Characteristics of X-ray attenuation in electrospun bis­muth oxide/polylactic acid nanofibre mats

The characteristics of the X-ray attenuation in electrospun nano(n)- and micro(m)-Bi₂O₃/polylactic acid (PLA) nanofibre mats with different Bi₂O₃ loadings were compared as a function of energy using mammography (i.e. tube voltages of 22–49 kV) and X-ray absorption spectroscopy (XAS) (7–20 keV). Results indicate that X-ray attenuation by electrospun n-Bi₂O₃/PLA nanofibre mats is distinctly higher than that of m-Bi₂O₃/PLA nanofibre mats at all energies investigated. In addition, with increasing filler loading (n-Bi₂O₃ or m-Bi₂O₃), the porosity of the nanofibre mats decreased, thus increasing the X-ray attenuation, except for the sample containing 38 wt% Bi₂O₃ (the highest loading in the present study). The latter showed higher porosity, with some beads formed, thus resulting in a sudden decrease in the X-ray attenuation.

1. Introduction

Nanoparticles, i.e. nanometre-sized particles, have attracted much attention amongst researchers in different fields of physics, chemistry, material science, medicine and biology because of their unique and often superior electronic, magnetic, optical, mechanical, physical and chemical properties (Patra et al., 2010; Huang & El-Sayed, 2010; Granmayeh Rad et al., 2011; Sahare et al., 2007; Lines, 2008; Haiwen et al., 2006; Popov, 2009). For example, in the medical field, nanoparticles have been widely used in diagnosis, tissue engineering and also as drug delivery devices (Storrie & Mooney, 2006). Gold nanoparticles are one of the most useful nanoparticles in industry and medicine (Granmayeh Rad et al., 2011). For instance, in the medical field, gold nanoparticles show significant improvement in the treatment of cancers by enhancing the sensitivity of radiation from a radiotherapy unit with minimal adverse effects on surrounding normal tissues (Patra et al., 2010; Huang & El-Sayed, 2010).

Additionally, this size effect has also become one of the virtues in designing materials for shielding of ionizing radiations. Some X-ray technologists believe that this effect will improve the X-ray attenuation ability of composites since nano-sized fillers can be dispersed more uniformly within a matrix with fewer agglomerations compared with micro-sized fillers and consequently affect the density and composition that modify the total attenuation coefficient of the composite (Botelho et al., 2011; El Haber & Froyer, 2008). The latest work by Buyuk et al. (2012) proved that decreasing the titanium diboride particle size in titanium diboride reinforced boron-carbide–silicon-carbide composites results in a higher linear attenuation coefficient for the energy of 0.662 MeV emitted by a Cs-137 gamma source (Buyuk et al., 2012). In a complementary finding, a recent study by Botelho et al. (2011) showed that nanostructured copper oxide (CuO) is more effective in attenuating lower X-ray beam energies (tube voltages of 26 and 30 kV), whilst no significant variation in the X-ray attenuation at higher energies (tube voltages of 60 and 102 kV) were observed. Künzel & Okuno (2012) also provided similar results, showing that the X-ray beam attenuation is greater for a nanostructured CuO compound compared with its microstructured counterpart for low energies (tube voltages of 25 and 30 kV) for a wide range of CuO concentrations incorporated into polymeric resins.

Electrospinning is a well established polymer processing technique which has proven to be a flexible and effective method for fabricating multilayers of microscale (>1 µm) to nanoscale (<1000 nm) fibres from different types of polymers used in a wide range of applications, such as in drug delivery, tissue engineering and protective clothing (Faccini et al., 2012; Prabhakaran et al., 2008; Sill & von Recum, 2008; Russo & Lamberti, 2011; Huang et al., 2012; Haroosh et al., 2012; Yiin-Kuen & Li-Chih, 2013; Rajeswari et al., 2012; Hu et al., 2010; Yu et al., 2009). This technique provides many benefits to industry with perhaps the most important being its versatility and simplicity, which means it is a very time efficient way to fabricate a variety of continuous nanofibrous structures. It is advantageous to use nanofibre webs in a layered structure together with a suited substrate material such that the final product offers sufficient strength and durability. Besides, the nanofibre layers should be flexible and also have a good adherence to the substrates without easily being broken or delaminated (Faccini et al., 2012; Lee & Obendorf, 2007; Brettmann et al., 2012). Moreover, some researchers have shown that electrospinning can improve the dispersion of nanoparticles within the polymer matrix thereby improving the properties of nanocomposites (Demir et al., 2004; Shanshan et al., 2010).

