Silk fibroin as a biotemplate for hierarchical porous silica monoliths for random laser applications

Moliria V. Santos *a, Édison Pecoraro a, Silvia H. Santagneli a, André L. Moura b, Maurício Cavicchioli a, Vladimir Jerez c, Lucas A. Rocha d, Luiz Fernando C. de Oliveira e, Anderson S. L. Gomes f, Cid B. de Araújo f and Sidney J. L. Ribeiro a
aInstitute of Chemistry, São Paulo State University (UNESP), Araraquara, SP 14801-970, Brazil. E-mail:
bNúcleo de Ciências Exatas – NCEx, Universidade Federal de Alagoas, Campus Arapiraca, Arapiraca, AL 57309-005, Brazil
cGrupo de investigación FIELDS, Universidad de Investigación y Desarrollo, Bucaramanga, Colombia
dFranca University – UNIFRAN, CP 82, Franca, SP 14404-600, Brazil
eNúcleo de Espectroscopia e Estrutura Molecular, Centro de Estudos de Materiais, Universidade Federal de Juiz de Fora, Juiz de Fora, MG, Brazil
fDepartamento de Física, Universidade Federal de Pernambuco – UFPE, Recife, PE 50670-901, Brazil

Received 8th August 2017 , Accepted 17th January 2018

First published on 17th January 2018

Bombyx mori silk fibroin offers unlimited opportunities for functionalization, processing, and biological integration. We describe in this work the design of structured organic inorganic hybrids based on silica and silk fibroin, taking advantage of the relationship between the structure and processing of the latter. In situ self-assembly of fibroin nanofibers along with hydrolysis and polycondensation of tetraethyl orthosilicate was employed. Structural characterization was performed by Raman and solid state NMR spectroscopies. Our findings demonstrated that fibroin precipitates in the reaction medium with prevailing β-sheet conformation. The transition from amorphous to crystalline state was observed to be favored by the increase of the fibroin concentration in the samples. The samples were obtained as robust and biocompatible monoliths, making them candidates for several applications, particularly in the biomedical field. As a novel development, the fibroin nanofibers were used as pore biotemplates to create mechanically robust silica monoliths with a hierarchical macro-mesoporous network in an easy templating process. The template was removed by thermal treatment and the as obtained silica based materials displayed surface area values ranging from 704 to 1057 m2 g−1 and a maximum pore volume of 0.621 mL g−1. The porous silica monoliths were then doped with rhodamine 6G and typical random laser action could be observed, with a minimum laser threshold of 9.7 μJ per pulse and a linewidth narrowing from 40 to 4 nm. In addition, it was shown that two coupled gain mechanisms were taking place, the random lasing and the stimulated Raman scattering, allowing us to observe Raman Stokes lines due to vibrational modes of the dye molecule.

1. Introduction

In nature, biopolymers play the role of controlling agents or templates for precipitation and crystallization of inorganic phases. Besides, they may also assemble into structures that serve as confining spaces or scaffolds in which the formation of the inorganic component takes place.1 From a materials science perspective, biopolymers can be a versatile source for the design of structured inorganic or inorganic/organic materials in the laboratory.2,3

Silk proteins represent a unique family of biopolymers due to their structural and biological properties. Fibroin extracted from the cocoons of the silkworm (Bombyx mori) is mainly composed of 43% glycine (Gly), 30% alanine (Ala) and 12% serine (Ser),4 and the structure consists of two chains interconnected by disulfide bonds.5 The chain with higher molecular weight (∼350 kDa) is formed by repetitive domains Gly-Ala-Gly-Ala-Gly-Ser and polyelectrolytes such as glutamic acid and aspartic distributed at the end of the chains. Jointly with a large number of hydrophobic residues, this chain has hydroxyl residues (serine and tyrosine) that promote its affinity for water.6

Silk fibroin may present an amorphous (random coil, water soluble) or crystalline (β-sheet, water insoluble) conformation, which allows its stability in water to be modulated.7 Stabilization can be further enhanced when the amorphous structure of silk is converted into the crystalline state8 through external stimuli, such as fibroin concentration, treatments with organic solvents, temperature, pH value, ion strength or mechanical stress.9 All these features have led to the transformation of this commodity material into a variety of new materials including hydrogels, ultrathin films, thick films, conformal coatings, 3D porous or solid matrices, fibers with diameters spanning the nano- to the macroscale and many related morphologies.9–13

In parallel, hierarchical porous materials are of high interest for applications where optimizing mass transport is an issue, as in sorption media, separation processes, heterogeneous catalysis, photocatalysis and filters, among others.14–17 In addition, these porous materials can also be used to optimize light diffusion, scattering and amplification, since light amplification through stimulated emission is possible in a random medium with a gain.18

The interplay between gain and scattering in disordered materials can sustain laser oscillations, named random lasers (RLs), which have been studied in a diverse range of materials, as reviewed in ref. 18–20, including dye colloids, porous materials21,22 and silk fibroin based materials.22,23 Ethanol solutions of rhodamine dyes containing TiO2 nanoparticles have been widely studied, mainly due to their simplicity, important scattering by the high refractive index particles and high gain provided by the dyes. One of the main drawbacks with liquid samples is the high probability of particle agglomeration, which results in degradable RLs,23 due to the precipitation of the scatterer particles. This problem can be mitigated by properly modifying the scatter nanoparticle, as recently demonstrated using specially designed TiO2 nanoparticles24 or engineering the nanoparticle as demonstrated by Jimenez-Villar et al.25

RLs are promising for applications in image systems, thanks to the low spatial coherence that allows speckle-free images to be obtained.26 Solid samples are required because precipitation of particles is avoided, they can be designed in several geometries, and they have decreased health and environmental hazards. In this context, structured inorganic or organic–inorganic hybrid (OIH) materials are promising for RL applications.27,28

In this work, structured OIH materials based on silica and silk fibroin were designed. The new materials were obtained through the in situ self-assembly of fibroin nanofibers along with hydrolysis and polycondensation of tetraethyl orthosilicate (TEOS). The alcoholysis and alcohol condensation reactions in TEOS solution produce, as a byproduct, ethanol, which induces the precipitation of fibroin, due to the much lower solubility of the protein in such solvent. Fibroin was observed to precipitate with prevailing β-sheet conformation. The amorphous to crystalline state transition was favored by the increase of the relative fibroin concentration. The samples were obtained as robust and biocompatible monoliths, making them a candidate for several applications, particularly in the biomedical field.29–33

We have used these materials to prepare porous silica monoliths for RL applications. In a templating approach, mesoporous and macroporous silica monoliths were obtained by using fibroin nanofibers as sacrificial scaffolds that were selectively removed from the monoliths via thermal treatment of the OIH. The porous silica monoliths were afterwards doped with rhodamine 6G (Rh6G) and the as obtained materials exhibited typical RL characteristics such as the nonlinear emission intensity dependence with the excitation intensity, large spectral bandwidth narrowing with the increase of the excitation intensity and excitation threshold behavior. In addition, by analyzing the emission spectra, the concurrence of two gain coupled mechanisms was shown: the random lasing and the stimulated Raman scattering (SRS), allowing us to observe Raman Stokes bands of the dye molecule. In fact, powder based random Raman lasers have recently been reported in the literature34,35 and they have been observed in porous materials such as vesicular polymeric films,36,37 silica foams38 and silica/titanate foams.39

