Influence of manganese ions in calcium silicate glass–ceramics on optical, mechanical, and magnetic properties

Nanocrystalline calcium silicate powder was synthesized by adding different ratios of MnO2 ranging from 0.00 to 2.00 wt% to detect its effect on the structure and physical properties. The pseudowollastonite triclinic and low combeite of hexagonal phase with nanocrystallite size less than 85.0 nm were confirmed by the XRD technique and average particle size ranging from 7.8 to 27.9 nm as detected by HR-TEM micrograph images. Stretching and bending vibration of the O–Si–O band were shifted to higher values upon the addition of MnO2 were verified by FT-IR. Increasing both the density and ultimate strength with a reduction in the porosity leads to an improvement in the mechanical properties with the addition of MnO2. Additionally, the increasing MnO2 content showed an improvement in magnetic and optical properties, which exhibited a decrement in the optical band gap Eg from 3.9 to 1.6 eV. Hence, the MnO2 acts as a structural network modifier of calcium silicate glass–ceramics. Furthermore, the estimated values of the Lande g-factor (2.01534–2.01731) for the d5 system of the Mn2+ displayed a negative shift from the free electron (2.0023), and the hyperfine splitting constant A value was 87 × 10−4 cm−1, indicating that the Mn2+ ions are in an ionic environment.

manganese (Mn 2? ) can reinforce the capability of the calcium silicate [2,3]. As prospective network modifiers, silicon and calcium stand for the improvement of bone substituent as therapeutic inorganic ions essential during the early stages of bone development, mineralization, and growth [4]. Calcium silicate is perceived as osteoinductive with the advantage of being safe (i.e., not being toxic to living cells) [5,6].
In material science, because of its characteristic physicochemical properties manganese dioxide (MnO 2 ) as a non-stoichiometric compound added an imperative consideration [7]. MnO 2 is chosen over other oxides because of its low cost, high activity, good stability, large specific surface area, structural flexibility, low toxicity, and sturdy oxidizing properties [8]. The thrilling physicochemical properties of MnO 2 are its dimensionality, crystallographic phases, particle size, and morphology. It acknowledged that manganese could be improved the tediousness of numerous functional materials as it accomplishes severe biomedical necessities [9]. Furthermore, the incapacitating MnO 2 can mend the dielectric properties of ceramics as exposed by investigators [10]. The effect of MnO 2 mostly influences the microstructure and energy storage properties of the calcium silicate. Xiu et al. established that the addition of a small amount of MnO 2 could improve the densification and homogenization of the material and it can well develop its microstructure [11]. Danewalia and Singh revealed that the occurrence of MnO 2 flourishes the disordering, which may improve the glass surface activity to form an appetite layer [12]. In addition, Kolmas et al. recognized that Mn doping possibly would intensify the density and compressive strength of the matrix [13]. Du et al. stated that scaffolds doped with Mn displayed developed compression strengths [14]. Calcium silicate doped with Mn could enrich bone regeneration more professionally than the undoped CS scaffold. For that reason, the CS scaffolds doped with bioactive elements are believed to be as noble candidates for bone tissue engineering aided by their adjustable biodegradation and bio-inductivity [15]. Researchers proposed that as the MnO 2 content rises in glass, the density has contradictory behavior with molar volume. Such finding was correlated to the larger atomic mass of the MnO 2 compared to another element, which contributes to the increasing trend of density glassy matrix [16].
The purpose of the present study was to incorporate Mn into the calcium silicate system and to assess the effect of Mn on the glass structure, composition, and morphology. Calcium silicate with several concentrations of MnO 2 was set up through the melt quench method. The addition of manganese oxide diverse from 0.25 to 2.00 wt%. The produced MnO 2 /calcium silicate glasses have described expressions of morphology, composition, and compressive strength. Description of the glasses was attained utilizing X-ray diffraction (XRD), while the diverse functional groups had investigated via FT-IR spectrophotometer. In the meantime, the crystallinity and nanoparticle nature had been practiced by employing a high-resolution transmission electron microscope (HR-TEM). Density had been restrained by an automatic helium automatic gas pycnometer.  GmbH 5657 Germany, Type S1, Volt 220/50 Hz] to obtain a powder of fine particle size. A melting step was as we aimed to obtain a glass-ceramic, not a ceramic material. Batches were sintered at 1200°C for 1 h (at a heating rate of 5°C/min). Glass powder had molded into discs of 1.00 cm diameter, using 7% polyvinyl alcohol (PVA) as a binder; at 20 kN for 20 s to measure the porosity and mechanical properties of the prepared samples.

