Introduction

In the textile industry, coloured fabrics are generally produced using the application of dyestuffs (Santamaría Razo et al. 2008). Many production processes using dyestuffs pollute the environment (Paul 2011). Structural colouration has been suggested as a solution to this problem (Hyoki et al. 2009, Gao et al. 2016a, b). Structural colours are also known as physical colours (Dufresne et al. 2009, Hyoki et al. 2009). Examples of structural colours can be observed in the natural world, for instance, the colour of some insect wings and bird feathers (Zi et al. 2003, Paul 2011, Gao et al. 2016a, b). In order to produce structural colour, many researchers have produced artificial photonic crystals using a range of materials (Xu 2000). Photonic crystals (PCs) were first proposed by John and Yablonovitch in 1987 (Yablonovitch 1987, Galisteo-Lopez et al. 2011). A PC is an artificial periodic structure with photonic band gaps (PBG) (Braun et al. 2001, Joannopoulos 2001, Gao et al. 2017a, b). Photonic crystals can preferentially reflect particular wavelengths of light by controlling photon propagation (Joannopoulos 2001). Structural colours are produced by artificial opals by a process of light diffraction (Paul 2011).

Self-assembled nanostructured materials have an interaction with particular wavelengths of light (Dufresne et al. 2009; Galisteo-Lopez et al. 2011). These materials provide a platform for a variety of applications in the photonics research field. In recent years, several articles have been published that introduce and suggest the advantages of photonic crystals used for the colouration and coating of materials such as textiles (Gao et al. 2017a, b). For instance, Shang et al. proposed that under external magnetic stimulation, a one-dimensional chain-like photonic crystal fibre structure can be formed by producing Fe3O4@C superparamagnetic colloidal nanocrystal clusters (SCNCs) embedded into a polyacrylamide matrix. In addition, in the chain-like fibre structures, the distance between each particle can create a reversible colour change because of the elastic deformation of the fibre. Owing to this reversible change, the chain-like fibre structures can show brilliant colours ranging from red to green. As these fibres are conductive, they can also be used as sensors to detect deformation (Shang et al. 2015). Yuan et al. suggested a new structural colouration technique to fabricate structurally coloured fibres using electrospinning. A monodispersed P(St-MMA-AA) solution of nanospheres and poly(vinyl alcohol) (PVA) were mixed. The PVA acted as a binding material to adhere the particles together. Green (220 nm), red (246 nm) and violet (280 nm) structurally coloured fibres were produced using this method (Yuan et al. 2015). Apart from producing colloidal fibres, researchers have also studied the fabrication of colourful photonic crystal structures on fabrics, such as cotton, polyester and nylon fabrics (Gao et al. 2017a, b). Liu et al. applied a vertical deposition self-assembly method to fabricate monodisperse P(St-MAA) colloidal particles on polyester fabrics. The colours produced were brilliant and vivid on the surface of fabrics (Liu et al. 2015a, b). Gao et al. used the gravitational sedimentation method to coat fabrics with silica nanoparticles (SNPs). This produced silica photonic crystals on the surface of cotton and nylon substrates. The samples were dried at 60 °C in the oven and structural colours were successfully produced (Gao et al. 2017a, b).

In this investigation, the solvent varying technique (SVT) suggested by Gao et al. (Gao et al. 2016a) has been used to create batches of “original” silica nanoparticles (O-SNPs). These O-SNP solutions have then been further processed using a solvent replacement and centrifuge technique to produce centrifuged silica nanoparticles (C-SNPs) (Lu and Owens 2018). Black polyester fabric was coated with SNPs using the gravitational sedimentation method. The self-assembly of SNPs produced coloured PC fabrics. The morphology of the SNPs on the surface of the polyester fabrics was recorded using a scanning electron microscope (SEM). The changes in particle size and the colour differences between fabrics coated using the original and centrifuged SNPs have been discussed. This paper reports that centrifuging the original SNP solutions can significantly improve the chroma of the coated fabrics. Three different batches of ethanol were used to create the O-SNPs in this paper (with concentrations of 99.97, 99.96 and 99.99%) and these batches were measured using the following techniques: transmission electron microscopy (TEM) (batch 1), scanning electron microscopy (SEM) (batch 2), dynamic light scattering (DLS) and spectrophotometry (batch 3).

