Skip to main content
NCBI home page
Royal Society of Chemistry logoLink to Royal Society of Chemistry
. 2026 Sep 30;28(40):24968–24978. doi: 10.1039/d6cp02653b

TEMPO-oxidised cellulose nanofibrils from bacterial cellulose: how the cellulose source influences the gelation mechanisms

Julien Schmitt a,†,✉, Vincenzo Calabrese a,‡, Kazi M Zakir Hossain a,§, Janet L Scott a,b, Karen J Edler a,¶,✉
PMCID: PMC13625619  PMID: 42813659

Abstract

TEMPO-oxidation of cellulose is used to form water-dispersible, negatively-charged nanofibrils. It is known that the type of cellulose source influences the yield of TEMPO-oxidation, the number of carboxylate groups and the crystallinity index. In this paper, we aim to demonstrate its influence on water-based formulations. Dispersions of TEMPO-oxidised bacterial cellulose nanofibrils (OBCNFs) in aqueous suspensions were studied using both small-angle X-ray scattering (SAXS) and rheology as a function of nanofibril concentration and addition of NaCl or sodium dodecyl sulphate (SDS), and the results were compared with those obtained previously for oxidised cellulose nanofibrils (OCNFs) from wood pulp. With the same oxidation procedure, OBCNFs exhibit a slightly higher oxidation degree than their OCNF counterparts (29% and ζ-potential = −56 ± 4 mV versus 25% and −48 ± 2 mV). They also exhibit a much larger length, as seen in TEM, with an almost ribbon-like architecture, whereas OCNFs form short stiff nanorods. As with OCNFs, OBCNF suspensions can form gels by increasing nanofibril concentration or upon addition of salt or SDS. Yet, the gels formed upon increased concentration have frequency dependent storage and loss moduli associated with a low yield strain (ca. 1%), consistent with the formation of a weak entangled network, while no interactions are observed in SAXS, whereas OCNF gels were formed due to an increase in repulsive interactions between nanofibrils. Upon salt/SDS addition, the gelation mechanisms consist of a gradual charge screening of the carboxyl groups and an increase of attractive interactions between OBCNFs with salt, and depletion attraction due to the presence of negatively-charged micelles for SDS, which are the same mechanisms observed for OCNFs. Yet, the rheological properties are strongly dependent on the type of nanofibril used. OBCNF gels exhibit, in most cases, higher viscosity and storage moduli than their OCNF counterparts under similar formulation conditions and a higher tolerance to salt before syneresis or to SDS before gelation, which allows for finer tuning of their rheological properties, relevant for the design of health-care products.

Hydrogels made with TEMPO-oxidised nanofibrils from bacteria are compared to those from wood-pulp.graphic file with name d6cp02653b-ga.webp

Introduction

In recent years, interest in bio-based materials has increased drastically, aiming to produce chemicals or products from renewable sources that are more environmentally friendly. Among these materials, cellulose, the most abundant natural polymer,1 is a promising candidate.2 Using chemical or mechanical processes, cellulose can be produced in various forms, such as nanofibrils2–5 or nanocrystals,6–8 and from various sources, such as wood pulp1,9,10 or bacteria.11,12 Cellulose is organised as fibrils made of glycosidic chains linked by β-1,4-glycosidic bonds and stabilised in a hierarchical organisation by strong hydrogen bonds. This structure prevents fibril dispersion in aqueous suspensions.13 Chemical modification of cellulose, for example, using sulfuric acid hydrolysis, allows the formation of needle-like cellulose nanocrystals (CNCs) that are dispersible in water thanks to the high negative charge at the surface of the nanocrystals.7,14 An alternative route is the so-called TEMPO-oxidation of cellulose, forming larger and partially amorphous nanorods, called oxidised cellulose nanofibrils (OCNFs).13,15 These OCNFs exhibit crystalline and amorphous regions and are also negatively charged and dispersible in water with long-term colloidal stability, thanks to the presence of carboxylate groups on their surfaces.

CNCs and OCNFs can be produced using various sources of cellulose: wood pulp,16,17 corncobs,18 bagasse,19,20 tunicates21 and bacteria,12 to name a few. The source has a strong influence on the properties of the obtained cellulose: for example, cellulose from wood pulp and other plant-based sources is mixed with hemicellulose and lignin, while cellulose obtained from bacteria is highly pure. As cellulose contains both crystalline and amorphous regions, its crystallinity also depends on the source, and the degree of polymerisation of the glycosidic unit is also different: 2000–4000 for cellulose from wood pulp versus 7000–16 000 for bacterial cellulose, resulting in very long microfibrils in the latter. This in the end influences the mechanical properties of cellulose fibrils, with an elastic modulus of 14–40 GPa for cellulose from wood pulp versus 60–115 for bacterial cellulose.22 After treatment such as TEMPO-oxidation or reaction with sulphuric acid, such differences are expected to influence the shape, size, crystallinity and surface charge of the particles. For example, CNCs made from wood pulp have lengths between 100 and 200 nm while those produced from bacteria have lengths between 100 and 1000 nm.22 Recently, Pratama et al. studied the structural properties of TEMPO-oxidised cellulose nanofibrils from various sources, such as corn cob, bagasse, waste-wood and bacterial cellulose.23 They showed that the type of source influences the yield of the TEMPO-oxidation procedure, the amount of carboxylate groups on the fibril surface and the crystallinity index. They notably showed that nanofibrils obtained from bacterial cellulose display a higher crystallinity index (ca. 83% compared to 73%), carboxylate content (almost 8500 mmol kg−1 compared to ca. 7500) and yield (31% versus 29%) than their counterparts obtained from wood waste, such as wood pulp.23 Yet, they have not studied the influence on the shape of the nanofibrils obtained or on the rheological properties of nanofibril suspensions.

OCNF aqueous suspensions exhibit shear-thinning properties, and can be used to form hydrogels, modulated by concentration,24 pH,25 addition of salt,26 surfactants27,28 or alcohol.29 We previously studied OCNFs that were prepared using wood-pulp as the source. We showed that OCNF suspensions undergo a liquid-to-gel transition at a nanofibril concentration of ca. 1.5–2 wt%, due to the increase of repulsion between the negatively-charged nanofibrils.30 Addition of salt to a 1 wt% OCNF suspension also induced gelation at NaCl concentrations of ≥0.1 M, this time due to charge-screening of the nanofibrils.30 Finally, surfactants such as sodium dodecyl sulphate (SDS, negatively charged) and dodecyltrimethylammonium bromide (DTAB, positively charged) can also trigger gelation, either by depletion attraction or by triggering nanofibril aggregation.31 Yet, the effect of the cellulose source on the hydrogel properties is less studied.