Characterization techniques
Crystalline phases of sintered glass ceramics were inspected using x-ray diffraction analysis via Empyrean diffractometer system with Cu-Ka radiation of k = 0.15418 nm. The objective of the XRD was to assess the phase purity and growth of crystalline phases and to consider the nano-size of the prepared samples. X-ray diffractometer functioned at 40 kV and 40 mA over a 2h range of 15°-70°. The crystalline phases had distinguished by comparing the diffraction patterns of the fabricated composites with Joint Committee on Powder Diffraction Standards (JCPDS) standards.
To yield identical disks, fine powder was assorted with potassium bromide (KBr) in a ratio of 1:100, before the mixtures were subjected to a load of 5 tons/cm 2 in an evocable die. Fourier transform infrared spectroscopy (FT-IR) was achieved to pattern the diverse functional groups using the FT-IR spectrophotometer [Type A FT/IR-6100, Jasco, USA]. FT-IR spectra were in the range of 4000-400 cm -1 .
As well, the nanoparticle nature and crystallinity were experienced by a high-resolution transmission electron microscope (HR-TEM, Joel model JEM-2100, operating voltage 200 kV, Japan). The aqueous dispersion of the particles was drop-casted onto a copper grid coated with carbon before being air dried at room temperature to be microscopically scanned. The average size of 40 particles was determined using the ImageJ program for each sample.
The morphology and microcrystalline structure of samples were verified by a field emission scanning electron microscope (FE-SEM, model FEJ Quanta 250 Fei, the Netherlands) at an operating voltage of 15 kV. Earlier to the SEM measurements, the samples' surfaces were coated with gold utilizing a [S150A sputter coater, Edwards, England] under 0.1-Torr vacuum 1.2-kV voltage, and 50mA current. This coating was to improve the scanning of samples.
Diffused reflection measurements had been carried out at room temperature for nano-powders using Jasco-V-570 optical spectrophotometer, Japan. The diffuse spectra were measured in the spectral range from 250 to 2000 nm.
The mechanical properties were studied by measuring the uniaxial compression test for all prepared nano-powders, molded into disks of diameter & 11.7mm and length & 20mm using a uniaxial pressure machine (Paul-Otto Weber Maschinen-und. Apparatebau, Germany; 20 kN/mm 2 ). Concerning the influence of MnO 2 on the mechanical properties of the prepared disks, the ultimate force (N), ultimate stress (MPa), and break force (N) were measured using a maximum force of 25kN/5000 lb universal testing machine (Tinius Olsen 25ST, Honeycrock Lane, Salfords Surrey RH15DZ, England). An automatic helium gas pycnometer (Ultrapyc 1200e [2014], Quantachrome, USA) was used to determine the porosity at room temperature [17].
The room-temperature magnetic hysteresis loop of the prepared nano-powder has been measured using a vibrating sample magnetometer (VSM), Lake Shore -7410, USA. (Westerville, OH, USA). This tool has maximum field strength of 31 kOe with output stability better than ± 0.05%, absolute accuracy better than 1% of reading ± 0.2%, and has ranged from 0.1 9 10 -6 up to 10 3 emu of moment measurement.
However, the room-temperature absorption spectra of electron spin resonance (ESR) were performed at X-band (9.685 GHz) using EMX spectrometer (Bruker, Germany). ESR is considered a useful tool to investigate the structure and dynamics of the molecular systems of unpaired electrons and In addition, the highest intensive peak is located at 2h = 34.149°for card no. (79-1089) and at 2h = 33.624°f or card no. (78-0364) as seen in Fig. 1. Thus, as it is clear in Fig. 1c and d, the highest peak is located at 2h = 33.624°, which matches with card no. (78-0364) rather than card no. (79-1089). On the other hand, the sample that contains a higher percentage of MnO 2 (G2.0M) is well matched with the