2. Materials and methods

2.1. Extraction of silk fibroin

Fibroin solution was obtained from silk cocoons produced by Bombyx mori silkworms, which were supplied by Bratac, Fiação de Seda S.A. (Bastos/SP, Brazil). The method was based on previous reports.40 Five grams of silk cocoon pieces was washed with 2 L of 0.02 mol L−1 Na2CO3 solution at 100 °C for 30 min and washed thoroughly with distilled water, in order to remove sericin (degumming). Degummed silk (fibroin fibers) was dissolved in 9.3 mol L−1 LiBr in a proportion of 1[thin space (1/6-em)]:[thin space (1/6-em)]4 (w/v) of fibroin to LiBr and the solution heated at 60 °C for dissolution for 4 h. Each 25 mL sample of the resulting solution was dialyzed against two liters of milli-Q water for 48 h in order to remove salts. An 8.5% (w/v) aqueous fibroin solution free of impurities was obtained after the centrifugation (twice) of the dialyzed solution at 20[thin space (1/6-em)]000 rpm at 4 °C for 20 min. The final solution was stored at 4 °C before use.

2.2. Preparation of organic–inorganic hybrid monoliths based on silica and fibroin

The steps of the sol–gel process exploited in the preparation of the samples are shown in Scheme 1. The tetraethoxysilane (TEOS, Sigma-Aldrish 98%) based sol was prepared by mixing TEOS, 0.3 mol L−1 aqueous HCl (Hydrochloric acid Synth 98%) and milli-Q water in a 1[thin space (1/6-em)]:[thin space (1/6-em)]0.008[thin space (1/6-em)]:[thin space (1/6-em)]3.82 molar ratio.
image file: c7tc03560h-s1.tif
Scheme 1 A schematic illustration of the self-assembly of fibroin nanofibers inside the bulk of silica composite materials and posterior calcination of fibroin nanofibers to form porous silica.

Silica/fibroin (SF) composites were prepared by combining the sol with a mixture of 8.5 wt% aqueous fibroin suspension and 1 mL of 0.6 mol L−1 aqueous NH4OH to achieve the desired proportion TEOS[thin space (1/6-em)]:[thin space (1/6-em)]fibroin (wt%): 1.00[thin space (1/6-em)]:[thin space (1/6-em)]0.00 (SF0), 0.985[thin space (1/6-em)]:[thin space (1/6-em)]0.015 (SF1), 0.97[thin space (1/6-em)]:[thin space (1/6-em)]0.03 (SF2), 0.925[thin space (1/6-em)]:[thin space (1/6-em)]0.075 (SF3), 0.85[thin space (1/6-em)]:[thin space (1/6-em)]0.15 (SF4), respectively. The mixtures were aged for 15 days and afterwards dried for 24 h under ambient conditions forming whitish monoliths, which were dried at 40 °C for 24 h in a furnace.

2.3. Preparation of porous silica monoliths

The thermal treatment of the SF organic–inorganic hybrid monoliths led to porous silica monoliths. The heating ramp used to calcinate the organic fraction is shown in Table S1 in the ESI. Silica monoliths so obtained were designated as S0, S1, S2, S3 and S4, and are related to their respective organic–inorganic hybrid monolith precursors SF0, SF1, SF2, SF3 and SF4.

2.4. Preparation of porous monoliths of silica containing Rh6G

Silica monoliths were immersed in 2 mL of an ethanol solution of 10−4 mol L−1 Rh6G and kept for 24 h at room temperature. Finally, the silica monoliths were removed from the dye solution and treated at 40 °C for 2 hfor evaporation of the ethanol.

2.5. Characterization of materials

The morphology of the materials was evaluated by scanning electron microscopy (SEM) images obtained with a JSM-7500F field emission scanning electron microscope from JEOL. Samples were covered with a platinum layer on copper supports.

Thermogravimetric (TG) curves of dried samples were obtained in an SDT Q600 from TA Instruments. Samples were heated in open alumina crucibles from 20 to 600 °C under an oxygen atmosphere at a flow rate of 100 mL min−1 and heating rate of 10 °C min−1.

Raman spectra of the samples were collected on an RFS 100 FT-Raman Bruker spectrometer equipped with a Ge detector using liquid nitrogen as the coolant and a Nd:YAG laser emitting at 1064 nm. The laser light power was varied from 30 to 150 mW. A back-scattering configuration was used. An average of 1024 scans was performed at a resolution of 4 cm−1 over a range from 3500 to 50 cm−1. The OPUS 6.0 (Bruker Optik, Ettlingen, Germany) software was used for Raman data acquisition. For all of the FT-Raman spectra obtained in this work, the samples did not undergo any previous preparation, and all spectra were obtained at least twice to be sure about position and intensity.

Solid state NMR of 29Si{1H} and 13C{1H} cross polarization (CP) experiments were conducted with Bruker Avance spectrometers at magnetic fields of 9.4 and 17.6 T. The 29Si{1H} CP-MAS spectra of the SiO2:fibroin composites were measured under speed spinning at 10 kHz, using a CP contact time of 2.5 ms and relaxation delay of 5 s. 13C{1H} CP-MAS spectra were measured under speed spinning at 10 kHz, using a CP contact time of 1.0 ms and relaxation delay of 5 s. All spectra were acquired with TPPM proton decoupling during the data acquisition applying decoupling pulses. The 1H experiments were acquired at magnetic fields of 17.6 T under speed spinning at 14 kHz, using a rotor-synchronized spin echo sequence to remove the probe background (total echo duration of 13 μs), and a saturation train of 56 pulses. The 13C{1H} CP heteronuclear correlation (HETCOR) experiments were collected using two CP contact times 0.5 × 2.5 ms. The experiments were carried out under speed spinning at 14 kHz and a recycle delay of 1 s. Frequency-switched Lee Goldburg (FSLG)41 homonuclear decoupling at a rf nutation frequency of 90 kHz (and an offset of ±30 kHz) was applied in the indirect dimension, leading to a scaling factor of 0.4986 (measured from a 1H–1H correlation spectrum collected with identical decoupling conditions in the indirect dimension). The indirect dimension was collected with a total of 64 increments and a spectral width (before scaling) of 10.1 kHz. Chemical shifts are reported relative to the TMS referencing standard.

Surface area and porosimetry measurements were performed using the Micrometrics ASAP 2020.

The random laser action in the silica monoliths containing Rh6G was achieved using the second-harmonic of a Q-switched Nd:YAG (wavelength: 532 nm; pulse duration: 6 ns) excitation source. A biconvex lens was used to focus the beam at the sample surface with a cross sectional area of 1.6 × 10−2 cm2. The emitted light was analyzed using a monochromator equipped with a CCD camera.

3. Results

3.1. Characterization of silica–fibroin (SF) organic–inorganic hybrids

Macroscopically homogeneous silica/fibroin hybrid monoliths showed a transparent to milky appearance, depending on the fibroin content (Fig. 1a). The final fibroin content for each sample was (in weight) 3.9, 7.9, 10.2 and 15.8%, according to TGA analysis, which is shown hereafter.
image file: c7tc03560h-f1.tif
Fig. 1 (a) Photograph of silica/fibroin hybrid monoliths. (b) SEM image of the surface of monolith SF3; (c) a closer look at the surface; (d) a fracture with the fibroin nanofibers connecting the two parts can be observed; (e) a closer look at the fibroin nanofibers connecting two silica bodies; (f and g) SEM image of the fracture of monolith SF4 where bigger fibers are observed; (h) a fracture with the fibroin nanofibers connecting the two parts can be observed.

