Efficient extraction of carboxylated spherical cellulose nanocrystals with narrow distribution through hydrolysis of lyocell fibers by using ammonium persulfate as an oxidant

Abstract

An efficient and low-cost approach to prepare spherical cellulose nanocrystals (SCNCs) is presented through chemical hydrolysis of lyocell fibers in an ammonium persulfate (APS) solution. The as-prepared cellulose nanoparticles were characterized by scanning electron microscopy, atomic force microscopy, laser light scattering particle analysis, wide angle X-ray diffraction, Fourier transform infrared spectrometry and thermal gravimetric analysis. Effects of hydrolysis conditions, such as reaction time and temperature, and APS concentration on the morphology, microstructure, and thermal stability of cellulose nanoparticles are discussed. Moreover, it is found that under mild reaction conditions, cellulose nanoparticles are spherical particles with a narrow diameter distribution, and have a cellulose II polymorphic crystalline structure with surface carboxyl groups. The optimal hydrolysis time was found to be around 16 h for hydrolysis at 80 °C with a 1 M APS aqueous solution.

---

1. Introduction

As a result of their nanoscale dimensions and intrinsic properties, cellulose nanocrystals (CNCs) have the potential of becoming an important class of renewable nanomaterials.^1–4 Apart from their potential use as reinforcing agents for industrial biocomposites, CNCs could provide impetus for their use in bioapplications, specifically enzyme immobilization, protein immobilization, antimicrobial and medical materials, drug delivery, green catalysis, and metallic reaction templates.^1,3–6 Furthermore, a lot of active functional hydroxyl groups exist on the surface of cellulose, which are easy to be modified into active groups such as carboxyl, amide and so on. By appropriate modification of CNCs, various functional nanomaterials with outstanding properties, or significantly improved physical, chemical, biological, as well as electronic properties can be developed, which would greatly expand their applications in various market sectors.^2,4,7

Several methods including chemical methods have been used to obtain highly purified nanocrystals from cellulosic materials, which were mainly carried out by acid hydrolysis,^8,9 enzyme-assisted hydrolysis¹⁰ and mechanical treatments. CNCs with different morphologies have been obtained by controlling the origin of the raw materials and reaction conditions of acid hydrolysis. CNCs are, in general, rigid rod-like nanocrystals with diameter in the range of 10–20 nm and length of a few hundred nanometers, while some spherical CNCs (SCNCs) are also produced through acid treatment.^8–11 SCNCs have been prepared from microcrystalline cellulose (MCC),^9,10,12 cotton,^8,13,14 Avicel pulp,¹¹ sesame husk,¹⁵ sweet potato residue,¹⁶ and cellulose fibers.^17,18 These raw materials were always swollen in NaOH solutions or NaOH–dimethyl sulfoxide mixtures,¹⁸ and then hydrolyzed using a mixture of sulfuric acid (H₂SO₄) and hydrochloric acid (HCl) under ultrasonic treatment. It has been found that for the hydrolysis of MCCs, spherical or rod-like CNCs could be selectively formed by varying the mixed acid concentrations of H₂SO₄ and HCl. Most interestingly, upon the sequential addition of H₂SO₄ and HCl in 1 to 2.5 molar ratio, only SCNCs were observed with the average size of 50 nm in narrow distribution.¹² The formation of SCNCs was ascribed to the combination of the ultrasonic treatment and mild acid concentration. Recently SCNCs with a bimodal size distribution (43 ± 13 and 119 ± 9 nm) have been prepared in a span of 7 days by controlled hydrolysis of MCC using an anaerobic microbial consortium.¹⁰

Ammonium persulfate (APS), a strong oxidizing agent with low long-term toxicity, high water solubility and low cost, was favored over its sodium and potassium counterparts.¹⁹ It has been demonstrated that more homogeneous rod-like CNCs with 5 nm diameter and 150 nm length have been prepared by using APS instead of conventional acid hydrolysis from MCCs²⁰ and bacterial cellulose,¹⁹ respectively. Especially, carboxylated CNCs could be obtained in these cases. It is well known that the carboxyl groups on the surface of the materials can provide active sites for template synthesis of nanoparticles, surface modification as well as protein/enzyme immobilization.^6,21 It has been noted that lyocell fibers have homogeneous distribution of fibrils and highly regular crystalline regions. Therefore, commercial lyocell fibers in this work were chosen as raw materials to obtain carboxylated SCNCs without ultrasonication. The effects of the hydrolysis conditions such as hydrolysis time, temperature and concentration of APS on the morphology, crystallinity, and thermal stability of the cellulose nanoparticles would be explored.

