It is known that stable, non-sedimenting aqueous suspensions of rod-like crystalline particles can be prepared by appropriate treatments of native cellulose or chitin with mineral acids [1
]. These colloidal particles, which have recently been called “nanowhiskers” of these polysaccharides, are shortened fragments of bio-synthesized “microfibrils” [5
], in which linear molecules of the polysaccharides are aligned in parallel to form long-ranging fine fibers. While the original microfibrils have lengths of an order exceeding tens of micrometers, exceeding areas of microscopic observations, nanowhiskers usually comprise shorter length scales. For example, lengths are 100–300 nm for cellulose nanowhiskers derived from higher plants and α
-chitin nanowhiskers from crab shells [6
], 1–2 µm for those from bacterial cellulose [12
] and up to several micrometers for nanowhiskers from tunicins or green algae [13
]. The widths of nanowhiskers also vary according to their origin, namely, 2–10 nm for cellulose nanowhiskers from higher plants and crab chitin nanowhiskers, and up to 20 nm for tunicin or algal cellulose nanowhiskers. The cross-sectional size of ribbon-like bacterial cellulose nanowhiskers is reported to be ca.
10 × 50 nm [12
These nanowhiskers and their colloidal suspensions have been attracting significant attention due to their striking characteristics, one of which is their lyotropic liquid crystallinity. Above a critical concentration, suspensions of these nanowhiskers are reported to spontaneously separate into upper isotropic and lower anisotropic phases, the latter containing chiral nematic [8
] or nematic [15
] ordering. The formed liquid crystalline domains are also known to align their chiral nematic axis parallel to applied magnetic field to form long-ranging anisotropic domains [9
]. Under some conditions, an unusual “birefringent glassy phase” was formed, irrespective of concentration [16
Another intriguing feature of the nanowhiskers is their outstanding mechanical properties, i.e.
, extremely high modulus and strength. The Young’s moduli of tunicin cellulose nanowhiskers and squid pen chitin nanowhiskers are both estimated as 150 GPa from a bending test using an atomic force microscope [18
] and a calculation from the moduli of nanocomposites [19
], respectively. These high values of modulus and strength are attributed to formation of an almost perfect single crystal by parallel alignment of rigid polysaccharide chains, which are bound to each other with many strong hydrogen bonds. Much attention has been focused recently on the utilization of these nanowhiskers as nano-sized fillers in composites, due to their excellent mechanical properties, as well as non-toxicity, biodegradability and ease of large-scale preparation. For example, in the last two decades there have been numerous reports on various types of nanowhisker composites, including films containing matrices of latex or synthetic polymers and fillers of nanowhiskers [19
], high-modulus poly(vinyl alcohol) wet-spun fibers containing uniaxially oriented nanowhiskers [22
], hydrogels reinforced by nanowhiskers [26
] and electrospun fiber mats incorporating nanowhiskers [28
Good dispersion of independent nanowhiskers in matrices is quite significant for the preparation of the above-mentioned nanocomposites, because well-dispersed nanofillers usually increase their specific surface area to enhance interactions with matrices and thereby improve mechanical properties, whereas aggregations of nanofillers often act as defects in nanocomposites or cause undesirable stress concentration resulting in a significant reduction in mechanical properties. Dispersion stability of the nanowhiskers is mainly dependent on repulsive forces caused by surface charge, i.e.