Fig. 1(b–h) show representative SEM micrographs. Fig. 1b shows the surface of silica/fibroin hybrids containing 10.2 wt% fibroin (SF3). Fibroin nanofibers with diameters of 20–30 nm emerging from the smooth silica surface are easily observed. Fig. 1c shows a closer look at the emerging fibers. The smooth surface is better observed here to be composed of spherical silica nanoparticles. Fig. 1(f and g) show the fracture of silica/fibroin hybrids containing 15.8 wt% fibroin (SF4). Bigger fibers, highlighted with yellow arrows can be observed. Fig. 1(d and e) show a fracture with the fibroin nanofibers connecting the two parts. This kind of connection is also clearly observed in Fig. 1h. The increase in the relative fibroin concentration results in higher precipitation and agglomeration of fibroin nanofibers, as shown in Fig. S1 (ESI).

Since the precipitation of fibroin fibers from water solutions does not occur unless specific processing methods are used (such as spinning, electrospinning or elongational flow in microfluidic devices),42–45 the present results suggest that hydrolysis and condensation of TEOS induce in some way the formation of fibroin nanofibers in the silica nanoparticle medium. These nanofibers cause scattering in the visible light range and lead to a gradual increase of optical density to an opaque white color with increasing fibroin concentration.

As already mentioned, fibroin can be present in different conformations with significantly different properties. The materials presenting β-sheet conformation exhibit high crystallinity, and they have high breaking strength and stiffness.46 Silk fibers have outstanding mechanical properties and they are insoluble in organic solvents and water due to the hydrophobic nature of the fibroin and the presence of a high content of β-sheet conformation. Thus, the preparation of materials with this conformation is of great interest because of the high mechanical performance that it can provide. In order to investigate the conformation of fibroin nanofibers in the silica/fibroin hybrids, thermal and structural studies were carried out using thermogravimetric analysis, Raman scattering and solid state NMR spectroscopy.

The TGA results are shown in the SI for the sake of clarity (Fig. S2 and Table S1, ESI). General features do not change from sample to sample and can in short be divided into three regions. The first region, in the temperature range of 35 to 100 °C, is attributed to the removal of physisorbed water and ethanol. The second one, in the range of 240 to 400 °C, is associated with the fibroin degradation, and the removal of ethanol and non reactant alkoxides present within closed pores. Above 400 °C, mass loss can be assigned to the dehydroxylation of silica.47

Silk fibroin degradation occurring between 313 and 318 °C is characteristic of fibroin with prevailing β-sheet conformation48 (Table S2, ESI). Qualitatively, these characteristics could be related to the higher stability induced by the crystallinity and orientated chain alignment of the β-sheet conformation.49 The results could indicate that fibroin assumes a prevailing β-sheet conformation during hydrolysis and condensation of TEOS in the samples. The same characteristics were observed by Motta et al.50 after inducing the crystallization (β-sheet formation) of fibroin films by treatment with ethanol.

Fig. 2 shows Raman scattering spectra obtained for the silica/fibroin hybrids. The pattern for all samples is characteristic of silk fibroin with prevailing β-sheet conformation.51 The amide I and amide III bands appear at 1666 cm−1 and 1233 cm−1, respectively. The band at 1085 cm−1 is a sensitive band for β-sheet conformation as well. The relative intensity of those bands compared to the lower energy bands clearly increases with the increasing fibroin relative content. Performing a deconvolution of the spectra from 1580 to 1730 cm−1, (see Fig. S4 in the ESI), attributed to stretching of the amide carbonyl groups (amide I) that are sensitive to protein conformation, it was possible to evaluate the increase of the β-sheet conformation formed.52 The intensity ratio of the bands at 1665 and 1650 cm−1, I(1665)/I(1650), increases from 1.75 to 3.87 for samples SF1 and SF2, respectively and to 18.34 for samples SF3 and SF4. These results show that the samples SF3 and SF4 are formed predominantly by β-sheet conformation. These results corroborate the TGA results and suggest that the increasing of silk fibroin concentration in the composite induces β-sheet conformation. Similar results were reported by Matsumoto et al.52 in their study on the mechanisms of silk fibroin sol–gel transition as a function of fibroin concentration. The authors associated the optical changes in fibroin gels with the precipitation of micrometer size aggregates with insoluble and crystalline β-sheet conformation.

image file: c7tc03560h-f2.tif
Fig. 2 Raman spectra of silica/fibroin hybrid monoliths: (a) SF1; (b) SF2; (c) SF3 and (d) SF4.

The silica condensation degree could be evaluated from the 29Si CP-MAS NMR spectra that are also shown in the ESI (Fig. S3). Three broad resonance lines are observed to be convoluted with chemical shifts in the range of −80 to −130 ppm, and they are assigned as Q2, Q3 and Q4 units (where Qn correspond to SiO4 tetrahedra in an amorphous network, and n represents the number of oxygen bonds to another tetrahedron).53 No important variation is observed among the samples, indicating that the degree of condensation within the silica monoliths is essentially independent of the silica/fibroin ratio.

Fig. 3 shows the 13C CP-MAS NMR spectra obtained for the silica/fibroin hybrid monoliths. Several structural studies of fibroin and (alanine, glycine)n polypeptides have been carried out using solid state NMR with a structual probe, thus allowing a better understanding of the effects of conformational change.54,55 The variation of the Cβ Ala chemical environment is directly related to the variation in the twist angle (φ, ψ, ω) of the peptide.56 The spectra of all samples show two resonance lines at 17.1 and 20.8 ppm, which are related to Cβ of alanine (Ala) amino acid, in random coil and β-sheet conformations, respectively. The presence of the two conformations can also be evidenced from the change in the chemical shift of the Ala Cα at 49.4 and 50.5 ppm attributed to β-sheet and random coil conformations, respectively.57 The 13C CP-MAS NMR spectra also show a resonance line at 62.5 ppm, assigned to Cβ and Cγ serine in β-sheet conformation. In addition to the significant change in the local field observed for Cβ alanine, the chemical enviroment of the carbonyl (C[double bond, length as m-dash]O) groups also reflected changes in the conformation of a secondary structure of the fibroin. An overlay of four carbonyl resonance lines at 168, 169, 173 and 175 ppm is observed. The 13C CP-MAS NMR data are in agreement with the residual signals observed from Gly and Ala in the β-sheet form (Ala–Gly)n.57,58 Futhermore, by analyzing the relative intensity of the two resonance lines at 17.1 and 20.8 ppm, related to Cβ of alanine (Ala) amino acid, in random coil and β-sheet conformations, it is noteworthy that the relative proportion of the β-sheet increases significantly with increasing fibroin ratio, confirming what was previously observed in the Raman spectroscopy and TGA results. Similar results were reported in a study of hydration of silk fibroin film in which the β-sheet conformation is predominant after 30 min of hydration,59i.e., confirming the transition from random coil to β-sheet conformation.

image file: c7tc03560h-f3.tif
Fig. 3 13C{1H} CP-MAS NMR spectra of silica/fibroin hybrid monoliths containing different fibroin concentrations: (a) SF1; (b) SF2, (c) SF3 and (d) SF4.