In this paper, we focused on the formation of hydrogels made using TEMPO-oxidised nanofibrils obtained from bacterial cellulose, referred to as oxidised bacterial cellulose nanofibrils (OBCNFs). We studied their morphological properties using TEM and SAXS, and compared their gelation mechanisms and gel properties with those of previously investigated formulations made using wood pulp, referred to as OCNFs. We studied the effects of nanofibril concentration and addition of salt (NaCl) and surfactant (SDS), as these are common methods used in the formulation of health and personal care products, where both OCNFs and OBCNFs could act as valuable rheological modifiers thanks to their renewability, eco-compatibility and biocompatibility. The larger dimensions of OBCNFs compared to OCNFs, coupled with their differences in shape and flexibility, have a clear influence on the rheological properties of the hydrogels, proving that the source of cellulose is also an important factor to consider in cosmetic formulation.

Materials and methods

Reagents

The Acetobacter organism was obtained from Happy Kombucha, UK, and used to grow bacterial cellulose. Glucose, yeast extract, peptone, anhydrous disodium phosphate and citric acid monohydride were purchased from Sigma-Aldrich, UK. For the TEMPO oxidation of bacterial cellulose and cellulose cleaning, a 2,2,6,6-tetramethylpipiridine 1-oxyl radical (TEMPO), NaBr, NaOCl 5.00 vol%, HCl (reagent grade) and NaOH (pellets) were purchased from Sigma-Aldrich. For sample preparation, NaCl and sodium dodecyl sulphate (SDS) were also purchased from Sigma-Aldrich. All reagents were used as received.

OBCNF production

Oxidised bacterial cellulose nanofibrils (OBCNFs) were obtained by TEMPO oxidation of bacterial cellulose. First, bacterial cellulose sheets were produced following an already established protocol,32 by adding Acetobacter to a 10 L aqueous suspension containing glucose (2 wt%), yeast extract (0.5 wt%), peptone (0.5 wt%), Na2PO4 (0.27 wt%) and citric acid (0.15 wt%). After ca. 2 weeks, a thick cellulose sheet is recovered, which was then sterilised and bleached by immersing it in a 2 L 5% sodium hypochlorite solution for 2 h. Then, bacterial cellulose sheets were rinsed extensively using deionised water (DI) before freeze-drying. TEMPO oxidation was then carried out as follows: TEMPO (0.016 g, 0.1 mmol) and NaBr (0.1 g, 1.0 mmol) were added to 200 mL of DI in an ice bath. 1–2 g of bacterial cellulose were added to the solution, under vigorous stirring. Another solution of NaOCl at 5 vol% was prepared and its pH was adjusted to 10 using 0.1 M HCl. After 10 min, the NaOCl solution was added dropwise to the suspension. The amount of NaOCl solution added was fixed to have 0.0.5–0.3 mol equivalents relative to anhydroglucose units (AGU) of the bacterial cellulose. The pH of the solution was kept at 10 by dropwise addition of 0.5 M NaOH when required. Once the pH of the suspension stabilised, the reaction was quenched by adding 10 mL of ethanol. The resulting oxidised bacterial cellulose nanofibrils (OBCNFs) were then dialysed against DI for 3 days, before freeze-drying. Then, a stock solution of OBCNF at 2.5 wt% in aqueous solution was prepared by ultrasonication using an Ultrasonic Processor, FB-505 sonication probe from Fisher. Ultrasonication was typically performed on 15 mL of suspension using a series of 1 s on/1 s off pulses for a total time of 2 min, at 40% amplitude, equivalent to 200 W.

OCNF production

Oxidised cellulose nanofibrils (OCNFs) made from wood pulp were also used for comparison. OCNFs were TEMPO-oxidised, and then they were purified by dialysis to remove salts and preservatives following an already described procedure.30 After purification, OCNFs were freeze-dried and then redispersed in a 2 wt% aqueous suspension via the same ultrasonication procedure used for OBCNFs.

Conductometric titration of OBCNFs

The carboxylate content of OBCNFs was determined by conductometric titration: 40 mg of OBCNFs were dispersed in 15 mL of 10 mM HCl standard solution for 10 min. A titrant (10 mM NaOH) was added in 0.5 mL aliquots and the conductivity was monitored using a Seven Multi Mettler Toledo conductivity probe. The degree of oxidation (DO) is calculated as follows:graphic file with name d6cp02653b-t1.webpwhere C is the concentration of NaOH (M), w is the weight of the OBCNF added (g), V is the volume of NaOH added (dm3), 162.15 g mol−1 is the MW of the AGU and 35.97 is the difference in MW between the AGU and the sodium salt of the glucuronic acid group.33 The DO of the OCNF was similarly determined, as previously reported.34

Transmission electron microscopy (TEM)

TEM measurements of OCNF and OBCNF were carried out using a JEM-2100 Plus (JEOL, USA) working at an operating voltage of 200 kV. The grid was prepared by depositing a drop of O(B)CNF suspension at 0.025 wt% with uranyl acetate (2 M) for enhancing contrast, before removing the excess of solution and allowing the grid to dry under ambient conditions. For OBCNFs, analysis of TEM micrographs using Image J (National Institutes of Health, USA) allowed extracting the length of the nanofibrils, using an average of 50 measurements. The large extension of the nanofibrils prevented more measurements to be carried out as some OBCNFs are longer than the field of view. For OCNFs, analysis was carried out on 175 measurements, and both the radius and length were extracted.30

Zeta-potential measurements

ζ-Potential measurements were carried out using a Malvern Zetasizer Nano ZSP (Malvern, UK) by preparing a suspension of OBCNF at 0.1 wt% in DI water at 25 °C. The value obtained was the average of 20 measurements of 30 s each, using the Smoluchowski model to convert the electrophoretic mobility into ζ-potential.

Sample preparation

OBCNF hydrogels studied by SAXS and rheology were prepared as follows: first, stock solutions of OBCNF 2.5 wt%, NaCl 1 M and SDS 100 mM were prepared. A series of concentrations were obtained by mixing OBCNF stock solutions with an appropriate amount of DI water with gentle stirring. The final OBCNF concentrations of 0.75, 1, 1.5, 2 and 2.5 wt% were obtained. The influence of salt on OBCNF hydrogels was studied using 1 wt% OBCNF suspensions with various NaCl concentrations: 0, 0.05, 0.1, 0.2, 0.3, 0.4 and 0.5 M. Finally, the influence of SDS was also studied on 1 wt% OBCNF suspensions with various SDS concentrations of 0, 20, 40, 60 and 80 mM. These concentrations were chosen as they allow some comparison with hydrogels obtained from TEMPO-oxidised cellulose nanofibrils obtained from wood pulp.30,31 Each sample was prepared with a volume of 5 mL. To complete the salt study, OCNF suspensions were also prepared at the same nanofibril (1 wt%) and salt (ranging from 0 to 0.5 M) concentrations using a 2 wt% OCNF stock solution.

