You are on page 1of 10

JCIS Open 3 (2021) 100014

Contents lists available at ScienceDirect

JCIS Open
journal homepage: www.journals.elsevier.com/jcis-open

Versatile nanocellulose-anatase TiO2 hybrid nanoparticles in Pickering


emulsions for the photocatalytic degradation of organic and aqueous dyes
Hugo Voisin a, Xavier Falourd a, b, Camille Rivard c, Isabelle Capron a, *
a
INRAE, BIA UR1268, F-44316 Nantes, France
b
INRAE, BIBS facility, F-44316 Nantes, France
c
Synchrotron SOLEIL, L’Orme des Merisiers, Saint-Aubin, 91192, Gif-sur-Yvette Cedex, France

A R T I C L E I N F O A B S T R A C T

Keywords: Hypothesis: Degradation of pollutants using photocatalysts is a surface strategy and their release is generally
Nanocellulose undesired in nature. Biobased particles with simultaneous activity in aqueous and organic phase could limit their
TiO2 amount for enhanced efficiency. Using photoactive materials as interfacial stabilizer for the preparation of
Pickering emulsion
Pickering emulsion should allow for its action on both phases.
Photocatalysis
Hydrophobicity
Experiments: We report the preparation of hybrid nanoparticles involving water-dispersible cellulose nanocrystals
(CNC) and anatase TiO2. TiO2 nanoparticles were grown in situ on the CNC surface resulting in hybrids that could
contain up to 40 wt % in TiO2 with a biobased core and interfacial properties. We measured their photocatalytic
efficiency in water on a hydrophilic dye and in organic phase on a hydrophobic dye. Their photocatalytic activity
was also observed on freeze-dried emulsions-templated aerogels.
Results: The hybrids led to improved dispersability of TiO2 NPs in water and photocatalytic efficiency on water-
soluble dye compared to ungrafted TiO2 NPs. When stabilizing the oil/water interface, they also simultaneously
degrade an organic dye. Finally, the emulsion-based aerogels retained the hybrids photocatalytic activity. It re-
sults in novel, essentially biobased photocatalytic nanoparticles, highly versatile and able to degrade model dyes
in aqueous suspensions, emulsions and aerogels.

1. Introduction hydrophilic [11], making it suitable for the degradation of aqueous


molecules. However, to be fully efficient, it should also be active on
The flexibility of metal oxides makes them ubiquitous in everyday life pollutants soluble in organic solvent. A strategy to answer such an issue is
and the subject of ever-growing innovations. The development of nano- the preparation of Pickering emulsions, in which the TiO2 NPs are fixed
technologies has opened a vast field of experimentation concerning their at the oil-water interface. Such emulsions have displayed encouraging
structure, at a wide range of scales. Amongst them, TiO2 in the anatase results for the degradation of hydrophobic molecules but while examples
form has been the subject of extensive research due to its photocatalytic stabilized by bare TiO2 NPs with naked [12] or dye-covered surface [13]
properties, making it attractive for water splitting [1], photovoltaics [2] have recently been described, they generally imply that the TiO2 surface
and organic molecule degradation [3,4]. Since these applications are needs to be further modified and hydrophobized to stabilize the interface
improved with a high surface contact area, the preparation of anatase [14–17], potentially using non-green chemicals and decreasing their
nanoparticles (NPs) has shown exciting results [5]. However, concern has water dispersability. The development of a versatile and environmentally
been expressed about the possible toxicity of such nanoparticles as they friendly platform for the simultaneous degradation of both water- and
are able to travel further in organs and cells [6–8]. Response strategies oil-soluble pollutants thus remains a challenge.
include fixing the particles by preparation of aerogels with large surface Furthermore, another growing concern of the scientific community is
areas and macroscopic size [9], or the immobilization of small TiO2 NPs the privileged use of green chemistry that uses less toxic and more sus-
on larger supports [10]. An important parameter for the application of tainable materials and reactants. In this regard, polysaccharides in gen-
photocatalytically-active anatase is its surface chemistry since it may eral have been the subject of extensive research due to their abundancy,
modify the type of pollutant it degrades. Indeed, TiO2 is intrinsically plant origin, low toxicity and biodegradability. Cellulose and, in

* Corresponding author.
E-mail address: Isabelle.capron@inrae.fr (I. Capron).

https://doi.org/10.1016/j.jciso.2021.100014
Received 21 February 2021; Received in revised form 5 May 2021; Accepted 25 May 2021
2666-934X/© 2021 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
H. Voisin et al. JCIS Open 3 (2021) 100014