2. Experimental section

2.1 Materials

Lyocell fibers without spin finishing were kindly supplied by Shanghai Lyocell Fibre Development Co., Ltd. APS and NaOH were purchased from Guoyao Group Chemical Reagent CO., Ltd.

2.2 Preparation

Lyocell fibers were first treated with a 1 M NaOH aqueous solution for 3 h to remove some impurities from the surface of the fibers. More importantly, the alkali treatment on cellulose had the effects of swelling, which would help in rapid oxidation degradation later. The swollen fibers were thoroughly washed with deionized water and then air-dried. Subsequently, 1 g of the treated lyocell fibers were cut into small pieces, and immersed into 100 mL of an aqueous solution of APS. The hydrolysis of lyocell fibers was carried out with continuous stirring under various reaction conditions. Furthermore, the reaction time, temperature and APS concentration were identified as three of the most important parameters in the hydrolysis process of cellulose. The suspension was diluted with cold distilled water to stop the reaction, and then centrifuged for 10 min at 12 000 rpm and 10 °C. The centrifugation step was repeated several times until a constant pH was attained. The cellulose nanoparticles were finally collected by vacuum freeze-drying, and the yield of cellulose nanoparticles was calculated as a percentage of the weight of the final products divided by the initial weight of the lyocell fibers.

2.3 Characterization

The morphologies of lyocell fibers and as-prepared cellulose nanoparticles were observed on a field emission scanning electron microscope (SEM) and a Digital Images Dimension 3100 atomic force microscope (AFM) (Veeco Metrology Inc., Santa Barbara, CA). All the AFM images were obtained using tapping mode with the scan rate of 1.0 Hz by using the cantilever with a spring constant of 40 N m⁻¹.

The particle size distribution and zeta potential of cellulose nanoparticles were measured with a Nano ZS Malvern Zetasizer, providing both multi-angle particle size analysis by dynamic light scattering and low-angle zeta potential analysis by electrophoretic light scattering. 2.5 mg mL⁻¹ aqueous suspensions of cellulose nanoparticles were prepared, and measurements were carried out in triplicate at 25 °C.

Chemical structures of lyocell fibers and as-prepared cellulose nanoparticles were characterized on a Nicolet 8700 Fourier transform infrared (FT-IR) spectrophotometer. The spectra were collected in a spectral range of 4000–400 cm⁻¹ with 64 scans and 4 cm⁻¹ resolution. Crystal structures were studied on a Philips PW2000 X-ray diffractometer (XRD) by using Cu Kα X-rays with a voltage of 40 kV and a current of 30 mA. Profile analysis was carried out with the peak fitting program using Gaussian line shapes to determine the crystallinity of samples.

Thermal stability was determined on a NETZSCH TG 209 F1 thermogravimetric analyzer (TGA). TGA was carried out from 40 to 600 °C at the heating rate of 20 °C min⁻¹ under a nitrogen flow of 20 mL min⁻¹. The degradation parameters, such as initial decomposition temperature (T₀) and maximum decomposition temperature (T_max) were calculated from the TGA curve.