Fig. 4 shows the 1H NMR spectra obtained for the hybrid SF4. A broad resonance line is observed at 4.7 ppm, characteristic of the SiOH–(H2O) species. It is known that water molecules exchange among silanol groups at the surface of the SiO2 particles.60 From the TGA data (ESI), it is known that the sample contains approximately 18% H2O in weight and these molecules jump among silanol groups with rates of 1–2 kHz. Despite the high mobility of the homonuclear dipolar interaction, this behavior leads to short transverse dephasing times of the water signal. The intensity of these signals can consequently be reduced by an appropriate echo duration, increasing the spectral resolution of the characteristic signals of fibroin H atoms (albeit at the expense of quantitativeness).

image file: c7tc03560h-f4.tif
Fig. 4 1H spin echo MAS spectrum of SF4 silica/fibroin hybrid monoliths. A saturation train of: (a) 28 pulses (2 ms) and (b) 7 pulses (0.5 ms).

Resonance lines of the Hβ Ala in the two conformations, random coil and β-sheet, may be observed at 1.4 and 1.2 ppm, respectively. This observation corroborates the results observed in the 13C CP-MAS NMR data. The resonance lines observed at 7.0 and 6.8 ppm can be attributed to the phenyl groups of tyrosine amino acids. The two resonance lines observed in the range of 7.5 to 8.5 ppm can be attributed to the NH protons of Ala and Gly amino acids, which form the peptide bonds of the protein.59,61

The NMR study allows not only the structure to be investigated, but the organic/inorganic interface to be understood as well. 13C{1H} CP-heteronuclear correlation (HETCOR) experiments were used to obtain information in spatial proximities between C and H atoms from heteronuclear dipolar interactions, and consequently understand the organic/inorganic interface of silica/fibroin composites. Fig. 5 shows the correlation map 1H–13C in the 2D spectra obtained with two different contact times, 0.5 and 2.5 ms, respectively. The advantage of this experiment is that the chemical shift information of 13C and 1H and the correlation of carbon and hydrogen, which involves only the dipolar coupling, can be simultaneously obtained. In this experiment, the correlation signal is only observed for close C and H atoms (<0.35 nm), as the polarization is transferred from each type of carbon and hydrogen in the vicinity. The resolution in the 1H (vertical) dimension is increased through the application of frequency-switched Lee Goldburg homonuclear decoupling, which reduces the strong residual dipolar interactions among protons that are not completely averaged by MAS, and thereby the associated broadening.

image file: c7tc03560h-f5.tif
Fig. 5 FSLG 13C{1H} CP HETCOR spectra of SF4 sample obtained with two contact times: (top) 0.5 and (bottom) 2.5 ms.

Fig. 5 (top) shows the results for the experiment carried out with a contact time of 0.5 ms, for which only H atoms with distances below 1.5 Å to the C atoms are observed.62 The αCH2 of Gly and the CH of Ala at 49.5 and 43.5 ppm, as well as the Hβ of Ala at 1.3 and 1.4 ppm on the β-sheet and random coil conformations are observed in the spectra. The chemical shift values of hydrogen of the CH2 groups of Gly and CH of Ala are clearly resolved, at 4.3 and 5.2 ppm, respectively. The correlation of hydrogen bonded to the Cβ of Ala with Cα (49.5 ppm) and with carbonyl groups (173 ppm) is also observed but with less intensity. An interesting observation is the presence of correlations of carbonyl groups that are not protonated. According to the literature, this correlation is observed due to the presence of serine and tyrosine peptides present in low concentration.63 Longer-range H → C[double bond, length as m-dash]O transfer among the Ala and Gly aminoacids cannot be ruled-out. Correlation with the Hβ of Ala was also observed. The correlations observed for carbonyl groups show a chemical shift in the range of 4–5 ppm. Other correlations related to the hydrogen have a chemical shift in the range of 8–9.5 ppm, from the dipolar coupling of carbonyl with amino (NH) groups.

It is possible in principle to observe the dipolar coupling correlation between atoms in longer ranges when the experiment is carried out with a contact time of 2.5 ms (Fig. 5 bottom). Correlations between all H and C atoms could then be observed.

The correlation of intensities of the Cβ and Cγ of serine in the β-sheet conformation with hydrogen at 5.2 ppm was observed to decrease while for other H and C atoms, a higher intensity was observed, indicating increasing intramolecular interaction. Moreover, a correlation was observed between carbonyl and SiOH–(H2O)OH. As described above, the chemical shift of the SiOH–(H2O) species is also observed in the range of 4.5 to 5.5 ppm, as shown in the 1H-MAS NMR spectrum (Fig. 4b). A strong dipolar interaction occurs between the Ala Cα (49.4 and 50.5 ppm) with a contact time of 0.5 ms, indicating that the SiOH–(H2O) species formed at the SiO2 interface have a strong dipolar interaction with the Ala Cα, that is, with distances up to 1.5 Å and with the carbonyl groups up to 3.5 Å.

These results show that the network formed over the sol–gel transition and aging steps (Si–O–Si condensation and particle coalescence) builds up a silica shell around the precipitated fibroin fiber.

Recently, Luo et al.64 reported a robust hydrogel with significant elasticity and mechanical performance formed by the strong interaction between silk fibroin and hydroxypropyl methyl cellulose (HPMC). The secondary structural transition of silk fibroin molecular chains from random coil to β-sheet was induced by hydrophobic interactions between HPMC and silk molecules forming small and uniform structures, which served as crosslinking sites evenly distributing throughout the hydrogel.

In the present work, the 3D network silica gel hybrids are reinforced by the interpenetrated fibroin nanofibers. An illustration of the self-assembly of the fibroin nanofibers inside the silica bulk is shown in Scheme 1.

Recently, the self-assembly of silk fibroin nanofibers by decreasing the hydrophilicity of the environment has been reported.65–67 In the present work, we demonstrated that the silk fibroin molecules self-assemble into nanofibrils by progressively decreasing the hydrophilicity of the environment via formation of ethanol in the sol–gel synthesis. The silica/fibroin hybrid monoliths are obtained by hydrolysis and polycondensation of TEOS producing ethanol as a secondary product. Ethanol gradually induces the self-assembly of the fibroin nanofibers with prevailing β-sheet conformation. The growth of fibroin nanofibers starts in the initial stages of gelation, when the average cluster size is very small, and continues during aging, when the 3D silica gel network is reinforced through further polymerization. We suggest here that these new hybrids could be applied as biocompatible mechanically resistant materials in biomedical applications. The easy and simple fabrication of the designed materials together with their biocompatibility, positive biological effects and synergism of their compounds, makes this contribution promising to biomedical applications, such as tissue engineering. Silica is an osteoinductive element in tissue regeneration,29 while silk fibroin is a biodegradable material, which in its β-sheet conformation can improve stability against water and mechanical properties of a designed material extending the range of its clinical applications. Besides, silk fibroin has the ability to support the differentiation of mesenchymal stem cells along the osteogenic lineage and it has been used as a favorable scaffold material for bone tissue engineering.68,69 Applications of osteoinductive biomaterials for bone regeneration combining silk fibroin with silica particles have been reported by the Kapplan group.33,70 Note, however, that in this work, we present a free-standing sample, which can be a viable alternative as a solid material for tissue engineering.