two cards number 74-0874 and 79-1089, remarking that the main phase is a hexagonal one. This result means that it is contrary to what could be proven for the two lower doped samples G0.5M and G1.0M. It is observed that the triclinic phase gradually decreases with increasing the percentage of MnO 2 as seen in Fig. 1b Notably, all patterns revealed as well indexed to a triclinic phase of Ca 3 (Si 3 O 9 ) (card no. 74-0874) and hexagonal low-combeite phase of Na 4 Ca 4 (Si 6 O 18 ) which matched with cards no. 79-1089 or 78-0364 dependent on the dopant percentage. This result indicates that there is no diffraction line detected for the MnO 2 dopant, i.e., the MnO 2 has no obvious influence on the host structure, which confirms that the manganese ions take interstitial occupancy inside the hosting calcium silicate materials [24]. The particle size indicated by the symbol (D) of all powder samples is evaluated from the highest intensive peaks of two phases (triclinic and hexagonal) using the Scherer equation [25]: where b is the full-width at half-maximum, k is a constant that denotes the shape coefficient, and k is the wavelength of the used X-ray radiation. The calculated particle size D of the two phases revealed that the prepared samples are formed in nano-size as presented in Table 2.
HR-TEM is one of the important tools through which it is possible to obtain important information about the size, shape, and structure of the particles of the prepared samples through its micrographs and electron diffraction patterns (EDP), respectively. The HR-TEM micrograph images of G0.0M, G0.25M, and G0.5M are depicted in Fig. 2a, c, and f. These micrographs show that there are different sizes of particles and there is no specific shape for them. Electron diffraction patterns (EDP) are exposed in Fig. 2b, d, e, and g for G0.0M, G0.25M, and G0.5M, respectively. The EDPs show bright spots arranged either in regular hexagonal shapes (Fig. 2d, e) or in rows intersecting with each other as revealed in (Fig. 2b) or arranged in the form of circles as shown in Fig. 2g. As it was proven in a previous part of the X-ray results that there is more than one crystallographic phase of the prepared samples, the EDP (Fig. 2d), and its invert elucidated in Fig. 2e of the G0.25M sample reveal a regular shape of crystallographic hexagonal phase, which represents one of the constituent phases. Figure 3a and c illustrates the micrograph images of the G1.0M and G2.0M samples and their EDP was clarified in Fig. 3b and d. The micrograph of G1.0M sample (Fig. 3a) reveals clear separated particles with a semi-circular shape, while Fig. 3c clarifies the micrograph of the G2.0M sample. The EPDs showing a bright spot also were arranged in parallel lines (Fig. 3b) and as a hexagonal shape (Fig. 3d). From the micrographs, the particle size was calculated taking into account the average size of 40 particles for each sample. The values of the calculated average particle size falls within the range of 7.8-27.9 nm as collected in Table 2. Based on the results obtained by studying X-rays and HR-TEM, it can be confirmed that the prepared samples were formed in the range of nano-size.
Field emission scanning electron microscope FE-SEM images by two magnifications (5 and 1 lm) of the studied nano-powder of representative samples G0.0M, G0.25M, and G2.0M are seen in Fig. 4. The surface of the samples were not treated chemically, but was coated with a thin-film layer of Gold. The powder showed sharp edges accumulated grains seen in Fig. 4a and b for undoped powder (G0.0M) of   In sample G2.0M some smaller and nearly spherical grains are seen which seem to contain nanoparticles with still some nano-rods observed (Fig. 4f). It could be observed that adding MnO 2 to pure powder has an effect on showing smaller nearly spherical grains of accumulated nanoparticles in the surface morphology of these samples.