Polylactic acid (PLA) fibres are polymeric in nature and provide the inherent performance of fibres together with the positive environmental advantages of being renewable and recyclable. In addition, production of PLA emits less CO₂ compared with other petroleum-based fibres. Disposal of PLA not only fits within the existing disposal systems but also includes the additional option of composting (Farrington et al., 2005). Meanwhile, Bi₂O₃ is non-hazardous and is also a relatively environmentally friendly compound that is used as a substitute for toxic lead oxide which is widely used in the production of lead glass. For example, Bi₂O₃-based glasses have gained fascination among glass researchers because of their non-linear optical properties which are important for the development of optical information processing technology (An et al., 2006). Thus, they have played an important role in the replacement of lead glass by radiation shielding glass (Chanthima et al., 2011).

In the present study, given the simplicity of the electrospinning technique, it is investigated whether electrospinning can also be used to produce nanofibre mats for the efficient shielding of ionizing radiation. In a recent work on WO₃-filled epoxy composites (Noor Azman et al., 2013), we investigated the X-ray attenuation effect of nano- and micro-sized fillers in epoxy composites over an X-ray tube voltage range of 22–127 kV. Our results showed that nano-sized WO₃ was more effective than its micro-sized counterpart in attenuation of low X-ray tube voltages (22–35 kV). This is explained by the domination of photoelectric interactions at low photon energy and also the number of W particles per gram in the nano-sized WO₃–epoxy composite being greater than that of the micro-sized WO₃–epoxy composite. The size effect was not apparent in the higher X-ray tube voltage range of 40–120 kV. Hence, the objective of our present study is to synthesize new radiation shielding materials using the electrospinning technique with the preparation of well dispersed Bi₂O₃ of different particle sizes in PLA fibre mats. The effectiveness of electrospun nano-Bi₂O₃/PLA nanofibre mats in radiation protection during diagnostic imaging using low X-ray energies is also reported, with the ultimate goal of offering a new approach to radiation protection, based on nanotechnology and the electrospinning technique to produce composites that are environmentally friendly.

2. Experimental procedure

2.1. Materials

Bismuth (III) oxide (Bi₂O₃) particles of sizes 90–210 nm and 10 µm were used as a filler for synthesizing electrospun Bi₂O₃/polylactic acid (PLA) nanofibre mats and Bi₂O₃/PLA thin films. Bi₂O₃, chloroform and methanol were obtained from Sigma-Aldrich. Meanwhile, PLA (3051D) pellets, with molecular weight M_n = 93500 g mol⁻¹ and glass transition temperature T_g = 338.50 K were supplied by NatureWorks USA.

2.2. Sample preparation

Electrospinning was carried out using 9% wt/v PLA solution by mixing with 8 ml of chloroform and 2 ml of methanol as solvents. The micro(m)-Bi₂O₃ and nano(n)-Bi₂O₃ suspension was added at 24–38 wt% to the polymer solution and was homogenized for 45 min under ultrasonication. For the electrospinning process, the solutions were transferred to a 10 ml syringe pump with 25-G needles. The flow rate of the polymer solution was 1 ml h⁻¹, and the applied positive voltage was ∼19 kV. The distance between the needle tip and the target was set at 12 cm. The resulting nanofibre mats were collected on a flat aluminium foil over ∼2 h to achieve an acceptable thickness for X-ray attenuation experiments. The nanofibre mats were removed from the aluminium foil and cut to a dimension of 2.0 cm × 1.5 cm or were folded together before cutting so that they had an acceptable thickness for a reliable X-ray attenuation experiment. Three sets of the same nanofibre mats were prepared. A list of prepared electrospun Bi₂O₃/PLA nanofibre mats with different Bi₂O₃ weight percentages are shown in Table 1.