particular, nanocellulose such as cellulose nanofibers (CNF), bacterial heated to reflux (oil bath at 120  C) under stirring for 2 h. One hundred
cellulose (BCN) and cellulose nanocrystals (CNC) have been integrated mL of a 0.1 M NaOH solution were then added for neutralization; the
into various materials and systems in which they have shown extremely suspension was refluxed for one more hour. The suspension was then
attractive properties besides being abundant, sustainable and cooled down in an ice bath and washed three times with deionized water
environmentally-friendly [18,19]. They can be used for the preparation using centrifugation (15000 g, 10 min).
of materials such as films, porous scaffolds, hydrogels [20–22] and Control TiO2 particles were prepared following the same experi-
emulsions [23–25] and have a reactive surface that makes them suitable mental protocol but without any CNCs, using a TiOSO4 concentration of
for the entrapment of pollutants [26] or the nucleation of metal oxides 9 g/L.
[27]. Indeed, numerous studies present examples of the use of nano- Hybrid characterization: TGA measurements were made using a TGA
cellulose as a template for metal oxides such as silica [28,29] zinc oxide 2050 instrument (TA Instruments), in an alumina crucible under air and
[30] and titania [31–35]. In the latter case, materials such as using a 10  C/min temperature ramp. Zeta potential and size measure-
self-cleaning membranes [36], hard films [37] and numerous photo- ments were performed using a Malvern NanoZS instrument. CNC dis-
active hybrids have been described. persions of 0.5-1 g/L at pH ¼ 2, 4 and 12 were prepared at 20  C and
Shandilya et al. previously described a way to graft rutile TiO2 NPs on sonicated.
sulfated CNCs [38]. CNCs are nanorods with a width of several nano- XRD measurements were carried out on dry crushed films obtained by
meters and a length of 100 to 2000 nm, depending on their origin. They drying the slurries. The resulting powder was placed on an adapted
present attractive mechanical properties and a flexible surface chemistry sample holder. The signals were corrected by subtraction of the sample
[39], giving them good dispersability and colloidal stability in various holder background, and the sample height was corrected using the
media. Moreover, they are surface-active since they readily adsorb at and anatase peak as an internal standard and then smoothed using a fast
stabilize air-water or oil-water interfaces and form so called Pickering Fourier treatment.
foams or emulsions. Furthermore, they are becoming increasingly X-ray absorption near edge structure (XANES) spectra were recorded
cheaper to obtain due to the large-scale industrial production of CNCs at the Ti K-edge on SAMBA beamline at synchrotron SOLEIL (Saint-
with homogeneous properties [40]. Rutile TiO2 possesses good Aubin, France). The Si (220) monochromator was calibrated to 4966.3
UV-screening properties that can be enhanced by its distribution on CNCs eV at the first inflection point of a Ti foil XANES spectrum. The dry
as a result of its improved dispersion state in suspension. However, it crushed films obtained by drying the hybrid samples slurries were mixed
presents little to no photocatalytic activity. In order to prepare with cellulose and pressed to obtain circular pellets with diameter of 13
safer-by-design photoactive TiO2 NPs with reduced toxicity while pre- mm with a controlled amount of TiO2 to be in the optimal conditions for
serving their photocatalytic activity, grafting of anatase NPs on CNC transmission mode. Several scan were collected and merged in trans-
presents clear advantages [41]. Liu et al. showed that co-synthesized mission and in continuous scan mode along the 4700 to 5800 eV energy
CNC-anatase hybrids exhibited photocatalytic activity [42]. The benefi- range with 5 eV/s monochromator velocity and 0.04 s/point integration
cial low price of CNC also permits the production of large quantities of time. Scans were normalized and background-subtracted using software
hybrids, which is of obvious interest for applications such as water package Athena [45].
depollution that require large volumes of active particles. With this in Emulsion preparation: suspensions at the adequate concentration of
mind, rather than using expensive commercial anatase TiO2 NPs, we hybrids, TiO2 NPs or native CNCs were prepared and sonicated (30 J/mg
prepared them in-situ using the cheapest precursor of TiO2 available: NP) prior to emulsion. The ionic strength was adjusted to 20 mM of NaCl
TiOSO4, in solid form. This resulted in hybrids where anatase TiO2 NPs to prevent repulsive interactions between NPs at the interface and to
were strongly fixed on the surface of a biobased organic CNC. Such hy- promote the limited coalescence mechanism [23]. 4 mL of suspension
brids could durably stabilize oil-in-water Pickering emulsions. The were then mixed with 1 mL hexadecane, and emulsified using ultra-
improved water dispersability and interfacial properties of the hybrids sonication (50 J/mL emulsion). The emulsions were left to equilibrate for
led to an increase of the anatase photocatalytical properties compared to one night before being characterized. Aerogels were obtained by
the native TiO2 NPs for the degradation of water (Rhodamine B) and freeze-drying emulsions which had been subjected to 2 min of centrifu-
oil-soluble (Nile Red) dyes. Finally, the emulsions could be freeze-dried gation at 4000g. Hydrophobic aerogels were obtained by adding typically
into aerogels which exhibited the same versatility for the photo- 178 μL of MTEOS to 7.821 mL of hybrid suspension, followed by 50 min
degradation of the two aforementioned dyes. The synthesis itself, using of prehydrolysis under stirring at 60  C before emulsifying and
only cheap, non-hazardous chemicals and biobased CNC, is readily freeze-drying following the same protocol that previously described.
up-scalable. Optical microscopy images were obtained by diluting a drop of
emulsion in water on a glass slide, and observed without cover. The
2. Experimental section average diameter of the oil droplets was determined using a Horiba LA-
960 laser light scattering particle size analyzer (Kyoto, Japan).
Chemicals: CNCs were provided by Celluforce, Inc., and were used as The photocatalysis experiments in water were conducted by dissolv-
received. Milli-Q water (R ¼ 18.2 MΩ) was used for the synthesis. ing Rhodamine B (RhB) in hybrid suspension while maintaining a con-
Ethanol was purchased from Merck, and methyltriethoxysilane (MTEOS), stant TiO2 concentration when varying the hybrid sample. 10 mL of the
titanium oxysulfate (TiOSO4), RhB, NR, sodium hydroxide (NaOH) and suspension with RhB were placed in a 5-cm-diameter Petri dish, main-
nitric (HNO3) and hydrofluoric (HF) acids from Sigma-Aldrich. The tained under stirring with a magnetic bar. The pH of the suspension was
chemicals were used as received. maintained at 3 to compare their properties since it is known to have a
Hybrid preparation: the hybrid synthesis was an adaptation of the so- strong impact on aqueous photocatalysis [46]. The suspensions were
called NAC-FAS synthesis (nanometer-sized crystallites from alcoholic then irradiated by a mercury-xenon LC8 Hamamatsu UV-lamp (200 W),
solution) used to prepare anatase NPs in suspension [43] or grafted on and 0.9 mL were sampled every 3 min before being centrifuged. The
graphene [44]. The reaction medium is a 60:40 vol:vol H2O/ethanol optical absorption of the supernatant, free of particles, was then
mixture. Typically, for a 250-mL reaction volume, 1.25 g of CNCs were measured using a Mettler-Toledo UV7 spectrophotometer equipped with
first dispersed in 150 mL H2O to achieve a final concentration of 5 g/L. a 10-mm plastic cell, and the absorbance at 544 nm was monitored to
The CNCs were dispersed through mechanical stirring followed by quantify the degradation of RhB.
ultrasonication (5 kJ/g). One hundred mL of absolute ethanol were then Photocatalysis experiments in the organic phase were conducted with
progressively added to the mixture in a round flask under stirring. A mass emulsions prepared with hexadecane in which Nile Red (NR) was dis-
of TiOSO4 was then added to achieve the desired precursor:CNC ratio, solved at a concentration corresponding to an absorbance at 492 nm of
and the mixture was vigorously stirred for 20 min. The mixture was then approximately 1.6. Ten mL of emulsions were poured into a 5-cm-