FT-IR absorption spectra
The FT-IR absorption spectra of base powder (G0.0M) and with the content of MnO 2 from 0.25 to 2.0 wt% measured in the wavenumber range 400-1600 cm -1 are presented in Fig. 5 [26,27], while the absorption band at 908 and 850 cm -1 is related to the Si-O-Ca stretching vibration [28,29]. The band at 715 cm -1 , assigned for Si-Ovibration, increased gradually in intensity and shifted to a higher wavenumber at 721 cm -1 with the rising of the MnO 2 content [29]. The absorption band

Optical properties
The study of the optical properties of glass ceramics based on wollastonite has not received sufficient attention from researchers, and it is almost very limited [32]. The optical properties of the prepared sintered glass-ceramic powders were studied by measuring the diffuse reflection, and the absorption coefficient was obtained, which in turn contributed to determining the optical bandgap. The diffuse reflection where F(R) is equivalent to the absorption coefficient (a) and their spectra are illustrated in Fig. 6b for all samples. The optical band gap (E g ) is analyzed from the Tauc relation [17]: where the exponent m value is contingent on the electronic transition type since it takes the values of 1/2 and 2 for direct and indirect allowed and 3/2 and 3 for direct and indirect forbidden transitions, respectively. The best linear fit of the straight line is obtained at m = 1/2 (i.e., for direct allowed transition). Figure 7 shows the relation of the square power of the product of the photon energy and the absorption coefficient (y-axis) versus photon energy (x-axis). The E g values are obtained from the intersection of the extension of the straight line with the x-axis. The determined values of E g are 3.9, 2.5, 2.35, 2.07, and 1.6 eV for G0.0M, G0.25M, G0.5M, G1.0M, and G2.0M nano-powders, respectively (Fig. 7). A red shift was observed in the optical bandgap (E g ) as the percentage of MnO 2 increased since it decreased from 3.9 to From the optical results, it could be distinguished that the values of both F(R) and E g of the nanopowders are strongly influenced by the presence of MnO 2. There is an increase in the values of F(R) intensity taken at wavelength 500 nm and a decrease in E g with the increase in the MnO 2 content from zero to 2.0 wt% as displayed in Fig. 8. This result is consistent with what was previously mentioned [24], as it is attributed to the breakdown in the silicon network due to the production of non-bridging oxygen's caused by MnO 2 defects. In this case, MnO 2 acts as a structural network modifier of glass ceramics, since their defects play an important role in the physical properties of glassy ceramic materials [24].

Mechanical properties
The mechanical compressive strength is measured to identify the behavior of the prepared nano-powders after applying a compressive load, to determine the extent they bear the compressive force, and to find out the value at which breakdown takes place. The Fig. 6 The plot of (a) reflection and (b) absorption spectra for the prepared nano-powders Fig. 7 Tauc plot of the prepared nano-powders measurement results are denoted in Fig. 9, which shows the uniaxial compressive strength values against the MnO 2 content, displaying a notable increase in their values with increasing MnO 2 concentration. This result indicates that the addition of manganese oxide led to a significant improvement in the mechanical properties of the nano-powders under test.
Although calcium silicate is considered one of the brittle materials that do not take a high load until the fracture is caused, it was observed that adding some percentages of manganese oxide increased the sample load bearing to reach a fracture or deform elastomer stage [35]. Once the compression test is measured, a complete condensation of the pores is made in the porous materials and they become nonporous.
In addition, the prepared nano-powders porosity and bulk density are measured. Table 3 displays the change in their values concerning MnO 2 content and the measured Ultimate uniaxial compressive strength. This collective data refers to that the bulk density generally increased, while the porosity values decreased in parallel with the rise in the percentage of MnO 2 . Since it is scientifically known that when density increases, it is normal for the porosity of the material to decrease and which leads to an increase in the material's compressive tolerance (i.e., increase in its compressive strength) [17,36]; this gives perfection in their mechanical properties. Moreover, some studies [36,37] have exposed the relationship between compressive strength and porosity governed by several factors related to the pore characteristics, for instance, pore size, shape, and distribution.
From this result, it could be concluded that the addition of MnO 2 to glass ceramics led to a marked improvement in its mechanical properties, which opens new fields of applications for these materials. In 2021 it was mentioned by Xiong et al., that the mechanical presentation is decisive for lattice constructions used in orthopedic applications since they have to simultaneously retain high strength and porosity [38].