Experimental

Materials

Tetraethyl orthosilicate (TEOS, 99.0%, Fisher Scientific Co., Ltd., UK); batch 1 ethanol (EtOH, 99.97%, VWR Science Co., Ltd.); batch 2 ethanol (EtOH, 99.96%, VWR Science Co., Ltd.); batch 3 ethanol (EtOH, 99.99%, Fisher Scientific Co., Ltd., UK); ammonia hydroxide (NH3·H2O, 25%, Fisher Scientific Co., Ltd., UK); distilled water (dispensed from USF-ELGA water purifier in the laboratory). All reagents were applied as received without any further purification. Black plain-woven polyester fabrics were sourced from a local fabrics market.

Synthesis of silica nanoparticles

The SNPs were produced following the solvent varying technique (SVT) suggested by Gao et al. The synthesis of the nanoparticles was performed in one 250-ml round-bottomed flask. The SNPs are created via two reactions, i.e. hydrolysis and condensation polymerization. Firstly, a mother solution was prepared by mixing 8 ml ammonia hydroxide, 3 ml distilled water and a certain initial volume of ethanol in the 250-ml flask. The flask was submerged in the water bath. The water bath was placed on a hotplate. Magnetic stirrers were placed both in the flask and the water bath. When the reaction temperature reached 60 °C, TEOS (6 ml) was added into the flask. The flask was sealed with silicone grease to mediate the temperature throughout the experiment. Stirring was maintained during this procedure. The whole reaction temperature was maintained at 60 °C for a duration of at least 3 h. After 3 h, the final solution was removed from the flask and transferred into a clean glass bottle.

Post processing of the silica nanoparticles

The SNP sample solutions were centrifuged (Thermo Scientific Sorvall Legend × 1 Centrifuge) twice with the speed of the centrifuge set at 14500 r/s. SNP sample solutions were centrifuged for 10 min. Firstly, the original colloidal suspensions (25 ml) were added into centrifugal tubes and then the tubes were inserted into the centrifuge. Once centrifuged, the silica nanoparticles collected at the bottom of the tube. The supernatant was removed and replaced with 25 ml of EtOH. The tubes were then placed into an ultrasonic water bath (Ultrawave U95) for 30 min to ensure the silica nanoparticles were adequately dispersed in the EtOH.

Deposition of SNPs onto black polyester fabrics

In this investigation, the black polyester fabrics were cut into circles with a diameter of 25 mm which would cover the bottom of a glass tube of the same diameter. The fabrics were washed using deionized water. This reduced the impact of the original dyestuffs and any remaining impurities on the sedimentation of the SNPs. Post washing, the fabrics were dried in the oven. The washed fabrics were placed into the transparent glass tubes and target volumes of colloidal suspensions were added. The samples were then dried in the oven. Four pieces of SNP-coated polyester fabric were coated using 3 ml of the same batch of original SNP solution. The coated fabric samples were dried in a Gallenkamp Hotbox oven at 40 °C, 60 °C, 80 °C and 100 °C, respectively. The corresponding centrifuged SNP solution was used to coat four pieces of woven polyester fabric and the procedure was then repeated in the same way as for the original SNP solution.

Characterization of SNPs and PCs

The SNPs were characterized using TEM, SEM and colorimetry. Micrographs were captured from a TEM (CM20) and an SEM (Zeiss Ultra 55). Reflectance measurements were obtained from a spectrophotometer (Datacolor SF 650). The morphology of the SNPs was investigated using TEM (CM20) and SEM micrographs. The diameters of a random sample of particles were measured using ImageJ software (National Institute of Health, USA) (Schneider et al. 2012). The average particle diameters and the colloidal dispersion in solution were measured using a DLS instrument (Malvern Zetasizer Nano S). The reflectance properties of the SNP-coated polyester fabrics were measured using a DataColor SF650 spectrophotometer.

