Materials Today Bio 5 (2020) 100032

Contents lists available at ScienceDirect

Materials Today Bio

journal homepage: www.journals.elsevier.com/materials-today-bio

Ceramics with the signature of wood: a mechanical insight

D. Bigoni a, *, R. Cavuoto a, D. Misseroni a, M. Paggi b, A. Ruffini c, S. Sprio c, A. Tampieri c
a
DICAM, University of Trento, Via Mesiano 77, Trento, Italy
b
IMT School for Advanced Studies Lucca, Italy
c
CNR-ISTEC, Via Granarolo 64, Faenza, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: In an attempt to mimic the outstanding mechanical properties of wood and bone, a 3D heterogeneous chemistry
Hydroxyapatite approach has been used in a biomorphic transformation process (in which sintering is avoided) to fabricate ce-
Mechanical properties ramics from rattan wood, preserving its hierarchical fibrous microstructure. The resulting material (called
Strength
biomorphic apatite [BA] henceforth) possesses a highly bioactive composition and is characterised by a multiscale
Fracture
Mechanical tests
hierarchical pore structure, based on nanotwinned hydroxyapatite lamellae, which is shown to display a lacunar
Fractal porosity fractal nature. The mechanical properties of BA are found to be exceptional (when compared with usual porous
hydroxyapatite and other ceramics obtained from wood through sintering) and unique as they occupy a zone in
the Ashby map previously free from ceramics, but not far from wood and bone. Mechanical tests show the
following: (i) the strength in tension may exceed that in compression, (ii) failure in compression involves complex
exfoliation patterns, thus resulting in high toughness, (iii) unlike in sintered porous hydroxyapatite, fracture does
not occur ‘instantaneously,’ but its growth may be observed, and it exhibits tortuous patterns that follow the
original fibrillar structure of wood, thus yielding outstanding toughness, (iv) the anisotropy of the elastic stiffness
and strength show unprecedented values when situations of stresses parallel and orthogonal to the main channels
are compared. Despite being a ceramic material, BA displays a mechanical behavior similar on the one hand to the
ligneous material from which it was produced (therefore behaving as a ‘ceramic with the signature of wood’) and
on the other hand to the cortical/spongy osseous complex constituting the structure of compact bone.

1. Introduction In this vein, ceramics implementing a microstructure and nano-

structure almost identical to wood can be easily believed to exhibit
Wood and bone are natural materials displaying a similar and excep- unprecedented mechanical performances and also to be ideal candi-
tionally performing mechanical behavior, which is rooted in their highly dates for bone replacement. Following these ideas, biotemplating pro-
hierarchical and fibrous microstructure [1,2]. Therefore, mimicking of cesses to fabricate 3D functional ceramics (such as apatites) from
this microstructure to produce superior materials has been the objective natural wood have been proposed, all terminating with sintering [5,7,8,
of a massive research investment [3–26] (see Figs. 23 and 24). 10]. Recently, Tampieri et al. [27] developed a new procedure1 to

* Corresponding author.
E-mail address: [email protected] (D. Bigoni).
1
Even if the details of the process are reported in the studies by Tampieri et al [27–29], it may be important to mention that the maintenance of the complex
architecture of wood was achieved through a chemical transformation process characterised by a very strict control of the reaction kinetics, which was obtained
through the use of the equipment permitting the regulation and control of the temperature and pressure of reacting gases, thus establishing the more suitable
thermodynamic conditions to reach a complete phase transformation in the whole solid [27]. Specifically, gas-solid reactions are strongly affected by various phe-
nomena related to the adsorption of the gaseous reactant by the solid, the kinetics of nucleation and growth of the newly forming inorganic phase at the surface, and,
most importantly, the penetration of the gaseous reactant in the inner regions of the structure so that only an effective control of all these processes yields the complete
transformation of the template into the desired phase [30]. This control becomes more and more critical when large ceramic pieces are to be obtained because
diffusive phenomena are predominant and regulate the rate of phase transformation. Hence, the strict control of the reaction kinetics is the key feature needed: (i) to
activate the chemical reactions throughout the whole solid without inducing deformations and relevant structural defects, (ii) to limit the grain growth, thus
maintaining the multiscale porosity, up to the nanosize, and (iii) to obtain highly reactive inorganic precursors, in turn facilitating subsequent transformation reactions
[27]. This complex and challenging chemical process has been used to produce the samples of sufficiently relevant dimensions for our mechanical experiments (and
also for bone replacement applications).

https://doi.org/10.1016/j.mtbio.2019.100032
Received 8 August 2019; Received in revised form 2 October 2019; Accepted 11 October 2019
Available online 24 October 2019
2590-0064/© 2019 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