3. Results and discussion

3.1 Morphology and dimension

It is well known that the dimensions and crystallinity of the CNCs were dependent upon the particular hydrolysis conditions and source of cellulose employed. Moreover, reaction time has been identified as one of the most important parameters to be considered during the acid hydrolysis process.²² Herein, SEM images of lyocell fibers and the resulting products prepared at various reaction times with reaction temperature of 80 °C and 1 M APS aqueous solution are shown in Fig. 1. Lyocell fibers were about 10 μm in diameter with a smooth surface. With the increase of reaction time, lyocell fibers would degrade into small pieces. In the first stage of the hydrolysis, the effect of reaction time on the size of degradation products was more apparent. Besides some cellulose nanoparticles with a near-spherical shape could be found; some fibrous particles also appeared in SEM images. However, after 3 h treatment, cellulose nanoparticles would be close to spherical in shape. SCNCs with relatively narrow diameter distribution could be observed after the reaction time of 8 h. With the increase of reaction time from 12 to 16 h, the average diameter (particle number > 20) would reduce from 45 to 40 nm, and then slightly decrease when the reaction time was further prolonged to 24 h as shown in Fig. 1. Herein, the particle size was determined by the number average particle size from a selected SEM image. Furthermore, the time needed for obtaining totally SCNCs reduced from 4 to 2 h when the reaction temperature increased from 70 to 90 °C. As for APS concentration, some rod-like or whisker cellulose nanoparticles could be observed at 0.5 M APS concentration for all the reaction times, while wide size distribution appeared for 2 M APS concentration.

Fig. 1 SEM images of lyocell fibers and cellulose particles prepared through chemical hydrolysis of lyocell fibers with reaction temperature of 80 °C and 1 M APS aqueous solution at various reaction times. (a) Lyocell fiber; (b) 8 h; (c) 16 h and (d) 24 h.

A similar change in the morphology of the cellulose nanoparticles can be observed in AFM topography images. Fig. 2 gives some isolated individual SCNCs obtained with reaction temperature of 80 °C and 1 M APS for 16 h. It was clearly recognized that the cellulose nanoparticles indeed were spherical particles with a regular shape and narrow size distribution, especially in three-dimensional images. The average diameter determined in the case of AFM was about 40 nm, being close to that from SEM images. It is worth pointing out that such a low-cost and easily scalable approach could be expected to provide a commercially viable method for extracting SCNCs with small diameter and narrow size distribution without requiring tedious separation steps such as differential centrifugation.

Fig. 2 AFM topography images of some isolated individual SCNCs obtained with reaction temperature of 80 °C and 1 M APS for 16 h.

Fig. 3 displays the schematic for the formation of the SCNCs. It is well known that lyocell fibers are composed of elementary fibrils with a cellulose II crystal structure partially separated by elongated voids, which run parallel to the fiber axis. Furthermore, all regenerated cellulose fibers have dislocations in the form of longitudinal and transverse intra- and interfibrillated micropores with diameters between 1 and 3 nm.²³ When lyocell fibers were immersed in an APS aqueous solution, SO₄˙⁻ free radicals produced through the thermal cleavage of the peroxide bond of APS would rapidly penetrate the amorphous parts to break down amorphous cellulose by cohydrolyzing the 1,4-β bond of the cellulose chain.^19,20 That is to say, by using lyocell fibers as cellulose raw materials, simultaneous hydrolysis of lyocell fibers would occur at the surface and in the inner amorphous regions due to fast penetration of free radical ions produced from APS through the elongated voids in the fibers. The hydrolysis reaction would occur in the amorphous region, and subsequently only the domain with a perfect crystalline structure remained after hydrolysis. More importantly, the SCNCs with relatively narrow diameter distribution could be extracted as shown in Fig. 3, benefiting from homogeneous distribution of oriented crystallites in lyocell fibers.²³

Fig. 3 Schematic for the formation of SCNCs through chemical hydrolysis of lyocell fibers in an APS solution.

3.2 Size distribution and yield

After the hydrolysis process in an APS aqueous solution, the lyocell fibers turned into a milky colloid suspension and no significant aggregation could be noticed for cellulose nanoparticles as shown in their morphology observation. The zeta potential for most SCNCs prepared through the hydrolysis of lyocell fibers by using APS was around −24 mV, which was an intermediate value between −69.7 mV for CNCs prepared by acid hydrolysis and −14.6 mV for microbial hydrolysis.¹⁰ This result indicates that some negatively charged functional groups were introduced into the surface of the SCNCs. The repulsive forces between the charged nanoparticles were generally sufficient to prevent the aggregation of particles, and thus the suspension would remain stable.