Table 1
Prepared electrospun Bi₂O₃/PLA nanofibre mats with different weight fractions (wt%) of filler (Bi₂O₃) and PLA

Filler (Bi2O3) (wt%)	PLA (wt%)
24	76
28	72
34	66
38	62

The solution casting method was performed to prepare Bi₂O₃/PLA thin films for verifying the X-ray attenuation results obtained for the electrospun Bi₂O₃/PLA fibre mats. In this method, PLA was mixed with chloroform without methanol (since methanol was only used in the electrospinning process to increase the conduction of the solution and will totally evaporate during the process) of the same amount as in the electrospinning process and was homogenized for 45 min under ultrasonication. Then, the solution was poured into a 5 cm-diameter beaker and left in a fume cupboard for 24 h to dry. Next, the thin film was removed from the beaker and three sets of the same thin film were cut into 2.0 cm × 1.5 cm pieces for X-ray attenuation experiments. The list of prepared Bi₂O₃/PLA thin films with different weight percentages of Bi₂O₃ are the same as those shown in Table 1.

2.3. Measurements of sample thickness and porosity

Since it is inherently difficult to produce all the electrospun fibre mats with a uniform thickness, which is an important factor in X-ray attenuation comparison, the nanofibre mat thickness for each sample was measured by determining the weight and surface area. The average thickness (t_ave) of the mats was then determined from equation (1), where m is the mass, A is the surface area and ρ is the apparent density of the electrospun mats,

The apparent density (ρ) was accurately measured using the density bottle method (ASTM D854) and an average of three measurements were taken for each mat. The porosity of the mats was calculated using equation (2). A calibrated single-pan electrical balance and distilled water were used for this purpose,

where ρ_theory is the bulk density of the Bi₂O₃/PLA composite.

Meanwhile, for the thin films the average thickness t_ave was measured using a vernier caliper. The apparent density (ρ) of the thin films was calculated using (1) with m and A being the film mass and surface area, respectively.

2.4. Measurement of X-ray attenuation

Two separate methods were used to characterize the X-ray attenuation of the samples. The first involved the use of the X-ray absorption spectroscopy beamline at the Australian Synchrotron. Experiments were carried out in the energy range 7–20 keV using a Si (311) monochromator, a beam size on the sample of about 0.25 mm × 0.25 mm and a photon flux of about 10⁹ photons s⁻¹. Transmission data were collected using ionization chambers before and after the sample. For each sample, 20 readings were recorded at each energy. To normalize the data, readings were also recorded for beam passing through the air.

The second method used a mammography unit (Siemens AG, model 2403951-4, GE Health Care) at Royal Perth Hospital, Western Australia. For the work with this mammography unit, the exposure was set at 10 mAs to obtain meaningful readings for the DIADOS diagnostic detector connected to the diagnostic dosimeter (PTW-Freiburg, Germany), and the X-ray tube voltage range 22–49 kV was selected. The dosimeter was a universal dosimeter for measuring simultaneous dose and dose rate for radiography, fluoroscopy, mammography, dental X-ray and computed tomography with a sensitivity of 0.01 microRoentgen (µR). Three different anode/filter combinations (Table 2) were used for filtering the X-ray beam produced by the chosen X-ray tube voltages used for the mammography machine, since the combination was controlled by the machine itself. The X-ray beams generated by these anode/filter combinations were mainly of the characteristic X-ray energies of molybdenum (17.5 keV and 19.6 keV) or rhodium (20.2 keV and 22.7 keV). For each sample the measurements were performed three times. The detector was placed 86 cm below the X-ray tube since this is the maximum distance that can be adjusted for the mammography unit, and the X-ray beam was collimated to the size of the sample to minimize the scattered X-rays produced by the sample.

Table 2
Anode/filter combination operated by the mammography machine

X-ray tube voltage (kV)	Anode/filter combination
22	Mo/Mo†
25
30
35	Mo/Rh‡
40
45	Rh/Rh§
49

†Molybdenum anode/molybdenum filter.
‡Molybdenum anode/rhodium filter.
§Rhodium anode/rhodium filter.

The incident intensity I₀ and transmitted intensity I (normalized to the case without a sample) were measured; the X-ray transmission (T = I/I₀) is related to the linear attenuation coefficient (μ) through equation (3),

Hence, equation (3) can be re-written as

where the X-ray transmission T is determined by μ which is the total linear attenuation coefficient of a PLA nanofibre mat or PLA thin film with Bi₂O₃ particles and pores expressed as (μ_{mat or film})x + (μ_Bi2O3)y + (μ_pores)z, where x + y + z = 1 is the weighted fraction contributions of the individual components.