chemically transform rattan wood into a biomimetic hierarchically 2. The lacunar fractal nature of porosity
structured hydroxyapatite (HA). Unlike in previous approaches where
the wood was infiltrated with HA slurries and finally sintered to elim- Nanotwinned HA lamellae are a unique nanostructure characterizing
inate the organic component and to consolidate the final ceramic [8, BA [28] and form a complex multiscale porosity, where large pores
10], the new procedure can directly transform wood pieces into large (approximately 300 μm in diameter) are surrounded by medium-sized
HA scaffolds, preserving the original multiscale structure through a pores (up to 50 μm in diameter) and coexist with a distributed fine
heterogeneous reaction under supercritical conditions directly in the 3D porosity (1 μm of diameter) (see the scanning electron microscopy image
state, without adopting any sintering process. This is an important in Fig. 2). The fine porosity is alveolar and closed, whereas the large- and
advantage because high-temperature treatments can represent a serious medium-sized porosity is two-dimensional so that longitudinal channels
drawback for the scaffold bioactivity.2 The resulting ceramic material are present, closely resembling the structure of the wood from which the
(referred henceforth as ‘biomorphic apatite’ [BA]) not only maintains a material was indeed originated.
highly bioactive composition and a multiscale pore hierarchy almost To show that the porosity of BA possesses a lacunar fractal nature [43,
identical to those of the parent wood (thanks to a careful control of the 44,51], images of microstructures captured by scanning electron micro-
reaction kinetics that prevents critical deformations at all scales and to scopy at different magnifications (38
, 100
, and 250
, in all cases
the absence of sintering processes) but also shows a nanosize structure using a variable pressure inside the scanning electron microscope (SEM)
and unprecedented mechanical properties owing to great cohesion of chamber in the 20  25MPa range) have been analyzed using an image
the neo-formed nanocrystals [27]. BA represents a significant recognition program ad hoc developed in MATLAB to filter the gray scale
improvement in the development of 3D inorganic devices with a com- scanning electron microscopy images and transform them into binary
plex microstructure and multiscale details, which are relevant for smart data (depicted in yellow or blue in Fig. 3). This program allowed a fast
functionality, which is still an open challenge owing to the ineffec- computation of the apparent porosity of the material based on the ratio
tiveness of the current ceramic fabrication processes. between the number of blue pixels and the total number of pixels. As
The objective of the present study is the systematic investigation of usually occuring with real images, an intensity threshold of gray (a
the mechanical properties of BA, which are found to be similar to the parameter ranging between 0 and 255) has to be introduced to filter the
mechanical properties of both wood and bone, so that BA becomes data and transform the gray scale image into black and white. Therefore,
promising for bone regeneration, especially for load-bearing regions. A the curves in Fig. 4 show the evolution of the apparent porosity vs. the
morphological investigation is reported, which shows that the porosity filter index, which is defined as the ratio between the aforementioned
of BA displays a fractal nature of the lacunar type, a feature that may intensity threshold and the value 255, for the three scanning electron
justify its excellent damage tolerance. In fact, the fractality of the microscopy images acquired at different magnifications. The porosity is
porosity explains our experimental observation that the fracture is not clearly resolution dependent, being an increasing function of the
abrupt and straight (as usual for ceramics) but evidences growth and magnification. Such a resolution dependency reveals a lacunar fractal
tortuosity (as usual for bone). The results from an experimental nature for BA. For a rigorous quantification, the box-counting method has
campaign are presented, based on multiple mechanical tests: uniaxial been applied to compute the local fractal dimension of the lacunar
compression (in both a standard configuration and in situ by scanning domain D. For each box of lateral size r, the number N of boxes containing
electron microscopy), three-point bending, ring tests, and ultrasound at least one black pixel is counted. This operation is repeated at varying r
evaluations. It is shown that BA performs better than the standard HA from 1 up to 512 lateral size divisions, with a geometric progression of 2.
[31,32] in several mechanical characteristics: elastic stiffness, strength, From the NðrÞ relation, the local fractal dimension D of the porous ma-
damage tolerance before failure, and related toughness. Moreover, terial has been finally obtained by differentiating logN with respect to
unlike in HA and similar to wood and bone, BA is found to exhibit a logr, thus obtaining DðrÞ ¼  dðlogNÞ
dlogr . The results are presented in Fig. 5
strength in tension often superior to that in compression, to evidence a where, for 500 computations equally sampled by varying the filter index
transversely isotropic behavior, which permits optimization of stiffness
between 0 and 0.3, the mean value of DðrÞ (with their error bars, cor-
in the direction of loading, a feature particularly useful for bone responding to s, where s is the root mean square of the computed
replacement.
values) is plotted as a function of logr. As expected, the local fractal
Our experiments allow the placement of BA in the Ashby charts, in dimension quantifying the effect of porosity is always less than 2, which
terms of Young modulus vs strength or vs porosity, Fig. 1.
corresponds to a Euclidean surface, over the whole range of r.
The Ashby charts reveal that BA occupies a virgin zone for ceramics The fact that the porosity of BA has a lacunar fractality has a deep
and displays a similarity to the ligneous material from which it was ‘born’
influence on the fracture propagation and trajectory, which (as shown
and to several different bones, which evidences that the material is later) is not abrupt and straight, but displays a growth with load in
particularly suited for several biotechnologies.
tortuous patterns which are unusual for ceramics and similar to bone [45,
It has to be finally mentioned that the results reported in the present 46].
study indicate that outstanding mechanical properties could be obtained
for materials different from BA but obtained with a process similar to that 3. Mechanical testings
used for BA, which can therefore be used as a guide for the fabrication of
a new generation of inorganic materials with significant improvement in Mechanical tests have been performed on 48 prismatic and 19 cy-
structural performance. This perspective is encouraged by previous re-
lindrical samples and also on 17 hollow cylinders of BA and, for com-
sults obtained with the chemical transformation of natural woods into parison, on 42 samples of two Engipore HAs [27].
various oxide (for instance, Al2O3, ZrO2, TiO2, and MnO) [33–37], and
Before testing, the chemical composition of the specimens has been
non-oxide (for instance, SiC, TiC, and ZrC) [38–42] ceramics, which are investigated to assess the successful and complete transformation of
particularly relevant for structural applications.
rattan. This was verified through X-ray diffraction tests, one of which is
presented inFig. 6, showing that the composition of HA is accompanied
by relatively small quantities (less than 5%) of beta-tricalcium phosphate
(β  Ca3 ðPO4 Þ2 ), represented by the black dots.
2
The drawback becomes particularly evident when the material's function- In the following section, a nomenclature borrowed from wood me-
ality is related to the presence of nanocrystalline, non-stoichiometric, or meta- chanics will be adopted for BA, namely, ‘parallel or orthogonal to the
stable phases, which can be easily degraded by high-energy processes. grain’ will denote alignment parallel or normal to the fibers (also
Moreover, sintering processes destroy the surface chemistry, reducing the denoting the direction of transverse isotropy) of the wood from which the
biocompatibility of materials.

2
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Fig. 1. Ashby charts reporting Young modulus vs strength (upper part) and vs porosity (lower part) for biomorphic apatite, loaded parallel (BA= = ) and perpendicular
(BA?) to the microtubule structure (‘grain’ in the following), for rattan wood (from which BA was obtained), and for bones. BA, biomorphic apatite; HA,
hydroxyapatite.

ceramic was made. stocks, so that the results presented in all tables (except Tables 8 and 10)
The results of (i) uniaxial compression tests, (ii) three-point bending, and in all figures (except Fig. 24) have been reported (with a linear
and (iii) ring tests are summarized respectively for the two materials in proportion) to the same value of the mass density, ρ0 ¼ 1:47 g=cm3 ,
Tables 1 and 2, where mean values and standard deviations are reported. assumed as the reference (as it was the most representative value for BA).
Note that the first two tests have been performed following the standards The possibility of using ultrasonic tests has also been analyzed (on 6
[47,48]. samples taken from stocks #1 and #2) to investigate the mechanical
Values of compression σ c , tension strengths σ t , and elastic modulus E properties of BA on a scale much smaller than that explored through the
(index ‘c’ in compression; index ‘f’ in flexure), are summarized in Ta- other mechanical tests and typical of the cells forming bone tissues
bles 1 and 2 (osteoblasts).
Different stocks of ceramics (four for BA and two for HA) have been An additional series of compressive tests were performed on prismatic
investigated; for BA, these will be denoted by a number (#1, #2, #3, samples by scanning electron microscopy, with different purposes: (i) to
#4), whereas for HA, it will be denoted by HA1 and HA2. The samples assess the role of friction between the steel platens and BA prismatic
have been ad hoc produced in different shapes to allow the execution of specimens, (ii) to document crack pattern evolution and sample exfoli-
multiple mechanical tests. The mass density was varying for different ation under uniaxial compression, and (iii) to evaluate the effect of the

3
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

platen to reduce friction, whereas the opposite interface has been left
with friction. Finally, a third experimental setup has been designed
(through the interposition of paper board layers) to induce eccentric
compression at low friction with the steel platens. Further evidence is
provided in videos filmed during different tests (see Support electronic
material).
Mechanical tests were conducted at the Instability Lab (University of
Trento) using a Midi 10 and a Beta 100 electromechanical testing ma-
chines (Messphysik Materials Testing) used in different configurations
for uniaxial compression tests, three-point bending, and ring tests (on
hollow cylinders), which are detailed in this section. Additional in situ
compression tests were conducted with the tensile/compression stage
DEBEN 5000S from GATAN, which is placed within the SEM EVO MA15

Fig. 2. An scanning electron microscopy image of the microstructure of

biomorphic apatite showing long open channels surrounded by finer alveolar
closed pores.

eccentricity in the axial load, which can be caused in real applications by

local imperfections in the specimen geometry and by inhomogenity of
the material's microstructure. In a first experiment, the BA specimen has
been tested under uniaxial compression with steel platens in direct
Fig. 4. Increase of apparent porosity for BA with the filter index at different
contact with the BA specimen, which leads to a contact problem with
magnification scales, showing the lacunar fractal nature of BA. BA, biomor-
high friction. In a second experiment, a high-density polyethylene layer
phic apatite.
has been interposed at the interface between the specimen and steel

Fig. 3. Scanning electron microscopy images of

cross sections of biomorphic apatite (BA) and the
corresponding apparent porosity identified on the
basis of the application of an ad hoc image anal-
ysis software with a filter index of 0.2. In partic-
ular, (a) 38
scanning electron microscopy image
of BA; (b) at 38
, the porosity is detected to be
14:6%; (c) 100
scanning electron microscopy
image of BA; (d) at 100
, the porosity is detec-
ted to be 19:0%; (e) 250
scanning electron mi-
croscopy image of BA; (f) at 250
, the porosity is
detected to be 27:0%. The image correlation
software shows that porosity is increasing with
the magnification, and this reveals the lacunar
fractality of the cross section domain. BA,
biomorphic apatite.