Rheology

Rheological studies were performed using a Discovery HR-3 stress-controlled rheometer (TA Instruments) using a sandblasted plate-plate geometry (12 mm in diameter and 1 mm gap). Approximately 1 mL of OBCNF suspensions was deposited between the plates, and the geometry was covered with a thin layer of mineral oil to avoid evaporation. To characterise the viscoelastic regime of the samples, amplitude strain sweeps of the storage and loss moduli were conducted at ω = 10 rad s−1. Frequency sweeps were carried out at a strain amplitude of 0.1% to be within the linear viscoelastic regime. The viscosity response of the samples to shear was monitored by steady flow measurements, with a shear rate ranging from 10−2 to 102 s−1.

SAXS

SAXS measurements were performed using the Swing beamline at SOLEIL synchrotron (Saclay, France) for the concentration series and at I22 beamline at Diamond Light Source (Didcot, Oxfordshire, UK) for salt and SDS measurements. At Swing, the X-ray beam had an energy E of 12 keV and a wavelength λ of 1.03 Å. The q-range probed was 4 × 10−3–0.4 Å−1, with Inline graphic the scattering vector depending on the scattering angle θ. Data were collected using a EigerX4M in vacuum (Dectris). At I22, the X-ray beam had an energy E of 12.4 keV and a wavelength λ of 1 Å. The detector used to collect the data was a Pilatus P3-2M (Silicon hybrid pixel detector, DECTRIS) and the probed q-range was 0.035–0.3 Å−1. In both cases, multiple 100 ms frames were recorded for a total of 1 s. This timeframe was chosen to avoid radiation damage, checked by the absence of variation of the SAXS patterns between frames. Samples were prepared using 1.5 mm diameter glass capillary tubes. In order to work in absolute scaling, scattering patterns of the empty capillary and the capillary filled with DI water were also recorded, as water exhibits a constant contribution in intensity in the SAXS regime equal to 0.016 cm−1. Capillaries containing water or saline solutions were also recorded for solvent subtraction.

SAXS data modelling

SAXS patterns were fitted using the model of (interacting) cylinders with an elliptical cross-section, previously used for oxidised cellulose nanofibrils from wood pulp.30

The scattered intensity I(q) (cm−1) is written as:I(q) = nΔρ2P(q)S(q) + bkgwhere n is the number of objects per unit volume, Δρ = ρc − ρ0 is the difference between the scattering length density of the dispersed particles (ρc) and the scattering length density of the solvent (ρ0) (in cm−2), P(q) is the form factor of the cylinders (in cm−6), and S(q) is the structure factor and describes interference of the X-ray beam with multiple objects, thus usually associated with particle–particle interactions. Finally, bkg is the background.

The form factor for OBCNFs was chosen as the one for rigid cylinders with a length L, an elliptical cross section (with a major radius Rmax and an ellipticity Inline graphic ratio between the minor and the major radii, Rmin and Rmax respectively) and a uniform scattering length density (ρc), in agreement with previous studies.30,31P(q) = P(q, Rmax, ε, L, σ) = 〈FCS2(q, Rmax, ε)〉σProd(q, L)with FCS the form factor associated with the elliptical cross-section:graphic file with name d6cp02653b-t4.webpand Prod the form factor of an infinitely thin rod of length L:graphic file with name d6cp02653b-t5.webpwith Inline graphic. We note that the value of L cannot be measured in the probed q-range, and was arbitrarily fixed at ca. 1 µm, in agreement with TEM measurements. The polydispersity in length was not taken into consideration, as again it lies out of the probed q-range and could not be accurately measured in our configuration.

〈·〉σ symbolises the polydispersity in size on the cross-section, which is described via the Schulz–Zimm distribution fSZ, i.e.:graphic file with name d6cp02653b-t7.webpwith Inline graphic

In contrast to spheres, calculating the exact structure factor is complex as the cylinder–cylinder interaction is strongly dependent on the particle relative orientation. The polymer reference interaction site model (PRISM) uses mean-field theories and Monte-Carlo simulations to obtain an analytical structure factor:35,36graphic file with name d6cp02653b-t9.webpwith Inline graphic characterising the strength of the interaction U(r) between cylinders normalised by kBT (kB is the Boltzmann constant and T is the temperature), and Inline graphic describing a local excluded volume of radius Rv around each point along the length of the cylinder. For repulsive interactions between nanofibrils, such as the ones experienced between two negatively charged nanofibrils, νRPA > 0; while, νRPA < 0 would be associated with attractive interactions between them. The parameter Rv ≥ Rmax ensures, similarly to the hard-spheres model for spherical particles, that two cylinders cannot be in closer contact than Rv, i.e. cannot occupy the same space.

Results

Characterisation of OBCNFs and comparison with OCNFs

OBCNFs present a degree of oxidation, measured by conductometric titration, of 29% (see Fig. S1, SI), representing the amount of carboxylate groups grafted onto the surface per anhydroglucose unit. This results in a strong negative charge on the surface of the nanofibrils, as evidenced by ζ-potential measurements, with a ζ-potential found at −56 ± 4 mV. Both the degree of oxidation and ζ-potential are found similar to OCNFs made from wood pulp (25% and −48 ± 2 mV, respectively).30

OBCNF morphology was observed using TEM and found to be of very long fibrils, with lengths that can reach up to a few µm, only with a very anisotropic cross-section, resulting in an almost ribbon-like shape (Fig. 1). TEM micrographs evidence an obvious flexibility of the fibrils, with frequent twisting. An estimation of the fibril length, via a number average on 50 measurements from TEM images, gives a length L = 830 ± 400 nm. This very broad standard deviation is associated with the presence of “short fibrils” of lengths typically around 500 nm along with much longer fibrils (>1 µm), as can be seen in the length distribution (Fig. 1c). Note that very long fibrils, often larger than the image size, could not be measured, inducing a bias towards short fibrils in the size distribution presented here.

Fig. 1. (a) and (b) TEM micrographs of OBCNFs at 0.025 wt%, stained with uranyl acetate. (c) Size distribution of the length of the fibrils over 50 measurements.

Fig. 1

Open in a new tab

Information about the cross-section of the fibrils was obtained via SAXS studies of OBCNF suspensions at different concentrations (Fig. 2 and Table 1). For all concentrations, the ribbon-like morphology of the fibrils could be well-modelled by the rigid cylinders model from Materials and methods with a length L fixed at 1 µm (as it could not be probed at this q-range), an elliptical cross-section with Rmax = 3.7 ± 0.1 nm, Rmin = 1.1 ± 0.2 nm and a polydispersity in radius fixed at σ = 30%. The large polydispersity in size could be associated with a broad-range distribution of cross-sections, similar to what is observed for the length. We note that models of flexible cylinders exist in the literature,30 yet these were not required to fit the data, probably as the persistence length where the flexibility of the fibril is observed is, similar to the overall length L, not accessible in the probed q-range.