2
H. Voisin et al. JCIS Open 3 (2021) 100014

diameter glass Petri dish and stirred using a magnetic bar. A quartz plate
was placed on top of the dish to prevent evaporation while transmitting
the UV. The emulsions were then irradiated as described above, and 1.5
mL were sampled every 5 min, starting at t0. The high stability of the
emulsions insured that they did not phase separate even when little
sample was left and stirred. Sampled emulsions were centrifuged at high
speed (15000 g, 10 min) to break them up and extract the organic phase.
The absorbance at 492 nm of the organic phase was measured with a
Mettler-Toledo UV7 spectrophotometer in a 10-mm quartz cell.
The degradation of dyes deposited on the aerogels was obtain by
irradiating the aerogels using the same UV-lamp during 10 min.
Hybrid films were dissolved for ICP-AES measurements using a
microwave-assisted method. Approximately 40 mg of materials were
added in a PTFE container with 4 mL 68% HNO3 and 1.5 mL 50% HF
placed in a high-pressure digestion bomb. The microwave transparent
bomb was then placed in a microwave oven and exposed to 900 W for 40
s. The resulting solution was cooled down, diluted to a factor of 37.5, and
then measured in triplicate with an ICP-AES ICAP 6300 spectrometer.
Solid-state NMR experiments were carried out on an Avance III-400
MHz spectrometer (Bruker; Champs-sur-Marne, France) operating at
100.62 MHz for 13C, equipped with a double-resonance H/X CP-MAS 4- Fig. 1. TGA profiles of initial CNCs, control TiO2 NPs and Ti20 and Ti32.
mm probe for CP-MAS (Cross-Polarization Magic Angle Spinning) solid-
state experiments. The samples were packed in a 4-mm-diameter rotor
anatase NPs (5-6 [47]), meaning that TiO2 will bear a positive charge at
after rehydration to 11.5 ( 2.8) % w/w.
pH ¼ 3, balancing the negative charge of the CNCs.
The samples were spun at 9000 Hz at room temperature. CP-MAS
Control of the fraction of TiO2 grafted. The inorganic content was
spectra were acquired with a contact time of 2.5 ms and an over-
first measured as the total amount of material remaining above 530  C
accumulation of 5120 scans at intervals of 10 s as a recycling delay.
using TGA, while taking in account the remaining mass of ash from the
The carbonyl carbon was set to 176.03 ppm through external glycine
cellulose (Fig. 1). The result was confirmed using atomic absorption
calibration. NMR spectra deconvolution was developed using the PeakFit
spectrometry. Using precursor:CNC mass ratios from 0.8 to 3, we ob-
(v.4.11) software program (Systat Software, Inc., USA). The crystallinity
tained TiO2 contents of between 20 and 42 wt%, and it is likely that it
index (CI) was calculated as the ratio of the sum of the peak area from 86
could be further increased by using an even higher precursor:CNC ratio.
to 91 ppm (corresponding to the crystalline part of C4) to the sum of the
Among the various samples prepared, we focused our study on the for-
peak area from 80 to 91 ppm (corresponding to all of the C4). The 1H
mulations containing 20 and 32 wt% of TiO2 (Ti20 and Ti32), corre-
spectra were acquired over eight accumulations separated by a recycling
sponding to precursor:CNC ratios of 0.8 and 1.8, respectively. Those two
delay of 5 s.
compositions were chosen because they presented the maximum pho-
Scanning transmission electron microscopy (STEM) imaging: 10 μL of
tocatalytic efficiency per gram of TiO2, as discussed later in this article.
hybrid suspension at 0.5 g/L were deposited onto glow-discharged car-
Their inorganic content was further confirmed in an accurate way
bon-coated grids (200 meshes, Delta Microscopies, France) for 2 min
using ICP-AES, and the resulting amount of TiO2 was found to be in good
before the excess was removed with filter paper. The grids were dried
agreement between the two techniques. The compositions of the other
overnight in air and a 0.5-nm platinum layer was then deposited by an
formulations are detailed in Table S1. While the TiO2 loading seemed to
ion-sputter coater. Images were recorded at 10 kV with a Quattro Scan-
linearly follow the TiOSO4 concentration when at low values, it reached a
ning electron microscope (Thermo Scientific) using a STEM detector.
plateau at around 40 wt%, leading to a loss of efficiency of the synthesis.
Films of CNC and hybrids were prepared from a suspension at 5 g/L of
This may be related to a preferential nucleation on the CNC surface, but
particles, deposited on a glass slide using doctor blade technique. Contact
above a critical concentration, simultaneous nucleation of TiO2 NPs
angles were measured by depositing a 15 μL water droplet on their sur-
face and immediately imaging it using a Digidrop instrument from GBX
Scientific Instruments.

3. Results and discussion

CNC–TiO2 hybrid preparation. Native CNCs are stable in aqueous


suspension because sulfate half ester groups on the surface impart elec-
trostatic repulsions. The initial CNCs were then typically easily redis-
persed in water with sonication. After TiO2 grafting, hybrid washing and
concentration by centrifugation, concentrated white slurries (50–80 g/L)
of hybrid particles were obtained. An immediate effect of the grafting of
TiO2 NPs on CNC was the clear improvement of their colloidal stability in
water. It was easy to redisperse the hybrids without the sonication step.
Furthermore, control TiO2 NPs prepared following the same synthesis
route precipitate in a matter of minutes, whereas the hybrids remain in
suspension for several days (Fig. S2). The electrostatic origin of this
stability was validated by zeta potential measurement since it was be-
tween -20 mV and -35 mV for pH values between 3 and 11. These values
were compared to zeta potential values of native Celluforce CNCs, which Fig. 2. XRD diffractogram of hybrids and TiO2 NPs prepared following the same
are approximately -40 mV for this entire pH range. The higher zeta po- protocol. White arrows indicate cellulose crystalline peaks; black arrows indi-
tential of the hybrids at low pH is related to the isoelectric point of cate the anatase.

3
H. Voisin et al. JCIS Open 3 (2021) 100014

Fig. 3. STEM pictures of native CNC (a), Ti20 (b) and Ti32 (c). The scale bar is 1 μm. Inset on (c) shows Ti32 at a higher magnification.

occurs in the bulk of the suspension. suspensions when the pH is decreased to less than 2. Such an acidification
In addition to the inorganic content of the hybrids, TGA measure- under reflux could have had an influence on the crystallinity of the CNCs,
ments revealed interesting discrepancies in the weight loss profiles of the but the H2SO4 concentration was probably too low for that. Furthermore,
hybrids. For native CNC, major weight loss occurred between 300  C and Shandilya et al. showed that in aqueous conditions, increasing the con-
350  C. Hybrids displayed a first weight loss below 200  C, a temperature centration of H2SO4 would promote TiO2 in the rutile crystalline phase.
much lower than for the CNCs. This loss of about 40% for both samples However, no such influence of the sulfuric acid concentration was
was probably related to a weakening of the cellulose structure in contact observed in NAC-FAS synthesis when increasing the TiOSO4 content.
with the anatase due to the generated radicals. A second degradation Lastly, XANES measurements confirmed the absence of rutile form but
peak occurred at 380  C or 420  C for Ti32 and Ti20, respectively, a also revealed that all the studied samples were not only anatase. The
higher temperature than that of the final degradation of the CNC. Given presence of an additional pre-peak appeared at 4970.3 eV, the attenua-
the relatively high heating rate used in these experiments (10  C/min), tion of the main peak at 4986.7 eV and the comparison with high-
this effect could arise from a retarding action from the TiO2 for the heat pressure modified nano-anatase [49] or amorphous TiO2 spectrum
conduction to the cellulose, as was already reported for clay platelets on a from literature [50,51] suggest the presence of an amorphous or poorly
CNF network [48]. The final plateau value corresponds to the remaining crystalline phase (Fig. S3), not detectable with XRD.
refractory TiO2. The TiO2 thermal resilience was checked using the CNC morphology. A previous study on the functionalization of CNC
control TiO2 NPs described in Material and method part, which only lost by rutile TiO2 highlighted the crucial importance of the distribution of
adsorbed water and the adsorbed surface entities, probably mainly sul- the inorganic particles on the resulting properties [52]. We thus char-
fate species. acterized the distribution of the TiO2 particles on the CNC using STEM
Hybrid crystallinity. We performed XRD measurements to check the (Fig. 3). These images show homogeneously covered, relatively
crystalline structure of the synthesized titanium species (Fig. 2). The well-dispersed hybrids, with no large aggregates of TiO2. This is of prime
peaks at 25, 38 and 48 validate an anatase structure, which is the importance for the photocatalytic efficiency. Since the main photo-
crystalline form of interest for photocatalytic applications. It is inter- catalyst parameter is the area of contact of anatase with targeted mole-
esting to note that the fraction of TiO2 on CNC did not have any influence cules, homogeneous distribution will improve its activity. Furthermore,
on the crystalline pattern of the systems, except for relative intensities the increase of anatase TiO2 content in the hybrid (Ti32 compared to
between cellulose peaks (at 14 and 22 ) and anatase peaks related to the Ti20) did not lead to the formation of larger TiO2 structures but to a more
larger anatase content of Ti32 compared to Ti20. TiOSO4 precursor continuous coating of the CNCs by the nanoparticles.
contains sulfuric acid as free acid, which is released during its dissolution TiO2 grafting interaction. We used FTIR and solid-state 13C NMR
during heating. This leads to a visible increase of viscosity of the sus- spectroscopies to probe the interaction between organic and inorganic
pension at the start of the reaction, a well-known phenomenon for CNC counterparts. FTIR did not show specific interaction peak variations

Fig. 4. 13C NMR spectra of native CNCs, control CNCs that have been exposed to reaction conditions similar to those of the hybrids but without TiOSO4, and the
hybrids Ti20 and Ti32 (Cr ¼ crystalline, Am ¼ amorphous). The CI is indicated under each spectrum.