, electrostatic stabilization. Negative charges of surface sulfate groups on cellulose nanowhiskers generated during sulfuric acid treatment [6
] and positive charges of inherent surface amino groups on chitin nanowhiskers [10
] are believed to be responsible for their stability. This electrostatic stability, however, was significantly reduced by addition of electrolytes via the shielding effect. Therefore, aqueous nanowhisker suspensions easily form aggregates to precipitate in the presence of electrolytes. The effect of electrostatic repulsion is also minor in media with low dielectric constants, because the range of electrostatic repulsion is proportional to the dielectric constant of the medium. Dispersion of nanowhiskers in organic solvents has not been achieved, except for those possessing high dielectric constants, such as dimethyl sulfoxide (DMSO) and dimethylformamide (DMF) [29
Another approach to solve the problem of electrostatic stabilization described above is so-called “steric stabilization”, in which polymers grafted or adsorbed on particle surfaces prevent the approach of the particle to prevent aggregation, resulting in a stable dispersion [30
]. In 2000 and 2001, two pioneering approaches to steric stabilization of cellulose nanowhiskers were examined by different research groups; one was preparation of cellulose nanowhiskers with surface-adsorbed surfactants, which gave a good dispersion stability in toluene or tetrahydrofuran [31
], while the other was surface grafting of poly(ethylene glycol) (PEG) onto cellulose nanowhiskers [33
]. Starting from these two researches, many literature articles have reported preparations of a wide variety of sterically stabilized nanowhisker suspensions, including ring opening living polymerization of °ε-caprolactone initiated by surface hydroxyls on cellulose nanowhiskers [35
], introduction of polystyrene [37
] or poly(N
-dimethylaminoethyl methacrylate) [38
] via atom transfer radical polymerization, single electron transfer living radical polymerization of N
-isopropylacrylamide on whisker surfaces [39
], and introduction of amine-terminated ethylene oxide-propylene oxide copolymers via amidation [40
]. Steric stabilization of nanowhiskers derived from chitin, a polysaccharide consisting of crab/shrimp shells and having a linear structure similar to that of cellulose, were also realized very recently [41
]. Although these samples can be useful for applications such as nanocomposite preparation, such a coexistence of steric and electrostatic stabilization may sometimes disturb a detailed investigation on a genuinely steric effect; namely, most of the above-mentioned studies use stable nanowhisker suspensions with surface charge groups, which remain after surface grafting and/or adsorption of polymers. When the carboxyl groups on nanowhiskers were used as the binding site for polymers, significant amounts of carboxyls remained unreacted and acted as surface charge groups [33
]. As a result, stability of the obtained suspensions is dependent not only on the steric repulsions of surface polymers but also on the charge repulsions by residual charge groups. To independently evaluate the effect of grafted or adsorbed polymers on stability, preparation of nanowhiskers that are free of charge and possess surface polymers, is required. Kloser and Gray [42
] first examined a preparation of such a sample by desulfation of nanowhiskers with 0.1 M NaOH treatment and subsequent grafting of poly(ethylene oxide), although very trace amount of sulfate groups (sulfur content of 0.04 mmol/g cellulose) could not be removed.
“Charge-free” cellulose nanowhiskers can be prepared by hydrolysis of native cellulose with hydrochloric acid and subsequent mechanical homogenization [33
]. In the present study, the authors examined preparation of a sterically stabilized cellulose nanowhisker suspension according to Scheme 1
, via esterification between surface hydroxyl groups of the “charge-free” nanowhiskers and terminal carboxyl groups of monomethoxy PEG (mPEG), mediated by 1,1'-carbonyldiimidazole (CDI).
Procedures for mPEG1000-COOH grafting onto “charge-free” nanowhiskers via CDI-mediated esterification.
Procedures for mPEG1000-COOH grafting onto “charge-free” nanowhiskers via CDI-mediated esterification.
Reagents and Conditions: (i) NaClO, NaBr, pH = 10–11, r.t., 1 h; (ii) DMAc or DMSO, r.t., 2 h; (iii) DMAc or DMSO, 80 or 100 °C, 1–7 day(s).
3. Experimental Section
Purified fibrous cotton (CF11, Whatman International Ltd., Maidstone, Kent, UK) was used as starting cellulose. 1,1'-carbonyldiimidazole (CDI) and the free radical form of 4-oxo TEMPO (4-oxo-2,2,6,6-tetramethylpiperidine-1-oxyl) was purchased from Tokyo Chemical Industry Co. Ltd. (Tokyo, Japan) and Wako Chemical Ltd. (Miyazaki, Japan), respectively. 3-Aminopropyl-functionalized silica gel (40–63 µm, ~1 mmol/g NH2 loading) and polyethylene glycol 1000 monomethyl ether (mPEG1000) were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). Other reagents were from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). Dimethylacetamide (DMAc) and dimethyl sulfoxide (DMSO) were dehydrated by reduced distillation after treatment with CaH2. Dry methanol was obtained after distillation following treatment with magnesium and iodine. Other reagents were used without further purification.