4
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

(owing to its fibrous nature inherited from wood) so that compression

has been applied both parallel and orthogonal to the fibers. (Both tests
have been performed only on prismatic samples, whereas cylindrical
samples have been tested only parallelly to the fibers.) The sample, in
direct contact with two circular steel platens (40-mm diameter and 5-mm
thick), is compressed by imposing displacements to the upper platen.
Fig. 7 reports a sequence of photos taken during a typical uniaxial
compression test on a BA cylindrical sample and related to the corre-
sponding stress/strain curve (reported in the following section). For
comparison, in Fig. 8, a uniaxial compression test on standard HA is re-
ported, which abruptly fails as soon as a splitting crack (visible in the
Fig. 5. Local fractal dimension D (which quantifies the porosity) as a function of
photo) is enucleated.
r (the lateral side division of the image in the box-counting algorithm).
The photos show a progressive failure with continuing exfoliation for
the BA sample compressed parallel to the grain, in sharp contrast with the
typically brittle failure displayed by HA (Fig. 8 and the study by Libonati
et al [49]). The exfoliation failure pattern is similar to that documented
by Gei et al. [50] for silicon nitride at high temperature but is now much
more pronounced and yields a much more evident softening in the
stress/strain curve (see Fig. 9) where a result for BA is reported (on the
left) and contrasted with an analogous result for HA (on the right). Note
that the stress/strain curves have been both reported to the reference
density of 1.47 g/cm3.
Note that BA evidences a higher peak strength than HA, both occur-
ring practically at the same strain level, but while HA is abruptly crushed
at that strain, exfoliation takes place for BA, which continues to carry
load and attains a final strain of more than 35%, thus displaying a great
toughness for a ceramic.
Further results from compression tests parallel to the grain are re-
ported in Table 3, for four different stocks of BA, each one characterized
by different geometries and dimensions. The results are contrasted with
the analogous for the two standard HAs, also reported in Table 3.
It can be observed in Table 3 that the fibrous microstructure of BA
implies that during compression parallel to the grain, the sample is not in
perfect contact with the loading platens, an effect which produces a
pronounced scattering of experimental results and is detrimental to the
Fig. 6. XRD (x-ray diffraction) pattern of one sample of bioapatite used for
measured strength. This effect is much less pronounced for HA samples,
mechanical tests, where the dots point to reflections addressing the β-TCP phase.
where parallelism and planarity of the surfaces can be better obtained
All the unmarked peaks point to the HA phase. HA, hydroxyapatite; β-TCP, beta-
than for BA.
tricalcium phosphate.
Prismatic samples from stock BA #4 have been compressed orthog-
from Zeiss, both available in the MUSAM-Lab (the experimental labora- onally to the grain (i.e., normally to the weak plane), with the same
tory of the Multi-scale Analysis of Materials Research Unit at the IMT experimental configuration used for other compression tests. A sequence
School for Advanced Studies Lucca). of photos during this compression is reported in Fig. 10, showing the
progressive formation of a set of inclined fractures, something that would
3.1. Uniaxial compression test be hardly visible for standard HA.
In the direction orthogonal to the grain, BA is weaker than HA, as a
BA and HA prismatic and cylindrical samples and BA tubular samples consequence of the highly oriented microstructure of BA, providing
have been tested under uniaxial compression. While HA is isotropic and strong transverse isotropy in strength and also in elastic stiffness. This
does not show any orientation characteristic, BA is strongly anisotropic can be observed from the stress/strain curve reported in Fig. 11, typical

Table 1
Summary of the mechanical properties of biomorphic apatite (BA), obtained from uniaxial compression, three-point bending, and
compression ring tests on samples of different geometry (prismatic, cylindrical, and tubular). All the mechanical properties are expressed
in terms of their average  standard deviation: σ c , strength in compression; σ t , strength in tension; Ec , Young modulus in compression; Ef ,
Young modulus in flexure; index, ‘ k ’ and ‘ ? ’ stand for parallel and orthogonal to the grains, respectively. Finally, ρ=ρ0 is the ratio
between the density and the reference density of 1:47 g=cm3 to which all mechanical values have been reported.

5
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Table 2
Summary of the mechanical properties of commercial hydroxyapatite (HA), obtained from uniaxial compression and three-point bending tests on prismatic samples.
Sample set Uniaxial compression Three-point bending

σ c [MPa] Ec [MPa] ρ=ρ0 [] σ t [MPa] Ef [MPa] ρ=ρ0 []

HA 1 6:2  2:5 239  131 0.36 3:5  0:6 3686  1087 0.36
HA 2 5:6  1:6 79  23 1.41 5:7  0:8 5235  886 1.27

HA, hydroxyapatite.
All the mechanical properties are expressed in terms of their average  standard deviation: σ c , strength in compression; σ t , strength in tension; Ec , Young modulus in
compression; Ef , Young modulus in flexure. ρ=ρ0 is the ratio between the density and the reference density of 1:47 g=cm3 , to which all mechanical values have been
reported.

Fig. 7. A sequence of photos taken during a uniaxial compression test of a cylindrical sample of BA. The instances when the photos have been taken are marked on the
stress/strain curve (Fig. 9 on the left) reported in the following. Note the progressive exfoliation of the sample, strongly enhancing toughness and related to the peaks
in the stress/strain diagram. BA, biomorphic apatite.

Fig. 8. A sequence of photos taken during a uniaxial compression test of hydroxyapatite samples. The instances when the photos have been taken are marked on the
stress/strain curve (Fig. 9 on the right) reported in the following. Note the abrupt failure of the specimen immediately after the appearance of a splitting crack.

Fig. 9. Stress/strain behavior of a BA sample (#1, on the left) and of a HA sample (#1, on the right) subject to uniaxial compression, parallel to the grain for BA. The
green spot identifies the peak strength, and a straight line drawn through the two indicated blue spots was used to evaluate the Young modulus E ¼ tanα. The superior
toughness of BA is evidenced by the slow load fall, contrasting with the sharp jump to zero displayed by HA.

for compression orthogonal to the grain of BA prismatic samples. How- Compression orthogonal to the BA fibers was performed on prismatic
ever, although strength and stiffness orthogonal to the grain are notice- samples only. In Table 4, the geometric properties of such samples are
ably inferior to those measured parallel to the grain, BA still retains an reported, along with the mechanical results of the tests.
excellent toughness.

6
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Table 3
Mechanical characteristics from uniaxial compression tests on different stocks of BA and HA (compression parallel to the grain for BA).
Sample set Geometrical properties Uniaxial compression test

ϕ [mm] H [mm] A ½mm 

2
ρ ½g=cm 
3
σ ck [MPa] Eck [MPa] ρ=ρ0 []

#1 14.30 19.89 160.62 1.39 6:5  3:0 216  99 0.95

#2 15.21 19.83 181.62 1.49 7:7  1:6 79  13 1.01
#3 29.35 31.13 511.29 1.56 10:8  2:9 1011  448 1.06
#4 19.47 31.22 213.27 1.44 12:3  3:2 2201  147 0.96
HA 1 10.07 14.29 79.59 0.54 6:2  2:5 239  131 0.36
HA 2 10.03 14.75 79.05 2.07 5:6  1:6 79  23 1.41

HA, hydroxyapatite; BA, biomorphic apatite.

On the left, the averages of the geometrical properties of each stock are shown: ϕ is the main dimension of the samples' base (which corresponds to the outer diameter of
samples #1; #2; #3 and to a side of the prismatic sample #4); H is the height; A is the average cross-sectional area, and ρ is the average density. On the right, the
following average properties ( standard deviation) are reported (from left to right): compressive strength and Young modulus. Finally, ρ=ρ0 is the ratio between the
density and the reference density of 1:47 g=cm3 to which all mechanical values have been reported.