In addition, we point out that the fibroin nanofibers inside the composite materials can also be used as organic spacers and can be selectively removed through careful optimized heat treatment (Table S2, ESI). The heat treatment leads to meso/macroporous silica with pores displaying the memory of the fibroin fiber morphology, as will be shown hereafter.

3.2. Characterization of porous silica monoliths

After heat treatment, all samples were obtained as robust monoliths, with gradually increasing optical density changes to an opaque white color with increasing fibroin initial concentration (Fig. 6a).
image file: c7tc03560h-f6.tif
Fig. 6 (a) Photograph of obtained silica monoliths after heat treatment; (b) SEM image of the porous surface of monolith S3; (c–f) Images of the worm-like pore morphology.

Fig. 6 also shows representative SEM micrographs of porous silica monolith S3. Fig. 6b shows relatively large worm-like pores dispersed at the surface. A closer look at the walls of the pores displays a morphology that can be seen as keeping the memory of the leaving fibroin fibers (Fig. 6c–f). The interstitial channel-like spaces between the particles with typical dimensions of hundreds of nanometers could be related to the larger fibers observed in the initial hybrids (Fig. 1).

The porous structure of the monoliths was analysed by nitrogen adsorption experiments. The nitrogen adsorption/desorption isotherms are reported in Fig. 7(a). As the fibroin concentration increases, the hysteresis loop shifts toward higher pressure, characteristic of an increase of material porosity. All the samples presented a type IV shape, characteristic of mesoporosity, with a desorption loop due to nitrogen condensation within the mesopores.71 Furthermore, textural information (e.g., pore size distribution, pore geometry, and connectivity) can be obtained by analyzing the shape of the hysteresis loop.71–74 In this way, the S0 sample presented a type H4 hysteresis, indicating narrow slit pores. Concerning the other samples, type H2 hysteresis (three-dimensional connected network)71 is observed. This H2 hysteresis is often associated with the disordered pores, indicating that the distribution of pore size and shape is not well defined.71,74,75 These results demonstrated that after the calcination of fibroin contained in the hybrids, hierarchical porous silica monoliths are created, which present mesopores besides the macropores observed in the SEM images above.

image file: c7tc03560h-f7.tif
Fig. 7 (a) Nitrogen adsorption/desorption isotherms, (b) pore size distribution, (c) plot of log(ln(ρ)) versus log(θ), (d) and evolution of the specific surface area, pore diameter and surface fractal dimension of the samples.

The physico-chemical properties of the samples are given in Table 1. The results demonstrate that the calcination of fibroin in the hybrid monoliths leads to an increase of surface area and pore volume with a clear dependence on the fibroin content. The increase of fibroin concentration in the samples SF2 and SF3 implies the formation of more fibroin nanofibers dispersed into the sample. Thus, after the calcination, the silica monoliths S2 and S3 presented a great amount of smaller pores. However, aggregation is observed for higher fibroin content (SF4), leading to larger pores in the silica monolith S4.

Table 1 Physico-chemical properties of calcinated samplesa
Samples S BET (m2 g−1) V p (cm3 g−1) D p (Å) D s
a D p, SBET, Vp and Ds stand for BET pore diameter, surface specific area, pore volume and surface fractal dimension, respectively.
S0 704 0.379 17 ± 12 2.63
S1 819 0.496 32 ± 13 2.72
S2 970 0.601 31 ± 13 2.68
S3 1047 0.621 33 ± 13 2.68
S4 1057 0.525 40 ± 14 2.76

The fractal nature of the silica surface is seen as a direct consequence of the fractal structure of the fibroin–silica interface. In this work, we used the Frenkel–Halsey–Hill (FHH) theory for adsorbent–adsorbate interactions in multilayer coverage.74 The fractal nature of the surface was studied by several authors, who adopted the FHH theory to evaluate their materials.71,73,76 The thickness of the adsorbed film as given by the classic FHH equation provides the characteristic length scale for the fractal-FHH analysis:

θ = [−ln(ρ)/k]−1/s(1)
where ρ = p/p0, θ = n/nmonolayer and s and k are constants. Smith et al.74 calculated the surface fractal dimension of the samples from the slope m of the log–log plot of ln(ρ) versus θ. For the nitrogen adsorption, the values assumed were s = 2.24 and k = 2.27, and the surface fractal dimension (Ds) was obtained from eqn (2).
Ds = (1/m) + 3(2)
Ds is a dimensionless number between 2 and 3: Ds = 2 describes a perfect smooth surface, 2 < Ds < 3 describes an irregular surface and Ds = 3 describes a theoretical volume in which all points would be assigned to the surface.74 Furthermore, important information about surface roughness can be obtained by the Ds interpretation.75

Fig. 7(c) is a log–log plot of θ versus ln(ρ) for the calcined samples. The addition of fibroin in the initial sol promotes changes in observed slopes. The S0 sample presents a slope of −2.73 ± 0.03 whereas values observed for S1, S2, S3 and S4 samples were −3.5 ± 0.1, −3.12 ± 0.05, −3.12 ± 0.05 and −4.10 ± 0.07, respectively. It was observed that the addition of 15.8 wt% fibroin caused an increase in the Ds value from 2.63 up to 2.76. However, this non-integer value implies that the surface exhibits self-similarity, with a slight increase in the surface roughness promoted by the fibroin added in the system before calcination. The evolutions of Ds values, pore diameter and specific surface area with fibroin concentration for the silica/fibroin hybrid monoliths are shown in Fig. 7(d).

3.3. Characterization of the random laser action in Rh6G containing silica monoliths

The designed hierarchical porous silica monoliths are an attractive alternative as a solid material for applications in the areas of catalysis, adsorption and separation. In addition, during recent years, porous silica materials have also been used as disordered scattering media to optimize light diffusion and amplification in random laser (RL) applications. Meng et al.77 investigated random lasing properties in scattering systems composed of a macroporous silica disk immersed in a dye solution with a mixture of two alcohols as the solvent. The proportion between the two alcohols allowed the control of scattering strength and consequent tuning of RL properties. Recently, the coexistence and competition of random lasing and stimulated Raman scattering were predicted in active disordered random media formed by dye-infiltrated macroporous silica38 and silica-/titania films.39 The authors have shown that the existence of both nonlinear effects depends on the scattering strength, in which random lasing has been shown to dominate in samples with a large transport mean free path, while strongly scattering samples favored SRS at reasonable pump fluence.

Herein, we demonstrate RL and SRS action from the porous silica monoliths. Interestingly, these solid and robust materials present the possibility for the control of scattering strength, in which porosity as scattering media can be tuned by adjusting the density and size of the pores by changing the fibroin concentration before the calcination.

To characterize the RL emission in our samples, porous silica monoliths doped with Rh6G were exposed to the excitation beam at 532 nm, and the emission spectra, centered at the RL emission of 561 nm, were recorded for different excitation pulse energies (EPEs). The samples consisted of rods (see inset in Fig. 8) and were excited at the longest face at a 30° angle, and the detected emission was obtained perpendicularly to the sample face. The excitation beam area was 1.6 × 10−2 cm2.