Fig. 2. SAXS patterns of OBCNF suspensions at concentration ranging from 0.75 to 2.5 wt%. Fits of the data are given as black lines.

Fig. 2

Open in a new tab

Table 1. Dimensions of the cellulose nanofibrils found by SAXS (radii and polydispersity) and TEM (length) of OCNF and OBCNF dispersions in water. OCNF dimensions were studied in ref. 30.

Cellulose R max (±0.1 nm) R min (±0.2 nm) σ (±5%) L (nm)
OCNF 5.7 1.4 — 160 ± 60
OBCNF 3.7 1.1 30 830 ± 400
Open in a new tab

By comparison, OCNFs, made from wood pulp instead of bacteria, are much smaller nanofibrils. Their dimensions were previously characterised using a combination of TEM and SAXS,30 and TEM micrographs and SAXS pattern (at 1 wt%) are also presented in Fig. 3. Table 1 gives a summary of OCNF dimensions from a previous study. Not only do they exhibit shorter dimensions, but OCNFs are also much less polydisperse in size and much more rigid, without the twisting observed for OBCNFs. Notably, their cross-section could be described without using a polydispersity in radius.

Fig. 3. (a) and (b) TEM micrographs and (c) SAXS pattern at 1 wt% and its associated fit of OCNFs. These nanofibrils were previously studied in ref. 30.

Fig. 3

Open in a new tab

Results indicate that the cellulose source has a strong influence on the final fibril dimensions, while the TEMPO oxidation procedure results in a similar degree of oxidation. This variation of shape, however, does influence the rheology and gelation properties of OBCNF suspensions when compared to OCNFs.

Effect of concentration

Fig. 2 presents the SAXS patterns of OBCNF suspensions for concentrations ranging from 0.75 to 2.5 wt%. Outside of an overall increase of the signal with OBCNF concentration, associated with an increase in the number of fibrils, no change in the SAXS pattern, due to an arising fibril–fibril interaction, was observed at this q-range and for the concentrations studied. All data could be fitted using the same form factor and particle dimensions. In contrast, OCNFs were previously found to exhibit a change in the SAXS patterns when increasing concentration, modelled by increasing repulsive interactions between nanofibrils, evidenced already at 1.5 wt%.30 We note that due to the much larger nature of OBCNF, interactions may appear at smaller q than available in our measurements.

The rheology of these OBCNF dispersions was measured, showing shear-thinning properties at all concentrations (with a zero-shear viscosity at ca. 100 Pa s for a 2 wt% OBCNF suspension, Fig. 4a), but also very weak gel-like properties for concentrations of ≥1.5 wt% (G′ ≥ G″). Indeed, G′ and G″ remain largely frequency-dependent even at 2.5 wt% (Fig. 4b), and all the gels exhibit a low yield strain (stress where the G′ value deviates from ca. 5–10% from the plateau in the amplitude sweeps) at ca. 1% (Fig. 4c). These data, associated with the lack of interaction observed in SAXS, indicate that OBCNFs, even at 2.5 wt%, form only a loosely entangled network, easily disentangled even at low strains. The corresponding OCNFs were also found to exhibit shear-thinning properties (with a zero-shear viscosity at almost 200 Pa s for a 2 wt% OCNF suspension) and a gel-like behaviour arising at 1.5 wt%, yet with little frequency dependence and a high yield strain (ca. 30% for a 2 wt% OCNF suspension). In this case, the liquid–gel transition was found to correspond to the emergence of fibril–fibril interactions measured in SAXS.30 Hence, the OCNF for the same concentration range forms a more rigid, percolated network driven by repulsion between nanofibrils, explaining the low frequency dependence and high yield strain. The flexible nature of OBCNF compared to OCNF, coupled with their higher aspect ratio, may be responsible for the rheological observations: at the same concentration, OBCNFs exhibit fewer objects than OCNFs, which are both larger in their length and more flexible as seen by the twisting in TEM. Repulsion between fibrils is hence less present, and the flexibility of the particles creates some transient entanglement points, easily rearranging with time, resulting in a weakly entangled network. On the other side, OCNFs, which are stiff rods, have a higher number of nanofibrils, meaning more junction points in the network fixed by fibril–fibril repulsion. The stronger repulsion between nanofibrils avoid aggregation and give more permanent junction points with time than entanglement. In summary, study of SAXS and rheology data suggests that OBCNFs form a weakly entangled network, easily sliding upon shear; while rigid OCNFs form a tougher percolated network relying on nanofibril electrostatic repulsion.

Fig. 4. Rheology measurements of OBCNF suspensions at concentrations ranging from 0.75 to 2.5 wt%: (a) shear viscosity, (b) frequency sweeps and (c) amplitude sweeps. For (b) and (c), solid symbols correspond to G′ values, whereas open symbols correspond to G″ values.

Fig. 4

Open in a new tab

Effect of ionic strength

We previously studied an OCNF suspension at 1 wt% without (0 M) and with NaCl (0.1 M).30 We showed that the addition of salt results in the formation of a stiff gel, as seen in rheology. This gelation was associated with a surface charge screening seen in ζ-potential that resulted in the emergence of attractive interactions seen in SAXS. To go deeper in this analysis and draw some comparison between OCNFs and OBCNFs, both systems were studied using 1 wt% suspensions and with NaCl concentrations ranging from 0 M to 0.5 M. Rheological studies of OCNFs at 1 wt% with salt are provided in Fig. S2, SI. A suspension without salt shows a very weak shear thinning behaviour and a value of G′ ∼ G″ over the whole frequency sweep, with a strong variation of both parameters with frequency. However, the addition of salt results in much stiffer, gel-like, suspension already at 0.1 M; with values of G′ that only weakly vary with ionic strength. From 0.2 M onwards, the system sees little variation with salt concentration, in terms of viscosity (with a viscosity η = 400 Pa s at Created by potrace 1.16, written by Peter Selinger 2001-2019 = 0.1 s−1), elastic modulus (with G′ at ca. 100–200 Pa), tan δ = G″/G′ (0.7–0.8) or yield strain (observed at ca. γ = 50%).