4
H. Voisin et al. JCIS Open 3 (2021) 100014

properties of the hybrids by quantifying the rate of degradation of an


aqueous dye, Rhodamine 6B (RhB). A suspension of a hybrid solution
with RhB was exposed to UV light and the discoloration was measured
through light absorption at 492 nm. The rates of degradation of RhB in
the presence of the Ti20 and Ti32 hybrids were compared to CNCs and
control TiO2 NPs (Fig. 5). For comparison of photocatalytic perfor-
mances, the concentration of TiO2 was normalized in all the tests (except
for CNC). Their behavior could be modeled following a pseudo first-order
model (see SI for further details). It is visible that in the absence of TiO2
(curve for CNC alone), the rate of degradation of RhB was extremely slow
and no photocatalytic activity is observed. Furthermore, the rate of RhB
degradation was much higher for the hybrid than for the TiO2 NPs. This is
certainly due to the much better dispersion state of the hybrid compared
to the naked particles, as highlighted by the very fast rate of decantation
of those compared to the hybrids. Among the hybrids, the Ti40 formu-
lation showed poorer performances despite its high TiO2 content (40 wt
%). Ti32 displayed a slightly better photocatalytic behavior compared to
Ti20. This tendency was also observed for higher TiO2 loading (see
Table S1). This result highlighted the existence of adapted precursor
concentration during synthesis to obtain optimal photocatalytic proper-
ties related to the TiO2 structure and contact surface, and to the colloidal
Fig. 5. Evolution of the absorbance of RhB solution over time in the presence of stability of the hybrids. This efficiency is thus ultimately not linearly
various NPs and under UV light irradiation. The hybrid concentrations were
related to the inorganic loading. It furthermore shows a much higher
chosen in order to characterize all suspensions at a TiO2 concentration of 0.25 g/
efficiency of the hybrid containing 80% of biobased product and 20% of
L TiO2 and pH ¼ 3. Lines are a guide for the eyes.
TiO2 than pure TiO2 NPs.
Properties as interfacial stabilizer. The interfacial properties of
(Fig. S4). Solid-state NMR showed that the crystalline structure of cel-
native CNCs have been extensively studied, especially for the stabiliza-
lulose was preserved and revealed similarity with the control CNCs that
tion of oil-in-water emulsions forming the so-called Pickering emulsions
were subjected to the same reaction process as the hybrids but without
[23–25,55,56]. It was furthermore demonstrated that their interfacial
TiOSO4 (Fig. 4). The 60-68 ppm region was assigned to the hydrox-
properties were preserved when functionalized with TiO2 [52]. The ca-
ymethyl C6 carbons and split into two regions: 60-63.5 ppm are the
pacity of the synthesized hybrids to stabilize emulsions is very interesting
disordered chains mainly at the surface of the CNCs, and 63.5-68 ppm are
since it might allow them to photocatalytically degrade pollutants not
the chains in the crystalline core [53]. In particular, the C6 peaks around
only in the aqueous phase but also in the organic phase. We thus
60-63.5 ppm are virtually unchanged, whereas some authors previously
compared emulsions prepared with native CNC with the two hybrids
described covalent bonding between TiO2 and cellulose through esteri-
including 20 and 32% of TiO2 (Ti20 and Ti32 respectively).
fication, changing these peak intensities [54]. This absence of difference
Since we chose a 20:80 oil:water ratio and an oil lighter than water,
could arise from the small amount of modified bonds or, more likely, the
the oil droplets tended to cream, concentrating at the surface. After
fact that the interaction in question is not covalent but mainly electro-
decantation, both hybrid-based emulsions presented a volume of emul-
static. This assumption was further confirmed by the observation that the
sion significantly larger than the emulsions prepared with native CNC
hybrids separated into CNCs and TiO2 NPs at high pH values (>12),
(Fig. S5). This volume increase corresponded to a larger fraction of water
where both cellulose and TiO2 surfaces are negatively charged and
retained in the emulsions stabilized by the hybrid. It is possible that the
strongly repulse each other. At lower pH values, the interaction is
higher roughness of the hybrids (Fig. 3) translated into a different film
resistant enough to hold the composite together even under sonication
structure. It could thus present a larger fractal dimension with an
(several minutes at 30-40 W), highlighting the high strength of the
increased ability to trap water. Investigating this aspect, 1H NMR showed
bonds. The hybrids were thus expected to hold under various conditions
an increase of the peak full width at half maximum (FWHM) after
as long as the pH remained below 12. Furthermore, the relative quanti-
hybridation with TiO2. FWHM depends on the T2* value [57] through the
fication of the crystalline and amorphous parts of CNCs for the different
relationship: FWHM ¼ π1T * . The higher the T2* value is, the higher the
samples (Fig. 4) revealed no clear change of the CNC crystallinity 2

measured at 62% before modification and found between 59% and 56% water mobility and the lower the FWHM will be. Thus, the increase of
for all of the samples after reaction. FWHM by an order of magnitude observed for both hybrid NPs (Fig. S6)
Photocatalytic activity. We characterized the photocatalytic expressed a lower mobility of water, perhaps related to the porous

Fig. 6. Effect of hybrid concentration on drop diameters visualized by optical microscopy for Ti32-based emulsions at 0.25, 1.25 and 5 g/L of hybrids. The scale bar is
25 μm.

5
H. Voisin et al. JCIS Open 3 (2021) 100014

interstitial film that retains less water [24,59]. The final droplet diameter
is thus only a function of the concentration of Pickering particles as well
as their properties (Fig. 6).
After emulsification, the Pickering emulsions are known to follow a
limited coalescence process [60] that produces monodisperse droplets
that can remain stable for months and resist concentration and dilution
[23,24]. At low particle content, the droplets coalesce until their surface
is sufficiently covered to stop the phenomenon (Fig. 7). Above a certain
threshold, the droplet diameter remains almost constant, whereas the
interface density increases as a result of the increasing CNC concentra-
tion. The hybrids presented the same droplet diameter profile as native
CNC, reflecting similarities in the aspect ratio and interfacial properties
of the particles. There is, however, a difference of mean average diameter
for a given particle concentration between native CNC, Ti20 and Ti32
that was confirmed by several repetitions (error bars are most of the time
smaller than the symbols). In a similar way as for the difference in
emulsion volume, the larger diameter of the CNC-stabilized droplets
arose from the strong cellulose-cellulose interaction between smooth
CNC surfaces, decreasing their ability to spread evenly at the interface
and efficiently stabilizing it. The TiO2 decreased this type of interaction,
allowing for a better coverage distribution and decreased droplet diam-
eter. Considering the lower droplet diameter of Ti32 compared to Ti20,
this may be related to a difference in the aggregation state in those
Fig. 7. Evolution of average droplet diameter of the emulsions at various par- conditions, as has been demonstrated for CNCs with different surface
ticle loadings.
charges [59] where desulfated CNCs stabilize larger droplets than
sulfated ones [24]. This is highlighted when plotting 1/D vs. the con-
structures that entrapped water. Indeed, the correlation between T2 centration (see inset in Fig. 7). The differences in the slopes correspond to
relaxation time and porous structure in cellulose was shown [58]. Con- a difference in the density or thickness of the interface: a steeper slope
tact angle measurements correlated this increase of roughness of the was linked to a more efficient coverage of the interface by the particles,
particles aggregates: the angle formed by a water droplet deposited on a which also correlated well with the intersection of the two lines repre-
film prepared from a CNC suspension is 24  3 , while it increases to 34 senting the limit between the two regimes of surface coverage. A higher
 4 on a film formed by Ti20 or Ti32 (Fig. S7). This difference likely concentration value to reach this limit implied that more particles were
arises from a rougher structure of the hybrid films compared to the pure required to stabilize the droplets before increasing the interface density
cellulose ones, rather than a difference of intrinsic hydrophilicity of CNC by aligning the CNC rods.
or TiO2. Between the two hybrids tested, Ti32 presented a lower emul- The dotted lines correspond to the concentrations chosen for Fig. 8.
sion volume than Ti20 (Fig. S5) and, consequently, a smaller fraction of The inset represents the evolution of the inverse of the droplet diameter
water retained. This may not be explained by a larger repulsion between with particle loadings. Lines are a guide for the eyes.
the droplets in the Ti20 case, since the effect should be balanced by the Photocatalytic properties of emulsions. We studied the photo-
presence of salt in solution. The difference in emulsion volume could catalytic properties of CNC–TiO2 hybrids in emulsion systems by dis-
then be related to the fact that the Ti32 particles are more aggregated solving a lipophilic dye, Nile Red (NR), in hexadecane prior to
than Ti20 (already visible in suspension) which could lead to a denser emulsification. The change of color observed from orange to pink varying