A graph of the X-ray linear attenuation coefficient (μ) as a function of X-ray energy was plotted for each sample.

2.5. Scanning electron microscopy (SEM)

The depth profile of the samples was examined using a Zeiss Evo 40XVP scanning electron microscope at a voltage of 15 kV with the working distance between 8.0 and 9.0 mm. Both secondary electron and backscattered electron techniques coupled with an energy-dispersive X-ray spectroscopy (EDS) probe were performed after standard coating with platinum to minimize charging in order to show the difference between the images due to the different atomic number of PLA and Bi. A pure electrospun PLA nanofibre mat was also examined as a benchmark.

3. Results and discussion

3.1. Thickness and porosity measurement

The average thicknesses t_ave for all the electrospun Bi₂O₃/PLA nanofibre mats measured were in the range 0.006–0.014 cm while for Bi₂O₃/PLA thin films they were in the range 0.04–0.08 cm. All of these t_ave measurements will be used in equation (4) to calculate the value of μ for each sample.

From Table 3, the apparent density ρ of each thin film (the last two columns in the table) of the same filler (Bi₂O₃) category increased by the increment of filler content in the PLA solution. The density found in the thin films illustrates that m-Bi₂O₃/PLA thin films have a lower density compared with n-Bi₂O₃/PLA thin films though neither are significantly different from their theoretical value. However, the apparent density of the electrospun Bi₂O₃/PLA nanofibre mats underestimated the theoretical values due to the nanofibres being highly porous, randomly oriented and aligned. As can be seen, the density of the electrospun Bi₂O₃/PLA nanofibre mat increased with the filler loading except for the electrospun Bi₂O₃/PLA nanofibre mat of 38 wt% Bi₂O₃ loading; there is a sudden decrease in the density due to the higher porosity found (Table 4) and also the formation of PLA beads.

Table 3
Density of electrospun Bi₂O₃/PLA nanofibre mats and Bi₂O₃/PLA thin films

	Density (g cm^−3)
Filler (Bi2O3) weight		Electrospun nanofibre mat		Thin films
fraction (wt%)	Theoretical	Nano	Micro	Nano	Micro
24	1.56	0.27	0.25	1.54	1.50
28	1.63	0.40	0.37	1.59	1.52
34	1.75	0.49	0.45	1.69	1.64
38	1.84	0.30	0.24	1.77	1.72

Table 4
Porosity of electrospun Bi₂O₃/PLA nanofibre mats

	Porosity (%)
Filler (Bi2O3) weight fraction (wt%)	Nano-Bi2O3/PLA nanofibre mat	Micro-Bi2O3/PLA nanofibre mat
0	88.8	88.8
24	83.0	84.0
28	75.4	77.0
34	72.1	74.3
38	83.6	87.0

Porosity is an important parameter when preparing the absorbing material for X-ray attenuation experiments. As can be seen from Table 4, the porosity of both electrospun n- and m-Bi₂O₃/PLA nanofibre mats was >70%. This is likely to be caused by the entangled structure of the randomly oriented nanofibres, indicating that they were highly porous and thus not ideally suited for X-ray attenuation, especially with the filler ≥38 wt%. Further increases to filler loading, beyond 38 wt%, was not performed owing to the decreased density and increased porosity, which is unlikely to give any advantages for future X-ray transmission experiments. The porosity found from the control electrospun PLA nanofibre mat was 88.8%. Thus, if a further investigation was performed for greater filler loadings the electrospun Bi₂O₃/PLA nanofibre mat would probably have similar or even higher porosity than this controlled sample.

3.2. X-ray attenuations

Fig. 1 shows the XAS linear attenuation coefficient (μ) results of electrospun nanofibre mats for an X-ray energy of 7–20 keV for 0–38 wt% Bi₂O₃ loading. It clearly shows a big difference in μ between electrospun n-Bi₂O₃/PLA nanofibre mats and electrospun m-Bi₂O₃/PLA nanofibre mats at the same filler loadings as the X-ray energy increased. Additionally, Fig. 2 also shows a distinct difference in μ between electrospun n-Bi₂O₃/PLA nanofibre mats and electrospun m-Bi₂O₃/PLA nanofibre mats at the same filler loadings as the X-ray energy increased, which is related to the mammography X-ray effective energy operated at an X-ray tube voltage of 22–49 kV (Fig. 3). These effective energies were determined using half-value layer (HVL) experiments (Suk et al., 2012) on the mammography unit. With kilovoltage X-rays, determination of the HVL of the X-ray beam can be used to characterize the effective energy by converting the measured HVL to the linear attenuation coefficient. The effective energy of a polyenergetic beam is equal to the energy of a monoenergetic X-ray beam that is attenuated at the same rate as the polyenergetic beam which can be determined from tabulated data (Berger et al., 2010).