Results and discussions

A comparison between the original and centrifuged SNPs

Figure 1a, b, c show TEM micrographs of the original SNPs. It can be observed that there are impurities on the surface of the particles and these impurities affect the final particle diameters, as measured by the DLS instrument, and observed colour quality. This suggests that not all of the chemicals have been exhausted during the reaction using the SVT recipes. Figure 1d, e, f show the morphology of particles after being centrifuged twice. A comparison of the micrographs between the original particles and the centrifuged particles suggests that after centrifuging, impurities were removed. Since the impurities have been removed from original colloidal suspensions, the average diameters of centrifuged particles has also been reduced. In Fig. 1, the average diameter of the original SNPs was 246.70 nm whilst the average diameter of the processed particles (C-SNPs) was approximately 221.10 nm. These differences between the particle diameters, due to the unused reactants, produce photonic crystals with a different morphology and hence a different colour appearance. This has been quantified and discussed in the “Ordered structure of SNPs on black polyester fabrics” section.

Fig. 1
figure 1

TEM images of SNPs made using 70 ml batch 1 ethanol (99.97%), magnification, a and d × 10,000, b and e × 15,000, c and f × 27,500. Micrographs 1 a–c show particles from the original SNP solution and micrographs d–f show particles from the centrifuged solution

Full size image

Application of centrifuged SNPs to polyester fabric

The comparisons in the “A comparison between the original and centrifuged SNPs” section are based on the diameter of silica spheres and the following discussions are about the effect of particle size on colour appearance. Figures 2 and 3 show two black polyester fabrics coated using original Fig. 2a and centrifuged Fig. 3a SNPs, respectively. Making a visual comparison between these two images, it can be seen that the black polyester fabric coated using the centrifuged SNPs produced a highly reflecting surface which is more chromatic (ΔC* = 6.29) than that produced using the original SNPs. The different colours of the fabric samples shown in Figs. 2a and 3a can be easily discriminated (ΔE*(ab) = 19.82).

Fig. 2
figure 2

a Image of polyester fabric (25-mm diameter) coated using the original SNPs (IPhone 6 Plus). SEM micrographs (b–d): micrographs of polyester fabrics coated with the original SNPs produced using 75 ml of batch 2 EtOH (99.96%) (blue-green appearance), b × 1000, c × 2000, d × 5000

Full size image
Fig. 3
figure 3

a Image of a polyester fabric (25-mm diameter) coated using centrifuged SNPs (iPhone 6 Plus). SEM micrographs (b–d): centrifuged SNPs produced using 75 ml of batch 2 ethanol (99.96%) (green appearance), b × 2000, c × 5000, d × 30,000

Full size image

Figure 2b–d show micrographs of the morphology of the original SNPs on the black polyester fabrics as imaged from directly above. Figure 3b–d show the SEM images of the morphology of the black polyester fabrics coated with the centrifuged SNPs. When comparing the original and centrifuged SNPs at the same magnification (Figs. 2d and 3c), it can be observed that there is a reduction of the surface contaminants which coat the photonic crystals. The photonic crystal produced from the centrifuged SNPs on the black polyester fabrics are more systematically ordered than the original SNPs. An example of face-centred cubic (fcc) structures on the surface of the coated fabric is indicated by a red hexagon (the (111) planes) (Fig. 3d)

Ordered structure of SNPs on black polyester fabrics

Five recipes were produced using the SV method and these solutions provided uniform (PDI < 0.1) colloidal batches of SNPs, as shown in Table 1. From the DLS measurements of the SNP solutions, it can be observed that for both the original and centrifuged SNPs the average particle diameter decreases with increasing ethanol and this relationship is consistent with that reported by Gao et al. (Gao et al. 2016a, b). The diameter differences between the original and centrifuged particles also reduce with increasing volumes of initial ethanol.