Table 1 lists particle size and distribution for cellulose nanoparticles prepared under different reaction conditions. Herein, the size distribution of the cellulose nanoparticles in 2.5 mg mL⁻¹ aqueous suspensions was calculated through a DLS particle size analyzer, and the reported size values are mean intensity diameters obtained with the General Purpose algorithm of the Zetasizer Nano software. The average of the five diameter measurements is listed in Table 1. It is noted that compared with the diameter determined from the AFM images, smaller diameter would be obtained by dynamic light scattering. This can be attributed to image widening because the convolution of the tip and the particle would result in an overestimation of the dimension of particles.²⁴ As expected, the diameter and yield tended to decrease more or less continuously with the elongation time, and the diameter could be controlled by selecting suitable reaction conditions. Under the APS concentration of 1 M and the reaction temperature of 80 °C, all the products extracted from lyocell fibers would exhibit spherical nanoparticles when the reaction time reached 3 h. The yield and average diameter of the SCNCs would further decrease along with the reaction time, and finally tended to a stable value after the reaction time of 16 h. It is worth noting that the polydispersity, as a percentage of standard deviation of the particle size distribution divided by the average size in diameter, decreased significantly with the increase of hydrolysis time. This indicates that more homogeneous nanoparticles can be obtained by increasing the duration of the hydrolysis process. This was a very important finding, since the uniformity and small diameter of such nanocrystals are critical for their intended applications as nanofillers or for drug delivery. The average diameter of the SCNCs prepared for 16 h was about 36 nm with the polydispersity of 43%, which was noticeably smaller than 60 nm for those obtained by hydrolyzing MCCs with the mixture of sulfuric acid and hydrochloric acid under ultrasonic conditions.⁹

Table 1 Particle size and distribution and yield of SCNCs prepared from lyocell fibers under different reaction conditions

Reaction condition			Diameter (nm)	Yield (wt%)
APS concentration (M)	Temperature (°C)	Time (h)
1	70	4	96 ± 47	75
		8	74 ± 41	51
		12	59 ± 29	41
		16	53 ± 25	39
		20	53 ± 25	37
		24	50 ± 23	35
1	80	4	75 ± 38	57
		8	60 ± 26	45
		12	45 ± 19	38
		16	35 ± 13	35
		20	35 ± 13	33
		24	35 ± 14	32
1	90	4	54 ± 33	41
		8	47 ± 31	37
		12	36 ± 29	34
		16	30 ± 19	30
		20	30 ± 18	29
		24	28 ± 16	28
2	80	4	56 ± 33	52
		8	46 ± 27	40
		12	35 ± 20	32
		16	33 ± 17	25
		20	30 ± 19	23
		24	29 ± 18	22

---

Moreover, higher concentration of free radicals would be produced at the same reaction temperature when the APS concentration varied from 1 to 2 M. This implies that a strong oxidative reaction has taken place, which would result in an obvious decrease in the diameter and yield of the SCNCs, especially in the early stage of the hydrolysis as seen in Table 1. As expected, similar phenomena could be observed by raising the reaction temperature from 80 to 90 °C. These results can be attributed to the fact that more of the amorphous region in lyocell fibers would be removed when they were attacked by more free radicals, resulting in large weight loss of cellulose in the early stage of the hydrolysis. Obviously, longer reaction time corresponded to lower yield and smaller diameter as shown in Table 1. However, excessive oxidizing nanoparticles would also lead to a dark feature. When the reaction was carried out by applying larger APS concentration or higher reaction temperature, smaller diameters, wider diameter distribution and lower yield of SCNCs would be obtained as listed in Table 1. In contrast, when a low APS concentration of 0.5 M was used, lyocell fibers did not fully convert to cellulose nanoparticles with a spherical shape; thus the dimensions of these nanoparticles are not included in Table 1. Although high yield could be obtained by decreasing reaction temperature, there is a very broad size distribution for SCNCs as given in Table 1. The optimum hydrolysis time was found to be 16 h for the production of SCNCs with relatively uniform diameter and good yield value; meanwhile reaction temperature of 80 °C and APS concentration of 1 M were believed to be relatively mild reaction conditions.