Figure 1 Linear attenuation coefficient as a function of synchrotron radiation energy operated by XAS (7–20 keV) for all Bi2O3 loading (0–38 wt%) of the electrospun n-Bi2O3/PLA and m-Bi2O3/PLA nanofibre mats.

Figure 2 Linear attenuation coefficient as a function of effective energy operated by various X-ray tube voltages of the mammography unit (22–49 kV) for all Bi2O3 loading (0–38 wt%) of the electrospun n-Bi2O3/PLA and m-Bi2O3/PLA nanofibre mats.

Figure 3 Effective energy as a function of the X-ray tube voltages operated by the mammography unit determined from half-value layer measurements.

As can be seen from both Figs. 1 and 2, μ increased with an increase of the filler loading within the PLA matrix for both electrospun n-Bi₂O₃/PLA and m-Bi₂O₃/PLA nanofibre mats, except for the 38 wt% Bi₂O₃ loading where there is a sudden decrease in μ. However, the difference in μ between electrospun n-Bi₂O₃/PLA and m-Bi₂O₃/PLA nanofibre mats becomes higher as the Bi₂O₃ loading increased. These findings support the density and porosity results discussed previously, including the 38 wt% of Bi₂O₃/PLA nanofibre mats which showed low density and high porosity, thus leading to decreased μ. Meanwhile, a comparison of thin films between m-Bi₂O₃/PLA and n-Bi₂O₃/PLA does not totally support the results found from this study for the electrospun Bi₂O₃/PLA nanofibre mats. From Fig. 4, as the X-ray effective energy increased to >17.4 keV (i.e. X-ray tube voltages > 35 kV), for the same wt% of Bi₂O₃ filler within this thin film sample, μ for m-Bi₂O₃/PLA thin film and μ by n-Bi₂O₃/PLA thin film become comparable. They only show significant differences in μ for lower X-ray effective energy, <17.4 keV (i.e. X-ray tube voltage 22–35 kV), operated from the mammography unit.

Figure 4 Linear attenuation coefficient as a function of the effective energy operated by various X-ray tube voltages of the mammography unit (22–49 kV) for all Bi2O3 loading (0–38 wt%) of the n-Bi2O3/PLA and m-Bi2O3/PLA thin films.

As in our previous study for the comparison of different sizes of WO₃ particles–epoxy composites (Noor Azman et al., 2013), we also obtain similar results for electrospun Bi₂O₃/PLA nanofibre mats which showed that the attenuation by electrospun n-Bi₂O₃/PLA nanofibre mats is higher than the attenuation by electrospun m-Bi₂O₃/PLA nanofibre mats with the same filler loading in the low-energy regime (22–35 kV operated from the mammography unit). However, by increasing the X-ray tube voltage beyond 35 kV (i.e. X-ray effective energy > 17.4 keV), the attenuation by electrospun n-Bi₂O₃/PLA nanofibre mats is still higher than the attenuation by electrospun m-Bi₂O₃/PLA nanofibre mats. In contrast, in our previous work, the differences in the attenuation by micro-sized WO₃–epoxy composites and nano-sized WO₃–epoxy composites become indistinguishable when the X-ray tube voltage was increased beyond 35 kV (Noor Azman et al., 2013). Only Bi₂O₃/PLA thin film has a good agreement with our previous results for X-ray tube voltages greater than 35 kV, which shows the indistinguishability in attenuation between m-Bi₂O₃/PLA thin film and n-Bi₂O₃/PLA thin film.

Fig. 5 presents the value of μ for electrospun n-Bi₂O₃/PLA nanofibre mats for all the X-ray beam energies generated by XAS and the mammography unit. The results found from XAS were correlated with those from the mammography unit since the mammography unit produced effective energies of 15–21 keV with the Mo and Rh anode/filter characteristic X-ray energies of 17.5–22.7 keV, while the X-ray energy used with XAS was ∼7–20 keV.