3.2. Preparation of Cellulose Nanowhiskers
An aqueous suspension of “charge-free” cellulose nanowhiskers was prepared according to previous reports [33
]. Briefly, air-dried CF11 (10 g) was treated with 2.5 M HCl (100 mL) at 105 °C for 15 min, followed by successive washing by filtration using deionized water until neutrality. The hydrolyzed sample was suspended in deionized water to form a 5%–10% slurry and homogenized with a Waring-type blender for 30 min. The non-sedimenting nanowhisker suspension was collected as a turbid supernatant through repeated centrifugation cycles (1600 g
, 5 min) of the homogenized sample and further dialyzed against deionized water to give a 0.5%–1% nanowhisker suspension. The initial ungrafted nanowhiskers and their suspension are designated as “ungrafted” hereafter. Yield from starting CF11 was around 40%–50%.
3.3. Synthesis of Terminal-Carboxylated mPEG (mPEG-COOH) via Oxidation Using Silica Gel Supported TEMPO
Terminal hydroxyl groups of mPEG1000 were oxidized into carboxyl groups according to previously published procedures [47
], with the modification of using silica gel particles carrying the TEMPO moieties instead of aqueous TEMPO solution. For this purpose, silica gel supported TEMPO was prepared via reductive amination of the amino groups on the silica gel and 4-oxo TEMPO, according to the literature [50
]. The silica gel product carrying TEMPO (1.00 g) was dispersed in a solution of mPEG1000 (10.0 g, 10.0 mmol -OH) and sodium bromide (100 mg, 0.972 mmol) in deionized water (100 mL). After adjustment of the pH to around 6 with 1 M HCl, sodium hypochlorite (NaClO) solution (21.1 mL, adjusted to contain 40.0 mmol of NaClO) was slowly added. The reaction mixture was stirred for 1 h at room temperature, while keeping the pH around 7–8 with dropwise addition of 2 M NaOH aqueous solution. After oxidation, 10 mL of ethanol was added to quench the oxidation through consumption of excess NaClO. The silica-gel catalyst was removed by filtration. The obtained filtrate was adjusted to pH 1 by addition of 1 M HCl, followed by extraction of oxidized mPEG with dichloromethane (100 mL × 3) and concentration by rotary evaporation. Yield of the oxidized mPEG1000, named mPEG1000-COOH hereafter, was 92.7%. The conversion ratio of the terminal hydroxyl groups was determined by pH titration using 0.1 M aqueous NaOH and a phenolphthalein indicator.
3.5. Qualitative Evaluation of Dispersion Stability of the PEG-Grafted Suspensions
The ungrafted and the PEG-grafted suspensions were adequately diluted with deionized water and/or 1 M NaCl aqueous solution to give equal nanowhisker content of 0.1% (excluding the weight of PEG in the cases of the PEG-grafted samples, by considering the weight increase, as in the next section) and NaCl concentrations of 0 or 0.1 M. The suspensions were observed between crossed polarizers to check the appearances of flow birefringence. Further, all of the suspensions were allowed to stand at room temperature for up to 24 h for visual observation of precipitation.
3.6. Evaluation of Amount of Grafted PEG
Evaluation of the amount of grafted PEG was performed by two methods. The first method was by the increase in sample weight by determining the concentration of suspensions before and after PEG grafting. From preliminary experiments, the final dialysis step in the preparation was found to completely remove unreacted mPEG1000-COOH. The other method was based on alkali hydrolysis of the ester moiety between nanowhiskers and mPEG1000. The PEG-grafted samples with a known value of solid content (i.e., known weight of nanowhiskers + PEG) were mixed with an equal amount of 1 M aqueous NaOH, followed by overnight stirring at room temperature. Through this alkali treatment, the ester moiety between nanowhiskers and mPEG1000 was believed to be completely hydrolyzed, whereas the nanowhiskers remained intact. The hydrolyzed samples were carefully neutralized with 3 M HCl, washed with deionized water by repeated centrifugation, and dried to yield the weight of “bare” nanowhiskers. The decrease in weight was determined by subtraction from the initial weight. The values of the grafted PEG weight determined by these two methods were converted into the PEG weight grafted to 1 g of nanowhiskers.
Size and shape of the nanowhiskers were observed by transmission electron microscopy (TEM). A drop of very dilute suspension was deposited on a TEM grid coated with a Formvar film, which was pretreated with 0.2% aqueous bacitracin solution. The dried grid was observed using a JEOL JEM-2100 instrument (JEOL Ltd., Tokyo, Japan) at 80 kV using defocus contrast technique. Attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectra of the freeze-dried samples were recorded on a Shimadzu IRPrestage-21 spectrometer with a DurasampleIR II ATR accessory (SensIR Technologies, Danbury, CT, USA) at a resolution of 4 cm−1 using 32 scans.