3.3 Crystalline structure

Fig. 4 shows X-ray diffraction patterns of lyocell fibers and cellulose nanoparticles produced under different reaction conditions. The diffraction patterns of cellulose nanoparticles exhibited two intense peaks at 2θ = 20.2° for the (110) peak and 21.8° for the (200) peak, and a weak peak at 12.2° for the (11̄0) peak, where θ is the corresponding Bragg angle. Such patterns were similar to those of lyocell fibers, indicating that the crystal structure of cellulose II in lyocell fibers remained.^13,25,26 However, lyocell fibers created an overlapped diffraction peak around 20.8 °C and a large amorphous diffuse peak, which illustrated that crystallinity of lyocell fibers was relatively low. That is to say, cellulose nanoparticles had comparatively high crystallinity due to preferential hydrolysis of amorphous regions of lyocell fibers in APS solutions.

Fig. 4 WAXD patterns of lyocell fibers and cellulose nanoparticles prepared through chemical hydrolysis of lyocell fibers under different reaction conditions.

The crystalline index (CrI) was estimated by calculating the height ratio between the intensity value (I₂₀₀ − I_am) and total intensity (I₂₀₀) after subtraction of the background signal measured without cellulose, i.e. CrI (%) = ((I₂₀₀ − I_am)/I₂₀₀) × 100, where I₂₀₀ is the intensity obtained from the (200) peak, while I_am is the minimum intensity found between the (110) and (200) peaks.^20,27–29 The average crystallite sizes were estimated using the well-known Scherrer equation, D_hkl = Kλ/B_hkl cos θ, where D_hkl is the crystallite size in the direction normal to the hkl family of lattice planes, K the Prague constant (0.89 in general), λ the wavelength of the X-rays (Cu target of 0.154056 nm) and B_hkl the full width at half-maximum in radians of the reflection of that family of lattice planes.^27–29 All above-mentioned results are summarized in Table 2.

Table 2 The average crystallite size, crystallinity and thermal parameters for the cellulose nanoparticles obtained under different reaction conditions

Sample			(11̄0)^a (nm)	(110)^a (nm)	(200)^a (nm)	CrI^a (%)	DO^b (%)	T 0 (°C)	T max (°C)
a Crystallite size and crystallinity index (CrI) were calculated from the WAXD patterns. b Degree of oxidation (DO) was estimated from FT-IR spectra. c T 0 and Tmax were obtained from the TGA curves.
Lyocell			5.43	5.06	5.12	67.0	—	327.2	360.5
Reaction conditions
4 h	80 °C	1.0 M	5.20	4.88	4.99	91.2	0.093	279.9	313.1
8 h			4.96	4.83	4.81	92.1	0.113	285.6	321.4
12 h			4.78	4.71	4.71	92.7	0.102	295.8	324.1
16 h			4.77	4.69	4.69	93.4	0.083	298.5	330.8
20 h			4.77	4.68	4.69	93.9	0.078	297.6	330.8
24 h			4.76	4.65	4.65	93.1	0.081	297.3	330.9
16 h	70 °C	1.0 M	4.93	4.87	4.87	87.5	0.140	261.5	298.0
	90 °C		4.62	4.59	4.57	89.5	0.058	287.3	321.5
16 h	80 °C	0.5 M	5.22	5.01	4.93	91.1	0.065	280.5	318.8
		2.0 M	4.55	4.53	4.51	89.4	0.132	247.3	281.7