Magnetic investigation
Microwave irradiation with a certain frequency is used to irradiate nano-powders doped with varying percentages of MnO 2 while sweeping the external magnetic field. There is a net absorption of energy when the external magnetic field meets the resonance condition, which is measured using electron spin resonance (ESR), and the obtained result is analyzed using the first derivative of the absorption spectrum [39].
It is known that manganese (Mn 2? ) has a nuclear spin I = 5/2 as a magnetic element with an electron configuration of 3d 5 and produces sextet splitting lines of hyperfine (hf) interaction [40]. Also, it was previously proved that the ESR absorption spectra depend strongly on the Mn content [39,41] since the hf interaction appears in the lower Mn concentration  [41] even though it disappeared at higher concentrations.
The first derivative of energy absorption spectra of ESR measured at room temperature for G0.25M, G0.5M, G1.0M, and G2.0M nanocrystalline samples are depicted in Fig. 10. The obtained ESR signal for each sample reveals a sextet lines of hyperfine splitting (hfs) interaction overlap on a broad line as viewed in Fig. 10. Furthermore, with increasing the MnO 2 content, the hyperfine interaction signals have a slight decrease, particularly in the G2.0M sample. This is attributed to the fact that the splitting hyperfine interactions (hfs) are predominant in samples with low Mn concentration due to the Mn 2? electron cloud and nucleus interactions. The hfs was previously detected at a low concentration (B 0.15Mol %) of MnO 2 -doped lead phosphate glasses [42] and 0.5-mol% MnO 2 -doped piezoelectric ceramics [43].
However, at higher concentrations, the Mn 2? -Mn 2? exchange dipolar interactions cause the hfs to vanish and merge into a broad resonance signal [41,44], which may be attributable to the ferromagnetic (FM) resonance. This FM resonance originated as a result of the transition inside the ground state of ferromagnetic domains [45]. The broad signal is a result of the reduction in the atomic distance of Mn 2? -Mn 2? and increasing their dipolar interactions resulting from increasing Mn content inside materials [45][46][47].
The splitting Lande g-factor signify a significant key factor to recognize the ESR signal, and this is evident that the unpaired electron has a Lande gfactor that leads to determining the position of the ESR signal at a certain value of the magnetic field strength [41]. The g-factor can be estimated using the next formula [40,48]: where H is the magnetic field in Millitesla (mT), b is the Bohr magneton, m is the radiation frequency (in GHz), and h is the Planck's constant (6.62607015 9 10 -34 JÁHz -1 ).
For the samples under study, the calculated g-factor of a broad signal is presented in Fig. 11, and it has values ranging from 2.01534 to 2.01731 as documented in Table 4. Commonly, the classical value of the Lande g-factor is about 2.0000 and it mostly has a possible error of 0.0002-0.0005, as well it has a value of 2.0023 for free electron [40,49]. The obtained gvalues reveal some deviation from that of the free electron by a value ranging from -0.015 to -0.013 as recorded in Table 4; this deviation is owing to spinorbit interaction [40]. The obtained values of the spectroscopic Lande g-factor are nearly consistent with that previously reported ones [50,51]. Worthily, the importance of determining the Lande g-factor is also represented in the fact that its value leads to knowing and determining the type of bonding in the material since the negative and positive shifts of its value with respect to the value of the free electron (2.0023) pointed to the formation of ionic and covalent bonds, respectively [52,53]. Therefore, the covalent bond becomes stronger with increasing the deviation of the g-factor toward the positive direction and becomes ionic if it has a negative shift [53,54]. In the present work, the determined g-factor demonstrates a negative shift ( Table 4) from that of the free electron and then the ionic bond is predominant.
To ensure the bonding type, the hyperfine splitting constant (A) is calculated from the following formula [55,56]: where D Opp À Á and D Ott denote the difference between peak to peak and trough to trough, respectively, for the first and sixth peak positions. Consequently, D Mpp and D Mtt symbolize the differences between the second and fifth peak position, whereas D Ipp and D Itt represent the difference between the third and fourth peaks. The estimated value of the constant A is found to be 86.98 9 10 -4 cm -1 , demonstrating the existence of an ionic bond between Mn 2? and its neighbors. This result is consistent with that previously reported [52,57,50], where the ionic bonds were identified when the hyperfine splitting constant was determined in the range from 87 9 10 -4 to 90 9 10 -4 cm -1 .
The Mn 2? oxidation state was formerly mentioned by Srisittipokakun et al. [51] for MnO 2 -doped sodalime silicate glass which was characterized by two resonance signals, one at g = 2.0 with a six hyperfine interaction and the other of less intensity positioned at g = 4.3. Similarly, Zhou et al. [50] reported that the sextet lines of hfs is generated from unpaired electron and Mn 2? nucleus interaction for the Mn-doped ceramics, and they proved that the Mn has an oxidation state of Mn 2? . Furthermore, the number of free electrons per unit volume can be calculated from the following equation: where k is a constant equal 10 13 , H o is the field at the center of signal, DH is the difference of magnetic field between peak maximum and minimum of the signal, A is the height of the peak, H m , ffiffiffiffiffiffi P H p , and G are the modulation frequency, power, and gain, respectively. Table 4 provides the calculated density of free spins for the nano-powders doped with MnO 2 .
Room-temperature magnetic hysteresis loops of the prepared nano-powders displayed in Fig. 12. The curves illustrate that the prepared nano-powders behave like the diamagnetic materials, although the effect of the presence of manganese oxide appeared in the ESR study, but here no effect detected.
However, the observation that should be taken into account is that the moment values of the prepared nano-powders gradually decreases with the increase in the proportion of the manganese oxide, and it indicates that there is a tendency to rotate toward a positive direction. This appears clearly in the (G2.0M) Fig. 11 The spectroscopic Lande g-Factor of the nano-powders doped with different percentages of MnO 2 sample, since its curve is nearly close to the magnetic field axis (x-axis). But in order that a complete rotation to positive direction happens clearly using VSM instrument, a higher percentage of manganese oxide ([ 2.0 wt%) must be added. The hysteresis curve of G2.0M is expanded in the inserted panel in Fig. 12 and indicates a very weak ferromagnetic behavior around the origin point. Earlier Lakshmi et al. considered the weak ferromagnetic behavior of the 2% concentration of Mn-doped nano-zinc sulfide (ZnS) and they stated that the broad signal of ESR was ascribed to the ferromagnetic resonance and that the VSM measurements disclosed at room temperature for the ZnS:Mn was very profound for ferromagnetism at 2% concentration of Mn [45]. From the obtained results of ESR and VSM, it is inferring that the ESR technique is most sensitive to detect the magnetic properties of the small percentage of materials doped with manganese oxide. Previously, Abdel-Hameed et al. decided that rising the percent of MnO 2 appended to the glass ceramic would ease the mobility of different ions accordingly initiating a relatively larger crystallite size as embossed from XRD and TEM results [58].