Fig. 8. a) Visual aspect of emulsions from hexadecane with NR at various NP contents. The inset shows a Ti32-based emulsion at 2.5 g/L after mild centrifugation. b)
Evolution of the absorbance of hexadecane dyed with NR dispersed in emulsions stabilized with different particles at various loading contents (g NPs/L aqueous
suspension), exposed to UV irradiation. Lines are a guide to the eyes.

6
H. Voisin et al. JCIS Open 3 (2021) 100014

Fig. 9. Schematic of a droplet of the Pickering emulsion, with the associated radical generation routes as well as the studied dyes.

with the particles concentration relates to the change of droplet diameter, the dynamics of diffusion and homogenization during the experiment,
which is known to influence the reflected light of a colored emulsion was sufficient to balance any acceleration of degradation due to a higher
[61]. Some solvatochromic from the dye molecules at the oil-water density of the interface. Furthermore, it is interesting to note that a
interface may also play a role in this color change. The photocatalytic higher loading of TiO2 led to a higher intrinsic efficiency of the hybrid for
activity was determined measuring in time the degradation of NR in the organic photocatalysis: Ti20 at 1.25 g/L presented slower photo-
oil phase. This activity was expected to occur at the oil-water interface degradation than Ti32 at 0.375 g/L, despite presenting both a higher
where the radical-generating TiO2 particles are located. We performed absolute concentration of TiO2 and a larger area of contact. This
the photocatalysis experiments on emulsions prepared at three different discrepancy could arise from differences in surface structure of the TiO2
concentrations of particles in the aqueous phase: 0.375, 1.25 and 2.5 g/L crystallites, as well as a difference in crystalline structure not detectable
(Fig. 8b). The control TiO2 NPs in aqueous suspension mixed with oil in XRD, implying subtle differences in the NP structures. It is interesting
degraded the dye significantly slower than the hybrids (k ¼ 0.017 min-1 to note that this difference only manifested itself in the organic photo-
for the control NPs and k ¼ 0.140 min-1 for Ti20 at 0.375 g/L, while the catalysis, but not in the aqueous phase.
latter system contain 20 times more TiO2 by weight than the former). The The positioning of the hybrids at the interface thus allowed for the
difference was directly related to the much lower area of contact between photodegradation of both NR and RhB as the radicals generated at the
the aqueous and the organic phases since native TiO2 NPs did not sta- TiO2 NPs surface were able to react with both dyes (Fig. 9). Furthermore,
bilize emulsions but produced macroscopic phase separation. Further- the CNC alone are responsible for the stabilization of the oil-water
more, the adsorbing properties of nanocellulose for various pollutants is interface, leaving the TiO2 surface intact for catalysis as well as impart-
well known [26], which may have contributed to the photocatalytic ac- ing persistent interfacial properties.
tivity of the hybrids by trapping the dye molecules close to the photo- Photocatalytic aerogels. The impressive stability of the CNC-
active centers. stabilized emulsions opens the way to other architectures. Notably,
Those hybrid concentration values were chosen to correspond to Pickering emulsions can be freeze-dried into aerogels for which the
three different regions on the diameter curve (Fig. 7) and surface porosity is controlled by the drop dimensions forming materials with
coverage. Comparing CNC to Ti20 and Ti32, the concentration of TiO2 in multi-scale properties [62,63] Such aerogels can thus be prepared by
the emulsions was thus not identical between the hybrids since their freeze-drying centrifuged Ti20 and Ti32-based emulsions. The very low
inorganic contents are different. It was thus expected that Ti32 would amount of water after centrifugation limits the formation of ice crystals,
display a higher photocatalytic performance than Ti20 for identical NP and the resulting materials present very high porosity (Fig. 10a). The
concentrations. The increase of particle concentration from 0.375 to 2.5 aerogels wetted by organic NR-colored hexadecane (Fig. 10b) displayed
g/L increased the rate of degradation of NRs since the oil/water interface immediate and strong solvatochromism with a color change from orange
was expanded following the decrease of the average droplet diameter to purple due to the contact between the dye and the hydration layer on
(19.4 μm to 4.6 μm and 9.9 μm to 4.7 μm for Ti32 and Ti20, respectively). the aerogel surface. The hexadecane was degraded during UV exposure
However, when the concentration was further increased to 10 g/L, we (Fig. 10c) in a similar fashion as previously observed in the emulsions,
observed no improvement of the degradation rate for Ti20 and even a demonstrating the preservation of the hybrids photocatalytic properties
slowdown for Ti32. The interface area size did not increase between in the aerogel form. When the aerogels were partially hydrophobized
those two concentrations since the average droplet diameter is kept MTEOS, no solvatochromism was observed with the NR-hexadecane. The
almost constant. Moreover, a larger concentration translated into a oil phase was not interacting closely with the CNC–TiO2 hydrophilic
higher emulsion viscosity, especially in the case of Ti32, which became surface preventing the change of energy gap hence of color of the NR
clearly viscoelastic at this concentration. This effect, which slowed down (Fig. 10d). However, the aerogels photocatalytic properties were

7
H. Voisin et al. JCIS Open 3 (2021) 100014

Fig. 10. Close-up on the surface of an aerogel ob-


tained by freeze-drying of centrifuged emulsions,
showing its porosity – scale bar is 2.5 mm (a); on (b)
a drop of NR-died hexadecane was placed on an
aerogel; (c) is the same aerogel after UV exposure,
showing the discoloration of the hexadecane stains
and the yellowing of the aerogel due to the photo-
catalytic degradation of the cellulose network; (d)
hydrophobically modified aerogel with a drop of
NR-died (orange) and a drop of RhB solution (pink);
(e) the two same aerogels after UV exposure.