3.2. Three-point bending tests

Three-point bending tests have been carried out to evaluate the ten-
sile stiffness and strength of BA and for comparison of HA. The tests have
been performed by imposing the vertical displacements of a steel tip
placed at the centerline of a prismatic sample, supported by two cylin-
drical elements placed at a distance L. The values of the stress σ, strain ε,
and elastic Young modulus in tension Ef are determined, respectively,
from the value of the applied load P and deflection at centerline v and
from a difference between two selected values of the load ΔP (measured
during elastic response of the sample), together with the corresponding
difference between midspan deflections Δv. The specimen is treated as a
linear elastic beam [52] as follows:

Lh 6h ΔP L3
σ¼P ; ε¼v ; Ef ¼ ; (1)
8I L2 Δv 48 I
where I is the moment of inertia of the cross section, L is the distance
between the supports, and h is the height of the cross section.
The three-point bending test with tension parallel to the grain was Fig. 11. Stress/strain behavior of a BA sample compressed parallel to the grain.
carried out on BA and HA samples for comparison to evaluate the tensile The green spot identifies the peak strength, and a straight line drawn through
elastic stiffness and strength in the direction parallel to the grain. A the two indicated blue spots was used to evaluate the Young modulus E ¼ tanα.
BA, biomorphic apatite.
sequence of photos taken during a three-point bending test on BA is re-
ported in Fig. 12, showing a feature rarely visible in a ceramic sample,

Fig. 10. A sequence of photos showing progressive failure during uniaxial compression orthogonal to the grain of a BA prismatic sample (from set #4, geometrical and
mechanical properties reported in Table 4). BA, biomorphic apatite.

7
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Table 4
Mechanical characteristics from uniaxial compression tests orthogonal to the grain.
Sample set Geometrical properties Uniaxial compression test

B [mm] H [mm] S [mm] A ½mm 

2
ρ ½g=cm 3
σ c? [MPa] Ec? [MPa] ρ=ρ0 []
#4 40.51 10.21 19.72 799.01 1.41 3:9  0:1 59  18 0.96

All samples are prismatic, with the following average dimensions: B width; H height; S depth; A cross-sectional area; and ρ density. On the right, the following average
properties ( standard deviation) are reported (from left to right): compressive strength and Young modulus. Finally, ρ=ρ0 is the ratio between the density and the
reference density of 1:47 g=cm3 to which all mechanical values have been reported.

Fig. 12. A sequence of photos taken during a three-point bending test on the prismatic sample of biomorphic apatite (BA set # 1), tested in the direction parallel to the
grain. The growth of a tensile fracture is clearly documented. (See the insets marked blue and green; the inset marked red shows a detail of the loading blade.) BA,
biomorphic apatite.

namely, the growth of a tensile crack, something which cannot be related to the dimensions of the samples, which are too thick to be rep-
observed in a HA sample and again demonstrating the superior damage resented as beams, so that the failure shown in Fig. 14 is more related to
tolerance of BA. the shear force applied by one support than to flexure.
A typical stress/strain curve obtained from a three-point bending test
is shown in Fig. 13, on the left for BA (with tension parallel to the grain)
and on the right for HA. 3.3. Ring tests
Further results on mechanical properties measured in bending (with
tension parallel to the grain for BA) for different stocks are presented (in The ring test is performed on tubular samples that are compressed
terms of average stress and stiffness) in Table 5. along the diametral direction until failure occurs. This setup is similar to
Table 5 shows that the mechanical properties, in particular the elastic the so-called ‘Brazilian test,’ except that the samples for the latter test are
Young modulus of BA, are noticeably different from stock to stock. This solid cylinders. Both tests are used to investigate the tensile mechanical
effect can be related to the different aspect ratio (H=L) of the specimens properties of brittle materials. Consistently, the ring test has been thor-
as this parameter influences the failure mechanism of the sample. In fact, oughly used for rocks as these materials are brittle (as ceramics) so that
stock #4 is characterized by an aspect ratio 2.5 times higher than that for the tensile properties can be inferred from a compressive experimental
the other stocks (BA #1; #2). Therefore, a shear mechanism of failure setup.
prevails, leading to an underevaluation of the Young modulus. Many works [53–56] have shown that the tensile strength of a linear
The three-point bending test with tension orthogonal to the grain was elastic, perfectly brittle material evaluated in the ring test is a function of
carried out on BA samples to evaluate the tensile stiffness and strength in the ratio between the internal and external radii r. Nonetheless, for
that direction. A sequence of photos taken during a test is reported in values of r more than 0.3, Hudson [57] proved that the evaluated tensile
Fig. 14, showing once more the growth of a tensile crack, again some- strength tends to become a constant quantity, in other words, the
thing rarely visible in ceramics and documenting the great toughness of strength results unaffected by such ratio. Furthermore, Srinath and
BA. Acharya [58] demonstrated that the beam theory [52] is effective in the
A typical stress/strain curve obtained from a three-point bending test evaluation of the tensile strength of ring specimen for values of r ranging
for tensile stress orthogonal to the grain is shown in Fig. 15. between 0.6 and 1, in other words, when the ring thickness is very small.
Further results on mechanical properties measured in bending with Because in our study r ¼ 0:512  0:009, the stress/strain relation σ/ε has
tensile stress orthogonal to the grain are presented (in terms of average been obtained by the following equation:
stress and stiffness) in Table 6.
hR 8h
Table 6 shows that the mechanical properties of BA, evaluated σ¼P ; ε¼v 2 ; (2)
2π I ðπ  8ÞR2
orthogonally to the grain, in particular the elastic Young modulus, are
significantly lower than the same properties evaluated in the direction
where v is the vertical displacement of the steel plate applying the force P
parallel to the grain. The fact that the measured values are much smaller
to the specimen, R is the average between external and internal radii, h is
than those referred to tests parallel to the grain (and also are smaller than
the thickness of the cylinder wall, and I is its moment of inertia.
the values obtained with the ring test presented in the next section) is
It should be noted that the ring tests are loading the sample in the

8
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Fig. 13. Tensile stress/strain behavior of BA (set #1, on the left) and of HA (set # 2, on the right) from a three-point bending test, parallel to the grain for BA. The
green spot identifies the peak strength, and a straight line drawn through the two indicated blue spots was used to evaluate the Young modulus E ¼ tanα. The superior
toughness of BA is highlighted by the failure strain much higher than in the HA sample. BA, biomorphic apatite; HA, hydroxyapatite.

Table 5
Mechanical characteristics from three-point bending tests on the different stocks of BA (with tension parallel to the grain) and HA.
Sample set Geometrical properties Three-point bending parallel to the grain

B [mm] H [mm] L [mm] Р ½g=cm 

3
σ tk [MPa] Ef k [MPa] ρ=ρ0 []

#1 11.04 5.72 37.33 1.22 8:7  3:1 3426  1448 0.83

#2 9.94 5.21 35.00 1.47 13:8  2:5 3341  884 1.00
#4 19.49 10.10 26.00 1.42 6:4  1:5 359  157 0.97
HA 1 10.03 5.11 40.00 0.53 3:5  0:6 3686  1087 0.36
HA 2 10.16 5.00 35.00 1.81 5:7  0:8 5235  886 1.27

BA, biomorphic apatite; HA, hydroxyapatite.

All the samples are prismatic, with the following average dimensions: B depth; H height; L length; and ρ density. On the right, the following average properties (
standard deviation) are reported (from left to right): tensile strength and Young modulus. Finally, ρ=ρ0 is the ratio between the density and the reference density of
1:47 g=cm3 to which all mechanical values have been reported.

‘weak’ direction, orthogonal to the grain, so that these tests are partic- and a second time at the formation of the horizontal crack. Before
ularly suited to reveal the less stiff and less resistant behavior of the cracking, the sample is a ring loaded with two opposite concentrated
material. forces, but after the first crack has formed, the static scheme changes to
A sequence of photos taken during the testing of BA is reported in an arch loaded with two opposite forces acting on its springer points. This
Fig. 16, showing initially the growth of two vertical tensile cracks and new configuration reminds that of the half ring test [59], where the
later of two horizontal tensile cracks, terminating the load-carrying ca- tensile strength of the specimen is obtained by using the Beer's formu-
pacity of the specimen. This growth of a tensile fracture, as already lation of a prismatic circular arch.
observed for three-point bending tests, is unusual for ceramic materials,
which usually display a sudden failure.
3.4. Ultrasonic tests
A typical stress/strain curve obtained from a ring test (tensile stresses
orthogonal to the grain) is shown in Fig. 17.
The multiscale structure of BA requires a mechanical characterization
Further results of the mechanical tests are shown in Table 7, together
both in the macroscale, for load-bearing applications, and nanoscale, for
with the geometrical properties of the samples.
the interaction of BA with bone cells. Therefore, in addition to the me-
The results in Table 7 report good values of strength for BA subjected
chanical tests performed at the scale of the sample, ultrasonic measure-
to tension orthogonal to the grain, values better than those obtained from
ments were taken at high wavelength in an attempt to characterize the
three-point bending. Fig. 17 displays an interesting feature, namely, a
elastic properties of the material at a microscale. P-waves were used
double peak in the stress/strain curve, which shows that the maximum
(generated and recorded using a pressure wave transducer, Olympus
tensile strength is achieved twice: first time at the first vertical cracking
A102S, frequency 1 and 0.5 MHz, connected to an ultrasonic square wave

Fig. 14. A sequence of photos taken during a three-point bending test on biomorphic apatite prismatic samples (set #4) tested in the direction orthogonal to the grain.
Note the growth of a tensile fracture.