---

The results listed in Table 2 clearly indicate that the CrI of cellulose nanoparticles was at least 20% higher than that of the lyocell fibers (67%). This is consistent with the reports about the trend in the change of the total crystallinity caused by acid degradation of raw materials. Furthermore, the average size of crystallites for cellulose nanoparticles became smaller as a whole after the reaction. When the reaction occurred with 1 M APS concentration at 80 °C, the CrI of cellulose nanoparticles gradually increased from 91.6 to 93 when the reaction time increased from 4 to 20 h. However, further increase in hydrolysis time appeared to reverse this trend, leading to a decrease in CrI. These findings could be ascribed to the partial destruction of cellulose crystalline regions in the cellulose nanoparticles. Moreover, at the same reaction time of 16 h, when the reaction time and APS concentration changed in turn, diffraction peaks of cellulose nanoparticles prepared did not exhibit an obvious change. However, the highest crystallinity could be obtained at 80 °C as shown in Table 2, which could be attributed to the insufficient degradation of cellulose amorphous parts at 70 °C; whereas partial damage of the crystal structure occurred at 90 °C. Similarly, due to more amorphous components remaining for 0.5 M APS concentration, whereas local damage in the crystal structure occurring for a high concentration of 2 M, it is more suitable for the preparation of high-crystallinity SCNCs with an APS concentration of 1 M.

3.4 Chemical structure

Fig. 5 shows the FT-IR spectra of lyocell fibers and cellulose nanoparticles prepared under different reaction conditions. It was observed that O–H stretching vibrations around 3445 cm⁻¹, C–H stretching vibrations at 2898 cm⁻¹, the H–C–H and O–C–H in-plane bending vibrations at 1423 cm⁻¹, and the C–H deformation vibrations at 1376 cm⁻¹ represented characteristic peaks of cellulose.¹⁰ The cellulose nanoparticles exhibited similar infrared peaks compared to lyocell fibers, indicating that no significant changes in the conformation of the cellulose structure occurred. However, a carboxyl peak at 1735 cm⁻¹ appeared in the infrared spectra of cellulose nanoparticles as shown in Fig. 5, because cellulose experienced the carboxylation in the process of oxidation degradation of APS.^19,30,31 It has been reported that the degree of oxidation (DO) could be obtained by calculating the ratio of the intensity of the carbonyl peak to that for the strongest band near 1060 cm⁻¹, relating to the backbone structure of cellulose. More clearly, the relationship deduced from previously published results, DO = 0.01 + 0.7(I₁₇₃₅/I₁₀₆₀), could be used in this case.³² The calculated results are listed in Table 2.

Fig. 5 FT-IR spectra of lyocell fibers and cellulose nanoparticles prepared through chemical hydrolysis of lyocell fibers under different reaction conditions.

The formation of carbonyl groups on the surface of cellulose nanoparticles would make these particles more reactive and improve their flexibility and processability in composites. In contrast, CNCs prepared by acid hydrolysis have only hydroxyl groups (when HCl is used) or up to 2% sulfonyl groups with use of H₂SO₄ only. Fig. 5 shows that when the reaction took place with 1 M APS concentration at 80 °C, the formation of carbonyl groups was promoted for the long interaction time between cellulose and free radicals generated from APS during hydrolysis until the reaction time reached 8 h, and then the peak intensity of carboxyl groups gradually became weaker from 8 to 20 h. However, with the further increase of reaction time, the carboxylic content of particles would increase slightly as shown in Table 2. This result could be attributed to the partial damage of the cellulose crystalline region during 24 h treatment, resulting in the increase of carboxyl content. Moreover, the highest carboxyl content could be obtained with 2 M APS concentration, whereas the lowest at the reaction temperature of 90 °C, although these cellulose nanoparticles had similar crystallinity as seen in Table 2. As shown in Fig. 5, peak intensity of carboxyl groups reduced with the increase of temperature, which indicated that oxidation degradation at the low temperature was conducive to carboxylation of cellulose nanoparticles. In addition, under the reaction at 80 °C for 16 h, with the increase of APS concentration, the peak intensity for carboxyl groups also increased as shown in Fig. 5, implying that amorphous parts of cellulose were more accessible to the carboxyl groups.