preserved as both aqueous RhB and organic NR solutions would fade well distributed on the surface of the particles. Consequently, the radicals
when exposed to UV light on the surface of the aerogel (Fig. 10e). formed upon illumination of the semiconductor could react with dyes
present in both water and alkane. Furthermore, they considerably
4. Conclusion decrease the environmental impact since their core is biobased, forming
NPs composed of 80% cellulose. Ultimately, the hybrid particles are more
We prepared CNC–TiO2 hybrids where crystal nucleation was ori- efficient than similarly synthetized TiO2 NPs and at the end of life of the
ented in order to graft TiO2 in its photocatalytic anatase form with material, mainly biobased material is released after use. This straight-
various inorganic contents following a safer-by-design approach. The forward synthesis gives then sustainable, low cost and versatile particles
prepared anatase NPs were grafted on the CNC surface in a homogeneous for photocatalytic materials such as emulsions, foams, aerogels or com-
fashion. The resulting hybrid presented a clear improvement in photo- posites. This flexible grafting strategy on a biobased substrate with
catalytic properties in suspension compared to the as-prepared control interfacial properties can easily be extended to other active inorganic
anatase TiO2 due to a better dispersability and colloidal stability. The particles to produce new types of multifunctional materials.
interfacial properties of CNC were also maintained allowing for degra-
dation of organic dye in Pickering emulsions. Differences of cellulose
surface interaction with water were revealed by color changes when the Declaration of competing interest
anatase content was changed. The emulsions could be freeze-dried into
versatile aerogels efficiently degrading both aqueous and organic dyes. The authors declare that they have no known competing financial
The photocatalytic process considering photon adsorption is a process interests or personal relationships that could have appeared to influence
occurring on the surface of the NPs. In the present study, all the TiO2 is the work reported in this paper.

8
H. Voisin et al. JCIS Open 3 (2021) 100014

Acknowledgments [17] M.F. Nsib, A. Maayoufi, N. Moussa, N. Tarhouni, A. Massouri, A. Houas,