Conclusion
Addition of MnO 2 to glass ceramic improves most of the studied parameters such as its mechanical and magnetic properties that enhance their use in the orthopedic and various applications. The XRD and HR-TEM revealed nanostructure crystalline samples with triclinic Ca 3 (Si 3 O 9 ) as main phase and the secondary phase of hexagonal Na 4 Ca 4 (Si 6 O 18 ), with crystallite size less than 85 nm. Addition of MnO 2 content (%) has a great effect which was shown on the FT-IR spectra. The wavenumbers of both stretching and bending vibrations of O-Si-O band shifted to higher wavenumber by introducing MnO 2 to calcium silicate as detected by FT-IR spectra. Moreover, the addition of MnO 2 led to a significant enhancement in the mechanical properties of the tested nano-powder. The mechanical investigations revealed lowering the porosity and enhancement of bulk density and the compressive ultimate strength by the addition of MnO 2 . The optical study showed a remarkable decrease of the band gap value with introducing the MnO 2 because of the invention of non-bridging oxygens, which guides the breakdown of the SiO 4 network structure. The magnetic properties showed an enhancement by investigating using ESR and VSM studies by introducing 2-wt. % MnO 2 to the calcium silicate glass-ceramics. The presence of the ionic bond between Mn 2? ions and their neighbors was ascertained from the assigned negative shift of the spectroscopic Lande g-factor and the determined value of the hyperfine splitting constant.
contributed to conceptualization, resources, and reviewing and editing of the manuscript. HHAS contributed to data curation, formal analysis, investigation, and writing of the original draft.

Funding
Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB).

Data availability
All data generated or analyzed during this study are included in this article.

Declarations
Conflict of interest The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licen ses/by/4.0/.