3.5 Thermal stability

Fig. 6 shows TG curves for lyocell fibers and cellulose nanoparticles prepared under different reaction conditions, and the thermal parameters are listed in Table 2. The thermal degradation process of lyocell fibers exhibited mainly a one-step degradation process with T₀ of 327.5 °C and T_max of 360.5 °C, and could be divided into four stages: the loss of water absorbed within cellulose before 120 °C, appearance of a plateau until 260 °C, degradation of cellulose at temperatures ranging from 250 to 400 °C, and carbonation of the residual products over 400 °C.²⁰ As shown in Fig. 6(a) and (b), an obvious decrease in the thermal stability could be found for all the cellulose nanoparticles, owing to their small particle size, high specific surface area and active surface groups. Fortunately, most cellulose nanoparticles were stable when the temperature was less than 280 °C. It is important for cellulose nanoparticles in thermoplastic applications since the processing temperature is often above 200 °C. In comparison, cellulose nanoparticles obtained from sulfuric acid hydrolysis have lower thermal stability ranging from 150 to 200 °C due to the presence of sulfate groups.

Fig. 6 TGA curves for lyocell fibers and cellulose nanoparticles prepared through chemical hydrolysis of lyocell fibers under different reaction conditions. (a) Reaction time; (b) reaction temperature and APS concentration.

The difference in thermal degradation behaviors resulted from the differences in outer surface structure, dimension and crystallinity of cellulose nanoparticles. When the reaction was performed with 1 M APS concentration at 80 °C, the degradation behaviors were close to each other for SCNCs prepared at various reaction times as shown in Fig. 6(a), and a slight increase in stability could be achieved for longer reaction time as shown in Table 2. In contrast, more obvious influence of reaction temperature and APS concentration on the thermal stability can be found as seen in Table 2. A two-step degradation process could be observed at relatively low temperature (70 °C) and with large APS concentration (2 M). As for the former, this result could be attributed to the wide distribution in diameter of particles produced at 70 °C, while for the latter the decomposition of cellulose with a large amount of carboxyl groups at low temperature usually facilitated the formation of a char residue, which would act as flame retardants to some extent.⁹

4. Conclusions

A low-cost and easily scalable approach for extracting SCNCs was presented through hydrolysis of lyocell fibers by using APS as an oxidant. Effects of reaction conditions such as reaction time, temperature and APS concentration on the dimensions, morphology, crystal structure, surface functional groups and thermal stability were discussed. By using lyocell fibers as cellulose raw materials, simultaneous hydrolysis of lyocell fibers would occur at the surface and in the inner amorphous regions due to fast penetration of free radical ions produced from APS through the elongated voids in the fibers. More importantly, relatively homogeneous distribution of crystalline regions in lyocell fibers would be of great benefit to the extraction of SCNCs with narrow diameter distribution. Furthermore, it is found that when the reaction occurred with 1 M APS concentration at 80 °C, SCNCs with a regular shape and narrow diameter distribution could be prepared when the reaction time was optimized to be around 16 h. Under optimized hydrolysis conditions, uniform SCNCs with an average diameter of 35 nm and surface carboxyl groups were successively obtained, meanwhile cellulose II crystal structure and relatively high thermal stability remained, which would make these nanoparticles more reactive and improve their flexibility and processability.

Acknowledgements

We greatly acknowledge financial support from the Fundamental Research Funds for the Central Universities (12D10610), and Program for Changjiang Scholars and Innovative Research Team in University (IRT1221).