Similar to OCNFs, OBCNFs also form gels with the addition of salt (see Fig. 5a and Fig. S3), with, for a suspension at 1 wt%, already a gel-like behaviour in 0.05 M NaCl. The gel is defined by G′ ≫ G″ (tan δ = G″/G′ = 0.16 ± 0.04 on average) and very weak frequency dependence of G′. Upon addition of NaCl, the resulting gels become stiffer, with a higher storage modulus G′ while the ratio G″/G′ remains the same. At 0.5 M of NaCl, G′ is found at 2000 Pa; which is more than ten times higher than values obtained for OCNFs under the same conditions. Yet, let us note that the yield strain is found lower for OBCNFs than OCNFs (γ = 2–3% even at 0.5 M), again probably related to the more flexible nature of OBCNF that allows for disentanglement already at low strain. Hence, not only are OBCNF gels made with salt stiffer than their OCNF counterparts for the same weight fraction of nanofibril and ionic strength, the formulation also exhibits more flexibility in tuning their rheological properties by varying the amount of salt added.

Fig. 5. Results for a 1 wt% OBCNF suspension with various concentrations of NaCl (ranging from 0.05 to 0.5 M): (a) frequency sweeps (G′ values are shown as solid symbols and G″ values as open symbols) and (b) SAXS patterns (fits of the data are given as black lines).

Fig. 5

Open in a new tab

These differences are also observed in SAXS (see Fig. 5b and Fig. S4 for OBCNFs and OCNFs, respectively). Indeed, addition of salt results in the increase of an attractive interaction between nanofibrils, modelled using the interaction parameter νRPA < 0 and the excluded radius around the fibril Rv (nm), which are observed for both OCNFs and OBCNFs. The stronger the attraction, the more negative νRPA becomes, while Rv becomes comparable to the nanofibril cross-section. For OCNFs, both νRPA and Rv are found constant between 0.1 and 0.4 M of NaCl, while both parameters decrease for OBCNFs, inducing stronger interactions, and a smaller excluded volume upon salt addition (see Fig. 6). Let us note that the actual value of νRPA is highly dependent on the fibril dimensions, and notably their length, which has been arbitrarily fixed at 1000 nm for OBCNFs as it is outside of the probed q-range. This prevents a direct comparison of νRPA between OCNFs and OBCNFs, but allows for studying the evolution of the parameter with ionic strength for each series. At 0.5 M, the SAXS pattern for OCNFs cannot be modelled accurately only with attractive fibrils, as an extra signal is observed at a small angle (see Fig. S4b). Particle aggregation is a hypothesis to explain this extra signal, as 0.5 M corresponds to the emergence of syneresis of the gel for this system (observed via visual observation after ca. 2–3 days). In contrast, the SAXS pattern for OBCNFs is still well modelled simply by attractive particles at this concentration. The continuous increase of attraction forces between OBCNFs combined with the decrease of the excluded radius Rv around fibrils may induce a stiffer network with ionic strength, in agreement with the rheological data.

Fig. 6. Evolution of (a) νRPA and (b) Rv as a function of NaCl concentration from fitting of the SAXS patterns of OCNF (blue curve) and OBCNF (red) suspensions at 1 wt%. For OCNFs, values at [NaCl] = 0.5 M are not given as the fit does not reproduce the data well. The lines are for guiding the eye only.

Fig. 6

Open in a new tab

Addition of SDS

Another trigger for OCNF gelation lies in the use of surfactants. Notably, we previously showed that an OCNF suspension at 1 wt% could form gels in the presence of SDS already at 20 mM.31 Gels were studied from 20 to 80 mM and exhibited strong frequency dependence, with G′ values ranging from 1.5 to 3.5 Pa upon the frequency sweep (ω = 0.1 to 100 rad s−1) at the highest SDS concentration probed, 80 mM. Small-angle neutron scattering showed that SDS micelles were still present in solution, and that gelation was probably due to depletion attraction between OCNFs due to the presence of SDS micelles.

According to Liang et al., the critical volume fraction of depletant to trigger depletion forces should decrease when increasing the size of the particles, in the case of spherical particles.37 A similar effect may happen when increasing the length of rod-like particles. We hence studied suspensions of OBCNFs at 1 wt% with SDS, using concentrations ranging from 20 to 80 mM (Fig. 7a and Fig. S5). Like OCNFs, OBCNF gelation in suspension can be triggered by SDS addition. Nonetheless, the gels obtained are very weak, and are clearly measurable only above 60 mM of SDS added. Yet, we note that the elastic modulus G′ is always found to be above 1 Pa even at 20 mM SDS and reaches up to 10 Pa at 80 mM SDS. These values are much higher than their OCNF counterparts. Furthermore, similar to OCNFs, attractive interactions are measured between OBCNFs in SAXS when SDS is added (Fig. 7b), with νRPA reaching −0.42 ± 0.02 at [SDS] = 80 mM. This suggests that although a depletion attraction is present when SDS is added, its effect on the rheological properties of the gel is much less pronounced, again probably due to the differences between the entangled network formed by larger and more flexible OBCNF compared to the percolated network of the rigid OCNF counterparts.

Fig. 7. Results for a 1 wt% OBCNF suspension with various concentration of SDS (ranging from 20 to 80 mM): (a) frequency sweeps (G′ values are shown as solid symbols and G″ as open symbols) and (b) SAXS patterns (fits of the data are given as black lines).

Fig. 7

Open in a new tab

In summary, OBCNFs, similar to OCNFs, can form hydrogels in aqueous solution by using the concentration increase and addition of salt or surfactants. Their viscosity upon shear exhibit shear-thinning properties, while frequency sweeps of the hydrogels show more or less pronounced frequency dependency, characteristic of physical hydrogels. Similar results were obtained with microcrystalline cellulose (MCC) in the presence of a base to partially dissolve MCC38–40 or by cross-linking carboxymethyl cellulose (CMC).41,42 O(B)CNFs have the main advantage that they are directly water-dispersible since they are negatively charged, which allows the use of the electrostatic interactions to tune the hydrogel properties, without the need of cross-linking or working at highly basic pHs. Furthermore, our findings show that the source of cellulose does not influence only the yield of the TEMPO-oxidation reaction and the amount of carboxylate groups under similar oxidation conditions, but that it has a drastic influence on the particle shape and flexibility, and ultimately the rheological properties of dispersions. OBCNFs are much longer and flexible particles compared to OCNFs from wood pulp, with an almost ribbon-like morphology, which when dispersed in water give rise to an entangled physical gel with increasing concentration. This size and shape difference compared to OCNFs allows the OBCNF dispersion to accommodate higher concentrations of SDS before showing signs of gelation, or higher concentration of salt before showing signs of syneresis. It also modulates the rheological properties of the gels formed, with storage moduli always found similar or higher for OBCNF formulations than OCNF under similar formulation conditions. Indeed, as both OBCNF and OCNF exhibit clear negative surface charges (ζ-potential = −56 ± 4 mV versus −48 ± 2 mV), these rheological differences can clearly be related to variations in nanofibrils morphologies. As the source of cellulose has a strong influence on its purity, crystallinity and overall morphology even before TEMPO-oxidation,22 it remains a key parameter to control the overall size and shape of the TEMPO-oxidised nanofibrils, as seen with the comparison between OCNFs and OBCNFs. Other parameters such as pre-treatment (i.e. mechanical grinding) could also influence the overall morphology of the nanofibrils. These results are key for the use of OBCNFs or other bio-based products in healthcare or personal care products, which often relies on the addition of rheology modifiers but also salt and surfactants, but to a wider extent it can also influence the properties of OBCNF-based aerogels43 or Pickering emulsions44 compared to OCNFs.45