Fig. 8 shows the peak intensity and bandwidth, defined by the full width at half maximum (FWHM), as a function of the excitation pulse energy (EPE) for samples S1–S4. The RL action is characterized by the bandwidth narrowing and an increase in the slope efficiency as the EPE is increased. According to Fig. 8b, all samples displayed bandwidth narrowing. However, according to Fig. 8a and its inset, which is the zoomed in area for small values of EPE, the increase in the slope efficiency was displayed only for samples S2–S4. This indicates that sample S1 supported only amplification of the spontaneous emission, which is due to the weak scattering strength.

image file: c7tc03560h-f8.tif
Fig. 8 (a) Peak intensity vs. excitation pulse energy (EPE) showing the nonlinear behavior of porous silica monoliths containing Rh6G. (b) Full width half maximum (FWHM) bandwidth vs. EPE. The minimum bandwidth was ≈4 nm for EPE above the threshold.

The EPE threshold of the RLs can be determined either from the peak intensity × EPE (input–output) curves or from the EPE corresponding to the medium value of the FWHM, that is, the EPE corresponding to the medium value of the bandwidth equidistant from its initial and final saturated values.24,78 To determine the EE threshold from the input–output curves, one should fit a linear curve to the data of the large slope efficiency and extrapolate this curve to the horizontal axis. These curves and the EPE threshold values are represented in the inset of Fig. 8a. To help the visualization of the EPE thresholds from the bandwidth narrowing, the inset of Fig. 8b shows the data of this figure with the EPE axis in a logarithm scale, and with sigmoidal fits to guide the eyes. The threshold values are indicated by vertical arrows. Table 2 shows the RL EPE thresholds determined by using both procedures, as well as the highest slope efficiencies. The agreement between both EPE threshold values is satisfactory. Using the whole range of EPE in Fig. 8a, one can overestimate the threshold values because in exploiting the large range of EPE, it was necessary to use a large number of optical filters to avoid saturation of the detectors due to the large growth of the output intensity, and the corrections could not be precisely determined.

Table 2 Random laser threshold and slope efficiency
Sample Random laser threshold (mJ) by bandwidth narrowing Random laser threshold (mJ) by input–output curve Slope efficiency (arb. units)
S1 11
S2 0.053 0.050 23
S3 0.016 0.022 160
S4 0.011 0.011 280

According to Fig. 8a and Table 2, the increase of the sample porosity implies higher RL performance, with sample S4 presenting a larger slope efficiency and lower EPE threshold.

The emission spectra obtained for S4 sample are shown in Fig. 9. Typical RL behavior is observed with smooth and nonspiky emission characteristic of dye based RLs and sometimes referred in the literature as due to nonresonant feedback. On increasing the EPE, typical RL emission behavior was observed, as described before, for an EPE up to ∼0.55 mJ, which is well above the determined threshold of 0.011 mJ (see Table 2). Upon increasing the EPE beyond 0.55 mJ, where the spectral bandwidth has already collapsed to its minimum value of ∼4 nm, the spectra redshifts due to reabsorption/reemission processes, and some narrow spikes appeared superimposed on the RL spectra. These spikes are attributed to the interplay between the RL emission, the dye broadband emission and Stokes line of the dye, similar to the reports in ref. 36 and 37. The overall broadening for the EPE beyond 0.55 mJ arises due to these combined effects. However, it can be seen that narrow spikes, due to the stimulated Stokes emission, start to appear with higher emission intensity beyond those EPE values, which follows the Raman spectra of Rh6G shown in ref. 38. This is clearly shown in the inset of Fig. 9, which displays the spectrum at 1.2 × 10−2 J with its Raman lines at 572 nm, 573.8 nm and 578.6 nm corresponding to 1316 cm−1, 1370 cm−1, 1514 cm−1 Raman shifts that can be attributed to aromatic C–C stretching vibrations of Rh6G molecules.79

image file: c7tc03560h-f9.tif
Fig. 9 Spectral behavior of S4 containing Rh6G as a function of the excitation pulse energy (EPE). Minimum bandwidth observed is about 0.5 nm. The insets show the yellow emission colour of the sample S4 doped with Rh6G excited with a laser at 532 nm and the spectrum obtained for an EPE of 12 mJ.

4. Conclusions

Multifunctional organic inorganic hybrids were prepared based on silica and silk fibroin extracted from the cocoons of the silkworm (Bombyx mori). Robust and biocompatible monoliths were obtained with different contents of fibroin. The SEM images demonstrated in situ self assembly of fibroin nanofibers dispersed into the IOH monoliths. We suggest that the formation of fibroin nanofibers is induced by the ethanol formed during the hydrolysis and polycondensation of tetraethyl orthosilicate (TEOS). Raman and solid state NMR spectroscopies show that the prevailing protein conformation of the fibroin nanofibers is the β-sheet type. Futhermore, we demonstrated that the fibroin nanofibers can be used as biotemplates, acting as a sacrificial material, which allowed us to prepare porous silica monoliths in one step, without addition of surfactant or organic particles and supercritical drying processes. When infiltrated with rhodamine dye, the porous monoliths demonstrated efficient random laser (RL) action with low excitation pulse energy (EPE) thresholds and bandwidth narrowing. Further reduction in the RL threshold can be obtained by optimizing the dye concentration and the excitation beam dimensions at the sample. From the spectral behavior, it is inferred that the RL operates in the diffusive regime in the hierarchical macro-mesoporous network. This work opens the possibility of new applications of the RL phenomenon in specklefree laser imaging. In addition, when higher values of EPE are incident on the samples, the emitted spectra present stimulated emission and stimulated Raman scattering simultaneously, suggesting that the prepared materials can also be exploited as a random Raman laser.

Author contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflicts of interest

The authors declare no competing financial interest.


We acknowledge financial support from the Brazilian Agencies: Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Fundação de Amparo à Ciência e Tecnologia do Estado de Pernambuco (FACEPE) and Fundação de Amparo à Ciência e Tecnologia do Estado de São Paulo (FAPESP). This work was performed in the framework of the National Institute of Photonics (INCT de Fotônica), PRONEX-CNPq/FACEPE, and CASADINHO/PROCAD-CNPq/CAPES projects. Moliria Vieira dos Santos acknowledges FAPESP for a doctoral fellowship, grants #2014/12424-2, and FAPESP for a postdoctoral fellowship, grants #2016/11591-8. André de Lima Moura acknowledges CNPq for a postdoctoral fellowship. We acknowledge Danilo Manzani for the photography taken and reproduced in the the graphical abstract and in Fig. 9. Financial support from the TGIR-RMN-THC Fr3050 CNRS for conducting the high-field NMR measurements (with technical assistance from Sylvian Cadars) is gratefully acknowledged.