References

1. L. Brinchi, F. Cotana, E. Fortunati and J. Kenny, Carbohydr. Polym., 2013, 94, 154–169.
2. B. Holt, S. Stoyanov, E. Pelan and V. Paunov, J. Mater. Chem., 2010, 20, 10058–10070.
3. K. Male, A. Leung, J. Montes, A. Kamen and J. Luong, Nanoscale, 2012, 4, 1373–1379.
4. B. Peng, N. Dhar, H. Liu and K. Tam, Can. J. Chem. Eng., 2011, 89, 1191–1206.
5. T. Liebert, M. Kostag, J. Wotschadlo and T. Heinze, Macromol. Biosci., 2011, 11, 1387–1392.
6. H. Liu, D. Wang, S. Shang and Z. Song, Carbohydr. Polym., 2011, 83, 38–43.
7. A. Way, L. Hsu, K. Shanmuganathan, C. Weder and S. Rowan, ACS Macro Lett., 2012, 1, 1001–1006.
8. P. Lu and Y.-L. Hsieh, Carbohydr. Polym., 2010, 82, 329–336.
9. N. Wang, E. Ding and R. Cheng, Polymer, 2007, 48, 3486–3493.
10. P. Satyamurthy and N. Vigneshwaran, Enzyme Microb. Technol., 2013, 52, 20–25.
11. P. Filson and B. Dawson-Andoh, Bioresour. Technol., 2009, 100, 2259–2264.
12. N. Wang, E. Ding and R. Cheng, Langmuir, 2008, 24, 5–8.
13. X. Li, E. Ding and G. Li, Chin. J. Polym. Sci., 2001, 19, 291–296.
14. R. Xiong, X. Zhang, D. Tian, Z. Zhou and C. Lu, Cellulose, 2012, 19, 1189–1198.
15. B. Purkait, D. Ray, S. Sengupta, T. Kar, A. Mohanty and M. Misra, Ind. Eng. Chem. Res., 2011, 50, 871–876.
16. H. Lu, Y. Gui, L. Zheng and X. Liu, Food Res. Int., 2013, 50, 121–128.
17. M. Ibrahim, W. El-Zawawy and M. Nassar, Carbohydr. Polym., 2010, 79, 694–699.
18. J. Zhang, T. Elder, Y. Pu and A. Ragauskas, Carbohydr. Polym., 2007, 69, 607–611.
19. A. Leung, S. Hrapovic, E. Lam, Y. Liu, K. Male, K. Mahmoud and J. Luong, Small, 2011, 7, 302–305.
20. E. Lam, A. Leung, Y. Liu, E. Majid, S. Hrapovic, K. Male and J. Luong, ACS Sustainable Chem. Eng., 2013, 1, 278–283.
21. M. Salajkov, L. Berglund and Q. Zhou, J. Mater. Chem., 2012, 22, 19798–19805.
22. H. Kargarzadeh, I. Ahmad, I. Abdullah, A. Dufresne, S. Zainudin and R. Sheltami, Cellulose, 2012, 19, 855–866.
23. H. Öztürk, A. Potthast, T. Rosenau, M. Abu-Rous, B. MacNaughtan, K. Schuster, J. Mitchell and T. Bechtold, Cellulose, 2009, 16, 37–52.
24. S. Beck-Candanedo, M. Roman and D. Gray, Biomacromolecules, 2005, 6, 1048–1054.
25. R. Ibbett, J. Taylor, K. Schuster and M. Cox, Cellulose, 2008, 15, 393–406.
26. G. Sèbe, F. Ham-Pichavant, E. Ibarboure, A. Koffi and P. Tingaut, Biomacromolecules, 2012, 13, 570–578.
27. J. Han, C. Zhou, A. French, G. Han and Q. Wu, Carbohydr. Polym., 2013, 94, 773–781.
28. A. Oudiani, Y. Chaabouni, S. Msahli and F. Sakli, Carbohydr. Polym., 2011, 86, 1221–1229.
29. Y. Yue, C. Zhou, A. French, G. Xia, G. Han, Q. Wang and Q. Wu, Cellulose, 2012, 19, 1173–1187.
30. S. Fujisawa, Y. Okita, H. Fukuzumi, T. Saito and A. Isogai, Carbohydr. Polym., 2011, 84, 579–583.
31. S. Montanari, M. Roumani, L. Heux and M. Vignon, Macromolecules, 2005, 38, 1665–1671.
32. Y. Habibi, H. Chanzy and M. Vignon, Cellulose, 2006, 13, 679–687.

---