Table 1 Average diameters of O-SNPs and C-SNPs (DLS results)
Full size table

Discussions in the “A comparison between the original and centrifuged SNPs” section were based on a comparison of particle diameters of the SNPs pre- and post-processing. The “Ordered structure of SNPs on black polyester fabrics” section discusses the effect of applying the centrifuged SNPs to circular pieces of black polyester fabric. The optimum sedimentation conditions such as the oven temperatures on the colour appearance of the PCs on the surface of black polyester fabrics are investigated. The SNPs were synthesized using 40 ml, 45 ml, 55 ml, 65 ml and 85 ml of initial ethanol. These solutions were deposited onto black polyester fabrics (diameter 25 mm) in glass tubes (also with a diameter of 25 mm). For each batch of SNPs (3 ml), four different oven drying temperatures (40 °C, 60 °C, 80 °C and 100 °C) were used. Similarly, the corresponding centrifuged SNP solutions were also applied to the surface of the black polyester fabrics and dried at the same four temperatures.

According to previously published results (Gao et al. 2017a, b), the SNPs which were created using 40 ml and 45 ml of initial ethanol did not show much colour or appeared white. However, in this investigation, post centrifuging, the black polyester fabrics coated using the centrifuged SNPs did present colours, as shown in Fig. 4. According to the chroma difference data in Table 2, colours of the coated fabrics created using the centrifuged SNPs are more chromatic than that of fabrics coated using the original SNPs. The higher chromaticness of these colours can also be observed in the reflectance spectra (Fig. 5). A narrower band of wavelengths around the peak reflectance value of the fabrics coated using the centrifuged SNPs is observed unlike the flatter reflectance spectra of the fabrics coated using the original SNPs. The reflectance spectra of the coated fabrics from the original batches of SNPs show a wider wavelength bandwidth around the peak wavelength of reflectance. Figure 5 also indicates that the fabrics coloured using centrifuged SNPs synthesized, using 40 ml and 45 ml of initial ethanol, produce bright, chromatic colours with C*(ab) values of 20.44 and 22.82 respectively (both samples dried at 80 °C).

Fig. 4
figure 4

Black polyester fabrics coated using the original and centrifuged SNPs and dried at different temperatures (40 °C, 60 °C, 80 °C and 100 °C). Each fabric was coated using 3 ml of the SNP solution (batch 3 ethanol, 99.99%)

Full size image
Table 2 Chroma differences (ΔC*) of coated fabrics (comparison of C-SNPs to O-SNPs)
Full size table
Fig. 5
figure 5

Reflectance spectra of SNP-coated polyester fabrics using original and centrifuged SNPs which were created using 40 ml (a) and 45 ml (b) of the batch 3 ethanol (99.99%) and dried at 40 °C, 60 °C, 80 °C or 100 °C

Full size image

In the “A comparison between the original and centrifuged SNPs” section, post centrifugation, the SNP diameters were reduced compared to the SVT O-SNPs and the colour appearance of the coated fabrics consequently changed. Figure 4 illustrates that the black polyester fabrics coloured using the original SNPs which were created using 40 ml, 45 ml and 55 ml of initial ethanol show a slight red hue. However, the fabrics coated using the same centrifuged SNPs display red-orange colours. In a similar trend, the fabrics coated using O-SNPs which were produced using 65 ml and 85 ml of ethanol show green and green-blue colours respectively, whilst the corresponding C-SNPs produce yellow-green fabrics and green fabrics respectively. The colour differences between the original and centrifuged SNPs are shown in Table 3, ΔE*(ab) values. The ΔE*(ab) values between fabrics coated using the original and centrifuged particles are large (> 5). The coloured fabrics created using the original and centrifuged SNPs show different colours.

Table 3 Colour differences (ΔE*(ab)) of coated fabrics (comparison of C-SNPs to O-SNPs)
Full size table

Table 4 indicates that the differences in the hue of the fabrics coated using the original and centrifuged SNPs reduced with the decrease in the particle diameter of the synthetic particles. Theoretically, the reduction in particle diameters should cause the particles to show the colour of the smaller particles; for example, if the original particles show yellow, then they should show yellow or green after centrifugation; however, the experimental results show that this is not the case. This is because the impurities in the original SNP solutions form white films on the surface of the fabrics. These films cover the surface of coloured fabrics which leads the coloured fabrics appearing less saturated. This phenomenon can be observed from the light reflectance spectra shown in Figs. 5 and 6. The reflectance spectra of fabrics coated using the original SNPs were closer to neutral reflectance spectra than those of the fabrics coated using the centrifuged SNPs, especially the SNPs created using 40 ml, 45 ml and 55 ml of ethanol (Fig. 7). Therefore, the diameters of the original particles do not provide a reliable estimate of the colour using Bragg’s law. These samples appear to have a white film covering the PC that diffusely scatters the incident light. The corresponding centrifuged SNPs are colourful in comparison.