Conclusion

OBCNFs, oxidised cellulose nanofibrils obtained from bacteria, were prepared, characterised and used for the formulation of hydrogels. Their characteristics and the properties of the hydrogels formed as a function of concentration and addition of NaCl and SDS were studied and compared with results previously obtained for OCNFs, oxidised cellulose nanofibrils obtained from wood pulp. Compared to OCNFs, OBCNFs exhibit much larger lengths and overall more polydisperse dimensions. They exhibit some clear flexibility as seen by TEM, notably exhibiting some twisting, whereas OCNFs are shorter, better defined but stiff. Yet, the TEMPO-oxidation procedure does result in similar carboxylate contents on the surface of OBCNFs when compared to OCNFs (albeit slightly higher for the former, 29% versus 25%), with both nanofibrils exhibiting highly negative surface charges as seen in ζ-potential measurements. This change in morphology induces some clear differences in the gel properties.

All aqueous formulations display a shear-thinning viscosity, consistent with what is observed for OCNFs or other fibril-like systems. Without additives, the gel-like behaviour of OBCNF suspensions is observed above 1.5 wt% in frequency sweeps. The high frequency dependence and low yield strain even at 2.5 wt% suggest that OBCNFs form a weak physical gel, due to fibril entanglement. OCNF formulations also showed the formation of a gel at similar concentrations, but with weak frequency dependence and higher yield strain. Furthermore, although gelation for OCNFs could be related to the increase of repulsive interactions between nanofibrils as seen in SAXS, such a behaviour was not observed for OBCNFs. Yet, similar to OCNFs, stiff gels can be formed when adding a salt such as NaCl. Compared to OCNFs, OBCNFs in the presence of salt exhibit stiffer gels with higher values of G′, and show a wider range of variations in their rheological properties as a function of NaCl concentration. Indeed, while OCNF gels show no difference above 0.1 M NaCl, the G′ values of OBCNF gels keep increasing up to 0.5 M NaCl. In SAXS, this gelation is clearly related to the increase of an attractive interaction between nanofibrils, similarly to what was evidenced for OCNFs, and can be hence related to a screening of the negative charges at the surface of nanofibrils. Finally, gelation can also be triggered by the use of the surfactant, e.g. SDS; although, this time, the gelation using OBCNFs is much less obvious, even at 80 mM. SAXS analysis still evidences some fibril–fibril attraction, suggesting, like for OCNFs, that gelation is due to depletion attraction. Hence, these results suggest that, although OBCNF hydrogels can be formed upon increasing the concentration or by adding salt or surfactant, as with OCNFs, the nature (for concentration) or the strength (for salt or surfactant) of the physical interactions and the macroscopic rheological properties observed are highly dependent on the overall shape and size of the nanofibrils. As both OBCNFs and OCNFs could be used for the formulation of skin-care products, these findings allow using the source and thus the size of the nanofibrils as an extra tool to finely tune the rheological properties of the hydrogels formed.

Conflicts of interest

There are no conflicts to declare.

Supplementary Material

CP-028-D6CP02653B-s001

Acknowledgments

The authors thank the EPSRC for funding this project (grant EP/N033310/1). Vincenzo Calabrese thanks the University of Bath for PhD studentship funding. Dr Marcelo A. da Silva is gladly thanked for his help in the SAXS data measurements at Diamond and Soleil. Dr Nick Terril and Dr Andy Smith on I22 are thanked for their assistance with SAXS experiments at Diamond Light Source Ltd (experiments no. SM3691 and SM16364-1). Dr Thomas Bizien and Dr Javier Pérez on Swing at SOLEIL are also thanked (proposal 20170502). Mrs Ursula Potter (Microscopy and Analysis Suite, University of Bath) is thanked for her assistance in TEM measurements.

Data availability

Data for this article, including SAXS data files and fits and rheology (amplitude and frequency sweeps and viscosity upon shear) are available at Science Data Bank at https://doi.org/10.57760/sciencedb.42823.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d6cp02653b.