Figure 5 Comparison of the linear attenuation coefficient for the electrospun n-Bi2O3/PLA nanofibre mats for all X-ray beam energies generated by XAS and the mammography unit.

In essence, the total X-ray attenuation by the absorbing material is determined by three energy-dissipative mechanisms, namely the photoelectric effect, Compton scattering and pair production. Pair production was not considered in this study because this mechanism will only occur when the photon energy is beyond 611 keV. In the photoelectric absorption process, a photon undergoes an interaction with an absorber atom in which the photon completely disappears. In its place, an energetic photoelectron is ejected from one of the bound shells of the atom. The photoelectric process is the predominant mode of photon interaction at relatively low photon energies and high atomic number Z, i.e. (Z/E)³. Meanwhile, Compton scattering takes place between the incident photon and one of the outer-shell electrons of an atom in the absorbing material. The probability of Compton scattering is almost independent of atomic number Z and X-ray energy E. It is most dominant as the photon energy increases due to a concomitant decrease in the photoelectric effect.

In addition, the number of Bi particles per gram in both n-Bi₂O₃/PLA fibre mats and thin films is higher than their micro-sized counterparts. Hence, the probability of X-rays with lower energies (i.e. 7–20 keV and 22–35 kV) interacting and being absorbed by n-Bi₂O₃ filler is higher when compared with their micro-sized counterparts since the photons interact with the absorbing materials mainly by the photoelectric effect.

As a consequence, electrospun n-Bi₂O₃/PLA nanofibre mats are superior to their micro-sized counterparts in terms of X-ray attenuation for all the X-ray beam energies (i.e. 7–20 keV and 22–49 kV generated by XAS and mammography, respectively). In contrast, n-Bi₂O₃/PLA thin films are a good X-ray shielding candidate only for the mammography unit at 22–35 kV when compared with m-Bi₂O₃/PLA thin films. Both can be chosen as X-ray shielding materials for voltages greater than 35 kV. The observed similarity in the attenuation results for PLA thin films of n-Bi₂O₃ and m-Bi₂O₃ at the X-ray tube voltage of the mammography unit of more than 35 kV may be attributed to: (a) the decrease of the photoelectric effect, and (b) the domination of the Compton scattering effect when the photon energy increases. The latter effect results in less interaction and absorption of the photons by Bi particles of these thin films; thus the X-ray attenuation of the films is similar. In contrast, electrospun n-Bi₂O₃/PLA nanofibre mats are superior to their micro-sized counterparts in terms of X-ray attenuation at the same energy range. The observed large differences in X-ray attenuation of these fibre mats may be explained by the difference in uniformity of dispersion between nano- and micro-sized Bi particles within the PLA matrix where large agglomerations tend to occur in the latter, thus affecting the final density of the composite.

Hence, the electrospun n-Bi₂O₃/PLA nanofibre mats of all filler loadings (24–38 wt%) are potential candidates in X-ray shielding for all the incident X-ray energies studied either by XAS (7–20 keV) or mammography (22–49 kV) when compared with their micro-sized counterparts. However, the latter may still be a suitable candidate for X-ray attenuator of scattered radiation which requires lower energy. These electrospun Bi₂O₃/PLA nanofibre mats can be used as a coating material for X-ray shielding because PLA nanofibres provide many benefits such as higher tenacity, resistance to degradation and mechanical properties (Farrington et al., 2005; Haroosh et al., 2012). Besides, since PLA has a higher tenacity, these electrospun Bi₂O₃/PLA nanofibre mats can be fabricated as gloves to be worn by radiation workers when holding radioactive materials or waste (non-clinical use), or used as a liner on examination tables to help attenuate scattered X-rays coming out from patients during medical diagnosis or therapy (using higher X-ray or γ-ray energies in the MeV range) performed by specialist radiographers and radiologists (clinical use).

Meanwhile, n-Bi₂O₃/PLA thin films are good X-ray shielding candidates only for mammography at 25–35 kV, when compared with m-Bi₂O₃/PLA thin films. Both can be chosen as X-ray shielding materials for X-ray tube voltages greater than 35 kV.