Table 4 Hue angles of coated fabrics for O-SNPs and C-SNPs
Full size table
Fig. 6
figure 6

Reflectance spectra of the coated fabrics coated using original and centrifuged SNPs which were created using 55 ml (a) and 85 ml (b) of the batch 3 ethanol (99.99%), respectively. The samples were dried at 40 °C, 60 °C, 80 °C or 100 °C

Full size image
Fig. 7
figure 7

CIE a*b* plot of a* (x-axis) and b* (y-axis) of SNP-coated fabrics using the original and centrifuged SNPs (created using 40 ml, 45 ml and 55 ml of batch 3 ethanol (99.99%))

Full size image

In addition to the colour changes, the brightness of SNPs on fabrics also increases post centrifugation. In Fig. 5, the peak wavelengths of the coated coloured fabrics using the centrifuged SNPs are visibly sharper than those of the coated fabrics using the original SNPs. The influence of the purity of the dried particles on the colour appearance, but also the effect of temperature on the colour of the particles has been investigated. Figures 5 and 6 demonstrate that according to the peak wavelength, it can be seen that for the centrifuged SNPs created using 40 ml and 55 ml of ethanol, 100 °C produces the highest reflectance value at peak wavelength. Similarly, 80 °C is the suggested drying temperature for centrifuged SNPs produced using 45 ml of ethanol. 100 °C is the suggested drying temperature using 85 ml of ethanol as these conditions both lead to the highest peak wavelength values.

The temperature not only affects the colour appearance of the fabrics coated with SNPs but also affects the spatial uniformity of the colouration. Figure 4 shows that the original and centrifuged particles do not completely cover the surface of the fabrics when dried at 40 °C and the particles collect at the centre of the fabrics. The first column of fabrics, shown in Fig. 4, do not have a spatially uniform coating of SNPs. When the optimum temperature is reached, the particles will adhere evenly to the surface of the fabrics by natural gravity sedimentation. Figure 8 shows a representation of the PC formation. This process is affected by the SNP diameter. There is a negative relationship between particle diameter and the initial volume of ethanol. With a decrease of particle diameter, it is observed that the particles tend to accumulate in the middle of the surface of the fabrics (Fig. 4). It can be observed that in the recipes with an initial ethanol volume of 40 to 85 ml, the particles gradually accumulate in the centre of fabrics when the drying temperature is 40 °C.

Fig. 8
figure 8

Process of gravity sedimentation. Fabrics were uniformly coated with the SNPs and the darker red areas above represent the deposition of less SNPs on the fabric post drying

Full size image

Conclusions

A colorimetric comparison of photonic crystals created from original and centrifuged SNPs has been presented. The post-processing of the SNPs produced using only the SVT technique produced smaller particles with less variation in the average particle diameter in all cases. The centrifuge and replace technique provides more uniform PCs that produce brighter colours when coating polyester fabric. Drying the original SNP solutions produced from the SV technique produces PCs that display desaturated colours. This appears to be due to the presence of unused reactants on the surface of the dried SNPs. The centrifuge and replace solvent technique can effectively improve the quality of SNPs, not only in the final arrangement of the photonic crystals but also in the removal of impurities from the surface of the SNPs. The reduction of unused reactants allows the colloidal crystal films created on the surface of fabrics by gravity sedimentation to be more uniform and more vibrant in colour (chroma increases on average 15 C* units) when compared to fabrics coated with the original SNPs. The impact of drying temperature on the colour appearance of the fabric samples using the centrifuged SNPs shows little variation between 60 and 100 °C. The centrifuged particles present bright colours when dried at temperatures between 60 and 80 °C which is higher than previously reported. The optimum temperature to produce spatially uniform photonic crystals from the centrifuged SNPs varies between solutions based on the target particle diameter.