle rearrangement during compaction. Before sintering, the obtained density values indicate that the best results were achieved in the samples pressed at 139 MPa (Fig. 9a). For this purpose, the dilatometry results of the aforementioned samples were studied to determine the sintering parameters (Fig. 9b). Based on the dilatometry results, sintering temperatures of 1200 °C, 1230 °C, 1250 °C, and 1270 °C were selected. These temperatures correspond to critical points in the densification process. At 1200 °C, rapid densification begins, initiating significant structural rearrangements. The temperatures of 1230 °C and 1250 °C represent intermediate stages where sintering progresses, allowing for the evaluation of phase evolution and microstructural changes. Finally, 1270 °C represents a near-final sintering stage, at which the material is expected to achieve its highest density and mechanical strength.

Fig. 9.

Theoretical densities (a), dilatometry curves (b), geometrical (c) and Archimedes (d) densities, and shrinkage (e) of pressed sintered samples

From Fig. 9c and d, it is evident that better densification was achieved at 1250 °C and 1270 °C. Additionally, Fig. 9e shows significant shrinkage at 1270 °C, which is consistent with the dilatometry results (Fig. 9b). However, according to the XRD results, at 1270 °C (Fig. 10a, b), a secondary phase of β-TCP was detected, indicating partial decomposition of HA. Therefore, the optimal sintering temperature was determined to be 1250 °C (Fig. 9c, d, e), where high densification was achieved without significant phase degradation. The density at 1250 °C reached approximately 2.9 g/cm³, which is notable as it approaches the theoretical density of HA (3.16 g/cm³). This level of densification indicates efficient particle packing and strong intergranular bonding, contributing to improved mechanical properties.

Fig. 10.

XRD patterns of sintered samples at 1250 °C (a) and 1270 °C (b)

Microstructural analysis of sintered samples at 1250 °C

Figure 11a–c presents the microstructural analysis of the pressed samples. The sample derived from bovine and porcine bone powder displays a smooth surface with minimal porosity which is aligned with the results presented on Fig. 9, indicating a dense and compact structure. In contrast, the ostrich sample (Fig. 11b) shows a highly porous surface, which could enhance bioactivity but may compromise mechanical strength.

Fig. 11.

Microstructure of pressed samples (139 MPa) sintered at 1250 °C, derived from bovine (a), ostrich (b), and porcine (c) bone powders

Mechanical properties of sintered samples at 1250°C

Table 7 summarizes the mechanical properties—specifically porosity, compressive strength, and hardness—of sintered cylindrical samples. The differences in the mechanical performance can be attributed to the inherent microstructural characteristics and porosity of the samples.

Table 7.

Mechanical properties of sintered samples after pressing

Animal	Cylinder 11 × 11.5 mm (Ø × H)
	Porosity (%)	Compressive test (MPa)	Hardness (GPa)
Bovine	4.1 ± 0.2	17 ± 0.5	0.4 ± 0.02
Ostrich	7.1 ± 0.3	10 ± 0.25	0.3 ± 0.01
Porcine	3.3 ± 0.1	13 ± 0.4	0.5 ± 0.02

Among the three, bovine and porcine-derived samples exhibited the highest compressive strength. The low porosity may have contributed to the higher hardness, indicating a denser material. In contrast, ostrich-derived samples had the highest porosity and the lowest mechanical strength and hardness. The increased porosity likely resulted in a more fragile and less compact structure. Despite having larger initial particle sizes, the porcine powder demonstrated a high internal porosity and surface area before sintering, which likely facilitated enhanced particle rearrangement and diffusion during densification. This efficient packing and strong intergranular bonding resulted in improved density, smaller porosity, and higher hardness compared to the other samples.

This combination of mechanical properties could offer a balanced performance between strength and durability, making bovine and porcine-derived ceramics versatile for both structural and biomedical applications.

Overall, the mechanical behavior of these samples aligns with typical trends where lower porosity correlates with higher mechanical strength and hardness.

Conclusions

This study demonstrates that the origin of animal bone affects the structural, thermal, and mechanical properties of HA derived from waste bones.

Bovine and porcine bones produced HA with higher density, lower porosity, whereas ostrich-derived HA exhibited higher porosity and greater carbonate substitution, leading to lower crystallinity and reduced mechanical performance.

Optimal processing conditions—calcination at 700 °C, pressing at 139 MPa with lubrication, and sintering at 1250 °C—allowed the production of dense, mechanically robust HA powders without significant phase degradation.

These findings highlight the importance of selecting bone sources based on the desired powder characteristics and illustrate a sustainable pathway for valorizing animal bone waste.

By transforming an abundant biological by-product into high-value bioceramics, this work supports circular economy principles and offers practical guidance for developing environmentally friendly HA for biomedical and industrial applications.

Author contribution

Liana Mkhitaryan: formal analysis, investigation, visualization, writing—original draft. Lilit Bagdasaryan: investigation, formal analysis, writing—original draft. Khachik Nazaretyan: investigation, formal analysis. Zhenya Khachatryan: investigation, formal analysis. Arevik Khachatryan: investigation, formal analysis. Mikayel Torosyan: investigation, formal analysis. Marina Aghayan: supervision, project administration, funding acquisition. Miguel A. Rodríguez: writing—review and editing, supervision, investigation, conceptualization. Viktorya Rstakyan: conceptualization, funding acquisition, project administration, supervision, writing—original draft, writing—review and editing.

Funding

This work was supported by the Higher Education and Science Committee Ministry of Education, Science, Culture and Sport RA [grant number No. 22rl-050, 2022].

Data availability

The data supporting the findings of this study are available within the paper.

Declarations

Ethical approval

This is not applicable for that specific section.

Consent to participate

This is not applicable for that specific section.

Consent for publication

This is not applicable for that specific section.

Competing interests

The authors declare no competing interests.