9
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

material.
It is worth mentioning that knowing the frequencies of the applied P-
wave and its speed allows checking the wavelength of the applied
exciting pulse as follows:

vp 2500 m=s
λp ¼  ¼ 2:5 mm; (5)
fp 106 Hz

Which is higher than the typical dimensions of osteoblasts (on the

order of 20 μm), so that the elasticity detected using the ultrasound tests
is not believed to be representative of the scale of a single cell, but may
still be acceptable to characterize the biological response of the material
because these osteoblasts tend to interact with bone in large groups
rather than individually.

3.5. In situ mechanical tests and further analysis of crack patterns

Fig. 15. Tensile stress/strain behavior of BA (set #4) from a three-point A series of compression tests were performed using a micro-
bending test, with tensile stress parallel to the grain. The green spot identifies mechanical compressive stage placed inside a SEM (Zeiss EVO MA15
the peak strength, and a straight line drawn through the two indicated blue
equipped with the DEBEN5000S tensile/compressive stage) with the
spots was used to evaluate the Young modulus E ¼ tanα. The high toughness of
purpose of analyzing the effects of boundary conditions on the samples
BA is highlighted by the slow load fall. BA, biomorphic apatite.
and the fracture paths typical of BA.
Compression is applied under displacement control with a gearbox
speed of 0.1 mm/min to achieve quasi-static conditions. The equipment
has a 10-mm linear extensometer for position readout, with 300-nm
Table 6 resolution. The combined use of the stage and the SEM allows inspec-
Mechanical characteristics from three-point bending tests with tension orthog- tion of the evolution of the crack pattern and the role of the material
onal to the grain. microstructure during the application of axial displacement. Three BA
Sample Geometrical properties Three-point bending prismatic specimens, whose geometrical data are collected in Table 9,
set orthogonal to the grain were tested under different boundary conditions to assess the sensitivity
B H L ρ ½g= σ t? Ef ? ρ=ρ0 of the material response and of the fracture pattern upon variations from
[mm] [mm] [mm] cm3  [MPa] [MPa] [] a reference testing configuration characterized by uniaxial compression
#4 37.00 10.13 12.07 1.44 1:9  1:2 11  7 0.98 with steel platens in direct contact with the specimen (where friction is
present at the interface between steel and BA).
All the samples are prismatic, with the following average dimensions: B depth; H In the first test, the BA specimen is inserted between the steel platens
height; L length; ρ density. On the right, the following average properties (
without introducing measures to reduce friction and is loaded under
standard deviation) are reported (from left to right): tensile strength and Young
uniaxial compression. The overall stress/strain mechanical response is
modulus. Finally, ρ=ρ0 is the ratio between the density and the reference density
of 1:47 g=cm3 to which all mechanical values have been reported.
represented by the red curve in Fig. 19, labeled as ‘high friction.’ The
progress of the crack pattern inspected by scanning electron microscopy
imaging shows the appearance of cracks predominantly aligned with the
axial direction, parallel to the grain (Fig. 20). The stress/strain curve
pulser/receiver unit, Olympus 5077 PR, combined with a NI PCI-5152 shows a peak stress of 9 MPa, corresponding to first cracking, followed by
Digitizer/Oscilloscope). The tests were performed by measuring the a plateau and a progressive hardening continuing up to σ  10  11 MPa
time tv needed for P-waves to travel along the height H of the sample and and a significant toughness at failure so that the failure strain is ε  0:05.
reach a sensor so that the velocity of the wave is given by vp ¼ H= tv , The failure strength observed from the test is consistent with that ob-
which, for an isotropic elastic material, is linked to the Lame constants tained on similar specimens using the Messphysik Materials Testing,
through reporting σ  12:3 MPa (Table 1, first column, fourth row).
In the second test, a high-density polyethylene layer (visible in
sffiffiffiffiffiffiffiffiffiffiffiffiffi
λ þ 2μ Fig. 21) is inserted at the contact between one of the BA bases and the
vp ¼ ; (3) steel platen so that friction is reduced on one base but not on the opposite
ρ
base. The overall stress/strain curve is shown by the blue curve in Fig. 19,
so that the elastic modulus E may be evaluated, once the Poisson's ratio ν labeled as ‘low friction.’ The perturbed boundary conditions have not
is known, from the following equation. significantly modified the values of the failure strength and of the cor-
responding deformation, when compared with the ‘high friction’ case.
ðλ þ 2μÞð1 þ νÞð1  2νÞ However, the crack pattern leading to failure was found substantially
E¼ : (4)
1ν different, with the appearance of a major exfoliating crack perpendicular
to the grain direction and clearly visible on the specimen surface
The result of two typical ultrasonic tests on BA (right) and HA (left) is
(Fig. 21). Inspection of the specimen after failure also showed the exis-
shown in Fig. 18 at two different frequencies (0.5 and 1.0 MHz), in terms
tence of many cracks aligned parallel to the grain direction, again as
of signal amplitude (‘Amp’) versus time.
observed in the ‘high friction’ case (see Fig. 22).
Mechanical properties derived from ultrasonic tests (parallel to the
In the third test, the BA specimen was inserted with paperboard layers
grain) are reported in Table 8, in terms of mean values of uniaxial strain
at both its bases to reduce friction on both sides. Moreover, the loading
elastic modulus λ þ 2μ and Young modulus E, the latter evaluated for two
platens were modified with the purpose of inducing an eccentricity e ¼
extreme values of Poisson's ratio ν. Note that the data collected in Table 8
h=4 (h, the specimen height) in the compression load N. This testing
have not been reported to the reference density because in an ultrasonic
condition leads to a linear variation of the stress σ orthogonal to the cross
test, the measurements are believed to be referred to the skeleton
section of the specimen (and thus parallel to the grain):

10
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Fig. 16. A sequence of photos taken during a ring compression test on biomorphic apatite (BA) tubular samples (set #3) tested in the direction orthogonal to the grain.
The growth of tensile fractures is clearly documented (see the insets).

N Ne
σ¼ þ  y; (6)
A bh3 12

where A is the cross-sectional area of the specimen, b is its width, and y

picks up the position along the cross section. At the extrados, y ¼  h=2,
which corresponds to the upper side of the BA specimen observed in the
SEM, the axial stress reaches its maximum positive value, corresponding
to a tensile stress, whereas at the opposite side, compression prevails.
This stress state leads to a very peculiar failure mode, where exfoliation
of a thick layer of BA parallel to the grain takes place, followed by the
final failure by transverse cracking on the side subjected to tension (see
Fig. 22). The corresponding stress/strain curve represents the evolution
of the tensile stress at the extrados, which is denoted with a black line in
Fig. 19, labeled as ‘eccentric compression.’ The failure strength (parallel
to the grain) approaches 6 MPa in tension, which is consistent with data
reported in Table 9 obtained from three-point bending tests.