Y. Chevalier, TiO2 modified by salicylic acid as a photocatalyst for the degradation
of monochlorobenzene via Pickering emulsion way, J. Photochem. Photobiol.
This work is a contribution to the Labex SERENADE (n ANR-11- Chem. 251 (2013) 10–17, https://doi.org/10.1016/j.jphotochem.2012.10.007.
LABX-0064) funded by the “Investissement d’Avenir” French Govern- [18] B. Thomas, M.C. Raj, B.K. Athira, H.M. Rubiyah, J. Joy, A. Moores, G.L. Drisko,
ment program of the French National Research Agency (ANR) through C. Sanchez, Nanocellulose, a versatile green platform: from biosources to materials
and their applications, Chem. Rev. 118 (2018) 11575–11625, https://doi.org/
the A*MIDEX project (n ANR-11-IDEX-0001-02). Solid state NMR ex- 10.1021/acs.chemrev.7b00627.
periments and STEM images were performed at the BIBS facility [19] S.J. Eichhorn, A. Dufresne, M. Aranguren, N.E. Marcovich, J.R. Capadona,
(UR1268 BIA, IBiSA, Phenome-Emphasis-FR (grant number ANR-11- S.J. Rowan, C. Weder, W. Thielemans, M. Roman, S. Renneckar, W. Gindl, S. Veigel,
J. Keckes, H. Yano, K. Abe, M. Nogi, A.N. Nakagaito, A. Mangalam, J. Simonsen,
INBS-0012)). We acknowledge SOLEIL for provision of synchrotron ra- A.S. Benight, A. Bismarck, L.A. Berglund, T. Peijs, Review: Current International
diation facilities and we would like to thank Gautier Landrot for his help Research into Cellulose Nanofibres and Nanocomposites, 2010, https://doi.org/
during data acquisition. We are also very grateful to Benoit Duchemin for 10.1007/s10853-009-3874-0.
[20] K.J. De France, T. Hoare, E.D. Cranston, Review of hydrogels and aerogels
the XRD experiments (CASMOC Universite du Havre – Fr) and Bruno containing nanocellulose, Chem. Mater. 29 (2017) 4609–4631, https://doi.org/
Novales (BIBS plateform BIA-INRAE) for the STEM pictures. 10.1021/acs.chemmater.7b00531.
[21] N.E. Zander, H. Dong, J. Steele, J.T. Grant, Metal cation cross-linked nanocellulose
hydrogels as tissue engineering substrates, ACS Appl. Mater. Interfaces 6 (2014)
Appendix A. Supplementary data 18502–18510, https://doi.org/10.1021/am506007z.
[22] K. Syverud, S.R. Pettersen, K. Draget, G. Chinga-Carrasco, Controlling the elastic
Supplementary data to this article can be found online at https:// modulus of cellulose nanofibril hydrogels???scaffolds with potential in tissue
engineering, Cellulose 22 (2015) 473–481, https://doi.org/10.1007/s10570-014-
doi.org/10.1016/j.jciso.2021.100014.
0470-5.
[23] I. Kalashnikova, H. Bizot, B. Cathala, I. Capron, Modulation of cellulose
References nanocrystals amphiphilic properties to stabilize oil/water interface,
Biomacromolecules 13 (2012) 267–275, https://doi.org/10.1021/bm201599j.
[1] M. Ni, M.K.H. Leung, D.Y.C. Leung, K. Sumathy, A review and recent developments [24] F. Cherhal, F. Cousin, I. Capron, Structural description of the interface of pickering
in photocatalytic water-splitting using TiO2 for hydrogen production, Renew. emulsions stabilized by cellulose nanocrystals, Biomacromolecules 17 (2016)
Sustain. Energy Rev. 11 (2007) 401–425, https://doi.org/10.1016/ 496–502, https://doi.org/10.1021/acs.biomac.5b01413.
j.rser.2005.01.009. [25] I. Capron, B. Cathala, Surfactant-free high internal phase emulsions stabilized by
[2] Y. Lan, Y. Lu, Z. Ren, Mini review on photocatalysis of titanium dioxide cellulose nanocrystals, Biomacromolecules 14 (2013) 291–296, https://doi.org/
nanoparticles and their solar applications, Nano Energy 2 (2013) 1031–1045, 10.1021/bm301871k.
https://doi.org/10.1016/j.nanoen.2013.04.002. [26] H. Voisin, L. Bergstr€ om, P. Liu, A.P. Mathew, Nanocellulose-based materials for
[3] K. Pirkanniemi, M. Sillanp€a€a, Heterogeneous water phase catalysis as an water purification, Nanomaterials 7 (2017), https://doi.org/10.3390/
environmental application: a review, Chemosphere 48 (2002) 1047–1060, https:// nano7030057.
doi.org/10.1016/S0045-6535(02)00168-6. [27] M. Kaushik, A. Moores, Review: nanocelluloses as versatile supports for metal
[4] Y. Xia, Q. Li, K. Lv, M. Li, Heterojunction construction between TiO 2 hollowsphere nanoparticles and their applications in catalysis, Green Chem. 18 (2016) 622–637,
and ZnIn 2 S 4 flower for photocatalysis application, Appl. Surf. Sci. 398 (2017) https://doi.org/10.1039/c5gc02500a.
81–88, https://doi.org/10.1016/j.apsusc.2016.12.006. [28] K.E. Shopsowitz, H. Qi, W.Y. Hamad, M.J. MacLachlan, Free-standing mesoporous
[5] S.M. Gupta, M. Tripathi, A review of TiO2 nanoparticles, Chin. Sci. Bull. 56 (2011) silica films with tunable chiral nematic structures, Nature 468 (2010) 422–426,
1639–1657, https://doi.org/10.1007/s11434-011-4476-1. https://doi.org/10.1038/nature09540.
[6] E. Peira, F. Turci, I. Corazzari, D. Chirio, L. Battaglia, B. Fubini, M. Gallarate, The [29] E. Dujardin, M. Blaseby, S. Mann, Synthesis of mesoporous silica by sol-gel
influence of surface charge and photo-reactivity on skin-permeation enhancer mineralisation of cellulose nanorod nematic suspensions, J. Mater. Chem. 13 (2003)
property of nano-TiO2 in ex vivo pig skin model under indoor light, Int. J. Pharm. 696–699, https://doi.org/10.1039/b212689c.
467 (2014) 90–99, https://doi.org/10.1016/j.ijpharm.2014.03.052. [30] K. Lefatshe, C.M. Muiva, L.P. Kebaabetswe, Extraction of nanocellulose and in-situ
[7] I. Fenoglio, J. Ponti, E. Alloa, M. Ghiazza, I. Corazzari, R. Capomaccio, D. Rembges, casting of ZnO/cellulose nanocomposite with enhanced photocatalytic and
S. Oliaro-Bosso, F. Rossi, Singlet oxygen plays a key role in the toxicity and DNA antibacterial activity, Carbohydr. Polym. 164 (2017) 301–308, https://doi.org/
damage caused by nanometric TiO2 in human keratinocytes, Nanoscale 5 (2013) 10.1016/j.carbpol.2017.02.020.
6567–6576, https://doi.org/10.1039/c3nr01191g. [31] S.A. Jabasingh, D. Lalith, M.A. Prabhu, A. Yimam, T. Zewdu, Catalytic conversion of
[8] J. Wu, W. Liu, C. Xue, S. Zhou, F. Lan, L. Bi, H. Xu, X. Yang, F.D. Zeng, Toxicity and sugarcane bagasse to cellulosic ethanol: TiO2 coupled nanocellulose as an effective
penetration of TiO2 nanoparticles in hairless mice and porcine skin after subchronic hydrolysis enhancer, Carbohydr. Polym. 136 (2015) 700–709, https://doi.org/
dermal exposure, Toxicol. Lett. 191 (2009) 1–8, https://doi.org/10.1016/ 10.1016/j.carbpol.2015.09.098.
j.toxlet.2009.05.020. [32] X. Chen, D.H. Kuo, D. Lu, N-doped mesoporous TiO2 nanoparticles synthesized by
[9] G. Dagan, M. Tomkiewicz, TiO2 aerogels for photocatalytic decontamination of using biological renewable nanocrystalline cellulose as template for the degradation
aquatic environments, J. Phys. Chem. 97 (1993) 12651–12655, https://doi.org/ of pollutants under visible and sun light, Chem. Eng. J. 295 (2016) 192–200,
10.1021/j100151a001. https://doi.org/10.1016/j.cej.2016.03.047.
[10] M.J. Bessa, C. Costa, J. Reinosa, C. Pereira, S. Fraga, J. Fernandez, M.A. Ba~ nares, [33] B. Liu, Y. Pan, G. Sun, J. Huang, The preparation and characterization of the Ni-
J.P. Teixeira, Moving into advanced nanomaterials. Toxicity of rutile TiO2 NiO/TiO2 hollow composite materials on micro-nano cellulose fibers, Vacuum 155
nanoparticles immobilized in nanokaolin nanocomposites on HepG2 cell line, (2018) 553–558, https://doi.org/10.1016/j.vacuum.2018.06.044.
Toxicol. Appl. Pharmacol. 316 (2017) 114–122, https://doi.org/10.1016/ [34] Z. Li, C. Yao, Y. Yu, Z. Cai, X. Wang, Highly efficient capillary photoelectrochemical
j.taap.2016.12.018. water splitting using cellulose nanofiber-templated TiO2 photoanodes, Adv. Mater.
[11] D.S. Kommireddy, A.A. Patel, T.G. Shutava, D.K. Mills, Y.M. Lvov, Layer-by-layer 26 (2014) 2262–2267, https://doi.org/10.1002/adma.201303369.
assembly of TiO 2 nanoparticles for stable hydrophilic biocompatible coatings, [35] W. Wang, J. Wang, X. Shi, Z. Yu, Z. Song, L. Dong, G. Jiang, S. Han, Synthesis of
J. Nanosci. Nanotechnol. 5 (2005) 1081–1087, https://doi.org/10.1166/ mesoporous TiO2 induced by nano-cellulose and its photocatalytic properties,
jnn.2005.149. BioResources 11 (2016) 3084–3093, https://doi.org/10.15376/biores.11.2.3084-
[12] N. Fessi, M.F. Nsib, Y. Chevalier, C. Guillard, F. Dappozze, A. Houas, L. Palmisano, 3093.
F. Parrino, Photocatalytic degradation enhancement in pickering emulsions [36] H. Zhan, N. Peng, X. Lei, Y. Huang, D. Li, R. Tao, C. Chang, UV-induced self-
stabilized by solid particles of bare TiO 2, Langmuir 35 (2019) 2129–2136, https:// cleanable TiO2/nanocellulose membrane for selective separation of oil/water
doi.org/10.1021/acs.langmuir.8b03806. emulsion, Carbohydr. Polym. 201 (2018) 464–470, https://doi.org/10.1016/
[13] Q. Li, T. Zhao, M. Li, W. Li, B. Yang, D. Qin, K. Lv, X. Wang, L. Wu, X. Wu, J. Sun, j.carbpol.2018.08.093.
One-step construction of Pickering emulsion via commercial TiO 2 nanoparticles for [37] C. Schütz, J. Sort, Z. Bacsik, V. Oliynyk, E. Pellicer, A. Fall, L. Wågberg, L. Berglund,
photocatalytic dye degradation, Appl. Catal. B Environ. 249 (2019) 1–8, https:// L. Bergstr€om, G. Salazar-Alvarez, Hard and transparent films formed by
doi.org/10.1016/j.apcatb.2019.02.057. nanocellulose-TiO2 nanoparticle hybrids, PloS One 7 (2012) 1–8, https://doi.org/
[14] N. Fessi, M.F. Nsib, Y. Chevalier, C. Guillard, F. Dappozze, A. Houas, L. Palmisano, 10.1371/journal.pone.0045828.
F. Parrino, Pickering emulsions of fluorinated TiO2: a new route for intensification [38] C.E. Zubieta, P.V. Messina, C. Luengo, M. Dennehy, O. Pieroni, P.C. Schulz, Reactive
of photocatalytic degradation of nitrobenzene, Langmuir 36 (2020) 13545–13554, dyes remotion by porous TiO2-chitosan materials, J. Hazard Mater. 152 (2008)
https://doi.org/10.1021/acs.langmuir.0c02285. 765–777, https://doi.org/10.1016/j.jhazmat.2007.07.043.
[15] M. Nawaz, W. Miran, J. Jang, D.S. Lee, Stabilization of Pickering emulsion with [39] Y. Habibi, L.A. Lucia, O.J. Rojas, Cellulose nanocrystals: chemistry, self-assembly,
surface-modified titanium dioxide for enhanced photocatalytic degradation of and applications, Chem. Rev. 110 (2010) 3479–3500, https://doi.org/10.1021/
Direct Red 80, Catal. Today 282 (2017) 38–47, https://doi.org/10.1016/ cr900339w.
j.cattod.2016.02.017. [40] M.S. Reid, M. Villalobos, E.D. Cranston, Benchmarking cellulose nanocrystals: from
[16] T. Burdyny, J. Riordon, C.T. Dinh, E.H. Sargent, D. Sinton, Self-assembled the laboratory to industrial production, Langmuir 33 (2017) 1583–1598, https://
nanoparticle-stabilized photocatalytic reactors, Nanoscale 8 (2016) 2107–2115, doi.org/10.1021/acs.langmuir.6b03765.
https://doi.org/10.1039/c5nr05859g. [41] F. Wesarg, F. Schlott, J. Grabow, H.D. Kurland, N. Heßler, D. Kralisch, F.A. Müller,
In situ synthesis of photocatalytically active hybrids consisting of bacterial