  1. M. Darder, P. Aranda and E. Ruiz-Hitzky, Adv. Mater., 2007, 19, 1309–1319 CrossRef CAS.
  2. D. Van Opdenbosch and C. Zollfrank, Adv. Eng. Mater., 2014, 16, 699–712 CrossRef CAS.
  3. M. Giese, L. K. Blusch, M. K. Khan and M. J. MacLachlan, Angew. Chem., Int. Ed., 2015, 54, 2888–2910 CrossRef CAS PubMed.
  4. S. Inoue, K. Tanaka, F. Arisaka, S. Kimura, K. Ohtomo and S. Mizuno, J. Biol. Chem., 2000, 275, 40517–40528 CrossRef CAS PubMed.
  5. K. Tanaka, N. Kajiyama, K. Ishikura and S. Waga, Biochim. Biophys. Acta, 1999, 1432, 92–103 CrossRef CAS.
  6. K. S. Hossain, E. Ohyama, A. Ochi, J. Magoshi and N. Nemoto, J. Phys. Chem. B, 2003, 107, 8066–8073 CrossRef CAS.
  7. K. Numata, P. Cebe and D. L. Kaplan, Biomaterials, 2010, 31, 2926–2933 CrossRef CAS PubMed.
  8. X. Liu and K. Zhang, Oligomerization of Chemical and Biological Compounds, 2014, pp. 69–102 Search PubMed.
  9. L. D. Koh, Y. Cheng, C. P. Teng, Y. W. Khin, X. J. Loh, S. Y. Tee, M. Low, E. Ye, H. D. Yu, Y. W. Zhang and M. Y. Han, Prog. Polym. Sci., 2015, 46, 86–110 CrossRef CAS.
  10. H. Tao, D. L. Kaplan and F. G. Omenetto, Adv. Mater., 2012, 24, 2824–2837 CrossRef CAS PubMed.
  11. R. F. P. Pereira, M. M. Silva and V. De Zea Bermudez, Macromol. Mater. Eng., 2015, 300, 1171–1198 CrossRef CAS.
  12. H. Y. Wang, Y. Y. Chen and Y. Q. Zhang, Mater. Sci. Eng., C, 2015, 48, 444–452 CrossRef CAS PubMed.
  13. L. R. Lima, M. L. Moraes, K. Nigoghossian, M. F. S. Peres and S. J. L. Ribeiro, J. Lumin., 2016, 170, 375–379 CrossRef CAS.
  14. B. Coasne, New J. Chem., 2016, 40, 4078–4094 RSC.
  15. G. Guiochon, J. Chromatogr. A, 2007, 1168, 101–168 CrossRef CAS PubMed.
  16. K. K. Unger, R. Skudas and M. M. Schulte, J. Chromatogr. A, 2008, 1184, 393–415 CrossRef CAS PubMed.
  17. A. Galarneau, A. Sachse, B. Said, C. Pelisson, P. Boscaro, N. Brun, L. Courtheoux, N. Olivi-tran, B. Coasne and F. Fajula, C. R. Chim., 2016, 19, 231–247 CrossRef CAS.
  18. D. S. Wiersma, Nat. Phys., 2008, 4, 359–367 CrossRef CAS.
  19. F. Luan, B. Gu, A. S. L. Gomes, K. T. Yong, S. Wen and P. N. Prasad, Nano Today, 2015, 10, 168–192 CrossRef CAS.
  20. M. A. Noginov, Solid-State Random Lasers, Springer, New York, 2005, pp. 198–221 Search PubMed.
  21. X. Shi, Y. Wang, Z. Wang, Y. Sun, D. Liu, Y. Zhang, Q. Li and J. Shi, Appl. Phys. Lett., 2013, 103, 23504 CrossRef.
  22. X. Shi, Y. Wang, Z. Wang, S. Wei, Y. Sun, D. Liu, J. Zhou, Y. Zhang and J. Shi, Adv. Opt. Mater., 2014, 2, 88–93 CrossRef.
  23. A. M. Brito-Silva, A. Galembeck, A. S. L. Gomes, A. J. Jesus-Silva and C. B. de Araújo, J. Appl. Phys., 2010, 108, 33508 CrossRef.
  24. P. I. R. Pincheira, A. F. Silva, S. I. Fewo, S. J. M. Carreño, A. L. Moura, E. P. Raposo, A. S. L. Gomes and C. B. de Araújo, Opt. Lett., 2016, 41, 3459–3462 CrossRef PubMed.
  25. E. Jimenez-Villar, V. Mestre, P. C. de Oliveira and G. F. de Sá, Nanoscale, 2013, 5, 12512–12517 RSC.
  26. B. Redding, M. A. Choma and H. Cao, Nat. Photonics, 2012, 6, 355–359 CrossRef CAS PubMed.
  27. S. García-Revilla, J. Fernández, M. A. Illarramendi, B. García-Ramiro, R. Balda, H. Cui, M. Zayat and D. Levy, Opt. Express, 2008, 16, 12251–12263 CrossRef.
  28. S. García-Revilla, M. Zayat, R. Balda, M. Al-Saleh, D. Levy and J. Fernández, Opt. Express, 2009, 17, 13202–13215 CrossRef.
  29. D. Arcos and M. Vallet-Regí, Acta Biomater., 2010, 6, 2874–2888 CrossRef CAS PubMed.
  30. Y. Kishimoto, F. Ito, H. Usami, E. Togawa, M. Tsukada, H. Morikawa and S. Yamanaka, Int. J. Biol. Macromol., 2013, 57, 124–128 CrossRef CAS PubMed.
  31. A. J. Mieszawska, N. Fourligas, I. Georgakoudi, N. Ouhib, J. David, C. C. Perry and D. L. Kaplan, Biomaterials, 2011, 31, 8902–8910 CrossRef PubMed.
  32. A. Hou and H. Chen, Mater. Sci. Eng., B, 2010, 167, 124–128 CrossRef CAS.
  33. A. J. Mieszawska, L. D. Nadkarni, C. C. Perry and D. L. Kaplan, Chem. Mater., 2010, 22, 5780–5785 CrossRef CAS PubMed.
  34. B. H. Hokr, J. N. Bixler, M. T. Cone, J. D. Mason, H. T. Beier, G. D. Noojin, G. I. Petrov, L. A. Golovan, R. J. Thomas, B. A. Rockwell and V. V. Yakovlev, Nat. Commun., 2014, 5, 4356 CAS.
  35. X. Shi, Q. Chang, J. Tong, Y. Feng, Z. Wang and D. Liu, Sci. Rep., 2017, 7, 1–8 CrossRef PubMed.
  36. V. P. Yashchuk, Laser Phys., 2015, 25, 75702 CrossRef.
  37. V. P. Yashchuk, A. O. Komyshan, E. A. Tikhonov and L. A. Olkhovyk, Quantum Electron., 2014, 44, 921–927 CrossRef.
  38. N. Bachelard, P. Gaikwad, R. Backov, P. Sebbah and R. A. L. Vallée, ACS Photonics, 2014, 1, 1206–1211 CrossRef CAS.
  39. P. Gaikwad, N. Bachelard, P. Sebbah, R. Backov and R. A. L. Vallée, Adv. Opt. Mater., 2015, 3, 1640–1651 CrossRef CAS.
  40. D. D. N. Rockwood, R. R. C. Preda, T. Yücel, X. Wang, M. L. Lovett and D. L. Kaplan, Nat. Protoc., 2011, 6, 1–43 CrossRef PubMed.
  41. A. Bielecki, A. C. Kolbert, H. J. M. De Groot, R. G. Griffin and M. H. Levitt, Adv. Magn. Reson., 1990, 14, 111–124 Search PubMed.