Fig. 17. Tensile stress/strain behavior of BA (Set #3) from a ring compression 4. BA vs. wood and bone: the Ashby plot
test, with tensile stress parallel to the grain. The green spot identifies the peak
strength, and a straight line drawn through the two indicated blue spots was The mechanical behavior of BA is now compared with that of its
used to evaluate the Young modulus E ¼ tanα. The high toughness of BA is
parent material, rattan wood, used to produce the BA, and with that of
highlighted by the slow load fall. BA, biomorphic apatite.
human bones. Tests have been conducted on rattan wood, obtained from
a plant belonging to the subfamily of Calamoideae, mechanical proper-
ties of which have been investigated through compression and three-
point bending experiments (performed following the standards

Table 7
Mechanical characteristics from ring tests on BA samples (tensile stresses are orthogonal to the grain).
Sample set Geometrical properties Ring test

H [mm] ro [mm] ri [mm] ρ ½g=cm 

3
σ t;? [MPa] Ef ;? [MPa] ρ=ρ0 []
#3 30.85 12.79 6.54 1.572 6:0  0:94 2851  1152 1.1

BA, biomorphic apatite.

All the samples are tubular, with the following average dimensions: H depth; ro outer radius; ri internal radius, and ρ density. On the right, the following average
properties ( standard deviation) are reported (from left to right): tensile strength and Young modulus. Finally, ρ=ρ0 is the ratio between the density and the reference
density of 1:47 g=cm3 to which all mechanical values have been reported.

11
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Fig. 18. Amplitude of the signal versus time

for two ultrasonic tests on a BA sample (set
#1, on the left) and on a HA sample (set #1,
on the right) at two different frequencies (0.5
and 1.0 MHz). Note that the points that have
been identified as the representative of the
time of flight tv have been identified in
agreement with the standards [60] so that
the arrival of the P-wave has been marked at
the instance when the leading edge of the
first peak was reached at 80%. (The evalua-
tion of the time of flight was performed
automatically using a Mathematica routine
ad hoc developed.) BA, biomorphic apatite;
HA, hydroxyapatite.

Table 8
Mechanical characteristics from ultrasonic tests on BA samples (parallel to the grain) and HA.
Set Geometrical properties Ultrasonic test

H [mm] ϕ [mm] A ½mm2  ρ ½g=cm3  vp [m/s] λ þ 2μ [MPa] E [MPa]

ν ¼0 ν ¼ 0:25
# 1,2 20.84 14.19 158.08 1.42 2723  212 10398  1661 10398  1661 8665  1384
HA 1,2 15.01 10.07 79.61 0.54 2191  72 2583  290 2583  290 2152  241

BA, biomorphic apatite; HA, hydroxyapatite.

All the samples are cylindrical, with the following average dimensions: H height; ϕ diameter; A cross-sectional area; and ρ density. On the right, the following average
properties ( standard deviation) are reported (from left to right): wave velocity; uniaxial strain elastic modulus; and Young modulus for two different values of
Poisson's ratio. All reported values have not been normalized to the reference density.

[61–64]), with the same configurations used for ceramics. A series of

Table 9 photos taken during a typical compression test are shown in Fig. 23, and
Geometrical properties for the BA specimens tested in situ using a SEM and
the relative stress/strain curve is shown in Fig. 24 on the left, and the
subjected to different loading conditions.
same figure presents a tensile stress/strain curve obtained from a
Loading condition Geometrical properties three-point-bending test on the right.
H w t ρ=ρ0 p (%) Further results from the mechanical compression and three-point
[mm] [mm] [mm] [] bending tests on rattan wood (with the stress always aligned parallel to
Uniaxial compression (high 31.47 20.52 10.11 0.91 57.29 the grain) are shown in Table 10, together with the geometrical prop-
friction) erties of the samples.
Uniaxial compression (low 31.29 21.29 10.53 0.93 56.21 Fig. 23 shows a failure process characterised by exfoliation by
friction)
buckling, a long process highlighting the outstanding resistance of the
Eccentric compression 31.55 21.46 10.22 0.95 55.83
material, while Fig. 24 shows a flat postelastic regime, characterized by
SEM, scanning electron microscope; BA, biomorphic apatite. large deformations. In rattan wood, failure may take different definitions
H is the height of the sample, w is the width, and t is the thickness. ρ= ρ0 is the so that it can be identified with the first load corresponding to visible
ratio between the density and the reference density of 1:47 g=cm3 , and p is the
local buckling of the fibers (I), with maximum curvature in the diagram
porosity.
(II), with the beginning of the plateau (III), with the peak (IV) in the
curve, or, finally, (V) with the beginning of softening. Therefore, when

Fig. 19. Stress/strain curve of BA specimens tested in situ using a SEM under uniaxial compression at two levels of friction (at the specimen/platen contacts) and for
eccentric loading. BA, biomorphic apatite; SEM, scanning electron microscope.

12
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Fig. 20. Specimen tested under uniaxial compression with high friction. On the left: view of the specimen at failure; on the right: cracks at failure.

Fig. 21. Specimen tested under uniaxial compression by scanning electron microscopy under conditions of low friction. On the left: view of the specimen at failure
(note the high-density polyethylene layer in black colour); on the right: cracks at failure.

Fig. 22. BA specimen tested under eccentric compression in situ using a SEM and shown at failure on the left (note the paperboard layers in contact with the sample);
cracks at failure are detailed on the right. BA, biomorphic apatite; SEM, scanning electron microscope.

Fig. 23. A sequence of photos showing progressive failure (involving exfoliation by fiber buckling) of a rattan wood specimen during uniaxial compression.

failure is identified with the first appearance of local buckling of fibers, Bones are highly hierarchical composite materials so that their me-
their strength in compression becomes lower than that in tension. It can chanical behavior is controlled by its two main constituents: cancellous
be concluded that the behavior of rattan wood has several points in bone (soft and deformable) and cortical bone (stiff and strong). Both of
common with that of its derived material BA. these materials are transversally isotropic [65–67], with stiffness and

13
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

Fig. 24. Stress/strain behavior of a rattan wood sample subject to uniaxial compression (on the left) and three-point bending (on the right), parallel to the grain. The
red spots (in the compression plot) identify different stress levels corresponding to different definitions of failure. A straight line drawn through the two indicated blue
spots was used to evaluate the Young modulus E ¼ tanα.

sintered HA, but also to other biomorphic HAs obtained from different
Table 10
types of woods that are present in the literature [7,8,10]. Moreover, BA is
Mechanical characteristics from uniaxial compression tests and three-point
transversely isotropic so that it can be effectively used in all the situations
bending (in both cases, stress parallel to the grain) on rattan wood.
(for instance, bone replacement) where the ceramic piece is subjected to
Tests Geometrical properties Mechanical parameters
a load having a privileged direction. The fact that BA is produced without
H, L ϕ ρ σ [MPa] E [MPa] sintering and its superior mechanical performance suggest use of this
[mm] [mm] ½g=cm3  material for several challenging biotechnologies.
Compression 29.18 21.90 0.46 23:7  2:7 1326  164
3-point 240 19.87 0.46 50:0  11:8 2127  415
bending Declaration of competing interest
All the samples are prismatic, with the following average dimensions: H height; L
span; ϕ base dimension; and ρ density. On the right, the following average The authors declare that they have no known competing financial
properties ( standard deviation) are reported (from left to right): compressive interests or personal relationships that could have appeared to influence
strength ad Young modulus. the work reported in this paper.