3.3. Microstructure analyses

Microstructure analysis of the electrospun nanofibres was performed using both backscattered electrons and secondary electrons to clearly show the difference between the PLA polymer and the embedded Bi elements. Fig. 6(a) shows a SEM image of the control electrospun PLA nanofibres without any Bi₂O₃ filler. The average diameter of these fibres was 854 ± 35 nm. The average diameter of PLA nanofibres increased when 24–34 wt% Bi₂O₃ was added to the PLA solution. For instance, Fig. 6(b) shows the homogeneous nanofibres of 28 wt% n-Bi₂O₃ filler with an average diameter of 971 ± 22 nm. Meanwhile, in Fig. 6(c) the average diameter of 28 wt% m-Bi₂O₃/PLA nanofibres is 911 ± 41 nm with a large variation in fibre diameter and particle agglomerations can be observed.

Figure 6 SEM images using, for each figure, the backscattered electron technique (left) and the secondary electron technique (right) for (a) the control electrospun PLA nanofibres without any particles (0 wt% of Bi2O3); (b) 28 wt% Bi2O3 of electrospun n-Bi2O3/PLA nanofibres; (c) 28 wt% Bi2O3 of electrospun m-Bi2O3/PLA nanofibres; (d) 38 wt% Bi2O3 of electrospun n-Bi2O3/PLA nanofibres; and (e) 38 wt% Bi2O3 of electrospun m-Bi2O3/PLA nanofibres.

This indicates that, by increasing the solution viscosity while both the conductivity and the surface tension decreased with the increment of filler (Bi₂O₃) loadings within the PLA solution, the average nanofibre diameter was increased. The significant increase in the viscosity of the solution with increasing filler loadings was due to the increased molecular entanglement which enabled the charged jet to withstand a larger stretching force (from the Coulombic repulsion) resulting in the coarsening of the nanofibres.

In contrast, a further increase of Bi₂O₃ loading to 38 wt% caused a decrease in the average diameter to 496 ± 32 nm for n-Bi₂O₃/PLA nanofibres which have a maximum diameter of 2.06 ± 0.49 µm and a minimum of 114 ± 2 nm (Fig. 6d). Meanwhile, the average diameter of m-Bi₂O₃/PLA nanofibres is 406 ± 71 nm with a maximum diameter of 1.13 ± 0.23 µm and a minimum of 111 ± 2 nm (Fig. 6e). The unexpected decrease in the average fibre diameter may be attributed to the domination of electrical conductivity when filler loading or viscosity of solution increases. As a result, this leads to the production of fibres with non-uniform diameters, as well as formation of beads and particle agglomerations.

Figs. 7(a)–7(c) illustrate the EDS results for the electrospun n-Bi₂O₃/PLA nanofibre mat with 34 wt% filler loading which confirms the existence of the Bi element from the filler together with elements C and O from the PLA matrix. SEM images of m-Bi₂O₃/PLA and n-Bi₂O₃/PLA thin films with 28 wt% filler loading are shown in Fig. 8, indicating the presence of agglomerations within the PLA matrix.

Figure 7 (a) SEM image used for EDS analyses on 34 wt% Bi2O3 of the electrospun n-Bi2O3/PLA nanofibres to prove that only Bi particles are detected other than C and O which is the composition of PLA. (b) EDS analyses for point 1 and (c) for point 2 marked in (a).

Figure 8 SEM images of a thin film of 28 wt% Bi2O3 for m-Bi2O3/PLA on the left (clear agglomerations can be seen) and n-Bi2O3/PLA on the right.

4. Conclusions

Electrospun nanofibre mats of n-Bi₂O₃/PLA and m-Bi₂O₃/PLA with filler loadings of 24–38 wt% have been successfully fabricated. From the analyses, the electrospun n-Bi₂O₃/PLA nanofibre mats of all filler loadings were found to be superior in attenuating X-rays compared with their micro-sized counterparts because n-Bi₂O₃ may provide more uniform materials since the particle size can affect the microstructure and consequently the density and composition that will then modify the attenuation coefficient of the composite. However, the electrospun Bi₂O₃/PLA nanofibre mats with 38 wt% loading are not recommended for X-ray shielding because of higher porosity compared with the lower filler loadings. The n-Bi₂O₃/PLA thin films are good X-ray shielding candidates only for the mammography unit at 22–35 kV when compared with the m-Bi₂O₃/PLA thin films. The particle size effect on X-ray attenuation diminished as the X-ray tube voltage exceeded 35 kV.