9
H. Voisin et al. JCIS Open 3 (2021) 100014

nanocellulose and anatase nanoparticles, Langmuir 28 (2012) 13518–13525, [53] R.H. Newman, J.A. Hemmingson, Carbon-13 NMR distinction between categories of
https://doi.org/10.1021/la302787z. molecular order and disorder in cellulose, Cellulose 2 (1995) 95–110, https://
[42] S. Liu, D. Tao, H. Bai, X. Liu, Cellulose-nanowhisker-templated synthesis of titanium doi.org/10.1007/BF00816383.
dioxide/cellulose nanomaterials with promising photocatalytic abilities, J. Appl. [54] A. Henry, S. Plumejeau, L. Heux, N. Louvain, L. Monconduit, L. Stievano, B. Boury,
Polym. Sci. 126 (2012) E281–E289, https://doi.org/10.1002/app. Conversion of nanocellulose aerogel into TiO2 and TiO2@C nano-thorns by direct
[43] S. Ito, S. Inoue, H. Kawada, M. Hara, M. Iwasaki, H. Tada, Low-temperature anhydrous mineralization with TiCl4. Evaluation of electrochemical properties in Li
synthesis of nanometer-sized crystalline TiO2particles and their photoinduced batteries, ACS Appl. Mater. Interfaces 7 (2015) 14590–14592, https://doi.org/
decomposition of formic acid, J. Colloid Interface Sci. 216 (1999) 59–64, https:// 10.1021/acsami.5b00299.
doi.org/10.1006/jcis.1999.6275. [55] I. Kalashnikova, H. Bizot, B. Cathala, I. Capron, New pickering emulsions stabilized
[44] T. Soejima, Y. Maru, S. Ito, Facile low-temperature synthesis and photocatalytic by bacterial cellulose nanocrystals, Langmuir 27 (2011) 7471–7479, https://
activity of graphene oxide/TiO2composite, Bull. Chem. Soc. Jpn. 86 (2013) doi.org/10.1021/la200971f.
1065–1070, https://doi.org/10.1246/bcsj.20130093. [56] I. Capron, O.J. Rojas, R. Bordes, Behavior of nanocelluloses at interfaces, Curr.
[45] B. Ravel, M. Newville, ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray Opin. Colloid Interface Sci. 29 (2017) 83–95, https://doi.org/10.1016/
absorption spectroscopy using IFEFFIT, J. Synchrotron Radiat. 12 (2005) 537–541, j.cocis.2017.04.001.
https://doi.org/10.1107/S0909049505012719. [57] J.R. Alger, Quantitative HMRS and spetroscopic imaging of the brain: a didactic
[46] U.G. Akpan, B.H. Hameed, Parameters affecting the photocatalytic degradation of review, Top. Magn. Reson. Imag. 21 (2010) 115–128, https://doi.org/10.1097/
dyes using TiO2-based photocatalysts: a review, J. Hazard Mater. 170 (2009) RMR.0b013e31821e568f.Quantitative.
520–529, https://doi.org/10.1016/j.jhazmat.2009.05.039. [58] C. Zhang, P. Li, Y. Zhang, F. Lu, W. Li, H. Kang, J. feng Xiang, Y. Huang, R. Liu,
[47] M.D. Chadwick, J.W. Goodwin, E.J. Lawson, P.D.A. Mills, B. Vincent, Surface Hierarchical porous structures in cellulose: NMR relaxometry approach, Polymer 98
charge properties of colloidal titanium dioxide in ethylene glycol and water, (2016) 237–243, https://doi.org/10.1016/j.polymer.2016.06.036.
Colloids Surfaces A Physicochem. Eng. Asp. 203 (2002) 229–236, https://doi.org/ [59] F. Cherhal, F. Cousin, I. Capron, Influence of charge density and ionic strength on
10.1016/S0927-7757(01)01101-3. the aggregation process of cellulose nanocrystals in aqueous suspension, as revealed
[48] O. K€oklükaya, F. Carosio, L. Wågberg, Superior flame-resistant cellulose nanofibril by small-angle neutron scattering, Langmuir 31 (2015) 5596–5602, https://
aerogels modified with hybrid layer-by-layer coatings, ACS Appl. Mater. Interfaces doi.org/10.1021/acs.langmuir.5b00851.
9 (2017) 29082–29092, https://doi.org/10.1021/acsami.7b08018. [60] S. Arditty, V. Schmitt, J. Giermanska-Kahn, F. Leal-Calderon, Materials based on
[49] A.M. Flank, P. Lagarde, J.P. Itie, A. Polian, G.R. Hearne, Pressure-induced solid-stabilized emulsions, J. Colloid Interface Sci. 275 (2004) 659–664, https://
amorphization and a possible polyamorphism transition in nanosized TiO2: an x-ray doi.org/10.1016/j.jcis.2004.03.001.
absorption spectroscopy study, Phys. Rev. B Condens. Matter 77 (2008) 1–9, [61] W. Chantrapornchai, F. Clydesdale, D.J. McClements, Influence of droplet size and
https://doi.org/10.1103/PhysRevB.77.224112. concentration on the color of oil-in-water emulsions, J. Agric. Food Chem. 46
[50] H. Zhang, B. Chen, J.F. Banfield, G.A. Waychunas, Atomic structure of nanometer- (1998) 2914–2920, https://doi.org/10.1021/jf980278z.
sized amorphous TiO2, Phys. Rev. B Condens. Matter 78 (2008) 1–12, https:// [62] S. Tasset, B. Cathala, H. Bizot, I. Capron, Versatile cellular foams derived from CNC-
doi.org/10.1103/PhysRevB.78.214106. stabilized Pickering emulsions, RSC Adv. 4 (2014) 893–898, https://doi.org/
[51] S. Matsuo, N. Sakaguchi, E. Obuchi, K. Nakano, R.C.C. Perera, T. Watanabe, 10.1039/c3ra45883k.
T. Matsuo, H. Wakita, X-ray absorption spectral analyses by theoretical calculations [63] C. Jimenez-Saelices, B. Seantier, Y. Grohens, I. Capron, Thermal superinsulating
for TiO 2 and Ni-doped TiO 2 thin films on glass plates, Anal. Sci. 17 (2001) materials made from nanofibrillated cellulose-stabilized pickering emulsions, ACS
149–153, https://doi.org/10.2116/analsci.17.149. Appl. Mater. Interfaces 10 (2018) 16193–16202, https://doi.org/10.1021/
[52] N. Shandilya, I. Capron, Safer-by-design hybrid nanostructures: an alternative to acsami.8b02418.
conventional titanium dioxide UV filters in skin care products, RSC Adv. 7 (2017)
20430–20439, https://doi.org/10.1039/c7ra02506h.

10

You might also like