Acknowledgments

strength different in tension and in compression [66–68] (see also the The authors are grateful to Mr. Flavio Vinante (University of Trento)
study by Cuppone et al [69]). Overall, the mechanical behavior of bone is for the invaluable help with the experiments. D.B. and R.C. acknowledge
also affected by the amount of cancellous vs. cortical bone so that for flat, financial support from the PRIN 2015 ‘Multi-scale mechanical models for
long, and irregular bones, the mechanical properties are unique and the design and optimization of micro-structured smart materials and
different from one another [67,70–72]. metamaterials’ 2015LYYXA8-006. D.M., M.P., A.R., S.S., and A.T.
The Ashby charts expressed in terms of Young modulus vs. strength acknowledge financial support from ERC-2013-ADG-340561-
and Young modulus vs. porosity for BA, loaded parallel (BA= = ) and INSTABILITIES.
perpendicular (BA?) to the grain, for the rattan wood used to produce
the BA (parallel to the grain) and bones (data from the aforementioned References
literature) are shown in Fig. 1.
[1] P. Fratzl, R. Weinkamer, Nature's hierarchical materials, Prog. Mater. Sci. 52 (2007)
The Ashby charts show that BA is occupying a region usually not
1263–1334.
covered by ceramics; moreover, the Young modulus is fully compatible [2] R. Weinkamer, P. Fratzl, Mechanical adaptation of biological materials - the
with values typical of wood and bone (particularly flat bone). If, on the examples of bone and wood, Mater. Sci. Eng. C-Mater. Biol. Appl. 31 (2011)
1164–1173.
one hand, the strength of BA is lower than that of bone and wood, on the
[3] A.H. Hueuer, D.J. Fink, V.J. Laraia, J.L. Arias, P.D. Calvert, K. Kendall, G.L. Messing,
other hand, it is much higher than that of the HA usually used for bone J. Blackwell, P.C. Rieke, D.H. Thompson, A.P. Wheeler, A. Veis, A.I. Caplan,
replacement. These considerations, in addition to the fact that porosity Innovative materials processing strategies: a biomimetic approach, Science 255
and the hierarchical structure of BA are similar to those of wood and (1992) 1098–1105.
[4] C.E. Bryne, D.C. Nagle, Cellulose derived composites - a new method for materials
bone, suggest that BA is a ceramic with outstanding mechanical char- processing, Mater. Res. Innov. 1 (1997) 137–144.
acteristics, particularly suited for bone replacement and biological [5] P. Greil, Biomorphous ceramics from lignocellulosics, J. Eur. Ceram. Soc. 21 (2001)
applications. 105–118.
[6] G.W. Kooistra, V.S. Deshpande, H.N.G. Wadley, Hierarchical corrugated core
sandwich panel concepts, J. Appl. Mech. 74 (2) (2007) 259–268.
5. Conclusions [7] C.R. Rambo, T. Andrade, T. Fey, H. Sieber, A.E. Martinelli, P. Greil, Microcellular
Al2O3 Ceramics from wood for filter applications, J. Am. Ceram. Soc. 91 (3) (2008)
852–859.
BA, a ceramic obtained through a biomorphic transformation process [8] J. Qian, Y. Kang, W. Zhang, Z. Li, Fabrication, chemical composition change and
from rattan wood, presents a nanostructure and microstructure that re- phase evolution of biomorphic hydroxyapatite, J. Mater. Sci. Mater. Med. 19 (2008)
tains elements inherited from its parent material, the ligneous essence. 3373–3383.
[9] W.R. Legant, A. Pathak, M.T. Yang, V.S. Deshpande, R.M. McMeeking, C.S. Chen,
This structure has been shown to generate a lacunar fractal porosity,
Microfabricated tissue gauges to measure and manipulate forces from 3D
which yields the outstanding mechanical properties that we have microtissues, Proc. Natl. Acad. Sci. 106 (25) (2009) 10097–10102.
measured for BA. In particular, BA shows a damage and defect tolerance, [10] C. Eichenseer, J. Will, M. Rampf, S. Wend, P. Greil, Biomorphous porous
hydroxyapatite-ceramics from rattan (Calamus Rotang), J. Mater. Sci. Mater. Med.
strength, and elastic stiffness superior to its competitor material, a porous
21 (2010) 131–137.