References

  1. Klemm D. Heublein B. Fink H. Bohn A. Cellulose: Fascinating Biopolymer and Sustainable Raw Material. Angew. Chem., Int. Ed. 2005;44(22):3358–3393. doi: 10.1002/anie.200460587. [DOI] [PubMed] [Google Scholar]
  2. Johnson R. K. Zink-Sharp A. Renneckar S. H. Glasser W. G. A New Bio-Based Nanocomposite: Fibrillated TEMPO-Oxidized Celluloses in Hydroxypropylcellulose Matrix. Cellulose. 2009;16(2):227–238. doi: 10.1007/s10570-008-9269-6. [DOI] [Google Scholar]
  3. Crawford R. J. Edler K. J. Lindhoud S. Scott J. L. Unali G. Formation of Shear Thinning Gels from Partially Oxidised Cellulose Nanofibrils. Green Chem. 2012;14(2):300–303. doi: 10.1039/C2GC16302K. [DOI] [Google Scholar]
  4. Nechyporchuk O. Belgacem M. N. Bras J. Production of Cellulose Nanofibrils: A Review of Recent Advances. Ind. Crops Prod. 2016;93:2–25. doi: 10.1016/j.indcrop.2016.02.016. [DOI] [Google Scholar]
  5. Siró I. Plackett D. Microfibrillated Cellulose and New Nanocomposite Materials: A Review. Cellulose. 2010;17(3):459–494. doi: 10.1007/s10570-010-9405-y. [DOI] [Google Scholar]
  6. George J. Sabapathi S. N. Cellulose Nanocrystals: Synthesis, Functional Properties, and Applications. Nanotechnol., Sci. Appl. 2015:45. doi: 10.2147/NSA.S64386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Eichhorn S. J. Cellulose Nanowhiskers: Promising Materials for Advanced Applications. Soft Matter. 2011;7(2):303–315. doi: 10.1039/C0SM00142B. [DOI] [Google Scholar]
  8. Habibi Y. Lucia L. A. Rojas O. J. Cellulose Nanocrystals: Chemistry, Self-Assembly, and Applications. Chem. Rev. 2010;110(6):3479–3500. doi: 10.1021/cr900339w. [DOI] [PubMed] [Google Scholar]
  9. Calabrese V. Muñoz-García J. C. Schmitt J. da Silva M. A. Scott J. L. Angulo J. Khimyak Y. Z. Edler K. J. Understanding Heat Driven Gelation of Anionic Cellulose Nanofibrils: Combining Saturation Transfer Difference (STD) NMR, Small Angle X-Ray Scattering (SAXS) and Rheology. J. Colloid Interface Sci. 2019;535:205–213. doi: 10.1016/j.jcis.2018.09.085. [DOI] [PubMed] [Google Scholar]
  10. Ma H. Burger C. Hsiao B. S. Chu B. Fabrication and Characterization of Cellulose Nanofiber Based Thin-Film Nanofibrous Composite Membranes. J. Membr. Sci. 2014;454:272–282. doi: 10.1016/j.memsci.2013.11.055. [DOI] [Google Scholar]
  11. Hestrin S. Schramm M. Synthesis of Cellulose by Acetobacter Xylinum. 2. Preparation of Freeze-Dried Cells Capable of Polymerizing Glucose to Cellulose. Biochem. J. 1954;58(2):345–352. doi: 10.1042/bj0580345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Kono H. Sogame Y. Purevdorj U.-E. Ogata M. Tajima K. Bacterial Cellulose Nanofibers Modified with Quaternary Ammonium Salts for Antimicrobial Applications. ACS Appl. Nano Mater. 2023;6(6):4854–4863. doi: 10.1021/acsanm.3c00616. [DOI] [Google Scholar]
  13. Saito T. Nishiyama Y. Putaux J.-L. Vignon M. Isogai A. Homogeneous Suspensions of Individualized Microfibrils from TEMPO-Catalyzed Oxidation of Native Cellulose. Biomacromolecules. 2006;7(6):1687–1691. doi: 10.1021/bm060154s. [DOI] [PubMed] [Google Scholar]
  14. Xu Y. Atrens A. Stokes J. R. A Review of Nanocrystalline Cellulose Suspensions: Rheology, Liquid Crystal Ordering and Colloidal Phase Behaviour. Adv. Colloid Interface Sci. 2020;275:102076. doi: 10.1016/j.cis.2019.102076. [DOI] [PubMed] [Google Scholar]
  15. Isogai A. Saito T. Fukuzumi H. TEMPO-Oxidized Cellulose Nanofibers. Nanoscale. 2011;3(1):71–85. doi: 10.1039/C0NR00583E. [DOI] [PubMed] [Google Scholar]
  16. Hossain K. M. Z. Calabrese V. da Silva M. A. Schmitt J. Bryant S. J. Islam M. T. Felfel R. M. Scott J. L. Edler K. J. Microstructural, Thermal, Crystallization, and Water Absorption Properties of Films Prepared from Never-Dried and Freeze-Dried Cellulose Nanocrystals. Macromol. Mater. Eng. 2021;306(1):2000462. doi: 10.1002/mame.202000462. [DOI] [Google Scholar]
  17. Mao Y. Liu K. Zhan C. Geng L. Chu B. Hsiao B. S. Characterization of Nanocellulose Using Small-Angle Neutron, X-Ray, and Dynamic Light Scattering Techniques. J. Phys. Chem. B. 2017;121(6):1340–1351. doi: 10.1021/acs.jpcb.6b11425. [DOI] [PubMed] [Google Scholar]
  18. Silvério H. A. Flauzino Neto W. P. Dantas N. O. Pasquini D. Extraction and Characterization of Cellulose Nanocrystals from Corncob for Application as Reinforcing Agent in Nanocomposites. Ind. Crops Prod. 2013;44:427–436. doi: 10.1016/j.indcrop.2012.10.014. [DOI] [Google Scholar]
  19. Carneiro Pessan C. Silva Bernardes J. Bettini S. H. P. Leite E. R. Oxidized Cellulose Nanofibers from Sugarcane Bagasse Obtained by Microfluidization: Morphology and Rheological Behavior. Carbohydr. Polym. 2023;304:120505. doi: 10.1016/j.carbpol.2022.120505. [DOI] [PubMed] [Google Scholar]
  20. Mamo K. A. Andualem T. L. Reddy Prasad D. M. Narayanan M. M. Bitew N. A. Debase E. A. Sustainable Production of Cellulose Nanocrystals from Sugarcane Bagasse via Statistically Optimized Acid Hydrolysis. Sci. Rep. 2026;16(1):10682. doi: 10.1038/s41598-026-46269-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Dunlop M. J. Clemons C. Reiner R. Sabo R. Agarwal U. P. Bissessur R. Sojoudiasli H. Carreau P. J. Acharya B. Towards the Scalable Isolation of Cellulose Nanocrystals from Tunicates. Sci. Rep. 2020;10(1):19090. doi: 10.1038/s41598-020-76144-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Seddiqi H. Oliaei E. Honarkar H. Jin J. Geonzon L. C. Bacabac R. G. Klein-Nulend J. Cellulose and Its Derivatives: Towards Biomedical Applications. Cellulose. 2021;28(4):1893–1931. doi: 10.1007/s10570-020-03674-w. [DOI] [Google Scholar]
  23. Pratama A. W. Piluharto B. Mahardika M. Widiastuti N. Firmanda A. Norrrahim M. N. F. Comparative Study of Oxidized Cellulose Nanofibrils Properties from Diverse Sources via TEMPO-Mediated Oxidation. Case Stud. Chem. Environ. Eng. 2024;10:100823. doi: 10.1016/j.cscee.2024.100823. [DOI] [Google Scholar]
  24. Wu Q. Meng Y. Wang S. Li Y. Fu S. Ma L. Harper D. Rheological Behavior of Cellulose Nanocrystal Suspension: Influence of Concentration and Aspect Ratio. J. Appl. Polym. Sci. 2014;131(15):40525. doi: 10.1002/app.40525. [DOI] [Google Scholar]