  42. X. Hu, D. Kaplan and P. Cebe, Macromolecules, 2006, 39, 6161–6170 CrossRef CAS.
  43. J. H. Yeo, K. G. Lee, Y. W. Lee and S. Y. Kim, Eur. Polym. J., 2003, 39, 1195–1199 CrossRef CAS.
  44. F. Vollrath and D. Porter, Polymer, 2009, 50, 5623–5632 CrossRef CAS.
  45. J. G. Hardy, L. M. Römer and T. R. Scheibel, Polymer, 2008, 49, 4309–4327 CrossRef CAS.
  46. D. Kaplan and K. McGrath, in Protein-based materials, ed. K. McGrath and D. Kaplan, Birkhäuser Basel, Boston, 1st edn, 1996, pp. 103–133 Search PubMed.
  47. C. Pereira, S. Patrício, A. R. Silva, A. L. Magalhães, A. P. Carvalho, J. Pires and C. Freire, J. Colloid Interface Sci., 2007, 316, 570–579 CrossRef CAS PubMed.
  48. X. Hu, K. Shmelev, L. Sun, E. S. Gil, S. H. Park, P. Cebe and D. L. Kaplan, Biomacromolecules, 2011, 12, 1686–1696 CrossRef CAS PubMed.
  49. H.-J. Jin and D. L. Kaplan, Nature, 2003, 424, 1057–1061 CrossRef CAS PubMed.
  50. A. Motta, L. Fambri and C. Migliaresi, Macromol. Chem. Phys., 2002, 203, 1658–1665 CrossRef CAS.
  51. P. Monti, P. Taddei, G. Freddi, T. Asakura and M. Tsukada, J. Raman Spectrosc., 2001, 32, 103–107 CrossRef CAS.
  52. A. Matsumoto, J. Chen, A. L. Collette, U. J. Kim, G. H. Altman, P. Cebe and D. L. Kaplan, J. Phys. Chem. B, 2006, 110, 21630–21638 CrossRef CAS PubMed.
  53. E. Lippmaa, M. Mägi, A. Samoson, G. Engelhardt and A. R. Grimmer, J. Am. Chem. Soc., 1980, 102, 4889–4893 CrossRef CAS.
  54. T. Asakura, J. M. Yao, T. Yamane, K. Umemura and A. S. Ultrich, J. Am. Chem. Soc., 2002, 124, 8794–8795 CrossRef CAS PubMed.
  55. T. Asakura, Y. Nakazawa, E. Ohnishi and F. Moro, Protein Sci., 2005, 14, 2654–2657 CrossRef CAS PubMed.
  56. P. Zhou, G. Li, Z. Shao, X. Pan and T. Yu, J. Phys. Chem. B, 2015, 105, 12469–12476 CrossRef.
  57. Q. Dang, S. Lu, S. Yu, P. Sun and Z. Yuan, Biomacromolecules, 2010, 11, 1796–1801 CrossRef CAS PubMed.
  58. H. Saitô, R. Tabeta, T. Asakura, Y. Iwanaga, A. Shoji, T. Ozaki and I. Ando, Macromolecules, 1984, 17, 1405–1412 CrossRef.
  59. H. S. M. Ishida, T. Asakura and M. Yokoi, Macromolecules, 1990, 94, 88–94 CrossRef.
  60. J. Trébosc, J. W. Wiench, S. Huh, V. S.-Y. Lin and M. Pruski, J. Am. Chem. Soc., 2005, 127, 3057–3068 CrossRef PubMed.
  61. T. Asakura, M. Endo, M. Hirayama, H. Arai, A. Aoki and Y. Tasei, Int. J. Mol. Sci., 2016, 17, 1517 CrossRef PubMed.
  62. S. P. Brown and H. W. Spiess, Chem. Rev., 2001, 101, 4125–4155 CrossRef CAS PubMed.
  63. J. L. White and X. Wang, Macromolecules, 2002, 35, 2633–2639 CrossRef CAS.
  64. K. Luo, Y. Yang and Z. Shao, Adv. Funct. Mater., 2016, 26, 872–880 CrossRef CAS.
  65. Z. Yin, F. Wu, Z. Zheng, D. L. Kaplan, S. C. Kundu and S. Lu, ACS Biomater. Sci. Eng., 2017, 3, 2617–2627 CrossRef CAS.
  66. D. Su, M. Yao, J. Liu, Y. Zhong, X. Chen and Z. Shao, ACS Appl. Mater. Interfaces, 2017, 9, 17489–17498 CAS.
  67. P. Tseng, B. Napier, S. Zhao, A. N. Mitropoulos, M. B. Applegate, B. Marelli, D. L. Kaplan and F. G. Omenetto, Nat. Nanotechnol., 2017, 12, 474–480 CrossRef CAS PubMed.
  68. B. Kundu, N. E. Kurland, S. Bano, C. Patra, F. B. Engel, V. K. Yadavalli and S. C. Kundu, Prog. Polym. Sci., 2014, 39, 251–267 CrossRef CAS.
  69. J. Melke, S. Midha, S. Ghosh, K. Ito and S. Hofmann, Acta Biomater., 2016, 31, 1–16 CrossRef CAS PubMed.
  70. A. J. Mieszawska, N. Fourligas, I. Georgakoudi, N. M. Ouhib, D. J. Belton, C. C. Perry and D. L. Kaplan, Biomaterials, 2010, 31, 8902–8910 CrossRef CAS PubMed.
  71. S. Lowell, J. E. Shields, M. A. Thomas and M. Thommes, Characterization of Porous Solids and Powders: Surface Area, Pore Size and Density, Springer, Netherlands, Dordrecht, 1st edn, 2004, vol. 16 Search PubMed.
  72. J. N. M. Batista, E. H. De Faria, P. S. Calefi, K. J. Ciuffi, E. J. Nassar, W. R. Cunha, J. M. A. Caiut and L. A. Rocha, Appl. Surf. Sci., 2012, 258, 5111–5116 CrossRef CAS.
  73. L. A. Rocha, J. do, C. Freiria, J. M. A. Caiut, S. J. L. Ribeiro, Y. Messaddeq, M. Verelst and J. Dexpert-Ghys, Nanotechnology, 2015, 26, 335604 CrossRef PubMed.
  74. M. A. Smith and R. F. Lobo, Microporous Mesoporous Mater., 2010, 131, 204–209 CrossRef CAS.
  75. E. Prouzet, C. Boissière, S. S. Kim and T. J. Pinnavaia, Microporous Mesoporous Mater., 2009, 119, 9–17 CrossRef CAS.
  76. L. Liu, Y. Hu, L. Song, X. Gu, Y. Chen and Z. Ni, Microporous Mesoporous Mater., 2010, 132, 567–571 CrossRef CAS.
  77. X. Meng, K. Fujita, S. Murai, J. Konishi, M. Mano and K. Tanaka, Opt. Express, 2010, 18, 12153–12160 CrossRef CAS PubMed.
  78. J. Kitur, G. Zhu, M. Bahoura and M. A. Noginov, J. Opt. A: Pure Appl. Opt., 2010, 12, 9–14 Search PubMed.
  79. H. Watanabe, N. Hayazawa, Y. Inouye and S. Kawata, J. Phys. Chem. B, 2005, 109, 5012–5020 CrossRef CAS PubMed.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c7tc03560h

This journal is © The Royal Society of Chemistry 2018