14
D. Bigoni et al. Materials Today Bio 5 (2020) 100032

[11] F.P.W. Melchels, K. Bertoldi, R. Gabbrielli, A.H. Velders, J. Feijen, D.W. Grijpma, [40] S. Binghe, F. Tongxiang, Z. Di, T. Okabe, The synthesis and microstructure of
Mathematically defined tissue engineering scaffold architectures prepared by morph-genetic TiC/C ceramics, Carbon 42 (2004) 177–182.
stereolithography, Biomaterials 31 (2010) 6909–6916. [41] C.R. Rambo, J. Cao, O. Rusina, H. Sieber, Manufacturing of biomorphic (Si, Ti, Zr)-
[12] N.A. Fleck, V.S. Deshpande, M.F. Ashby, Micro-architectured materials: past, carbide ceramics by sol–gel processing, Carbon 43 (2005) 1174–1183.
present and future, Proc. R. Soc. A 466 (2010) 2495–2516. [42] L. Esposito, D. Sciti, A. Piancastelli, A. Bellosi, Microstructure and properties of
[13] L. Bardella, A. Sfreddo, C. Ventura, M. Porfiri, N. Gupta, A critical evaluation of porous b-SiC template from soft woods, J. Eur. Ceram. Soc. 24 (2004) 533–540.
micromechanical models for syntactic foams, Mech. Mater. 50 (2012) 53–69. [43] V.E. Tarasov, Elasticity of fractal materials using the continuum model with non-
[14] A. Vigliotti, D. Pasini, Mechanical properties of hierarchical lattices, Mech. Mater. integer dimensional space, Comptes Rendus Mechanique 343 (1) (2015) 57–73.
62 (2013) 32–43. [44] V. Kunin, S. Yang, Y. Cho, P. Deymier, D.J. Srolovitz, Static and dynamic elastic
[15] F. Libonati, A.K. Nair, L. Vergani, M.J. Buehler, Mechanics of properties of fractal-cut materials, Extreme Mech. Lett. 6 (2016) 103–114.
collagen–hydroxyapatite model nanocomposites, Mech. Res. Commun. 58 (2014) [45] X. Zhai, J. Gao, Y. Nie, N. Kedir, B. Claus, T. Sun, K. Fezzaa, X. Xiao, W.W. Chen,
17–23. Real-time visualisation of dynamic fractures in porcine bones and the loading-rate
[16] C.T. Bauer, E. Kroner, N.A. Fleck, E. Arzt, Hierarchical macroscopic fibrillar effect on their fracture toughness, J. Mech. Phys. Solids 131 (2019) 58–371.
adhesives: in situ study of buckling and adhesion mechanisms on wavy substrates, [46] D. Yin, B. Chen, W. Ye, J. Gou, J. Fan, Mechanical test and fractal analysis on
Bioinspiration Biomim. 10 (2015), 066002. anisotropic fracture of cortical bone, Appl. Surf. Sci. 357 (2015) 2063–2068.
[17] A. Buehrig-Polaczek, C. Fleck, T. Speck, P. Schueler, S.F. Fischer, Biomimetic [47] ASTM C1424 - 04 Standard Test Method for Monotonic Compressive Strength of
cellular metals-using hierarchical structuring for energy absorption, Bioinspiration Advanced Ceramics at Ambient Temperature.
Biomim. 11 (4) (2016). [48] ASTM C1161 - 18 Standard Test Method for Flexural Strength of Advanced
[18] M. Fraldi, G. Perrella, M. Ciervo, F. Bosia, N.M. Pugno, A hybrid deterministic- Ceramics at Ambient Temperature.
probabilistic approach to model the mechanical response of helically arranged [49] F. Libonati, A.K. Nair, L. Vergani, M.J. Buehler, Fracture mechanics of
hierarchical strands, J. Mech. Phys. Solids 106 (2016) 338–352. hydroxyapatite single crystals under geometric confinement, J. Mech. Behav.
[19] F. Ongaro, E. Barbieri, N.M. Pugno, The in-plane elastic properties of hierarchical Biomed. Mater. 20 (2013) 184–191.
composite cellular materials: synergy of hierarchy, material heterogeneity and cell [50] M. Gei, D. Bigoni, S. Guicciardi, Failure of silicon nitride under uniaxial
topologies at different levels, Mech. Mater. 103 (2016) 135–147. compression at high temperature, Mech. Mater. 36 (2004) 335–345.
[20] F. Ongaro, P. De Falco, E. Barbieri, N.M. Pugno, Mechanics of filled cellular [51] M. Brun, L. Casnedi, G. Pia, Bending strength of porous ceramics tiles: bounds and
materials, Mech. Mater. 97 (2016) 26–47. estimates of effective properties of an Intermingled Fractal Units' model, Ceram. Int.
[21] F. Libonati, G.X. Gu, Z. Qin, L. Vergani, M.J. Buehler, Bone–inspired materials by 44 (9) (2018) 10241–10248.
design: toughness amplification observed using 3D printing and testing, Adv. Eng. [52] S. Timoshenko, J.N. Goodier, Theory of Elasticity, Ed. 2, McGraw-Hill, 1951.
Mater. 18 (8) (2016), 1299-1299. [53] S. Timoshenko, On the distribution of stresses in a circular ring compressed by two
[22] G.X. Gu, F. Libonati, S.D. Wettermark, M.J. Buehler, Unraveling the role of nacre's forces acting along a diameter, London, Edinburgh, Dublin Philos. Mag. J. Sci. 44
mineral bridges, J. Mech. Behav. Biomed. Mater. 76 (2017) 135–144. (1922) 1014–1019.
[23] Z.L. Yu, N. Yang, L.C. Zhou, et al., Bioinspired polymeric woods, Sci. Adv. 4 (2018) [54] Ripperger, E.A., Davids, N., Critical stresses in a circular ring, Trans. Am. Soc. Civ.
eaat7223. Eng., Paper No. 2308, pag. 619-629.
[24] L. Zorzetto, D. Ruffoni, Wood-inspired 3D-printed helical composites with tunable [55] J.C. Jeager, R. Hoskins, Stresses and failure in rings of rock load in diametral
and enhanced mechanical performance, Adv. Funct. Mater. 28 (2018) 1804510. tension or compression, Br. J. Appl. Phys. 17 (1966) 685–695.
[25] F. Auricchio, A. Bacigalupo, L. Gambarotta, M. Lepidi, S. Morganti, F. Vadalac, [56] D.W. Hobbs, The tensile strength of rocks, Int. J. Rock Mech. Min. Sci. 1 (1963)
A novel layered topology of auxetic materials based on the tetrachiral honeycomb 385–396.
microstructure, Mater. Des. 179 (2019). [57] J.A. Hudson, Tensile strength and the ring test, Int. J. Rock Mech. Min. Sci. 6 (1968)
[26] U. Sabu, G. Logesh, M. Rashad, A. Joy, M. Balasubramanian, Microwave assisted 91–97.
synthesis of biomorphic hydroxyapatite, Ceram. Int. 45 (2019) 6718–6722. [58] Srinath, L.S., Acharya, Y.V.G., Stresses in a circular ring, Appl. Sci. Res., 1954, Sec.
[27] A. Tampieri, A. Ruffini, A. Ballardini, M. Montesi, S. Panseri, F. Salamanna, M. Fini, A Vol. 4, pag. 189-194.
S. Sprio, Heterogeneous chemistry in the 3-D state: an original approach to generate [59] B.H. Choi, Y.K. Lee, C. Park, C.H. Ryu, C. Park, Measurement of tensile strength of
bioactive, mechanically-competent bone scaffolds, Biomater. Sci. 7 (2019) brittle rocks using a half ring shaped specimen, Geosci. J. 23 (4) (2019) 649–660.
307–321. [60] UNI EN 583-3 - Non Destructive Testing - Ultrasonic Examination - Transmission
[28] A. Tampieri, S. Sprio, A. Ruffini, Study of the hydrothermal transformation of wood- Technique.
derived calcium carbonate into 3D hierarchically organized hydroxyapatite, Chem. [61] BS EN 408-2010-A1, Timber Structures. Structural Timber and Glued Laminated
Eng. J. 217 (2013) 150–158. Timber. Determination Of Some Physical and Mechanical Properties, 2012.
[29] A. Tampieri, S. Sprio, A. Ruffini, G. Celotti, I.G. Lesci, N. Roveri, From wood to [62] ISO 8375, Timber Structures – Glued Laminated Timber – Test Methods for
bone: multi-step process to convert wood hierarchical structures into biomimetic Determination of Physical and Mechanical Properties, 2009.
hydroxyapatite scaffolds for bone tissue engineering, J. Mater. Chem. 19 (2009) [63] ISO/TR 22157-1, Bamboo – Determination Of Physical and Mechanical Properties –
4973–4980. Part 1: Laboratory Manual, 2004.
[30] J. Szekely, Gas-solid Reactions, first ed., Academic Press, 1976, ISBN [64] ISO/TR 22157-2, Bamboo – Determination Of Physical and Mechanical Properties –
9780323151399. Part 2: Requirements, 2004.
[31] R.I. Martin, P.W. Brown, Mechanical properties of hydroxyapatite formed at [65] A.A. Abdel-Wahab, K. Alam, V.V. Silberschmidt, Analysis of anisotropic
physiological temperature, J. Mater. Sci. Mater. Med. 6 (1995) 138–143. viscoelastoplastic properties of cortical bone tissues, J. Mech. Behav. Biomed.
[32] A.J. Wagoner Johnson, B.A. Herschler, A review of the mechanical behavior of CaP Mater. 4 (5) (2011) 807–820.
and CaP/polymer composites for applications in bone replacement and repair, Acta [66] Y.C. Fung, Biomechanics: Mechanical Properties of Living Tissues, second ed.,
Biomater. 7 (2011) 16–30. Springer, 1993.
[33] J. Cao, C.R. Rambo, H. Sieber, Preparation of porous Al2O3-Ceramics by [67] Y.H. An, R.A. Draughn, Mechanical Testing of Bone and the Bone-Implant Interface,
Biotemplating of wood, J. Por. Mat. 11 (2004) 163–172. CRC Press, 2000.
[34] C.R. Rambo, H. Sieber, Novel synthetic Route to biomorphic Al2O3 ceramics, Adv. [68] S. Li, E. Demirci, V.V. Silberschmidt, Variability and anisotropy of mechanical
Mater. 17 (8) (2005) 1088–1091. behavior of cortical bone in tension and compression, J. Mech. Behav. Biomed.
[35] M. Singh, Bo-Moon Yee, Reactive processing of environmentally conscious, Mater. 21 (2013) 109–120.
biomorphic ceramics from natural wood precursors, J. Eur. Ceram. Soc. 24 (2) [69] M. Cuppone, B.B. Seedhom, E. Berry, A.E. Ostell, The longitudinal young's modulus
(2004) 209–217. of cortical bone in the midshaft of human femur and its correlation with CT
[36] J. Cao, O. Rusina, H. Sieber, Processing of porous TiO2-ceramics from biological scanning data, Calcif. Tissue Int. 74 (2004) 302–309.
preforms, Ceram. Int. 30 (7) (2004) 1971–1974. [70] K.V. Arun, K.K. Jadhav, Behaviour of human femur bone under bending and impact
[37] X. Li, T. Fan, Z. Liu, J. Ding, Q. Guo, D. Zhang, Synthesis and hierarchical pore loads, Eur. J. Clin. and Biomed. Sci. 2 (2) (2016) 6–13.
structure of biomorphic manganese oxide derived from woods, J. Eur. Ceram. Soc. [71] J.H. McElhaney, J.L. Fogle, J.W. Melvin, R.R. Haynes, V.L. Roberts, N.M. Alem,
26 (2006) 3657–3664. Mechanical properties of cranial bone, J. Biomech. 3 (1970) 495–511.
[38] P. Greil, T. Lifka, A. Kaindl, Biomorphic cellular silicon carbide ceramics from [72] M. Biggemann, D. Hilweg, P. Brinckmann, Prediction of the compressive strength of
wood: I. Processing and microstructure, J. Eur. Ceram. Soc. 18 (1998) 1961–1973. vertebral bodies of the lumbar spine by quantitative computed tomography, Skelet.
[39] P. Greil, T. Lifka, A. Kaindl, Biomorphic cellular silicon carbide ceramics from Radiol. 17 (1988) 264–269.
wood: 2. Mechanical properties, J. Eur. Ceram. Soc. 18 (1998) 1975–1983.

15