  25. Fall A. B. Lindström S. B. Sundman O. Ödberg L. Wågberg L. Colloidal Stability of Aqueous Nanofibrillated Cellulose Dispersions. Langmuir. 2011;27(18):11332–11338. doi: 10.1021/la201947x. [DOI] [PubMed] [Google Scholar]
  26. Fukuzumi H. Tanaka R. Saito T. Isogai A. Dispersion Stability and Aggregation Behavior of TEMPO-Oxidized Cellulose Nanofibrils in Water as a Function of Salt Addition. Cellulose. 2014;21(3):1553–1559. doi: 10.1007/s10570-014-0180-z. [DOI] [Google Scholar]
  27. Quennouz N. Hashmi S. M. Choi H. S. Kim J. W. Osuji C. O. Rheology of Cellulose Nanofibrils in the Presence of Surfactants. Soft Matter. 2016;12(1):157–164. doi: 10.1039/C5SM01803J. [DOI] [PubMed] [Google Scholar]
  28. da Silva M. A. Calabrese V. Schmitt J. Hossain K. M. Z. Bryant S. J. Mahmoudi N. Scott J. L. Edler K. J. Impact of Wormlike Micelles on Nano and Macroscopic Structure of TEMPO-Oxidized Cellulose Nanofibril Hydrogels. Soft Matter. 2020;16(20):4887–4896. doi: 10.1039/D0SM00135J. [DOI] [PubMed] [Google Scholar]
  29. da Silva M. A. Calabrese V. Schmitt J. Celebi D. Scott J. L. Edler K. J. Alcohol Induced Gelation of TEMPO-Oxidized Cellulose Nanofibril Dispersions. Soft Matter. 2018;14(45):9243–9249. doi: 10.1039/C8SM01815D. [DOI] [PubMed] [Google Scholar]
  30. Schmitt J. Calabrese V. da Silva M. A. Lindhoud S. Alfredsson V. Scott J. L. Edler K. J. TEMPO-Oxidised Cellulose Nanofibrils; Probing the Mechanisms of Gelation via Small Angle X-Ray Scattering. Phys. Chem. Chem. Phys. 2018;20(23):16012–16020. doi: 10.1039/C8CP00355F. [DOI] [PubMed] [Google Scholar]
  31. Schmitt J. Calabrese V. da Silva M. A. Hossain K. M. Li P. Mahmoudi N. Dalgliesh R. M. Washington A. Scott J. L. Edler K. J. Surfactant Induced Gelation of TEMPO-Oxidised Cellulose Nanofibril Dispersions Probed Using Small Angle Neutron Scattering. J. Chem. Phys. 2022;5:0129276. doi: 10.1063/5.0129276. [DOI] [PubMed] [Google Scholar]
  32. Dufresne A., Nanocellulose: From Nature to High Performance Tailored Materials, De Gruyter, 2012 10.1515/9783110254600 [DOI] [Google Scholar]
  33. Zaman M. Xiao H. Chibante F. Ni Y. Synthesis and Characterization of Cationically Modified Nanocrystalline Cellulose. Carbohydr. Polym. 2012;89(1):163–170. doi: 10.1016/j.carbpol.2012.02.066. [DOI] [PubMed] [Google Scholar]
  34. Calabrese V. da Silva M. A. Schmitt J. Muñoz-Garcia J. C. Gabrielli V. Scott J. L. Angulo J. Khimyak Y. Z. Edler K. J. Surfactant Controlled Zwitterionic Cellulose Nanofibril Dispersions. Soft Matter. 2018;14(38):7793–7800. doi: 10.1039/C8SM00752G. [DOI] [PubMed] [Google Scholar]
  35. Schweizer K. S. and Curro J. G., PRISM Theory of the Structure, Thermodynamics, and Phase Transitions of Polymer Liquids and Alloys, in Atomistic Modeling of Physical Properties, ed. Monnerie, L. and Suter, U. W., Advances in Polymer Science, Springer Berlin Heidelberg, 1994, vol. 116, pp. 319–377 10.1007/BFb0080203 [DOI] [Google Scholar]
  36. Cannavacciuolo L. Pedersen J. S. Schurtenberger P. Monte Carlo Simulation Study of Concentration Effects and Scattering Functions for Polyelectrolyte Wormlike Micelles. Langmuir. 2002;18(7):2922–2932. doi: 10.1021/la010884f. [DOI] [Google Scholar]
  37. Liang W. Tadros Th. F. Luckham P. F. Effect of Volume Fraction and Particle Size on Depletion Flocculation of a Sterically Stabilized Latex Dispersion Induced by Addition of Poly(Ethylene Oxide) J. Colloid Interface Sci. 1993;155(1):156–164. doi: 10.1006/jcis.1993.1021. [DOI] [Google Scholar]
  38. Pereira A. Duarte H. Nosrati P. Gubitosi M. Gentile L. Romano A. Medronho B. Olsson U. Cellulose Gelation in NaOH Solutions Is Due to Cellulose Crystallization. Cellulose. 2018;25(6):3205–3210. doi: 10.1007/s10570-018-1794-3. [DOI] [Google Scholar]
  39. Brattelli A. Gentile L. Adsorbent Semi-Transparent Cellulose-Based Self-Standing Thin Films. Carbohydr. Polym. 2025;359:123584. doi: 10.1016/j.carbpol.2025.123584. [DOI] [PubMed] [Google Scholar]
  40. Kundu D. Banerjee T. Development of Microcrystalline Cellulose Based Hydrogels for the in Vitro Delivery of Cephalexin. Heliyon. 2020;6(1):e03027. doi: 10.1016/j.heliyon.2019.e03027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Enoch K. Somasundaram A. A. Rheological Insights on Carboxymethyl Cellulose Hydrogels. Int. J. Biol. Macromol. 2023;253:127481. doi: 10.1016/j.ijbiomac.2023.127481. [DOI] [PubMed] [Google Scholar]
  42. Kono H. Characterization and Properties of Carboxymethyl Cellulose Hydrogels Crosslinked by Polyethylene Glycol. Carbohydr. Polym. 2014;106:84–93. doi: 10.1016/j.carbpol.2014.02.020. [DOI] [PubMed] [Google Scholar]
  43. Panahi-Sarmad M. Alikarami N. Guo T. Haji M. Jiang F. Rojas O. J. Aerogels Based on Bacterial Nanocellulose and Their Applications. Small. 2024;20(44):2403583. doi: 10.1002/smll.202403583. [DOI] [PubMed] [Google Scholar]
  44. Razavi M. S. Nematollahzadeh A. Carullo D. Ghaani M. Farris S. Bacterial Nanocellulose-Stabilized Pickering Emulsions: A Review on Stability Factors and Industrial Applications in the Food Sector. Food Res. Int. 2026;232:118875. doi: 10.1016/j.foodres.2026.118875. [DOI] [PubMed] [Google Scholar]
  45. Courtenay J. C. Jin Y. Schmitt J. Hossain K. M. Z. Mahmoudi N. Edler K. J. Scott J. L. Salt-Responsive Pickering Emulsions Stabilized by Functionalized Cellulose Nanofibrils. Langmuir. 2021;37(23):6864–6873. doi: 10.1021/acs.langmuir.0c03306. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

CP-028-D6CP02653B-s001
CP-028-D6CP02653B-s001.pdf (708.2KB, pdf)

Data Availability Statement

Data for this article, including SAXS data files and fits and rheology (amplitude and frequency sweeps and viscosity upon shear) are available at Science Data Bank at https://doi.org/10.57760/sciencedb.42823.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d6cp02653b.

Articles from Physical Chemistry Chemical Physics are provided here courtesy of Royal Society of Chemistry

RESOURCES